Abstract
ISMRM and SMI do not imply that such financial interests or relationships are inherently improper or that such interests or relationships would prevent the presenter from making an objective presentation. However, it is imperative that such financial interests or relationships be identified by the presenter so that participants at the CME activity may have these facts fully disclosed prior to the presentation, and may form their own judgments about the presentation. Towards this end, the information provided by each presenter can be found at the bottom of each abstract.
Every speaker, abstract presenter, organizer or anyone else who has control over any content in this meeting has been required to submit a Declaration of Financial Interests or Relationships, even if there is no conflict or relationship to declare.
Keynote Presentation
Abstract ID: 001
Progenitor cells with stem cell-like properties can be generated from embryonic stem cells and isolated from the developing and adult CNS. These cells, which can be maintained and propagated in cell culture, has provided a new tool for brain repair, in particular for replacement of lost neurons in neurodegenerative diseases. Cell therapy in Parkinson's disease (PD) is based on the idea that transplants of immature dopamine neuroblasts or dopamine neuron precursors can substitute, both structurally and functionally, for the lost dopamine neurons. In rodent and primate models of PD, fetal dopamine neurons grafted to the denervated striatum can re-establish a new functional dopaminergic innervation, restore striatal dopamine synthesis and release, and reverse impairments in motor behavior. Clinical trials have shown that embryonic mesencephalic dopamine neurons can survive and function also in the brains of patients with ongoing PD. The grafted neurons remain functional for at least 10 years after transplantation. Long-lasting symptomatic improvement has been observed in a majority of the grafted patients, and in the most successful cases L-DOPA treatment has been able to be withdrawn. For further development of the cell replacement technique there is a need to find alternative sources of cells for intracerebral transplantation. Recent studies suggest that in vitro expanded immature progenitors or dopamine neuron precursors, generated from the developing nervous system or from embryonic stem cells, may be developed as a virtually unlimited source of cells for cell replacement therapy in PD.
Further reading: Winkler C, Kirik D, Bjorklund A. (2005) Cell transplantation in Parkinson's disease: how can we make it work? Trends in Neurosciences, 28(2):86–92.
Plenary I: Advances in Biological Molecular Imaging
Abstract ID: 002
The transfer of experimental results and of derived pathophysiological concepts into clinical application is still limited by inappropriate experimental models and inadequate methods for the comparison of data. A few concepts from experimental models became very influential for the management of clinical stroke: in experimental models the dynamics of the development of ischemic tissue damage were studied and the time dependency of irreversible morphologic destruction as well as the potential of recovery of functionally impaired, but morphologically preserved tissue, were observed. This critically perfused tissue compartment, coined the penumbra, was later on the basis for reperfusion strategies, which with recombinant tissue plasminogen activator (rtPA) is still the only therapy of ischemic stroke with proven efficiency. In contrast, many efforts to save ischemic tissue by drugs affecting metabolism, reducing excitotoxicity or effects of free radicals, interfering with inflammation or preventing apoptosis - which all were successful in experimental models - failed to be effective in clinical trials. Therefore, for further clinically relevant animal studies models resembling clinical conditions and investigative procedures which can be applied and yield comparable results in experimental models and in stroke patients must be applied. As an example of translational research a parallel study of middle cerebral artery occlusion in cats and of malignant infarction in patients is described, where in both settings neuromonitoring and PET imaging could clearly differentiate between benign and malignant course and could identify early subjects at risk for formation of space occupying edema. From these comparative results of translational studies criteria can be defined to select patients who could benefit from invasive therapeutic strategies, e.g. early hemicraniectomy.
Abstract ID: 003
The emerging fields of genomics and proteomics have led to a better comprehension of the pathophysiology of cancer and the identification of novel signaling pathways. These pathways offer novel ‘targets’ which have led to the development of ‘lead molecules’ designed to inhibit the signaling derived from these pathways. However, this poses a tremendous challenge for selecting and/or validating these targets and for broad profiling of lead molecules for candidate selection. Molecular imaging technologies have the potential to address these scientific and technological challenges. A central goal of molecular imaging has been to develop strategies wherein activation or inhibition of key pathways in tumor formation as well as in the response of tumors to therapies can be non-invasively imaged. This presentation overviews imaging strategies for detection of for example, DNA damage, protease activation and signal transduction events. Molecular imaging reporter strategies are important tools that which are invaluable for testing the efficacy of targeted therapeutic agents as well as in optimization of dose, schedule and development of combination therapies. Overall, imaging strategies may provide important rationale for initiation of clinical oncology trials with molecularly targeted therapies. These combined with imaging biomarkers of early treatment response assessment should play a key role in future advances of individualization of patient treatment and care.
Abstract ID: 004
The increasing use of mouse models in drug discovery has resulted in a need for more efficient methods for studies of morphologic phenotypes. Conventional histopathology remains the gold standard. But survey of an entire mouse is prohibitively expensive. We have previously described the use of magnetic resonance histology as a complement to traditional histopathology. We report here reduction of that method to routine practice with improvements in efficiency of more than 25 × over our initial efforts. Mice are perfusion fixed with a mixture of formalin/ProHance. MR images are acquired at 7.0 T using fixtures and rf coils designed for high throughput. A novel scanning method using expanded dynamic range and an automated pipeline reconstruction yields a 256 × 256 × 1024 array @ 125-micron isotropic resolution in 35 minutes and a 62-micron isotropic array in 3.66 hours. The method has been extended to arrays of 1024 × 1024 × 4096 with isotropic resolution of 31 microns and to isolated organs with spatial resolution down to 10 microns. To date more than 100 specimens have been scanned that provide a baseline to understand mouse anatomy and normal variability.
Acknowledgements: All work was performed at the Duke Center for in Vivo Microscopy, an NIH/NCI National Resource (P41 RR005959, R24 CA092656).
Plenary II: Chemistry: Novel Probes and Activation Stategies
Abstract ID: 005
G-protein coupled receptors are important targets for the development of new drugs. They are characterized by a seven-transmembrane configuration. Members of this family include receptors for a variety of hormones, neurotransmitters, sensory receptors, chemokines. The interest to target these receptors with molecular imaging techniques and targeted therapy options is based on their plasma membrane location being easily accessible. In addition, many of the hormone receptors are highly overexpressed on several human tumors compared to normal cells. The prototype of receptors is the somatostatin receptor which is found mainly in neuroendocrine tumors. DOTA- and DTPA-modified analogs have been labelled with a variety of radiometals for γ-imaging (111In, 67Ga), positron emission tomography (PET, 68Ga) and radionuclide therapy (90Y, 177Lu, 149Pm, 153Sm, 213Bi). Examples of the clinical superiority (diagnostic sensitivity) and therapeutic efficacy will be discussed. In addition, other peptides will be discussed which target major human tumors like bombesin analogs (prostate, breast cancer), GLP-1 (insulinoma), NPY (breast cancer), gastrin (medullary thyroid cancer, neuroendocrine tumors).
Coordination chemical factors, namely the choice of the chelator, the mode of its coupling to the peptide and the size and charge of the radiometal influence the biologic and targeting properties of the vector and may be used to tailor suitable pharmacokinetic properties.
Abstract ID: 006
Different strategies have been developed that allow modulation of sensitised lanthanide luminescence to provide information about the local environment using emissive probe complexes. Key photochemical issues have been the nature and size of the sensitiser's singlet/triplet energy gap, the facility of inter-system crossing and the degree of control over quenching processes involving the three salient excited states that limit the emission quantum yield, i.e. the singlet and triplet states of the sensitising chromophore and the longer lived lanthanide excited state. Complexes engineered to report on the local concentration of bioactive species such as hydrogencarbonate, citrate, and phosphorylated Tyr sites have been devised [1]. In addition, various charged complexes have been studied that exhibit a marked tendency to be taken up by different cell types. Some of these complexes bind reversibly to DNA and localise selectively in the cell nucleus [2]. Strategies that may serve to allow their practicable usage in vitro or in cellulo will be aired and the key chemical and photophysical challenges described.
Abstract ID: 007
Superparamagnetic nanoparticles appear as good reporters for molecular imaging agents as compared to paramagnetic molecules. Indeed, although their r1 relaxivity tends to drop at high magnetic field, their r2 relaxivity reaches a plateau characterized by a very high r2/r1 ratio.
The synthesis of a vectorized superparamagnetic reporter designed for molecular imaging is a multistep procedure which should, at every stage, be perfectly reproducible.
This presentation will describe some of the problems related to the production of such nanoreporters and illustrate some applications in the context of molecular Magnetic Resonance Imaging (MMRT).
Symposium I: Molecular and Functional Imaging in Cancer Theranostics
Abstract ID: 008
A hallmark of cancer is increased proliferation with upregulation of glycolysis. Conversion of glucose to lactic acid in the presence of oxygen (aerobic glycolysis) is uniquely observed in cancers. In addition, increased activity of glycolytic enzymes has been observed. Moreover, hexokinase and other glycolytic enzymes are higher in metastases than in some primary tumors, suggesting an association between glycolysis and tumor progression. In addition to generate energy, a higher glycolytic flux in cancer cells is needed for increased production of precursor molecules for biosynthesis of nucleotides, phospholipids, and other components required for cell division. Increased activities of lactate dehydrogenase and pentose phosphatase pathway enzymes result in reduced amounts of pyruvate entering the mitochondria leading to low rates of oxidation by the tricarboxylic acid cycle and generation of ATP. This low rate of ATP production coupled with high-energy requirements leads to an increased need for glycolysis and therefore glucose uptake.
Glucose uptake can be visualized using PET and the glucose analogue 18FDG. Increased FDG uptake, seen in many tumors, relates to upregulation of glucose transporters and/or hexokinases. This in vivo characterization and measurement of biological processes is unique to PET.
An exciting area of PET is the possibility to quantify molecular processes in vivo over time. This enables monitoring response early during therapy, ultimately avoiding toxicity of ineffective treatment and optimizing therapy in individual patients. This is highly relevant, as monitoring response to therapy using anatomical imaging techniques is restricted by limited accuracy, specificity and delay between initiation of therapy and visible effects.
In this presentation, above-mentioned processes and the role of several biomarkers involved in FDG uptake will be reviewed, demonstrating that tumor tissue correlates for variable FDG uptake can be found and that these variables are indicative for the in vivo behavior of cancer in the individual patient.
Abstract ID: 009
Using the fraction of tumor pixels with PIs greater than 12.5ml/min/100g and defining 11% as a critical threshold level, therapy responder and therapy non-responder could be clearly separated except in eight cases. The predictive value to predict non-responders was 100%.
Abstract ID: 010
Non-invasive MRI methods for assessing the functional and morphological characteristics of tumour vasculature, a critical determinant of tumour growth, metastatic potential and therapeutic response, are continually being sought. The ultimate aim is the identification, development and validation of quantitative clinical MRI indices to assist in planning of individual patient treatment protocols.
Blood oxygenation level dependent (BOLD) MRI images are sensitive to changes in deoxyhaemoglobin concentration through the transverse MRI relaxation rate R2* of tissue water, hence the quantitative measurement of tumour R2* may provide a sensitive index of tissue oxygenation. We have shown that GH3 prolactinomas and RIF-1 fibrosarcomas exhibit very different baseline R2* and carbogen (95% O2 /5% CO2)-induced ΔR2* responses, and inferred that the magnitude of the ΔR2* response to carbogen is determined principally by the tumour blood volume, itself a determinant of the hypoxic fraction (JMRI
The prognostic potential of tumour R2* and carbogen-induced ΔR2
The data suggest that tumour R2* and carbogen-induced ΔR2* provide non-invasive prognostic indicators of potential acute radiotherapeutic response.
Symposium II: Novel Reporter Genes
Abstract ID: 011
Molecular imaging provides visualization of normal as well as abnormal cellular processes at a molecular or genetic level. In general, molecular genetic imaging means molecular imaging methods using an imaging reporter gene expression. Most molecular genetic imaging paradigms involve the use of reporter-transgene technology and a complementary reporter probe. For nuclear medicine molecular imaging, several reporter genes were developed in PET and gamma camera system. Current PET reporter genes fall into two categories: enzyme based (HSV1-thymidine kinase: HSV1-tk) and receptor based (dopamine 2 receptor: D2R) methods. PET imaging reporter genes, such as HSV1-tk and D2R require the synthesis of complicated substrates and expensive PET scanner. We and other groups proposed that sodium/iodide symporter (NIS) gene may serve as an alternative imaging reporter gene. Thyroid cells take iodide through a specific transporter, so called sodium/iodide symporter. The driving force for iodide uptake is the transmembrane concentration gradient of sodium ion, which is generated and maintained by the sodium-potassium pump. This transporter co-transports sodium ion and iodide ion and named sodium/iodide symporter. NIS gene was firstly identified in 1996 from rat. The transfer of NIS gene and expression of functional NIS protein would enable cells to concentrate radioiodide including Re-188 and Tc-99m and therefore offer the possibility of reporter gene imaging measured by gamma-camera. NIS has many advantages as an imaging reporter gene due to the wide availability of its substrates including radioiodines, Tc-99m and Re-188, and the well-understood metabolism and clearance of these substrates in the body. Reporter gene imaging using NIS allows the visualizations of the endogenous gene expression, and of nuclear receptor activities and can be used to monitor target cells such as cancer or stem cells.
Abstract ID: 012
The ability to determine the level and duration of gene expression in-vivo is critical for the clinical use of gene therapy. Positron emission tomography (PET) can be used to monitor gene expression using tracers that image the spatial distribution of the therapeutic transgene product. We have used PET to monitor gene expression in a primate model of Parkinson's disease (PD). Our therapeutic strategy is aimed at increasing striatal aromatic L-amino acid decarboxylase (AADC) levels. Striatal neurons infected with the AADC gene by an adeno-associated viral (AAV) vector can convert low doses of systemically administered L-dopa to dopamine resulting in clinical improvement without the side effects typically association with higher doses of L-dopa. We have shown that PET measures of striatal uptake of the AADC tracer, 6-[18F]fluoro-L-m-tyrosine (FMT), is substantially increased following the AAV delivery of AADC gene in parkinsonian monkeys, and striatal FMT uptake correlates with histological measures of the extent of gene expression. We have also performed longitudinal PET studies and have demonstrated sustained gene expression and clinical improvement for over 5 years. These findings have lead to the development of a human clinical trial which began last year.
Abstract ID: 013
Reporter genes provide information on the biodistribution and viability of transfected cells. Magnetic resonance (MR) imaging is suitable for non-invasive cell tracking, and lately, monitoring gene expression. However, existing MR reporters rely on the presence of (super)paramagnetic substances. As such, they all use water relaxation to gain contrast, making it impossible to distinguish differentially-labelled target cells. In this study we demonstrate a genetically engineered reporter that provides frequency-selective contrast based on the Chemical Exchange Saturation Transfer (CEST) mechanism, in which radiofrequency saturation is transferred from the reporter's exchangeable protons to water protons. Rat glioma (9L) cells were transfected with Lysine Rich Protein (LRP), which was detected by immunostaining. The LRP expression showed no effect on cell proliferation and metabolic rate.
In in-vitro MR studies of protein extracts from cells expressing LRP showed significantly higher average change in MR signal of 0.87% (n=5; p-value < 0.05) then protein extract from control cells (0.24%; n=4). Moreover, in pixel-by-pixel t-test maps only pixels in the capillary containing the LRP showed a significant change in MR signal, which was not detected in the control capillaries (Fig 1B). This is a first example of an artificial MR reporter gene providing contrast that can be switched “on and off” by applying RF pulses at the proper amide proton frequency. This new concept offers a general approach for constructing a family of endogenous reporter genes suitable to detect multiple cell populations by MRI.
Figure 1: A. in-vitro system: poly-L-lysine 30kDa 10μM (1) and 100μM (2). Protein extract from LRP expressing cells (3), protein extract from control cells (4) or PBS only (5). B. t-test map, color coded pixels represent p-values < 0.05.
Symposium III: High Sensitive, Metal-Based Imaging Probes
Abstract ID: 014
Water-soluble fullerene derivatives possess potential for biomedical applications as antioxidants, anti-HIV drugs, X-ray contrast agents, bone-disorder drugs and photosensitizers for photodynamic therapy. In addition, endohedral metallofullerenes (M@C2n) have been suggested as nuclear medicines (M = Ho3+), fluorescent tracers (M = Er3+) and MRI contrast agents (M = Gd3+)1,2 largely because the closed fullerene cage insures against toxic metal-ion release in vivo. Water-soluble members of the Gd@C60 family of metallofullerenes have recently been shown to achieve their large proton relaxivities (r 1 ) though pH-controlled self-aggregation.2 In fact, self-aggregation may well be a common feature of all water-soluble fullerene chemistry, and its understanding is, therefore, of general importance for fullerene-based drug delivery. Proton nuclear magnetic relaxation dispersion (NMRD) measurements for the water-soluble metallofullerenes, Gd@C60(OH)x (x≈27), and Gd@C60[C(COOH)2]10 can be used as a reporter to probe the aggregation (or disaggregation) characteristics of water-soluble fullerene materials in the presence of physiologically-encountered agents.
Molecular models of Gd@C60(OH)26 (top), and Gd@C60[C (COOH)2]10 (bottom).
Abstract ID: 015
CEST agents represent an emerging class of MRI contrast media of huge potential. They act as negative agents by reducing the signal intensity of the water protons through a saturation transfer mediated by chemical exchange. The saturation transfer occurs through the selective irradiation at the absorption frequency of the exchanging protons of the agent. The great potential of CEST agents lies on possibility of switching “on” and “off” the contrast at will, making possible the detection of more agents, each uniquely characterised by the specific absorption frequency of their mobile protons, within the same specimen. Basically, the saturation transfer efficiency is proportional to the exchange rate (kex) and to the number of mobile protons of the CEST agent. However, in order to carry out selective irradiation of the mobile protons, kex must be smaller than the frequency separation (αω) between the resonance of bulk water and of the exchanging protons of the agent. For this reason, efficient CEST agents may be found in the class of paramagnetic metal complexes (the so-called PARACEST agents), where the presence of unpaired electrons induce large αω values.
In this contribution, two novel classes of highly-sensitive paramagnetic CEST systems containing a large number of exchangeable protons, SUPRACEST and LIPOCEST agents, will be described.
SUPRACEST agents are supramolecular adducts formed between a paramagnetic shift reagent and substrates containing a high number of exchangeable protons. Upon interaction, the resonance frequency of the mobile protons of the polymer are shifted, thus allowing their selective saturation. These systems display a sensitivity in the micromolar range. A further considerable sensitivity improvement has been obtained by using liposomes entrapping a water proton paramagnetic shift reagent (LIPOCEST agents), where the sensitivity limit is decreased down to the picomolar range (expressed as liposome concentration).
Abstract ID: 016
The 68Ge/68Ga generator (68Ge, T1/2 = 270.8 d) provides a cyclotron-independent source of 89% positron-emitting 68Ga (T1/2 = 68 min). Although 68Ga is separated as trivalent cation, the eluate needs to be pre-concentrated and purified from 68Ge, Zn(II), Ti(IV), Fe (III) prior to labelling.
We developed a system for efficient processing of 68Ge/68Ga generator eluates. The main component is a micro-chromatography cation exchange column. Purification and concentration of 68Ga(III) are carried out in hydrochloric acid-acetone systems. For labelling, DOTA-and DFO-conjugated (DOTA = 1,4,7,10-tetraazacyclododecane-1,4, 7,10-tetraacetic acid, DFO = desferrioxamine-B-succinyl) octreotides such as DOTA-DPhe1-Tyr3-octreotide (DOTATOC) and DOTA-DPhe1-Nal3-octreotide (DOTANOC) were used. DOTATOC binds with high affinity only to hsstr2, while DOTANOC provides high affinities to all subtypes of the human somatostatin receptor.
More than 97% of 68Ga are finally obtained in < 0.5 ml of a 0.05 M HCl / acetone solution. The 68Ge contamination was reduced by a factor of 1000. In addition, initial amounts of Zn(II), Fe(III) and Ti(IV) are reduced significantly. High-yield labelling was achieved with this pure fraction. In a clinical environment, injectable radiopharmaceuticals, e.g. 68Ga-DOTATOC, are available routinely within 25 min and with up to 60% yield referred to the initially eluted 68Ga. Under optimised conditions, specific activities of the labelled peptides of up to 400 MBq/nmol can be achieved.
The developed system represents a simple and efficient way for labelling of biomolecules with generator-produced 68Ga(III). 68Ga-DOTATOC and 68Ga-DOTANOC were successfully used in a series of somatostatin receptor-expressing tumour diagnosis with PET/CT. The impact of molecular imaging of neuroendocrine tumours and metastases is illustrated. This aproach of targeting tumour membrane receptors with 68Ga-labelled DOTA-conjugated peptides might be considered as a prototype for the future application of generator produced 68Ga.
Symposium IV: Multimodal Imaging - Instrumentation
Abstract ID: 017
This presentation will review approaches used to develop scanners capable of acquiring both PET and MR images simultaneously, using PET detectors readout by photo-multiplier tubes (PMT). Despite the inherent sensitivity of PMTs to magnetic fields, there are many reasons to continue with the PMT approach in a PET-MR scanner. These include gain, energy resolution, timing, stability and cost. The pioneering work of Shao et al [1,2] demonstrated that a PET-MR scanner based upon PMT readout could be achieved by the use of long fibre optics connecting the scintillating crystals and PMTs. To improve upon the low sensitivity of the original McPET it has recently been extended to a multi-layer design [3]. Most recently, split MR designs have been proposed that can accommodate a multi-ring PET scanner in order to have fully 3D PET performance [4,5]. The presentation will consider the relative benefits of adapting the MR to accommodate a conventional PET geometry over adapting the PET geometry to fit into a conventional magnet.
Abstract ID: 018
Abstract ID: 019
Positron emission tomography (PET) reveals functional processes with a very high sensitivity in the pico-molar range. The disadvantage of PET is the lack of morphological information which makes it very hard to localize exactly the regions of tracer uptake. PET combined with X-ray computed tomography (CT) has shown in clinical applications to be a very valuable multimodality imaging device revealing functional and anatomical information. However, in applications involving animal imaging, CT provides only limited soft tissue contrast and exposes the animal to a relatively high radiation dose. In sharp contrast, magnetic resonance imaging (MRI) provides much better soft tissue contrast than CT without applying contrast agents and does not require ionizing radiation. Several groups are currently working on the combination of PET and MRI into one single device for simultaneous functional and morphological imaging. First encouraging results show the feasibility of this approach. The talk will provide an overview about the work done in this area and will point-out the achieved milestones.
Symposium V: Imaging of Protease Activity in Disease
Abstract ID: 020
Proteases are enzymes that have the ability to cleave proteins and peptides by irreversible catalytic hydrolysis of peptide bonds. Human genome sequence analysis has indicated that more than 500 genes are encoded for proteases or protease-like proteins, which explain the diversity and complexity of proteases involve in physiological processes. Clinically, the activity of proteases can not only act as biological markers for various diseases, but also predict the disease outcome. A detail understanding of the protease activity in vivo will lead to better pharmacological approaches for disease control. Several new types of protease-mediated optical probes recently developed in our group will be discussed. These optical probes are designed to have minimal fluorescence signal, and becoming brightly fluorescent when they are recognized and degraded by the target proteases. Recently we have applied these probes to study tumor, arthritis, and cardiovascular diseases. The acquired information of in vivo protease activity could potentially revolutionize current approaches in disease detection, screening, diagnosis, staging, drug development and treatment evaluation.
Abstract ID: 021
Protease activity is tightly regulated in both normal and disease conditions. However, it is often difficult to monitor the dynamic nature of this regulation in the context of a live cell or whole organism. Recently, a number of elegant methods have been developed to image protease activity using fluorescently labeled reporter substrates. However, these methods rely on substrates which are often processed by multiple protease targets and that can diffuse away from the initial site of enzyme processing making analysis of imaging data difficult. We have developed a series of fluorescently quenched small molecule probes that become fluorescent upon covalent modification of a protease target. These reagents freely penetrate cells and can be used to directly image protease activity in live cells and whole animals. Targeted proteases can be directly identified and monitored biochemically by virtue of the resulting covalent tag thereby allowing unambiguous assignment of protease activities observed in imaging studies. We have applied this method to develop probes for cysteine cathepsins and caspases. Examples of applications of this methodology to imaging of both monolayers and 3D cultures of cancer cells will be presented. In addition application of these small molecule probes to in vivo models of angiogenesis and tumorigenesis will be discussed.
Abstract ID: 022
The ubiquitin/proteasome system (UPS) is a complex machinery that is dedicated to the degradation of misfolded proteins present in the cytosol, nucleus and endoplasmic reticulum (ER) of cells. Conformational diseases, such as Alzheimer's and Parkinson's disease, are characterized by the accumulation of misfolded proteins. The question remains why the UPS fails to destroy these misfolded proteins. To address this issue, we have developed a set of fluorescent protein-based reporter substrates that allows functional analysis of UPS in vivo. With these reporters, we studied the effect of ER stress, a condition found in conformational diseases, on the functionality of the UPS. Notably, loss of ER integrity causes misfolding of ER proteins, which are translocated to the cytosol where they are cleared by the UPS. We found that ER stress had a subtle inhibitory effect on the degradation of three unrelated cytoplasmic/nuclear UPS substrates. ER stress also primed the cells to the inhibitory effect of an aberrant ubiquitin associated with conformational disease. These data suggest that ER stress compromises the UPS and may contribute to the accumulation of misfolded proteins in conformational diseases.
Symposium VI: Therapeutics
Abstract ID: 023
Abstract ID: 024
Cancer continues to be a major humanitarian problem. Its incidence is likely to increase because of a growing and older population. Yet, mortality rates remain constant. In some cases pretreated logoregionally advanced tumors significantly interfere with quality of life. However, treatment options are limited. After extensive pre-clinical testing we have used nanomagnetic drug-targeting in 14 cancer patients with advanced malignant disease to verify the technology, identify possible problems and to prepare further clinical trials. This technology implies a ferrofluid with particle diameters of 100–500 nm (depending on the protocol), the reversible binding of cytotoxic drugs to a carbohydrate coating of these particles and the intravenous infusion of this compound while a high-engery permanent magnetic field is being created at the tumor site to be treated. Through this technique very high concentrations of the cytotoxic drug can be created within the tumor, while, on the other hand, systemic toxicity can be minimized. Current animal experiments and planned cinical trials are aimed at identifying the optimal particle size for use in patients, different desorption strategies for the drug from the ferrofluid within the organism and patient-tailored treatment protocols with possible highly toxic and/or expensive and/or drugs with a short half-life.
Abstract ID: 025
Metal nanoshells are a new class of optically-tunable nanoparticles that can be designed to have strong optical resonance in the near infrared, the region of the spectrum where penetration of light through tissue is maximal. Additionally, particles can be designed such that extinction is predominantly absorption or scattering, or somewhere between those extremes. When gold is used as the surface metal, conjugation to targeting moieties such as antibodies, aptamers or peptides is easily accomplished. We have developed both molecular imaging using reflectance confocal microscopy and photothermal therapy technologies based on gold nanoshells. In photothermal therapy, we have achieved 100% tumor regression and survival in a colon carcinoma model. We have further demonstrated that it is possible to integrate imaging and therapy using particles with both absorption and scattering properties to facilitate a ‘see-and-treat’ approach to cancer therapy.
Plenary III: Imaging Signal Transduction, Transcription and Cell Cycle In Vivo
Abstract ID: 026
New reporters for imaging protein trafficking, protein-protein interactions, and enzyme activity in the live cell context will be described. Application to the study of AMPA receptor function and its relation to synaptic plasticity will be described.
Abstract ID: 027
Notch signaling is an evolutionary conserved mechanism that is used by metazoans to direct cell fate decisions, proliferation and apoptosis during development and in self-renewing adult tissues. Mammalian cells contain four Notch receptors and several Delta-like and Jagged ligands. Notch mediated cell fate decisions depend on ligand-induced proteolytic cleavage of Notch which leads to ectodomain shedding followed by presenilin mediated release of the Notch intracellular domain which acts as a transcription factor. Currently it is not well understood which cell types/lineages depend on Notch signaling in vivo during vertebrate development, in the adult or during disease processes. We have designed two methods to visualize Notch activity in vivo. The first is a novel Notch1 allele (N1::Cre) to indelibly mark cells and their descendants after they have undergone ligand mediated Notch1 activation. In mice expressing this Notch allele, ligand-mediated proteolysis of Notch1 results in release of nuclear-targeted Cre (nlsCre) from the membrane. By generating compound heterozygotes with N1:Cre and reporter lines, Notch1 activation leads to an irreversible genetic switch resulting in marking of descendent cells. In another approach we used a luciferase complementation strategy to visualize the interaction between Notch and its nuclear partner, RBPjk in real time.
Abstract ID: 028
Antimalarial drug resistance is a horrendous global problem. Elucidating mechanism(s) of antimalarial drug resistance requires high resolution sub cellular information. We have developed novel approaches for three dimensional vs. time (“4D”) imaging of living intraerythrocytic malarial parasites in the presence vs. absence of antimalarial drugs. Individual frames of 4D movies (e.g. a full 3D z stack at one time point), at 200–250 nm resolution, are obtained in less than 1 second for living parasites under continuous perfusion. Since parasites “wiggle” inside red cells, this allows us to examine of a number of clinically important phenomena at significantly improved resolution (both spatial and temporal) relative to conventional scanning confocal microscopic techniques. Initial experiments to be discussed have focused on organellar pH and volume control as well as biomineralization of heme within the parasite digestive vacuole.
Plenary IV: Imaging in Neuroscience
Abstract ID: 029
Newer in vivo imaging modalities, including bioluminescence, fluorescence tomography and magnetic resonance (MR) imaging provide a means to assess gene delivery, therapeutic action and regression in brain tumors. Dual bioluminescence imaging can be used to track gene delivery into tumors via virus vectors or migration of neural stem cells to tumors, while in parallel, monitoring changes in tumor volume over time. A new potent coelentrazine luciferase from Gaussia allows monitoring of protein release into the tumor environment. One of our goals has been to image therapeutic activity of proteins, including apoptotic activity and lethal influx of sodium with near infra-red imaging probes and fluorescent dyes. Another goal has been to couple cell labeling with targeted therapy by expressing a metabolically biotinylated protein on the surface of tumor cells which can be imaged by MRI using avidin-conjugated magnetic nanoparticles, and at the same time provide a target for avidin-toxin moieties. The presented methods will allow rapid assessment of the diagnostic and therapeutic potential of novel interventions to speed the search for effective therapy.
Abstract ID: 030
Microglia constitute 10–20 % of glial cells and are ontogenetically related to cells of mononuclear phagocyte lineage. They respond to brain injury by expressing cytokines and other immunologically relevant molecules and have been used as an early marker of active brain disease. Activated microglia express mitochondrial peripheral benzodiazepine binding sites (PBBS). (R)-PK11195 is a selective isoquinoline ligand for PBBS and, when labelled with [11C], can be used as a PET tracer. In Alzheimer's disease 11C-PK11195 PET detects 1.5–2 fold increases in microglial activation in cortical association areas. These increases correspond to areas of amyloid plaque deposition demonstrated with 11C-PIB PET. In mild cognitive impairment focal cortical areas of microglial activation and amyloid deposition can be detected with PET. Subcortical degenerations, such as Parkinson's disease and multiple system atrophy, show increased 11C-PK11195 uptake in basal ganglia and brainstem which appears relatively stable over 1–2 years of follow up. We are currently relating the distribution of brain 11C-PK11195 uptake in Alzheimer's and Parkinson's diseases to rates of disease progression and testing the effects of putative neuroprotective agents, such as the PPARg agonist minocycline, on glial activity.
Abstract ID: 031
Songbirds share with humans the capacity to produce learned vocalizations (songs). Learning songs and singing is controlled by steroid-sensitive
Since the early discovery of the SCN by F. Nottebohm in 1981, birdbrain neuroscience has become a research area with a high impact factor that remained however until 2002
Introduction of in-vivo MRI to study neuroplasticity in the birdbrain represents a stepping stone in transferring in-vivo technology from mammalian to bird neurosciences and creates novel opportunities in the most intriguing domains of neuroscience. On the other hand the songbird system clearly provides a natural model to
Symposium VII: Mechanisms of Oncogenesis and Disease Progression in Tumor Models
Abstract ID: 032
Although sophisticated regimens of conventional therapies are being carried out to treat patients with gliomas, the disease invariably leads to death over months or years. Before new and potentially more effective treatment strategies can be effectively implemented in the clinical application certain prerequisites have to be established. First, the exact localization, extent and metabolic activity of the glioma must be determined to identify the biologically active target tissue for a biological treatment regimen such as gene therapy. This is usually performed by imaging the expression of upregulated endogenous genes coding for glucose or amino acid transporters and cellular hexokinase and thymdine kinase genes, respectively. Second, neuronal function and functional changes within the surrounding brain tissue have to be assessed in order to save this tissue from therapy-induced damage. Third, pathognomonic genetic changes leading to disease have to be explored on the molecular level to serve as specific targets for patient-tailored therapies. Last, a concerted non-invasive analysis of both endogenous and exogenous gene expression in animal models as well as the clinical setting is desirable to effectively translate new treatment strategies from experimental into clinical application.
Supported in part by MWF516-40000299, ZMMK-TV46, DFG JA 981/1–2, EU-6thFW-EMIL.
Abstract ID: 034
The phenotypic changes of increased motility and invasiveness of cancer cells are reminiscent of the epithelium-to-mesenchyme transition (EMT) that occurs during embryonic development. TGF-β is a well-characterized inducer of EMT in embryonic development, certain fibrotic diseases and cancer (progression). In contrast, another member of the TGF-β superfamily, bone morphogenetic protein 7 (BMP-7), is involved in the maintenance of the epithelial phenotype by induction of the mesenchyme-to-epithelium transition (MET). This study was designed to investigate whether: 1) Tumorigenicity and invasive behavior is associated with a modification of BMP-7 expression in clinical breast and prostate cancer and in human breast and prostate cancer cell lines 2) BMP-7 treatment affects the growth of human breast and prostate cancer cells in orthotopic and bone metastasis tumor models. 3. BMP-7 modulates TGF-β-mediated intracellular signaling.
In most invasive primary breast and prostate carcinomas BMP-7 expression was low or absent, while it was found to be expressed in non cancerous tissue. In line with these observations, BMP-7 expression in tested cancer cell lines is inversely related to the expression of the mesenchymal marker vimentin and to tumorigenicity in vivo. Daily systemic administration of rhBMP-7 significantly inhibited the orthotopic and intraosseous growth of human MDA-MB-231 and PC-3M-Pro4 cancer cells in nude mice, as monitored by whole body bioluminescent imaging (BLI).
Taken together, our data suggest that BMP-7 controls the epithelial homeostasis in the human mammary and prostate gland by preserving the epithelial phenotype. Loss of BMP-7 expression during breast and prostate cancer progression could trigger the TGF-β-stimulated EMT and, thus, contribute to the acquisition of an invasive phenotype. However, exogenous BMP-7 can still counteract the EMT process in the primary and metastatic tumor. Therefore, BMP-7 may represent a novel therapeutic molecule for repression of local and systemic breast and prostate cancer progression.
Abstract ID: 035
We have designed, synthesized, and validated a novel photodynamic therapy (PDT) agent combining its original role in treating cancer with the ability to directly assess PDT-induced apoptosis by monitoring caspase-3 activity in the treated cancer cells.
This molecule contains: 1) a fluorescent photosensitizer that produces singlet oxygen upon irradiation; 2) a fluorescence quencher; 3) a peptide linker that is a caspase-3 substrate; and 4) a folate used as a tumor specific delivery vehicle. Once localized inside the tumor cells, the light activated photosensitizer produces cytotoxic singlet oxygen that destroys the cells. If cell death occurs primarily by apoptosis, caspase-3 will be activated and will then cleave the peptide linker between the photosensitizer and the fluorescence quencher. Separation of the fluorescence quencher allows cancer cells undergoing apoptosis to be visualized by near-infrared (NIR) fluorescence imaging. We have validated this concept both in vitro and in vivo (Fig. 1) using folate receptor positive and negative tumors. Folate receptor specific targeting and PDT-induced apoptosis were successfully observed simultaneously.
Abstract ID: 036
Noninvasive molecular imaging has become an important tool in monitoring tumor growth and therapy. In this project, a versatile, potent technique for targeting and imaging of tumors is explored using the strong interaction of biotin to chicken egg avidin or bacterial streptavidin. Recombinant DNA constructs were engineered for synthesis of a cell surface receptor (22 kDa) incorporating the biotin acceptor peptide (BAP) of the prokaryotic Propionibacterium shermanii 1.3S transcarboxylase domain (PSTCD) between an N-terminal signal sequence and the PDGF receptor. This expression construct was incorporated into a lentivirus vector to facilitate delivery to tumor cells. During synthesis of this fusion protein, the endogenous mammalian biotin ligase covalently attaches a single biotin moiety to a specific lysine residue in the BAP region. Western blot analysis of cell homogenates and immunocytochemistry of non-permeablized cells demonstrated that this fusion protein is efficiently biotinylated in mammalian cells and that the biotin is displayed on the cell surface. These engineered tumor cells can be imaged in culture and in vivo with fluorescence molecular tomography (FMT) or magnetic resonance (MR) using streptavidin-Alexa680 (Molecular Probes) or avidin-CLIO (developed in our laboratory), respectively. Using T2-weighted spin echo pulse sequences, MR imaging of cultured cells incubated with avidin-CLIO showed a significant decrease in the T2 value of biotinylated cells with respect to control cells. Similarly, in vivo FMT imaging of subcutaneous tumors expressing the biotinylated receptor showed a 10-fold increase in fluorescence over control tumor after i.v. injection of streptavidin-Alexa680. The versatility of this technique allows non invasive imaging of tumor cells or any other cell type, in real-time, with any imaging agent coupled to avidin or streptavidin.
Symposium VIII: Vascular Endothelium and Atherosclerosis
Abstract ID: 038
In up to two third of patients, acute coronary events result from the rupture of the atherosclerotic plaque. The plaque vulnerable to rupture is characterized by distinct histologic characteristics that include large lipid cores, attenuated fibrous caps of less than 65 micron thickness, and intense inflammation of the fibrous caps. An extensive apoptosis of inflammatory cells is associated with rupture of the plaque. Annexin V, a naturally occurring protein has a nanomolar affinity for binding to phosphatidyl serine (PS), a pospholipid expressed on the outer leaflet of cell membrane of apoptotic cells. 99mTc labeled Annexin V has been successfully used for noninvasive imaging of apoptosis in atherosclerotic plaques in various experimental atherosclerotic models and also in the detection of apoptosis in carotid artery atherosclerotic disease in patients undergoing carotid endarterectomy. Annexin imaging has also demonstrated that dietary modification and statin therapy lead to histologic stabilization of the plaques that is associated with marked reduction in apoptosis.
Abstract ID: 039
Cardiovascular diseases are the most common cause of death and disability in industrialised countries. Most of the morbidity and mortality is due to atherosclerosis of the coronary arteries, resulting in coronary heart disease (CHD) and its principal manifestations, angina pectoris, myocardial infarction, sudden cardiac death and heart failure. The main pathologic feature of atherosclerosis is the atherosclerotic lesion. The potentially most dangerous lesions are unstable and prone to rupture. At autopsy, they are characterised by a thin fibrous cap overlying cell-rich regions with a central core of extracellular lipid. Rupture of those plaques mainly occurs in their shoulder where active macrophages accumulate which locally destabilise the plaque shoulder by secreting a variety of matrix-degrading proteolytic enzymes such as metalloproteinases. Therefore, there is a great need for non-invasive diagnostic techniques that discriminate vulnerable from stable plaques within coronary arteries, to early identify patients at high risk of major acute coronary events. Since techniques such as MRI are based on morphological analysis of plaque structure, composition and size, none of them is able to elucidate the characteristic metabolic profile of potentially unstable (vulnerable) plaques.
Positron emission tomography (PET) provides the most sensitive and selective means for imaging molecular interactions non-invasively in the living body and could therefore prove a potent approach to the identification of the metabolically active plaque that is vulnerable to rupture. Targets for PET imaging are among enzymes (e.g. matrix-metalloproteinases, caspases) and receptors, whose expression an activation destabilizes the plaque and increase the risk for plaque rupture and subsequent lifetheatening events.
Different strategies for developing new radiopharmaceuticals for imaging the MMP activity in vulnerable plaques and first results will be presented.
Abstract ID: 040
Abstract ID: 041
Symposium IX: The Truth about Target to Noise: Physical and Chemical Determinants
Abstract ID: 042
The ability to image (and measure) molecular events in vivo depends on the signals which are generated when imaging a molecular target, event or process, and by the sensitivity of the instrument(s) to record those signals. Two imaging strategies (direct and indirect) involving optical, positron emission and magnetic resonance imaging are currently being used. Both imaging strategies depend on the relationship between tissue concentration of the biological target (direct) or the reporter gene product (indirect) and the readout of the chosen imaging modality. Namely, the difference in signal intensity/activity (contrast) between specific vs non-specific signal recorded within target and non-target tissues.
This presentation will explore the biological determinants of image contrast and noise in molecular-genetic imaging studies. It will focus on the relationship between target concentration and location, and the selection of an appropriate imaging probe - imaging modality combination. For direct imaging strategies, knowledge of target concentration and the physical/biological characteristics of the probe (coupled with the sensitivity of the recording instrumentation) are essential considerations. For indirect reporter gene studies, enzymatic signal amplification provides an advantage; however, similar issues relate to the physical/biological characteristics of the probe and the sensitivity of the recording instrumentation.
Several general points emerge: 1) total measured signal intensity/activity and “specific binding” decrease proportionately with lowering the number of target molecules per cell; 2) the fraction (%) of measured activity due to “specific binding” decreases as the specific activity of the probe decreases; 3) total measured activity increases when nonspecific binding or metabolites increase; 4) the fraction (%) of the measured activity due to “specific binding” decreases as nonspecific binding or metabolites increase. The physical and biological determinants of image contrast and noise should be considered prior-to initiating imaging studies in order to determine/estimate whether imaging a particular molecular-genetic target or process is feasible.
Abstract ID: 043
Altered central nervous system serotonin (5-HT) function is involved or has been implicated in a number of debilitating mental health disorders, including depression. Assessment of 5-HT synaptic integrity across cerebral regions of interest (ROIs) of living brain may be accomplished by dynamic positron emission tomography (PET) imaging with the analysis of the distribution density of the presynaptic serotonin transporter (SERT) protein marker. The in vivo evaluation of neuron cell surface expressed SERT demands potent and selective SERT inhibitor-based PET tracers possessing appropriate ROIs kinetic profiles. Human postmortem autoradiography studies of subjects diagnosed with depression have found distinct SERT distribution patterns across the neocortex and limbic system ROIs, which are characterized with moderate to low SERT density patterns. Recent PET imaging studies utilizing select carbon-11 labeled SERT tracers have attempted to quantify moderate-low neocortical and limbic SERT distribution densities in primates. The efforts have been stymied by insufficient radioligand signal in select ROIs. Thus, more sensitive PET tracers for neocortical and limbic ROIs are required. To overcome the radioligand signal problem while preserving target-to-nontarget (specific-to-nonspecific) tracer kinetic binding profiles, efforts in our laboratory have focused on employing fluorine-18 ([18F]) labeled radioligands with high SERT binding affinity. Our medicinal chemistry strategy has encompassed the rational design and development of new SERT PET [18F] tracers possessing: a) increased plasma availability for enhanced brain penetration, b) greater SERT binding potency suitable for moderate-low SERT density ROIs characterized with association-dissociation rates favorable for the [18F] nuclide, and c) specific structural attributes lending to more optimal target-to-nontarget tissue ratios. Our radioligand design has been based upon SERT inhibitor ligand pharmacophore models and select tracer structure-correlated pharmacokinetic-pharmacodynamic analyses. The design and advancement of a new [18F]6-nitroquipazine SERT PET tracer with significantly enhanced rodent and non-human primate signal-to-noise qualities for critical SERT ROIs will be discussed.
Abstract ID: 044
Cell-based therapy using pluripotent stem cells has demonstrated restorative capability of the injured myocardium. Both animal and clinical investigations have shown that cell-based therapy improves cardiac function following myocardial infarction. Although the exact mechanism of functional restoration of the injured myocardium is not known, survival, proliferation, and transdifferentiation of embryonic stem cells (ESC) play an important role in the restorative effects. We have been studying the feasibility of dual in vivo MRI using both 1.5 and 4.7 T scanners for simultaneous detection of transplanted mouse embryonic stem cells (mESC) and assessment of functional restoration in a mouse myocardial infarction model. Dual in vivo MR detection of transplanted mESC and evaluation of the mESC-treated myocardium was achieved over 4-week duration demonstrating the following by the mESC treated mice: 1) Significant improvement of the left ventricular (LV) function, 2) Attenuation of fibrosis and calcification in the infarct region, and 3) Serial assessment of the transplanted mESC.
However, the negative MRI dephasing signal induced by the magnetically labeled stem cells frequently obscures the contrast generated from the cells and the surrounding artifacts. In order to address this issue, we successfully generated positive signal from the magnetically labeled stem cells using an off-resonance (OR) MRI sequence. This OR scheme excites and refocuses a narrow band of off-resonance water within the dipole field surrounding the magnetically labeled cells to generate positive contrast. The net integrated signal intensity scales linearly with the volume of labeled cells to enable a measurement of signal area that increases monotonically with growing cell population. This capability allows high contrast to noise ratio to potentially enable molecular and cellular MR imaging of stem cell survival, proliferation and transdifferentiation.
Abstract ID: 045
Molecular imaging has great potential for studying pathophysiological processes in brain diseases, such as stroke or neurodegenerative diseases. In preparation for human studies, we investigated the technical feasibility of near-infrared fluorescence (NIRF) imaging for optical imaging of the brain in a pigmented mouse strain that is commonly used in brain research. We implanted a capsule containing a near-infrared fluorescent dye and the MRI contrast agent Gadolinium into the left hemisphere of anesthetised C57Bl6 mice in vivo. MRI was used to obtain anatomical information, to exactly locate the implanted capsule and to exclude animals with abnormalities (e.g. haemorrhages). Sensitivity and contrast of optical imaging was examined using planar (transillumination fluorescence imaging (TFI) and fluorescence reflectance imaging (FRI)) and fluorescence-mediated tomography (FMT).
The results show that the detection limit for TFI and FRI was 10−12 mol of NIRF dye implanted into the brain of pigmented C57Bl6 mice. TFI yielded a higher contrast than FRI, but required longer acquisition times. With FMT, contrast could be further improved and spatial information -essential for imaging the brain-was gained. To simulate background fluorescence varying amounts of NIRF dye were injected intravenously. The capsule could still be detected with TFI and FRI when the blood concentration of the NIRF dye was less than one third of the dye concentration in the capsule.
This work was funded by the Europäischer Fonds für regionale Entwicklung (EFRE) and Hermann and Lilly Foundation.
Abstract ID: 046
Cancer research has long sought a magic bullet that would selectively target and destroy malignant cells. Here, we show that bacteria injected intravenously into live animals selectively target solid tumors by employing optical imaging technique. LuxCDABE or GFP has been cloned into pUC19 plasmid to engineer pUC19lux or pUC19gfp. Engineered plasmid was transformed into varying kinds of wild type (MG1655) or mutant E. coli (DppGpp, PurE−, Crp−, FlhD−, RpoS−, etc.) strains. These light emitting bacteria were injected through tail vein into mouse bearing tumor. In vivo bioluminescence image was monitored periodically up to 14 days. The imaging signal from wild type E.coli was detected initially at liver (6 hrs) but at the necrotic region of tumor mass (day 3), and thereafter expanded to entire tumor mass (from day 4). No imaging signal was observed after injection of flagella mutant strain (FlhD−). Mutant defective in crp gene, upstream activator of flhD, reached necrotic region although slowly (day 5) but failed to expand to proliferative tumor region. Similarly the mutant strains defective in the expression of stress-related genes (DppGpp, rpoS−) could reach the necrotic region on day 3, but failed to expand further. Thus, we concluded that the targeting of tumor by E. coli is an active process: small number of circulating E. coli reach tumor through leaky vasculature via Chemotaxis first to necrotic region and later translocate to proliferative region. It is speculated that the translocation requires bacterial ability to sense a certain stress in the necrotic region that may although be nutritionally abundant.
Plenary V: Cell Trafficking and Cell Biology
Abstract ID: 047
Mesenchymal stem cells (MSCs) can give rise to cells of the osteoblast, chondrocyte and adipocyte lineage, upon the appropriate conditions in vitro and in vivo. These cells have great potential in bone tissue engineering and gene therapy. However, application of these cells is still hampered due to our incomplete understanding how cell fate is controlled.
Stable cell lines have frequently been used for the study of gene function, either by overexpression of the gene product, or by gene knock down experiments using RNA interference. Generating stable cell lines, however, is time consuming and tedious because integration of DNA in the genome is uncontrolled. Stable transformants often show extreme variability in expression of the introduced transgene due to the highly variable number of copies integrated into the genome and from positional effects on gene expression. In addition, integration can occur at a genomic locus involved in the regulation of differentiation or other cellular functions, thereby interfering or even disrupting the function of these genes independently of the transgene. All this compromises direct comparison of different expression constructs in clones obtained from the same parental cell line. The drawbacks in generation of stable cell lines could be circumvented by site-specific integration of DNA at a pre-selected site in the genome.
In this presentation the use of site specific integration of gene constructs for either gene overexpression or knock-down (by shRNA) will be discussed. Using this method we can not only rapidly study gene function in relation to mesenchymal differentiation but can also make all kinds of gene-reporters for e.g. differentiation, and cell signaling. This allows them to be used for High Throughput Screening in vitro. Since these cells carrying gene reporters can also be transplanted into mice, they can also be used to follow their fate by in vivo imaging.
Abstract ID: 048
The aim of this study was to assess the efficacy hMSC for targeting microscopic tumors and suicide gene or cytokine gene therapy. Immunodeficient mice were transplanted subcutaneously (s.c.) with human colon cancer cells of HT-29 Inv2 or CCS line and 3–4 days later intravenously with “tracer” hMSCs expressing herpes simplex virus type 1 thymidine kinase (HSVTK) and eGFP reporter genes. Subsequently, these s.c. tumors were examined for the specificity and magnitude of HSVTK+, eGFP+ stem cell engraftment and proliferation into tumor stroma by in vivo positron emission tomography (PET) with 18F-FHBG. In vivo PET images of tumors growing for 4 weeks demonstrated the presence of a significant portion of HSVTK+ tumor stroma with an average of 5.78 ± 4.79 % ID/g 18F-FHBG accumulation. In vivo imaging results were validated by in situ correlative histochemical, immuno-fluorescent, and -cytometric analyses, which revealed eGFP expression in vWF+ and CD31+ endothelial cells of capillaries and larger blood vessels, in germinal layer of dermis and hair follicles proximal to the s.c. tumor site. These differentiated HSVTK+, GFP+ endothelial cells had limited proliferative capacity and a short life-span of less than 2 weeks in tumor fragments transplanted into secondary hosts. We conclude that hMSCs can target microscopic tumors, subsequently proliferate and differentiate, and contribute to formation of a significant portion of tumor stroma. PET imaging should facilitate clinical translation of stem cell-based anti-cancer gene therapeutic approaches by providing the means for in vivo non-invasive whole body monitoring of trafficking, tumor targeting, proliferation of HSV1-tk expressing “tracer” hMSCs in tumor stroma.
Plenary VI: Molecular Imaging in the Drug Discovery Process (Cancer-related)
Abstract ID: 050
Abstract ID: 051
Many anticancer agents exert their lethal effect by inducing a series of different cellular responses which ultimately result in the initiation of apoptosis mainly through the activation of mitochondrial pathway. Death signals converging onto mitochondria, are integrated by proteins of Bcl-2 family that regulate the permeability of mitochondrial membrane and the release of pro-apoptotic soluble proteins. Here we discuss whether 99mTc-labeled lipophilic cations are able to detect activation of mitochondrial pathway and eventually alterations in the control of this pathway.
Several anti-cancer agents from main standard and experimental chemotherapeutic regimens were selected and used to induce apoptosis in control and Bcl-2 overexpressing breast cancer cells. Drug exposure caused an early increase of tracer uptake in Bcl-2 overexpressing breast cancer cells whereas control cells showed an opposite change in tracer uptake. To elucidate the molecular mechanism underlying such changes of tracer uptake, we tested the involvement of Bcl-2 phosphorylation, Bax translocation to mitochondria and the effect of Bad transient cotransfection. Finally, the observations obtained from cells have been translated to patients with non-Hodgkin's lymphoma by designing and performing clinical imaging studies before and immediately after treatment. In agreement with in vitro results, visualization of new lesions and increase of tumor-to-heart ratios were found in the early post-treatment scan of Bcl-2 positive patients. Our findings indicate that activation and dysregulation of mitochondrial pathway may be detected in vivo and predict tumor response.
Abstract ID: 052
Cellular interactions with the extracellular matrix and the migration therein are fundamental to tissue inflammation, repair, and cancer progression. Using time-resolved bright field, confocal, and multi-photon microscopy, we have reconstructed molecular live-cell dynamics in tumor during tissue invasion. Cell interactions with collagen fibers in 3D tissue culture models and the mouse dermis were obtained by time-lapse confocal reflection and multiphoton second harmonic generation imaging, showing the traction and remodeling of matrix structures by cancer cells (proteolytic migration mechanisms) as well as cellular shape changes adapting to tissue scaffolds (physical migration mechanisms). 5D semiquantitative fluorescence resconstruction was used to monitor (i) cytoskeletal dynamics of filamentous actin by invading cancer cells, (ii) the subcellular topography of protease resdistibution and proteolytic cleavage of the extracellular matrix structures, and (iii) structural changes and dislocation of ECM fibrils by moving cells. These approaches have provided the novel insight into salvage pathways in tumor cell migration after molecular interrerence with protease and adhesion receptor systems, termed the mesenchymal-amoeboid transition. After pharmacotherapeutic interference with protease function and concommitant loss of pericellular proteolysis, tumor cells switch to an non-proteolytic, amoeboid migration mechanisms driven by amoeboid shape-changes and cell deformation rather than ECM remodeling. Likewise, using different strategies to interfere with β1 integrin-meditate cell-matrix interactions, an integrin-independent amoeboid migration mechanism rescues motility via low-stringent interactions with extracellular matrix. Multimodal interference with alternative adhesion systems showed, that persisting migration was provided by β3 integrins and glycosaminoglycans from the cell surface, suggesting a role of the glycocalyx in residual cell migration after the loss of intergrin function. In conclusion, time-resolved imaging in 3D tissue models and in vivo represent a ‘dynamic pathology’ approach to monitor molecular cell and tissue patterning relevant to basic and applied cancer research.
Symposium X: Molecular Targets in Cardiovascular Disease
Abstract ID: 053
The concept of the vulnerable coronary plaque has recently emerged to explain how quiescent atherosclerotic lesions evolve to cause clinical events. Vulnerable plaques are generally non obstructive, asymptomatic lesions that may abruptly rupture and induce thrombotic occlusion leading to tissue ischemia and its attendant sequelae. Coronary plaque rupture is the most common type of plaque complication, accounting for approximately 70% of fatal acute myocardial infarctions and/or sudden coronary deaths. Thrombotic complications that arise from rupture or erosion of a vulnerable plaque may be clinically silent yet contribute to the natural history of plaque progression and ultimately luminal stenosis. Non-invasive molecular imaging of vulnerable atherosclerotic lesions has attracted much interest recently to identify patients at risk for thromboischemic vascular events. Many radionuclide imaging approaches have been evaluated to detect atherosclerotic lesions including lipoproteins, fibrinogen, cytokines, MMP-inhibitors, and ligands for the vitronectin receptor alphavbeta3. Moreover, platelets and their soluble collagen receptor glycoprotein (GP) VI have been shown to detect vulnerable vascular lesion that are proned for thrombotic events, thus acute thrombotic vessel occlusion. Molecular imaging of atherosclerotic plaque may have important clinical implications in the near future such as studying the natural history of atherosclerosis, evaluating novel drug or genetic therapies on progression and regression of atherosclerosis, evaluating plaque stability, screening and serial follow-up of high-risk individuals, non-invasive imaging of vulnerable plaques, and assessing the clinical efficacy of new treatments of atherosclerosis.
Abstract ID: 054
Incremental increases of HDL levels are strongly associated with ameliorated cardiovascular outcome. Beyond its role in reverse cholesterol transport and anti-atherogenesis, HDL exerts a variety of direct vasoprotective effects via activation of the endothelial nitric oxide synthase (eNOS). We have demonstrated that this mechanism is involved in HDL-induced vasodilation in vitro and its increase of myocardial perfusion in vivo. Furthermore, administration of human HDL reduced myocardial infarct size after ischemia/reperfusion by approximately 20%. This cardioprotective effect was mediated by both eNOS-dependent inhibition of leucocyte adhesion by HDL in vitro and in the infarcted area, and direct inhibition of caspase-mediated apoptosis in cardiomyocytes in vitro and in the injured myocardium.
In all of these studies, we could attribute the vaso- and cardioprotective functions of HDL to several lysophospholipids associated with the HDL particle such as sphingosine-1 phosphate (S1P). Furthermore, by using mice deficient in the lysophospholipid receptor S1P3, we identified S1P3 as the responsible receptor. S1P receptor agonists mediate a variety of physiological processes from immunity to cardiogenesis via their cognate G protein-coupled receptors, of which S1P1, S1P2 and S1P3 are abundantly and differentially expressed in the vasculature and heart. Interestingly, the first general pharmacological agonists of S1P receptors are currently undergoing Phase III clinical trials in transplantation medicine. We have shown that one of these, FTY720, that posesses high affinity for 4 of the 5 known lysophospholipid receptors completely and potently mimics the vasodilatory effect of S1P and HDL. As even more small molecular and, especially, receptor subtype-specific S1P receptor agonists are currently being developed by the pharmaceutical industry, their potential for labelling and imaging cardiovascular diseases in which either the receptor expression levels, their functionality or the effects of their natural ligands and/or HDL are altered may open a new area of both diagnostic and therapeutical applications.
Abstract ID: 055
Stem cells in current human trials include circulating (endothelial) progenitor cells, unfractionated bone marrow, and skeletal muscle myoblasts. Stem cells in preclinical studies include bone marrow mesenchymal stem cells, multipotent cells from other sources, and novel progenitor or stem cells discovered in the adult myocardium. Also, the potential of pluripotent embryonal stem cells is currently being evaluated using different approaches.
Existing trials use direct injection during cardiac surgery, intra-muscular delivery via catheters (e.g., the NOGA system for electro-mechanical mapping), or intracoronary delivery routes (over-the-wire balloon catheters). In addition, there is the theoretical potential for systemic delivery, suggested by the homing of some cell types to infarcted myocardium, and strategies to mobilize endogenous cells from other tissue sites to the heart.
Molecular imaging gains increasing importance in two directions: It can be used to achieve cell trafficking looking at the fate of stem cells investigated. Molecular imaging also allows for the evaluation of certain functional endpoints of stem cell therapy such as myocardial (or other organ) function and perfusion using appropriate biological markers.
Symposium XI: Imaging in Drug Discovery and Drug Development (not Cancer-related)
Abstract ID: 056
Xenogen's technology is based on the detection and quantification of light transmitted through a living animal. Microorganisms such as bacteria, fungi and viruses, as well as tumor cells are tagged with a luciferase gene and their growth is non-invasively monitored via photon counting over time. In addition, gene expression in living transgenic animals can be monitored by using promoters of interest to drive luciferase, thus creating LPTAs [Light Producing Transgenic Animals]. The promoters that are chosen drive the expression of proteins in a pathway that we wish to study. The animals are then used for a variety of purposes including metabolism and toxicology studies as well as for drug evaluation. For example, we have placed the promoters for cytochrome P-450 enzymes, iNOS, insulin, GFAP, NF-kB, VEGFR2 and IL-2 in front of the luciferase gene and generated transgenic mice. We then use these animals to detect the transcriptional activation of these genes during and after treatment with various drugs. Our results indicate that in-vivo biophotonic imaging can be used to monitor and quantify gene expression and to study processes in which these genes are induced.
Abstract ID: 057
Neurochemical imaging studies can identify molecular targets of abused drugs and link them to the underlying pathology associated with behaviors such as drug dependence, addiction, and withdrawal. Positron Emission Tomography (PET) is opening new avenues for the investigation of the neurochemical disturbances underlying drug abuse and addiction and the in vivo mechanisms by which medications might ameliorate these conditions. PET can identify vulnerable human populations, treatment strategies and monitor treatment efficacy. Thus, with this tool and the knowledge it provides, the potential for developing novel drugs and treatment strategies for drug addiction is now close at hand.
Symposium XII: Advanced Optical Technologies
Abstract ID: 058
Conventional intravital confocal microscopes provide a window for real-time viewing of the complex and dynamic behavior of individual cells interacting within the host environment. However, the large size of the objective lenses used often requires externalization of the tissue imaged. Recent advances in miniaturization of objective lenses and scanning mechanisms have allowed for the development of fiber-optic instruments whose reduced size improves access to tissue, and can be used for both small animal and medical imaging. In this presentation, the challenges associated with miniaturization of intravital confocal microscopes will be discussed in the context of existing single and dual axes prototypes. Single axis confocal microscopes using either a single optical fiber or a fiber optic bundle have been developed for the collection of fluorescence with 488 nm excitation. Examples of this approach to monitor lymphocyte trafficking patterns in CXCR6-GFP knock-in mice and detection of colonic neoplasia during endoscopy will be presented. The dual axes confocal architecture uses two low numerical aperture objectives oriented with the illumination and collection beams crossed at an angle. This configuration significantly reduces the axial resolution, and achieves a dynamic range that allows for collection of vertical cross-sections over 1 mm deep. Furthermore, light scattered along the illumination path is rejected, resulting in improved image contrast. In addition, the long working distance allow for a MEMS (micro-electro-mechanical systems) mirror to be placed on the tissue side of the objective, creating a large field of view. Examples of this architecture to image normal and diseased colon, esophagus, and skin at 1300 nm will be presented. While the initial results appear promising, the future scientific and clinical impact of this imaging modality will be determined by continued advances in instrument miniaturization and improvements in imaging performance, such as resolution, tissue penetration depth, and field of view.
Abstract ID: 059
A new implementation of the theta principle [1,2] takes advantage of parallel recording. This high-resolution light microscope has been developed for modern life sciences. It is designed to image large samples (embryos, three-dimensional cell cultures) down to the subcellular level. The fundamental principle of Selective Plane Illumination Microscopy (SPLM) remains the detection of fluorescence light perpendicular to the illumination axis. The illumination system selectively excites fluorophores within an entire plane, which is the focal plane of a detection system consisting of a long working distance lens and a CCD camera. SPIMaging provides optical sectioning directly. Bleaching outside the thin volume of interest is avoided. Since it performs well with long working distance lenses and has a good penetration depth millimetre-sized specimens can be observed in their entirety. To further increase the resolution and the information content of the data stack, rotation of the sample changes excitation and detection axes with respect to the sample. Parts of the sample that would otherwise be hidden or obscured become accessible. Data stacks recorded at different angles are combined in post-processing steps to yield high-resolution images of complete samples [3]. The 3D resolution is then dominated by the lateral resolution and resolution becomes identical along all directions. For the past years various different species (Medaka, drosophila, yeast, …) were observed with the instrument. SPIM technology opens up new dimensions in many disciplines. In the emerging field of 3D cell cultures SPLM is an ideal instrument. It enables researchers to image whole populations of cells in their 3D context, to visualize their morphology in a matrix and to track cells through the context of surrounding tissue.
Abstract ID: 060
Novel techniques in fluorescence light microscopy provide deeper insight into sub-cellular processes. However, it is still challenging to measure the size and positions e.g. of specific domains of the human genome on the nano-scale with high precision in three dimensionally conserved cell nuclei. Spatially Modulated Illumination (SMI) is a light optical technique which allows the measurement of certain features of individual fluorescent structures well below the conventional optical resolution limit inside fixed cells [2].
An automated size determination procedure using Spatially Modulated Illumination Microscopy (SMI) [1] in combination with reference objects of known size and shape is applied to determine the sizes of single and up to hundreds of individual fluorescent objects within a single cell.
This and other advanced microscope systems, like e.g. 4Pi microscopy, have been applied to measure the size of macromolecular complexes [2] and of individual small gene regions [3]. The potential of such “light nanoscopy” approaches extends to the “in situ” analysis of cellular protein-protein and protein-nucleic acid interactions, even for objects whose volumes are about two orders of magnitude smaller than the illumination or observation volume, respectively, in a confocal laser scanning microscope using a high numerical aperture objective lens [4]. To achieve these highest structural topological and size resolution, questions on the influence of the fluorescence labelling procedures, e.g. FISH and anti-body labelling on the detected objects, especially the labelling efficiency and specificity have to be addressed.
Symposium XIII: Infectious Agents and Host Response
Abstract ID: 061
The methods of molecular imaging have now been extensively employed to study the kinetics of infection in small animals, revealing the location of many diverse infectious agents throughout the course of disease. To study the kinetics of the pathogenic process in a controlled way under circumstances amenable to extensive genetic, cellular, and immunological manipulation, we have brought imaging to bear on one of the most well characterized models of infection, murine listeriosis. In the many decades since its inception, this model has yielded molecular details of actin-based intracellular motility, epitope presentation and T cell response, and species specificity of bacterial-epithelial interaction. Murine listeriosis also provided the very first described demonstration of specific cellular immunity to any infection, by Mackaness in 1962. This model has thus led to discoveries that far transcend the boundaries of the disease itself, and reach into the basic biology of hosts and pathogens. The causative agent of listeriosis, the Gram-positive bacterium Listeria monocytogenes, infects macrophages and many other host cells, leading to a systemic disease in humans with a high fatality rate of 20% despite the use of antibiotics. Immunocompromised individuals such as AIDS patients and transplant recipients are at particular risk, as well as the fetuses of pregnant women. We genetically labeled L. monocytogenes with a bacterial lux operon encoding luciferase and enzymes for the production of the bacterial luciferin, and employed this strain to image listeriosis in mice using in vivo bioluminescence imaging. The results demonstrated that L. monocytogenes grows in previously unsuspected tissue sites, including the gall bladder. Here, we present the most recent observations, including the surprising persistence of highly attenuated mutants, and discuss the implications of imaging as a window into the process of bacterial infection.
Abstract ID: 062
Studies of viral and host factors that influence pathogenesis largely have used experimental mouse models that rely upon sacrifice of infected mice to determine distribution and titer of virus. While this experimental paradigm has provided important data, it precludes real-time investigations of the same animal over the entire course of disease progression. Research by our laboratory and others demonstrates several significant advantages of investigating viral-host pathogenesis with imaging. Spatial and temporal progression of infection can be quantified in the same animals, identifying animal-to animal variations in viral replication and dissemination and host immunity. Imaging greatly increases data obtained about viral pathogenesis in living mice compared with other global assays of disease progression, such as weight loss or external signs of disease. Imaging data also allow relative amounts of viral replication in various anatomic sites to be quantified over time. This presentation will focus on studies of two different DNA viruses, Herpes simplex virus type 1 and vaccinia virus, as models of host-pathogen interactions. We have developed reporter viruses that enable viral replication and dissemination to be monitored with bioluminescence and fluorescence imaging techniques. These viruses have been used to investigate mechanisms of viral spread in vivo and effects of interferons in dissemination of these pathogens. Collectively, this research demonstrates the potential of cellular and molecular imaging to advance studies of viral pathogenesis and response to therapy.
Abstract ID: 063
Evolving trends in multi-drug resistance in microbial pathogens that commonly infect humans (e.g., S. aureus) underscore the crisis in the current and future ability to treat serious microbial infections. This is particularly relevant to vascular system infections. Therefore, there is an urgent need to identify, understand and target critical molecular components that modulate the activation of key virulence gene networks of such pathogens. Several approaches have been developed to detect microbial gene expression within the host during experimental infection, including: reverse transcription (RT)-PCR; signature-tagged mutagenesis; and green fluorescent protein (GFP) reporters. Although these approaches are intriguing, they suffer many problems, including: i) require animal sacrifice; ii) don't provide real-time serial temporal assessments of gene expression; and iii) need large number animals. Recently a novel photon capture imaging system (IVIS) has been developed as a particularly attractive technology for potentially tracking microbial gene expression in vivo. We have been used the IVIS system, a luciferase operon (lux) and an experimental rat endocarditis model to real-time monitor efficacy of conventional antibiotics in treating bacterial infection and to detect S. aureus virulence gene expression profiles in living animals in our laboratory. Using this system, living animals can be serially queried over whatever time-frame is desired for a particular gene promoter-of-interest. Additionally, for constitutively-expressed promoters, once a robust correlation between photon signals and bacterial densities within a target tissue is established, bioluminescence can serve as a faithful and precise surrogate for assessing genetically-active tissue bacterial counts in vivo. In conclusion, this new imaging technology has several significant and unique features, including a unique platform technology for basic life-sciences research (i.e., real-time assessments of microbial virulence gene expression in living animals); and disclosing new avenues towards novel antimicrobial development in terms of agents which target virulence gene regulation or expression.
Symposium XIV: Design Methods for Imaging Probes
Abstract ID: 064
Accurate and rapid detection of diseases will facilitate early medical intervention to prevent or eradicate the disease. While many human diseases have been studied successfully by using differences in the intrinsic optical properties of normal and pathologic tissues, molecular imaging of the expression of aberrant genes, proteins, and other pathophysiologic processes would be enhanced by the use of highly specific exogenous molecular probes. Accordingly, we have developed a variety of molecular contrast effectors for imaging tumors. The carrier molecules are typically small peptides, whose structures were optimized for their biological target by molecular modeling methods. Labeling of the compounds possessing high receptor binding affinity with near infrared fluorescent probes afforded receptor-specific molecular probes for in vitro and in vivo applications. Some examples include new GRD peptides for imaging beta-3 integrin and octreotate analogues for targeting somatostatin receptors (neuroendocrine tumors). The results show that these probes are highly specific for their target receptors. In addition to tumor imaging, the molecular probes are also useful for monitoring the response of the tumors to treatment. Incorporation of antennas for other imaging modalities such as radiolabeled derivatives provides complementary diagnostic and prognostic information.
Abstract ID: 065
The integrin αvβ3 plays an important role in angiogenesis and tumor cell metastasis, and is currently being evaluated as a target for new therapeutic approaches. Several techniques for noninvasive determination of αvβ3 expression have been suggested. We developed several cyclic peptides, peptidomimetics, carbohydrated and hydrophilic peptides as well as homo- and heteromultimeric tracers for imaging αvβ3-integrins with PET and SPECT. All compounds were evaluated in vitro and in suitable animal models using IHC and μ-PET imaging. These and subsequent studies in patients show, that [18F] Galacto-RGD-PET enables noninvasive quantitative mapping of the αvβ3 expression on tumor and endothelial cells. Furthermore, new labeling strategies were developed and used for the synthesis of multimeric RGD-constructs, which resulted in a further improvement of affinity and target-to-nontarget accumulation ratios. These tracer may offer insights into molecular processes during development and dissemination of tumors in preclinical as well as clinical settings, and will be a helpful tool in planning and controlling novel αvβ3-directed therapies.
Abstract ID: 066
The development of targeted imaging and therapeutic agents has involved numerous decisions into the selection of components, which comprise these macromolecular structures. Over ten years ago, we developed a novel perfluorocarbon agent, which was derived from blood substitutes and initially used for targeted acoustic imaging. Since that time, we have further modified the platform for MRI, SPECT, optical and CT imaging. Beyond diagnostic applications, we have extended the platform to encompass therapy with a variety of compounds to treat tumors, atherosclerotic disease, prevent post angioplasty restenosis, and for vascularly constrained thrombolysis. To achieve these goals, we have employed a variety of in vitro analytical techniques and phantom models to accumulate data to allow evidence based decisions. Although the highest throughput and most important screening technique employed is the Gedanken experiment, we will present a variety of experimental designs used routinely in the development of these novel agents.
Symposium XV: Molecular and Cellular Mechanisms of Angiogenesis
Abstract ID: 067
Influence on tumor angiogenesis, is emerging as a promising target in the treatment of malignancies. Therefore, monitoring of angiogenesis-relevant approaches with functional imaging and histomorphometrical analyses are desirable to evaluate the biological effects.
Using a bicistronic retroviral vector for angiopoietin-2 (Ang-2) Morris hepatoma (MH3924A) cell lines with Ang-2 expression were generated (Ang-2-MH3924A). In human umbilical vein endothelial cells (HUVEC) cocultured with Ang-2-MH3924A, the proliferative action with/without growth factor was determined using a Coulter counter. Furthermore, in vivo experiments were done in order to measure the effects on tumor growth and perfusion. Finally, the tumors were examined by immunohistochemistry (including computer-assisted morphometry) and DNA chip analysis.
Stable Ang-2-expressing hepatoma cells enhanced bFGF-mediated endothelial cell proliferation in vitro. In vivo, perfusion, as measured by H215O PET, was increased in genetically modified tumors, whereas the blood pool (measured by 68Gallium-labeled albumin) was not different between the two groups. However, despite an increased perfusion in Ang-2-MH3924A, the tumor growth was similar to wild type MH3924A (WT-MH3924A). Consistent with the increased perfusion, the immunohistochemically quantified micro- and macrovascularization (indicated by CD31- and a-actin immunoreactive area) were increased in Ang-2-MH3924A. DNA chip analysis of tumors with gene transfer showed an induction of angiogenesis-promoting genes, as well as genes related to extracellular matrix, apoptosis, signal transduction and oxidative stress.
Our results suggest that Ang-2-expression increases perfusion / 4growth, accompanied by an enhanced expression of pro-angiogenic genes. Enhanced expression of genes for cellular matrix, apoptosis, signal transduction and angiogenesis indicates the divergent role of Ang-2 in destabilizing signals on vessels and angiogenesis.
Abstract ID: 068
The luminal surface of vascular endothelium is inherently accessible to imaging and therapeutic agents circulating in the blood. This interface has specialized transport structures or invaginations called caveolae for endocytosis and transcytosis. Here, we investigate in vivo the endothelial processing of specific antibodies targeting caveolae vs. the general endothelial cell surface, primarily in the neovasculature of solid tumors. Our comprehensive proteomic mapping of endothelial plasma membranes and caveolae in normal organs and various solid tumors from rodents and humans reveals >1000 proteins with about 5–10% showing restricted expression among the tissues. Using a variety of imaging techniques including fluorescence intravital microscopy and dynamic SPECT/CT, we find very rapid binding and transport across the endothelial cell barrier of caveolae-targeting antibodies which facilitates effective penetration throughout solid tumors within minutes to hours of intravenous injection. This transport can occur against a significant concentration gradient and thus by definition the caveolae appear to mediate active transport. Significant differences in antibody processing exist in normal vs neoplastic tissue. Interstitial transport appears more compartmentalized and significantly slower in tumors. Normal tissues are thoroughly flooded with antibody within minutes of intravenous injection whereas tumors may require so far up to 4 hrs. The new neovasular targets and their antibodies provide a means to image angiogenesis potentially in multiple diseases and to deliver more sophisticated imaging and therapeutic agents into specific tissues of the body.
Poster Session: Biological Molecular Imaging, Part 1
Abstract ID: 070
Angiogenesis describes the process of new blood vessel formation and is a critical component of development, wound healing and tumor growth. In an effort to non-invasively monitor blood vessel growth in vivo by bioluminescent imaging, we introduced a luciferase transgene into the endogenous VEGFR2 [vascular endothelial cell growth factor receptor 2] locus of murine ES cells. VEGFR2 is a major receptor for VEGF on the surface of proliferating endothelial cells. Undifferentiated murine ES cells do not express VEGFR2 and correspondingly our targeted ES cells did not bioluminesce. When we let these ES cells differentiate into embryoid bodies, however, we observed an increase in bioluminescence, which could be inhibited by treatment with suramin. Injection of these ES cells into nude mice resulted in the formation of teratomas. The initial inoculum did not glow, but we observed an induction of luciferase expression once differentiated into a teratoma. Histological examination of the teratomas showed that a proportion of the ES cells had differentiated into vascular beds. We generated mice from the VEGFR2-luc ES cells, such that they expressed luciferase in their endothelial cells. To enable tumor xenograft studies, we crossed the VEGFR2-luc allele onto a nu/nu genetic background. These mice were then injected with non-bioluminescent tumor cells, thus enabling bioluminescent measurement of de novo tumor angiogenesis using an IVIS® Imaging System and Living Image® software. As tumor size increased, we observed a concomitant increase in bioluminescence signal. Histological analysis of the tumor correlated luciferase expression with markers associated with blood vessels. Taken together, our data suggests that this VEGFR2–luc transgenic mouse will be a valuable reagent with which to non-invasively monitor tumor angiogenesis in vivo as well as to measure the effects of drugs on the angiogenic process.
Abstract ID: 071
Vector imaging in vivo has remained a challenge, particularly by MRI. We describe here a technique for the visualization of viral vector delivery by magnetic resonance imaging (MRI) in vivo. By conjugating avidin-coated baculoviral vectors (Baavi) with biotinylated ultra-small (50 nm) superparamagnetic iron oxide particles (USPIO), we were able to produce vector-related MRI contrast in the choroid plexus cells of rat brain in vivo (Image 1. Timepoints as days). 10 μl of 2.5 × 10E10 pfu/ml nuclear-targeted LacZ-encoding Baavi with bUSPIO coating was injected into rat brain lateral ventricles and visualized by MRI at 4.7T. Altered MRI contrast was detected in the injected ventricles. No specific signal was detected when wild-type baculoviruses or intact biotinylated USPIO particles were injected into the lateral ventricles. Cryosectioned brains were stained for nuclear targeted beta-galactosidase gene expression, which was found to co-localize with MRI contrast to choroid plexus cells in the injected ventricle. This study provides the first proof of principle for robust and non-invasive viral vector MRI by using avidin-displaying baculoviruses in vivo. Considering the ubiquitous role of MRI in current medical imaging, the approach is likely to provide useful applications for in vivo imaging of therapeutic gene transfer.
Abstract ID: 072
After an initial response to chemotherapy, most cancers develop resistance to treatment. The anticancer alkylating agent temozolomide (TMZ) showed an initial inhibitory effect on the growth of wild type breast carcinoma MCF-7wt tumors in mice, and at later time points the tumor volume gradually increased suggesting development of drug resistance. In contrast, no growth inhibitory effect of TMZ was observed in MCF-7adr adriamycin resistant phenotype tumors. In vitro TMZ exposure demonstrated that no P-glycoprotein expression was detected for a TMZ dose range used in vivo. Effect of TMZ chemotherapy on the tumor vasculature was evaluated by dynamic MRI of albumin-GdDTPA conjugates. Vascular volume (VV) and permeability surface area (PS) of wild type tumors decreased during TMZ chemotherapy for 5 weeks (2 mg/kg/day × 2 doses/week), while VV and PS of the drug resistant phenotype increased with TMZ chemotherapy ( Figure 1 ). Vascular endothelial growth factor (VEGF) expression of MCF-7wt was suppressed by TMZ at a dose of 50 μg/ml, suggesting this as one possible mechanism of the anti-vascular effect of TMZ within the tumor. In fact, even in the wild type non-treated tumors 1H/13C magnetic resonance spectroscopic imaging and gadolinium enhanced dynamic MRI demonstrated non-uniform distributions of 13C-labeled TMZ and the contrast agent, respectively. This functional MRI study revealed that the vascular effects of TMZ chemotherapy can contribute to physiological drug resistance in wild type tumors. In drug resistant tumors, increase in VV and PS caused by TMZ treatment did not result in the sensitization of the tumor to the therapy possibly due to an efficient efflux of the drug from the cells by multidrug resistance proteins.
We acknowledge support from NIH RO1 CA097310.
Three-dimensional images of vascular volume (VV) and permeability surface area (PS) distributions in the MCF-7wt and MCF-7adr tumor xenograft.
Abstract ID: 073
Abstract ID: 074
The purpose of this study was to evaluate the feasibility of 17O as a label for an MRI agent. Previously reported studies have demonstrated that 17O-labeled water is effective in measuring tissue perfusion or oxygen utilization. In contrast to these studies, our previous studies showed that 17O-labeled phosphoric compounds can be detected by using heteronuclear multiple quantum coherence (HMQC) between 17O and 31P. In this report, we present a new scalar-decoupling sequence to improve the signal-to-noise ratio (SNR) from 17O measurement and validate this improved SNR by phantom experiments.
Experiments were performed using an 11.7-T NMR system (Bruker AMX-500) equipped with 17O-31P-1H-2H quadruple RF coils and a z-gradient coil. A sample tube with 25.7 atom% randomly 17O-labeled phosphoric acid was measured. A new scalar-decoupling sequence based on an asymmetric spin-echo sequence and 17O decoupling pulse (on/off) was used. Figure 1 shows the resultant two signal decays with increasing TE; the signal decay without decoupling shows a unique non-simple exponential decay. The difference between the two signals for scalar-decoupling on and off, which represents the signal from 17O-labeled molecules, reaches its maximum at TE of 4 ms. Figure 2 shows two signal-intensity profiles measured by using the scalar-decoupling sequence and the HMQC sequence at the same TR of 2 s and an accumulation number of 128. Each TE was set to obtain maximum signal intensity; that is a TE of 1.8 ms for HMQC. The graph shows that the scalar-decoupling sequence improves SNR by about 20% for this sample.
Abstract ID: 075
Self-assembled peptide amphiphile (PA) nanofibers have been investigated for their potential use as scaffolds for tissue engineering and drug delivery applications. Nanofiber networks formed from PAs contain targeting epitopes can be useful for the preparation of specific biomaterial scaffolds. We report on the synthesis of a new magnetic resonance (MR) peptide amphiphile molecules that self-assemble into spherical and fiber-like nanostructures enhancing T1 relaxation time. We have modified these MR detectable peptide-amphiphiles placing the contrast agent in different positions to test the properties of the supramolecular assemblies formed. These MR active PAs mix with a variety of different epitope bearing PAs and PAs with different structural characteristics. We have found that adding the MR contrast agent PAs produces a gel with three times enhancement of T1 relaxation compared to control gels at a concentration of 0.1 mM at 400 Mhz. This new class of MR contrast agents can potentially be used to combine high-resolution three-dimensional MR fate mapping of tissue-engineered scaffolds with targeting of specific cellular receptors.
Abstract ID: 076
Integrins such as αvβ3 have been described as potential targets for imaging tumor neovascualture in vivo. Peptidic substrates like the RGD-peptide have been identified to exhibit high binding affinity to these targets.
An αvβ3 receptor specific fluorophore was synthetized by labelling the RGD-peptide cyclic-(Cys-Arg-Gly-Asp-Cys)-Gly-Lys with a cyanine dye (Cy 5.5). M21 human melanoma cells were incubated with non modified Cy 5.5 or RGD-Cy 5.5, respectively and visualized by fluorescence microscopy. Binding specificity was assessed by adding unlabeled RGD-peptide prior to the incubation with RGD-Cy 5.5. For in vivo studies M21 tumor bearing mice were i.v. injected with either 2 nmol of Cy 5.5 or RGD-Cy 5.5 respectively. In order to specifically block the αvβ3-related signal, 250 nmol of the non-labeled RGD-peptide were administered prior to the injection of RGD-Cy 5.5. Fluorescence reflectance imaging (FRI) and fluorescence mediated tomography (FMT) were performed at different time points after injection of the fluorescent tracer.
In vitro, the peptide-dye conjugate showed a distinct binding affinity to αvβ3-positive M21 melanoma cells. Binding of RGD-Cy 5.5 could be blocked by co-incubation of the conjugate with the unlabeled RGD-peptide. In vivo M21 human melanoma xenografts were clearly visualized up to 24 h after i.v. injection of RGD-Cy 5.5 revealing an up to 40 % higher target-to-background ratio as compared to the non-modified dye (p < 0.05). The RDG-Cy 5.5 signal could partially be blocked by pre-injection the unlabeled RGD-peptide.
RGD-Cy 5.5 may be used as a specific optical tracer for in vivo detection of αvβ3-expression and may thus facilitate tumor detection in vivo.
Abstract ID: 077
We developed the amphiphilic chelate Gd-EPTPA-C16, as a potential MRI CA, with a good water exchange rate and a favorable rotational correlation time due to micelle formation, thus displaying a reaonably high relaxivity. In this work we studied the micelle size distribution of the complex in solution using Dynamic Light Scattering (DLS). Gamma-imaging and biodistribution studies were made of 153Sm-EPTPA-C16 on male Wistar rats. The data shows that its the targeting organs were lung, liver and spleen. These facts can be explained by the presence in these organs of particulate filtration systems (alveolar macrophages in lung and Kupffer cells in liver and spleen) able to retain particles, like those formed by the complex. The low activity in the brain indicates that the complex does not cross the BBB. A MRI study of the Gd-EPTPA-C16 complex was performed on male Wistar rats at 7.0-T on a Bruker Pharmascan system. T1-weighted spin-echo anatomical images were acquired in sagittal, horizontal and axial orientations. Baseline images were obtained prior to administration of our CA, injected into the catheterized tail vein as a bolus of 250 μl containing 0.2 mmol kg−1. In the first animal, a pharmacokinetic study was constructed with the same T1-weighted sequence in an axial orientation, during 90 min, to obtain the CA time-response curve. The maximum signal intensity in the liver occurs within 9–10 min after the injection. Therefore, T1-weighted acquisitions were performed around this time point after injection of the contrast agent. An enhancement was observed in the liver upon injection of the Gd-EPTPA-C16 contrast agent, which was much higher than that observed in the same conditions using the commercial agent Gd-DTPA (Magnevist®, Schering, Germany).
We thank the support from F.C.T., Portugal (project POCTI/QUI/47005/2002), EU COST Action D18 “Lanthanide chemistry for diagnosis and therapy” and the European-funded EMIL programme (LSHC-2004-503569).
Abstract ID: 078
Gastrin-releasing peptide receptors (GRPR) are overexpressed on human breast cancers. Bombesin is a fourteen amino acid peptide that binds with high affinity to GRPR and can be labeled with radiometals through the use of a bifunctional chelator that is conjugated to the peptide. The purpose of this study was to evaluate bombesin (BN) analogs radiolabeled with copper-64 (64Cu) that contain various sized linkers between the chelator and the peptide for targeting breast cancer tumors that express GRPR. The chelator 1, 4, 7, 10-tetraazacyclododecane-1, 4, 7, 10-tetraacetic acid (DOTA) was conjugated to the eight C-terminal amino acids of BN (BN(7–14)) using five different alkyl linkers to form DOTA-linker-BN(7–14). The linkers were 4, 5, 6, 8, and 11 carbons in length (Aba, Ava, Ahx, Aoc, and Ado, respectively) Human T-47D breast ductal carcinoma cells were used for the evaluation of the analogs in vitro and in vivo. Whole cells were used for a competitive binding assay to give IC50 values of 82, 42, 17, 7, and 26 nM for the analogs with the Aba, Ava, Ahx, Aoc, and Ado linkers respectively. The analogs were then evaluated by internalization assays using 64Cu-labeled BN(7-14) (64Cu-DOTA-linker-BN). The rates and maximal amounts of internalized 64Cu-DOTA-linker-BN varied depending on the linker. In addition, in vivo biodistribution studies of 64Cu-DOTA-BN in SCID mice bearing T-47D xenografts showed a similar effect, with the amount of uptake being dependent on the linker. The 64Cu-DOTA- Aoc-BN showed the highest tumor uptake of 4.6 ± 0.4% ID/g. Also, uptake was found to be receptor-mediated. Based on these results, these analogs will be used for microPET imaging of T-47D breast cancer in an animal model. In addition, by determining which of the linkers provides the highest receptor affinity, internalization, and tumor uptake, we will expand the use of these analogs to other tumors that express GRPR.
Abstract ID: 079
Change in the number of labeled cells has been reported to change the signal intensity during in vivo and ex vivo MRI, but no study to our knowledge has been done to compare the effect of intracellular versus free iron on MRI. Furthermore, there is no definite quantitative model for estimating the number of incorporated cells in the tissue of interest. The purposes of this study were to compare the effects of intracellular ferumoxides (FE) to those of free FE on T2-weighted MRI, and to make a phantom to quantify the number of cells. Gliosarcoma (9L) and T-cells were labeled with FE-protamine sulfate. Labeled cells (from 3.75 × 103 to 2 × 106) were put into tubes containing 1 ml of 4% gelatin, mixed thoroughly and quickly solidified. T2-weighted images were acquired using both 7T and 3T MR systems. Signal intensities were calculated from T2-weighted images and R2 maps were created for all tubes that contained cells and free FE. There was linear correlation between the number of labeled cells and the value of R2. However, despite linear correlation between the R2 values and cell number, there was a significant difference in the R2 values for T-cells and gliosarcoma cells for an equivalent number of labeled cells. The R2 values of gliosarcoma cells were twice those of T-cells at each cellular concentration. Additionally, as expected, both the 3 and 7 Tesla systems showed almost identical R2 values for both cell types. Iron oxide in gel showed higher R2 values than did intracellular iron at each concentration. Intracellular iron both in T-cells and gliosarcoma cells, showed lower R2 values than did the corresponding free iron concentration in gel. These preliminary data propose method to differentiate free iron from intracellular iron and prepare a model to quantify the number of cells for in vivo MRI studies.
Abstract ID: 080
Abstract ID: 081
Membrane Phantom: Diameter=1.8 mm. 1.5T, TE=10 ms, % gradient rephasing=2.5.
Abstract ID: 082
Mouse carotid arteries (CA) were studied in an ex vivo setup (images 1A-B). An isolated artery is mounted in a home-built perfusion chamber, pressurized (80mmHg transmural pressure), and imaged using a water dipping objective (60×, NA=1.0). Resolution=0.5μm.
For in vivo imaging, an anesthetized rat was positioned on the microscope stage with a renal artery exposed (same objective). To reduce motional disturbances due to heart beat and respiration, a triggering system was used to acquire images at the same stage of the cardiac cycle. Combined with short exposure times, this resulted in stable optical sections.
TPLSM (combined with a triggering system) enables us to study the development of vascular diseases in intact large arteries at a sub-cellular level.
Examples of optical sections of an axial series of mouse CA. Figure 1A: Elastin network in CA of control mouse, stained with eosin. Figure 1B: Atherosclerotic plaque in CA of apoE-/- mouse. Cytoplasm is stained with acridin red, nuclei are stained with SYTO13.
Abstract ID: 083
A new macrocyclic DOTA-like ligand (BPAMD) for bone imaging and therapy containing a mono-amide bis(phosphonic acid) bone-seeking group was designed and synthesized. Its lanthanide(III) complexes were prepared and characterized by 1H and 31P NMR spectroscopy. The Gd(III)-BPAMD complex was investigated in detail by 1H and 17O relaxometric studies to inspect parameters relevant for its potential application as a MRI contrast agent. Sorption experiments were conducted with Gd(III) and Tb(III) complexes using hydroxyapatite (HA) as a model of bone surface. Very effective uptake of the Gd-BPAMD complex by the HA surface was observed in NMR experiments. Radiochemical studies with the (160Tb-BPAMD)-HA system proved the sorption to be remarkably fast and strong on one hand and fully reversible on the other hand. The strong (Gd-BPAMD)-HA interaction was also supported by 1H NMRD measurements in the presence of a hydroxyapatite slurry, which showed an increase of the rotational correlation time upon adsorption of the complex on the HA surface resulting in a significant relaxivity enhancement. The amidebis(phosphonate) moiety is the only factor responsible for the binding of the complex to HA.
Abstract ID: 084
Abstract ID: 085
Abstract ID: 087
Abstract ID: 088
To detect and quantify arthritic inflammations in an experimental arthritic mouse-model population using a newly-developed high-resolution multi-pinhole (MPH) single-photon emission computed tomography (SPECT) technique.
Eight IL-1Ra deficient mice (Balb/c background, weeks 12–16) with arthritis of the front and back paws and three healthy control Balb/c mice were imaged with the MPH-SPECT system after being scored macroscopically for arthritis. SPECT imaging was performed under short general anaesthesia with a conventional gamma camera upgraded with a pyramidal lead collimator affixed with MPH apertures. Full-body projections were acquired two hours after injection of 10 MBq Tc-99m-MDP/g given intravenously. All images were reconstructed with a modified MLEM-Algorithm and uptake in the paws was determined by regional analysis using the imaging tool AMIDE.
MPH-SPECT technique revealed arthritic lesions in all IL-1Ra deficient mice. The high resolution three-dimensional tomographic images of bone metabolism in units of measured counts/weight and injected activity provided both, the exact localization and the degree of increased bone metabolism. MPH results of the inflamed joints were highly correlated with data from macroscopic scoring and histological joint analysis post mortem. Even in macroscopically normal joints the MPH-technique could detect differences in bone metabolism. Compared to the mice with clinical scores greater than 0 healthy controls did not show arthritic alterations at all.
MPH-SPECT technique represents a new diagnostic tool in the detection of joint pathology for this arthritis mouse model. Compared to macroscopic scoring and other imaging modalities, this new method provides a higher-precision quantitative measurement and a three-dimensional image of joint inflammation, visualizing biochemical inflammatory processes in vivo. The results strongly suggest that the MPH-SPECT technique can be used in longitudinal studies. Furthermore, we have extended this approach to a human study with SPECT and arthritic patients.
Abstract ID: 089
Abstract ID: 090
The iron oxide nanoparticles have been used for MRI negative contrast agent due to its distinctive contrast effect in T2 sequence. For the design of a suitable targeting strategy of USPIO, extend the in vivo half-life time in circulation and to avoid or decrease their being uptaken by macrophage or Kupffer cells is essential. In this report, we have synthesized the aqueous iron oxide nanoparticles with -NH2 functional group, and modified with a capping agent in aqueous phase. The modified USPIO presented excellent biocompatibility in both in vitro and in vivo analysis.
In MR imaging study, we demonstrated the liver contrast images of mice after injected the iron oxide nanoparticles with or without the capping agent by I.V. administration. As shown in the results, we found that the MR imaging and contrast signal of the liver became dark and decrease when the mice were administrated with the USPIO without the capping agent, and the contrast signal decreased with prolonged observation time in 3T Bruker MR imager in ether T2 or T2*-weighted imaging. On the other hand, when the modified USPIO with capping agent was used as the contrast, the MR imaging and contrast signals of liver did not change significantly in the T2*-weighted imaging and T2-weighted imaging sequence.
In this report, we provided a new strategy for the biological stealth of the USPIO as the negative MRI contrast agent that could avoid or decrease being uptaken by the Kupffer cells in the liver and prolong the half-life time in the in vivo study. It also provided future application for diagnostic tracking, drug deliver and targeting of iron oxide nanoparticles.
Poster Session: Novel Probes and Activation Strategies, Part 1
Abstract ID: 091
Abstract ID: 092
Many fluorescent probes are available for measurement of intracellular compounds or membrane potential, but we can not observe trafficking of intercellular signaling molecules, such as nitric oxide (NO) with them. In this study, we visualize the intercellular diffusion of NO by using fluorescent probes anchored to the outside of the cell membrane. NO is an important signaling molecule involved in the regulation of a wide range of physiological and pathophysiological mechanisms, and is thought to act on neighboring cells after being produced by intracellular NO synthase and permeating through the cell membrane.
To anchor the probe in the membrane, one approach would be to introduce a hydrophobic anchor into the probe, while another would be covalent binding of the probe to a bio-molecule on the cell surface. We employed the former approach, and combined dipalmitoylphosphatidylethanolamine (DPPE) with negatively charged fluorescein or DAF, an NO probe that we previously developed. The fluorophore moiety of the probes on the cell membrane is expected to be mostly located outside of the cell due to the negative charge, and to react with NO there to yield a highly fluorescent form. We examined the fluorescence of cells loaded with the synthesized probes under an inverted microscope and a confocal laser microscope. The observation of ring-shaped fluorescence images confirmed that the probes were localized on the cell membrane. Moreover, we successfully observed the transmission of NO from a cell to its neighboring cells with DAF-DPPE.
Abstract ID: 093
Much progress has been made in the development of small molecule imaging agents in recent years, however there remain some drawbacks. One problem is the inability to amplify a signal from an imaging agent. We are overcoming this problem by using enzymes to amplify an optical signal from a small molecule. Small molecules that mimic an enzyme's natural substrate were designed to be acted upon by a promiscuous enzyme, causing a change in the electronic configuration of the agent. This resulted in changes in their fluorescent properties. Thus, in this manner, enzyme activity would lead to increased optical signal. Using these guidelines, a sensitive, selective fluorogenic assay to monitor medium-chain acyl-CoA dehydrogenase (MCAD) activity has been developed. MCAD is the first enzyme in the β-oxidation pathway and carries out the α, β-dehydrogenation of fatty acids. We replaced the saturated fatty acid carbon chain with an aromatic system, which resulted in a substrate in which dehydrogenation extends the π-conjugation. Synthesis and screening of a small library yielded three fluorogenic substrates. MCAD deficiency has an occurrence rate of 1:15,000 newborns in the US. This fluorogenic assay could be an efficient method for MCAD deficiency neonatal screening. Further, as MCAD is involved in fatty acid metabolism, the fluorescent assay could be used to diagnose chronic diseases, such as diabetes and metabolic syndrome.
Abstract ID: 094
Novel fluorescent photoprobes for the imaging of endothelin (ET) receptors were developed. Based on the nonpeptide, highly affine and selective ETA receptor antagonist (3-Benzo[1,3]dioxol-5-yl-5-hydroxy-5-(4-methoxyphenyl)-4-(3, 4, 5-trimethoxybenzyl)-5H-furan-2-one (PD 156707), a modification of the lead structure with a PEG-spacer containing an amino moiety was performed. Labelling of this precursor with fluorescent markers like FITC and Cy 5.5 was accomplished by modification of the known peptide-labelling procedure. The Cy 5.5-labelled receptor antagonist revealed high binding affinity to ET-positive human breast carcinoma cell lines. Specific binding could be blocked using the parent antagonist PD 156707 as a competing inhibitor. The results suggest that the modified fluorophore tightly binds to ETA receptors and thus may be a potent candidate for the in vivo imaging of ET-overexpressing tissues.
Abstract ID: 095
The synthesis of a new class of stable microporous materials has been reported, consisting of mixed tetrahedral-octahedral framework silicates containing stoichiometric amounts of Ln3+ ions in the framework rather than in the pores, fundamentally different from conventional zeolites. These zeo-type silicates (AV-9 materials) have composition (Na4K2)(Ln2Si16O38)·10H2O (Ln = Sm, Eu, Nd, Tb, Gd, Dy) [1].
29Si and 23Na MAS NMR spectra of SmAV-9 indicate the presence of 6 and 2 distinct sites, respectively, in agreement with the solid state structure proposed for this material [1]. The magnetic susceptibility of Gd-AV9 exhibits a simple paramagnetic behavior (Curie law) with an effective magnetic moment (μeff) of 7,97 μB. For Eu-AV9, the magnetic susceptibility is much lower, with a very shallow temperature dependence (Van Vleck paramagnet behavior) [2]. The water 1H longitudinal (r1) and transverse (r2) relaxivities were obtained for aqueous suspensions of several Ln-AV-9 materials with different Ln3+ ions. r1 values are very small (0.04–0.09 s−1.mM−1) and almost independent of the Ln3+.r2 values are much larger (16–60 s−1.mM−1) and proportional to μeff2. While r1 of Gd-AV-9 suspensions is almost independent of the magnetic field (Bo), r2 increases proportionally to Bo2. The μeff2 and Bo2 dependence of r2 was rationalized using the chemical exchange model considering outer-sphere diffusion of water around the particles in the long-echo limit [3].
We thank the support from F.C.T., Portugal (project POCTI/QUI/47005/ 2002), EU COST Action D18 “Lanthanide chemistry for diagnosis and therapy” and the European-funded EMIL programme (LSHC-2004–503569).
Abstract ID: 097
We developed a novel type of photosensitizer with OFF/ON switch of singlet oxygen (1O2) generation by recognizing molecular targets. Our strategy focused on the existence of various environments from polar environment to non-polar environment in biological sample, which could be detected by environmental sensitive fluorescence probes. We designed and synthesized a series of environmental sensitive photosensitizers (ESPers), which generate 1O2 by recognizing the target hydrophobic environment (i.e., lipid environments or hydrophobic pocket around ligand binding site) in order to reduce non-specific damage in the cytosol (i.e., polar environment). Our ESPers consist of photosensitizing chromophore and an electron donor moiety, in which the property of photosensitization could be controlled by quenching of singlet-excited state by photoinduced electron transfer (PeT). The efficiency of 1O2 generation of ESPers changed according to the HOMO energy level of electron donor moiety and solvent effect of PeT. Then, we synthesized ESPerconjugated ligand (ESPer-IP3) for inositol 1, 4, 5 trisphosphate receptor, which has been suggested to have hydrophobic pocket around ligand binding site. ESPer-IP3 in cytosolic environment had no photosensitization effects in spite of light illumination and restored the ability of 1O2 generation by binding to the receptor, resulting in receptor photoinactivation. Our ESPer strategy provides a powerful tool to control photosensitization in biological system.
Abstract ID: 098
Transglutaminases (tTG) form a family of enzymes that have evolved for covalent cross-linking of proteins in stabilization of atherosclerotic plaques, and generation of clots, providing mechanical strength during tissue remodeling, and in association with angiogenesis at the front of invading tumors. We recently reported the development of a novel peptide based contrast material (b-TGS-GdDTPA) as a substrate for tTG for specific MRI mapping of tTG activity (1). DTPA-dianhydride was conjugated to the b-TGS peptide via one of the two lysine residues on the carboxy end of the peptide. The conjugation occurs through the lysine side chain amine Ne, which after conjugation becomes an amide proton. We hypothesized that the binding site of GdDTPA could explain the high relaxivity of b-TGS-GdDTPA (14.3 mM−1s−1 compared to 4.2 mM−1s−1 of GdDTPA). We used 2D 1H-NMR spectroscopy (mainly TOCSY and NOESY), combined with mass-spectrometry (MS/MS). Sequential assignment of the peptide signals was based on 1H-1H NOE connectivities between spin systems of sequentially neighboring residues. According to the NMR data the b-TGS peptide does not adopt a well-defined conformation, assuming a random coil structure, as expected for small peptides. There was no evidence of random mix of DTPA conjugated to both of the lysines, since we could not detected 2 sets of spin systems, which would be mostly notified in the lysine peaks. 2D NMR methods combined with mass-spectrometry revealed that the DTPA was conjugated only to the first lysine in the peptide sequence. We hypothesize that relaxivity may be enhanced though exchange of NH protons on both lysine side chains.
Abstract ID: 099
Gd(III) chelates are commonly used as contrast agents in magnetic resonance imaging (MRT) for clinical diagnostics, due to their paramagnetic properties. One approach to the optimization of their in vivo efficacy is to try to increase their relaxivity through formation of supramolecular structures by inclusion of Gd(III) chelates in cyclodextrins. This approach has so far been limited by the weak host-guest interactions observed in aqueous solution.[1,2] We report here the effect of supramolecular interactions on the relaxivity of host-guest complexes between highly charged Gd(III)-tetraazamacrocyclic ligands and symmetrical per-functionalized cyclodextrins bearing a high charge of the opposite sign. The systems chosen are: a) [Ln(DOTAM)3+] and negatively charged β-cyclodextrin per-functionalized at the C(6) positions with SCH2COO− groups (per-CO2−-β-CD); b) Ln(DOTP)5− and Ln(DOTA)− and positively charged cyclodextrins per-functionalized at the C(6) positions with amino groups (per-NH3+-β-CD and per-NH3+-γ-CD). The frequency dependence of the water relaxivities (NMRD curves) for Ln = Gd was measured as a function of pH and temperature. The stability and stoichiometry of the association process of the suparmolecular entities was studied. High resolution 1H NMR (Ln = Nd, Eu, Tm, Gd) gave paramagnetic dipolar contributions to the proton shifts and relaxation rates of the cyclodextrin protons, which allowed the definition of the geometries of the supramolecular systems. The supramolecular association in solution was also studied using Mass Spectrometry.
We thank the support from F.C.T., Portugal (project POCTI/QUI/47005/2002), EU COST Action D18 “Lanthanide chemistry for diagnosis and therapy” and the European-funded EMIL programme (LSHC-2004-503569).
Abstract ID: 100
DOTASA is a DOTA-like macrocyclic ligand showing a carboxymethyl -CH2COOH substituent moiety at a Cα carbon of one of the four acetate pendant arms, present as a racemic mixture of R and S configurations.[1] The protonation constants of the ligand determined by potentiometry were close to DOTA except for the extra pK3 value of 5.35 assigned to protonation of the extra carboxylate group in the succinyl arm. The thermodynamic stability constant of its Gd3+ chelate was determined by a Potentiometrie method, and the value obtained, log KML = 27.2 (0.2), is higher than for the [Gd(DOTA)(H2O)]− complex. The solution structure of the asymmetrical Ln3+ chelates of DOTASA was studied by 1H NMR spectroscopy, indicating the presence of four isomers, corresponding to the combination of the antiprismatic (M) and twisted antiprismatic (m) helicities of the pendant arms and to the R and S configurations of the substituted pendant arm Cα atom. The m/M isomer ratio decreases along the lanthanide series, with the m isomer decreasing from 90% at La to about 50% from Eu-Lu.[2] This shows that the expected m isomer population of the Gd3+ complex with DOTASA is higher than for the unsubstituted Gd3+-DOTA (~15%) but lower than for a Gd3+ chelate of a RRRR tetrasubstituted DOTA (~70%). Thus the stabilisation of the m isomer by Cα mono-substitution at the DOTA acetate pendant arms in [Gd(DOTASA) (H2O)]2− is responsible for its increased water exchange rate and higher relaxivity.[1]
We thank the support from F.C.T., Portugal (project POCTI/QUI/47005/2002), EU COST Action D18 “Lanthanide chemistry for diagnosis and therapy” and the European-funded EMIL programme (LSHC-2004-503569).
Abstract ID: 101
Magnetic Resonance is entering the molecular imaging phase of its evolution. The major drawback for MR imaging is the inherent insensitivity of the NMR phenomenon, i.e. MR imaging offers incredible resolution, but poor sensitivity. To overcome this problem we are taking an accumulation and activation strategy. The signal from the contrast agent is amplified by some biological event, i.e. an enzyme-activated contrast agent is employed. The acetoxymethyl esters of the pendant carboxylates result in a charge neutral complex capable of crossing cell membranes. In serum the complex will be q = 0, i.e. the inner-sphere water molecules will be displaced by HCO3. Once internalised, the acetoxymethyl esters are susceptible to hydrolysis by intracellular esterase enzymes. On esterase activity the carboxylate moieties will be unmasked yielding a 3- charged complex. This will slow the outflow from the cell and result in a large increase in contrast within the cell due to accumulation and activation of the agent, switching from a complex with a hydration state of q = 0 to q = 2, as binding HCO3 is now disfavoured due to the increased negative charge on the complex. Initial results concerning the luminescence (Eu) and proton relaxivity (Gd) of the contrast agent will be presented. These complexes demonstrate reversible binding of hydrogencarbonate and a ~ 300% increase in proton relaxivity of the complex when moving from pre- to post-enzyme activity.
Abstract ID: 102
Magnetic Resonance Imaging (MRI) provides the acquisition of highly detailed and three-dimensional images of the human body. The use of contrast agents (CA's) can greatly improve the quality of a MRI scan. Since the clinical introduction of the first MRI contrast agent (Gd-DTPA), efforts have been made to synthesize new products with improved relaxivity and specificity.
The theory of contrast agents demonstrates that numerous parameters affect relaxivity (efficiency) with high relaxivities being obtained for complexes with long molecular reorientational time (tR) and fast exchange (kex) of the water molecule(s) coordinated to the paramagnetic metal ion.
Our goal was the synthesis of a new ligand EPTPA-C16 (Figure), whose Gd(III) chelate could simultaneously optimize both parameters: (i) water exchange rate (kex) due to steric compression around the water binding site in the coordination sphere of the metal ion and (ii) molecular reorientational time(tR) upon formation of micelles in solution.1
The synthetic procedure as well the characterization of the Gd(III)-chelate concerning its potential application as a MRI contrast agent will be presented. These include the determination on the critical micelle concentration (CMC) of the [Gd(EPTPA-C16)(H2O)]2− chelate by protonic relaxivity measurements, and a global analysis of the data obtained from variable-temperature and -field 17O NMR measurements, and 1H NMRD variable-temperature measurements, which allowed the determination of the relaxivity and its parameters. The rotational dynamics of the micelles was analysed using the Lipari-Szabo approach. This is the first time that the effect of the conjugation of both parameters tR and kex on the relaxivity is studied for paramagnetic micelles.
We thank the support from F.C.T., Portugal (project POCTI/QUI/47005/2002), EU COST Action D18 “Lanthanide chemistry for diagnosis and therapy” and the European-funded EMIL programme (LSHC-2004-503569).
Abstract ID: 103
Chemical Exchange Saturation Transfer (CEST) is a novel MRI technique which relies on altering the water signal intensity by presaturating a contrast agent proton pool and subsequent proton exchange. CEST has two properties attractive for molecular imaging: the contrast enhancement is actively controlled by the saturation, and since the proton exchange and thus the extent of water signal decrease is often pH dependent CEST can be used for pH mapping.
Both Sherry and Aime with coworkers have shown that paramagnetic complexes with exhangeable amide protons are very suitable for CEST. However, for a significant decrease of the water signal, still CEST agent concentrations in the order of mM are needed. One approach to improve the efficiency of a CEST agent is to attach multiple paramagnetic complexes to a carrier. We have covalently attached DOTA-tetraamide (DOTAM) ligands to poly(propylene imine) (PPI) dendrimers, and subsequently carried out a complexation with Yb3+. The CEST effect, i.e. the water signal decrease, was studied on a Philips Achieva 3T whole body MR scanner. The figure below shows an image of a phantom containing vials with the dendritic CEST agent carrying multiple Yb-DOTAM complexes in buffer at different pH, compared with vials with a single Yb-DOTAM complex (with equal Yb3+ concentration). The maximum CEST effect is 54% for both materials, but it has shifted to lower pH for the dendritic agent relative to the single complex. This is because the basic core of the dendrimer accelerates the base-catalysed proton exchange. We will show that Yb-DOTAM-dendrimer is a suitable CEST agent for pH mapping using a novel ratiometric method for concentration dependence elimination without the need for a second exchangeable proton pool.
Abstract ID: 104
The ability of MRI to visualize events between biology and the clinic, and to simultaneously and non-invasively image tissue merits its use as a principle clinical and molecular imaging modality of the future. MRI contrast agents are potential tools for studying cellular events. Our previous work1 has shown that the integration of gadolinium lipids into liposomal membranes produces paramagnetically labelled liposomes that significantly increase proton relaxivity and lead to brightly contrasted MR images, allowing for the use of liposomes as labelling vehicles in molecular imaging.
The synthesis and liposome formulation of a novel gadolinium lipid designed for cellular uptake is described. Liposome formulation and liposome dose were optimised in order to allow for maximum cellular entry of the gadolinium probe, resulting in a largely enhanced MR signal. The optimised formulation showed low cytotoxicity and, furthermore, was able to mediate pDNA transfection at levels comparable with commercially available transfection agents such as Trojene™ (Avanti Polar Lipids, USA). The addition of fluorescently labeled components to the liposome formulation results in a competent dual-imaging system, results of which will be presented.
Abstract ID: 105
Gold nanoparticles (AuNP) may be suitable imaging probes for x-ray based imaging modalities. This presentation describes ongoing efforts within our institution to develop AuNP as targeted imaging agents. A series of experiments were performed wherein AuNP are injected into weaner pigs (< 20 kg). These experiments were chosen to determine the biodistribution of nanoparticles and to visualize contrast enhancement in these regions through sequential computed tomography scans. A swine model was chosen due to availability and similarity to human physiology. The pigs were placed under anesthesia (isoflorane) and a catheter was placed in the right femoral artery.
Computed Tomography scans were acquired prior to the injection of AuNP. The animal was positioned and anesthetized approximately two hours prior to the scan; physiologic changes are thus stabilized. Approximately 30 mg of AuNP were injected within 30 seconds. Subsequent scans were acquired at various times post injection. Images were analyzed using a standard display program using manually drawn regions of interest at the same anatomic level for each image acquisition time. In the example are shown CT images of the lung at two different times. AuNP uptake was greater in the lungs than in other tissues, causing an increase in the CT number (higher attenuation) of approximately 75 Hounsfield units at 30 minutes post injection. This result is consistent with biodistribution data of AuNP for this animal model.
Abstract ID: 106
Noninvasive imaging in the near-infrared spectral range offers the potential for highly sensitive tumor detection and tracking. Targeting a tumor-enriched cell surface receptor with a ligand-conjugated fluorescent probe specifically allows detection the tumor relative to the negligible animal autofluorescence. In this study, we targeted EGF receptor, which is ~100-fold overexpressed in most solid tumors, using EGF conjugated to IRDye™800CW (maximal emission at 800nm). To confirm specific binding and bioactivity of EGF-IRDye™800CW, two human prostate cancer cell lines, PC3M-LN4 and 22Rv1, were evaluated in vitro by In-Cell Western. Cells stimulated with 150 ng/mL labeled EGF (dye to protein ratio=1.5) or unlabeled EGF showed equivalent efficacy in triggering downstream phosphorylation of ERK, indicating the label had no significant effect on biological activity of the probe. Specificity of labeled EGF for the EGF receptor was further demonstrated by competitions with unlabeled EGF or C225, a monoclonal antibody that blocks the EGF receptor, which effectively blocked binding at 1.25 or 9.4 μg/mL, respectively.
We then examined the imaging efficacy of EGF-IRDye™800CW in a longitudinal study of subcutaneous tumor growth in SCID mice injected with PC3M-LN4 (n=4) or 22Rv1 (n=4) cells. The EGF-IRDye™800CW biomarker (1 nmol) was delivered by tail vein injection. Using a prototype 2D imaging system optimized for IRDye™800CW detection, we first determined systemic dye clearance kinetics. Calculations of tumor signal-to-noise ratio indicated the greatest sensitivity at 96 hours post dye injection. Tumors of <1mm in diameter were detected after 1 week. Tumor growth was tracked following weekly injections of EGF-IRDye™800CW for 5 weeks. Tumor signal-to-noise ratios were found to be representative of size as measured by wet weight at the endpoint analysis. This characterization of EGF-IRDye™800CW as a biomarker is the first study of tumor growth kinetics tracked by near-infrared imaging.
Abstract ID: 107
Myeloperoxidase (MPO) is a prognostic indicator of inflammation, an enzyme present in unstable plaque that contributes to plaque rupture causing acute cardiovascular events. It has previously been shown that MPO is secreted by neutrophils and macrophages; MPO activates matrix metalloproteinases, contributes to foam cell formation and reverses protective function of HDL. Herein, we describe the properties of a novel Gd-based molecular probe di-serotonin-DTPA(Gd) bisamide [1] which due to its specific functionalization is able to efficiently reduce oxidized forms of peroxidases [2], including MPO [3].
The MPO-specific MR signal increase is based on the ability of MPO to generate 5-hydroxytriptamide (serotonin)-derived radicals which can have various fates in vivo: 1) these radicals can oligomerize with the resultant increase in relaxivity (r1 and r2) [2,3]; 2) these radicals can undergo cross-linking with other macromolecules present in the inflammatory lesion. We hypothesized that the combination of these effects could enable to specifically image MPO activity. Preliminary MRI results obtained in mouse models of inflammatory disease have shown an increase of SNR (3:1) in MPO-containing areas, in contrast to nonspecific paramagnetic chelates or the case when no MPO was present. Therefore, by co-localizing with MPO, the probe yields improvement not only in local signal contrast but also in signal persistence in vivo.
Abstract ID: 108
Citrate is not only the key intermediate of the tricarboxylic acid cycle but it is also a key component of some metabolisms. Because of its metabolic significance, citrate levels have been linked to the characteristics of several diseases. There are several commonly used techniques for the determination of citrate, e.g., chromatography, and are most frequently applied to the analysis of beverages, food or pharmaceutical formulations. Analyses based on absorption or emission spectroscopy offer enhanced sensitivity. However, a chemoselective citrate sensor, that is not compromised by pH fluctuations and affords a ratiometric method of analysis (to preclude additional calibration protocols), still needs to be developed for in vitro or in vivo assays. Herein we report a europium complex-based citrate sensor. Titration of citrate to the aqueous solution of the complex resulted in very significant changes in its luminescence spectrum at very low added citrate concentrations. The ratio of the intensity of the ΔJ = 2 bands to that of the ΔJ = 0 transition chemoselectively respond to the concentration variation of citrate, but not to some other common anions in biological circumstance, such as bicarbonate, chloride, phosphate, AMP, ADP, ATP, and most of amino acid. It is also free of interference from pH variation. Initial studies indicate that NIH 3T3 cells have a good permeability to this complex, indicating its promising application.
Abstract ID: 109
The purpose was to synthesize and characterize very small superparamagnetic iron oxide particles (VSOP) based on (in)organic silica coating as MR imaging contrast agent.
A two-step method was used for the synthesis of (in)organic silica-coated VSOP. First the magnetic core was synthesized by coprecipitation of ferric and ferrous salts in sodium hydroxide solution. Then the particles were coated by the organic (3-aminopropyltrimethoxysilane (APS) and N-(2-aminoethyl)-3-aminopropylmethoxysilane) (AEAPS) and inorganic silica through a dehydration process at high temperature. Furthermore, it was tested whether IgG-antibodies can be bound to the new type of particles. The time-dependent (20min, 3h, 6h and 24h) and dose-dependent (0.03, 0.3 and 3μmol/L iron concentrations) uptake of modified and IgG-coupled VSOP by a rat stem cell line and its effect on cell viability were evaluated by MRI, histology and transmission electron microscopy (TEM).
The size of the resulting particles is about 10nm without a significant increase in diameter after the coating step. X-Ray diffraction patterns of uncoated, silica coated, APS coated and AEAPS coated VSOP show peaks corresponding to a inverse spinel structure, which is consistent with the crystal characteristics of magnetite. T2 relaxation rates of AEAPS-, APS-, Silica-coated and IgG-coupled VSOP (AEAPS) were 84.8±0.1, 134.4±0.1, 339.8±0.2 and 85.2±0.1s−1mM−1 and T1 relaxation rates were 1.05±0.1, 1.17±0.1, 0.55±0.1 and 0.20±0.1s−1mM−1, respectively. TEM located the VSOPs in lysosomes and pinocytotic vesicles inside the cells. Although there was an intense intracellular uptake of modified and IgG-coupled VSOPs no inverse effects on the cell viability were observed, even at high iron concentration (3μmol/L) and long incubation periods (24h).
In summary a promising type of VSOPs is presented, which shows excellent magnetic properties and biocompatibility. Antibodies can easily be bound to the particles, which make them suited as matrix of specific molecular imaging probes.
Abstract ID: 110
Recent modifications to strategies for developing activatable optical probes involve use of smaller, peptide-based scaffolds and single quencher-fluorophore pairs. Recently, we reported the synthesis and initial characterization of a cell permeable, quenched imaging probe, TcapQ647, comprised of a Tat-peptide based permeation peptide sequence, an effector caspase recognition sequence, DEVD, the flanking far-red quencher, QSY 21, and a paired fluorophore, Alexa Fluor 647. Here we present further biochemical and cellular analysis demonstrating the utility of TcapQ647 to image apoptosis via intracellular effector caspase proteolysis.
The selectivity of TcapQ647 cleavage by effector caspases was determined by recombinant enzyme assays utilizing caspases 1 to 10. TcapQ647 cleavage rates were highest for the effector caspases 3, 6 and 7 with the calculated rates over background being 0.4 ΔFU*min−1 U−1, 10 ΔFU*min−1U−1, and 23 ΔFU* min−1U−1, respectively. Assays with initiator and inflammatory caspases resulted in slow cleavage rates, with most caspases having no proteolytic activity toward TcapQ647. Using IC50 values determined from inhibition assays with DEVD-CHO, Km values for caspases 3, 6, and 7 were calculated from enzyme kinetic equations to be 890 nM, 200 nM, and 83 nM, respectively.
Two epithelial cell lines, KB 3–1 and HeLa, and a breast carcinoma cell line, MCF-7, were utilized to determine intracellular activation of TcapQ647 during apoptosis. Following induction of apoptosis with doxorubicin, TcapQ647 was added to each cell line and activation was monitored by fluorescence confocal microscopy. A commercial, general caspase substrate, D2 R, was co-incubated with TcapQ647 as an independent marker of apoptosis. Significant TcapQ647 fluorescence activation in all doxorubicin-treated cell lines was observed with minimal activation observed in cells incubated without drug. Furthermore, incubation of doxorubicin-treated cells with the non-cleavable, all D-amino acid dTcapQ647 resulted in detection of minimal fluorescence activation. Thus, TcapQ647 is a promising probe for monitoring effector caspases in cellulo and potentially in vivo.
Abstract ID: 111
Furthermore the different affinities of the two functional groups to Ca2+ permit to obtain free coordination sites at gadolinium. The generation of these coordination sites, which are mandatory for water relaxivity, depends on the structure of the complexes and the Ca2+ concentration. Gadolinium complexes
This work was supported by the Louis-Jeantet Foundation, and the Hertie Foundation.
Gadolinium complexes of the bis-macrocyclic compounds with variable length of side chain.
Relaxivity studies of the complexes in the presense and absence of Ca2+.
Abstract ID: 112
‘Smart’ contrast agents (CA) exhibit dynamic and reversible modulation of their relaxivity by specific physiological or biochemical triggers such as changes in pH, Ca2+ concentration or enzymatic activity (1–3). The extracellular concentration of Ca2+ plays important role in physiological and pathological processes in the nervous system. This led to the designing of a chelating system in which relaxivity is influenced as a function of Ca2+ concentration by changing coordination number around the paramagnetic metal ion. We synthesized a novel bifunctional bismacrocycle [Gd-(DO3A-DTPA-DO3A); Fig.] based on DO3A-EA [{4,7-Bis-carboxymethyl-10-(2-aminoethyl)-1, 4, 7, 10-tetraaza-cyclododec-1-yl}-acetic acid] coupled to DTPA-bis-anhydride via a flexible alkyl spacer to form amide linkages. The overall yield of the four step synthesis starting from cyclen was 54%. This gadolinium-based agent has two limiting conformational states with different Ca2+ concentrations. It is hypothesized that in the absence of Ca2+, the carboxylates of the DTPA ligand interact with the Gd3+ ions which were held in DO3A, but in the presence of Ca2+, these carboxylates rearrange to chelate Ca2+ thereby allowing water to bind directly to Gd3+.
(Work was supported by the Hertie Foundation.)
Abstract ID: 113
Chemical Exchange Saturation Transfer (CEST) agents are a new type of non-metallic MR contrast agents in which contrast depends on the chemical exchange rate of the agent protons with water [1–5]. Recently, Sherry, Aime and coworkers produced a series of small paramagnetic agents with custom tailored exchange rates to produce maximal contrast[4–6]. Polypeptides can produce large contrast[2] in the micromolar range, and in principle could also be tailored to maximize this contrast. Previous studies of protein folding have produced equations which predict the exchange rates in proteins [7–9]. In this study, we test these equations predicting exchange rates for use in the design of optimal CEST contrast agents and verify the calculations using a set of polypeptides. The HX (X= O or N) exchange rates were predicted based on the neighbors to the left and right using the combined base-catalyzed and acid-catalyzed equation produced by Englander and coworkers [7]. The MT contrast increases exponentially with chemical exchange rates as discussed previously[3], provided a sufficient saturation field such that the exchangeable group is fully saturated is assumed. Using these equations we are able to predict the proper amount of contrast-enhancement. These principles should prove useful for future design and use of CEST agents.
Poster Session: Molecular and Functional Imaging in Cancer, Part 1
Abstract ID: 114
Clinical 18F-FDG/PET studies encountered difficulties in detecting early stage lung cancers due to accumulation of 18F-FDG in inflammatory lesions. This study evaluated the ability of O-2-18F-fluoroethyl-L-tyrosine (18F-FET), a radiolabeled amino acid, and 2-Deoxy-2-18Ffluoro-D-glucose (18F-FDG) to differentiate the inflammation and lung carcinoma in a mouse model. Fifty-eight C57BL/6 mice were inoculated with 2 × 106 LLC1 lung carcinoma cells at right hind flank at day 0 and then injected with 100 μl turpentine at left hind flank at day 3. The progress of inflammation and growth of tumor was longitudinally monitored by 18F-FDG/microPET. Biodistribution, pharmacokinetics, microPET image and whole body autoradiography (WBAR) of 18F-FET and 18F-FDG were performed in the mouse model. Both the tumor uptake and the inflammation uptake are 4.1-fold and 3.2-fold significantly higher than that in muscle at 90 min after injection of 18F-FDG at day 8 after tumor inoculation. The tumor uptake of 18F-FET at 90 min after injection was 2.0-fold higher than that in muscle. However, no significantly uptake of 18F-FET in inflammation lesions compared with that in muscle (Inflammation-to-muscle equals to 0.9). Results of biodistribution and WBAR corroborated with quantitative microPET images confirmed that 18F-FET was revealed only in lung carcinoma. The distribution half-life (T1/2α) and elimination half-life (T1/2β) of 18F-FET was 39.23 min and 204.96 min, respectively in the mouse model. In Conclusion, 18F-FET utilized only in lung tumor, whereas 18F-FDG delineated both tumor and inflammation. The 18F-FET owned the potential to distinguish the inflammation from lung carcinoma in vivo.
Abstract ID: 115
CT Perfusion has proven to be a powerful clinical tool in detecting and diagnosing cancer in patients by providing quantitative assessment of blood kinematics in tumour angiogenesis. Due to recent advances in molecular biology, there has been an increased interest in the development and imaging of numerous mouse tumour models to study cancer. Current micro-CT systems are capable of spatial resolution as high as 10μm. However, in order to perform dynamic perfusion imaging in mice, quantitative CT image data must be acquired at a rate of at least 1 Hz. Through our collaboration with GE Healthcare, our lab has obtained a novel research prototype pre-clinical CT scanner, the eXplore Locus Ultra, designed specifically for dynamic perfusion imaging in small animals. Using a digital flat-panel detector and a clinical slip-ring gantry, this system is capable of acquiring high-resolution volumetric image data at a rate of 1 Hz, while covering the entire thoracic region of a mouse. The performance of the scanner was demonstrated through the in vivo imaging of 4 C3T mice with KHT tumours. A bolus of 1.5 μl/g of Omnipaque 300 was injected into the tail vein at the beginning of each scan. The arterial input function was obtained from carotid arteries, aorta or pulmonary arteries and tissue time-density curves were deconvolved to generate functional maps of blood flow, blood volume and vascular mean transit time. The values for all three functional maps are increased in the region of the tumour when compared to healthy tissue, thus indicating the presence of increased vasculature in the tumour. The blood flow rate increased from 18.5±2.3 ml/min/100g in the healthy tissue to 32.4±4.1 ml/min/100g in the tumour region, showing a significant difference (p<0.5). Micro-CT has multiple applications in cancer research by providing high-resolution anatomical information and functional information through dynamic perfusion imaging.
Abstract ID: 116
Lymphotrophic nano-particle enhanced MRI (LNMRI) has shown to be an accurate lymph node staging technique in patients with various primary cancers. Currently the accuracy of LNMRI is largely qualitative and is dependant on the experience of the radiologist evaluating the study. Quantitative estimation has been hampered due to the limitations of the MRI sequences as the spatial resolution required for capturing the nodes is limited. Even when the nodes are adequately covered the heterogeneity of the nodes is a stumbling block for quantification. As LNMRI becomes widely available for nodal staging, it has become necessary to overcome the limitations by developing technical algorithms that not only allow fast throughput reading but also partially render independence from the currently required reader experience. The goal of the current study is to optimize the sequences so as to maximize the spatial resolution and to conduct a multi-parametric study to optimize the quantitative detection aspect, enabling semi automated detection and characterization of lymph nodes enhanced with nano-particles. It has been shown [1] that with a dual echo T2* weighted sequence it is possible to perceive malignancy from the computed T2* values, however in the case of heterogenous nodes this can have its limitations due to the approximation of the exponential decay. We have extended the approach to multiple echoes so as to improve the estimation and detection of the nodes. Also by careful selection of the echo placement we propose to improve the predictability of the quantification. We also propose to use a multi-parametric approach to constrain the estimation. We hypothesize that this technique will reduce the time required by radiologists to analyze LNMRI and will also aid the inexperienced radiologist in analysis.
Abstract ID: 117
Affibody® molecules, small non-immunoglobulin affinity ligands, can be selected to virtually any target protein from combinatorial libraries based on a 58 amino acid protein A domain scaffold. Their small size and stability, high affinity and specificity, rapid biodistribution kinetics and clearance make them ideal for in vivo targeting and detection of tumors using SPECT or PET imaging.
Affibody molecules can be made either by recombinant techniques or by conventional peptide synthesis. Synthetic Affibody molecules fold spontaneously and bind their targets both in vitro and in vivo. Thus, Affibody molecules, site-specifically modified with metal chelators like DOTA can be made in a single chemical synthesis step.
Data on the use of the HER2-specific Affibody molecules (ZHER2) for targeting of HER2-positive tumor xenografts will be presented. ZHER2 was labeled with 125I, 111In or 68Ga and analyzed by binding assays, biodistribution and imaging (Figure 1). Rapid and specific delivery of labeled tracers to target tumors was obtained within one to four hours post injection. Specificity of tumor targeting was proven by competition with unlabeled Affibody molecules or modulation of HER2 expression using the drug 17-AAG as shown by molecular imaging. Tumor to blood ratios in the range of at least 7 and up to 200 were obtained within one to twelve hours post injection. Due to the very low background, tumors could be visualized by SPECT imaging as early as one hour post injection.
The radiolabeled DOTA-ZHER2 Affibody molecule for imaging allows for novel theranostic disease management of HER2 positive tumors.
Abstract ID: 118
Abstract ID: 119
Radiolabeled peptides serving as substrates for caspase-3 were tested to image apoptosis. One requirement of the peptides serving as caspase-3 substrates is the specific recognition sequence DEVD. A series of peptides containing this sequence was synthesized and radiolabeled with 131I and 99mTc. Cell uptake studies were performed using normal controls and apoptotic Jurkat cells. The substrate specificity of two peptides, YDEVDG-NH2 and Tat49–57-yDEVDG-NH2, was investigated with recombinant caspase-3 and with lysate of apoptotic cells. In vitro time course experiments of the radioiodinated peptides DEVDGY-NH2, YDEVDG-NH2, NQVNGY-NH2 and YNQVNG-NH2 revealed no uptake preference in apoptotic cells. To achieve an increase in cellular uptake these peptides were combined with the transporter peptide Tat49–57 and its reversed sequence Tat57–49. Two of six peptides investigated here, Tat49–57-yDEVDG-NH2 and Tat57–49-yDEVDG-NH2, showed preferential uptake in apoptotic cells (Fig. 1). Compared with the former peptide the 99mTc complexes obtained with Tat49–57-C(Acm)GC(Acm)DEVDG-NH2 demonstrated also higher uptake in apoptotic cells (Fig. 2). Differences in cell uptake of the two isomeric forms resulting from this reaction illustrate the sensitive sterical requirements for being transportet and/or accepted by the enzymes. Proof of caspase-3 substrate specificity was received with Tat49–57-yDEVDG-NH2 by a competitive caspase-3 assays and MALDI-MS. Within the series DEVDG derivatives the radiolabeled peptides obtained with Tat49–57-yDEVDG-NH2, Tat57–49-yDEVDG-NH2 and Tat49–57-C(Acm)GC(Acm)DEVDG-NH2 were found to exhibit significant uptake differences in favour of apoptotic cells. The enhanced retention is interpreted with the interaction of the labelled peptide or peptide fragment with activated caspases. Support from the European Union (FP6 Program, EMIL) is gratefully acknowledged.
Abstract ID: 120
Human estrogen receptor-positive breast cancer cells typically display elevated levels of CCND1 mRNA and insulin-like growth factor 1 (IGF1) receptors. We designed and synthesized a chelator-antisense peptide nucleic acid (PNA)-peptide chimera complementary to CCND1 mRNA (WT4348), with a cyclic peptide analog of IGF1 at the C-terminus for intracellular delivery via IGF1 receptors, and a DOTA chelator at the N-terminus to bind 64Cu to enable microPET imaging. 8-Amino-3,6-dioxaoctanoate spacers were inserted at each end of the PNA to minimize steric hindrance of mRNA hybridization to the antisense PNA by DOTA and the cyclic peptide. The entire WT4348 probe was assembled in a single run by solid phase Fmoc coupling on a peptide synthesizer. The two cysteines of the IGF1 analog were oxidized/cyclized to a disulfide bond by iodine while the chimera was still attached to the solid phase support. Two control chimeras, one with four mismatched PNA bases and another with two mismatched amino acid residues, were prepared by the same method. Biodistribution of the three 64Cu-labeled chimeras administered via the tail vein into NIH NCR immunocompromised female mice showed high uptake and long retention in the liver and kidneys, and efficient clearance from muscle and blood. MCF7 tumor xenografts (200–700 mg) in NCR mice were imaged with the three 64Cu-labeled PNA chimeras by CT and microPET at 4, 8 and 24 hr after administration. 64Cu-WT4348 showed the most intense tumor images at all time points studied, with a tumor/contralateral tissue intensity ratio of 2.89 at 4 h post-injection. Mismatch sequences yielded weak images. These studies establish the proof-of-principle that antisense PNAs can be utilized as PET imaging probes for in vivo detection of cancer gene mRNA expression, dependent on delivery of antisense radiodiagnostics across cell membranes by a receptor-specific targeting peptide. Supported by DOE ER63055.
Abstract ID: 121
Experimental and clinical studies have shown the feasibility of 99mTc-sestamibi for detecting breast cancer and monitoring response to therapy. Several reports have also described the possible clinical application of this agent for lung cancer. Alteration of radiopharmaceutical kinetics in small-animal tumor models may be an important marker of tumor response to therapy. The purpose of this study was to compare the kinetic properties of 99mTc-sestamibi in SCID mice with A549 lung-cancer or MCF7 breast-cancer xenografts by in vivo imaging using a small-animal SPECT system, FastSPECT.
Abstract ID: 122
Malignant melanoma is the most lethal form of skin cancer with an increasing incidence worldwide. Early detection is crucial as melanoma metastases are resistant to current therapeutic regimens. Radiolabeled alpha-melanocyte stimulating hormone (a-MSH) peptides are attractive due to their nanomolar binding affinities to melanocortin-1 (MC-1) receptor over-expressed on most human and murine melanoma cells.
Abstract ID: 123
Abstract ID: 124
Abstract ID: 125
Abstract ID: 126
Diffusion Weighted Imaging (DWI) was performed after treatment initiation using a multi-shot EPI-DWI sequence. Apparent Diffusion Coefficient (ADC) maps were calculated and ADC values were measured in ROIs. After treatment the Vascular Volume Fraction (VVF) was obtained by using iron oxide enhanced MRI. Moreover, tumor volume and histological grade of tumor necrosis were determined.
Abstract ID: 127
As anionic phospholipids, principally phosphatidylserine (PS), become exposed on the external surface of viable vascular endothelial cells in tumors, possibly in response to oxidative stresses present in the tumor microenvironment, this PS might be considered as a target for molecular imaging of those tumors using an adequate targeting vector.
We tested the hypothesis that a new chimeric IgG3 monoclonal antibody (mab) ch3G4 (Tarvacin®), directed against anionic phospholipids and labeled with radioactive arsenic isotopes, can be used for the vascular targeting and molecular imaging of solid tumors in rats in vivo.
The two longer-lived arsenic radioisotopes 74A and 77As were produced, isolated and reacted to the no-carrier-added labeling synthon AsI3. A method for the labeling of antibodies with radioarsenic isotopes was developed and the labeled antibody was tested for its ability in terms of localization to tumor vessels and imaging qualities.
The mab ch3G4 recognized anionic phospholipids on the external membrane of R3227 Dunning prostate solid tumors in male Copenhagen rats. The tumors could be imaged with planar scintigraphy techniques and with Positron Emission Tomography in vivo, showing excellent and antigen-specific localization. In addition, the experiments proved the concept of the radiochemical separations applied and the developed labeling chemistry and demonstrated the potential benefits of the use of arsenic radioisotopes for molecular imaging of antibodies.
The biomedical use of radioactive arsenic isotopes was exemplified for the first time in a multi-modality molecular imaging approach in vivo.
Supported in part by NCI Pre-ICMIC CA86354, the DFG and the Boehringer Ingelheim Foundation.
Abstract ID: 128
FLT, as a cell proliferation tracer, has been shown to be more specific for monitoring tumor growth and treatment response. The distinct dose dependent FLT kinetic pattern has been reported previously. In this study, we analyzed the effects of fractionation using kinetic modeling. Only the implanted right breast tumors in mouse were irradiated. The left tumor was used as a control. The fractionated protocol (2.5Gy twice a day for 2 days) was applied (n=5). Single dose irradiations (n=1, 3, 3 and 4 for 2.5, 5, 10 or 20 Gy, respectively) were performed to examine the difference. Dynamic microPET images were acquired for 60 minutes after injection. Regions of interests (ROIs) were defined at 60 min. The rate constants (K1, k2, k3, k4) were estimated using the kinetic image system developed at our laboratory. The FLT uptake constant, KFLT, was defined as K1*k3/(k2+k3). 60-min FLT ROI value was normalized by 10-min value as a semiquantitative parameter (FLT 60min/10min). No distinct FLT kinetic pattern was observed in 10Gy fractionated studies as compared to those with single dose (Figure 1). FLT 60min/10min of the fractionated dose did not fit the dose response curve of non-fractionation protocols (Figure 2, left panel). The KFLT values of fractionated studies were significantly higher than that of the corresponding single dose studies (Figure 2, right panel). The results implied significantly different responses of tumor after fractionated dose versus single dose irradiation. FLT microPET could be useful to reveal potential difference in thymidine kinase activity due to fractionated irradiation.
The distinct dose-rate dependent FLT kinetic pattern, with increased uptake shortly after injection and faster clearance afterwards, was shown at the irradiated tumor with non-fractionated protocol. For tumor with fractionation dose, such pattern was not observed. (The y-scale at Lt-side is for non-fractionated tumors while that at Rt-side is for fractionated tumors.)
Different responses of tumor cells after fractionated dose versus single dose irradiation when using semiquantitative method (FLT 60min/10min, Lt panel) and kinetic modeling (K FLT, Rt panel).
Abstract ID: 129
Abstract ID: 130
Visualization and quantification of incipient molecular processes during cancer treatment are crucial for developing and delivering successful tumor therapy. Both sodium and diffusion are promising cancer biomarkers. In our previous study, during efficacious tumor treatment in a rat glioma model sodium and diffusion time courses correlated in time and their maximum responses corresponded to a decreased tumor cellularity. Both modalities, many days in advance predict future tumor shrinking. To test if these correlated responses yield a general phenomenon, dual modality imaging was performed during chemotherapy of a subcutaneous 9L tumor. BCNU (1,3-bis (2-chloroethyl)-1-nitrosourea) was administered (IP, 26 mg/Kg), when the tumor size reached ~ 300 μl, in six Fisher tumor bearing rats, while six other untreated animals served as a control. The first signs of tumor treatment were evident by both imaging modalities in two days after initiation of therapy. Maximal increase in sodium and diffusion throughout the whole tumor were detected in treated animals in ~8 days (Fig.1). In a non-treated tumor diffusion and sodium remained at low pre-treatment levels.
Correlated responses of sodium and diffusion during chemotherapy for subcutaneous tumors suggest that these changes may be a general phenomenon reflecting the same process of tumor cell destruction, being observable and predictive for therapy outcomes even beyond gliomas. The results also support the current clinical testing of ADC tumor treatment evaluation and demonstrate the complementary insights of sodium MRI for oncology applications.
Abstract ID: 131
The LAT1 amino acid transport system is over-expressed in many tumour cell lines but not in normal tissue. We and other authors noticed that besides neutral lipophilic L-amino acids LAT1 can also transport the D-analogs. We compared the uptake properties of [123I]-2-Iodo-D-phenylalanine (D2IPHE) with those of the L analog (L2IPHE) which already showed a high specific uptake in several tumour types.
Poster Session: Novel reporter genes and Acoustics/therapeutics
Abstract ID: 132
Abstract ID: 133
Molecular imaging on small animals has attracted widespread attention. In ultrasound, imaging needs to be done at frequencies higher than those for clinical ultrasound (> 20MHz) in order to provide sufficient resolution on small animals. The higher operating frequencies also lead to the need to reduce the size of contrast agents, which are typically encapsulated microbubbles that can be further modified as molecular probes. In this regard, critical tasks in ultrasonic small animal molecular imaging include the production of appropriate contrast agents that are suitable for high frequency imaging and the development of efficient imaging methods for detection of the molecular probes. In this paper, pulse inversion based fundamental imaging is proposed for the enhanced detection of liposome microbubbles that are used as a contrast agent and thus a candidate for ultrasonic molecular probes. Imaging was done at 25–45 MHz with spatial resolution at the order of 100 microns. The liposome microbubbles were produced in-house and they outperformed commercially available contrast agents. The pulse inversion technique involves two firings with inverted waveforms. When the returning echoes from the two firings are summed, the residual signal is limited to even-order harmonics for tissue. However, when the returning echoes are from microbubbles, the fundamental signal is not completely cancelled because the reaction of the bubbles under compression is different from that under rarefaction. Thus, with the application of the pulse inversion technique, the signal in the fundamental band can be used to enhance the contrast-to-tissue ratio. Phantom experiments were performed. Compared to conventional imaging, the contrast-to-tissue ratio was enhanced by 7~18 dB when the transmit signal was at 25–45MHz. With this technique, the microbubbles can be further combined with ligands for small animal molecular imaging.
Abstract ID: 134
Among molecular imaging modalities, ultrasound has a unique feature that it is applicable not only for diagnosis but also for therapeutic purposes. Considering applications in tissues such as tumors, contrast agents should be smaller than about 200 nm to leak from blood vessels to be delivered into tissues. However, such small agents are an order of magnitude too small to produce high contrast with ultrasound at 1–10 MHz. We propose the use of nano-sized emulsion systems which change their phase from liquid to gas, generating microbubbles, upon exposure of ultrasound pulses. They can be delivered into tissues due to the small size of the emulsion and produce high contrast due to high echogenicity and characteristic non-linearity of the microbubble. Moreover, highly selective ultrasonic tumor treatment should be possible due to the ultrasonic-energy deposition in the vicinity of microbubbles.
In in vitro experiments, it was confirmed that our nano-sized emulsion system changes its phase by ultrasound exposure at relatively low acoustic intensity (less than 10 W/cm2). Further, in vivo experiments using subcutaneously implanted murine tumor showed that the phase can be shifted by ultrasound exposure under the same conditions as above in vitro experiments. The emulsion particles remained in the tumor tissues at least for two hours after intravenously administered while maintaining the ability to shift the phase. It was also found that focused ultrasound at an acoustic intensity of 20 W/cm2 contiguously exposed after the phase shift is established in the tumor tissues induce therapeutic effects at the focus.
Our results indicated the potential usefulness of the nano-sized emulsion system as a versatile agent for ultrasonic diagnosis and therapy.
Part of this work was supported by the Japanese Ministry of Education, Culture, Sports, Science and Technology through a grant-in-aid for the creation of innovations through business-academic public sector cooperation.
Abstract ID: 135
Molecular imaging with high frequency ultrasound (> 15 MHz) has potential for applications in ophthalmology, cardiology and small animal imaging. Reports to date have detected targeted microbubble agents with 20–40 MHz intravascular ultrasound (IVUS) by means of echogenicity enhancement. For low ultrasound frequencies (< 10 MHz) it has been shown that targeted bubble detection can be improved using harmonic imaging. In this study we investigate second harmonic and subharmonic (respectively, twice and half transmit frequency) imaging of targeted microbubbles for high frequency ultrasound with prototype IVUS instrumentation.
A harmonic IVUS system was developed to perform conventional fundamental imaging at 20 MHz (F20) and 40 MHz (F40), second harmonic imaging at 40 MHz (H40), and subharmonic imaging at 20 MHz (SubH20). The agent examined was an experimental biotinated, lipid encapsulated formulation comprised substantially of micron to sub-micron bubbles (BG3039; Bracco Research, Geneva). The agent was targeted to an avidin coated agar substrate and imaging experiments were performed with the agar situated at 3 mm from the transducer. A block of tissue mimicking phantom was positioned within the imaging plane to provide reference tissue signals.
Fundamental imaging resulted in low contrast-to-tissue signal ratios (CTR): 3 and 1 dB for F20 and F40 respectively. Harmonic imaging improved the CTR significantly. H40 imaging had a CTR of 15 dB, an improvement of 12 dB relative to F20 mode. SubH20 imaging suppressed tissue signals below the noise floor, with a signal to noise ratio of up to 16 dB, depending on transmit amplitude. These results provide an in vitro demonstration of the feasibility of harmonic imaging as a strategy for improving the sensitivity of targeted contrast agent detection at high ultrasound frequencies.
Abstract ID: 136
Radiation therapy of solid tumors using radiolabeled antibodies is often compromised because the therapeutically effective dose is close to the maximum tolerated dose. The therapeutic effective dose is a function of the target antigen levels at the tumor site, the Kon and Koff rates of the antibody, and the nature of the radioisotope. Recently, we developed a novel antibody, 19G9, targeting the prostate associated extracellular target mindin/RG-1, which is highly effective in inhibiting growth of LNCaP tumors in nude mice when labeled with 90Y. Here we report dosimetry calculations based on biodistribution studies with 111In labeled 19G9 antibody in mice.
Using the MIRDOSE 3.0, we calculated the estimated radiation dose in human from mice biodistribution data with the assumption that the kinetics of this therapy agent would be the same in human as in mice. These calculations predict a dose deposition to the critical organ of 19.4 rad/mCi to the testes and of 12.4 rad/mCi to the kidneys. The whole body dose was 1.75 +/- 0.07 rad/mCi. The radiation dose to the implanted tumors was calculated using the nodule module from MIRDOSE 3.0. The average tumor mass was 0.174g and the residence time was 3.49 hr. For such tumors, the radiation dose was 19900 rad/mCi.
Based on these assumptions, a therapeutic dose is predicted for a radiolabeled 19G9 antibody that is well below the MTD. Furthermore, immunohistochemistry analysis of mindin/RG-1 expression in prostate tumors reveals that the target antigen expression is frequently higher in tumor tissues when compared to the expression in LNCaP xenografts. Consequently, the calculation with the LNCaP model may represent a low estimate of the dose delivered to the human tumors. We conclude then, that the radiolabeled antibody 19G9 may deliver an effective anti-tumor dose of radioactivity for therapy of prostate cancer without causing major toxicity.
Abstract ID: 137
Abstract ID: 138
The androgen receptor (AR) is a key cellular regulator in normal and malignant prostate cells and a prime therapy target. Whereas its inhibition in the hormone naïve tumors by androgen ablation and antiandrogens is efficient and allows controlling tumor growth for some time, targeting it after development of hormone-refractory disease remains a challenge. A promising approach is antisense AR knockdown. The main obstacle is the problem of specific and efficient delivery of antisense molecules into the tumor. We studied ultrasound contrast microbubbles as carriers for antisense molecules and their delivery into tumor cells xenografts by ultrasound triggered bursting.
Antisense molecules were charge-coupled to cationic perfluorocarbon gas-filled microbubbles and added to LNCaP prostate cancer cells cultured in Opticells chambers. In a water bath the loaded microbubble were then bursted with ultrasound. For in-vivo testing digoxigenin labeled antisense oligonucleotides were loaded to cationic microbubbles and delivered into xenograft tumors by ultrasound bursting after intra tumural or intravenous application. 24 hours after the last of three treatments tumors and organs isolated and analyzed by anti-digoxigenin immunohistochemistry.
Delivery of 50 pmol of AR antisense oligonucleotide or siRNA loaded on 1 × 107 cationic microbubbles resulted in a significant uptake of fluorescence labeled antisense molecules (more than 50% positive cells) and a significant down regulation of AR protein in LNCaP cells. Treatment was associated with induction of apoptosis and inhibition of proliferation. In the xenograft model uptake of labeled antisense oligonucleotides in tumors was confirmed.
We conclude that ultrasound contrast microbubbles are suitable as carriers for small antisense molecules and can be used in combination with ultrasound bursting for efficient delivery of these drugs into tumor cells in-vitro and in vivo.
Supported by the Austrian Research Funds (FWF).
Abstract ID: 139
Whereas gene delivery using TNF-alpha has been indicated as one of the few promising therapies for treating squamous cell carcinoma of the head and neck, limited success has been achieved due to insufficient distribution of the plasmid, consequent expression and therapeutic efficacy. Previous studies using pulsed-HIFU have shown how these noninvasive exposures can enhance systemic and local delivery of a variety of agents including naked DNA.
Squamous cell carcinoma (SCCVII) tumors were grown in the bilateral flanks of C3H mice. Approximately one week post-inoculation, tumors were exposed to pulsed-HIFU and given a single intratumoral injection of naked TNF-alpha plasmid. Tumor growth rates were then monitored and compared to tumors receiving plasmid injections only. Monitoring was also carried out in untreated tumors and those receiving either pulsed-HIFU exposures or saline injections only. Laser scanning confocal microscopy was also performed on treated tumors injected with fluorescent nanoparticles to illustrate the manner in which pulsed-HIFU enhances the distribution of locally injected substances. In addition, treated and untreated tumors were observed using transmission electron microscopy (TEM).
Growth rates of tumors treated with either pulsed-HIFU or saline injection only were not different than controls whereas regression wasfound in tumors treated with TNF-alpha only. Enhanced regression was observed when TNF-alpha was combined with pulsed-HIFU. Confocal microscopy data indicated improved distribution of nanoparticles which was supported by ultrastructural analysis with TEM.
It is hypothesized that radiation force-induced displacements resulting from pulsed-HIFU exposures created structural changes in tumor architecture allowing for increased distribution of TNF-alpha plasmid and consequent expression. This type of noninvasive adjunct therapy could potentially be implemented in the clinic for treating head and neck cancer and other solid tumors currently being evaluated for gene therapy.
Abstract ID: 140
Ferritin was previously proposed by our group and others as an endogenous reporter protein which would directly change MRI image contrast, in its expression site, without the need to administer exogenous contrast materials [1,2]. Over-expression of ferritin leads to physiological compensation augmenting iron uptake. Sensitivity for detection by MR can be further enhanced by changes in relaxivity due to redistribution of iron within a larger ferritin pool [3]. We report here the generation of transgenic mice which over-express HA-taggedferritin and EGFP in a tissue specific manner under tetracycline regulation. Feasibility for MR detection was evaluated in mice in which transgene expression was selectively induced in all vascular endothelial cells. Brain MRI spin echo images of the transgene expressing mice and matched control siblings, were acquired at 4.7T with 8 different TE values and used for derivation of R2 values. Significant elevation in R2 was detected in all regions tested including the cortex, hypocampus and the facial muscle. The ability to detect significant contrast changes from endothelial cells, which comprise less than 1 % of the tissue demonstrates the feasibility and sensitivity of this approach, furthermore suggesting that ferritin can be used as an endogenous reporter for MRI tracking endothelial cell migration during angiogenesis. This approach provides a platform for selective expression of the reporter in other target organs.
* first two authors contributed equally to this work.
This work was supported by the Minerva Foundation.
Abstract ID: 141
Abstract ID: 142
Several modalities are currently available for noninvasive imaging in vivo, such as bioluminescence, positron emission tomography (PET), and fluorescence, each with its own strengths and limitations. In order to take advantage of the strengths of each and facilitate multi-modality imaging, a reporter construct was engineered in which firefly luciferase (FLuc) was fused to the N-terminus of a mutant herpes simplex virus thymidine kinase previously enhanced for imaging (mNLS-SR39TK). Expression of the fusion protein was driven by the ubiquitin promoter and a 12 amino acid linker was inserted between the two proteins to minimize steric constraints on enzyme activity. A second construct was developed in which Monster green fluorescent protein (mGFP), a recently available enhanced fluorescent protein from Promega, was introduced into the fusion vector downstream of an internal ribosome entry site (IRES) to allow analysis by fluorescence microscopy or flow cytometry. Fusing a fluorescent protein onto the FLuc-mNLS-SR39TK fusion reporter may compromise the activity of the two enzymes, while providing no quantitative advantage, as the qualitative nature of fluorescence imaging does not allow for quantitative correlation of its signal with other modalities.
Functionality of each of the enzymes was validated in cell culture; FLuc activity was measured by a cooled CCD camera (IVIS) and mNLS-SR39TK activity by 3H-penciclovir uptake. While HeLa cells transiently transfected with the FLuc-mNLS-SR39TK-IRES-mGFP construct had approximately 50% of the penciclovir signal of HeLa cells transiently transfected with the FLuc-mNLS-SR39TK construct alone, the ratio of mNLS-SR39TK:FLuc activity was comparable, indicating that introduction of the IRES-mGFP did not affect the specific activity of the mNLS-SR39TK.
This second-generation fusion construct may be useful for concurrently obtaining semi-quantitative bioluminescence and quantitative PET imaging data as well as qualitative fluorescence data in cell culture and living animals.
Abstract ID: 143
Currently, different PET tracer-reporter gene systems possessing varying characteristics are making big strides in non-invasive evaluation of gene therapy. Considering the significance of safer human use, versatile tissue permeability and effective gene expression, we have developed an adenoviral PET ligand-reporter gene system utilizing 18F-Estradiol (FES) as PET ligand, human estrogen receptor a ligand binding domain (h
Cultured Cos-7 cells were infected with varying titers of recombinant Ad5CMV-TIER virus. ECL Immunoblotting showed increasing hTP and hERL expression levels, corresponding to increased viral titers. A periodic gene expression assay performed over 7 days post infection (p.i.) of Ad5CMV-TIER virus showed optimum hERL expression at 2 days p.i., thereafter gradually decreasing over 7 days p.i. However, hTP levels remain unchanged up to 7 days p.i. hERL/hTP bands were not seen in control virus expression. An in vitro 3H-Estradiol specific binding assay was performed in Cos-7 cells 2 days p.i., with 10 min uptake and the unlabeled ligand at 200-fold concentration of the labeled ligand. With TIER virus, the uptake was 44-fold higher in radioligand only added cells, than in cells treated with both labeled and unlabeled estradiol, indicating high specificity of radioligand to the expressed hERL receptor. The uptake in LacZ expressed radioligand only added cells were 69-fold lower than that of corresponding TIER expressed cells, indicating very low non-specific binding. Radioligand disassociation study performed at three different wash conditions showed uptake in TIER expressed cells decreasing by 26%, whereas in LacZ expressed cells it decreased by 97%, indicating high affinity of the radioligand to the expressed hERL protein. Considering the positive results in cell model, we are moving to animal model to further validate our system for prospective clinical use.
Abstract ID: 144
We designed a new PET reporter gene system with 18F-fluoroestradiol(FES) and human estrogen receptor ligand binding domain(hERL) which would work in wide variety of tissues including brain, have minimized physiological effect brought by the reporter expression, and be safe for human. We constructed a plasmid, pTIER, to co-express a model therapeutic gene and hERL and evaluated the potential of the reporter gene system in gene therapy monitoring in vitro and in vivo. Cos7 cells transfected with pTIER expressed hERL along with the therapeutic gene. When the pTIER transfected cells were injected into mouse abdominal cavity and the radioligand, either 3H-estradiol or 18F-FES, was injected from the tail vein, the cells took up c.a. 20 times higher amount of the radioligand than the control cells, which demonstrates that hERL expressing cells can take up and accumulate radioligand efficiently from circulation. When pTIER was transfected into mouse calf muscle by in vivo electroporation, higher accumulation of FES was observed with the pTIER transfected side than the other, control plasmid transfected side. These data support the prospect that our in vivo reporter gene system would be useful in gene therapy monitoring.
Abstract ID: 145
The deoxyribonucleoside kinase of Drosophila melanogaster (Dm-dNK) is closely related to herpes simplex virus thymidine kinase (HSV1-tk). Dm-dNK shows broad substrate specificity to both pyrimidine and purine nucleosides but preferentially phosphorylates pyrimidine nucleosides. In this report, for the first time we explore the use of Dm-dNK as a PET reporter gene using a pyrimidine analog [F-18] FIAU (1-(2′-deoxy-2′-fluoro-1-β-D-arabinofuranosyl)-5-iodouracil).
Subcutaneous xenografts on opposite flanks of CD1 nu/nu mice were generated using MH3924A and MH3924A-dNK cell lines. For biodistribution and microPET imaging studies mice were injected IV with 80 and 150 μCi of [F-18] FIAU, respectively. In the first group, three mice underwent dynamic imaging from 5–65 min followed by static imaging at 65 min and 120 min post injection (Concorde R4 microPET scanner). Then, all animals were sacrificed at 150 min. Second group of five animals were sacrificed at 60 min post tracer injection. Samples of blood and selective tissue were collected for biodistribution and metabolite analysis. Regions of interest were drawn on both the tumors in the microPET images to calculate the time activity curves. All the activity measurements from ex vivo counting were converted into standard uptake values.
MicroPET imaging studies showed a clear accumulation of the tracer in MH3924A-dNK tumors. Biodistribution studies showed highest uptake of activity in the MH3924A-dNK tumors. The tissue/muscle ratios for MH3924A-dNK tumors were 3.1 ±0.36, 4.30±0.61 (mean±SE) and that of MH3924A tumors were 1.15±0.07, 1.44±0.18 at 60 and 150 min respectively. FIAU showed rapid accumulation in MH3924A-dNK tumors and slowly reached a plateau over 30 min.
Our studies illustrated a clear uptake and retention of [F-18] FIAU in the MH3924A-dNK tumors preferential to control tumors. This study demonstrates that it is practical to image the Dm-dNK gene expression using [F-18] FIAU as a reporter probe with microPET.
Abstract ID: 146
Reporter genes offer the ability to monitor gene expression in living systems and thereby directly observe cellular physiology in vivo. In order to apply reporter gene techniques to hypoxia, the behavior of reporter genes in low oxygen conditions, in terms of gene transcription, translation, and activity of the resulting protein product, must be rigorously assessed. We exposed 293T cells stably expressing a triple fusion reporter consisting of renilla luciferase, monomeric red fluorescent protein, and truncated HSV1-sr39 thymidine kinase under a CMV promoter to environments containing 21, 2, 0.5, or ≤0.02% O2. After 24 hours the cells were either assayed for bioluminescence while still under controlled oxygen conditions, for fluorescence levels using fluorescence microscopy, or assayed for thymidine kinase activity using a tritiated penciclovir assay. Western blotting was performed to assess any variations in luciferase protein levels induced by exposure to varying oxygen levels. Our findings suggest that bioluminescent light production, fluorescence, and thymidine kinase activity are hampered at low oxygen concentrations. A similar effect was observed in a solid tumour model. The pancreatic cell line MiaPaca2, stably transfected with the firefly luciferase gene was inoculated subcutaneously into nu/nu mice. Once tumours reached 7–8 mm in diameter, luciferin was administered and tumour luminescence was measured. To visualize the effect of oxygen levels on bioluminescence, blood flow to the tumour was temporarily occluded by use of a clamp. During 10 minutes of occlusion, the luminescence signal in the tumour decreased by approximately 40-fold, as shown in Figure 1. When the clamp was removed, the signal immediately returned to pre-clamping levels. These findings should be considered when devising reporter gene strategies to monitor tumor hypoxia.
Abstract ID: 147
Several reporter genes available include Aequorea-derived fluorescent proteins (AFPs), luciferases, and herpes simplex virus type 1 thymidine kinase (HSV1-tk) genes. Optical reporter genes have limited clinical applications because of poor tissue penetration and resolution. Thymidine Kinase reporter genes are limited by non-specific background activity of endogenous enzyme and the requirement for expensive radio-labeled probes for imaging. Tc99m O4- is widely used as a radiopharmaceutical in nuclear medicine because of its excellent imaging properties, low toxicity and cost, and easy availability. Enzymatic reduction and intracellular trapping of Tc99m O4- has been reported in Escherichia coli (E. coli), sulfate-reducing bacteria Desulfovibrio Desulfuricans (DD) and in the eukaryotic human parasite Trichomonas vaginalis (TV). Two genes, HydA and HydB encode for the hydrogenase activity in DD. We demonstrate that this microbial reductase is able to mediate enzymatic reduction and intracellular trapping of Tc99m O4- in bacteria and mammalian cells. HydA and HydB genes from (DD) were cloned and expressed in bacteria and Hela cells. The mRNA and protein expression were confirmed using RT-PCR, SDS-PAGE and Western blotting. Tc99m O4- trapping/reduction assays were performed on E. coli and Hela cells over-expresing the HydA/B genes. Both E. coli and Hela cells demonstrated 10–15 fold higher association of radioactivity over controls These results of in vitro assays along with imaging of cell pellet and in vivo imaging bacterial and tumor imaging in mice will be presented. Overall, we present a novel reporter-probe system for in-vivo SPECT imaging.
Abstract ID: 148
Incremental truncation screens recently identified fragments of Firefly (Photinus pyralis) luciferase that enabled a protein complementation assay useful in studying protein-protein interactions. Firefly luciferase and the family Click Beetle (Pyrophorus plagiophthalmus) luciferases share high domain homology fostering the possibility of creating split luciferase pairs with varying spectral emissions, while utilizing the same substrate. The utility of a two color system would expand the range of protein-protein interaction assays.
Each protein was split such that a critical overlap region was retained on the N-Luc and C-Luc fragments. This overlap region spans residues 397–416 of Firefly luciferase and residues 394–413 of Click Beetle Red and Click Beetle Green (PNAS 101:12288, 2004). Combinations of all possible fragments were assayed using rapamycin-inducible interaction of FRB with FKBP fused to N-Luc and C-Luc fragments, respectively. All the pairs, when interacted with each other, yielded comparable activity to the original Firefly luciferase N-Luc and C-Luc complementation pair. While some pairs displayed better overall intensity, others demonstrated a higher rapamycin dependent fold-induction than the original split luciferase pair. However, contrary to our primary hypothesis, the emission spectra of all the pairs collapsed to a maximum emission near 580 nm. This suggests that a less restrictive binding pocket within the enzyme limits the range of possible color variants. This collapse to yellow-green emission may be due to several factors such as susceptibility of internal residues to the external environment and/or the rotational angle between aromatic groups of the excited state of the substrate.
Abstract ID: 149
Bioluminescent proteins have become essential tools for molecular imaging, primarily as reporter genes in the context of small animal imaging. Recently, their use has been expanded to include bioluminescent tagging, in which the protein itself is conjugated to a ligand and the resultant fusion protein is used as an imaging probe. We have been developing mutants of Renilla luciferase (RLuc) that possess improved characteristics for both reporter gene applications and bioluminescent tagging.
Candidate mutations were chosen using a consensus driven approach, via homology to bacterial haloalkane dehalogenases, and screened for light output and stability. A mutant (RLuc8) combining a total of 8 mutations exhibited a 150-fold in vitro serum stability increase compared to RLuc (>100 hrs versus 0.7 hrs) and a 4-fold increase in light output. Transient transfection of RLuc8 in mammalian cells showed a 15-fold increase in signal compared to RLuc after 24 hours. A double mutant (RLuc-M185V/Q235A) was developed that exhibited a 50 percent reduction in stability compared to RLuc but increased light output 4-fold. When used with the substrate analogs coelenterazine-cp, coelenterazine-n, and bisdeoxycoelenterazine (DeepBlueC), RLuc8 showed 6, 9, and 60-fold increases in light out-put, while RLuc-M185V/Q235A showed 14, 7, and 20-fold increases. Only a portion of these increases could be explained by enhancements in quantum yield.
Ongoing work is attempting to alter the emission spectrum of RLuc, by targeting mutations based on a homology model and an estimate of the substrate orientation in the active site. One mutant with a 10 nm red shifted emission showed little drop in light output, while another mutant showed a 20 nm red shift but with a large drop in output.
In summary, RLuc8 represents a significant improvement for the bioluminescent tagging of ligands, and both RLuc8 and RLuc-M185V/Q235A are significant improvements over the native luciferase for reporter gene applications.
Abstract ID: 150
We reported previously the in vivo detection of ectopic and transient expression of creatine kinase gene (ck) in the liver by phosphorus-31 magnetic resonance spectroscopy (31P MRS). Here we demonstrate the feasibility of utilizing ck as a reporter gene to monitor the transfer of low density lipoprotein receptor (LDLr) gene to LDLr-/-mice, a preclinical model for familial hypercholesterolemia. A recombinant adenovirus was generated which contains creatine kinase gene (ck) and human LDL gene (hLDLr) linked by an internal ribosomal entrance site (IRES) sequence. Intravenous injection of the adenovirus into LDLr-/-mice (E11 viral particles/per mouse) resulted in transduction of over 90% hepatocytes in the liver. Simultaneous expression of ck and LDLr was confirmed by western analysis of the transduced livers. Through precise regulation of transgene expression in hepatocytes in vitro, an excellent correlation (R2=0.96) between LDLr and ck expression was demonstrated over a wide range of viral dose. In vivo 31P MRS was employed to detect the metabolic product (i.e., PCr) of the CK reaction. CK activity, which is a true measure of ck gene expression, was quantified in vivo by magnetization transfer technique. In conclusion, as a liver-specific reporter, ck is useful to monitor any
Abstract ID: 151
Apoptosis is an essential event for normal physiology and has become an area of intense exploration specifically in the arena of cancer. Though advances made in elucidating the intracellular networks that govern the apoptotic pathway are astonishing, the ability to examine this pathway from a global perspective in an in vivo model is lacking. As previously reported, a recombinant luciferase reporter molecule, activated during an apoptotic event through proteolytic cleavage by caspase-3, has been developed for bioluminescence imaging of apoptosis. The ability to non-invasively image apoptosis allows for unique observation of this event in vivo, which would be useful in many areas of research including development of compounds targeting the apoptotic pathway especially in the area of cancer therapeutics.
5-fluorouracil (5-FU), a common chemotherapeutic, has been recently reported to enhance sensitivity of certain cell lines to TRAIL (tumor necrosis factor alpha-related apoptosis inducing ligand) induced apoptotic death. In this study, we examined the effects of 5-FU in sensitizing subcutaneous D54 tumors to TRAIL induced cellular death and validated the observed effects in vivo using our apoptosis reporter system. D54 tumors constitutively expressing our reporter construct were established subcutaneously in athymic mice and 5-FU (60mg/kg) was administered on day 0 and 5 while TRAIL (4mg/kg) was administered daily through i.v. for 10 days.
In Figure 1, treatment with TRAIL in conjunction with 5-FU demonstrated marked sensitivity to TRAIL specifically after the second administration of 5-FU. This enhancement in apoptotic activity correlated with increased tumor growth delay as shown in Figure 2.
Poster Session: Design Methods for Imaging Probes
Abstract ID: 152
Magnetic nanoparticles are usually considered by their core size, often in the range of 4–20nm in diameter. Here we show in both experimental and numerical data, that increasing coating size results in decreased R2 as well as decreased clustering effect.
PEG-modified, phospholipid micelle-coated nanoparticles were developed with three coating sizes. Monte Carlo simulations were performed with an excluded, spherically-shaped space into which protons were not allowed to diffuse. Figure1(A) depicts the magnetic field of a particle core. When surrounded by a large coating (blue), protons are excluded from high field gradients near the core, significantly reducing R2.
Figure1(B) shows experimental results indicating radius changes of only a few nanometers can have significant effects on R2. Simulation data (Figure1[C]) show R2 decreasing with increasing exclusion volume. Similarity of the results indicates that at least some of the experimentally observed decrease in R2 is due to physical exclusion of protons from the immediate neighborhood of the magnetic core.
Particle clustering has been used as a mechanism for switchable contrast since it increases R2. Therefore, we investigated the effect of coating on change in R2 due to clustering (ΔR2). Simulation data (Figure1[D]) demonstrate reduced ΔR2 as coating thickness is increased for a given core size, indicating a reduced clustering effect. As clustering effects increase with core size, particles with larger cores and minimal coating thickness are optimal for maximizing the clustering effect.
In constructing functionalized particles using proteins such as antibodies and receptor ligands, effective coatings can become quite large. These results indicate the need for maintaining small coatings in nanoparticle probe design and are useful for designing complex, activatable magnetic nanoprobes.
Abstract ID: 153
The study analyses the hepatic uptake of anionic magnetic nanoparticles (AMNP), a new class of particulate MRI contrast agent. AMNP are 35nm, DMSA stabilized maghemite particles. It was previously demonstrated that AMNP is effective for cellular tracking because of high affinity for cellular-internalization ex vivo. Application of AMNP can be extended to cell-specific in vivo targeting by conjugating AMNP with biological effectors.
Prior longitudinal studies by ferromagnetic resonance (FMR) showed that 55–74% of AMNP reside in intact liver (parenchyma, blood and bile) while 25% reside in liver parenchyma 20–30min after AMNP intravenous injection (IV). In this study, we performed longitudinal, NMR relaxometry quantification of bile and blood to account for the above difference. We delivered 16.5–35 μmol Fe/kg into rats by IV. Bile samples were obtained by cannulating the common bile duct over 240min. T2 of samples were obtained by relaxometry at 0.47T, from which the Fe mass in bile and blood were calculated.
Abstract ID: 154
Complex high-molecular-weight bioactive chemical structures, such as peptides, proteins and oligonucleotides are proposed as radiopharmaceuticals and their applications are rapidly gaining importance in nuclear medicine. Based on recent advances in nucleophilic hetero-aromatic fluorinations, two new [18F] fluoropyridine-based reagents have been designed for the prosthetic labeling of these macromolecules : (a) [18F]FPyME (1-[3-(2-[18F]fluoropyridin-3-yloxy)propyl]pyr-role-2,5-dione) dedicated to the labeling of peptides and proteins via selective conjugation with a thiol function as borne by a cysteine residue and (b) [18F]FPyBrA (2-bromo-N-[3-(2-[18F]fluoro-pyridin-3-yloxy)propyl] acetamide) dedicated to the labeling of oligonucleotides via selective conjugation with a phosphorothioate monoester group.
[18F]FPyME and [18F]FPyBrA were both efficiently prepared using a three-step radiochemical pathway: (1) a high-yield nucleophilic hetero-aromatic ortho-radiofluorination on appropriated ammonium trifluoromethanesulfonate precursor; (2) a rapid and quantitative TFA-induced removal of the N-Boc-protective group and (3) an optimized maleimide or bromoacetamide formation. Conjugation of [18F]FPyME with peptides and proteins (performed in a mixture of DMSO and buffer at room temperature for 10 min) was achieved in 60–70% non decay-corrected and isolated yields and routinely, 1.85 GBq of LC-purified conjugated [18F]proteins were obtained starting from 2.96–3.70 GBq of [18F]FPyME. Conjugation of [18F]FPyBrA with oligonucleotides (in a mixture of MeOH and buffer at 120°C for 15 min) was achieved in 20–30% yields and routinely, 0.37–0.55 GBq of HPLC-purified conjugated [18F] oligonucleotides were obtained starting from 1.85 GBq of [18F]FPyBrA. As demonstrated by LC analysis the labeled compounds were more than 95% radiochemically pure.
[18F]FPyME and [18F]FPyBrA both represent new valuable, thiol-selective, fluorine-18-labeled reagents for the prosthetic labeling with fluorine-18 of peptides, proteins and oligonucleotides. These reagents have been used to label two 8-kD proteins (c-AFIM-0 and c-STxB), currently developed as tumor imaging agents as well as 9- and 18-mer natural and chemically-modified oligonucleotides.
Supported by EMIL (European Molecular Imagaing Laboratories) EU contract LSH-2004-503569.
Abstract ID: 155
The paramagnetic liposomes are increasingly used as magnetic reporters for magnetic resonance imaging (MRI). The liposomal structure allows to concentrate paramagnetic lipophilic derivatives in the membrane and so, to target specific receptors. However, the evaluation of the relative contributions of the inner and outer layers of the liposomes to the relaxivity has not been carried out so far.
In this work, two kinds of liposomes incorporating Gd-DTPA-BC14A have been synthesized: liposomes with Gd-complexes inside and outside the phospholipidic membrane, prepared as described by Lasic [1] and liposomes with Gd complexes only in the external layer. To obtain this last type of paramagnetic system, liposomes incorporating lanthanum complexes have been prepared and submitted to a transmetallation with gadolinium ions (the evolution of the transmetallation has been followed by relaxometry). The relaxivity of the liposomes with Gd-DTPA-BC14A in external and internal membrane layers is lower than that of the liposomes carrying this complex only on the external part of the membrane. It thus appears that complexes in the internal layer of the membrane contribute less to the global relaxivity of the paramagnetic liposomes than complexes in their external layer.
Abstract ID: 156
Fluorescence tomography has emerged as a promising technique forsmall animal imaging with many applications in medicine and biology. Recentattention has been given to resolving engineered fluorochromes with molecular specifity that emit in the near-infrared to achieve penetration depths of several centimeters. It would be however highly desired to develop appropriate technologies to tomographically resolve fluorescent proteins in-vivo in small animals. Tomography can account for the non-linear dependence of light intensity to depth and tissue optical heterogeneity to provide accurate maps. Such development can resolve fluorescent proteins with higher sensitivity and less sensitivity to tissue auto-fluorescence and can provide information on important biological processes in-vivo in small animals.
We have developed a Fluorescence-mediated Molecular Tomography system that facilitates three dimensional in-vivo imaging and quantification offluorescent signals in small animals. This system enables the acquisition ofcomplete 360° angle projections in a non-contact geometry, as well as the capability of 3D surface reconstruction of bodies of arbitrary shapes. It consists of a rotating sample holder and a lens coupled CCD camera in combination with a fiber coupled laser-scanning device, and a custom-made back illumination system for the surface reconstruction. An Argon-ion laser is used as the source and different filters are used for the detection of fluorophores or fluorescing proteins. With a non-contact tomographer a large measurements dataset can be achieved while the use of inversion models give a high capacity for quantitative 3D reconstruction of fluorochrome distributions as well as high spatial resolution. The system has been used in the observation of the distribution of GFP expressing T-lymphocytes in-vivo for the study of the function of the immune system in a murine model.
Abstract ID: 157
Backbone Metal Cyclization (BMC), a novel approach for radiolabeling peptides, was recently reported[1]. This approach enables the fast discovery of backbone cyclic analogs based on the sequence of biologically active linear peptides containing a metal ion in the bridge. It is anticipated that backbone cyclization will improve the metabolic stability of peptides[2], thus opening new avenues to peptide drug development. We have applied the BMC methodology to develop GnRH analogs bearing a technetium-99m oxide or rhenium oxide as lead compounds for cancer molecular imaging and radiotherapy. Cyclo(Re(O)1–10)[Cys-Gly]1[D-Ala]6[Nα(η-Cys-amino hexyl)]10GnRH (cyclo[Re(O)-Gn-2]) was synthesized and its binding affinity to the GnRH receptor was measured (IC50 = 50nM).
The aims of this study were to optimize the synthesis of radiolabeled BMC GnRH analog, cyclo[99mTc(O)-Gn-2], to determine the extraction partition from blood at different time points and to evaluate the stability of cyclo[99mTc(O)-Gn-2] in blood.
We have improved the radiosynthesis and developed an efficient procedure for the purification of the radiolabeled peptide suitable for biological studies. The extraction partition experiments showed constant high activity in plasma of 90%. The stability results indicated that about 70% of cyclo[99mTc(O)-Gn-2] remained intact in blood after 4 hours of incubation in comparison to the native GnRH which degrades within minutes in blood.
Abstract ID: 158
Abstract ID: 159
Integrins are heterodimeric transmembrane cell adhesion molecules involved in tumor angiogenesis and metastasis. Integrin αvβ3, in particular, is significantly upregulated on invasive tumor cells and tumor vasculature. Quantum dots (QDs) with size- and composition-tunable fluorescence emission have high quantum yields and photo-stability suitable for optical imaging and multiplexing. We have labeled QD655 (emission wavelength 655 nm) with monomeric RGD peptide c(RGDyK) (potent integrin αvβ3 antagonist) and the resulting conjugate QD655-RGD was tested for in vitro staining of several cancer cell lines (U87MG glioblastoma, MDA-MB-435 breast cancer and C6 rat glioma which have decreasing levels of integrin αvβ3 expression), ex vivo tissue staining, and in vivo tumor targeting.
The in vitro cell staining experiments successfully demonstrated the integrin targeting ability of QD655-RGD, as the QD655 stained cells showed no visible fluorescence signal while the cells with high integrin αvβ3 expression level (U87MD & MDA-MB-435) were lit up by QD655-RGD and the fluorescence could be effectively blocked when unconjugated c(RGDyK) was added (see Figure). Moreover, the fluorescence intensity correlated well with the integrin level of the cell lines. Based on the initial results shown here, a combinatorial approach which takes advantage of the multiplexing ability of the quantum dots, high integrin targeting efficacy of the cyclic RGD peptides, and emission wavelength in the near infrared window (700–900 nm) will have great potential in cancer diagnosis and imaging as well as imaging-guided surgery and therapy.
In conclusion, for the first time we demonstrated that suitably labeled quantum dots are capable of specific integrin αvβ3 targeting. This will have great potential and may open up new perspectives in integrin-targeted optical imaging.
Abstract ID: 160
A variety of human tumors of epithelial origin express a high affinity for vitamin B (Folic acid), a water-soluble vitamin composed of a pterin ring, para-aminobenzoic acid, and glutamic acid. They achieve this affinity through the expression of the high affinity folate receptor (hFR) that binds folate with a dissociation constant of about 10−9. Several cancer diagnostic and therapeutic agents have been targeted to the hFR by attaching these agents to folate molecules. Attachment of these molecules must occur at the gamma carboxylate of the glutamate moiety in order to achieve similar affinities for the hFR. Once bound to the hFR, these agents enter the cells via receptor-mediated endocytosis.
We recently developed a new strategy to target the hFR based on the hypothesis that any alpha-amino acid linked to the para-aminobenzoic acid and pterin ring such that the alpha acid remains free will target the hFR. We prepared a pteroyl azide by cleaving off the glutamate moiety from folic acid with trifluoroacetic anhydride followed by treatment with hydrazine, and then trifluoroacetic acid with KCN, and tert-butylnitrous acid. The pteroyl azide was then reacted with a polyamidoamine point dendron with either cysteine or cystamine linked to the attachment point via a disulfide bridge and 8 surface fluoresceines. The azide-cysteine conjugate specifically binds to hFR-positive cells; this binding is significantly higher in cells expressing higher levels of the FOL1-R RNA; binding increased with upregulation of the hFR, is inhibited by free folic acid, and is nonspecific when using the alpha amino acid free analog cystamine. This approach broadens the types of functional groups that can be used to target a diagnostic or therapeutic agent via the hFR.
Abstract ID: 161
Macrocyclic metal chelates using 1,4,7,10-tetraazacyclododecane-N-N′-N″-N‴-tetraacetic acid (DOTA) are used to create or enhance contrast in MRI, SPECT, and fluorescence imaging. More recently, metal-DOTA chelates have been conjugated to peptides that bind to specific cell surface receptors, penetrate cell membranes, interact with the extracellular matrix, and alter physiological clearance rates. To date, DOTA and succinimide and isothiocyanato DOTA derivatives have been conjugated to the tertiary amines of peptides, including the N-terminal amine, the side chain amine of lysine, and unnatural amine-derivatized amino acid residues such as p-NH2-phenylalanine. Although these conjugation methods can be used to synthesize some types of DOTA-peptides for molecular imaging, a more robust synthetic methodology is required for conjugating DOTA to any natural or unnatural amino acid side chain, and for conjugating DOTA at any residue position.
To address this unmet need, we have developed a new synthetic methodology to conjugate DOTA directly to Glycine to create DO3A-α-amino-Glycine. We've incorporated tBOC and CBZ protecting groups in our syntheses, and we've used protected DO3A-Glycine to incorporate DOTA at different peptide residue positions through solid-phase peptide syntheses and solution-phase peptide syntheses, respectively. We've synthesized peptides containing both Eu-DO3A-Gly and Gd-DO3A-Gly to demonstrate the utility of this methodology for developing multimodality imaging probes. We've also demonstrated that DOTA-peptides can be selectively detected via the mechanism of MR Chemical Exchange Saturation Transfer. Results of ongoing studies will be presented that incorporate natural or unnatural side chain spacers between the DOTA and the peptide backbone, in order to demonstrate that this synthetic methodology has the flexibility to synthesize any type of DOTA-peptide for molecular imaging.
Abstract ID: 162
Abstract ID: 163
Targeted imaging requires the site specific accumulation of contrast agents. Cell type specific receptors exposed on the surface of cells can be recognised by antibodies. Coupling of contrast agents to antibodies could represent a strategy for cell type specific imaging. However, the presence of a single complex of contrast agent is usually not sufficient for detection by MRI or other imaging techniques. The number of complexes which have to be present per voxel is dependent on the nature of the contrast agent. It is therefore important to have an estimate of the number of exposed receptors on the surface of the target cells prior to starting imaging experiments because the number of contrast agent complexes that could accumulate might be below the detection limit of the instrument. Integrin receptors are considered to be particularly promising targets due to their accessibility from the blood stream and to their involvement in tumour growth. We have estimated the number of the receptors (alpha-v, beta-3, EGFR) on the surface of stable cell lines (U251, U87, A549, A431, THP1) and primary cells (UVEC) by flow cytometry analysis with mono-biotinylated antibodies and biotinylated microbeads as standards. The data obtained indicated that the number of exposed receptors were below the detection limit of MRI with oligomeric Gadolinium chelates that had been synthesised by us. We attempted therefore to use two different types of iron oxide-streptavidin beads complexed with biotinylated antibodies as targeted contrast agents. Adherent cells were labelled with antibody-beads in culture, cells were collected and analysed as agarose phantoms by MRI. Labelled cells appear as dark spots due to the strong negative contrast exerted by the iron oxide beads. We observe a good correlation between the estimated number of receptors and the number of spots obtained with the antibody specific for the corresponding receptor.
Abstract ID: 164
Fluorescence imaging is the most powerful technique currently available for continuous observation of dynamic intracellular processes in living cells. Suitable fluorescence probes are naturally of critical importance for fluorescence imaging, but only a very limited range of biomolecules can currently be visualized because of the lack of flexible design strategies. At present, design is largely empirical. Recently, we demonstrated that the fluorescein molecule, which has been widely employed as a core of fluorescence probes, could be understood as a directly linked electron donor - fluorophore acceptor system. By using the concept of intramolecular photoinduced electron transfer (PeT), we could easily control and anticipate the fluorescence properties of fluorescein derivatives. Further, we found that the purportedly indispensable carboxylic group of fluorescein could be replaced with other substituents, and could develop novel fluorescein derivatives, called TokyoGreens. TokyoGreen dyes had great advantages over traditional fluorescein molecule in some aspect of the development of new probes. By precisely controlling the oxidation potential of the benzene moiety of TokyoGreens, we could construct the first and totally rational design strategy for novel fluorescence probes based on a strict photochemical basis. The value of this approach is exemplified by its application to develop novel fluorescence probes for various reporter enzymes. The first example is a probe for beta-galactosidase, named TG-betaGal. TG-betaGal is far more sensitive than FDG and X-gal, which are so-far known the most sensitive fluorescence probe and the most widely used staining dye, respectively. Furthermore, TG-betaGal is the first membrane-permeable fluorescence probe which enables us to monitor the beta-galactosidase activity in living cells. The second example is a novel, probably most sensitive fluorescence probe for alkaline phosphatase (ALP), named TG-Phos. TG-Phos is the first fluorescein-based fluorescence probe applicable for Western blot analysis, which allows extremely sensitive detection of ALP activity on the blot membrane.
Abstract ID: 165
The contrast agents currently used in clinical MRI protocols act catalyzing the relaxation rate of water protons. The most important class of such systems is represented by Gd(III) chelates. In the last two decades, a huge work has been done aimed at improving the efficacy of these agents by modulating their structural and dynamic features. On the other hand, on the side of MRI instrumentation, the achieved detection sensitivity makes now possible the acquisition of MR images from substrates different from water. In analogy to what done for water, improved MRI acquisition procedures for either endogenous or exogenous substrates may rely on the use of agents that decrease the relaxation times of their NMR-active nuclei. In this contribution, we show that high relaxation enhancements of substrate nuclei can be attained if the Gd(III) chelate form reversible ternary adducts with the molecule of interest.
The obtained results demonstrate that Gd(III)-based probes may offer the opportunity to improve substantially the diagnostic potential of the MRI technique allowing the direct visualization of molecules different from water present at millimolar concentration. This is possible thanks to the huge relaxation enhancement attained when the substrate molecule enters the inner coordination sphere of the Gd(III) ion. Such a relaxation enhancement depends on a number of dynamic and structural features of the ternary complex whose optimization may further decrease the detection threshold for the substrate imaging.
Abstract ID: 166
Abstract ID: 167
NMRD of micelles in blood at 25°C.
CNR of blood containing mixed micelles and GdDTPA. 2A) Small and large micelles in human blood at 1.5T, TSE sequencce, TR=150 ms, TE = 7.7 ms. 2B) Comparison of large micelles and GdDTPA. Concentration of large micelle multiplied by 3 to allow for plotting on same graph.
Abstract ID: 168
Chemical Exchange Saturation Transfer (CEST) is a new method of generating image contrast in MR images. It operates by presaturating the exchangeable protons of a contrast agent. As these protons exchange with the bulk solvent the apparent concentration of bulk solvent observed in the image is reduced and thus image contrast is generated. Some of the most effective agents reported to date are those of paramagnetic lanthanide complexes, known as PARACEST agents. As these agents approach clinical medicine one goal is to increase the sensitivity of the agent, in other words reduce the required dose. In this poster we report the results of our investigations into systems containing multiple hydroxyl ligating groups, that potentially offer a larger number of exchangeable sites, and therefore enhanced sensitivity of the agent. The complex EuCNPHC3+ has been studied and a representative CEST spectrum is shown. Fitting of this spectrum to a six-site exchange model using the Bloch equations allows a vast amount of information to be gathered about the exchange processes in the complex. In particular, the extraordinarily slow rate of exchange of some if these protons (kex ~ 102 s−1) can be calculated.
Abstract ID: 169
The development of targeted imaging agents will enable physicians to detect cancer at earlier stages thus improving the treatment decisions and outcome of cancer patients. We are developing multivalent ligands that will target specific combinations of cell surface epitopes expressed on cancer cells but not on normal cells. In order to improve the specificity of these multivalent ligands we are in the process of designing heterodimers that can simultaneously bind two different receptors types. For proof-of-principle experiments, we have synthesized heterodimers of [Nle4, D-Phe7]-α-melanocyte stimulating hormone (NDP-α-MSH) and cholecystokinin-8 (CCK-8). NDP-α-MSH and CCK-8 are peptide hormones that bind melanocortin receptors (hMC4R) and cholecystokinin receptors (CCK-BR), respectively. These ligands have been linked together via synthetic linear linkers of varying lengths. The linker cores are constructed of dipeptide repeats with PEGylated groups between the linkers and the respective ligands. Ligand binding was evaluated using a lanthanide-based competitive binding assay [1] using cells that express one or both of the cognate receptors (CCK-BRand MC4R for CCK and MSH, respectively). With one receptor, binding is assumed to be monomeric and the presence of both receptors makes dimeric ligand binding possible. Three heterodimers were tested to date with linker lengths of 90, 112 and 148 atoms. The 148 atom linker had the largest effect. Comparing dimeric to monomeric binding, this construct showed a 9-fold increase in affinity at the hMC4R and a 3-fold increase in binding affinity at the CCK-BR. These initial results suggest that these heterodimers are simultaneously binding the melanocortin and CCK receptors, although additional experiments will have to be completed to confirm this observation. If validated, this would be the first demonstration of synthetic hetero-crosslinking of two cell surface receptors.
Poster Session: Multimodal Imaging-Instrumentation
Abstract ID: 170
The purpose of the study is to evaluate the performance characteristics of a new GE eXplore Vista pre-clinical PET scanner - a stationary ring-type “depth-of-interaction” system designed for both high sensitivity and high resolution imaging of rodents. The system sensitivity is measured using a line source placed along the central axis of the system. Image resolution is measured across the entire field-of-view (FOV) using a 0.5 mm Na-22 point source. A FORE+FBP and a 3D iterative OS-EM algorithm are used for image reconstructions. Phantom and rodent studies are also performed to evaluate the overall system performance. Our results show that the system sensitivity are 2.1%, 4% and 6.5 % for the 100–700, 250–700 and 400–700 keV energy windows respectively. At the center of central axis, both tangential and radial resolutions are 1.4 mm with FORE+FBP, while the axial resolution is 1.5 mm, yielding a volume resolution of 2.94 mm3. At 1 and 2 cm off-center in radial direction, the resolutions in all three directions are better than 1.7 mm and 2.1 mm, respectively. The 3D OS-EM algorithm, which incorporates accurate system modeling, significantly improves the reconstructed image resolution and a volume resolution of ~1 mm3 is obtained at the center. For the 250–700 keV energy window, the peak NEC rates are 151 kcps at 140 kBq/ml and 197 kcps at 444 kBq/ml, with scatter fractions of 36.7% and 28.4%, for a rat-size and a mouse-size uniform cylindrical phantom, respectively. In the phantom study using a ultra-high resolution phantom, ‘hot’ rods with diameters down to 1.0 mm can be distinguished, especially in images obtained with the 3D OS-EM. In conclusion, the system provides superior performance characteristics as compared to previous generation of pre-clinical PET scanners and is suitable for high resolution and high throughput rodent imaging to evaluate physiology in vivo.
Abstract ID: 171
A novel small-animal SPECT system called FastSPECT-II, the new version of FastSPECT, has been constructed in our laboratory recently. FastSPECT-II provides list-mode acquisition and more flexibility in choosing magnification, resolution and field of view than FastSPECT. This study was designed to test the capability of FastSPECT-II in detecting ischemia-reperfusion injury in rat hearts by 99mTc-tetrofosmin imaging.
Abstract ID: 172
Our goal is to design and develop a high resolution, high sensitivity multi-ring PET scanner integrated in a 7 Tesla Bruker Biospec small animal magnetic resonance imaging system. In our approach we propose to couple the scintillator elements to position sensitive avalanche photodiodes (PSAPDs) using short bent optical fibers. In this way we would greatly reduce the interference between the two systems without significantly degrading their performance.
The prototype PET module consisted of an 8 × 8 LSO array (crystal size 1.5 × 1.5 × 10 mm3), coupled through 10 cm long straight optical fiber bundles to a 14 × 14 mm2 PSAPD. Two similar modules were positioned 6 cm apart and the following measurements were performed using a Ge-68 point source: timing, energy and spatial resolution. A FWHM timing resolution of 9.2 ns was determined. All 64 crystals were clearly identifiable in the resulting flood histograms and the average energy resolution was 25%. To assess the spatial resolution two capillary tubes with 0.5 mm inner diameter were filled with 50 μCi F-18 and placed between the modules. The flood histograms were used to create look-up-tables that relate coordinates in the histogram, to the crystals of interaction.
As a next step the PET modules and their associated electronics enclosed in an aluminum box, were placed inside the magnet. The LSO crystals were located in the center of the MR scanner field of view outside the RF coil and the above measurements were repeated. In addition, we imaged an MR phantom to assess the effect of the PET modules on the MR data acquisition.
A complete PET system would require 16 modules in a ring around the RF coil for an axial field of view of 1.2 cm. We are currently working on bending the optical fiber bundles and designing the readout and processing electronics for the system.
Abstract ID: 173
High resolution X-ray microtomography (micro-CT) was applied for the detection of emphysema in living mice. In emphysema a permanent enlargement of air-spaces distal to the terminal bronchiole can be observed. For an in vivo analysis of emphysema non-invasive imaging techniques are required. It was shown [1] that micro-CT of the lung area in living mice was possible despite cardiac and respiratory motion. The major aim of the present report study was to analyze whether micro-CT can detect and quantify the extent of emphysema in living mice. Emphysema was induced in C57BL/6J mice by intra-tracheal instillation of different amounts of porcine pancreatic elastase. The presence of emphysema could be detected by micro-CT. Image analysis showed that the degree of emphysema was correlated to the level of elastase treatment. By comparing 3D models the areas of emphysema could be correlated to the volume of the entire lung. This is another advantage of non-invasive imaging as the original shape of the lungs was preserved. These results open future perspectives for the assessment and quantification of emphysema. Longitudinal studies of the pathogenesis and treatment of emphysema in living mice become possible.
Abstract ID: 174
Abstract ID: 175
The objective of the study is to evaluate the performance of x-ray CT integrated with SPECT system (X-SPECT, Gamma Medica, Inc. Northridge, CA) for small animal imaging. The system consists of a CT unit with an x-ray tube (200 μm focal spot) and a 12 cm × 12 cm CMOS detector that provides anatomical CT images for correlating with functional SPECT images. The CT was evaluated for x-ray exposure reproducibility, linearity, beam energy, image uniformity, homogeneity, spatial and contrast resolution. Measurement accuracy of various size objects, radiation exposure for typical scan protocol and modulation transfer function was determined. In the absence of CT numbers, a calibration method was developed to determine CT numbers based on pixel brightness for objects of varying densities. This will aid in quantifying CT images since the current system is used only for localization of isotope uptakes.
Abstract ID: 176
Modern imaging technologies like NMR, PET, CT and Optical Imaging are basic tools for molecular imaging research. Each of them gives limited information and so rises the need to serially perform several experiments in different scanners in a short time interval. During those imaging experiments several physiological parameters have to be controlled and monitored (body temperature, ECG and breathing). But most important is the position of the animal must be the same in all experiments, to ease matching of the different images.
To realize those needs, we developed an animal carrier unit (pic.1) which is fully NMR and PET compatible and is now being adapted for Optical Imaging as well. The system in picture 1 is a mice version which works with NMR and PET scanners with a minimum bore of 120 mm.
The locking system of the carrier unit is the same in all imaging systems, so the animal has to be placed at the unit once and then can be plugged into the different scanners in one session.
Fixation of the mice head is done by ear plugs and bite bar. The body is fixed with a vacuum cushion. Using a rectal temperature control the frontal part of the unit is heated or cooled with water by a dedicated temperature control unit (stability < 0,2°C). Anaesthesia is supplied by a mask in combination with the bite bar. Surplus gas is removed with a vacuum line. ECG is none invasively done by fore pore platin electrodes.
For NMR investigation a 30 mm Helmholtz detector can be added to the setup and removed without interfering with the animal.
Abstract ID: 177
The potential of small animal PET for drug development is immense as it allows for the accurate measurement of the in vivo kinetics of a pharmaceutical. This accuracy is dependent upon the image reconstruction algorithm and the corrections employed. In this work, we evaluate the accuracy in the measurement of the activity concentration in small animal's organs imaged with 18F-FDG-microPET with comparison to bio-distribution data.
Abstract ID: 178
Abstract ID: 179
For a conventional clinical scanner equipped with a multi-pinhole collimator (MPH), an approaches is presented that facilitates the identification of intracerebral target regions. In contrast to clinical scanners most dedicated small-animal SPECT systems do not allow simultaneous data acquisition with two or more radionuclides. Instead, energy windows have to be separately adjusted, and measurements must be performed successively. In contrast, conventional clinical scanners using the novel MPH technique allow the simultaneous data acquisition within different energy windows. This study presents results indicating the feasibility of fusing receptor/transporter images using the DAT ligand 123I-FP-CIT or D2 receptor ligand 123I-IBZM with bone metabolism scans (99mTc-labeled MDP or DPD). These measurements were performed either separately or simultaneously in different non-overlapping energy windows (dual-isotope technique). DAT and D2 receptor blockade was induced by methylphenidate and haloperidol, respectively. Bone metabolism was imaged using the 99mTc-labeled tracers DPD or MDP. Both retrospective image fusion and dual tracer imaging (Figure 1) may prove helpful during evaluation of novel radioligands, whose distribution patterns are yet unknown. Moreover, both methods may facilitate orientation when receptor/transporter binding is to be quantified under pharmacological blockade leading to deliberate reductions of radioligand binding in striatal regions. Thus, quasi-anatomical landmarking provided by synchronized acquisition or by retrospective fusion of 99mTc-DPD scans, may provide a valuable alternative, if a dedicated animal CT or MR scanner is unavailable.
Dual-isotope study of a rat head. Anatomical information is displayed as transversal (left), coronal (middle) and sagittal (left) slices, provided by the bone metabolism marker MDP (first row). The second row shows the corresponding slices of the receptor ligand IBZM, followed by the fusion in row three.
Abstract ID: 180
For clinical applications as well as for biomedical research multimodality imaging like combined positron emission tomography (PET) and computed tomography (CT), revealing functional and morphological information, becomes more and more important. Clinical and preclinical multimodality systems already showed the great benefit of fused image data. In contrast to CT magnetic resonance tomography (MRT) has much better soft tissue contrast and no additional radiation dose. First trials with manually fused images of a stand alone small animal PET and high field MRT systems using fiducial markers were promising and helped elucidate the uncertain interpretation of the functional image information. However, studies involving several hours scan time lead to misaligned image data. Also, the relocation of the animals from one imaging device to the other made often a co-registration, especially of small regions like lymph nodes, impossible.
Thus, combined PET-MRT would be the system of choice especially for biological studies in order to avoid radiation induced changes of the animal model.
We are developing a PET detector based on avalanche photodiodes (APDs) fitting in the limited radial space between gradient coil and RF-coil of a 7 Tesla magnet (Bruker Biospin, Germany). Each detector consists of a 12 × 12 LSO crystal block (CTI-Concorde, Knoxville, USA) coupled to a 3 × 3 APD array (Hamamatsu, Japan). The APDs are read out by an ASIC 9-channel charge sensitive preamplifier (CTI-Concorde, Knoxville, USA). Preamplifier and APD are located on a printed circuit board meeting the space limitation and using only non magnetic parts. First test of the detector electronics showed very encouraging results and proved the feasibility of such a combined scanner for simultaneous imaging. MRI images acquired in a 7 Tesla scanner with the detector in place showed no degradation. Coincidence measurements using more detectors up to a full ring will be conducted to reconstruct first PET images.
Abstract ID: 181
Figure 1 displays 3 out of 32 consecutives sagittal slices (FISP sequence, TE/TR=10/21ms, 430μm in plane, slice thickness 1mm, total acquisition time 32minutes) of a 30g living mouse.
Abstract ID: 182
Micro-CT has the potential to non-invasively study organ structure in rodents with high resolution and fast image acquisition. Technological advances in micro-CT scanners, including the introduction of dynamic, flat-panel scanners, allows the acquisition of volume images in under 8 seconds. However, motion artefacts associated with the normal respiratory motion of the animal may arise when imaging the abdomen or chest cavity due to the high respiratory rates of rodents. To reduce these artefacts and the accompanying loss of spatial resolution, we propose a respiratory gating technique for volume micro-CT imaging of anaesthetized, free-breathing rodents.
A custom-made bed with an embedded pressure chamber was connected to a pressure transducer. Anaesthetized animals were placed in a prone position on the bed with their abdomens located over the chamber. Inhalation motion caused an increase in the chamber pressure, which was converted into a voltage signal by the transducer. Images of the thorax were acquired using a GE Locus Ultra volume micro-CT scanner at 80 kVp, 50 mA and reconstructed with an isotropic voxel spacing of 0.15 mm. During the imaging protocol, the x-ray signal was monitored and recorded, along with the respiratory pressure waveform. Projection images were acquired with up to 10 rotations in less than 1 minute. Respiratory gating was performed retrospectively by reconstructing only the projections that were obtained during the same portion of the respiratory cycle. Gated images in mice and rats exhibit improved definition of the diaphragm boundary and increased conspicuity of the airways and other small structures within the lungs. From these results, we are confident that retrospectively gated volume micro-CT will be a useful tool for imaging the lungs of rodents, as it would be possible to produce images at different respiratory phases enabling dynamic depictions of lung function.
Abstract ID: 183
We have developed a Micro-CT system to provide high-resolution and anatomic information to combine with a microPET R4 system. This micro-CT system consists of a charge coupled device (CCD) detector couple via fiber optical taper to a phosphor screen, an Oxford X-ray tube and a rotational gantry. The Micro-CT system is a cone-beam computed tomography with a cone angle of 12 degrees. In the work, the system performance including system resolution, the scatter-to-primary fluency ratio (SPR), and the effect of X-ray scatter on the contrast in the reconstructed image, were evaluated. Detector response affected by the tube voltages, currents, and exposure time were measured to find the optimized combination for scan(45 keV, 0.5 mA). Two acrylic cylindrical phantoms with 30 and 50 mm diameters were used to simulate mice and rat, respectively. Five lead foils with diameters of 3, 6, 9, 12, and 15 mm were pasted on the surfaces of phantoms for scan individually. The scattered and total counts were measured with and without lead foils, respectively. The amount of scatter at the center was estimated by polynomial-fitting (2nd-order) extrapolation to the 0 mm size of lead foil. Therefore, the SPRs at the center of the cone-beam system are 64.1%, 20.1% for rats, and mice, respectively. The SPRs are reduced to 21.0% for rats, and 11.3% for mice, while the cone angle was narrowed to be 3° by a slit collimator. We have evaluated the spatial resolving power of the micro-CT system using the slant-alit method to calculating the modulation transfer function (MTF) curve. The effects of number of projection view and the interval angle on the MTF curve will be presented at the conference.
Appearance of the INER's Micro-CT scanner.
A coronal slice of a normal mouse (no contrast agent was used).
Scattered counts with/without slit collimator versus various sizes of lead foils.
Abstract ID: 184
Abstract ID: 185
A Micro-PET/CT scanner was developed by combining an in-house Micro-CT and a microPET R4, The PET/CT scanner was constructed to provide accurate image alignment between the PET and CT image volumes, and to benefit both the qualitative and the quantitative applications. The gantry of the Micro-CT was designed to be coaxial with the scanner of microPET R4. It can be moved in and out for the convenience of mounting the Ge-68 point-source holder for PET's calibration. The Micro-CT system consists of an x-ray tube with 4–50 kV voltage, 0–1 mA anode current and 40 μm focal spot size, and a 70 mm GOS/CCD detector with 24 μm pixel size. A CT volume with 512 × 512 × 512 voxels is reconstructed by using Feldkamp algorithm. Two centers of scanners have been calibrated carefully. A lead curtain was added between the two modalities to shield off x-ray effect on PET detectors. The x-ray load on PET detectors was measured to be 7.8% increased without lead curtain between PET and CT. In order to fuse PET and CT images, a 3D co-registration technique has been designed and evaluated. The PET and CT can be operated separately, and the image volumes obtained from two modalities can be registered by the internal spatial transformation function. Phantoms and mice were imaged with both modalities to show the availability of the combined Micro-PET/CT system. The aligned PET and CT images are used both for qualitative and quantitative studies. The combined Micro-PET/CT system has been operated successfully. Results of merging the anatomical and functional images, and the improvement over FDG-PET without CT will be discussed.
A Micro-CT has been developed to combine with the microPET R4 scanner.
A transaxial slice of CT, PET/CT and PET images of a mouse with hepatoma lesions.
Abstract ID: 186
Near-infrared diffuse optical tomography (DOT) provides excellent functional information directly related to tumor patho-physiology (e.g. angiogenesis and blood concentration), which is complementary to structural and functional information provided by conventional imaging, such as MRI. The functional parameters, together with morphological parameters from MR can be suitably combined and correlated to the absolute diagnosis from histopathology.
Fusion of the multimodal datasets will eventually lead to a significant improvement in the sensitivity and specificity of breast cancer detection. Fusion may also allow a priori structural information derived from MRI to be incorporated into the reconstruction of DOT images.
There are very few known systems that combine clinical DOT and MRI data, and many of them physically integrate the DOT device into the MR scanner. We provide a novel technique for registration and fusion of non-concurrent 3D DOT with MR data. This technique will provide a pathway for comparing non-concurrent MRI and DOT data, and will contribute to our long-term goal of providing a significant improvement to the diagnosis and treatment of breast cancer patients by a) improving and enabling optical breast imaging, b) providing combined quantitative and qualitative analysis of structural and functional data, derived from both modalities, and c) advancing the development cycle of a DOT system, which has the potential of increasing the sensitivity and specificity of breast cancer detection through the use of future molecular contrast agents.
The figure below shows initial results for two invasive breast cancer patients: superimposed MRI and DOT (total hemoglobin concentration) datasets (3D renderings, and 2D fused images) are shown before and after registration (breast is compressed axially for DOT imaging, and sagittally for MR imaging).
Abstract ID: 187
We have developed a non-contact Fluorescence Molecular Tomography (FMT) system to detect in-vivo fluorescent signals in bodies of arbitrary shapes making use of a composite model for light propagation in free space and diffusive (highly scattering) media. In order to obtain accurate results, this approach needs a priori knowledge of the 3D boundary. We have thus implemented in our FMT system an approach for measuring the geometry combined with a 3D surface reconstruction algorithm. The complete system (i.e. FMT and surface acquisition setups) consists of a rotating sample holder and a lens coupled Charged Coupled Device (CCD) camera in combination with a fiber coupled laser scanning device, and a custom made homogeneous back illumination system for the surface reconstruction.
The surface reconstruction is obtained by adding all taken projections (shadows) of the objects using an inverse Radon transform and keeping those voxels in space that include all projections. The obtained surface resolution is principally limited by the pixels size of the camera and the number of projections. We here present a detail analysis of the dependence of surface resolution with number of projections, pixel size, and binned detector area. Finally, by making use of the obtained 3D surface we contrast the results obtained from a non-contact CCD measurement such as those employed in our FMT setup with an in-situ fiber measurement.
A 3D surface reconstruction of a mouse, A) the triangulated mesh that defines the surface, B) the surface rendered with the white light reflection images.
Abstract ID: 188
Anatomical variable phenotype in mutant mice could be assessed ex vivo. By adequate use of μCT, we are able to analyze non-invasively abnormalities in living subjects. The Cdx2 heterozygote mutants, the model chosen, exhibit skeletal abnormalities as well as early development of intestinal polyps [1].
3D reconstruction of the ribs: left, normal animal, right; Cdx2 mutant mouse.
Polyps in the caecum, as described in reference 1 (large arrow); small flat one (small arrow).
Abstract ID: 189
Tumor-induced neovascularization or angiogenesis plays an important role in solid tumor progression and metastasis. This process has been shown to be dependent upon expression of receptors for vascular endothelial growth factor (VEGF). The following study was undertaken to identify suitable fluorescent labels for studying expression of VEGF receptors in vivo, with the ultimate goal of identifying strategies for inhibiting the VEGF signaling pathway. Xenografts were established in female BALB/c mice by injecting 2 × 104 4T 1luc cells subcutaneously on the mouse backs. For imaging, mice were anesthetized and the area to be imaged was shaved and further treated with Nair® to completely remove all hair. The immobilized animals were placed on the platen of the Kodak IS4000MM with the tumor facing down toward the camera. VEGF receptor expression was evaluated using several VEGF-probes, labeled with Near IR reagents including Cy5.5 (GE Healthcare), that were introduced via tail vein injection. Images were captured using a 10 sec exposure and either 625nm excitation/700nm wide angle emission filtering or 755nm excitation/830nm wide angle emission filtering. X-ray imaging was performed in the same instrument using a 30 Kev, 150μA source to generate a digital image that was used as a base to overlay fluorescent images. Results demonstrate that VEGF receptors associated with subcutaneous primary solid tumors can be easily detected in vivo at one week post injection using VEGF labeled with NearIR reagents (Figure 1A). Localization of metabolized label could also be identified in the liver (Figure 1B).
Abstract ID: 190
Abstract ID: 191
Following the success of PET-CT scanners, there is increasing interest in developing a single scanner that combines PET with anatomical and functional/physiological images from MRI (Shao 1997, Grazioso 2005, Shaw 2005, Pichler 2004, Mackewn 2004). We are ‘fusing’ hardware for simultaneous PET-MRI.
Our approach locates conventional, magnetic-field sensitive photomultiplier tubes (PMT's) in a region of low (~0.02T) magnetic field via fibre optic bundles through a 80mm ‘split’ in the MRI system (Figure 1). The PMT's are enclosed in magnetic shields which reduce the magnetic field at the PMT below the threshold (~0.1mT) for standard operation.
The design for a novel 1 Tesla, ‘split’-coil MRI magnet was created using a genetic algorithm approach (Shaw 2002). The magnet design was refined in collaboration with Magnex Scientific before construction.
High resolution images with voxel dimensions 200umx200umx 1.2mm have been acquired using the magnet and conventional MRI hardware (Shaw 2005). We are now developing a novel, ‘split’ gradient set to enhance MRI performance and accommodate a conventional microPET® system.
In parallel, we are testing our design concept with single PET detector measurements and a coincidence-mode, dual-detector set-up. Specifically, we will test the influence of the MRI (static field, gradients, RF), optical fibre geometry (length, bending, cut-angles), PMT location and shielding on the detector crystal map (Figure 2 for 10cm (left) and 1.2m optical fibres), energy spectra, sensitivity and image resolution.
Information from these measurements will feedback into the design of the complete 4-ring FOCUS microPET®-MRI system.
Abstract ID: 192
Genomic and proteomic analyses of human and mouse tissue is increasingly used to 1) identify novel molecular imaging targets, 2) credential existing targets in large cohorts and 3) phenotype diseases more accurately than is currently possible (often in combination with imaging). Unfortunately, there are currently no IT tools that enable storage and analysis of imaging and genomic data in a highly integrated fashion. A medical image IT content management system, MIPortal, has previously been designed by us to store, share and analyze DICOM imaging data. Here we describe the extension of MIPortal to integrate genomic data. Gene expression data fit into the existing MIPortal hierarchy and are stored alongside imaging and other experimental data, to enable cross searching. The system provides a common interface for retrieval of genomic data from both public and private sources. Storage is currently based on standards for data object model (MAGE-OM, Microarray Gene Expression - Object Model) and for data exchange format (MAGE-ML, Microarray Gene Expression - Markup Language). Gene expression is analyzed using a suite of common tools, thus enabling a meaningful comparison of data from different sources. The above described tools are useful to perform gene- and experiment-centric analysis. The graphical user interface for data storage and sharing is identical for imaging and genomic data. The system provides secure access based on user privileges. Overall, the developed IT infrastructure provides a highly integrated workspace for storing and cross analyzing imaging and genomic data. The tool should be useful for phenotyping, biomarker discovery and validation, and drug development.
Abstract ID: 193
We have upgraded a dual-headed Picker Prism 2000 with HiSPECT, a high-resolution high-sensitivity multiplexing multi-pinhole SPECT module with dedicated reconstruction software. The system was initially intended exclusively for small-animal studies, though given the increase in sensitivity and resolution gains of this approach we recently designed and implemented multi-pinhole apertures for clinical studies of hands and feet. Specifically, we will report on our initial study in which we have imaged Patients with osteoarthritis, varying stages of rheumatoid arthritis as well as controls. On average, twenty projections of the hand were acquired three hours after 550 MBq of technetium-99m labelled dicarboxy propane diphosphonate (Tc99m-DPD) were administered intravenously to the patient. In addition to the SPECT measurements, all patients received conventional radiographs and three-phase bone scintigraphy of the relevant hand. Additionally, MRI of the appropriate hand was performed on two of the rheumatoid arthritis patients. The HiSPECT images were evaluated in each patient category and the results were correlated with and validated against the other imaging modalities. Patient hand position was fixed using a specially developed plexi-glass hand holder, which in turn allowed us to fuse the MR and HiSPECT images. The first clinical application of the HiSPECT technique indicates that there exists an increase in bone metabolism information as compared to conventional bone scintigraphy. Figures: Imaging results collected from a patient with early rheumatoid arthritis using three different imaging modalities: Planar X-ray (A), coronal STIR-sequence (MRI) (B) and multi-pinhole SPECT fused with MRI. The image fusion shows areas of increased bone metabolism in the 2nd and 3rdinterphalangeal joints which show no pathologies in bone marrow signal in the MR image.
Abstract ID: 194
More than 50 studies can be displayed simultaneously in orthogonal views : true-color-visualization is slice-oriented, optimized for real-time-interaction and supports both multi-monitor-setups and small displays (projector for live presentations).
We believe that with additional preprocessing (manual or automatic masking) we can also register certain PET and MRI data sets from animal scanners with a similar technique.
Landmark-based registration has been added: the plugin registers two images by a least-square-fit of two sets of user defined points: a Generalized Procrustes Problem we solve with Singular Value Decomposition.
Manual co-registration is also supported, online fusion (of up to fourimages) and contour rendering is instrumental for this technique and for evaluation of automatic registration results.
Abstract ID: 195
A simple, inexpensive, and multifunctional imaging technique for imaging metastatic lesions in small animal will be very useful for noninvasive molecular characterization of metastasis and for assessing responses to molecularly targeted therapies. Near-infrared (NIR) optical imaging offers unique advantages due to its high sensitivity, versatile fluorescent probe design, and the ability to penetrate the skull and bone. Here we report the in vivo NIR optical imaging of human melanoma grown intracranially and human prostate cancer grown intratibially in nude mice using Kodak Image Station multimodal imaging system. To imaging melanoma brain metastasis, mice were injected with NIR dyes targeted to integrin αvβ3 and matrix metalloproteinases 8 days after tumor inoculation. To image prostate bone metastasis, NIR dyes were injected 3 days after tumor inoculation. NIR images were acquired 24 hours after intravenous injection of each NIR dye at a dose 15 nmol per mouse. In vivo NIR imaging showed selective accumulation of the contrast agents in the tumor after the contrast injection. The merged images from NIR and high-resolution X-ray imaging clearly pinpointed the primary tumor at the site of tumor cell inoculation. Histological examination confirmed the presence of tumors that were less than 1 mm in size. Our data demonstrate that multimodal imaging technique combining NIR and X-ray is a particularly suitable for molecular imaging of tumor at sites involving the brain and the bone in small animals (Supported by NIH grants R01 EB00174, Prostate Cancer SPORE Career Development Award CA90271).
Abstract ID: 196
At present the application of radiation to small animal models is limited to planar fields shaped using lead blocks. The ability to apply therapeutic doses of radiation to specific volumes within small animals would facilitate studies of radiotherapy in orthotopic and spontaneous tumor models, as well as investigations using radiation-activatable genes. Towards this end, we have built a variable aperture hexagonal collimator that can be installed on the eXplore RS150 microCT scanner (GE Medical Systems, Milwaukee, WI). The collimator is similar in structure to a camera iris, consisting of six hexagonal blocks mounted in a rigid frame. By sliding the blocks using embedded pins driven by a motor-controlled rotating plate, the size of the hexagonal aperture may be adjusted. In its fully open state the collimator aperture measures 6 cm, allowing full passage of the RS150 X-ray beam. Correspondingly, when fully closed the central portion of the collimator completely blocks the beam. Transmission through a single collimator plate is shown in Figure 1. Some radiation bleed between the blocks is noticable in Figure 1f, with an estimated transmission of 10%. This will be improved by cutting new blocks with interior angles aligned against the X-ray beam. Two parallel-driven collimator plates, offset by 30°, will be used in the final collimator apparatus to achieve a more circular beam profile. The weight of the complete collimator will be approximately 20 pounds, and will be installed on the gantry of the RS150 scanner with appropriate counterbalancing to ensure rotational accuracy. This unit will allow clinically-relevant conformal radiotherapy to be delivered to small animal models.
X-ray transmission through the collimator for several aperture sizes.
Abstract ID: 197
The unique capabilities of the Kodak Image Station have enabled the merging of differing modes of imaging small animals in vivo. A number of laboratories have successfully applied the Image Station to the different modes of planar optical imaging (bright-field, fluorescence and luminescence) over the broad spectral range of 400 to 850 nm. The recently released version of the Image Station (IS In Vivo EX) has extended the system's application to the practice of co-registering/merging X-ray and isotopic images with optical images, and doing so with enhanced sensitivity and resolution, and animal management accessories.
The necessity of conserving the system speed (for luminescence and isotopic imaging) and spatial resolution (for anatomic X-ray imaging) with the flexibility of variable magnification was achieved with the development of a 2K×2K cooled CCD and a 10×-zoom lens. A system speed consistent with the significant measure of thousands of emitted photons/sec./mm2 of object is possible and necessary for animal luminescence and isotope imaging. For X-ray imaging, considerable attention was paid to optimizing a sliding phosphor screen that matches the resolution enabled by a 50 μm micro-focus X-ray system from 10–35 KVP.
The optical configuration of the Image Station, together with the application software, permits a convenient quantification of emitted light that is reasonably constant with image magnification. The 16-bit linear dynamic range with extended time reciprocity inherent to a cooled CCD is an extremely reliable/reproducible system, whose full potential is now extended to X-ray and isotope imaging. Isotopes have included beta-emitters (14C and 32P) and X, gamma-emitters 99Tc, 111In, and 18Fl (FDG-PET).
The numerous demonstrated capabilities of the system include such diverse applications as the kinetic bio-distribution of isotopes and the skeletal localization of fluorescently-tagged pre-metastatic lesions.
Abstract ID: 198
The effect of each variable on system performance
Abstract ID: 199
Pinhole SPECT is an important molecular imaging technique for in vivo antibody imaging. One of the most favorable radioisotopes for labeling antibodies is indium-111. However, indium-111 emits high energy gamma rays (171 keV and 245 keV) resulting in a large penetration depth in the collimator material which can reduce contrast and spatial resolution. Therefore, we have developed a small animal pinhole camera for indium-111 imaging using a 20 × 20 cm2 cadmium zinc telluride detector enclosed in a lead collimator with 2.5 cm wall thickness. The collimator is equipped with an interchangeable pinhole which is machined from tungsten and has an opening angle of 65° providing low septal penetration at 245 keV.
Currently, we are evaluating the performance of the pinhole camera. We have used a 0.5 mm pinhole to acquire 80 projections of a hot rod phantom with an average of 2 × 105 counts. In addition, we have performed simulations with a ray tracing program using the earlier described parameters to determine the best spatial resolution of the pinhole camera. Our experimental data (Figure 1A) shows that the spatial resolution of our pinhole cameras is better than 1.2 mm for both In energy lines. This is supported by our simulations (Figure 1B) indicating that a spatial resolution of 0.8 mm can be achieved.
In conclusion, we have successfully built a pinhole camera for indium-111 imaging, which will be incorporated in a SPECT/CT imaging system currently under construction in our laboratory.
Reconstructed experimental (A) and simulated (B) projection data of a hot rod phantom to evaluate the system spatial resolution that can be achieved for imaging indium-111 with our pinhole SPECT system.
Abstract ID: 200
The purpose of the project is to develop high performance pinhole SPECT imaging techniques that provide high resolution, contrast, detection efficiency and quantitative accuracy for cardiovascular imaging in small animals. We developed calibration methods to determine system geometric misalignments and incorporated them into image reconstruction algorithms for accurate and artifact-free 3D pinhole image reconstructions. Three-dimensional (3D) iterative pinhole reconstruction methods were developed that incorporate accurate attenuation correction using transmission CT images and pinhole collimator-detector response correction for improved quantitative accuracy and image quality. We developed specially designed multi-pinhole collimation and 3D iterative multi-pinhole reconstruction methods for improved detection efficiency with minimal image degradation and artifact generation. Also, we implemented cardiac gating techniques for small animals and developed 4D image reconstruction methods for an increased number of timing gates with improved temporal resolution while maintaining low reconstructed image noise. We demonstrated that accurate and artifact-free images can be obtained with our system calibration methods. The attenuation and pinhole collimator-detector correction techniques allowed absolute quantitative accuracy to be within 10% in pinhole SPECT imaging of rodents and improved spatial resolution from 1.1 mm (without correction) to 0.75 mm (with correction) in mouse imaging using a 1 mm pinhole aperture. A 3.7-fold and a 4.7-fold increase in detection efficiency were realized using specially designed 4- and 5-pinhole collimators for a Gamma Medica X-SPECT system with minimal resolution degradation and reconstruction image artifacts. The cardiac gating technique and 4D image reconstruction methods reduced image blurring due to cardiac motion. The high performance pinhole SPECT imaging techniques were further demonstrated in phantom studies, a myocardial perfusion study using Tc-99m Sestamibi, and an atherosclerotic plaque study using Tc-99m Annexin V in mice. We conclude that high quality cardiovascular imaging can be achieved using the high performance pinhole SPECT imaging techniques.
Abstract ID: 201
Multiple pinhole SPECT has the potential to improve the detection efficiency over conventional small animal radionuclide imaging. These methods can produce images in studies in which both the image acquisition time and the injected radioactivity are reduced. They therefore have significant promise for tumor detection and radionuclide therapy studies were radiation dose is a concern. We propose a method for small animal SPECT that uses a rotating array of multiple detectors, each with a multipinhole collimator to acquire projection data from an animal lying along the rotational axis. Each detector has a different pinhole configuration designed to produce projection data that are mathematically independent of one another. The array can be rotated to provide complete angular sampling of the object with each detector and in a way that allows equivalent single pinhole Radon projections can be calculated. The resulting projections then can be reconstructed using a conventional ML-EM iterative algorithm. We have conducted numerical simulations to validate the method and to analyze its sensitivity and stability under noisy conditions. Our simulation study assumed 5 detectors each having a collimator with a 5 × 5 array of 1.5 mm diameter pinholes (left figure). This method was tested by simulating multipinhole SPECT imaging of a 2 cm diameter hot rod phantom (middle figure) with target to background ratio of 2:1, and with lesion sizes of 0.5, 1.5, 2.25, and 5.0 mm. The proposed multipinhole system has a nominal sensitivity of approximately two orders of magnitude higher than a single 1.5 mm diameter pinhole collimator for a mouse-sized reconstruction volume (i.e., 2.5 cm diameter, 6 cm length).
Poster Session: Biological Molecular Imaging, Part 2
Abstract ID: 202
The aim of this in vitro study was the evaluation of the Drosophila melanogaster multisubstrate deoxyribonucleoside kinase (DM-dNK) as a new potential suicide and in vivo reporter gene.
Abstract ID: 203
Recently we have found that clodronate can decrease proliferation of many thyroid cancer cell lines including SW579, WRO and TT cells. However this drug has not been tested on papillary thyroid cancer (PTC) cells which are the major thyroid carcinoma. In this study, we further compared the drug potency of clodronate on PTC cells, CG3 cell line and primary culture from patients with PTC. The calcium signaling and detail mechanism caused by clodronate were also investigated in this study. Although we did not find any change in cell progression of SW579 cells when treated with clodronate, we compared this cell cycle differences between cells tested in this study. The F-actin cytoskeleton of SW579 and PTC cells was exam both by confocal laser scanning (CLS) and total internal fluorescence reflection (TIR) illumination methods before and after clodronate treatment. Only the one located near cell surface of clodronate-treated SW579 and PTC cells was found to decrease the amount. We used two mitochondrial indicators, rhod-2 and TMRE, to measure the physiological status of mitochondria within clodronate-treated thyroid cancer cells. Similar to the results in F-actin, the activity of mitochondria both on calcium uptake and membrane potential was shown to be affected only on clodronate-sensitive thyroid cancer cells. Furthermore, we injected GFP labled thyroid cancer cells into mouse thyroid gland for in vivo monitoring and also treated animals with clodronate and other drugs. Together with results in this study, the mechanism of clodronate to anti-PTC cells was found and it's clinical application might worth to be further noticed.
Abstract ID: 204
Abstract ID: 205
Various adjutants are administered along with cancer DNA vaccine to enhance immune response. It has been reported that cytokines such as MIP1a and Flt3L were helpful to recruit and mature antigen presenting cells to DNA vaccine injection sites. In this study, we visualized gene expression patterns of DNA vaccine and its protective effects against Fluc expressing CT26 (CT26/Fluc) in mice immunized with Fluc plasmid or Fluc plus cytokine expressing plasmid co-injection.
We intradermally immunized Balb/c mice using following plasmid; pcDNA, pcDNA-Fluc and pcDNA-Fluc plus pcDNA-MIP1α plus pcDNA-Flt3L plasmid. The bioluminescence images were serially obtained using IVIS system (Xenogen Technology) at 1, 2, 3, 7 days after DNA injection. Following last immunization, each mice were subcutaneously challenged with 1 × 105 CT26/Fluc and total cells of draining lymph nodes were prepared for flow cytometry assay using specific CD8 and IFN-γ antibody (BD pharmingen).
Tumor growth of each group was monitored with bioluminescence images and measured with caliper at 1, 7, 12, 19 days after tumor challenge.
Bioluminescence images were detected in each group and maintained at injection site until 3 days. However, bioluminescent signals from Fluc plus cytokine co-injection group were not detected at injection site on 7 days. FACS analysis showed increased IFN-γ secreting CD8+ T cell of Fluc group than both pcDNA and Fluc plus cytokine co-injection group; pcDNA(5.40±0.19%), Fluc (7.58±0.52%), Fluc plus cytokine co-injection (5.42±0.56%) group. We could also detect decreased bioluminescence signals or CT26/Fluc tumor retardation in Fluc vaccination but not in pcDNA or Fluc plus cytokine plasmid co-injection group.
Optical imaging using Fluc could be useful for monitoring not only the gene expression pattern but also protective effect of cancer DNA vaccines in living animals. Furthermore, Fluc DNA could be used as the alternative antigen in substitute of tumor-specific antigen, for DNA vaccine development and protocol design.
Abstract ID: 206
Abstract ID: 207
A) A gradient-echo image from WHHL rabbit thigh B) H MR spectrum from biceps femoris muscle of a normal WHHL rabbit.
IMCL concentrations (mmol/kg w.w.) from normal and diabetic NZW, and normal and diabetic WHHL rabbits.
Abstract ID: 208
Abstract ID: 209
Animal thrombosis models reflecting the clinical setting can be of great help to guide the development of novel anti-thrombotic and thrombolytic compounds to treat and prevent thrombus formation. We aimed to set up an imaging model for real-time detection of thrombus formation and growth in large arteries.
Thrombus formation in mouse carotid artery was induced by a mechanical injury and monitored in real-time by optical imaging of platelets that were preloaded with the fluorescent Acridine-Red in vivo.
This imaging platform provided detailed information about the spatio-temporal characteristics of platelet adhesion and aggregation at the injured vessel wall.
Sequential imaging showed that platelets started to accumulate 8 to 20 min after injury. These findings were confirmed by histological evaluation. Cross-sections of the brain showed significant thrombosis at the brain, at the same side as the injured vessel.
Evaluation of thrombus formation in response to pretreatment with Acetylsalycic acid (25mg/kg) showed a 11% final reduction of thrombus size and a slight prolongation of the tail bleeing time. Intravenous injection of streptokinase in occluded carotid vessels resulted in a gradual, dose-dependent resolution of the thrombus, up to 28% reduction of the thrombus size (Figure 1). Pretreatment with the experimental compound P4pal10 (PAR-4 receptor inhibitor), resulted in a significant 39% reduction of thrombus growth (p<0.05) and markedly delayed the tail bleeding time (600±360 sec vs. 102± 21 sec in controls, n=5, P<0.05).
Hence, we report a novel approach for real time rapid evaluation and analysis of thrombus formation in large arteries.
Abstract ID: 210
Imaging of tumors and metastasis is a prerequisite for successful treatment. Peptides have many advantages which makes them good imaging tracers. New peptides need to be discovered to promote tumor targeting modalites.
A peptide that binds to MCF7 cells and FRO82–2 cells could be identified through phage display. The 111In labeled peptide showed strong binding to MCF7 and FRO82–2 cells, which shows a maximum after 3–4 h and is maintained almost unchanged for more than 24 h. Benign cells like HUVEC showed no binding. The binding to MCF7 and FRO82–2 cells was inhibited to more than 95 % by the addition of unlabeled peptide, and the competition decreases with decreasing competitor concentration. BSA did not disturb the binding. Binding experiments also showed internalization of 111In-FROP-1-DOTA starting after 1h. Incubation of the peptide in human serum showed that the peptide is stable for 30 min and then degradation started slowly. Biodistribution studies with mice bearing human FRO82–2 thyroid tumors showed a strong enrichment in the tumor which is maintained for 45 min before the peptide is most likely degraded. The liver values indicate degradation via liver.
Thanks to its binding capacity the peptide is a promising candidate for the development of a tumor specific peptide, which can be used for tumor imaging. Experiments that will identify which of the 12 amino acids is important for binding and to elucidate the secondary structure are currently being performed. This will allow the development of improved peptides for tumor targeting and imaging.
Abstract ID: 211
Prostate cancer is one of the most frequent cancer in males and a leading cause of death from cancer worldwide. Peptides as tracers with high sensitivity and specificity for prostate tumor cells would be beneficial for diagnosis and therapy.
The peptide DUP-1 (FRPNRAQDYNTN) showed in in vitro binding studies binding to DU-145 and PC-3 prostate tumor cells, while the peptide did not bind to primary endothelial cells HUVEC. Biodistribution studies using radiolabeled DUP-1 with s.c. transplanted DU-145 and PC-3 cells in nude mice revealed the targeting potential of DUP-1. A high accumulation rate of up to 7 % of the injected does per gram tumor tissue could be observed, lower concentrations were observed in all of the organs except for the kidneys, from where the peptide is rapidly cleared. The addition of a DOTA at the C-terminus to allow labeling the 111-indium increased the in vitro binding dramatically up to 100 % of the % applied dose/106 cells after 2 h and was maintained for several hours. Competition of the DUP-1-DOTA by DUP-1 was unchanged. Stability experiments demonstrated that DUP-1 (without the DOTA) was degraded rapidly in human serum. Several modification of DUP-1 have been performed to investigate binding and stability. The circularization of the peptide via two cysteine removed most binding activity, as did the substition of all L-amino acids with D-amino acids. The exchange of the N-terminal amino acids L-phenylalanine with a D-phenylalanine was sufficient to prevent the peptide from binding while the addition of a D-tyrosine at the C-terminus did not disturb the binding.
The effect of the DOTA on stability and in vivo biodistribution will be the next steps towards the use of DUP-1 for tumor imaging.
Abstract ID: 212
Abstract ID: 213
Reports of successful tumor targeting by radiolabeled antisense DNAs are increasingly common, but further studies are needed to establish the mechanisms involved. In particular, an explanation is needed for specific accumulations of about 106 antisense DNAs per cell over 10 h in culture that are much higher than expectations based on steady-state mRNA levels. This investigation used in situ transcription and immunofluescence to confirm antisense targeting and intracellular migration and a radioactivity method to estimate transcription rate. As before, the antisense phosphorothioate DNA was against the mdr1 mRNA coding for the multidrug resistance protein Pgp in KB-G2 (Pgp + +) and KB-31 (Pgp+) carcinoma cells while the sense DNA was used as control. Direct evidence for hybridization of antisense DNA to mRNA was obtained by in situ transcription showing cDNA elongation in KB-G2 cells exposed to antisense but not sense DNA. Thus, the antisense DNA served as an intracellular PCR primer and therefore must have hybridized to mdr1 mRNA. Besides showing higher accumulations of antisense vs sense DNAs in KB-G2 cells, immunofluorescence staining showed migration of only the former to the nucleus To estimate transcription rate, 99mTc-antisense DNA was used as a probe along with 99mTc-sense DNA as control on total mRNA extracted by polyT affinity chromatography from cells previously saturated with unlabeled antisense DNA and, as control, unlabeled sense DNA. Compared to controls, mdr1 mRNA levels were found to be initially low following saturation and to recover at about 2,000 copies per min per cell, or about 106 in 10 h. The higher than expected cellular accumulations of the antisense DNA are therefore explainable if the target is pre-mRNAs present at high concentrations in the nucleus rather than mature mRNAs in the cytoplasm. Furthermore, the migration to the nucleus could then be explained by simple diffusion to higher concentrations.
Abstract ID: 214
Gene therapy requires tight control of both spatial and temporal expression of the therapeutic transgene. Heat-shock promoters, especially the human hsp70 promoters, have been proposed for gene therapy strategies because they are both heat-inducible and efficient. A noninvasive physical approach which permits local temperature control was developed in our laboratory. Automatic execution of a predefined temperature-time evolution is performed using MRI-guided focused ultrasound (MRI-FUS) with real-time feedback control in a whole-body clinical MRI system (1.5T). In the present work, this procedure was used for the control of expression of a gene driven by an hsp70B promoter following a preliminary account (Guilhon et al., 2003). A stable clonal modified C6 glioma cell line was used to induce subcutaneously implanted tumors in young rats (age 3-weeks). The cell line was engineered in such a way to contain the green fluorescent protein (GFP) gene under the control of the human hsp70B promoter. In vivo, selective areas within the generated tumors were submitted to a 5-min weak and constant temperature elevation using the MRI-FUS system. Treated animals were sacrificed 6 h following the heating procedure. Tumors were resected, snap-frozen and cut into 20-μm slices for fluorescence imaging. Results show that cells in heated region are alive. Fluorescence image analysis show strong local induction of GFP expression in regions heated above 43°C with a strong decrease in expression levels in regions with lower temperatures. A good correlation was found between MRI-temperature maps and histological pictures of GFP expression. Guilhon E, Quesson B, Moraud-Gaudry F, de Verneuil H, Canioni P, Salomir R, Voisin P, Moonen CT, 2003. Image-guided control of transgene expression based on local hyperthermia. Mol Imaging 2 (1); 11–17.
Abstract ID: 215
Docosahexaenoic acid (DHA), a saturated fatty acid, recently found is a specifically selective ligand for retinoid acid X receptor in which activation can increases the survival rate of dopamine neurons. In this study, we reported that DHA can up-regulate the RNA expressions of thyrosine hydroxylase (TH) and nurr1 in human neural stem cells (hNSCs). To further investigate the role of DHA in dopaminergic-lineaged differentiation of hNSCs, green fluorecein gene (GFP) was retovirally transduced into hNSCs. Under in vitro GFP-labeling and immunofluorescein assays, DHA-treated hNSCs could increase 3-fold number of TH-positive neurons than those without DHA treatment. DHA-treated hNSCs can further be in vivo differentiated into TH-positive cells displaying GFP activity and functionally recovery in Parkinson's disease mice. These in vitro and in vivo GFP imaging systems provide a new imaging modality for understanding the differentiation process and the effective expression in stem cell study.
Abstract ID: 216
Abstract ID: 217
Abstract ID: 218
Abstract ID: 219
Magnetic cell labelling by superparamagnetic iron oxide nano-particles (SPIO/USPIO) is a pre-requisite for MRI-based cell tracking in vivo. After labelling quantification of the iron content per cell shows the success of the procedure. This is necessary to detect the cells by MRI after re-infusion in vivo. Currently, atomic absorption spectrometry (AAS) is mainly used to quantify iron. The present investigation has been undertaken to quantify the iron content by spectrophotometry.
Peripheral mononuclear cells were separated from whole-blood samples of healthy donors by density-gradient-centrifugation. 1–2 × 106 cells/mL were cultivated for 2 hours under standard conditions in the presence of 140μg/mL SPIO or 140μg/mL USPIO. The labelled cells were counted by fluorescence flow cytometry technique and a semiconductor laser (Sysmex XE-2100). Then the samples were stained by potassium ferrocyanide (K4Fe(CN)6·3H2O), and the absorption of the colored complex has been measured by spectrophotometry (Lambda 12 spectrophotometer, Perkin Elmer). Additionally, iron quantification has been done by NMR relaxometry on a MiniSpec PC100 (Bruker), on a whole body 3T system (Achieva Intera, Philips), and by SpectrAA-10Plus AAspectrometer (Varian).
The absorption spectra of SPIO and USPIO were different: the smaller the diameter of the particles the lower the wavelength (nm) of the peaks. The standard curve generated on the basis of 10 samples with known iron concentrations (range 0.5 to 10μg Fe/ml) was linear. Kinetic study done by spectrophotometry showed after an initial equilibration constant values from 20 to 60 min. The iron content per cell has been calculated after measurement of the absorption by a standard curve. Comparing the results of spectrophotometry with either T2 or AAS values resulted in significant correlations.
Our results suggest that quantification of iron after Prussian blue reaction is possible by spectrophotometry, and comparable to NMR relaxometry, and AAS. The method is user-friendly, cost saving, and can be performed within few hours.
Abstract ID: 220
We sought to characterize the sensitivity of information-theory-based signal receivers versus conventional signal power analyses for delineation of early neovascularization in the ears of transgenic mice, which is driven by the papilloma virus.
This study was funded by NIH EB002168, HL042950, and CO-27031. The research was carried out at the Washington University School of Medicine.
Abstract ID: 221
The use of high-relaxivity, intracellular contrast agents has enabled MRI monitoring of cell migration through and homing to various tissues. Here we show that MRI can detect single cells, in vivo, homing to tissue, following cell labeling and transplantation. Primary mouse hepatocytes were double-labeled with green-fluorescent, 1.63-micron diameter iron-oxide particles (MPIOs) and red-fluorescent endosomal labeling dye, and injected into the spleens of recipient mice. This is a common hepatocyte transplantation paradigm and one in which hepatocytes migrate to the liver as single cells. More than 50 MPIOs were counted in most cells, and at 1.1 pg iron/MPIO, cells were labeled with >50 pg iron. One month following transplant, gradient-echo MRI at 100 × 100 × 300 microns was performed, followed by harvesting of the tissue for histological analysis. The experimental livers showed dark, punctate contrast (Figure 1A), while livers from uninjected mice showed no susceptibility-induced dark contrast (Figure 1B). Livers from control experiments (dead labeled cells, free particles, labeled fibroblasts) all showed grainy contrast, indicative of isolated, single MPIOs, not grafted cells. Confocal fluorescence microscopy co-localized green-fluorescent MPIOs and red-fluorescent cell tracker in single, isolated grafted cells in the experimental livers, which was corroborated by iron-stained histology (Figure 1C,D). This can only happen if transplanted cells graft to and survive in the liver the full one month experimental duration. Measured distances between single cells in histology were consistent with measurements of contrast spots obtained from MRI. Histological analysis of control livers showed only scattered, isolated MPIOs, with no co-localized red- or green-fluorescence. Lastly, comparisons of the sizes of the contrast regions in the experimental liver samples with an agarose phantom containing labeled hepatocytes were identical, and averaged 7.5 pixels/cell.
A) MRI of experimental liver showing punctate dark constrast, B) MRI of control liver showing homogeneous signal intensity, C) Co-localized red- and green-fluorescence from single, grafted, transplanted cell in histological section, D) Prussian Blue iron-stained histology of isolated, grafted cell.
Poster Session: Novel Probes and Activation Strategies, Part 2
Abstract ID: 222
An activatable nanoparticle probes that is specific for caspase-3 was designed to obtain the image of apoptosis. Caspase-3 is known as one of the major enzyme that is activated during the process of apoptosis. A NIR cyanine fluorochrome was served as an optical reporter probe and was attached to the amino terminal of a 9 amino acid peptide sequence that showed specificity for caspase-3. The peptide-cyanine dye conjugate was subsequently attached to polyethylene imine-FITC nanoparticles for efficient cellular uptake. The close spatial proximity of the multiple fluorochromes resulted in an efficient quenching of fluorescence in the bound state. The quenched nanoparticle probes were then cleaved and activated by recombinant human caspase-3. A 10-fold signal amplification was observed post cleavage in the presence of caspase-3 in vitro. However, these polymeric nanoparticles was not be activated by caspase-3 in the presence of caspase-3 inhibitor, DEVD-CHO. In live cell imaging system, we confirmed that efficient cell permeation and activation of the polymeric nanoparticle probes in Trail treated HeLa cells. In summary, a novel quenched polymeric nanoparticle probes were activated by recombinant caspase 3 as well as apoptotic cells, in vitro.
Abstract ID: 223
Herpes simplex virus thymidine kinase has been previously used as PET reporter gene with [I-124]FIAU(1-(2′-deoxy-2′-fluoro-1-β-D-arabinofuranosyl)-5-iodouracil). We have developed a synthesis of FIAU utilizing [F-18]. Here, we report HSV-tk gene expression imaging with [F-18] FIAU, which is a more desirable isotope than [I-124] because it is more widely available, has a more favorable energy profile, and superior dosimetry. Also, we account for activity uptake in HSV1-tk tumors due to [F-18] FAU, a de-iodinated metabolite of [F-18] FIAU.
CD-1 nu/nu mice with subcutaneous MH3924A and MH3924A-tk8 xenografts on opposite flanks were used for biodistribution and imaging studies. Ten mice were injected IV with 80 μCi of either FIAU or FAU. In the first group, three mice underwent dynamic imaging with each tracer for 60 min followed by static imaging at 65 min and 120 min post injection (Concorde R4 microPET scanner). Then, the animals were sacrificed at 150 min post injection. A second group of five animals for each tracer was sacrificed at 60 min post injection. Samples of blood and tissue were collected for biodistribution and metabolite analysis. Regions of interest were drawn on both the tumors in the microPET images to calculate the time activity curves.
Biodistribution and imaging studies showed highest uptake of FIAU in the MH3924A-tk8 tumors. The mean SUV±SE for MH3924A-tk8 were 1.8±0.42, 4.25±1.45, and that of MH3924A tumors were 0.16+0.04, 0.39+0.09 at 60 and 150 min respectively. FIAU showed continuous accumulation in HSV-tk tumors. Imaging with FAU demonstrated SUV of less than one at both time points. At 60 min nearly 50% of blood activity is present as FAU.
We have successfully used [F-18] FIAU to image HSV1-tk gene expression. The contribution FAU to the total uptake seen in HSV-tk tumors is minimal.
Abstract ID: 224
Dependence of the complex relaxivity of the Ca2+ ions concentration.
Abstract ID: 225
Upon initial diagnosis, approximately 80–90% of prostate cancers are androgen-dependent, and endocrine therapy of prostate cancer is directed toward the reduction of serum androgens and inhibition of the androgen receptor (AR) [2]. On the other hand, some very aggressive forms of prostate cancer were shown to be androgen-independent and thus insensitive to inhibition of the AR [1,2]. Specific treatment of prostate cancer is done by inhibition of AR using anti-androgens such as flutamide (pro-drug of hydroxyflutamide), nilutamide and bicalutamide. Thus, since it is critical to determine the role of tumor dependency on the AR in each individual patient, molecular imaging modalities are important components for monitoring and guidance of therapy. (R)-3-Bromo-N-(4-fluoro-3-(trifluoromethyl)phenyl)-2-hydroxy-2-methyl-propanamide
18F-radiolabelled nonsteriodal derivatives.
Abstract ID: 226
In recent years there has been significant interest in the development of monofunctional, water-soluble near-infrared fluorochromes with improved optical properties for use in a myriad of imaging applications. The majority of these compounds require multi-step synthesis, are difficult to purify, or show a tendency to aggregate in aqueous solution, necessitating the development of new NIR fluorochromes. In this report, we detail the development of a new class of monocarboxylate functionalized cyanine derivatives that address these issues. The synthetic strategy relies on efficient nucleophilic attack of alkyl thiols on commercially available cyanine dyes bearing chloro substituted polymethinic linkers. Monocarboxylate derivatized fluorochromes (CyTE dyes) can be prepared in one step in greater than 90% yield without the need for purification. Several of the fluorochromes prepared by this route show no tendency to aggregate in aqueous solution and have excitation and emission maxima greater than 800 nm. To demonstrate the potential of the CyTE fluorochromes, we chose to monitor endothelial vascular adhesion molecule-1 (VCAM-1) mediated uptake of phage displaying the VCAM-1 targeting peptide (VHSPNKK) in murine cardiac endothelial cells (MCECs). VCAM-1 is upregulated in the endothelium under atherosclerotic conditions and the utility of VCAM-1 targeted probes for imaging atherosclerotic lesions was demonstrated previously with a modified magnetofluorescent nanoparticle. In this study, we labeled phage containing the VCAM-1 targeting peptide with the CyTE-777 fluorochrome and were easily able to monitor the internalization of the peptide displaying phage, demonstrating the utility of this fluorochrome for imaging of biological processes.
Abstract ID: 227
Targeted imaging of disease-associated cell surface markers can lead to development of personalized treatment regiments. The lack of a reliable technology for “loading” contrast agents onto targeting proteins inhibits the advances in this area. Currently, loading is achieved by random chemical cross-linking of cargo to targeting proteins, which damages proteins, requires expensive customized development, and yields heterogeneous products.
We have developed a new platform technology for loading contrast agents onto targeting proteins that circumvents these problems. The strategy is shown on Figure 1 and is based on expression of targeting protein with a 15-aa humanized tag containing cysteine in position 4 (C4-tag). The tag is fused to a targeting protein via a short G4S linker.
In our experience, C4-tagged recombinant proteins produced in bacteria and refolded in typical red-ox buffers contain varying numbers of free thiol group (5–40%), while the majority of the C4 thiol groups require “deprotection” under mild reducing conditions. After deprotection, C4 thiol group can be conjugated either directly to contrast agents (e.g. near-infrared dye) or to contrast agent carriers(e.g. chelators for radionuclides).
Alternatively, contrast agents or their carriers can be conjugated to a ~10 kDa humanized adapter protein that can bind to C4-tag (Fig. 1B). To avoid dissociation of the complex, adapter protein is engineered to form an intermolecular disulfide bond with C4 thiol group upon the complex formation. Importantly, formation of this disulfide bond does not require prior “deprotection” of C4 thiol group.
We expect that these facile procedures for loading contrast agents onto proteins will provide new opportunities for targeted imaging.
Loading Contrast Agents on Targeting Proteins via Cystein-containing Tag (A) or Complimentary Adapter Protein (B).
Abstract ID: 228
Rupture of atherosclerotic plaques is the most common cause of ischemic episodes. These instable plaques are characterized by the presence of apoptotic cells, mainly vascular smooth muscle cells and macrophages. Current diagnosis methods do not discriminate between stable and instable plaques. Detection of instable plaques would improve in-time treatment.
Annexin A5 (AnxA5) is a ligand known to specifically bind to apoptotic cells. To specifically visualize apoptotic cells, we conjugated AnxA5 with fluorescent QDs. These are bright, non-bleachable nano-particles with narrow emission and broad excitation bands. Two conjugation methods were examined. We coupled biotinylated AnxA5 to streptavidin-coated QDs as well as AnxA5 directly to QDs by thiol chemistry.
For both conjugation methods AnxA5 retained its specific apoptotic cell binding property as measured in vitro. The AnxA5/QD conjugate is therefore a promising probe for apoptosis visualization. Visualization of QDs by two photon laser scanning microscopy allows inspection of a complete vessel at subcellular level, as described by van Zandvoort et al. [van Zandvoort et al. (2004)]. In the future carotid arteries from ApoE knock-out mice will be examined for instable plaques using AnxA5/QD conjugates.
left: early apoptotic jurkat cell labeled with AnxA5-QD using streptavidin/biotin system; right: late apoptotic jurkat cell labeled with AnxA5-QD using direct coupling system; fragmented nucleus shown by propidium iodide.
Abstract ID: 229
Since it is widely accepted that the formation of senile plaques (SPs) is associated with the development of Alzheimer's disease (AD), quantitative evaluation of SPs in brain using non-invasive nuclear medicine techniques would allow presymptomatic detection of AD as well as therapeutic effect monitoring. In this study, we synthesized a novel series of benzofuran derivatives (BZFs) as to be isosteric analogs of thioflavin T, and evaluated the possibility as PET tracer targeting SPs in AD brain.
Using the AD brain gray matter homogenates, the affinities of BZFs were measured by competitive inhibition study of [123I]IMPY. Consequently, 5-hydroxy-2-(4-methylaminophenyl)benzofran (HMBZF) showed a remarkable Ki value of 0.7 nM, which is lower than [123I]PIB (4.3 nM), a feasible SPs imaging SPECT tracer of benzothiazole structure. [11C] HMBZF was successfully synthesized from corresponding N-desmethyl precursor with [11C] methyl triflate (radiochemical yield: >99%, specific activity: 32GBq/μmol). When injected intravenously to ddY mice (6-week-old), [11C]HMBZF displayed fast and high brain accumulation (4.8%ID/g at 2 min after injection) and rapid clearance (0.5%ID/g at 30 min), which is comparable to [123I]PIB (7.0 and 0.6%ID/g at 2 and 30 min, respectively).
In summary, [11C] HMBZF is characterized with a high affinity to SPs and good kinetics (high accumulation into and rapid washout from normal brain) and therefore, appears to be a promising candidate for imaging SPs in brain with PET.
Abstract ID: 230
Cell labelling: Flt3+CD11b+progenitor cells were isolated from mice [3]. Cells were incubated with Gd-DTPA-FA (20mM) for up to 24hr, washed and either used for in vitro testing or implanted in the rat brain.
MRI: T1-weighted MR images were acquired on a Bruker Biospin 7.0T scanner using 3D gradient echos (FLASH, TR=120ms, TE=5ms, spatial resolution 78μm).
Animal model: 100,000-1,000,000 cells in 2μl were implanted in Wistar rats (n=4) [1]. Animals were imaged immediately after implantation and 4–14 days thereafter.
In vivo detection limits were 50,000 cells, which were detectable for up to 7 days after implantation. 200,000 cells were also detectable after 14 days.
We were able to show that insoluble Gd-chelates are a suitable CA for conditional activation by intracellular lipases. The chelate can easily be modified to be targeted by other enzymes expressed during specific changes of cell status, making it suitable for functional cellular MR imaging.
Abstract ID: 231
Robust methods are now available for fluorine 18 labelling of oligonucleotides leading them suitable for imaging by Positron Emission Tomography (PET). Here we validated a general strategy to isolate nucleic acids-based antibodies (aptamers) to directly target transmembrane receptors in their native cell membrane inserted configuration. Our goal is to design these aptamers as specific radiotracer probes for in vivo imaging of cancer.
To this purpose we addressed to the transmembrane Ret receptor tyrosine kinase as a very suitable model system. Germline mutations in the RET gene are responsible for constitutive activation of the receptor and for inheritance of multiple endocrine neoplasia (MEN) type 2A and 2B syndromes and of familial medullary thyroid carcinoma. The C634Y mutation in the extracellular domain (95% of MEN 2A) results in constitutive dimerization whereas the M918T mutation in the intracellular domain (100% of MEN 2B) activates constitutively the receptor without dimerization. Therefore, using the extracellular domain of Ret as target would have the double advantage: first, to overcome any problem related to the delivery of ligands through the cell membrane, and second, to enrich for aptamers able to preferentially target the conformational variants caused by the activating mutations located within the extracellular domain. We raised RNAse resistant 2′-F pyrimidine aptamers against PC12/MEN2A and counter selected with PC12 cells. One of the characterized aptamers, binds specifically to PC12/MEN2A cells with an approximate Kd of 50nM and can antagonize Ret action as measured by: a) inhibition of phosphorylation of RetC634Y and downstream targets; b) inhibition of GDNF-dependent Retwt signaling; c) reversion of the morphological phenotype of GDNF-stimulated cells; d) reversion of the transformed morphology of NIH3T3 cells expressing the RetC634Y receptor. Our results constitute a proof-of-principle that ex-vivo selection can be effective for the generation of oligonucleotide ligands that are potentially suitable for in vivo imaging studies.
Abstract ID: 232
Partial evidence suggests that in vivo stability of oligonucleotides is dependant on their sequence, but so far no correlation between sequence and stability has been evidenced. We attempted to isolate sequences exhibiting increased stability in vivo by performing SELEX (Systematic Evolution of Ligands by Exponential enrichment) directly in living animals.
A random library of single strand DNA was subjected to repeated cycles of SELEX including IV injection in nude mice, recovery of the sequences resistant to degradation after 60 minutes and amplification by PCR. After six cycles, four major sequences were selected and the one with the best stability (seq. 22) was further studied. The AUC (0–1h) of seq. 22 in plasma was higher by a factor 10 than that of the random sequence (4 against 0,4 respectively).
No binding of seq. 22 to circulating blood cells could be evidenced. Electrophoretic Mobility Shift Assay (EMSA) and affinity chromatography showed binding of seq. 22 to a serum protein. However, the naive random library bound similarly to the same protein, outruling the possibility that binding to this protein was responsible for increased stability. In contrast, kinetics of the in vitro degradation by S1-nuclease, a single strand specific endonuclease, showed an improved resistance of seq. 22 to S1-nuclease.
These preliminary results will have to be confirm by using other nucleases but they suggest that the improved stability of seq. 22 is due to increased resistance to nucleases rather than to binding to circulating blood cells or serum proteins.
They confirm that the 3D structure of oligonucleotides, driven by their primary sequence, conditions in turn their interaction with other biomolecules such as nucleases. Sequences stable in vivo could prove useful as in vivo diagnostic tools or to protect therapeutic oligonucleotides from rapid degradation in blood.
Abstract ID: 233
In order to overcome the low sensitivity of MRI modality, it is necessary to use high sensitive imaging probes. In this study Gd-loaded apoferritin functionalized with biotin residues has been used to visualise endothelial tumor cells. Following a previously reported procedure, it has been possible to entrap about 10 units of Gd-HPDO3A complexes within the interior of the apoferritin cavity. Interestingly, the relaxivity shown by each Gd(III) complex in the cavity is very high and makes this “paramagnetic protein” a particularly efficient contrast agent.
Then Gd-loaded apoferritin has been biotinilated accordingly to a standard protein modification protocol using NHS-LC-BIOTIN in order to pursue its selective binding to a specific site of interest, using a streptavidin unit as a linker. The biotinylated apoferritin has been tested “in vitro” on immortalised microendothelial cells derived from a renal carcinoma (Eck-25). These cells overexpress a specific adhesion molecule called N-CAM molecule, able to bind selectively a particular peptide (N-CAM peptide). The N-CAM peptide has been biotinylated without loss of its targeting ability. Using this peptide as a vector for the Gd-loaded apoferritin, the microendothelial cells can be visualized. On this basis, “in vivo” MRI visualization of matrigel supported microendothelial cells inoculated on SCID mice has been assessed.
Abstract ID: 234
The anthracycline antibiotic doxorubicin (Adriamycin®) has been extensively used for more than 30 years in the treatment of different cancers. Despite doxorubicin's widespread use, its clinical application is limited by undesired side effects like cardiotoxixity, myelosuppression and nephrotoxicity [1].
To overcome these problems and to enhance its activity against tumors, several doxorubicin prodrugs have been synthesized. Often doxorubicin is linked to a carrier system which is capable of targeting the cells where the drug is activated. In all of these cases the activation of the drug can be followed only by ex-vivo analysis. The introduction of an MRI “procontrast-prodrug” agent would allow the monitoring of the drug activation on real time. Such agent, namely a conjugate between an MRI contrast agent and doxorubicin, would show a relaxivity change following the release of the active drug.
As a first step in this direction, we designed and synthesized the doxorubicin conjugate with a Gadolinium complexing agent shown in Figure 1. The MR agent and doxorubicin are connected through a hydrazone linker that is hydrolysed in mildly acidic conditions like those present in lysosomes [2].
In the conjugate the presence of the doxorubicin moiety is expected to influence either the hydration state of the gadolinium ion or the water exchange rate or both. This will give rise to a change in relaxivity following the release of the active drug.
The synthesis of the procontrast-prodrug agent will be presented together with the preliminary relaxometric characterization of the gadolinium complex.
Abstract ID: 235
Glycosaminoglycans (GAG) are naturally present in healthy cartilage tissue, and the loss of GAG from cartilage is an early marker of osteoarthritis. Magnetic resonance imaging (MRI) methods can evaluate the spatial distribution of GAG concentrations in cartilage by detecting the relative distribution of a negatively charged MRI contrast agent that is repulsed by negatively charged GAG. Improved quantification can be obtained by comparing distributions of a charged MRI contrast agent and a neutral agent, which accounts for variable pharmacokinetics of contrast agents in cartilage that are unrelated to charge properties. However, standard MRI agents can't be selectively detected, requiring the serial administration of charged and neutral agents, which creates technical challenges and can compromise the comparison of the MRI agents.
To address this problem, we've synthesized a negatively charged MRI contrast agent, Tm(DOTAM-Gly), and a neutral agent, Yb(DO3AM-Acetamide) that can be selectively detected via the mechanism of chemical exchange saturation transfer (CEST). New MRI methods developed in our laboratory were used to detect CEST agents within in vitro samples. Healthy and GAG-depleted ex vivo rabbit cartilage were soaked for 24 hours with a solution of one or both CEST agents at physiologically-relevant concentrations (0.25 mM–2.0 mM). The concentrations of each contrast agent in each cartilage sample are detected using our MR methods, and correlated with the MR results with histopathological measurements of GAG concentrations in each cartilage sample. Results from these studies will be presented to assess the potential for improved quantification of cartilage GAG concentrations using selectively–detectable CEST agents.
Abstract ID: 236
Abstract ID: 237
CEST probes represent an emerging class of MRI contrast agents of huge potential. They act as negative agents by reducing the signal intensity of the water protons through a saturation transfer mediated by chemical exchange.
The great potential of CEST agents lies on the possibility of switching “on” and “off” the contrast at will, making possible the detection of more agents, each uniquely characterized by a specific absorption frequency of their mobile protons, in the same specimen. The sensitivity of a CEST agent is directly dependent on the number of mobile protons which can be saturated. For this reason, in order to design CEST probes characterized by a well defined shifted resonance corresponding to a high number of exchangeable protons, we consider the use of liposomes entrapping a water proton paramagnetic shift reagent (SR) in their aqueous inner cavity (LIPOCEST agents). The resonance frequency of the water protons inside liposomes may be modulated by changing the nature and concentration of the SR unit, whereas the water permeability of the liposome may be controlled by modifying the chemical composition of the lipidic bilayer.
LIPOCEST agents display a noticeable sensitivity, in the range of picomolar concentration (if expressed as liposome concentration), and, very important, they do not require high intensity saturation fields for transferring saturation to bulk water.
Abstract ID: 238
Low Density Lipoproteins are liposomial-like particles (ca 22 nm diameter) that represent important endogenous transport vehicle in mammalian system. LDL receptors are overexpressed in tumor cells. Therefore LDL may be considered a good candidate for MRI visualization of tumor cells once loaded with Gd complexes.
The binding to LDL of several Gd-based Imaging Probes bearing lipophilic substituents has been investigated in detail. Then the system showing the highest binding constants (KA > 104) have been tested for the uptake in the tumor cells (HepG2 and B16) via the LDL receptor route.
The cellular labelling experiments proved that, after 24 hours of incubation in the presence of 30 μg/ml of LDL and 18μM of the Gd (III) containing probes, the amounts of internalised Gd is sufficient to generate hyper intense signals in the corresponding MR images. Work is in progress to assess the potential of the proposed procedure on tumor bearing animal models.
Abstract ID: 239
Therapy-induced tumor necrosis is associated with long-term treatment response. MR imaging probes that selectively accumulate in necrotic tissues would provide a means of non-invasively assessing necrosis with high spatial resolution, thus allowing the assessment of early treatment response in cancer patients. We have previously reported the synthesis and characterization of polymeric DTPA-Gd-poly(L-glutamic acid) (PG-Gd) as a potential blood-pool MR imaging agent (Wen et al., Bioconjugate Chem., 15: 1408, 2004). Here, we show that PG-Gd can be used to image regions of tumor necrosis. Mice bearing syngeneic murine ovarian OCA-1 tumors (~8-mm in diameter) were treated on Day 1 with taxane and subsequently injected with PG-Gd, Magnevist™, or a degradation product of PG-Gd (PG-Gd pretreated with cathepsin B). Each contrast agent was given intravenously at a dose of 0.2 mmol Gd/kg. T1-weighted MR images were acquired immediately after contrast agent injection and subsequently on Days 3 and 5. Rapid enhancement of the tumor was noted following Magnevist™ and degraded PG-Gd injection. Given the large molecular weight (~100 kD) of PG-Gd, very minimal contrast enhancement of the tumor was noted immediately after its injection. However, a heterogeneous enhancement pattern was observed within the tumors on Days 3 and 5. Comparison of MR images and microphotography of matched tumor sections stained with H&E clearly showed that areas of contrast enhancement in the MR images corresponded well with regions of necrosis. Neither Magnevist™ nor the degradation product of PG-Gd yielded detectable enhancement on Day 3 or 5 images. These data strongly suggest that the polymeric form of PG-Gd is required for the agent to accumulate in the necrotic areas, and that further studies are warranted to elucidate the specific mechanism of action. (Supported by U54 CA90810 and the John S. Dunn Foundation).
Abstract ID: 240
High quality magnetic iron oxide (Fe3O4) nanocrystal probes were developed for in vivo cellular magnetic resonance (MR) tracking. The nanocrystal surface properties were modulated by adopting either cationic or anionic charged ligand onto the nanocrystal surface. Subsequently, the surface charge dependent intracellular labeling capability of the iron oxide nanocrystals were examined to a variety of cell types including neural stem cells, tumor cells, fibroblast cells, and ovary cells. Systematic analyses including Prussian blue staining, MR imaging, inductively coupled plasma atomic emission spectra (ICP-AES), and transmission electron microscope (TEM) reveal that transfection of our cationic nanocrystals into the cells is highly efficient in the case of cationic charged iron oxide nanocrystals. Simultaneously, cytotoxicity of the nanocrystals was investigated through trypan blue exclusion assay. Our nanocrystals are reasonably nontoxic in a typical dose level for the intracellular labeling. Finally, these magnetic nanocrystals with nontoxicity and highly effective transfection capability were successfully utilized as MR imaging probes for the monitoring of neural stem cell migration in a rat spinal code.
Abstract ID: 241
As Alzheimer's disease pathogenesis is associated with the formation of insoluble aggregates of amyloid-peptide, approaches allowing the direct, noninvasive visualization of plaque growth in vivo would be beneficial for biomedical research. Here we describe the synthesis and characterization of a series of oxazine dyes that readily penetrate the intact blood-brain barrier and show preferential binding to amyloid plaques. In particular, the dye AOI987 has optical properties, such as maximum absorption/emission in the near infrared domain and high quantum yield, that are favorable for in vivo imaging. Using near-infrared fluorescence imaging, we demonstrated specific interaction of AOI987 with amyloid plaques in APP23 transgenic mice in vivo, as confirmed by postmortem analysis of brain slices. Quantitative analysis revealed increasing fluorescence signal intensity with increasing plaque load of the animals, and significant binding of AOI987 was observed for APP23 transgenic mice aged 9 months and older. Thus, the plaque-specific oxazine NIRF dye AOI987 is an attractive probe to noninvasively monitor disease progression in animal models of Alzheimer disease and to evaluate effects of potential Alzheimer disease drugs on the plaque load.
Abstract ID: 242
Abstract ID: 243
Among different matrix metalloproteinases, gelatinase A (MMP-2) and gelatinase B (MMP-9) have been the most consistently detected in malignant tumors and their level of expression has been associated with tumor angiogenesis and metastasis. In this work we describe an MMP-2-targeted NIR imaging agent developed on the basis of structure-activity relationship of cyclic peptides having inhibitory activity against MMP-2. We started from cyclo(CTTHWGFTLC) (CTT-peptide), which had previously been identified as a MMP-2-binding peptide through in vivo phage selection technique. Replacing the S-S bond formed between two Cys in CTT-peptide with an amide bond formed between the N-terminus of Ala and the side chain of Asp resulted in a significant increase in the stability of the resulting cyclic peptide, c(ATTHWGFTLD)NH2. Substitution of Asp-NH2 with β-Ala without changing the ring size of the cyclic peptide resulted in an increase in IC50 value from 11 μM of c(ATTHWGFTLD)HN2 to 84 μM of c(ATTHWGFTL-βAla), suggesting that the CONH2 group of Asp was one of critical sties. Results from Ala survey indicated that the first three amino acids ATT in (ATTHWGFTLD)HN2 could be substituted to introduce a functional group suitable for further derivatization. Final elaboration to reduce the number of amino acids led to a side chain-to-side chain cyclized peptide, c(KAHWGFTLD)HN2, which had an IC50 value of 7.6 μM. Conjugation of Cy5.5 to the α-NH2 of Lys in c(KAHWGFTLD)HN2 gave Cy5.5-c(KAHWGFTLD)NH2, a NIRF dye that bound to a variety of tumors cells expressing MMP-2 in vitro and to tumors inoculated either ectopically or orthotopically in vivo. (Supported by R01EB00174 and John S. Dunn Foundation).
Abstract ID: 244
Non-invasive detection of tumoral expression of hypoxia would permit selection of patients for hypoxia-targeted therapeutic strategies and improved management of cancer. Many studies have shown that there is a correlation between expression of the transmembrane enzyme carbonic anhydrase 9 (CA9) and tumor hypoxia. 2,3,5,6-Tetrafluoro-3′-sulfamoylbenzanilide(TFSB) has been shown to bind to and inhibit CA9 with high activity (ki (hCA9) = 0.8 nM) and high selectivity for CA9 relative to other cell-surface carbonic anhydrases (ki (hCA2)/ki (hCA9) = 26.3). To develop a CA9-selective probe for positron emission tomography (PET), 18F-2,3,5,6-tetrafluoro-3′-sulfamoylbenzanilide was prepared using an 18F-19F isotope exchange reaction. TFSB was synthesized as previously reported, with the labeling reaction performed at 110 °C in DMSO with the addition of Kryptofix2,2,2 and potassium carbonate. 18F-TFSB was separated by a semi-preparative HPLC followed by further C-18 Sep-Pak purification. The radiochemical yield was 8–15% with a synthesis time of 50 minutes, including HPLC separation. The radiochemical purity was higher than 95%. Due to limitations of the isotope exchange reaction, the specific activity is limited to 20–200 MBq/μmol. A no-carrier-added synthesis method to improve specific activity is currently being developed. The uptake of this tracer in CA9-overexpressing RCC4 cells and hypoxic and normoxic HT-29 cells has been measured. The results showed significant higher uptake in RCC4 cells and hypoxic HT-29 cells than that in normoxic HT-29 cells (Fig. 1). These results indicate the potential of 18F-TFSB as a PET tracer for detecting expression of CA9 in vitro, and we are currently pursuing application of this tracer in hypoxic animal tumor models.
Uptake of 18F-2,3,5,6-tetrafluoro-3′-sulfamcylbenzanilide in cells.
Abstract ID: 245
In the field of targeted imaging apoptosis has received broad attention since it is a fundamental mechanism for healthy organisms as well as for many diseases. While the caspase-3 activation is seen as a crucial bottleneck for the most cases of apoptosis induction, the externalization of negative phospholipids to the outer leaflet of the cell membrane during the early phase of apoptosis is more accessible. Therefore apoptosis imaging with radioactive annexin V, which binds to these exposed phospholipids, could progress to first clinical trials1.
Recently we have developed a near infrared fluorescent for optical2 and a superparamagnetic annexin V for MRI3,4, which were successfully used for imaging apoptosis in mouse models.
Here we will present results in the development of a new ultra small magnetic annexin V, which is based on monomer coated Very Small Iron Oxide Particles5 (VSOP, ~ 7 nm diameter, currently in a phase II clinical trial) in contrast to the dextran coated Cross-linked Iron Oxide Particles (CLIO) used so far. These VSOP's provide the opportunity to make a magnetic annexin V below 20 nm with a possible improved bioavailability in comparison to Anx-CLIO (~50 nm)3. We coupled an engineered annexin V via Sulfosuccinimidyl 4-(N-maleimidomethyl)cyclohexane-1- carboxylate and short poly-amines covalently to the VSOP's. To date we could bind up to 1.8 mol of engineered annexin V per mol VSOP. With this coupling procedure we could preserve the binding affinity of the resulting Anx-VSOP's.
Concluding, we could synthesize Anx-VSOP, a new ultra small magnetic annexin V for detection of apoptosis by MRI, which could further improve the method of annexin V based MRI of apoptosis in vivo.
Abstract ID: 246
Poster Session: Molecular and Functional Imaging in Cancer, Part 2
Abstract ID: 247
Abstract ID: 248
Abstract ID: 249
Abstract ID: 250
For the systemic treatment of colorectal cancer 5-fluorouracil (FU)-based chemotherapy is standard. However, only a subset of patients respond to chemotherapy. Breathing of carbogen (95% O2; 5% CO2) may increase uptake of FU by changes in tumor physiology. The aim of this study was to monitor in vivo in animal models the effect of carbogen breathing on tumor blood plasma volume, pH and energy status, as well as the effect on FU uptake and metabolism in C38 and C26a colon tumors that differ in their vascular structure. We used 31P magnetic resonance spectroscopy (MRS) to assess tumor pH and energy status and 19F MRS to follow FU uptake and metabolism. Advanced MR imaging (MRI) methods using ultra small particles of iron oxide were performed to assess blood plasma volume. Carbogen breathing significantly decreased extracellular pH, increased tumor blood plasma volume (Fig. 1), and increased FU uptake in the tumors. These effects were most significant in the C38 tumor, which has a larger relative vascular area than the C26a tumor. Carbogen breathing also enhanced systemic toxicity by FU.
Fractional change in the longitudinal relaxation rate R1, which is proportional to plasma blood volume, for C38 and C26a tumors 3 min before start carbogen breathing, during carbogen breathing and after terminating carbogen breathing. The error bar indicates the standard error of the mean.
Abstract ID: 251
18F-labeled amino acids represent a promising class of imaging agents in tumors visualized by means of positron emission tomography (PET). Because of the high uptake it is still problematic to clear differentiate between tumors and inflammation.
The high enrichment in tumor tissues assumed the uptake of the tracer via a tumor-specific amino acid transporter, which is not or different expressed in inflammatory cells suggesting a different endowment of neutral amino acids. As previously shown, L-type amino acid transporter 1 (LAT1) is playing a key role because of its high up-regulation in malignant tumors. For the functional expression of LAT1 a single membrane-spanning protein, the heavy chain of 4F2 antigen (4F2hc), essentially forms a heterodimeric complex via disulfide bonds.
The present study investigated the amino acid transport mechanism of LAT1 for 3-O-methyl-6-18F-Fluoro-L-DOPA (OMFD), a novel 18F-labeled phenylalanine derivative, into tumor cells.
For molecular characterisation of L-type amino acid transporters focusing on the LAT1-4F2hc subtype we used two different tumor cells like FaDu (squamous cell carcinoma)/HT29 (colorectal adenocarcinoma) and tumor bearing mice performing quantitative RT-PCR, Western-Blot, and immunhistochemistry. In vitro uptake assays with HT29 and FaDu were performed with OMFD under physiological amino acid concentrations.
OMFD demonstrated a saturable and sodium- and energy-independent accumulation in vitro in different tumor cell lines, suggesting its uptake to be mediated exclusively by sodium-independent LAT1.
Our data emphasize the relevance of OMFD as a PET tracer for imaging of specific amino acid transport via LAT1 in tumors. Furthermore, the identification and characterization of tumor specific amino acid transporters like LAT1 will be a helpful tool for therapeutic implications. The inhibition of LAT1 activity in tumor cells could be effective in the inhibition of tumor cell growth by depriving tumor cells of essential amino acids, too.
Abstract ID: 252
Histone deacetylase (HDAC) inhibitors (HDACIs) are emerging as a new and exciting class of anti-neoplastic agents. Our goal is to develop a noninvasive magnetic resonance spectroscopy (MRS)-based molecular imaging method to detect inhibition of HDAC.
We investigated the fluorinated HDAC substrate, Boc-Lys(Tfa)-OH (BLT). First we used 19F MRS to show that BLT is cleaved in vitro by recombinant HDAC-8, confirming that it is a substrate of HDAC and that its metabolism can be monitored by MRS. To test the value of BLT as a molecular marker of HDAC inhibition, we next investigated PC-3 human prostate cancer cells. Cells were treated, in the presence 1mM BLT, with 10μM of the clinically relevant HDACI SAHA, or with DMSO Treatment with SAHA for 24 hours lead to a significant drop in HDAC activity (41% relative to controls) and in cell proliferation (62% relative to controls). BLT did not significantly affect HDAC activity or cell proliferation. Using 19F MRS we found that treatment with SAHA lead to a significant increase in BLT levels from 8 fmol/cell in controls to 18 fmol/cell in HDACI-treated cells. This indicated that BLT levels are associated with HDAC inhibition. In addition, using 31P MRS, we determined that phosphocholine levels increased significantly by 200% in HDACI-treated cells indicating that phosphocholine could serve as a surrogate metabolic marker of response.
This study indicates that 19F MRS-imaging of fluorinated HDAC substrates, combined with 31P MRS-imaging, could be used to directly and noninvasively monitor HDAC inhibition in cells, and possibly in vivo.
Abstract ID: 253
In vivo-selected peptides specifically bind to PC-3 cells but do not bind to other cultured carcinoma cells.
Live cell monolayers (96 well plates) were incubated with increasing concentrations of I biotinylated peptide. Binding was detected fluorometrically by using NeutrAvidin-Texas Red. Affinity was determined from dose-dependant binding by using nonlinear regression analysis and the GraphPad Prism program.
NBD - The best-fit curve produced by the nonlinear regression algorithm did not fit the data, as evidenced by a combination of negative rate constants, wide confidence intervals and/or low R2 values.
In vivo phage display selections might identify peptides with superior in vivo stability and tumor-targeting ability. In vivo selections in mice heterotransplanted with PC-3 human prostate tumors identified two peptides (D4, G1) that selectively recognized tumor versus normal tissue and PC-3 human prostate carcinoma cells versus control cells. Saturation binding analyses (Table 1) determined that both peptides bound only to PC-3 cells, although G1 bound to its cognate molecular target(s) with higher affinity than did D4 to its cognate target(s). Competition analysis suggested that non-biotinylated D4 (7.5 nM), but not G1 (750 pM), completely abolished binding of both biotinylated D4 and G1 (Figure 1). Non-biotinylated G1 reduced G1 affinity and maximal binding but did not affect D4 binding. Fluorescein-labeled peptides were internalized maximally by PC-3 cells at 60 minutes incubation (Figure 2). Pharmacological inhibition indicated that the D4 target might be a lipid raft-associated molecule (Figure 3, methyl-β-cyclodextrin). Inhibition of endosome acidification (bafilomycin) reduced D4 internalization. A range of inhibitors reduced G1 internalization nonspecifically. These data indicated that both peptides have PC-3 targeting potential, although D4 may be more antigen-specific than G1. In summary, in vivo selections successfully identified peptides that specifically recognized and were internalized by carcinoma cells.
Abstract ID: 254
Abstract ID: 256
Abstract ID: 257
Nanometer-sized polymeric particles that carry drugs, genes and imaging probes are of particular interest in many fields of nano-biotechnology. However, the lack of target-specific delivery systems is a still bottleneck that must be overcome for genomics/proteomics-based pharmaceuticals as well as current chemotherapy drugs. In this study, we sought to answer the following important questions for the rational design of cancer-specific nanoparticles: (i) What particle size is optimal for targeting the openings between defective endothelial cells in tumors? (ii) is the targeting efficiency tumor-dependent, if so, what are the main decisive factors? (iii) what happens to nanoparticles in blood circulation, do they still maintain the nanostructure in vivo? (iv) what are the mechanisms of intracellular uptake and what are they related to? (v) where are the nanoparticles stored intracellularly? Despite their obvious importance to the rational design of nanoparticles, these questions have not been fully understood yet.
In this study, polymeric self-assembled nanoparticles exhibiting long stability in blood circulation were prepared from amphiphilic chitosan derivatives. The size of nanoparticles was varied from 150 to 470 nm in a mean diameter. FITC-labeled or 131I-labeled nanoparticles were injected into the tail veins of tumor-bearing mice, and their body distribution was examined by gamma camera imaging. Various factors involved in the tumoral distribution of nanoparticles were studied in terms of physiochemical characteristics of nanoparticles, tumor angiogenesis and vascular permeability. Angiogenesis in different tumors was examined in terms of the heparanase activity, the matrix metalloproteinase (MMP) activity, and the expression of vascular endothelial growth factor (VEGF), etc. Tumor vasculature was visualized by using an endothelial cell selective reagent, Rhodamine-labeled GSL1 (Griffonia simplicifolia lectin 1). This study may provide a new strategy for the rational design of target-specific nanoparticulate drug delivery systems.
Abstract ID: 258
The occurrence of apoptosis can be a good prognostic indicator for the outcome of cancer treatment. The progress in personalized molecular medicine, which may in the future provide a variety of treatment options, has stimulated a need for methods that permit the early noninvasive detection of tumour apoptosis. Magnetic resonance imaging (MRI) is one method that could be used to achieve this in the clinic [1]. Manganese, which can enter cells through calcium channels, enhances T1 relaxation and this property has been exploited in MRI of cellular activity [2]. Increased calcium uptake is an early event in apoptosis, so we sought to determine whether there was also an increase in Mn2+ uptake, which would allow the early detection of apoptosis using MRI.
Apoptosis was induced in mouse lymphoma (EL-4) cells (107 cells/ml) using etoposide (15μM), which were then incubated with 1mM MnCl2 for one hour before T1-maps and T1-weighted images were acquired from the cell pellets at 9.4T. In parallel experiments, the effect of apoptosis induction on calcium and Mn2+ uptake was determined fluorometrically using the Ca2+ probe, Fura-2 (Ca2+ uptake was measured directly from the increased Fura-2 fluorescence, whereas Mn2+ uptake was inferred from Mn2+ quenching of Fura-2 fluorescence). Both the MRI and fluorescence measurements indicated that apoptotic cells showed increased Mn2+ uptake by four hours of etoposide treatment, a time point at which histological staining does not yet reveal the morphological characteristics of the process.
These data show that MRI can detect early increases of Mn2+ concentration in cells undergoing apoptosis in vitro. Although issues of Mn2+ perfusion and washout are likely to complicate these measurements in vivo, these data suggest that Mn2+ uptake might be a MRI-detectable marker for apoptosis in tumours undergoing chemotherapy.
Abstract ID: 259
Apoptosis, programmed cell death, is induced in tumours following therapy, where the extent and speed of onset are good prognostic indicators for treatment outcome. The C2A domain of Synaptotagmin I is a 14 kDa protein that binds with high affinity (nM) to exposed phosphatidylserine (PtdS) on apoptotic cells.
Previous work from our group [1] has demonstrated that C2A conjugated to a superparamagnetic iron-oxide nanoparticle, can detect apoptotic cells, both in vitro and in a murine lymphoma tumour model (EL4-), using T2-weighted MRI. We are currently developing a new generation of smaller contrast agents (CAs), based on Gd-chelated compounds [2], which provide positive (T1-weighted) contrast. This may improve detectability in tumours, which are irregular structures that can show heterogeneous and time varying contrast.
GST-tagged C2A was directly conjugated to a gadolinium chelate to produce a 95 kDa CA, with a T1-relaxivity of ca. 30 mM−1s−1 (at 9.4 T). Other CAs were developed based on the interaction of avidin and biotinylated C2A. Initial modification of avidin with a Gd-chelate and subsequent complex formation with biotinylated C2A produced 100–150 kDa complexes, with T1-relaxivities in the range 60–75 mM−1s−1 (9.4 T). A third approach that we have adopted involved cross-linking of C2A to polyamidoamine (PAMAM) generation 4 dendrimers, whose 64 surface amino groups allow conjugation to numerous Gd-chelates. These Gd-conjugated dendrimers had relaxivities of up to 170 mM−1s−1 (at 9.4 T). Finally, the avidin/biotin interaction can be exploited in a system with biotinylated C2A and CA-loaded apoferritin.
Recent studies in vitro demonstrated that the directly-conjugated and avidin-based CAs were able to detect apoptotic EL4 cells, using T1-weighted MRI and flow cytometry. We are currently screening and validating these conjugates in vivo to select the most effective CAs.
Abstract ID: 260
The lack of success in many therapeutics of breast and other human cancers is a most frustrating, disheartening aspect of cancer prognosis and has stagnated the development of anti-cancer drugs. While tracking of cancer cells to the vascular bed of the lung and other target organs has been very highly developed, there has been no metric available which has given online prediction of the effect of the drug upon cancer metabolic activity, beacon delivery, etc. that could be monitored at any time during the course of drug therapy. The histogram display of heterogeneous metabolic states and genetic expression of beacons are readily displayed by 2- and 3D high resolution cryoimaging of cancers currently at 50 micron resolution with 1 photon imaging technology and prospectively with 2–5 micron resolution with 2 photon technology. One of the most responsive signals of cancer tumor activity is the histogram display of the redox state of mitochondria which can be detected by the ratio of the fluorescence signals of the flavoprotein moiety of the citric acid cycle, alphaketoglutarate and pyruvate dehydrogenase to the matrix space, NADH signal. Furthermore, the expression of beacon signals by the tumor genome itself can be imaged as GFP. With high resolution imaging, 104 voxel images can be made from which histogram displays of these values indicate widespread differences of their activities in different parts of the cancer mass. Furthermore, the redox state of the tumor mitochondria appears to reflect the apoptitic activity of cancer drugs, an intrinsic rather than extrinsic signal. The hypothesis proposed here is that the effect of drugs will be to inhibit a certain part of the total tumor heterogeneous population and that this histogram of the survivors tells the effectiveness of the therapies.
Supported in part by NIH CA72895, CA096016.
Abstract ID: 261
Microbubbles (MB) are intravenous contrast agents for ultrasound (US) and may be targeted to specific vascular beds. In the present study, we demonstrate that avidin can be incorporated into the shell of perfluorocarbon-exposed dextrose albumin (PESDA) MB and serve as an anchor for biotinylated monoclonal antibodies (mAbs). To demonstrate the efficacy of this novel system in vitro we chose to target endoglin (CD105), a glycoprotein expressed on the surface of proliferating endothelial cells (EC). As the targeting moiety we used monoclonal antibody MJ7/18, which is specific for murine CD105. MB targeted to CD105 specifically bound to EC but not to fibroblasts while MB conjugated with a control IgG did not bind either cell type. We then evaluated the utility of this reagent for visualizing tumor vasculature in vivo. Mice bearing subcutaneous pancreatic adenocarcinomas were treated with chemotherapy or placebo starting two weeks after injection of cancer cells. After 3 weeks of treatment mice were injected with both untargeted and targeted MB and tumors were imaged by US. Monoclonal antibodies directed against CD105 or Vascular Growth Factor Receptor-2 (VEGFR2) were used as targeting moieties. Quantitative acoustic videointensity (AVI) of tumor images was significantly increased with MB targeted with anti-CD105 or VEGFR2 mAbs compared to untargeted MB, 12.0±1.6, 12.4±2.4, 5.9±0.2 video intensity units respectively (p<0.001). AVI of tumor images measured after injection of targeted MB was significantly lower in mice treated with chemotherapy compared to placebo. Furthermore, AVI profiles generated with targeted MB were proportional to mean vascular density assessed by immunohistochemistry. We conclude that Av-PESDA-MB represent a novel and attractive tool to conjugate MB with biotinylated mAbs for the purposes of vascular targeting and imaging. MB conjugated with markers of endothelial proliferation (anti-CD105 or -VEGFR-2 mAbs) combined with US imaging allow non-invasive in vivo assessment of tumor vasculature and angiogenesis.
Abstract ID: 262
For the purposes of molecular diagnostics and therapeutics, it is important to know the optimal agent dose in order to avoid excess administration with respect to the expression level of targeted receptors or other disease markers. Yet to date, it is not clear what dose of molecular diagnostic or therapeutic agents causes saturation of the receptor binding sites in vivo. Herein, the dose dependence of the uptake of fluorescently labeled RGD peptide, Cy5.5-c(KRGDf) targeted to the αvβ3 integrin was assessed from xenografts bearing a subcutaneous human Kaposi's sarcoma (KS1767) and melanoma (M21) tumor. Dynamic fluorescence optical imaging with the three-compartment PK model was used to analyze fluorescence images acquired with an intensified charge-coupled device (ICCD) system.
The results show early and rapid uptake of Cy5.5-c(KRGDf), which was mediated by the administration of c(KRGDf) 1 hour before the administration of the conjugate agent. We found linear increases in PK uptake rates at doses between 0.75 to 1.5 nmol/animal in the KS1767 tumor model. However in the M21 model, PK parameters were not affected by doses ranging from 3 to 6 nmol / animal, or even by the pre-administration of c(KRGDf) as a competitive ligand. Furthermore, using the rapid millisecond image acquisition, we can dynamically image the first pass of agent through the vasculature, enabling an assessment of vessel permeability, and how it is impacted by binding. In summary, in vivo PK analysis as well as fast, dynamic NIR fluorescence optical imaging can be used not only to assess the expression level of αvβ3 but also to characterize tumor vasculature.
Abstract ID: 263
The activated, angiogenic tumor endothelium is associated with an up regulation of cell adhesion receptors, including the αvβ3-integrin. This integrin may therefore serve as a molecular marker for molecular imaging of angiogenesis. It has been established that a cyclic conformation of the RGD-peptide has high affinity for the αvβ3-integrin.
In this study we aimed to identify the activated tumor endothelium in tumor bearing mice with in vivo MRI and ex vivo fluorescence microscopy. To that end, we used recently introduced paramagnetic and fluorescent lipidic nanoparticles (LNP) that were conjugated with approximately 300 RGD-peptides per particle to introduce multivalent properties.
Tumor bearing mice were anesthetized and measured on a 6.3 T MRI scanner. The LNP were injected intravenously in the magnet via an infusion line in the tail vein. 4 Mice were injected with αvβ3-specific RGD-LNP and 4 mice were injected with non-specific RAD-LNP. Following MRI, the tumors were dissected and frozen for fluorescence microscopy. To further assess specificity of RGD-LNP, 3 mice were first injected with non-paramagnetic RGD-LNP, followed by the administration of paramagnetic RGD-LNP.
In vivo MRI revealed that RGD-liposomes localized to a large extent in the tumor rim, whereas RAD-liposomes showed a diffuse distribution pattern. Furthermore, it was established with fluorescence microscopy that RGD-LNP were exclusively associated with tumor blood vessels, while RAD-LNP were localized in the extravascular compartment.
We demonstrated that MR molecular imaging of angiogenesis is feasible by using a targeted contrast agent specific for the αvβ3-integrin. The multimodality imaging approach revealed the site and mechanism of LNP accumulation in the tumor.
Abstract ID: 264
A) Representative 1H MRSI spectrum and 0.9 ppm lipid (−CH3) maps from BT4C glioma bearing animal during ganciclovir treatment B) Concentration of saturated fatty acids (SFA), monounsaturated fatty acids (MUFA), polyunsaturated fatty acids (PUFA) and total unsaturated vinyl (−CH=) proton moieties of membrane phospholipids during ganciclovir induced cell death.
Abstract ID: 265
Three dimensional Optical Diffusion Tomography allows in vivo studies of tumour life without any stress or damage for the animal. We present hereafter theoretical and experimental results obtained with our system on phantoms and mice.
The experimental set-up consists of a laser source (685nm) coupled to a motorized stage, a CCD camera and a tank to receive the animal. The excitation and emission wavelengths of the system and the fluorophore are chosen to optimize transmission through the whole animal (leaver, lungs …).
The acquisition geometry is considered as infinite in (x, y) directions and of thickness z varying from 10 to 15mm. For mice acquisitions, this is achieved by immersing them in an index matching medium. Reconstruction is performed through an ART algorithm based upon a fine description of material-light interaction taking into account diffusion as well as absorption phenomena.
A first study on phantoms was conducted to evaluate depth resolution: two glass capillary tubes filled with Cy5 (φ=1mm, L=20mm) separated in z by various distances (axis to axis) are immersed in a diffusing medium. A depth resolution of 4mm is achieved on simulation and of 4,5mm on experimentation.
A second study is performed on mice. A mouse, with a capillary inserted in the colon, is reconstructed with a good estimation of the capillary 3D position. Absorption modelization improves significantly the localization. We present on the following figure two representations of a mouse with tumours reconstruction: z projections superimposed on the mouse (left), and z slices (right). Detection and localization of lungs tumours will be presented on the final poster.
These first results show our system performances for reconstructing whole body mice in slab geometry.
Abstract ID: 266
Vascular endothelial growth factor (VEGF, M.W. 42 kDa) secreted by cancer cells mediates angiogenesis through binding to its receptors (VEGFR1 & 2), which are overexpressed on surrounding vascular endothelial cells (VEC) resulting in VECs proliferation. In contrast, normal VECs express low levels of VEGFRs. Molecular imaging agents able to probe tumor angiogenesis could be valuable to assess the efficacy of anti-angiogenic therapy. To target the overexpressed VEGFRs on VECs of solid tumors, our aim was to conjugate VEGF165 with DTPA for labeling with Indium-111. For the conjugation, DTPA dianhydride was reacted with VEGF165 at a molar ratio of 10:1, then purified from free DTPA by size-exclusion chromatography. It was labeled with 111In acetate and the 111In-DTPA-VEGF165 purified from free 111In by size-exclusion chromatography. The specific radioactivity of 111In-DTPA-VEGF165 was 5.62 × 104 MBq/μmol with radiochemical purity at 91% measured by instant thin-layer chromatography. Biodistribution and imaging studies were carried out using athymic mice implanted with s.c. U87MG highly vascularized glioblastoma xenografts. Biodistribution studies revealed tumor uptake (10.86 ± 1.28 % i.d./g) at 72 h after injection. This uptake was comparable to that previously observed using the fusion protein 111In-hnTf-VEGF165 (M.W. 130 kDa) consisting of VEGF165 fused to the n-lobe of human transferrin (8.48 ± 1.78 % i.d./g, P>0.05)1. 111In-DTPA-VEGF165 showed faster blood clearance than 111In-hnTf-VEGF165 yielding a tumor-to-blood ratio at 72 h p.i. of 17.74 ± 3.31 and 10.67 ± 2.08 for 111In-DTPA-VEGF165 and 111In-hnTf-VEGF165, respectively. Imaging of the glioblastoma xenograft bearing mice showed tumor localization of radioactivity as early as 24 h after injection. We conclude that 111In-DTPA-VEGF165 may serve as a promising radiopharmaceutical for imaging VEGFR-mediated angiogenesis in solid tumors.
Acknowledgements: This research was supported by grant from the Natural Sciences and Engineering Research Council of Canada. 1J Nucl Med. 2005 (submitted).
Poster Session: Imaging in Infectious Diseases, Protease Activity and Oncogenesis
Abstract ID: 267
Abstract ID: 268
Color, light, and NIRF images of apo E -/- aortae. (top row) Mice that received the cathepsin K agent showed strong NIRF signal in aortic plaques (arrows). (bottom row) Mice injected with the D-configuration control agent had minimal aoritic NIRF signal. NIRF images are widowed identically.
Abstract ID: 269
Small animal models of respiratory disease have been used to examine the changes in morphology between normal and diseased animals. In particular, tissue inhibitor metalloproteinase-3 (TIMP-3) deficient mice exhibit enlarged alveolar spaces compared to wild type controls. In this study, we propose a method of measuring lung morphology of free-breathing mice using respiratory-gated micro-CT.
Wild type and TIMP-3-deficient C57/BL6 mice were anaesthetized with a mixture of ketamine and xylazine. For imaging, the free-breathing animals were positioned on a pressure chamber that monitored their respiration externally, using the change in pressure exerted on the chamber caused by the motion of the diaphragm. Respiratory-gated micro-CT images were acquired for each animal at peak inspiration and at end expiration. Images were acquired at 80 kVp, 0.4 mA, with an entrance dose of 0.24 Gy and a scan time of 25–30 minutes, and were reconstructed with an isotropic voxel spacing of 0.087 mm.
For each image, a region of interest containing the lungs was defined and minimum intensity projections through the ROI were generated. Diameter measurements of the trachea and bronchi were recorded, along with the length from the apex to the inferior aspect of the lower lobe of the right lung. An intensity-based seeded region-growing algorithm was used to extract the entire lung and the airways separately and calculate the respective volumes. The average lung density, calculated in Houndsfield units, was −432.87±5.09 HU for the wild type and −472.38±14.42 HU for the TIMP-3 deficient mice (p < 0.05). The functional residual capacity and the tidal volume were also computed for both groups.
Respiratory-gated micro-CT images of free-breathing mice had sufficient resolution and image quality to quantify differences in the mean lung density between wild type and TIMP-3 deficient mice and promises to become a valuable tool for studying lung disease in mouse models.
Abstract ID: 270
Abstract ID: 271
HAb18G is a new hepatoma-associated antigen recently cloned by hepatoma monoclonal antibody HAb18 screening from human hepatocellular carcinoma cDNA library, ant it is abundantly expressed in human hepatoma tissues and on the cell surface of several highly metastatic hepatoma cell lines as detected by immunohistochemistry using monoclonal antibody against
Abstract ID: 272
Whole animal imaging allows viral replication and localization to be monitored in intact animals, which provides significant advantages for determining viral and host factors that determine pathogenesis. To investigate effects of interferons on spatial and temporal progression of vaccinia infection, we generated recombinant viruses that express firefly luciferase or a monomeric orange fluorescent protein. These viruses allow vaccinia infection to be monitored with bioluminescence or fluorescence imaging, respectively. The recombinant viruses were not attenuated in vitro or in vivo relative to a control WR virus. In cell culture, reporters could be detected readily by 4 hours post-infection, showing that these viruses can be used as early markers of infection. The magnitude of firefly luciferase activity measured with bioluminescence imaging in vitro and in vivo correlated directly with increasing titers of vaccinia virus, validating imaging data as a marker of viral infection. Replication of vaccinia was significantly greater in mice lacking receptors for type I interferons (IFN I R−/−) compared with wild-type mice, although both genotypes of mice developed focal infections in lungs and brain after intranasal inoculation. IFN I R−/− mice had greater dissemination of virus to liver and spleen than wild-type animals even when mortality occurred at the same time point after infection. Protective effects of type I interferons were mediated primarily through parenchymal cells rather than hematopoietic cells as analyzed by bone marrow transplant experiments. Collectively, our data define a new function for type I interferon signaling in tissue tropism of vaccinia and validate these reporter viruses for studies of pathogenesis.
Abstract ID: 273
Leukotriene B4 (LTB4) is a potent proinflammatory mediator of acute and chronic inflammatory events. A radiolabeled LTB4 antagonist that is capable of binding to leukocytes in vivo should have broad applicability in the detection of inflammatory events, without the inherent disadvantages of radiolabeling leukocytes ex vivo. In previous studies we demonstrated that 111In-labeled divalent LTB4 antagonist DPC11870-11 rapidly visualizes inflammatory foci in a variety of rabbit models. In the present study we report the synthesis, 99mTc labeling, and screening of two HYNIC analogues of DPC11870-11. Divalent HYNIC conjugate BMS57868-81 and monovalent HYNIC conjugate BMS57868-88 were labeled efficiently with 99mTc with high specific activities. The biodistribution and infection-imaging properties of these new agents were examined in New Zealand White rabbits with E. coli infection in the thigh muscle, and compared with DTPA conjugate [111In(DPC11870-11)]. All three agents were able to clearly visualize the abscess within two hours post injection (p.i.). Abscess uptake at 8 h p.i. for the three compounds was 0.22±0.08 %ID/g, 0.36±0.13 %ID/g, and 0.33±0.18 %ID/g for the divalent HYNIC, monovalent HYNIC, and divalent DTPA compounds, respectively. Monovalent HYNIC-conjugate BMS57868-88 gave superior imaging characteristics. Clearance of BMS57868-88 was exclusively renal, in sharp contrast to the hepatic clearance of the original lead compound. This switch of the principal route of clearance was facilitated by the use of multiple cysteic acid pharmacokinetic modifier groups. We have also shown that the biodistribution properties of divalent LTB4 antagonist BMS57868-81 is affected by the coligand system used with the HYNIC chelator. A comparison of coligand systems tricine/TPPTS, tricine/isonicotinic acid, and tricine alone revealed that the tricine/isonicotinic acid ligand system gave the highest uptake of agent in the abscess and lowest uptake in non-target tissues. In conclusion, monovalent LTB4 antagonist BMS57868-88 shows great potential for the rapid imaging of inflammatory and infectious foci in patients.
Abstract ID: 274
We earlier reported on specific binding of the 99mTc-M13 phage to its bacterial host in vitro and in infected mice as a potential bacterial specific imaging agent. We have now extended this investigation further to include phages P-22, E-79, VD-13 and 60 with host bacteria S. typhimurium, P. aeruginosa, E. faecium, and K. pneumoniae, respectively. All phages were radiolabeled with 99mTc via MAG3 while P22 and E-79 were also labeled with Cy5.5 for fluorescence imaging. After 10 min of incubation in vitro, the bacteria were washed and binding measured by counting or optical imaging. By fluorescence, P22 and E-79 showed association only with their respective hosts while the 99mTc counts showed at least 3-fold higher binding to host bacteria compared to non-host. For example, 99mTc-E-79 bound to P. aeruginosa 10-fold higher than to E. coli, and 20-fold higher than to S. typhimurium, while 99mTc-60 showed 20-fold greater binding to K. pneumoniae compared to three non-host bacteria. Mice infected 3 h earlier with 107host or non-host bacteria in one thigh were administered one phage and were sacrificed 3 h later. When infected with its host bacteria, the infected to normal thigh ratio was 13.8 for 99mTc-E-79 compared to 5.1, 6.7, and 11.8 when radiolabeled phage 60, VD-13 and P22 were administered. Also the accumulation of 99mTc-P22 in the infected thigh was 1.3%ID compared to 0.72%, 0.42%, and 0.8696ID when phage 60, VD-13 and E-79 were administered. In the case of VD-13 and 60, accumulations in the infected thigh were either equal to or lower than that following administration of phage and thus without apparent specificity for the infecting bacteria. In conclusion, specific host binding was observed in vitro for each of the four 99mTc-phages. Since in vivo only 99mTc-P22 and E-79 showed specificity these phage merit further investigation.
Abstract ID: 275
Allowing simultaneous visualisation, identification, enumeration and localisation of individual microbial cells, fluorescence in situ hybridization (FISH) is useful for many applications in microbiology. We used this technique for the detection of pathogens in smears or tissue sections from humans or animals, e.g. showing the invasion of Treponema pallidum into tissues (Figure 1).
Since FISH not only detects culturable microorganisms, but also fastidious and yet-to-be-cultured bacteria, it has become an important instrument for the analysis of complex microbial biofilms. Most information on biofilm formation has derived from in vitro systems showing adhesion, coaggregation and communication between bacteria. However, biofilms of the human body are difficult to mimic since highly complex components like body fluids, cell surfaces and the immune system are involved.
We used FISH to study in vivo grown human biofilms, e.g. subgingival plaque in periodontitis, intestinal spirochaetosis or infective endocarditis. FISH showed the high numbers and spatial distribution of not-yet cultured oral spirochetes and oral pathogens directly within subgingival biofilms of periodontitis patients. We studied their distribution, biofilm architecture and interaction with human cells. Furthermore numerous streptococci and staphylococci were visualized in infected heart valves of endocarditis patients in culture negative cases.
This approach might contribute to our understanding of the pathogenesis of biofilm development in the human body.
Abstract ID: 276
Human T-cell leukemia virus type 1 (HTLV-1) is the etiological agent of adult T-cell leukemia/lymphoma (ATLL). The HTLV-I protein Tax is critical for viral replication/activation and plays a fundamental role in the transformation process leading to ATLL. We previously developed a transgenic mouse model wherein restricted expression of Tax to mature T- and NK-cells through transcriptional regulation of the granzyme B promoter (GzmB-Tax) resulted in spontaneous development of granular lymphocytic leukemia (LGL). This is the only available transgenic model for Tax-induced leukemia.
To monitor Tax-dependent transformation, tumorigenesis, progression of primary lesions and disseminated disease by bioluminescence imaging (BLI), two animal models were developed:
(1) LGL cells derived from GzmB-Tax mice were transduced ex vivo to constitutively express firefly luciferase (FLuc). Subcutaneous inoculation of these reporter cells in Rag-1 immunodeficient mice resulted in development of aggressive tumors that subsequently disseminated to distant organs. BLI detected both primary and distant lesions with high sensitivity and correlated with tumor burden, disease progression as well as with post-mortem pathological analyses.
(2) To monitor Tax-dependent tumorigenesis and tumor progression in real-time, mice carrying the LTR-FLuc cassette (a Tax-activated reporter) were crossed with GzmB-Tax mice. Like the parental LTR-FLuc mice, the double-transgenics showed minimal background. Tumors spontaneously developed similar to the parental GzmB-Tax mice, but bioluminescence was confined to tumor sites. In addition, aggressive osteolytic activity of these tumors was observed with X-ray imaging. In some animals, bioluminescence also co-localized with sites of injury, probably due to inflammation-dependent Tax expression from activation of the GzmB promoter. Treatment of tumor-free animals with the hepatic T-cell activator concanavalin-A induced a transient increase in liver bioluminescence in an LTR-dependent manner.
These models should serve as valuable tools for real-time, non-invasive analyses of induction and progression of Tax-induced leukemia, as well as for monitoring novel therapeutic interventions against ATLL.
Abstract ID: 277
Cells adapt to hypoxia by stabilizing the hypoxia inducible factor (HIF-1α). HIF-1α acts as transcriptional activator for several genes, which have hypoxia response elements (HRE). Placing green fluorescent protein (GFP) expression under the control of an HRE in the human prostate cancer cell line PC-3 (PC-3 HRE-GFP) enabled us to indirectly detect HIF-1 activity and hypoxia in PC-3 HRE-GFP tumors. Although increased total choline (tCho) and phosphocholine (PC) are hallmarks of cancer cells, a heterogeneous distribution of total choline is frequently detected in solid tumors. Combined magnetic resonance (MR) spectroscopic imaging (MRSI) and optical imaging of PC-3 human prostate cancer xenografts stably expressing GFP under hypoxic conditions (PC-3 HRE-GFP) revealed a coarse co-localization between total MRSI choline maps, and hypoxic fluorescing regions (e.g. Figure 1A-D).
A: Example of a CSI data set obtain from an HRE PC-3 tumor (550 mm3) with a spatial resolution of 0.5 × 0.5 × 4 mm. B: Spectrum from a single 0.5 × 0.5 × 4 mm voxel showing total choline at 3.2 ppm, and lactate/lipid at 1.3 ppm (TE=272 ms). C: Triplanar view of total choline map (from Fig. 1A) displayed in red fused with a co-localized spin density map. A heterogeneous distribution of total choline is apparent coarsely matching the GFP distribution in Fig. 1D. D: GFP expression in a fresh tissue slice matching the MRSI slice overlaid on a white light image. Comparison on Fig1C and D reveals a coarse co-localization between the GFP distribution and total choline.
To verify this in cell culture, we induced hypoxia in PC-3 HRE-GFP cells using a hypoxic chamber containing 0.3% O2. PC-3 HRE-GFP cells exhibited significantly increased PC and tCho levels following 24h of hypoxia (Figures 2A, D), as demonstrated by quantitative 1H MR spectroscopy of cell extracts. GFP and choline kinase expression were significantly elevated in hypoxic PC-3 HRE-GFP cells compared to normoxic controls, as determined by Western blotting and fluorescence microscopy (Figures 2B, C). The increase of choline kinase expression under hypoxia resulted in the elevated PC and tCho levels in hypoxic cells. Choline kinase most likely contains an HRE in its promoter region. The distinctly heterogeneous distribution of tCho frequently observed in solid tumors may therefore be partly due to hypoxia.
A: Expanded regions of the 1H MR spectra of normoxic (left) and hypoxic (right) PC-3 HRE-GPF prostate cancer cells. Assignments: Cho, free choline; GPC, glycerophosphocholine; PC, phosphocholine; tCho, total choline (Cho+GPC+PC). B: Corresponding phase contrast (top) and fluorescence (bottom) microscopic images (40× lens). C: Corresponding choline kinase (top) and GFP expression (bottom) probed with choline kinase or GFP antibodies, respectively. Actin was probed in loading controls. D: Summary of changes in PC, GPC and total choline for two independent cell extract studies. Values are mean ± standard deviation.
Abstract ID: 278
Multidrug resistance (MDR) remains a major obstacle to successful chemotherapeutic treatment of cancer and can be caused by over-expression of P-glycoprotein (Pgp), the MDR1 gene product. To further validate a knockdown approach for circumventing MDR, we developed a Pgp inhibition strategy using short hairpin RNA interference (shRNAi) and now demonstrate efficacy and target specificity in vivo. Two of eight tested shRNAi constructs targeted against human MDR1 mRNA inhibited expression of Pgp by >90%, whereas control shRNAi had no effect. Ablation of Pgp in cells stably transduced with retroviral-mediated shRNAi was documented by Western blot and functionally confirmed by increased sensitivity of MDR1-transfected cells toward the cytotoxic drugs vincristine, paclitaxel and doxorubicin as well as by transport of 99mTc-Sestamibi. ShRNAi-mediated downregulation of Pgp transport activity both in cultured cells and in tumor implants in living animals could be followed by direct non-invasive bioluminescence imaging using the Renilla luciferase fluorophore, coelenterazine, a known Pgp transport substrate. Furthermore, after somatic gene transfer by hydrodynamic infusion of a MDR1-Firefly luciferase (MDR1-FLuc) fusion construct into mouse liver, the effect of shRNAi delivered in vivo on Pgp-FLuc protein levels was documented with bioluminescence imaging using D-luciferin. ShRNAi against MDR1 reduced bioluminescence output of the Pgp-FLuc reporter 4-fold in vivo compared to mice treated with control or scrambled shRNAi. Targeted downregulation of a somatically transferred Pgp-eGFP fusion reporter also was observed using fluorescence microscopy. Our results show that shRNAi effectively inhibited MDR1 expression and function in cultured cells, tumor implants and mammalian liver, documenting the feasibility of a knockdown approach to reversing MDR in vivo.
Abstract ID: 279
Abstract ID: 280
Galectin-3 (gal-3) is a mammalian lectin that recognizes terminal β-galactose residues. By interacting with a number of tumor-associated glycoconjugates, gal-3 mediates tumor cell adhesion, which provides a rationale for the pursuit of anti- metastasis therapeutics. Several groups have explored carbohydrate ligands of gal-3 as anti-cancer agents with mixed success, in part because of their ability to bind other galectins. Our laboratory identified a gal-3-binding peptide (G3-C12) agonist using phage display. The peptide exhibited high affinity (Kd ≈ 40 nM) and remarkable specificity for the carbohydrate recognition domain (CRD) of gal-3, but not other galectins with similar CRDs. Moreover, the peptide down-regulated human breast cancer homotypic and heterotypic cell aggregation in a dose-dependent manner. To investigate the structural basis by which this inhibition occurred, we conducted nuclear magnetic resonance (NMR) experiments with gal-3 CRD (107–250aa), as well as the full-length protein (l-250aa), with and without G3-C12. Based on analysis of chemical shift perturbation and 15N backbone relaxations, the peptide-binding site on gal-3 resides in the apolar β-sheet region of the CRD, which differs from the carbohydrate-binding portion of gal-3. Importantly, peptide G3-C12 interacts with a region unique to gal-3 and does not inhibit carbohydrate binding.
Motivated by this unusual specificity for gal-3, we conducted experiments to determine if G3-C12 could reduce tumor growth in vivo. Preliminary 18F-FDG PET/CT imaging studies with MDA-MB-435 human breast carcinoma xenografted mice suggest that the peptide does reduce tumor growth. In vivo optical and micro-PET/CT imaging studies with human breast carcinoma heterotransplanted mice are being performed to directly test the anti-cancer and anti-metastatic effects of the gal-3-binding peptide. Our study suggests that non-carbohydrate agonists of gal-3 function may be viable new anti-cancer agents.
Abstract ID: 281
c-Myc is an oncogene that is amplified in a wide range of human neoplasms. Recent studies have suggested that overexpression of the c-Myc oncoprotein deregulates glucose metabolism through the activation of several components of the glucose metabolic pathway. This effect confers a survival advantage to tumor cells by facilitating rapid cellular proliferation in the setting of hypoxia and diminished glucose supply. Our goal was to determine the effect of c-Myc overexpression by observing hepatic glucose metabolism in c-Myc transformed and non-transformed cells using small animal positron emission tomography (PET). Utilizing 10 transgenic mice, five transfected with an intravenously injected volume of the adenoviral constructs Ad/Myc, and five control mice injected with Ad/LacZ (a Myc deletion mutant), dynamic FDG scans were acquired using a dedicated small animal PET scanner. Regions of interest were drawn on multiple image slices in the liver to generate time activity curves and standard uptake values were computed as the mean activity normalized for injected dose, weight and plasma glucose. The mice in the c-Myc transformed group demonstrated 34% lower plasma glucose level (p<0.05, students t-test) and 55% greater hepatic FDG uptake (p<0.05), as compared with the control group. It was confirmed through histological staining that c-Myc expression induced liver cell hypertrophy. These findings are consistent with the in vivo manifestations of c-Myc overexpression on glucose metabolism previously reported in the literature and confirm the utility of FDG PET in monitoring the in vivo metabolic effects of c-Myc non-invasively.
Abstract ID: 282
Brain metastasis occurs in approximately 25% of cancer patients and over 50% of brain tumors are metastatic. Factors that contribute to the development of brain metastasis are poorly understood, partly due to the lack of relevant animal models. We have developed a mouse model of brain metastasis by injecting a genetically modified human tumor cell line, DU145, into the left cardiac ventricle of nude mice. This model allows us to investigate three essential steps of metastasis: cell survival in the circulation, intravasation, and clonal growth in the brain. In this study, we used two MRI techniques to assess tumor cell intravasation and clonal tumor growth. First, to detect tumor cells intravasation into the brain, we labeled DU145 cells with micron-sized iron-oxide particles and performed MRI of the injected iron-oxide labeled cells in living mice by MRI. 3-D gradient echo images of 100pm isotropic resolution of whole brain were acquired. Single iron oxide labeled cells could be detected in the brain. Second, to monitor the growth of brain metastasis and angiogenesis, we performed SPIO based cerebral blood volume studies. Gradient echo images at 75μm resolution were acquired before and after iron-oxide infusion via tail vein. The blood volume inside the brain metastasis was measured. The data showed that less than 10% of tumor cells that extravasated to the brain developed into brain tumors. Further, when brain tumors occur they are characterized by increased blood volume in this model. To our knowledge, this is the first report utilizing MRI to detect the development of brain metastasis from the single cell level and to monitor further tumor growth by following changes in blood volume in living mice. Such models are not only essential for the study of the biology of brain metastases, but are also relevant for the evaluation of anti-brain metastasis therapies.
Abstract ID: 283
Effective development of new compounds and biologics for the treatment of cancer depends on the optimization of pre-clinical animal models with well-defined biomarkers. Of the in vivo molecular imaging modalities currently available, the development of optical imaging and their use in small animal cancer models provides significant opportunity to refine oncology drug discovery research.
We have therefore developed xenograft tumor models in nu/nu mice using rat MATBIII adenocarcinoma cells constitutively expressing green (GFP) and red (DsRed) fluorescent proteins or firefly luciferase as probes to report tumor size. Monoclonal cell lines with a high and stable expression of the reporting probes have been generated by stable transfection and FACS sorting. Cellular and tumor growth rates were slower for all selected clones compared with the parental cell line and the clones varied in their in vitro and in vivo detectability. For selected clonal lines a linear correlation was demonstrated between fluorescence or bioluminescence and mass of dissected tumors. Tail vein injection of tumor cells provided a metastatic model with the metastases detected mainly in lungs and lymph nodes. Orthotopic implantation in mammary glands (see Fig. 1) resulted, after removal of the primary tumor, in metastasis to the lungs, mammary glands, and lymph nodes. These xenograft tumor models are now in use to follow tumor development before, during and after drug treatment with potential anticancer agents.
MATBIII-GFP cells implanted in mammary glands of nu/nu mouse.
Abstract ID: 284
A majority of tumors as seen from PET imaging studies are characterized by high rates of glycolysis. Tumors expressing elevated aerobic glycolysis (Warburg Effect) are highly aggressive and frequently associated with poor prognosis in patient recovery. It is not yet fully understood what mechanism drives this phenotype, but several oncogenic factors have been implicated. HIF-1 (hypoxia inducible factor) has generated interest because it becomes activated in low oxygen conditions associated with tumors and subsequently triggers the production of survival factors including genes responsible for upregulating glucose uptake and metabolism. We have previously shown a possible causal relationship between HIF-1a expression and aerobic glycolysis in metastatic breast cell lines MDA-mb-231 and MDA-mb-435. These results were expanded in a similar study which investigated low passage breast tumor lines from 12 patients. Glucose uptake and lactate production along with mRNA levels of GLUT-1 and GLUT-3 glucose transporters were assessed and compared to HIF-1α expression. Our results suggest that a subset of these tumors demonstrate a glycolytic response to HIF-1α expressed in the nuclear fraction. The lines were further examined for the expression of other possible molecular factors associated with the Warburg Effect including phosphorylated Akt, and phosphorylated c-Myc.
Abstract ID: 285
Bone metastasis is a frequent complication, occurring in patients with advanced breast cancer. To monitor the metastasis during tumor development, the highly metastatic MDA-MB-435s human breast carcinoma cell line was engineered to constitutively express herpes simplex virus type 1 thymidine kinase (HSV1-tk) and luciferase. After establishment of MDA-MB-435s/tk-lu stable clone, MDA-MB-435s/tk-luc cells were implanted into the mammary fat pad of 6-week-old female SCID mice and monitored for tumor growth and spontaneous metastasis. For lung and bone metastases, cells were injected into 6-week-old female SCID mice via the lateral tail vein and the left ventricle of heart, respectively. Multi-modalities of molecular imaging were applied to assay reporter gene expression. Mice bearing MDA-MB-435s tumors were imaged for tumor growth and osteolytic lesions using Xenogen in vivo imaging system. The signal of optical imaging was subsequently confirmed by MicroSPECT and whole-body autoradiography. Therefore, this new MDA-MB-435s/tk-luc breast cancer animal model would be useful in understanding the molecular mechanism of disease progression and for the evaluation of the new developed drugs as well as the new therapeutic trials.
Poster Session: Imaging Signal Transduction and Angiogenesis
Abstract ID: 286
Abstract ID: 287
The transcription factor NF-κB is a key regulator of cellular activation, proliferation and apoptosis. Recent data suggest that defects in the NF-κB pathway contribute to a broad array of malignant, neurodegenerative and chronic inflammatory diseases. NF-kB activity is regulated by IKK-dependent phosphorylation and subsequent proteasomal degradation of its inhibitor Iκbα. Thus, this critical regulatory point is emerging as an important target for drug development. To directly monitor regulation of IKK activation in intact cells and living animals, we engineered an IκBα-firefly-luciferase (IκBα-FLuc) fusion reporter. In real-time, the reporter displayed rapid, ligand-induced degradation followed by a gradual increase in bioluminescence that shared the same phosphoserine-dependent kinetic pattern of degradation and re-synthesis as endogenous IκBα. In stably-transfected cells, in tumor xenografts and in living mice expressing hepatic IκBα-FLuc by somatic gene transfer, the reporter provided a non-invasive readout of the kinetics and concentration-dependent activity of selective inhibitory drugs of both IKK (i.e., Bay 11–7085, IKK-NBD peptide and PS-1145) and the 26S proteasome (i.e., MG-132 and Bortezomib). For drugs acting on or upstream to IκBα degradation, excellent correlation was shown between IκBα-FLuc degradation and conventional NF-κB reporter-gene assays. Finally, using IkBa-Fluc tumor xenografts, full time- and dose-response profiling of a novel IKK inhibitory drug candidate (PS-1145) was determined in vivo through use of implanted micro-osmotic pumps for continuous delivery of D-luciferin (Fig. 1).
This novel IκBα-FLuc reporter now enables continuous readout of IKK activation in real time in vivo, facilitates development and validation of target-specific therapeutics, and complements conventional NF-κB transcriptional reporters for more complete temporal and regional investigations of the NF-κB signaling pathway in health and disease.
Time- (A) and dose- (B) responsive inhibition of IKK in tumors of PS-1145-treated mice.
Abstract ID: 288
3′-18F-fluoro-3′-deoxy-fluorothymidine (18F-FLT) is a nucleoside analog being developed for non-invasive imaging of cellular proliferation. Imaging drug-induced changes in proliferation potentially could be used to identify tumors responding to molecularly targeted cytostatic therapies. To test this concept, the effect of erlotinib, a specific epidermal growth factor receptor (EGFR) inhibitor, on 3H-FLT uptake was evaluated in parental A431 and MDA-468 cells. Following 10 μM erlotinib treatment, 3H-FLT uptake was significantly suppressed in sensitive A431 cells (23 ± 12% relative to control treatment) as compared to resistant MDA-468 cells (57 ± 5%; p=0.03). Suppression of FLT uptake in A431 cells correlated with reduced thymidine kinase 1 (TK1) expression (13 ± 11% relative to control, p=0.002), and activity (11 ± 1% relative to control, p< 0.001). Consistent with the importance of TK1 activity for cellular FLT trapping, inhibition of TK1 expression by siRNA suppressed FLT uptake in both A431 (6 ± 2% relative to control; p<0.001) and MDA-468 cells (7 ± 5% relative to control; p=0.002). In mice bearing A431 xenografts, the effect of erlotinib treatment on 18F-FLT uptake was evaluated by PET imaging. While placebo treatment was associated with a median 1% increase in tumor SUV, erlotinib treatment resulted in a median 20% decrease in SUV (p = 0.001). Decreased FLT uptake following erlotinib treatment was associated with decreased proliferation (labeling index 75 ± 6% for placebo treatment vs. 58 ± 10% for erlotinib treatment; p=0.008) and significant inhibition of tumor TK1 kinase activity (420 ± 18 cpm/μg vs. 83 ± 44 cpm/μg; p=0.002). These data suggest that 18F-FLT PET could be used to monitor the effects of molecularly targeted cytostatic agents on tumor proliferation.
Abstract ID: 289
The Epidermal Growth Factor Receptor (EGFR/Her-1/) is involved in proliferation and differentiation of normal and malignant cells. Overexpression of EGFR is present in a variety of human cancers. Inhibition of EGFR-TK has emerged as a major approach for cancer targeted therapy. Consequently, there has been a considerable interest in the use of radioactively labeled EGFR inhibitors as bioprobes for Positron Emission Tomography (PET) molecular imaging of EGFR overexpressing tumors. While several classes of non-labeled and labeled EGFR inhibitors have been developed as potential anticancer drugs and molecular imaging agents, these inhibitors exhibited poor solubility, bioavailability, and biological stability. In order to overcome these problems, tetraethyleneglycol was attached at the 7-position of the quinazoline ring; in addition, fluorine was added to the PEG residue to enable labeling with fluorine-18. A multistep synthesis has been developed for synthesizing compound
Compound
Abstract ID: 290
Abstract ID: 291
Overexpression of the epidermal growth factor receptor (EGFR), a member of the ErbB receptor family, is a frequent hallmark of human carcinomas, and its aberrant signaling is associated with neoplastic cell proliferation, migration, stromal invasion, angiogenesis, resistance to apoptosis, and an overall more aggressive clinical course. Despite its detrimental role in the initiation and progression of epithelial tumors, no currently available assay can provide adequate in vivo quantitation of receptor levels. Consequently, low molecular weight EGFR tyrosine kinase inhibitors of the anilinoquinazoline family were harnessed for in vivo PET molecular imaging of EGFR overexpressing tumors.
Labeling these inhibitors with proper positron-emitting isotopes, such as C-11, F-18, Br-76 or I-124 could offer a four-dimensional, quantitative measurement of EGFR overexpressing carcinomas.
Previous reports on imaging of EGFR overexpressing tumors, using the labeled acrylamido anilinoquinazoline EGFR irreversible inhibitor, [11C]ML03, demonstrated a rapid metabolism of the tracer, which led to its low in vivo accumulation in these tumors [Ortu et al, 2002, Int J Cancer (4)].
In order to increase the biological stability and enhance the tumor uptake, the chemical structure of the compound was modified, and three new groups of EGFR inhibitors, with a wide range of chemical reactivities, were synthesized. Chemical reactivity assay of the compounds, performed with reduced glutathione (GSH), revealed that group 3 (4-dimethylamino-but-2-enoic amide) derivative was the least chemically reactive against the nucleophilic attack of GSH. Nonetheless, it demonstrated a high inhibitory potency and bound irreversibly to the EGFR. The superior stability of group 3 compounds was further demonstrated by comparative blood-stability assays, using I-124 as a radiolabel. Both the chemical and the blood stabilities of group 3 derivatives were significantly higher as compared to [11C]ML03. Since group 3 offered greater promise for tumor accumulation, it represents, to date, the most suitable candidate for radiolabeling with long-lived PET radioisotopes.
Abstract ID: 292
Abstract ID: 293
Conventional MR contrast agents are primarily extracellular and are constitutively active. The advent of entirely new classes of agents with increasing sophistication for targeting and biactivation is expanding the types of experiments performed by MRI. In this report, we introduce the first steroid hormone-MR contrast agent conjugate which is designed to monitor the activation of a specific signal transduction pathway by MR imaging. We have synthesized a contrast agent covalently attached to RU-486. The Gd(III) chelate conjugated to RU-486 will selectively bind to the mutant receptor and activate the gene expression pathway. In addition, it is expected to increase the rotational correlation time (τr) upon binding to the receptor protein resulting in further enhancement of the contrast.
The relaxivity of
Abstract ID: 294
Abstract ID: 295
Activation of the hedgehog (Hh) signaling pathway has been implicated in multiple cancers including glioblastoma, medulloblastoma, prostate cancer, breast cancer, and basal cell carcinoma. Inhibition of Hh signaling may therefore provide a novel therapy for those cancers and provides the impetus for discovery of new pharmacologic agents to inhibit Hh signaling. We used bioluminescence imaging (BLI) and planar gamma scintigraphy to assess the modulation of Hh inhibitors on the Hh pathway in vivo, through Gli, a downstream transcription factor in the Hh cascade. A tri-fusion multimodality reporter gene (fluc-mrfp-ttk) was engineered under control of the Gli promoter. In vitro co-transfection assays demonstrated that the reporter is specifically responsive to Gli. Murine U87 (glioblastoma) xenografts stably transfected with the Gli-trireporter could be imaged with BLI in vivo. Treatment of U87 cells stably transfected with the Gli-trireporter with Hh-Antag691, an inhibitor of Hh signaling, in vitro decreased luciferase signal by 70% relative to control while causing only a 30% cell kill. In 22RV1 (prostate cancer) cells, Hh-Antag691 decreased luciferase activity by more than 70%, but no cell kill was observed. Those findings are in keeping with the expected lower concentration of drug required to inhibit the pathway than to kill the tumor cells and suggests that we have developed a sensitive reporter system for assessing Hh activity modulation in vivo. Studies to assess the effect of Hh-Antag691 and other pathway inhibitors in vivo with BLI and gamma imaging are under way.
Abstract ID: 296
Epstein-Barr virus (EBV) is a human herpesvirus that is associated with a number of malignancies including certain forms of lymphoma and gastric cancer. We have proposed activation of the lytic cycle of EBV to treat EBV-associated tumors and describe a method by which that process can be imaged in vivo. We used [125I]-1-(2′-deoxy-2′-fluoro-beta-D-arabinofuranosyl)-5-iodouracil ([125I]FIAU) as a reporter probe for the EBV thymidine kinase (TK). In vitro cell uptake assays with [14C]FIAU and ex vivo bio-distribution studies with [125I]FIAU showed that uptake and retention of [125I]FIAU is highly specific for cells that are EBV-TK-positive upon lytic infection but not when virus is in the latent state. Planar gamma imaging of tumor-bearing SCID mice show high levels of [125I]FIAU localized within tumors during lytic induction with the proteasome inhibitor bortezomib (Velcade). That effect was demonstrated in xenografts derived from the human lymphoma cell lines rael and akata. We saw increased [125I]FIAU uptake within tumor of 0.85% ID/g in Akata at 96 hours after injection of [125I]FIAU. These results indicate the feasibility of imaging lytic induction during the treatment of EBV-associated tumors with Velcade in vivo, and may provide a simple way by which this therapy could be monitored in the clinic.
Abstract ID: 297
Evaluating the preclinical therapeutic responses of transgenic mice forming tumors in situ is a complex task because tumor onset is variable and tumor site may vary, often juxtaposed to bone. Contrast-enhanced small animal computed tomography is a promising economical and highly quantitative tool for serially assessing tumor vasculature in living mice; however, practical guidelines remain to be established. Using a long-acting iodinated triglyceride blood pool contrast agent, we developed optimized scanner acquisition parameters and volume rendering techniques for examining the intermediate and large vessels of tumors in transgenic mice. We found that multiple-frame, 360–720 view acquisitions were mandatory for clarifying bone and soft tissue from vessel contrast. This finding was consistent in volume renderings using a 1-dimensional transfer function where voxel color and opacity were assigned in proportion to density, and a 2-dimensional transfer function where voxel color and opacity were assigned in proportion to density and gradient magnitude. This study lays groundwork for the qualitative and quantitative assessment of antiangiogenesis therapies in preclinical studies using transgenic mice.
Abstract ID: 298
There are numerous studies correlating hypoxia and angiogenesis in tumor progression. However, recent reports show development of new vessels from the periphery of the tumor where tumor and the surrounding tissues interact. We hypothesized that in addition to hypoxia there are other factors in the tumor and surrounding tissues that may be important for differential migration of endothelial progenitor cells (EPCs) for tumor angiogenesis. To address this issue we implanted 1 × 106 tumor cells (glioma) that were pre-mixed with 105 iron-labeled EPCs (dead or alive) in the right flank of nude mice. In vivo and ex vivo MR images were acquired at different stages of tumor development. Following ex vivo MRI, tumors were sectioned at 6–10 μm, stained for iron (Prussian blue staining) and different factors, such as HIF1α, VEGF, MMP-2, human VEGFR2, human vWF and human CD31. MR images showed differential migration of implanted live EPCs towards the periphery of the tumor but dead labeled EPCs did not migrate. Iron positive cells were seen at the site of tumor invasion into muscles. The migration of the cells towards periphery, where tumor interacts with surrounding tissues, was most visible in larger tumors. Iron positive EPCs that had migrated peripherally showed expression of endothelial cell markers. MMP-2 was seen all over the tumor as well as in the surrounding tissues. The central as well as sub-peripheral tumor cells expressed HIF1α. Bands of tumor cells near the periphery showed HIF1α expression but few cells expressed VEGF. Migration of labeled cells was seen slightly away from the area of VEGF expression. In conclusion, we have shown that MRI can be utilized to detect the differential migration of implanted EPCs in tumor and to elucidate the cell signaling pathways in tumor angiogenesis.
Abstract ID: 299
In peripheral vascular disease (PVD), X-ray angiography can reliably detect the proliferation of mature vessels in response to L-arginine (LA) therapy about 40 days after the ischemic insult. The objective of this study was to determine whether molecular imaging with αvβ3-integrin-targeted paramagnetic nanoparticles (NP) could detect the neovascular response to LA earlier than traditional x-ray angiography.
Abstract ID: 300
The purpose of this study is to monitor metabolic changes and tumor progression of a rat glioma model following treatment with our non-viral gene delivery system using αvβ3 targeted antiangiogenic nanoparticle (NP).
Rats were divided into 3 groups. Treated and control groups (Groups 1 and 2) received RT2 tumor cells implanted intracranially into the striatum. Normal group (Group 3) received no tumor cells or treatment. Group 1 received an αvβ3 targeted NP that carried a mutated Raf gene (NP-ATPμ-Raf) complex. This gene disrupts the VEGF and FGF mediated signaling pathway. Treated animals were imaged with 18F-FDG PET to monitor glucose uptake, and with T2-weighted MR sequences to record tumor volume. Animals in the control and normal groups were imaged on microPET and MR at the same dates when the treated animals were imaged.
Treated animals showed an increased FDG uptake in brain during the treatment period, followed by a constant decrease till the end of study. Control animals had a high FDG uptake over the course of study; and most animals died 10 days post tumor implantation. On MRI, both the treatment group and the control group demonstrated tumors in the beginning of the treatment. The tumor in the controls continued to increase in size and signal intensity. Tumors in the treated group showed reduction in tumor size. Animals with no tumor had normal PET and MRI images.
MRI and microPET imaging combined can be powerful to monitor tumor progression and to follow therapy. MR images showed that the NP slowed tumor progression with subsequent encephalomalacia, while microPET images showed that this agent caused an initial increase then a significant decrease in metabolic activity in brain tumors.
Abstract ID: 301
Hypoxia is a common feature of various human tumor entities. Hypoxic tumor tissue can be studied by nuclear medicine imaging techniques like PET and SPECT. In this study we analyzed in vitro the influence of experimental hypoxia on the radiotracer uptake of various primary endothelial cells, which play an essential part in tumor angiogenesis.
Experimental hypoxia was induced by cultivating cells for 24 hours in presence of 2% oxygen in a special incubator. Cellular uptake of [99mTc]MIBI, [18F]MISO and [18F]FDG was determined after one or four hours incubation and measured after cell lysis with a gamma counter. Three types of primary endothelial cells were used: human umbilical vein endothelial cells (HUVEC), human dermal microvascular endothelial cells (HDMEC) and human aortic endothelial cells (HAEC) as well as the two tumor cell lines (FaDu and HT29). Cellular expression of hypoxia-inducible factor-1a (HIF-1a) and vascular endothelial growth factor (VEGF) was monitored with quantitative PCR.
In our experiments we found that under hypoxic conditions the uptake of [99Tc]MIBI was decreased in all cell types. The uptake of the hypoxia-specific radiotracer [18F]MISO was slightly increased in all primary endothelial cell types. Furthermore, primary endothelial cells incorporated significant higher amounts of [18F]FDG under experimental hypoxic conditions in comparison to normoxic conditions. Especially HDMECs showed a marked response to hypoxia with an approximately two-fold higher [18F]FDG uptake. Also HUVECs and HAECs responded to experimental hypoxia, whereas tumor cell lines FaDu and HT29 showed only moderate increase of [18F]FDG uptake under hypoxic conditions.
We conclude that experimental hypoxia represents a potent stimulus for primary endothelial cells to accumulate [18F]FDG and [18F]MISO in vitro. Our data emphasize the relevance of endothelial cells as one important part of the tumor micromilieu and stimulate further studies on the different patterns of radiotracer uptake in neoplastic or neovascularized lesions.
Abstract ID: 302
Changes in R2 and R2*, following USPIO administration, in murine brain tumors.
Abstract ID: 303
Endothelial cells in tumor vasculature overexpress VEGFR-2, a major receptor for vascular endothelial growth factor (VEGF). Since VEGF/VEGFR-2 pathway is crucial for tumor angiogenesis, it is a primary target for potential anti-cancer therapeutics. However, little is known about patterns of VEGFR-2 expression at various stages of tumor development or therapeutic treatments. In part, it is due to the paucity of targeted constructs that can selectively bind to VEGFR-2 in tumor vasculature. We reasoned that VEGF could be used as a targeting protein for selective imaging of tumor neovascularization and its responses to various anti-cancer treatments.
To optimize VEGF for molecular targeting, we cloned two fragments encoding VEGF121 monomers as a single-chain VEGF (scVEGF) that was expressed as a single polypeptide and refolded to mimic a VEGF dimer. To optimize construction of imaging and/or therapeutic conjugates, scVEGF was armed with an N-terminal peptide tag, containing a unique cysteine residue for site-specific conjugation. Near infrared dyes Cy5 or Cy5.5 were conjugated to either VEGF or scVEGF. The resulting VEGF/Cy conjugates retained functional activity comparable to that of parental VEGF in tissue culture assays, such as induction of VEGFR-2 tyrosine autophosphorylation and competition with a VEGF-toxin for binding to cellular VEGFR-2.
In several mouse tumor models, VEGF/Cy constructs selectively accumulated in the tumor area, as visualized by near-infrared fluorescent imaging. VEGFR-2 overexpressing cells are implicated in this effect, since it was dramatically decreased in mice pretreated with a VEGF-toxin that selectively eliminates such cells. Selective accumulation was detectable as early as four days after subcutaneous implantation of 10,000 mammary mouse adenocarcinoma luciferase-expressing 4T1luc cells and persisted for 4–5 days after VEGF/Cy injection providing opportunities for detection of vascular remodeling.
We expect that VEGF/Cy conjugates will be useful in development of anti-angiogenic compound in animal models and, perhaps, in monitoring “close-to-surface” angiogenesis in humans.
Abstract ID: 304
Calreticulin (CRET) is a calcium-binding protein with known anti-angiogenic activity. We investigated the effect of CRET gene transfer on endothelial cell proliferation in vitro and tumor growth, perfusion and glucose utilization in vivo.
A Morris hepatoma (MH3924A) cell line stably expressing CRET (MHCRET) was generated by transfection of the CRET and the hygromycin resistance gene and subsequent hygromycin selection. The proliferation of human umbilical vein endothelial cells (HUVECs) stimulated with basic fibroblast growth factor (bFGF) or vascular endothelial growth factor (VEGF) was tested in the presence of MHCRET or MH3924A wild type (MHWT) supernatant. After inoculation of 2 × 106 cells into the hindlimb of ACI rats tumor growth of MHCRET and MHWT was monitored. At tumor diameters >1 cm, perfusion and glucose metabolism were assessed by dynamic PET with H215O (40–60 MBq i.v.) and 18F-FDG (10–20 MBq i.v.), respectively. Regions of interest over the hearts and tumors were analyzed in 1-tissue and 2-tissue compartment models, respectively.
The proliferation of bFGF-or VEGF-stimulated HUVECs was significantly inhibited in presence of MHCRET-conditioned medium. In vivo, growth of MHCRET was inhibited >60% in comparison with MHWT. Kinetic modeling of the H215O perfusion data showed 2-fold higher median k1 and 1.5-fold higher median k2 rate constants in MHCRET (n=5) than in MHWT (n=5), respectively. Modeling of the FDG data showed a higher VB in MHCRET than in MHWT tumors whereas the rate constants K1, k2, and k3 were comparable for MHCRET and MHWT, respectively. SUVs were comparable between MHCRET and MHWT (2.3 vs. 2.5).
In vivo the perfusion of the MHCRET is increased in comparison with WT although growth of MHCRET is significantly inhibited and although in vitro experiments confirm direct antiangiogenic effects of CRET on endothelial proliferation. The growth inhibition of MHCRET seems to depend on mechanisms not related to perfusion.
Poster Session: Imaging in Neuroscience, Part 1
Abstract ID: 306
In vivo detection of aggregated amyloid peptides (Aβ) (amyloid plaques) presents targets for the development of biological markers for Alzheimer's disease (AD). In an effort to fabricate in vivo probes, polymeric n-butyl-2-cyanoacrylate (BCA) nanoparticles (NPs) were encapsulated with the radiolabelled amyloid affinity drug 125I-clioquinol (CQ, 5-chloro-7-iodo-8-hydroxyquinoline), and evaluated. 125I-CQ discriminately binds AD post-mortem brain tissue homogenates, versus control. Additionally, 125I-CQ-BCA NPs labeled the Aβ plaques from AD human post-mortem frontal cortical sections, on paraffin-fixed slides. Storage phosphor imaging verified preferential uptake by the AD brain sections, compared to cortical control sections. 125I-CQ-BCA NPs cross the blood/brain barrier (BBB) in the wild type mouse, with a significantly increased brain uptake (%ID/g), compared to 125I-CQ. Brain entry and clearance of 125I-CQ-BCA NPs was rapid, demonstrating ideal in vivo imaging characteristics for small animal modalities; good clearance of free and non-specifically bound radioisotope affords high-quality temporal resolution, and good signal-to-noise. 125I-CQ-BCA NPs have specificity for the Aβ plaques in postmortem tissue, and have a rapid brain entry and clearance. This combination makes 125I-CQ-BCA NPs a promising candidate as an in vivo SPECT (123I), or PET (124I) amyloid imaging agent. AD mouse models, double transgenic for Aβ −40 and −42, are currently being treated with 125I-CQ-BCA NPs. in vivo imaging experiments will be presented.
Phosphor screen image of human post-mortem AD frontal cortex (A) and control cortex (B) incubated with 125I-CQ-BCA NP, containing 2 × 106 CPM radioactivity (in 0.5 mL buffer per fixed slide; Incubation time: 30 min; Film exposure: 7 min).
Abstract ID: 307
Abstract ID: 308
Abstract ID: 309
Abstract ID: 310
MR is the test of choice to support clinical diagnoses of many neurodegenerative diseases, including multiple sclerosis (MS), the most common autoimmune demyelinating disease of the central nervous system (CNS). However, the lesions in the white matter of MS patients can have a similar macroscopic appearance in MR images to lesions from other etiologies, such as ischemia, sarcoidosis, and others. Furthermore, lesions in the gray matter are often undetected by standard MR methods.
There is a need for novel contrast agents more specific for cell types or ligands relevant to neurodegenerative diseases. One of our novel MR stains, luxol fast blue (LFB), is a widely-used histological stain for myelin. The molecular structure of LFB contains a paramagnetic Cu2+ core and influences the relaxation rates of water molecules within ex vivo samples. At 14T, we have observed selective changes in the longitudinal relaxation times: a 47% decrease in white matter and a 30% decrease in gray matter. Our preliminary results show the contrast-to-noise ratio (CNR) between white and gray matter for samples immersed in LFB to be enhanced by 1.74 over samples that have only been formalin-fixed. As shown below in Figure 1, this contrast enhancement is superior than the 1.27 increase for samples immersed in Gd-DTPA. Thus, LFB has the potential to be more sensitive to lesions in white matter or cortical demyelination than standard gadolinium-based contrast agents. Future work will use high-resolution MR to compare the structural changes in post mortem MS brains to the appearance of lesions resulting from other disorders.
Abstract ID: 311
“White Marker” (WMk) MRI [1] has recently been adapted to produce positive contrast from superparamagnetic iron-oxide (SPIO) labelled cells [2]. A similar approach, using an off-resonance spin-echo is also being explored [3]. The WMk technique uses a dephasing gradient to spoil signal throughout the sample, save regions where local gradients are of the right amplitude and orientation to refocus the lost signal. The resulting image is void everywhere except in small volumes surrounding paramagnetic sources (ie: iron-loaded cells), which are bright.
The advantage of WMk is the increased specificity that accompanies this positive label. In conventional MRI, the presence of cellularly compartmentalized iron-oxide is indicated by local signal hypo-intensities in T2 or T2* weighted images. A disadvantage of this form of cellular imaging is that endogenous signal loss can mimic the appearance of cellularly-compartmentalized iron oxide. We have chosen a rat model of SCI to showcase such difficulties, and the advantage WMk provides (Figure 1).
Because white marker uses a 2π dephasing gradient, it is necessary that the sample maintain uniform intra-voxel proton density for the spoiler to be fully effective - spins on one side of the voxel cancel those on the other. Thus, at tissue surfaces, partial voluming effects can create false positives (Figure 1). In this abstract we demonstrate that SPIO-refocused and “partial volume” signals disperse differently, suggesting that this property can be used to distinguish between the two effects (Figure 1,2).
(Left) Single MR slice of a rat spinal cord injury. At longer TE, hemorrhage causes signal loss. (Right) Same as left. At 24h post-injury, Ferridex was administered via tail-vein injection, accumulating in leukocytes at the site of inflammation. Unequivocal distinction between controls and experimental models can be difficult with conventional sequence, such as SPGR (a,b). The relaxation differences between tissues in WMk imaging (c,d,e) can be used to overcome such problems.
(Top) Average white marker signal vs. echo time for voxels containing a single 4 pg / 10 pg SPIO bead embedded in gelatin. (Bottom) Avg WMk signal vs. TE for voxels containing hemorrhage / partial volumes in a rat control (ie: non SPIO-injected) model of SCI. Error bars correspond to standard deviation.
Abstract ID: 312
Drug-delivery to central nervous system has long posed a major challenge. Direct intracranial delivery also called convection-enhanced delivery (CED) has been developed as a drug delivery strategy to overcome this difficulty. Monitoring drug delivery by means of MR imaging technology is of considerable interest for any clinical application of CED. This would allow tracking drug distribution in real-time, therefore avoiding undesired distribution, reflux and visual validation of successful drug delivery to target region. We have developed a MPI-based method to visualize the CNS distribution of liposomal gadolinium in rodents and non-human primates. MPI of liposomal gadolinium is highly accurate at determining tissue distribution, as confirmed by comparison with histological results from concomitant administration of fluorescent liposomes. To further develop this nanotechnology-based strategy for clinical application, we monitored liposomal distribution within certain anatomic structures and determined the lowest possible liposomal gadolinium concentration that would permit MPI visualization. Non-human primates were infused with liposomal gadolinium into brain stem, corona radiata and putamen. Co-infusion of liposomal gadolinium and chemotherapeutics was performed in dogs to monitor distribution. Results indicate that liposomal volume of distribution is largely affected by the structure infused. Mixing liposomal gadolinium with liposomal chemotherapeutics does not alter the volume of distribution on MPI. Real-time monitoring of infusion procedure greatly enhances the efficacy and safety of delivery. Current clinical CED applications in neuro-oncology and Parkinson's disease lack monitored drug delivery. Introduction of RCD would very likely reduce side effects seen in such applications, and moreover could enhance efficacy of drugs. RCD will give neuro-imaging a whole new active role in therapy, beyond merely confirmatory imaging of clinical outcomes.
Abstract ID: 313
The discovery of appetite-regulating neuropeptides is providing new insights into the mechanisms associated with appetite and satiety. Although c-Fos expression in vitro can identify regional brain activation, studying temporal response is still not possible. Manganese-enhanced MRI (MEMPI) has been proposed to delineate neuronal connectivity and ultrastructures. Here we demonstrate its novel use in determining neuronal activation associated with appetite.
Spin-echo T1-weighted sequence was carried out in male C57BL/6 mice before and after intravenous (IV) administration of 100 mM MnCl2 (5 ml/kg, 0.2 ml/h). Ghrelin (0.06 and 0.30 nmol/g), or the vehicle solution (PBS) was infused intraperitoneally (IP) to fed mice; while peptide YY (PYY3–36) (100 μg/kg) or PBS to fasted mice. After 3 baseline scans, peptides or PBS was infused with simultaneous IV MnCl2, and 63 scans continuously recorded. The signal intensities (SI) of the arcuate (Arc), the periventricular (Pe), the ventromedial (VMN), the paraventricular (PVN) hypothalamic nuclei were measured.
Ghrelin and PYY3–36 infusion led to significant region-specific modulation in the Mn2+ enhancement. In the Arc (Fig 1), higher enhancement was observed at both doses of ghrelin, as compared to PBS. In the VMN (Fig 2), a dose-response modulation by ghrelin was shown. In addition, PYY3–36 led to decrease enhancement compared to fasted controls. We conclude that Mn2+ enhancement can be modulated by the gut-peptides in a region-specific manner. More importantly, the findings correlate well with those reported by c-Fos. This suggests that MEMRI is an invaluable tool to evaluate hypothalamic activation in vivo.
Abstract ID: 314
Transplantation of neural precursor cells (NPCs) is a promising therapeutic strategy for the treatment of CNS injuries and neurodegenerative disorders because of the potential for cell replacement. We have been using neuronal-restricted precursors (NRPs) and glial-restricted precursors (GRPs), NPCs with lineage restrictions for neurons and glia, respectively, and have shown their robust survival, migration and differentiation following grafting into adult CNS. However, an important element of future clinical applications is the development of a non-invasive procedure to follow the fate of NPC transplants. We show here that NRPs and GRPs can be labeled in vitro with the superparamagnetic iron oxide contrast agent Feridex and poly-L-lysine. Following engraftment into the intact adult spinal cord (n=6), the labeled cells robustly survived in both white and gray matter and migrated selectively along white matter tracts for distances of up to 5 mm. The localization of the graft-derived cells was reliably established using ex vivo magnetic resonance imaging of recipient spinal cords, and the imaging coincided with histological detection of the human placental alkaline phosphatase transgene. Transplanted, labeled cells elaborated mature morphologies and differentiated into all 3 mature CNS cell types, neurons, astrocytes and oligodendrocytes, similar to grafts (n=3) of unlabeled NRPs and GRPs. In conclusion, Feridex-labeling does not interfere with normal neural differentiation. NPCs can be reliably tracked using MRI following engraftment into the adult CNS, which is important for the clinical application of transplantation strategies in the treatment of spinal cord injury and other CNS pathologies.
Abstract ID: 316
5-I-A85380 is a pyridyl ether based nicotinic acetylcholine receptor ligand, binding predominantly to the α4β2 subunit, that has been developed for use in SPECT molecular imaging studies (Figure1). SPECT imaging in depressed patients using 5-123I-A85380 could provide an understanding of the involvement of the nicotinic system in depression and antidepressant treatment. However before the 5-123I-A85380 images from SPECT studies investigating antidepressant therapy can be interpreted, investigation into the effect of antidepressants on 5-I-A85380 binding is required. The present study examined the effects of antidepressants on the binding of 5-125I-A85380 to human post mortem brain tissue and after chronic administration in rats using in vitro autoradiography.
5-125I-A85380 was synthesised via electrophilic iododestannylation with a high isolated radiochemical yield and high radiochemical purity. For the humans post-mortem study sections of hippocampus, thalamus, basal ganglia and cortex were incubated in the presence or absence of commonly prescribed antidepressant drugs. Chronic treatment of rats involved injection of venlafaxine (10mg/kg, i.p) for 14days. Autoradiographs obtained were analysed by densitometry.
None of the antidepressant drugs studied blocked the binding of 5-125I-A-85380 to nicotinic receptors in the human brain. In addition venlafaxine administration to rats had minimal effects on 5-125I-A85380 binding to nicotinic receptors in any region studied.
These findings indicate that data obtained from SPECT imaging studies of depressed patients who are taking antidepressant drugs at the time of the scan are unlikely to be confounded by concurrent medication. 5123I-A85380 SPECT imaging studies in medicated depressed patients can now be performed to explore the role of nicotinic receptors clinical depression and treatment response.
A saggital section of a 5-123I-A85380 SPECT scan in the human brain.
Abstract ID: 317
5-I-A85380 is a pyridyl ether based nicotinic acetylcholine receptor ligand, binding predominantly to the α4β2 subunit, that has been developed for use in SPECT molecular imaging studies. High affinity nicotinic receptors are concentrated in the thalamus, an area implicated in control of consciousness. Disturbances of consciousness (DOC) are a common feature of dementia with Lewy bodies (DLB), and one which impacts significantly on daily function. Results from a previous study found DOC to be related to an increased level of high affinity nicotinic receptors in the temporal cortex. This study uses 125I radiolabelled 5-I-A85380 to investigate thalamic nicotinic receptor density in post-mortem tissue from DLB patients with and without DOC and controls.
Radiolabelled 5-125I-A85380 was prepared from the corresponding trimethylstannyl compound by electrophilic iododestannylation. The radioiodination was achieved in good isolated yield (66.8 % ± 4.75, n = 3) and after careful formulation with reasonable stability.
In vitro autoradiography found 5-125I-A85380 binding was higher in patients with DOC compared to patients without DOC in all thalamic nuclei examined, reaching significance in the reticular and ventral anterior thalamic nuclei. Patients with DOC had similar binding levels to controls. Reduced binding was seen in patients without DOC compared to controls in all thalamic nuclei, reaching significance in the ventral anterior nucleus.
These findings demonstrate a relative preservation of nicotinic receptor density in DLB patients with DOC, thereby implicating nicotinic modulation of thalamo-cortical circuitry as a key component in the control of conscious awareness in DLB. This may be important in terms of treatment and highlights the need for intervention trials with cholinergic therapies in DOC. A SPECT imaging study in DLB patients is needed to confirm this result in vivo.
Abstract ID: 318
Alzheimer's Disease (AD) is primarly a disorder of brain metabolism. Disturbances in the amyloid metabolism can lead to the formation of amyloid plaques in the brain with selective vulnerability of cortex and hippocampus. This causes degeneration of neurons but also affects axons. Therefore it is reasonable to anticipate that white matter is also damaged.
Changes in the microstructural organization of the brain are reflected in the molecular motion of tissue water, detectable by diffusion measurements. The purpose of this study was quantifying changes in magnitude (=mean diffusion) and direction (=fractional anisotropy=FA) of the molecular mobility of brain tissue water in a mouse model of diffuse intracellular β-amyloid plaque deposition associated with AD (n=4) and age matched (20months) controls (n=4).
Diffusion Tensor Imaging (DTI) was performed on a 7T-system (Bruker) using multislice SE-EPI, a circular surface coil and diffusion sensitizing gradients along 30 directions. Mice were anaesthetized with isoflurane (0.4–0.7%). Core temperature(37°C) and breaths per minute(150) were monitored.
FA values (numerical value between 0–1) in white matter structures of APP mice were significantly reduced as compared to control values: corpus callosum (cc 0.64vs0.50) and cerebellum (0.34vs0.25). This clearly reflects the loss of neuronal connectivity. Due to the high resolution (136*136*700μm3) of the EPI images, we could demonstrate the apparently low anisotropy of gray matter and were able to identify a reduction in FA values for the cortex (0.22vs0.18) and hippocampus (0.22vs0.17). This reflects changes in underlying cellular architecture. The mean diffusion remained unaltered.
This work demonstrates that altered FA values could represent a reliable and non invasive MRI tractable biomarker for AD.
Note the strong similarity of anatomic features on the FA maps compared to the raw EPI image slice (right). Calculated FA maps (expressed as numerical value between 0–1) of two consecutive acquired slices at bregma −3mm and bregma −2mm show more detailed anatomy (ct: cortex, dhc: dorsal hippocampal commissura, ec: external capsula, ce: corpus callosum, hippo: CA layers of the hippocampus). The colormap shows the characteristic directions of anisotrophy on a pixel per pixel base.
Abstract ID: 319
Abstract ID: 320
High-temperature superconducting (HTS) coil is one of the best ways to increase the signal-to-noise ratio (SNR). Bi2Sr2Ca2Cu3Ox (Bi-2223) tapes were suitable to use because of the easier fabrications and lower cost. Diffusion tensor imaging (DTI) can map the fiber structures non-invasively. The accuracy of the DTI highly depends on the SNR. We could find more detailed structures revealed in the image of using the HTS tape coil. DTI reconstruction was shown in Figure. With the same SNR achieved by conventional copper coil, organized cerebral cortex structures, corpus callosum and the hippocampus regions were revealed in a much shorter scan time. In our case, the acquisition time is reduced from 88 min to 11 min, which is tremendous in terms of rats throughput in phenotyping study for transgenic rats. The benefit of reduced scan time can be used to improve the spatial resolution to map more detailed neural fiber structures. The limitation of the voxel size would lead to the heterogeneity of the DTI mapping. Acquisition time of MR diffusion tensor imaging (DTI) can be reduced by factor of 9 for the same signal-to-noise. In summary, with this HTSC system, a 3T MR system could reach the high signal-to-noise of 12 T MR system with the advantage of less T2 shortening effects at high field. Currents researches are focused on brain connectivity, molecular imaging and fMRI studies.
(A) DTI of the rat brain using HTS coil, where the organized cerebral cortex structures, corpus callosum and the hippocampus were shown in the enlarged image (B).
Abstract ID: 321
A) Fluorogold staining of retinae after intravitreal injections of various doses of MnCl2. B) Mean number of RGC as a function of dose.
Abstract ID: 322
Abstract ID: 323
Non-invasive, serial monitoring of human embryonic stem cell-derived neural precursor grafts (hES-NPCs) in the central nervous system of living animals is critical for optimizing transplantation conditions suitable for the long-term survival of these and more tissue-restricted populations. We utilized optical bioluminescence imaging (BIT) and a lentiviral triple reporter gene construct (TGL) to assess neural precursor graft survival, localization, and proliferation in immunosuppressed mouse models and early graft failure in immunocompetent models. hES-NPCs, derived from undifferentiated hES cells according to our published protocols, were transduced using the TGL construct. No alteration in cellular morphology or function was observed.
Lentiviral-transduced hES-NPCs (100,000–175,000 luc+ cells) were stereotactically implanted into the striatum of both NOD-SOD and immunocompetent (129/sv) mouse models. In 129/sv mice, a ten-fold or greater increase in brain signal was observed over the first 5 days, followed by a dramatic decline in signal to near background levels at about 2 weeks. Conversely, NOD-SCID mice, studied over a 2 month period, revealed initial signal increases (4–5 days), a gradual decline in signal over the next 3 weeks, followed by subsequent stabilization of the brain signal over the final 4 weeks.
The histologic distribution of viable cells co-labeled with a human marker and anti-GFP antibody, corresponded with the BLI findings in both animal models. Scant positive staining for both markers was found in largely fragmented and shrunken 129/sv striatal grafts, while viable striatal grafts in NOD-SCID mice demonstrated numerous cells co-labeled for both human and anti-GFP markers. Cellular migration was observed within the host tissue, including the corpus callosum and subventricular zone, perhaps reflecting incorporation into the endogenous stem cell pool, as seen previously in our immunosuppressed rat models.
Abstract ID: 324
Cell labelling: The murine embryonic stem cell-lines D3 and L1 as well as the neural progenitor cell-line C17.2 were labelled with commercially available Gd-chelates (Magnevist® and Prohance®) by endocytosis or electroporation. Toxicity and proliferation were tested in cell suspensions. Visualisation limits were determined in phantoms and after stereotactic implantation in the brain of Wistar rats.
MRI: T1-weighted MR images were acquired on a Bruker Biospin 7.0 Tesla small animal scanner using 3D gradient echo sequences (FLASH; TR=120ms; TE=5ms; spatial resolution 75μm in all three dimensions).
We were able to show that Gd-chelates are suitable contrast agents for the detection of small cell clusters in the brain. This is an important condition for the development of responsive contrast agents to monitor functional changes of implanted stem cells in vivo. Further improvement in sensitifity of contrast agents will allow detection of small clusters of migrating cells.
Abstract ID: 325
Cell transplantation is widely used to reduce behavioural deficits after focal cerebral damage. We hypothesize that the adult mesenchymal stem cells (MSCs) can migrate after intravenous application to the lesion after intracerebral hemorrhage (ICH) which this can be visualized by species specific markers and by in vivo tracking by MRI.
The functional deficit was monitored with the Rota-Rod Test which was performed everyday in the training period 3 weeks before and on day 0, 1, 3, 7, 10, 14, 17, 21, 24 and 28 after ICH induction. Experimental ICH was induced by stereotactic administration of collagenase into the striatum of 20 adult rats. After one day rats were randomly assigned into control or treatment group. Treatment group received 3 × 106 adipogenic MSCs by intravenous injection in the tail vein while the control animals received a placebo injection with Ringer solution. Morphometric analysis of the hemorrhage and imunhisto-chemical analysis was performed afterwards.
The average value in the Rota-Rod Test before ICH induction was 23,1+/-3 rpm which decreased to 7,3+/-3 rpm on day 1. There was a highly significant improvement in functional outcome for the MSC treated group (p<0,01) improving to 17,5+/-2,4 rpm while the control animals were 11,5+/-4,2 rpm after 4 weeks. Also we were able to label our MSCs in vitro by gadolinium liposomes. First results showed a stable T1 positive signal over 4 weeks in our in vitro experiment.
This is the first report using human adipogenic mesenchymal stem cells as a highly effective treatment to improve functional outcome after ICH. Labelling the cells with gadolinium liposomes shows excellent results in vitro and is promising for in vivo tracking of the MSCs in further experiments.
Abstract ID: 326
In temperate zone songbird species, seasonal morphological and functional plasticity of the neural substrate of song behavior is accompanied by seasonal changes in song output.
Cellular events underlying the seasonal changes of the two pathways emerging from two separate cell types in the central song control nucleus HVC (abbreviation used as proper name) differ both in neuronal replacement capacities and the expression of steroid receptor types. The motor pathway projecting to nucleus robustus arcopallii (RA) starts from replaceable cells that express only testosterone receptors. Non-replaceable neurons project to the basal ganglia region, area X, and express both estrogen and androgen receptors. Using in-vivo Manganese Enhanced (ME)-MRI, we aimed at comparing the ME-MRI measures of the song control system with its behavioral output (song) in 22 female starlings during a natural seasonal regression in song behavior and plasma steroid levels.
We showed that Mn2+-labeling as the result of HVC activity and subsequent axonal transport from HVC to both nuclei (RA and X) was dramatically reduced in July (post-breeding season) when birds did not sing, compared to March (breeding season) when birds produced song. The volumes of RA and area X decreased 46% and 50% respectively, whereas changes in telencephalon volume did not exceed 4% and were not significant.
In July, manganese dependent delineation of area X was only possible in 10 % of the birds while RA could be delineated in 50% of the population, probably reflecting reduced activity of these pathways. The birds in which RA could be delineated in July had in March a higher activity of the HVC to area X projection as reflected by the total amount of Mn2+ accumulated in area X. This indicates that the plasticity of the song motor pathway is regulated by the activity of an associated basal ganglia circuit.
Abstract ID: 327
Poster Session: Biological Molecular Imaging, Part 3
Abstract ID: 328
The YAP-(S)PET scanner is a small animals tomograph able to perfom both PET and SPECT. It is made up of four detector heads: each one composed of a 4 × 4 cm2 of YAlO3:Ce matrix of 20 × 20 elements, 2 × 2 × 25 mm3 each, coupled to PS-PMT (Hamamatsu R2486). The four modules are positioned on a rotating gantry: opposite detectors are 15 cm apart and in time coincidence when used in PET mode. In SPECT mode four parallel-hole lead collimators are simply positioned in front of the scintillators. The scanner has a field of view (FOV) of 4cm × 4cm Ø. In PET mode the volume resolution is below 8 mm3 and is nearly constant over the whole FOV. The maximum absolute sensitivity in PET mode, measured at the center of the FOV, is 1.9% (19 cps/kBq) for 50–850 keV energy window. This system has been widely evaluated for the in vivo imaging of mice and rats injected with 18F radiopharmaceuticals. However, no information on the use of YAP-(S)PET for quantitative studies with 11C ligands are yet available. We have investigated if YAP-(S)PET performances are adequate to obtain quantitative information in pre-clinical small animal studies using 11C ligands. In particular, YAP-(S)PET sensibility and spatial resolution have been evaluated using different radiopharmaceuticals: [11C] Choline, [11C]Raclopride and [11C]Carfentanil. In addition, [11C]Raclopride binding data have been compared with those obtained ex-vivo either using tissue sampling techniques or phospho-imager authoradiography. These studies showed that YAP-(S)PET sensitivity was sufficient to image [11C]Raclopride binding in rats and mice striata by injecting activities lower than 180 and 90 μCi, respectively. Furthermore, spatial resolution was enough to resolve small cerebral structures as striata and thalami. In conclusion, this work indicates that the YAP-(S)PET system could be adequately applied to the pre-clinical protocols using 11C-ligands.
Abstract ID: 329
Abstract ID: 330
Abstract ID: 331
Abstract ID: 332
Abstract ID: 333
MRI has been used in combination with superparamagnetic nanoparticulates to enhance contrast to provide visualization of atherosclerotic plaques. This study provides proof-of-feasibility of targeting iron oxide nanoparticulates to atheromatous plaques for specific identification of vulnerable plaque. While small molecule radionuclides and paramagnetic liposomes have been targeted to plaques for in vivo gamma imaging and MRI, respectively, we demonstrate the efficacy of tethering larger structure particulates that are imageable using magnetic resonance.
We assessed the ability of Annexin V-conjugated nanoparticles to direct contrast to atherosclerotic plaques by injecting Annexin V and control (protein G conjugated) nanoparticles into Watanabe rabbits and evaluating signal changes in a 4.7 T MRI unit. Nanoparticles were determined to be 98 nm diameter, 20 proteins per particle, and have an iron concentration of 0.05 mg/ml (Annexin V-conjugated), while protein G nanoparticles were 90 nm, had >2 proteins per particle, and an iron concentration of 0.5 mg/ml.
We detected signal cancellation by Annexin V-conjugated nano-particles in atherosclerotic plaques of the abdominal aorta with simultaneous morphologic delineation. Signal cancellation was observed visually and confirmed by ROI analysis even though 2000 times less iron was used than reported by untargeted studies. Signal cancellation was detected in a repeat round of imaging in the same animal, indicating that particles did not induce a response that precludes re-imaging. Conversely, MRI revealed no control particle accumulation in plaques, nor was Annexin V or control particle accumulation found in control (plaque-free) animal vasculature. We will also present a survey of antibody-conjugated reagents targeted to vulnerable plaques.
Our approach thus demonstrates the feasibility of delivering a particulate platform of size sufficient to carry multiple functional payloads, (e.g., imaging and/or therapeutic agents) to plaque sites. Additionally, this approach is repeatable and enables higher order morphological characterization and resolution than previous targeted radionuclide imaging studies.
Abstract ID: 334
Bone remodeling is a key event during normal bone growth and repair. Changes in bone remodeling can be used as a surrogate biomarker for numerous disease states. This communication outlines the use of OsteoSense, a near-infrared fluorescent (NIRF) bisphosphonate agent, to non-invasively assess bone remodeling in vivo and in realtime. OsteoSense was used to image bone remodeling as a result of therapeutic intervention and disease progression in two different experimental animal models. In the first model, calvarial bone growth induced by repeat administration of parathyroid hormone (PTH) was assessed using OsteoSense and NIRF imaging. Significantly higher bone growth was detected using OsteoSense and NIRF imaging in PTH- treated mice as compared to non-treated control mice (p <0.0001). OsteoSense NIRF signal correlated well (R2=0.76) with calvarial thickness as measured by traditional, more laborious histological methods. In a second animal model, bone erosion resulting from arthritis in a mouse collagen induced arthritis (CIA) model was assessed using OsteoSense and NIRF imaging. Significantly higher bone remodeling was detected by OsteoSense and NIRF imaging in CIA mice with established arthritis as compared to healthy control mice (p < 0.0001). The OsteoSense NIRF imaging results correlated with more traditional measures of disease such as clinical score (R2=0.74), paw thickness (R2=0.75) and histology. Taken together, these results demonstrate the utility of OsteoSense and NIRF imaging to assess bone growth and erosion in various animal models of disease.
Abstract ID: 335
Hyperspectral Optical Tomography can provide accurate 3D localization of bioluminescent probes in small animals. Images of the probes are constructed by solving an inverse problem which requires the solution of the optical forward problem that describes the photon flux on the surface for a light source inside by the diffusion equation. This equation in realistic geometries is solved by FEM and requires knowledge of the tissue optical properties throughout the volume. Solutions of the forward model and consequently of the inverse problem are dependent on the use of correct tissue properties. While the tissue properties of skin are well documented in literature, there is a wide range of values reported for bone or skull. Uncertainty of these optical properties can have an impact when imaging probes inside the brain.
We analyze the errors caused by variation in the skull optical properties and the skull thickness using realistic mouse models, block shaped and spherical phantoms. Figure 1 shows linear tomographic reconstructions for a cubic model of dimensions 20 × 20 × 20 mm using the correct optical properties (left) and overestimation of the skull properties by 50% (right). The simulated point source at location (10,10,10) mm mislocalized at (10,10,5) mm, because the larger skull absorption leads to weaker surface flux for sources at the same height as compared to the forward simulation using original optical properties. These results indicate the importance of accurate knowledge of optical properties to minimize localization errors that may otherwise arise in in-vivo tomographic bioluminescence studies.
Left: Absoption and scattering coefficients, original values in green, values used for reconstruction in blue. Center: reconstruction using original properties. Right: Localization with overestimated properties.
Abstract ID: 336
Abstract ID: 337
SCCVII-tumors revealed the lowest heat tolerance. 8 hours after the exposure to 42°C a 1208% higher, at 44°C a 16158% higher and at 46°C a 5.4% higher HSP70 expression increase was seen. The diffusion coefficient increased 6.9% at 42°C, 44% at 44°C and 34% at 46°C.
Abstract ID: 338
We have found that DCE-MRI and T1ρ-weighted MRI can differentiate two typical human melanoma xenografted in SCID mice, the aggressive C8161 and the indolent A375P. Here we report on 3D high spatial resolution cryo-fluorometry to measure the redox states of these two xenografts and correlations between the redox states of these tumors with their metastasis capacity, their metabolic rates, and with tissue oxygenation.
The tumor samples were freeze-trapped to preserve tissues' metabolic states and then dissected for multi-slice fluorescence imaging of mitochondrial NADH and flavoproteins (Fp). The redox ratio Fp/(Fp+NADH) was calculated as an index of mitochondrial respiration.
Distinct tumor heterogeneity in the spatial distribution of redox ratio has been observed. In the first slice (600 μm beneath the skin), on average, the aggressive xenograft was at least 2 times more oxidized than the indolent xenograft, indicating higher levels of mitochondria respiration. An interesting bi-modal spatial distribution of redox ratios was observed on the deeper slices of the aggressive tumor which is significantly more oxidized in the central region (mean 0.70 ± 0.14) than in the outer ring (mean 0.15 ± 0.10). This probably reflects much higher mitochondria respiration, greater hypoxia, higher metastatic potential and lower radiosensitivity in the tumor core than in the tumor rim.
However, an imaging scan of the relative oxygenation distribution in the same slice did not show significant spatial heterogeneity and did not indicate substantial differences between the two lines. More sensitive oxygenation measurement techniques are being initiated. In short, these preliminary data showed a promising way of using cryofluorometer measurements to characterize human melanoma and perhaps also other cancer cell lines to evaluate the underlying metabolic characteristics responsible for differences in metastatic potential.
Abstract ID: 339
It is generally accepted that dietary habits influence disease development. The mechanisms for many of the beneficial effects have been studied in cell cultures, and it has been demonstrated that dietary constituents can regulate expression of many genes associated with disease development. However, most of these in vitro findings have not been properly validated in more complex systems.
Transgenic reporter mice represent useful and important tools that can be used to validate many of the in vitro findings. Luciferase is frequently used as a reporter gene, enabling non-invasive optical imaging of a given gene regulatory element in a time- and spatial manner. In dietary research this has obvious advantages as bioavailability and tissue dependent bioactivity is largely unknown.
We have made reporter mice where luciferase expression is regulated through NF-kB response elements, retinoic acid response elements (RARE) or antioxidant response elements/Electrophile response elements (ARE/EpRE).
NF-kB is a master regulator of inflammation. RARE is bound by retinoic acid receptors, which are particularly important in embryonic development, but are also vital for cell differentiation and maturation. ARE/EpRE is bound by Nrf proteins, which regulate phase II detoxification and antioxidant defense genes. It is previously demonstrated that all these gene regulatory systems are influenced by nutrients in vitro.
In these studies we have a used a cooled CCD camera (IVIS-100, Xenogen Corp.) to measure luciferase activity in these transgenic mice and show firstly that NF-kB is down modulated by vitamin A; secondly that vitamin A is activating RARE dependent luciferase activity, in a tissue specific manner. Finally we demonstrate that ARE/EpRE dependent luciferase activity is increased by phytochemicals.
To our knowledge this is the first in vivo imaging demonstration of nutrient regulated gene expression.
Abstract ID: 340
Abstract ID: 341
Non-invasive in-vivo imaging of small animals is a rapidly growing field, with new technologies and techniques being constantly developed. Because of the high cost of MR- and CT-based systems, a great deal of effort has gone into developing optical imaging methods. Both bioluminescence and fluorescence imaging systems have been available on the market for some time. To date, in-vivo fluorescence imaging has not progressed as rapidly as bioluminescence imaging despite having several potential advantages such as the ability to multiplex fluorophores, the lack of a need for a timed injection, and the potential to use injected labeled antibodies. The ability to image and quantitate fluorescently labeled tumors and other fluorescently labeled markers in-vivo has generally been limited by the autofluorescence of the tissue. The presence of any auto-fluorescence reduces the sensitivity of detection and accuracy of quantitation of the labeled tumor. Autofluorescence therefore limits the usefulness of conventional in-vivo fluorescence imaging; expensive cameras do not help, since they simply capture the autofluorescence more efficiently. One solution to this problem is to use a multispectral imaging methodology to spectrally characterize and computationally eliminate autofluorescence, revealing otherwise invisible labeled targets, which now appear bright against a near-black background. This dramatic improvement in signal-to-noise can increase sensitivity by orders of magnitude, allowing much smaller targets to be detected. The application of multispectral imaging methods to several small animal and fluorescence microscopy will be shown. Recently developed automated methods for determining the spectral signatures of the fluorophores greatly simplify the application of this technology and make possible the analysis of multiple fluorophores in the simultaneously.
Abstract ID: 342
Abstract ID: 343
Abstract ID: 344
Abstract ID: 345
Presenilin 1 (PS1) is the critical component of the gamma-secretase complex that cleaves amyloid beta from the amyloid precursor protein (APP). More than 100 mutations in the PS1 molecule have been described that are linked to Familial Alzheimer's Disease (FAD).
Fluorescence Lifetime Imaging Microscopy (FLIM) is a FRET-based method that allows for the assessment of inter- or intramolecular interactions in intact cells. If two fluorophore-labeled epitopes are less than 10nm apart, the donor lifetime shortens proportional to the distance between the epitopes.
In this study, two separate FLIM assays were developed. In the first assay, the relative molecular distance between PS1 N and C termini was measured. We show that FAD-linked missense mutations lead to a similar alteration in the spatial relationship between PS1 N and C termini, resulting in an increased proximity of the two epitopes.
An opposite effect on the proximity of PS1 N and C termini has been observed by treatment with Abeta42 lowering nonsteroidal anti-inflammatory drugs (NSAIDs) (Lleo et al., 2004). Accordingly, treatment of PS1 mutant neurons with high-dose NSAIDs somewhat offsets the conformational change associated with the mutation.
In the second FLIM assay the relative distance between a PS1 loop epitope and the APP C terminus was monitored. We demonstrate that the FAD PS1 mutations are associated with a consistent change in the configuration of the PS1-APP complex that is characterized by a decrease in proximity between PS1 and APP.
We propose that the conformational change we observed using FLIM may therefore provide a shared molecular mechanism for FAD pathogenesis caused by a wide range of PS1 mutations.
Abstract ID: 346
Representative images of a control mouse versus an LPS-instilled mouse.
Abstract ID: 347
Poster Session: Novel Probes and Activation Strategies, Part 3
Abstract ID: 348
Fluorescence microscopy is an important tool to study the internalization and localization of biomolecules within a living cell. Common “biomolecular imaging” experiments are, however, affected adversely by background signal from extracellular material. To suppress background, we envisioned labeling a biomolecule with a fluorogenic substrate that remains nonfluorescent until cellular entry and reaction with intracellular enzymes. Many known fluorogenic substrates are unstable in aqueous solution and are unsuitable for use in biomolecular imaging experiments. We developed a novel class of latent fluorophores based on the “trimethyl lock” prodrug strategy. These compounds possess low background fluorescence and superior stability while maintaining high reactivity with the enzyme of interest. The modular design of the probe allows specificity for a particular enzyme to be incorporated into the molecule by simple functional group manipulations. We initially applied this concept to the amine-containing fluorescent dye rhodamine 110. Subsequent work with a urea-substituted rhodamine 110 showed a dramatic improvement of the kinetic parameters with esterase enzymes. The urea moiety was also exploited in the preparation of a thiol-reactive latent fluorophore. Attachment of this fluorogenic substrate to ribonuclease A yields a stable conjugate that can be used to image endocytosis of the protein.
Abstract ID: 349
Macromolecular Gd(III) chelates are superior MRI contrast agents as shown in animal models. However, clinical development of these agents is limited due to safety concerns related to their slow excretion. We have recently developed fast eliminating biodegradable macromolecular MRI contrast agents based on polydisulfide Gd(III) complexes, which are broken down by the endogenous thiols and rapidly excreted from the body after the MRI exam. Here we modified the structure of the polydisulfides by introducing different steric hindrance around the disulfide bonds to study how it altered the degradability and in vivo contrast enhancement. Gd-DTPA cystine bisamide copolymer (A), Gd-DTPA cystine bispropylamide copolymer (B) and Gd-DTPA cystine bisisopropyl amide copolymer (C) were prepared. The T1 relaxivity is 4.37, 5.28 and 5.56 mM−1sec−1 at 3T for agents A, B and C, respectively. The in vitro degradation rate of the copolymers decreased with increasing steric effect around the disulfide bonds in the order of A, B and C. The in vivo contrast enhancement of the agents with molecular weight about 20 KDa was evaluated in rats on a Siemens Trio 3T scanner. Strong contrast enhancement was observed in the blood pool at the initial period postinjection with all agents and the signal intensity gradual decreased, Figure 1. The agents also resulted in strong liver contrast enhancement. In conclusion, although the agents had different in vitro degradability, they resulted in similar contrast enhancement in the blood pool.
The coronal MR images showing rat heart contrast enhanced with the agents A, B and C before and at various time points after the injection of the agents at a dose of 0.1 mmol-Gd/kg.
Abstract ID: 350
The methanethiosulfonyl group (MTS) is known to readily react with thiol functions. The synthesis of gadolinium (III) chelates of several ditopic ligands featuring one or several MTS units will be reported. The MTS groups have been added to DTPA and to either the carboxyl groups or the tetraaza ring of DOTA (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid) ligands rigidified by one cyclohexyl or four methyl groups. The complexes are kinetically inert and are monohydrated as shown by luminescence spectroscopy. These chelates easily react with the thiol function of bovine serum albumin and the relaxivity increases from 6 to 25 s−1mmol−1(20 MHz, 25°C). These relaxivity effects are compared with those observed with analog Gd(III) chelates featuring hydrophobic benzyloxymethyl groups directly grafted on the tetraaza ring of DOTA that form non-covalent assemblies with BSA. The nuclear magnetic relaxation dispersion curves (NMRD) are interpreted quantitatively and show a large increase in rotational correlation times as expected and substitutions on the acetate arms or on the tetraaza ring modify differently the water exchange time and bring about different relaxivity increases. The MTS chelates have also been reacted with polythiolated silica particles (Sicastar) featuring 120.000 thiol groups per particles. Very large relaxivity increases are again observed and the relaxivity parameters are deduced from NMRD curves. The stability of the disulfur bonds is tested in presence of 1,4-dithiothreitol, glutathione and cysteine. The breaking of the disulfur bonds is easily observed by relaxometry and is clearly indicative of the presence of thiolated derivatives.
Abstract ID: 351
Dipolar field effects in solution, conveniently expressed as intermolecular multiple-quantum coherences (iMQCs), have been shown to be useful for enhancing imaging contrast in vivo. These challenging applications have traditionally been limited by relatively poor sensitivity (typically 10–20% of the full magnetization) and by issues associated with coregistering iMQC images with their conventional counterparts. We have demonstrated “multiCRAZED” sequences which simultaneously acquire as many as two conventional and three iMQC images in a single shot, each image having essentially its full intensity. The combination of the five images dramatically enhances signals, reduces coregistration issues, and permits clean separation of iMQC contrast differences from other effects such as diffusion, relaxation, and flow.
One particularly intriguing application is characterization of excised breast cancer tumors with advanced contrast agents (LHRH (luteinising hormone-releasing hormone)-conjugated nanoparticles. When human breast tumors are grown in nude mice, conjugated nanoparticles injected into the tail selectively deposit in the tumor; for mice without tumors, they deposit in the liver. These tumors are highly heterogeneous (thus limiting detection methods based on T2 or T2*-based contrast) but they exhibit large enough multiCRAZED contrast differences to enable in vivo applications. Specifically, we have shown that anisotropy in pumping iMQC coherences (that is to say, signal differences as the correlation gradient direction is changed) is dramatically enhanced by the local fields produced by nanoparticles, which permits enhanced detection of these particles. Figure 1 shows a specific example with excised tumors. This work is supported by the NIH under grant EB 2122.
Abstract ID: 352
Abstract ID: 353
Atherosclerosis is an inflammatory disease of the arterial walls and represents a significant health problem in developed nations. The current “gold standard” for imaging atherosclerosis remains angiography, which images the luminal space and not the arterial wall. With angiography no information can be gained about plaque composition, only constriction of the vessel wall. Magnetic Resonance Imaging (MRI) has emerged as a promising modality for imaging plaque formation, as it offers excellent soft tissue contrast and sub-millimeter in-plane resolution. Contrast enhanced MRI has been implemented using dextran coated iron oxide particles, which are nonspecifically taken up by macrophages in atherosclerotic plaques. We have developed a MRI contrast agent for in vivo imaging of early stage atherosclerosis that is targeted for specific uptake by macrophages. We hypothesize that by including moieties on the surface of the MR contrast agent that are recognized by macrophage specific receptors the specificity and sensitivity of plaque labeling can be increased.
Macrophages express a class of receptors, collectively known as scavenger receptors (SR), which are expressed primarily on macrophages, but not on normal arterial wall. We have previously shown that fluorescently labeled ligands to the macrophage SR can specifically target restenotic plaques in a rat model. We have developed iron oxide particles coated with a scavenger receptor ligand as a targeted MR contrast agent for atherosclerosis. Mouse P388D1 macrophages, which are known to express the SR, were used to determine MR contrast agent uptake. Targeted MR contrast agents offer a unique opportunity to visualize early plaque development in vivo with high sensitivity and resolution, allowing for early diagnosis of atherosclerosis.
Abstract ID: 354
The targeted delivery of therapeutics and imaging agents to specific tissues requires appropriate targets and specific molecular localization on the cell surface. Tumors are known to exhibit abnormal gene expression profiles and these genes have been utilized as diagnostic and therapeutic targets. Proteases expressed by tumor cells or proximal host cells in response to tumor infiltration offer a suitable target for specific delivery of molecular agents to a variety of pathologic tissues, and their enzymatic capabilities could conceivably be exploited to improve the level of specific targeting achieved. Molecules based on the arginine octamer protein-translocation domain (PTD) were adapted for selective uptake by cells that expressed prostate-specific antigen (PSA), a biomarker for prostate cancer. The peptides have three regions: PTD, PSA protease-recognition region and attenuating region. Experiments demonstrated that these peptides were targeted by PSA and upon cleavage at the protease-recognition site, the fluorescent cargo linked to the PTD was taken up by the cells. Two different cell lines Jurkat (suspension cells) and PC3M (adherent prostate cancer cells) were used for the experiments. These peptide-based molecules could conceivably be used for targeted delivery of therapeutics or imaging agents to PSA-expressing prostate cancers.
Abstract ID: 355
Imaging the subcellular localization, co-localization and organization of cytoplasmic and nuclear RNAs in living cells can provide essential information on RNA synthesis, processing, transport and functions. Here we report the direct visualization of subcellular localization of K-ras and GAPDH (Glyceraldehyde 3-phosphate dehydrogenase) mRNAs and U1, U2 and U3 nuclear RNAs in live HDF cells using molecular beacons together with membrane-permeabilization and peptide-based delivery. Unexpectedly, we found that both K-ras and GAPDH mRNAs co-localize with mitochondria. Extensive control studies were performed, including the use of FISH (fluorescence in-situ hybridization), negative-control beacons, and the detection of co-localization of 28S ribosomal RNA with the rough ER, suggesting that the mRNA localization and co-localization patterns observed in our study are true and specific. Further, using a probe composed of a single nuclear-localization-sequence peptide covalently linked to a molecular beacon, we demonstrated the specific and sensitive detection of the localization of small nucleolar RNA U3 in the nucleolus, and the localization and co-localization of small nuclear RNAs (snRNA) U1 and U2 at discrete foci in the nucleoplasm. In summary, our observation revealed intriguing sub-cellular associations of mRNA with organelles such as mitochondria, which may provide new insight into the transport, dynamics and functions of mRNA, and mRNA-protein interactions. Our novel approach for nuclear RNA detection allows for the imaging of transcriptional and post-transcriptional processing of RNAs in the cell nucleus. Taken together, the ability of molecular beacons to detect and image endogenous RNA in living cells in real time can offer tremendous opportunities for biological and disease studies, and significantly impact drug discovery and medical diagnostics.
Abstract ID: 356
Understanding viral pathogenesis is critical for prevention of outbreaks, development of antiviral drugs, and biodefense. Here we describe a novel method to directly detect the viral genome of a clinical isolate of bovine respiratory syncytial virus (bRSV) in living cells using molecular beacons, which are dual-labeled, hairpin oligonucleotide probes with a reporter fluorophore at one end and a quencher at the other. They are designed to fluoresce only when hybridized to a complementary target. By imaging the fluorescence signal of molecular beacons, the spreading of bRSV was monitored for 7 days with a signal-to-noise ratio of > 50, and the measured time-course of infection was quantified with a mathematical model for viral growth. We found that molecular beacon signal could be detected in single living cells infected with a viral titer of 2 × 103.6 TCID50/ml diluted a thousand-fold, demonstrating high detection sensitivity. Low background in uninfected cells, and simultaneous staining of fixed cells with molecular beacons and antibodies showed high detection specificity. Furthermore, using confocal microscopy to image the viral genome in live, infected cells, we observed a connected, highly three-dimensional, amorphous inclusion-body structure not seen in fixed cells. Taken together, this new live-cell-based method provides a powerful tool for rapid viral infection detection, basic virology studies of RNA viruses, and the design of new antiviral drugs.
Abstract ID: 357
Optical molecular imaging (MI) may provide a superior method for detection, diagnosis and characterization of breast cancer as it is noninvasive, requires breast stabilization but not necessarily compression, and lacks exposure to ionizing radiation. Furthermore, MI provides physiologically relevant information about the suspected lesion and can potentially serve in therapeutic efficacy monitoring.
The peripheral benzodiazepine receptor (PBR) represents an exceptional target for MI agents due to its overexpression in breast cancer. Capitalizing on this expression profile and the low absorption and increased photon penetration depth in the near-infrared (NIR) region, we have recently synthesized a novel PBR-targeted NIR fluorescent MI agent (NIR-PK 11195) by coupling the conjugable PK 11195 developed in our laboratory [1] to an NIR dye. To evaluate the ability of NIR-PK 11195 to target tumor cells in vivo, we implanted MDA-MB-231 human breast cancer cells into BALB/c mice and injected free NIR dye or NIR-PK 11195 into tumor-bearing and non-tumor-bearing mice three weeks post-implantation. The biodistribution of the probes was monitored in real-time for ~48 hours post-injection using the eXplore Optix (GE/ART). The mice were then sacrificed and harvested tissues were imaged.
It will be shown that direct comparison of the biodistribution and accumulation of the PBR-targeted and non-targeted probes in tumor-bearing mice demonstrates significantly different clearance profiles, resulting in preferential labeling of MDA-MB-231 tumors in vivo by NIR-PK 11195. At twelve hours post-injection of the NIR-PK 11195, the tumor regions exhibit an approximately 10-fold contrast enhancement over the free dye; this contrast persists for the duration of the study (48 hours post-injection). Non-tumor-bearing control mice also confirm that NIR-PK 11195 is preferentially labeling the MDA-MB-231 tumors. In addition, region of interest measurements and MATLAB algorithms evaluation will show that NIR-PK 11195 uptake kinetics can be quantitatively determined.
Abstract ID: 358
Abstract ID: 359
Abstract ID: 360
Luciferase complementation imaging (LCI) enables the real time imaging of protein-protein interactions in cellulo and in vivo. Previously, a library of incrementally truncated amino and carboxy-termini fragments of luciferase that were fused to FRB and FKBP (N-Luc-FRB and C-luc-FKBP respectively) was screened for rapamycin dependent induction of luciferase activity. This screen identified a number of N-Luc/C-Luc fragment pairs in which the active site of the enzyme was divided between the two fragments, but in the presence of rapamycin, were able to reconstitute 30% of native luciferase activity. Notably, all pairs identified shared at least a minimal 19 amino acid common, overlapping domain.
In order to test whether duplication of this overlapping domain is required for reconstitution of activity, N-Luc-FRB and C-Luc-FKBP clones with successive 4 amino acid truncations to the overlapping domain were created and coexpressed. A number of pairs showed rapamycin inducible activity comparable to that seen with the originally identified N-Luc-FRB and C-Luc-FKBP pair. A strict requirement for any degree of overlap was not observed. However, truncation beyond more than 12 amino acids from either N-Luc-FRB or C-Luc-FKBP resulted in significant loss of activity. One of the pairs examined N1-Luc-FRB and C3-Luc-FKBP, retained over 70% of the activity of the original pair in the presence of rapamycin, but had a greater than 67% reduction in background activity in the absence of rapamycin, resulting in a signal to background ratio of 9.4, a more than two-fold improvement over the original levels.
Abstract ID: 362
Abstract ID: 363
Abstract ID: 364
Histone deacetylase (HDAC) inhibitors (HDACIs) are emerging as a new and exciting class of anti-neoplastic agents. Our goal is to develop a noninvasive magnetic resonance spectroscopy (MRS)-based molecular imaging method to detect inhibition of HDAC.
We investigated the fluorinated HDAC substrate, Boc-Lys(Tfa)-OH (BLT). First we used 19F MRS to show that BLT is cleaved in vitro by recombinant HDAC-8, confirming that it is a substrate of HDAC and that its metabolism can be monitored by MRS. To test the value of BLT as a molecular marker of HDAC inhibition, we next investigated PC-3 human prostate cancer cells. Cells were treated, in the presence 1mM BLT, with 10μM of the clinically relevant HDACI SAHA, or with DMSO. Treatment with SAHA for 24 hours lead to a significant drop in HDAC activity (41% relative to controls) and in cell proliferation (62% relative to controls). BLT did not significantly affect HDAC activity or cell proliferation. Using 19F MRS we found that treatment with SAHA lead to a significant increase in BLT levels from 8 fmol/cell in controls to 18 fmol/cell in HDACI-treated cells. This indicated that BLT levels are associated of HDAC inhibition. In addition, using 31P MRS, we determined that phosphocholine levels increased significantly by 200% in HDACI-treated cells indicating that phosphocholine could serve as a surrogate metabolic marker of response.
This study indicates that 19F MRS-imaging of fluorinated HDAC substrates, combined with 31P MRS-imaging, could be used to directly and noninvasively monitor HDAC inhibition in cells, and possibly in vivo.
Abstract ID: 365
Formed through cooperative self-assembly of amphiphilic diblock copolymers and electronically conjugated porphyrinic near infrared (NIR) fluorophores, NIR-emissive polymersomes (50 nm–50 um polymer vesicles) define a family of organic-based, soft matter quantum dot analogues that are ideally suited for in vivo optical imaging. Membrane incorporation of a wide range of related multi-porphyrinic fluorophores has enabled precise emission energy modulation over a broad domain of the visible and near infrared spectrum (600–900 nm). Long-wavelength optical excitation of such assemblies generates intense, highly localized emissive signals capable of penetrating through the dense tumor tissue of live animals. We present the latest developments towards the generation of NIR-emissive polymersomes in which optical, chemical, and physical properties are optimized for in vivo fluorescence-based imaging through deep-tissue, as well as prototype self-assembled, fully-bioresorbable emissive polymersomes comprised of an amphiphilic polymer consisting of two previously FDA-approved building blocks.
Abstract ID: 366
Among different matrix metalloproteinases, gelatinase A (MMP-2) and gelatinase B (MMP-9) have been the most consistently detected in malignant tumors and their level of expression has been associated with tumor angiogenesis and metastasis. In this work we describe an MMP-2/9-targeted NIR imaging agent developed on the basis of structure-activity relationship of cyclic peptides having inhibitory activity against MMP-2. We started from cyclo(CTTHWGFTLC) (CTT-peptide), which had previously been identified as a MMP-2/9-binding peptide through in vivo phage selection technique. Replacing the S-S bond formed between two Cys in CTT-peptide with an amide bond formed between the N-terminus of Ala and the side chain of Asp resulted in a significant increase in the stability of the resulting cyclic peptide c(ATTHWGFTLD)NH2 as compared to CTT-peptide Substitution of Asp-NH2 with β-Ala without changing the ring size of the peptide resulted in an increase in IC50 value from 11 μM of c(ATTHWGFTLD)HN2 to 84 μM of c(ATTHWGFTL-βAla), suggesting that the CONH2 functional group of Asp in the peptide is required for its enzyme inhibition activity. Ala survey for the first three amino acids in c(ATTHWGFTLD)HN2 indicated that these amino acids could be replaced without impairing the peptide's activity. Final elaboration to introduce a functional group and to reduce the number of amino acids lead to c(KAHWGFTLD)HN2, which had an IC50 value of 7.6 μM. Conjugation of Cy5.5 to the α-NH2 of Lys in c(KAHWGFTLD)HN2 led to Cy5.5-c(KAHWGFTLD)NH2, designated WW88, which bound to a variety of tumors cells expressing MMP-2/9 in vitro and to tumors inoculated either ectopically or orthotopically in vivo.
Abstract ID: 367
Gadolinium-complexes are widely used as T1-contrast agents for aequeous solutions. Linked to target specific structures, they are very useful for molecular MR-imaging. However, variable tissue-T1 lead to variations of the background signal hence causing ambiguities of the detected contrast. This can be overcome by 19F-based contrast agents: since flourine has a negligible natural occurrence, 19F images are free of background signal. By selectively increasing the 19F T1-relaxation we aim to increase the signal-to-noise ratio per unit time. For this purpose, polyfluorinated Gd-complexes were synthezised with a Gd center ion within a polyfluorinated DTPA backbones.
DTPA-bis-anhydride, fluorinated benzylamines and benzylalcohols reacted in DMF at 60° C. After purification, GdC13 was added to an aqueous suspension of ligand and neutralized. Chemical structures were characterized by elektrospray ionisation-MS, matrix associated laser desorption ionisation (Maldi)-Tof-MS and multinuclear high resolution MRS. Inversion-recovery T1-measurements were performed in water at 360 MHz at 20° on samples containing 1mmol/L of the fluorinated Gd-complexes. Solutions of fluorinated chelates contained 10% of ethanol to improve the solubility.
All fluorinated Gd-DTPA-complexes shorten the T1-relaxation of water as compared to Gd-DTPA-BMA (see table). 19F-NMR spectra of compounds A, A1 and B, B1 exhibit a sharp signal at δ= −63.6 ppm, while C consists of a broad multiplett due to coupling to aromatic ring protons. 19F T1-relaxation of CF3-Groups in fluorinated DTPA-Gd complexes A and B is increased significantly (see table). Thus, fluorinated Gd-DTPA-contrast agents promise better SNR per unit time for 19F-imaging.
Poster Session: Molecular and Functional Imaging in Cancer, Part 3
Abstract ID: 368
Abstract ID: 369
Gene therapy for thymidine kinase (HSV-tk) gene transfected BT4C/BDLX rat glioma triggers programmed cell death (PCD) that leads to a number of NMR-detectable phenomena in vivo as a result of ganciclovir (GCV) treatment: accumulation of polyunsaturated (and other) lipids; gradual but controlled drop of concentrations of low molecular weight metabolites; changes in all common MR image contrasts (SD,T2,T1,T1ρ) and changes in water diffusion. It is well documented that in the early stages of GCV-treatment the main mechanism of cell death is apoptosis, but no clear indicator has been proposed for the detection of possible change in apoptosis- /necrosisratio in the later stages of treatment. In this work, we investigated the feasibility of creating regional choline (at 3.23 ppm) diffusion maps in vivo by utilizing diffusion weighted 1H MR spectroscopic imaging (MRSI). Necrosis leads to a sudden breakdown of plasma membrane and leakage of intracellular matter into the extracellular space, in which the viscosity is low. Since apoptosis is characterized by cell shrinkage and hence, higher viscosity of the intracellular space, regional choline diffusion could effectively discriminate between areas which are apoptotic or necrotic.
BDIX rats bearing BT4C tumors (n=4) with untreated controls (n=3) were studied. The animals were treated with GCV over 8 days to induce apoptosis and total tumor reduction. We performed diffusion MRSI (b-values of 0, 500, 1300, and 2000 s/mm2) at 4.7 T. Data was analyzed in the time-domain using jMRUI-software.
The ADC of choline in tumors was found to be 0.18–0.20 × 10−3 mm2/s, Figure 1. We show that imaging of the molecular diffusion of choline is feasible in tumors in vivo, and can potentially be used to detect favorable treatment response.
Typical dataset from BT4C/BDIX rat treated for 4 days. A) Anatomical image, B) Choline signal (at 3.23 ppm) intensity map and representative spectrum and C) choline ADC map from the same voxels.
Abstract ID: 370
Micro-computed tomography is an important method of imaging rodents non-invasively to achieve high-resolution, volumetric datasets with short acquisition times. Due to the poor soft-tissue contrast inherent to micro-CT imaging, injectable contrast agents have been introduced to increase the contrast between vasculature and background tissue. Our hypothesis was that by knowing the time-course of contrast enhancement, enhanced images of the vasculature and liver could be obtained with a single injection of contrast material.
To study contrast enhancement as a function of time, iodinated blood-pool contrast agent (Fenestra VC) was injected, via a tail vein, into immune deficient NIHIII mice. Imaging was performed using a GE Locus Ultra micro-CT scanner to acquire volume datasets in 8 seconds. Each image was reconstructed with an isotropic voxel spacing of 0.15 mm; the entrance dose to the animal was estimated to be 0.07 Gy. Images were acquired prior to and at 0, 0.5, 1, 2, 4, and 6 hours post contrast injection. From this study, we determined that peak enhancement of the vasculature occurs immediately following contrast injection and peak liver enhancement about 6 hours later.
This dual-phase imaging protocol was subsequently applied to NIHIII mice injected with HT-29 (human colon carcinoma) cells into the mesenteric vein to target tumor growth to the liver. Respiratory-gated images of free-breathing tumour-bearing mice were acquired using a GE explore RS micro-CT scanner immediately post injection and 6 hours later. The respiratory-gated acquisitions took approximately 25 minutes, with an estimated entrance dose of 0.24 Gy. In this group of mice, the vasculature was well visualized in the images acquired immediately post injection. After 6 hours, the normal liver parenchyma was enhanced, enabling delineation of the tumours, which remained at the baseline value. Dual-phase contrast-enhanced micro-CT can be used to study liver tumours and associated vasculature in mice.
Abstract ID: 371
Abstract ID: 372
Abstract ID: 373
An insufficient blood supply to a rapidly growing tumor leads to the presence of hypoxia, which is a well-known feature of solid tumors. Hypoxia is known to decrease the efficiency of current anti-cancer therapies such as chemotherapy and radiotherapy. Moreover, hypoxic conditions increase the aggressiveness of tumor cells by inducing expression of a variety of genes related to proliferation, and malignant progression such as invasion and metastasis. Hypoxia-inducible factor-1 (HIF-1) is a master transcriptional activator of such genes and significant associations between HIF-1 overexpression and patient mortality have been shown in many cancers. Thus imaging and targeting tumor cells where HIF-1 is active are important in cancer therapy. In this study, HIF-1 activity in solid tumors was monitored via an optical in vivo imaging system by using a luciferase reporter gene under the regulation of an artificial HIF-1-dependent promoter, 5HRE, which contained five copies of hypoxia-responsive element (HRE). Stable tansfectants of the reporter construct were isolated from several cancer cell lines and expressed more than 100-fold luciferase in response to hypoxic stress. By using these transfectants, we monitored the change in HIF-1 activity in the process of tumor development. We found that HIF-1 activity was correlated with the growth speed of tumors. We also monitored HIF-1 activity in solid tumors during the treatments with a hypoxia-targeting protein drug and radiotherapy. These results revealed that the imaging system allows sensitive, real-time and spatiotemporal analyses of tumor hypoxia in solid tumors and provides a useful method to evaluate the efficacy of a cancer therapy on hypoxia in malignant solid tumors.
Abstract ID: 374
Abstract ID: 375
Abstract ID: 376
Abstract ID: 377
The biodistribution studies (rat tumor AR4–2J model) showed a high and specific uptake in tumor and receptor positive tissue. [111In-DOTA]-TATE showed a tumor uptake of 4.1%id/g and a tumor-to-kidney ratio of 1.7:1, whereas [111In-DOTA]4-Apg3-Sar5-TATE showed a tumor uptake of only 1.9%id/g but no significantly different uptake into the somatostatin receptor positive organs. Compared to Gd(DOTA)−(r1=3.83 mM−1s−1/20 MHz/37°C), the relaxivity of [Gd-DOTA]4-Apg3-Sar5-TATE (r1=8.9 mM−1s−1/20 MHz/37°C/per Gd) increased by a factor of 2.3 which is due to its increased rotational correlation time of 490 ps (vs. 177 ps for Gd(DOTA)−).
Abstract ID: 378
Prostate cancer (CaP) and mammary cancer (CaM) metastasize frequently to bone and influence its remodelling. Their bone metastases can be either osteosclerotic or osteolytic, but the mechanism determining these features is unclear. Bone morphogenetic proteins (BMPs) are osteoinductive molecules. Extracellular antagonists modulate their activity. We investigated whether differential expression of BMP antagonists in human CaP and CaM cell lines may explain their osteosclerotic or osteolytic properties. Only noggin and at least one bone resorbing cytokine, either parathyroid hormone-related protein (PTHrP) or macrophage colony stimulating factor-1 (CSF-1), are constitutively expressed in all the osteolytic cell lines, either CaP or CaM. In contrast, all CaP and CaM osteoinductive cell lines do not express noggin or any of the resorbing cytokines. Endothelin-1 (ET-1), postulated as a major inducer molecule of osteosclerotic metastases, is mainly expressed by osteosclerotic CaM cells, but not by osteosclerotic CaP cells. Forced expression of noggin in the osteoinductive CaP cell line C4-2B4 abolished the osteosclerotic reaction induced by its experimental metastases, and only marginally affected basal bone resorption or kinetics of tumour growth. Collectively, the above results suggest that noggin blunts the bone formation that is coupled to physiological bone resorption. Accordingly, lack of noggin tilts the balance toward osteosclerosis. In contrast, presence of noggin allows full expression of the osteolytic effects of either PTHrP or CSF-1.
Abstract ID: 379
Abstract ID: 380
MR nodal staging with lymphotropic magnetic nanoparticles (LNMRI) has the potential to provide highly accurate cancer staging non-invasively. Images are currently assessed “visually” or by locally drawing ROI, a laborious and impractical approach given the multitude of lymph nodes. We have therefore developed automated image analysis including 1) identification, 2) segmentation, 3) calculation of tissue parameters (T2*, variance) and 3D display of results (color-coded tissue parameters superimposed on angiographic MIP). Specifically we developed an algorithm that simultaneously evolves parameters of an ellipse as well as parameters of registrations onto multiple image domains using differential equations. The algorithm extends 2D segmentations to 3D in an automatic fashion by carrying a converged contour onto next slice and using appearance and geometry constraints. Thus, a 3D representation of a lymph node is obtained by reconstructing a surface from the final set of 2D ellipses. The features extracted in the segmentation are subsequently used to classify the lymph node as malignant versus benign. This process utilizes features like T2*, δT2*, SNR and the pixel variance. Each feature is then preprocessed with an Entropy based discretizer and binned in either 2 or 3 segments, depending on the distribution of each feature. We combine these features by using an advanced Bayesian network classifier, which generates a graph model describing dependencies between the features on the one hand and the pathological status of the lymph node on the other hand. It also reveals underlying patterns not visible to a human reader. Based on this probabilistic model we can support a reader in his decision by classifying each lymph node and assigning also a probability to the decision.
Abstract ID: 381
The identification of the tumor specific signal transduction modules involved in promoting metastatic progression is of particular importance. Recent studies have underlined the importance of PKA as a gating element in a number of different processes including cancer motility and invasion and there has been a paradigm change in the understanding of PKA-dependent regulation in which specificity is secured by tight localization of signaling components. Through the use of new PKA subunit probes linked to fluorescent compounds it is possbile through measuring changes in Fluorescence Resonance Energy Transfer (FRET) and Fluorescence Lifetime Imaging (FLIM) microscopy analysis to both visualize the location and measure the amount of microcompartimentalized cAMP production and PKA activity. Here we compare the localization of cAMP generation and PKA activity in basal and in serum deprived conditions in human breast normal, MCF-10, and cancer, MDA-MB-435, cells. The results are consistant with an even distribution and activation of cAMP and PKA between the cell body and pseudopodia in the normal cells while in the cancer cells there are slightly higher levels of cAMP and PKA in the pseudopodia. Importantly, the mobilization of cAMP and subsequent activiation of PKA by serum-deprivation or by forskolin/IBMX occurs preferentially along the cell membrane of the pseudopodia that protrude in the direction of movement. This polarized expression and activation of the PKA signal system is particularly strong at the ends of the pseudopodia. This data support our previous data that serum deprivation-dependent activation of PKA phosphorylates pseudopodial RhoA and inhibits a pseudopodial RhoA/p160ROCK/p38 signaling module which activates NHE1 and increases invasion only in the cancer cells. These data further support the idea that invasion specific signal transduction modules in cancer cells are organized in invasion specific cell structures in-order-to be able finely orchestrate invasive activity.
Abstract ID: 382
Abstract ID: 383
Optical tomography using near-infrared (NIR) fluorescent contrast agents has been demonstrated in phantom studies but has not yet been demonstrated in patients. Thus, then influence of heterogeneous structure of tissue has yet to be tested. Herein, we investigate the effort of the natural distribution of endogenous optical properties as well as the non-uniform, background distribution of exogenous fluorescent agents upon optical tomography problem.
For image reconstruction, we developed an algorithm based on the penalty/modified barrier function (PMBF) method which utilizes the gradient based constrained truncated Newton method (CONTN) with trust region. Using this algorithm, three dimensional images of targets (0.5–1 cm3 dia., as deep as 2.8 cm) contrasted by fluorescence cross section and fluorophore lifetimes were reconstructed from experimental phantom boundary data measured with point excitation illumination and point collection geometries. While similar images were also reconstructed using planar wave illumination, all cases possessed a uniform, unstructured background.
To simulate the non-specific distribution of the fluorescent agent as well as the heterogeneity of endogenous and exogenous optical properties, we utilize the “lumpy-object” model developed Rolland and Barrett. With the same algorithm and measurement geometries, we show that it is possible to reconstruct images of the fluorescent target when the maximum value of the spatially heterogeneous exogenous and endogenous optical properties reaches 200% of the background average values. Figure 1A–C shows the true absorption map and fluorescent target recovery from the “lumpy” background. Additional results show that the algorithm is less forgiving of the heterogeneity of fluorophore cross section. This research demonstrates potential of fluorescence enhanced tomography in clinically relevant volumes of heterogeneous tissue.
Abstract ID: 384
Abstract ID: 385
Abstract ID: 386
The bacterial protein subunit B Shiga toxin from Shigella dysenteriae binds to a cellular receptor, the glycosphingolipid (GSL) globotriaosyl ceramide (Gb3 or CD77), that is expressed by some human cancers and in a transgenic mouse model of intestinal adenocarcinoma. We have explored the potentiality of Shiga toxin to be used for in vivo positron emission tomography (PET) contrast agent in intestinal tumoral development evaluation. The Shiga toxin was coupled to [F-18]fluoropyridine-based maleimide reagent via selective conjugation with a thiol (sulfhydryl) function. [18F] fluoropyridine-Shiga was injected into the tail vain of mice, which were imaged between 1 and 2 hours later using a microPET® FOCUS 220 scanner. Biodistribution showed that one hour post intravenous administration the bulk of radioactivity was found in the urinary tract as demonstrated by the high levels of radioactivity in the bladder and kidneys. Prominent uptake was also observed in the spleen and, to a lesser extent, in the liver. In transgenic mice analysis of tumoral load with [18F]fluoro deoxy glucose (FDG) confirmed the presence of adenocarcinomas in digestive track. In most of cases, FDG tumoral detection was correlated with a high level of [18F]fluoropyridine-Shiga uptake. This bimodal tumoral detection was confirmed by autopsy and Gb3 assays in biopsy from the same animal. In conclusion, [18F]fluoropyridine-Shiga as PET contrast agents has the potential of a tumor PET contrast agent for molecular imaging of digestive tumors and potential to delineate tumors expressing specifically the Gb3 receptor.
Abstract ID: 387
Whereas biotin-dependent enzymes are ubiquitous in nature, biotinylation, carried out by biotin ligase, is a rare enzymatic modification. We propose that metabolic (enzymatic) biotinylation of a cell surface receptor can serve as a method for targeted molecular imaging. We have illustrated this principle using modified PDGF receptor in a prostate cancer cell line (PC-3MM2). A biotinylation sequence (GLNDIFEAQKIEWHE) was inserted into the extracellular domain of PDGFR-β (wtBio-PDGFR-β). A mutated biotinylation sequence, in which the lysine residue was replaced with glycine, was used as a control (mBio-PDGFR-β). Stably transfected cell lines expressing high levels of fusion protein were selected and propagated. In the presence of E. coli biotin ligase (BirA), wtBio-PDGFR-β was highly biotinylated on the surface of PC-3MM2 cells as demonstrated using anti-biotin fluorescein conjugated antibody (Figure left - Green: anti-biotin, Blue: nuclei). In contrast, the biotinylation of mBio-PDGFR-β (right) was much lower and similar to the biotinylation level of wtBio-PDGFR-β in the absence of BirA (center). Due to the rapid turnover rate of the receptor, biotinylation is also observed in the cytoplasm. In a tumor model, coexpression of wtBio-PDGFR-β together with BirA will facilitate single or multi-step targeting of an MR imaging probe such as avidin-biotin-dendrimer-GdDTPA. It will also allow evaluation of the concentration of target molecule required for imaging and its accessibility to the imaging probe. If successful, this approach can de used to identify small primary tumors and metastatic lesions, and to monitor anti-PDGFR-β (STI571) or other anti-tumor therapies.
Abstract ID: 388
The pH-sensitive contrast reagent (CR), GdDOTA-4AmP, has been used previously to image extracellular pH (pHe) in kidneys [1] and gliomas [2] using the pharmacokinetics (PK) of a pH-insensitive CR for concentration correction. However, to transfer this methodology to a clinical environment, an alternative approach using a single injection to provide improved temporal resolution is needed. In this work, we investigate the use of T2 relaxivity induced by co-injection of a Dysprosium chelate to calculate the concentration of GdDOTA-4AmP [3]. In vitro data show that both T1 and T2 relaxivities induced by Gd-4AmP follow the same pH dependence. Consequently, we have investigated the use of an alternate lanthanide, Dy, as a doping agent to provide pH-independent T2 contrast [4]. In pH and concentration titrations in vitro, the r2/r1 ratios determined for Dy were 10 times larger than those for Gd. A similar effect was observed in C6 glioma xenografts models in vivo where a DyDOTP infusions showed no effect on T1 while producing large decreases in T2 as determined by both spin echo MRI and Proton Echo Planer Spectroscopic Imaging (PEPSI). These data suggest that a single injection of a defined mixture of DyDOTP and Gd-DOTA-4AmP could be used to measure pHe in vivo with high temporal and spatial resolution.
Abstract ID: 389
Abstract ID: 390
In recent years, bio-nanotechnology development in quantum dots (QDs) has been used to study cell behavior in vitro and in vivo for disease early diagnosis and therapeutic drug target. The size tunable QDs have opened a new world in molecular imaging research field. It provided new detection method with high resolution in the tiny space, but also generated huge amount of image data. How to process these molecular image data promptly is a challenging problem.
In this paper we will present the research work conducted in Bio-MIBLab (Bio-medical Informatics and Bio-imaging Lab, Georgia Tech/Emory University) using high performance computer to quantify and interpret the nanoimaging data in
Abstract ID: 391
An understanding of the tumor microenvironment, including tumor hypoxia, is essential in the design of new chemotherapies for the treatment of cancer which can influence the efficacy of drug treatment, the potential for metastasis, and the induction of an aggressive tumor phenotype. Under hypoxic conditions the hypoxia inducible factor, HIF-1α, is translocated to the nucleus of the cell where it dimerizes with HIF-1β, which then acts as a transcription factor responsible for upregulation of angiogenic factors and glycolytic enzymes. Elevated HIF-1α levels may lead to a glycolytic phenotype because of increased transcription of glycolytic enzymes, shown by PET imaging which utilizes the Warburg Effect, or the increased glycolytic nature of tumor cells, even in the presence of oxygen, and is often associated with poor prognosis. It has been shown that metastatic tumor cell lines often have elevated HIF-1α levels. A series of twelve primary breast tumor cell lines and three established breast tumor cell lines have been characterized for their HIF-1α levels, glucose consumption rates, expression of glucose transporters, and lactate production rates. Immunocytochemistry results suggested that subcellular localization of HIF-1α may be more indicative of cellular phenotype. Subsequent quantification of confocal image analysis and immuno-spot blot data are correlative and show differential subcellular localization of HIF-1α among these cell lines. The imaging data are also consistent with glucose uptake, lactate production, and glucose transporter expression which show two distinct populations of cells. This suggests that localization of HIF-1α plays a role in determination of a glycolytic phenotype in tumor cells.
Abstract ID: 392
Evaluation and optimization of fractionated chemotherapy dosage schedules using orthotopic brain tumor models has traditionally relied upon surgical inspection of lesions or survival as the quantitative endpoint. Diffusion MRI has been shown to be a sensitive non-invasive surrogate marker for evaluating therapy induced cellular changes. Diffusion values of water within a tumor are proportional to the tumor cellular density. Water movement within the tumor is measured as an apparent diffusion coefficient (ADC), the more extracellular water present the higher the ADC. When a cytotoxic dose has a positive impact on the tumor mass, tumor cells lose their membrane integrity and thus an increase in diffusion is evident.
In this study we compared the effects of systemic fractionated (2 mg/kg body wt daily, 10 doses total) doses of BCNU (carmustine) on intracerebral 9L tumors. Tumor ADC values were quantified pre- and post-treatment using a diffusion MRI sequence. Tumor growth, ADC value, and survival were calculated for each treatment group.
In Figure 1, the fractionated treatment animals resulted in a significant increase in ADC values compared to control (vehicle only) with a maximum diffusion value on day 6 which slowly began to decline during the second week of dosage which may be due to the tumor becoming resistant to BCNU. This study reveals that diffusion MRI is a powerful tool in optimization of dosage schedules for anti-neoplastic agents by providing timely information on the spatial effectiveness, and resistance before traditional volumetric analysis allowing for real time adjustments in therapeutic protocol planning.
Abstract ID: 393
Expression of the Gi-protein-coupled chemokine receptor CXCR4 has recently been linked to increased proliferation, invasion and migration of human pancreatic cancer cell lines. However, the relevance of CXCR4 for organ specific pancreatic cancer metastasis in vivo remains unclear. Here we have studied the role of CXCR4 in vivo using noninvasive bioluminescent imaging of metastasis in a mouse model of pancreatic cancer. Murine TD-2 cells, established from pancreatic carcinoma of a TGFα/p53+/− transgenic mouse, display poor metastatic potential, as demonstrated by tail vein metastatic assay in nude mice. CXCR4 mRNA and cell surface protein expression levels were at the lower detection limit in TD-2 cells. To study the role of CXCR4 in pancreatic cancer invasion and metastasis, we generated stable cell lines with similar CXCR4 expression levels as human pancreatic cancer cell lines derived from metastatic lesions. Addition of the CXCR4 ligand CXCL12 induced calcium influx and migration in a dose dependent manner in stable CXCR4, but not in mock or CCR7 transfected TD-2 cell lines. CXCR4, but not CCR7 expression dramatically increased the in vivo metastatic potential of TD-2 cells, resulting in liver and lung metastasis in nude mice. Systemic administration of the selective CXCR4 inhibitor AMD 3100 effectively blocked the enhanced metastatic potential of CXCR4 expressing pancreatic cancer cells. These results indicate that CXCR4 expression accelerates organ specific metastasis of pancreatic carcinoma cells and provide evidence that blockade of the CXCL12/CXCR4 axis is a target for anti-metastatic therapy.
Abstract ID: 394
We are using activatable cell penetrating peptides (ACPP's) to visualize the activity of matrix metalloproteinases (MMP's) in malignant tumors from live animals. Transgenic mice that spontaneously develop invasive breast tumors through the expression of the polyoma middle T oncogene were injected with a peptide that is cleavable by MMP-2 and MMP-9. As expected, cleaved peptide was recovered from the tumor specimen at 2 hours and at 6 hours. Further examination demonstrated this cleaved peptide to be primarily localized to the stromal compartments of the tumor, (shown in the attached figure), consistent with previous reports of MMP expression in this breast cancer model as well as our own antibody stains. These results indicate that ACPP's offer a reliable measure of MMP activity in vivo.
Abstract ID: 395
Poster Session: Imaging in Neuroscience, Part 2
Abstract ID: 396
Neuroinflammatory processes are involved in numerous neurodegenerative diseases such as stroke and Alzheimer disease. Neuroinflammation has an acute and long-term impact on brain damage extension, through induction of several pro and anti-inflammatory proteins. Among these, VCAM-1 (Vascular Cell Adhesion Molecule), a member of the immunoglobulin superfamily, mediates leukocyte adhesion to endothelial cells and therefore can be considered as a potential marker of neuroinflammation. Our work aims at developing a specific tracer of induced VCAM-1 in the brain, for in vivo Positron Emission Tomography (PET) imaging.
The SELEX technology produces aptamers from a library of oligonucleotides able to tightly bind a specific target. The enrichment process consists in repeated rounds of selection followed by Polymerase Chain Reaction (PCR) amplification of ligand sequences, thereby promoting those which recognize the target. Different selection strategies have been tested depending on:
i) the target □ native VCAM-1 protein expressed on cell surface, or purified fragment,
ii) the starting library of sequences □ totally random sequences or mutated variants of a specific sequence.
A sequence was isolated from a first selection, using only the purified fragment : the aptamer showed a high affinity for the fragment (Figure 1). However binding to cells overexpressing the target was not saturable (Figure 2).
Alteming rounds of selection against the purified fragment and the native protein are presently ongoing in order to improve recognition of VCAM-1 expressed on cells.
Supported by EMIL (European Molecular Imaging Laboratories), EU contract LSH-2004-503569.
Abstract ID: 397
Increasing prevalence of neurodegenerative diseases, such as Alzheimer and Parkinson diseases, or pathological conditions, such as stroke, in ageing population has led to an increasing need to develop new diagnostic and therapeutic follow-up tools. Neuroinflammation has been shown to be induced by and to induce neurodegeneration. Amongst potential markers of neuroinflammation, the peripheral benzodiazepine receptor (PBR, expressed mainly by microglial cells) has been chosen. Here, we aim at showing that PBR are induced in two different models of neurodegeneration and that μPET is a valid method to screen and validate pre-clinically new brain imaging agents.
PBR expression was imaged using a μPET imager following intravenous injection of the radiotracer [11C]PK11195 in rats (7days after intrastriatal injection of AMPA) or transgenic mice over-expressing both presenilin-1 and β-amyloid. In vitro autoradiography using [3H] PK11195 was used to confirm μPET imaging results, and correlated with histological analysis on brain sections.
In rats, a strong induction of PBR was imaged using both μPET (2.3-fold increase in ipsilateral vs contralateral side) and autoradiography (14-fold increase), and localised within and around the lesion delineated by histology. In transgenic mice we observed a similar induction (2.50±0.44 vs 1.28±0.04%ID/cc in transgenic WT/PS1 vs APP/PS1 mice respectively, data expressed as % of injected dose/cc of tissue, mean ±SD; 1.95-fold increase).
μPET images quantification of PBR through [11C]PK11195 uptake correlates well with in vitro autoradiography, although it shows a lower sensitivity, in a general model of neurodegeneration (eg. striatal lesion in rats, allowing rapid and accurate screening of future new compounds) and in a more specific model of disease (eg. transgenic model of Alzheimer disease in mice). Here we show that [11C]PK11195 and μPET may be used in animal models to evaluate in vivo new therapeutic strategies, and that μPET technique is suitable to screen and validate new radiotracers in small animals.
Abstract ID: 398
Multiple sclerosis (MS) is a chronic inflammatory demyelinating disease of the central nervous system (CNS). Most MS lesions are well detected by magnetic resonance imaging (MRI) which has significantly improved our understanding of this disease. Histopathological correlates of conventional MRI techniques, however, are not well defined. In the present study we correlated histopathology and MRI signal intensity changes in the experimental autoimmune encephalomyelitis (EAE) animal model of MS.
EAE was induced by the adoptive transfer of a PLP139–51 specific T-cell clone in SJL/J mice. EAE lesions were predominantly located in the spinal cord, lower and upper pons and cerebellum. Histopathology in the early disease phase was characterized by infiltrations composed of abundant polymorphnuclear granulocytes, macrophages and T-cells. Quantitative morphometry was obtained for inflammatory cell density, myelin and axonal density, immunoglobulin (Ig) deposition, astrogliosis and microglia activation. Lesional cellularity and myelin density correlated inversely with signal intensities on T1- and T2-weighted images, whereas IgG deposition led to a signal intensity increase on T1- and T2-weighed MR images. In the perilesional area, microglia activation, astrogliosis, and IgG deposition correlated positively with T1- and T2-weighted MRI signal intensities. Gd-DTPA enhancement was primarily observed in perilesional areas with activated microglia cells and correlated with Ig deposition.
In summary, we present a mouse EAE model suitable for the assessment of lesional evolution, therapeutic strategies and the development of cellular and molecular imaging techniques by MRI.
Abstract ID: 399
Identification of in vivo markers for neuropathology is important for early diagnosis but also for evaluation of therapeutic strategies. 123I-FPCIT single photon emission computerized tomography (microSPECT), revealing alterations in the binding capacity of the dopamine transporter, is an established diagnostic tool for presynaptic dopaminergic deficits such as in Parkinson's Disease (PD). Here we report on the results of a combined SPECT - MRI study study which aimed to investigate whether the Magnetization Transfer Ratio (MTR) as determined with MRI is able to discriminate the neuropathology in the unilateral 6-hydroxydopamine (60HDA) rat model for PD from normal controls. To that end control (n=5) and 6OHDA-lesioned animals (n=5) were subsequently imaged using 123I-FP-CIT pinhole microSPECT and MTR based MRI (Bruker,7T). All 6OHDA-lesioned animals displayed decreased DAT binding capacity in the affected striatum (left-right ratio: 0.23±0.05), indicating a successful lesion in the substantia nigra (SN). MTR was measured in the corpus callosum (CC), the left and right SN and striatum (STR) of all animals. No significant asymmetry was observed between left and right SN and STR of controls and lesioned animals (p>0.05) and the MTR measured in CC of 6OHDA-rats was not different from controls (p>0.05). However, MTR of the left and right striatum was significantly lower in 60HDA rats than in control animals (p<0.01, * Figure). MTR has been correlated to the degree of myelin loss and axonal destruction in animal and human studies and has thereby been related to degeneration of dopaminergic neurons in PD patients. Further investigation needs to elaborate the MTR findings in the intact striatum of 6OHDA-lesioned rats.
MTR in STR, SN and CC in normal controls and 60HDA-lesioned rats.
Abstract ID: 400
In contrast to rats, in cats it was difficult to visualize extra-striatal binding. The comparison of the relative tracer uptake in different brain regions demonstrated that in comparison to 11C-Raclopride the tracer 18F-Fallypride had a substantially higher affinity to its specific binding sites.
In mice bearing subcutaneous Gli36 gliomas, 18F-Fallypride allowed the differentiation between wild-type gliomas and D2-receptor expressing gliomas.
18F-Fallypride kinetics in striatum and blood resemble 11C-Raclopride and 18F-FDG kinetics.
Abstract ID: 401
ClearPET Neuro is a high resolution, high sensitivity PET system dedicated to brain studies on small animals. Besides the brain the heart is frequently characterized by high tracer uptake. This is likely a cause of additional scatter, both in the object as well as in the tomograph gantry, and random coincidences. To study the influence of these effects on slice dependent scatter fraction and NEC rate was the goal of the present investigations.
For standard measurements of the rat brain the rat's head is positioned in the center of the field of view (FOV) of the scanner. Since the body of the rat cannot be completely scanned the heart is most often lying at the edge of the FOV. To study the scatter effects a cylindrical plexiglass phantom was used as simple rat phantom. The cylinder had a drilling for a line source and a small cylindrical reservoir that simulates the heart. In various arrangements the phantom was scanned either with the line source only or with additional activity in the reservoir. The phantom was measured in the centre of the scanner but at different axial positions, so that the reservoir was located just inside or outside the FOV. Investigations of slice dependent scatter fractions for the line source showed that the scatter fraction is remarkably higher when the reservoir contains activity. It increases from ~25% in slices at the far end from the reservoir to ~40% at the close end. It was found that this increase holds true both for the reservoir placed at the border of the FOV as well as outside the FOV. These data clearly support the view that there is a need to detect scatter coincidences in rat brain studies and to develop means to correct for or suppress this potential source of error.
Abstract ID: 402
Neural precursor cells (NPCs) are capable of tracking glioma tumors and thus offer a potential to deliver therapeutic molecules. We have engineered both glioma cells and NPCs with lentiviral vectors (LVs) bearing fusions of bioluminescent and fluorescent markers and shown in real time that in vivo-marked neural precursors display self-renewal and have migratory capabilities. We have also created a human glioma cell that stably express Renilla luciferase (Rluc)-DsRed2 and firefly luciferase (Fluc)-GFP fusion proteins and shown a direct correlation between the subcutaneous tumor volumes and the photon count in mice. In addition we have also engineered NPCs to deliver regulatable forms of therapeutic proteins using FKBP based regulatable system in which the protein of interest is reversibly retained in the endoplasmic reticulum (ER) of the cell as an aggregate until selectively released by a cell-permeable CAD (conditional aggregation domains)-ligand, that dissolves CAD aggregates. We show that NPCs expressing recombinant s(secretable)-TRAIL (tumor necrosis factor-related apoptosis-inducing ligand), migrate towards the intracranially implanted gliomas in mice and have profound antitumor effects. These studies reveal the potential of NPCs and the engineering of regulatable systems for controlled gene delivery as therapeutically effective means for the treatment of gliomas and also provide important tools to evaluate the migration of NPCs and death of glioma cells in vivo.
Abstract ID: 403
Two-dimensional fluorescence imaging has proven effective for noninvasive in-vivo small animals studies in various contexts. It is however mostly restrained through the blurring effect of photon diffusion which easily leads to equivocal interpretation of the depth and extent of the signal. To complement the information easily retrieved from planar imaging, the combination of optical tomography and fluorescently labeled probes delivers the volumic distribution of fluorophore concentration by measuring surfacic fluorescent intensity as a function of illumination. So far this newcoming technique gives a satisfactory accuracy on white and nude rodents; its applicability for in-vivo studies on the central nervous system on pigmented mice needs to be assessed.
We developed a novel tomographic technique suitable for the localization and quantification of fluorescent reporters in the head of small animals. By illuminating from several angles using optical fibres and detecting the surface radiance from above with a filtered backilluminated CCD and low-f lens lens, we obtain a series of two-dimensional images of the phantom or mice lying horizontally. For the sake of coherence with the murine model of stroke in use in our laboratory, we measure on the C57BI6 pigmented mouse strain. In the present device, continuous wave excitation is obtained with 680nm laser light and fluorescence of Cy7-like dyes is collected at 780nm. The system is mounted inside a light-tight box.
Light propagation inside the animal upto the surface is modelled using a Monte Carlo method fed with MRI anatomical data of the mouse head, thus offering the possibility to take tissue inhomogeneities into account. We solve the inverse problem for the reconstruction of the fluorescent marker distribution using the efficient ART or SIRT.
(Support: Europäischer Fonds für regionale Entwicklung -EFRE-and Hermann & Lilly Stiftung.)
Abstract ID: 404
Abstract ID: 405
Recirculation induced transient recovery of CBF but not CMRO2, and ICP rose progressively until cerebral perfusion pressure decreased below 50 mmHG: At that stage, symptoms of transtentorial herniation were recognized. In the 8 remaining (benign) cats, volumes of severe ischemia during MCAO covered only 10–45% of the ipsilateral hemisphere, whereas volumes of moderate ischemia covered 15–30%. In these cats, ICP did not rise in the reperfusion phase above 20 mmHG, and secondary deterioration of CBF and CMRO2 in regions outside the ischemic territory was not observed.
Abstract ID: 406
A key component of medical image analysis, image registration essentially consists of bringing two images into spatial alignment. In particular, by registering each pair of consecutive slices in a stack, we can recover a geometrically coherent 3-D alignment and reconstruct a 3-D volume.
The large variety of registration applications (3-D volume reconstruction, cross-dye histology gene mapping, etc.) induce an equally diverse set of requirements in terms of accuracy and robustness. In turn, these directly translate into regularization constraints on the deformation model, which should ideally be specifiable in a user-friendly fashion. Classical techniques usually offer only limited (and often indirect) control over the characteristics of this regularization process. More over, little if any information is taken into account about the images to be registered.
We propose here an adaptive regularization approach whose rigidity is informed by the registration application at hand and whose support is controlled by the topology of the images to be registered. Our technique first computes a block-matching field, and then estimates affine or rigid transformations, in a robust fashion, over sliding circular neighborhoods cut to fit the contours of the histological tissues. By varying the regularization radius, we control the local rigidity of the regularized displacement field, and hence, that of the transformation. Such approach enables us to tackle a wide variety of registration tasks with the same generic framework.
Volume reconstructions from calbindin- (see Figure) and myelin-stained histological sections of the left and right hemisphere of a transgenic mouse brain and cross-modality registration between Lhx5, Penk1 and Drd4 gene expression slices demonstrated the effectiveness of our method (see in particular the olfactory bulb and the hippocampus).
Abstract ID: 407
The Parkinson's disease (PD) is a progressive neurodegenerative disease which is mainly due to the loss of dopaminergic neurons in nigrostriatal pathway. The dopaminergic pathway transmits the dopamine to the several areas in the brain from the substantia nigra and involves a lot of physiological functions. Divalent Manganese (Mn2+) has been shown as a useful MRI contrast agent tracing specific neuronal pathway. This paper provided a useful method of in vivo manganese-enhanced magnetic resonance imaging (MEMRI) to visualize the pattern of neural projection from the Mn2+ injection site. For this purpose, the 290g ± 10g Sprague-Dawley rats were injected with MnCl2 (1 μl, 100 mM) into the substantia nigra under the stereotactic unilateral microinjection. At 6h, 12h, and 24h post-injection, the T1 weighted images were acquired by NTU Bruker 3T MRI with FOV = 4 × 4 cm, TR/TE = 500/9 ms, matrix size = 256 × 256, and thickness is 1mm, yielding in-plane resolution of 156μm. The area mainly highlighted by Manganese accumulation in the Lateral Septal Intermediate (LSI) but less in Caudate-putamen complex (Cpu) can be observed at 24h postinjection(Figure 1). The enhancement of MRI signals can also be observed in the olfactory bulb (OB), showing the other neural pathway projecting from substantia nigra. Using the assessment of MRI with MnCl2, we can monitor the profile of neural projection from substantia nigra and evaluate the functions of neural circuits in the brain in vivo to provide more accurate information in the animal model study of PD.
Abstract ID: 408
Abstract ID: 409
Intracellular Ca2+ concentration gradients are highly dynamic. The spatio-temporal profile of Ca2+ signals in restricted cellular domains, critically defines the activation of different signal transduction pathways. Ca2+ signaling is also involved in driving cell-to-cell communication, for example at electrical or chemical synapses in the brain. Optical detection of Ca2+ signals in cellular microdomains and/ or in sub-populations of cells can therefore provide ‘real-time’ information about biological processes occurring within an integrated system.
To date, whole animal in vivo bioluminescence imaging has been largely undertaken with gene expression reporters using the firefly luciferase system. In an alternative approach, we have developed the genetically encoded bioluminescent reporter, GFP-aequorin (GA). Binding of Ca2+ to aequorin, results in an intra-molecular non-radiative energy transfer to GFP and emission of light (Λmax 510 nm). GA reporters can be molecularly targeted to sub-cellular domains and the dynamics of Ca2+ signals can be monitored with excellent signal-to-noise ratios in single-cells and tissue slices[1].
In more recent studies, we have constructed several transgenic animals with a Cre-mediated site-specific gene activation of the targeted GA reporters. Kinetic analysis of Ca2+ dynamics in different experimental paradigm show that changes in local Ca2+ concentrations accompanying specific physiological functions can be directly visualized in vivo in the living animal with temporal resolution in the millisecond range. ‘Real-time’ measurements of Ca2+ were undertaken using the Photon Imager, a new device for bioluminescence and fluorescence imaging, based on a photon counting technique in the visible and near infrared range.
High resolution mapping of biological events using GFP-aequorin will have important applications in the systematic analysis of physiological processes in animal models for human disease or in the development of new therapeutics.
Abstract ID: 410
Abstract ID: 411
Standard group analysis of human cortex in positron emission tomography (PET) suffers from missegmentation and misregistration issues. These degrade data quality through introduction of partial volume effects (PVE) from low uptake compartments; also, registration precision does not easily permit separation of adjacent banks of gray matter in cortical folds.
Cortical flattening allows sampling the gray matter and analysis in its natural 2D topology, avoiding introduction of additional PVE. To enhance statistical acuity of this new method, topologically correct smoothing of PET data in 2D space was evaluated.
Seven right-handed young healthy men received MRI for cortical segmentation and 18F-FDG-PET after injection of 370 MBq. Each subject's individual scans were co-registered using VINCI software (MPI Cologne). Left hemispheric cortex was segmented and reconstructed and then flattened and spherically registered to a standard atlas using SureFit and Caret software (van Essen, WUSTL). Individual metabolism was scaled to mean and surface-projected. 2D smoothing was applied using one generation of neighbours.
For comparison, standard 3D normalization was carried out using SPM2 and then overlaid on the atlas in SPM space. Data was scaled to mean metabolism, mean and standard deviation over all subjects were calculated vertex-by-vertex.
Cortical flattening alone decreases standard deviation of glucose metabolism by 17% compared to SPM2. When smoothing with one generation of neighbours, it decreases by 25%.
We have demonstrated the feasibility of fully surface-based PET data integration across subjects. Data fidelity is greatly improved in comparison to volume registration (Figure 1A), while standard deviation across subjects on a vertex-by-vertex level decreases by 25% (Figure 1B). Statistical power can be further increased by using a larger smoothing radius, compromising on absolute resolution as desired.
Abstract ID: 412
The subventricular zone (SVZ) is a neurogenic center in the rodent brain. Here, astrocyte-like neural stem cells (NSCs) lining the ventricles produce neural precursors which migrate within the rostral migratory stream (RMS) exclusively to the olfactory bulb (OB). To label the NSCs, green fluorescent 1.63-micron diameter iron-oxide particles were injected directly into the lateral ventricle. Once particles are endocytosed by the NSCs, through asymmetric cell division, the particles can be transferred to the daughter neural progenitors and carried along during their migration. In vivo, T2* weighted gradient-echo MRI at 100 micron isotropic resolution was used to monitor this migration. Dark, susceptibility induced contrast from the particles was detected in MR images along the entire migration route, in the ventricles, within a confined strip along the RMS, and fanning out inside the OB, as early as one week post injection (Figure 1A,B). This contrast increased and persisted until five weeks post injection, the longest time point used. These anatomic locations correspond to areas observed by microscopy to contain fluorescent particles. Prussian blue staining for iron demonstrated that particles were confined to the ventricle and RMS as well as in the periglomerular layer in the OB. Cells were observed containing as few as single particles, and as many as 15 particles. Immunohistochemistry identified that particles were inside GFAP+ NSCs in the SVZ (Figure 1C), migrating PSA-NCAM+ and Doublecortin+ neural precursors within the RMS and OB (Figure 1D), and Neu-N+ mature neurons in the OB. Electron microscopy confirmed the intracellular nature of these particles (Figure 1E). This is the first demonstration of in vivo cell labelling of stem cells for MRI and endogenous stem/precursor cell migration by MRI.
A) MRI of RMS (arrow) and OB immediately post-injection, B) MRI of RMS and OB five weeks post injection, C) particles in GFAP+ NSCs at the ventricle, D) particles in Doublecortin+(red) neural precursors in the RMS, E) Electron microscopy identification of intracellular particles in granule cell neuron.
Abstract ID: 413
Our research group has focused on mercaptoimidazole derivatives as the lead structure to design radioactive probes for in vivo imaging of 5-HT1A receptors. This type of compounds can be labelled directly with 11C or can be used to prepare tripodal ligands suitable to stabilize the fac-[M(CO)3]+ (M = Re, 99mTc) moieties. [1,2] Interestingly, this versatility allows an easy assessment of the influence of the bulkier metallic fragment on the biological properties of PET (11C) or SPECT (99mTc) radioligands bearing a common pharmacophore.
In this communication, we report on the radiosynthesis, characterization and biological evaluation of mercaptoimidazole piperazinyl derivatives labelled with 11C (
Abstract ID: 414
The purpose is to detect the expression of endogenous genes in live brains using phosphorothioate oligodeoxynucleotide (s-ODN) conjugated to dextran-coated monocrystalline iron oxide nanoparticles (MION) as a reporter probe in magnetic resonance (MR) imaging. We infused male C57black6 mice intracerebroventricularly with one of two conjugates of MION: (1) MION-A26, a targeting conjugate which contains an s-ODN of 26 bases with sequence complementary to c-fos mRNA, and (2) MION-Ran, a non-targeting conjugate of s-ODN with a random sequence as a negative control. We selected c-fos mRNA as the target, because Fos peptide, the product of c-fos mRNA, is an essential component of activator protein-1 (AP-1) which regulates many genes that are responsive to cells at stress in the central nervous system. Brain cells express a very minimal level of c-fos transcript at resting conditions, but they are activated to express c-fos mRNA at 1–2 orders of magnitude higher by cerebral ischemia (stroke or cardiac arrest); the transcript is mostly on the cortex, hippocampus. We demonstrated that MION-A26 significantly increased MR susceptibility relaxation rate (R2*) in a mouse model of cardiac arrest. The elevation was observed in brain regions of the somatosensory cortex and the hippocampus, corresponding to elevation in c-fos mRNA, but not b-actin mRNA, using molecular assay in post mortem samples. (NSR01045845, P41RR14075 and the MIND Inst).
R2* values in vivo with and without cardiac arrest using bilateral carotid occlusion for 30-minutes that induces forebrain is chemia (FbIR).
Postmortem MR Microscopy and Conventional Molecular Assay.
MR Microscopy and Conventional Molecular Assay.
Abstract ID: 415
A noninvasive method for detection and quantification of β-amyloid plaques (Aβ) in vivo would have wide utility for premortem diagnosis of Alzheimer's disease (AD) and monitoring anti-amyloid therapy. Candidate Aβ-targeted imaging agents would be expected to demonstrate five critical characteristics: a) binding to Aβ plaques; b) permeability across the blood-brain barrier (BBB); c) metabolic stability; d) efficient synthesis for rapid formulation; and e) use of a readily available isotope, such as technetium-99m, for widespread deployment. Initial efforts to design Tc-99m labeled chrysamine G (CG) and Congo Red (CR) derivatives or mixed functionalities such as isonitriles have been unsuccessful. While exploring peptide-based motifs for biomedical applications, we have synthesized a BBB permeation peptide
Abstract ID: 416
[18F]FM665017, a specific α7 nicotinic receptor antagonist, is a candidate ligand for investigating cerebral α7 receptors in vivo. For this study FM665017 PET scans were performed in two baboons. Each animal underwent two PET scans: a 120 minute control study and a 90 minute blocking study using 1 mg/kg pretreatment with AZ10456184–019, an α7 agonist. Arterial samples were acquired throughout each study to measure the metabolite-corrected blood input function. MRI scans were also acquired for anatomic coregistration with PET. Regions-of-interest were drawn for several cortical regions to generate time-activity curves (TACs), which were then quantified using a 1-tissue, 2-parameter compartmental model to measure the blood-to-tissue transport rate (K1) and the total distribution volume (DV). The mean parameter estimates are shown in Table 1. In comparison with the control studies, there was over a 30% increase in both K1 and DV following blockade. The implication is that blood flow increased in the blocking study, which produced a similar increase in DV. Peripheral metabolism was altered as well, as the percentage of FM665017 increased from 35% in the control studies to 57% in the blocking studies. The possible effects of the blocking agent on blood flow may confound the ability to detect blocking of FM665017 at the α7 receptor sites. We will investigate alternative blocking agents, and studies are also underway to better characterize peripheral metabolism and to determine if labeled metabolites might enter the brain.
Abstract ID: 417
Accurate evaluation of tumor margins during resection of intracranial neoplasms would aid neurosurgeons in performing a more complete resection while allowing greater sparing of normal tissue. We evaluated a near-infrared fluorescent iron oxide-based nanoparticle to 1) determine accuracy of tumor margin delineation using the agent, and 2) determine the degree of particle uptake in tumor cells verses host phagocytic cells.
The 9L gliosarcoma cell line, transfected to express GFP, was implanted into brains of Wistar rats and nude mice and allowed to grow for 7–10 days. Animals were evaluated 24hrs after intravenous injection of 15mg/kg CLIO-Cy5.5. For margin evaluation, multiple high resolution images were obtained, and using a fixed algorithm and location on each image, under or overestimation of the tumor boundary was measured. In separate animals, tumors were disaggregated into single cell suspensions and evaluated by FACS for particle uptake, CD11b expression (macrophage and activated microglia populations), and GFP expression (tumor cells). Confirmatory fluorescence microscopy was also performed.
Mean overestimation of tumor borders was 0.002mm in mice and 0.024mm in rats. Thus, in both models highly accurate margins were reported by the nanoparticle. The slightly larger boundary in rats was likely secondary to a more robust host response to the tumors. There was a marked increase in fluorescence in both CD11b+ cells and tumor cells in both rats and nude mice. In both model systems, tumor cells had a greater uptake, as evidenced by a greater increase in relative mean fluorescence, compared to CD11b+ cells. Uptake by both tumor cells and host phagocytic cells potentially increases the generalizability of the approach. Cells which were CD11b- and GFP-had minimal uptake of the nanoparticle. Fluorescence microscopy showed scattered clumps of particles within CD11b+ and GFP+ cells, suggesting lysosomal distribution of the particle at the 24hr time point.
Poster Session: Imaging in Atherosclerosis and Heart Disease
Abstract ID: 418
Abstract ID: 419
Angiogenesis is integral to the progression of atherosclerosis. αvβ3-targeted paramagnetic nanoparticles (NP) maybe used to quantify early neovascular expansion of the vasa vasorum and to deliver an anti-angiogenic drug (fumagillin) directly to atherosclerotic plaques. This study evaluates the predictive utility of MRI signal enhancement at the time of treatment to forecast therapeutic response.
Abstract ID: 420
Abstract ID: 421
The endothelial glycocalyx (EG) lines vascular endothelium and consists of a membrane-bound part, containing proteoglycans and glycoproteins, and a plasma-derived part incorporated in the membrane-bound network. Located between blood and endothelium, it acts as a mechanosensor and a selective barrier for large molecules. The EG is delicate and sensitive to environmental changes. Therefore, imaging of the EG to assess functional parameters requires physiological conditions and circumstances. Whereas in vivo imaging of the EG has been done in microvessels, conventional microscopic techniques do not allow in vivo imaging of EG in macrovessels. Two photon laser scanning microscopy (TPLSM) is a new technique that enables imaging of vascular structures in intact larger vessels.
We now apply TPLSM to image the fluorescently labeled (WGA-FLTC) EG in murine carotid arteries after mounting them in a perfusion chamber (Figure 1). This high resolution setup assures good penetration depth without extensive vessel processing. In future experiments, we will use this technique in vivo as well, ensuring physiological conditions when assessing functional parameters of the EG.
Endothelial glycocalyx in two different carotid arteries.
3D-reconstructions of a carotid artery mounted on a perfusion chamber stained with SYTO44, WGA-FITC and eosin imaged with a two photon microscope. Figure 1A shows the medial layer with on top the endothelial cell nuclei and the EG. Figure 1B shows a general overview of the adventitia, medial layer and the endothelial layer including the EG.
Abstract ID: 422
(A) Experimental setup with exposure of carotid plaque (arrow) and corresponding laser scanning microscopy images from CLIO-Cy5.5 injected mouse. A phantom (*) is used as reference during microscopy; (B) Epifluorescence images of excised carotid plaque from a CLIO-Cy5.5 injected mouse; (C) Fluorescent microscopy and immunohistochemistry, showing Cy5.5 signal that co-localizes with macrophage staining of an adjacent section.
Abstract ID: 423
Both, superparamagnetic iron particles and gadolinium compounds targeted to molecular structures have been investigated for molecular imaging. In the present investigation we compared four different iron based MRI contrast agents in respect to their magnetically and biological properties.
Abstract ID: 424
Abstract ID: 425
Figure A. shows enhanced uptake of TAA5 (arrow) at the symptomatic carotid artery region. Histologic analysis shows a vulnerable shoulder from thin fibrous cap atheroma (EVG staining, Figure B.) and extensive presence of AA5 (anti Annexin A5 staining, Figure C.)
Abstract ID: 426
Abstract ID: 427
Abstract ID: 428
Abstract ID: 429
Caspases (
Radiolabeled Annexin V derivatives were introduced as apoptosis detecting radiotracers, but they preferably bind to negatively charged phospholipids (i.e. phosphatidylserine) that externalize to the outer cellular membrane during apoptosis. In contrast 5-pyrrolidinylsulfonyl isatins represent a rare class of potent nonpeptidyl caspase inhibitors which bind selectively to the downstream effector caspases 3 and 7. Radiolabeled 5-pyrrolidinylsulfonyl isatins should result in nonpeptidyl
Non-radioactive (S)-(+)-5-[1-(2-methoxymethylpyrrolidinyl) sulfonyl]isatin that has been shown to possess cardioprotective potential in isolated rabbit hearts after ischemic injury [1] was chosen as lead structure. The radiosyntheses of two feasible imaging agents were performed yielding the PET-compatible radiotracers 11C-and 18F-CbR, putative biomarkers of apoptotic tissues [2].
In conclusion, a nonpeptidyl 11C-labeled compatible as well as a 18F-labeled caspase inhibitor based on 5-pyrrolidinylsulfonyl isatin, were synthesized for the first time as putative CbR for the exclusive in vivo detection of apoptosis.
Abstract ID: 430
Soluble CD40 has been shown to be elevated in serum in acute coronary syndromes and CD40 expression is increased in atherosclerotic plaques. Aim of the present study was to investigate whether CD40 could be used as an imaging agent to detect atherosclerotic plaques in Apo-E-/-mice.
Abstract ID: 431
Macrophages are critically involved in the pathogenesis of many diseases including, rheumatoid arthritis, atherosclerosis, and tumorigenesis. Macrophages are recruited and activated by many signals and indeed, activation of macrophages is crucial for their function, provoking an interest in them as targets for imaging. We hypothesized that phage display could be used to identify peptide ligands discriminatory of activated macrophages. The display of peptide libraries on the surface of bacteriophage offers a way of searching rapidly and efficiently for peptides with specific binding properties among the millions of peptides contained within such a library. Although initial selection of phage-displayed peptides is relatively straightforward, identifying and validating relevant hits from a selection process often remains a technical challenge because most screens yield 10–50 of such hits. We therefore developed a computational approach to display data from high-throughput enzyme-linked immunosorbent assay (ELISA) screens to visualize display affinity and specificity for identified peptides. Applying this approach, we show here the discovery and validation via fluorescence microscopy of two novel peptide sequences specific for activated macrophages. The described methodology and sequences should be useful in the design of new macrophage directed imaging agents.
Abstract ID: 432
Since myeloperoxidase is present at higher concentrations in vulnerable compared to stable plaque, the enzyme may by a useful target for diagnostic imaging of unstable plaque. A cysteine constrained phage display peptide library (f88–4/Cys6, G. Smith, University of Missouri) and phage display were used to select for peptides with affinity for myeloperoxidase immobilized on polystyrene plates. The consensus sequences of the four phage clones and one control were: (MPO-1) AMPVCSLDRKACTAYQ (2 clones)
(MPO-6) HVLGCSYATNSCAHNA (3 clones)(MPO-15) TVMLCAPMEQGCRWM C (2 clones)
(MPO-17) RSGQCHDDTWNCLACV (5 clones)
control QMKQCWVNTATCWGLS
Each phage was conjugated with NHS-MAG3 and radiolabeled with 99mTc. Myeloperoxidase binding was evaluated by adding the same molar excess of 125I-myeloperoxidase to each of the five 99mTc labeled phage clones with analysis by Sepharose 6B open mini-column. The 99mTc counts identified the phage fractions while the 125I counts therein identified the concentration of bound myeloperoxidase. After correcting for control phage binding, the order was MPO-17, 6, 15 and 1 with 8.0 %, 5.8% 4.5% and 3.9% bound, respectively. Thus, all clones showed some binding to myeloperoxidase but the highest was MPO-17. This peptide and a control peptide were synthesized with a GGC tripeptide tail for labeling with 99mTc. The 99mTc peptide was evaluated by size exlcusion HPLC which showed a shift to higher molecular weight upon addition of myeloperoxidase while no shift was observed for the control peptide. Binding of this peptide to myloperoxidase was also evaluated by surface plasmon resonance which showed a 3-fold higher binding affinity to myeloperoxidase of the MPO-17 peptide compared to the control peptide. Therefore in this investigation, one peptide was identified by phage display that shows affinity for myeloperoxidase. The peptide MPO-17 merits further study as a potential radiopharmaceutical for the scintigraphic detection of myeloperoxidase in vulnerable plaque.
Abstract ID: 433
Abstract ID: 434
Abstract ID: 435
Abstract ID: 436
Fibroblasts at the tumor-host interface can differentiate into myofibroblasts and pericytes, enhancing angiogenesis as well as endothelial cell motility, and contributing to guidance and stabilization of endothelial sprouts [1]. Thus, co-cultures of ovarian carcinoma SKOV-3 cells with myofibroblasts facilitated invasion of endothelial cells in three-dimensional cultures [2]. However, gaining deeper insight regarding their interactions with the endothelium during tumor angiogenesis requires specific in-vivo monitoring. With the motivation of tracking fibroblasts in a tumor model, we utilized caveolae dependent uptake of biotin-BSA-GdDTPA for ex-vivo labeling of fibroblasts. Spin echo images of control and labeled cells were acquired on a daily basis on a 400 MHz spectrometer. ΔR1 (ΔR1 = R1 (labeled)-R1 (unlabeled)) was determined from the T1 weighted images. The intracellular relaxivity (R IC ) was according to the equation: R IC = ΔR1/[Gd], where [Gd] is the intracellular concentration in pmol/cell as was measured by ICP-AES). Remarkably, MR visibility of Gd labeled cells showed a significant increase during the first four cell cycles, despite the expected dilution of the contrast material. This finding is consistent with an initial limited MR detection due to intracellular compartmentization of the CM in granules [3], and subsequent enhanced MR contrast as a consequence of intracellular processing/redistribution of the CM. Moreover, ex-vivo labeled fibroblasts which were co-inoculated with tumor cells into mice, were easily detectable in-vivo by MRI. In accord with a higher cell density, the contrast detected from labeled cells in-vivo was greater than the one brought about in-vitro. This study shows that labeling of fibroblasts with biotin-BSA-GdDTPA is feasible and provides the basis for future use of in-vivo MRI to track the recruitment of pre-labeled fibroblasts, during tumor progression.
Abstract ID: 437
Abstract ID: 438
Anticancer immunotherapies using dendritic cells (DCs) are currently being explored in the clinic. For these therapies accurate delivery and subsequent migration of vaccinated DCs to regional lymph nodes (LNs) is of great importance for effective stimulation of the immune system. Using magnetically labeled DCs, we investigated the potential of magnetic resonance imaging (MRI) for cell tracking to monitor DC therapy. Autologous monocyte-derived DCs were labeled with a clinically approved superparamagnetic iron oxide (SPIO) formulation and 111In-oxine separately, and co-injected into the draining LNs of stage III melanoma patients who were scheduled for radical LN dissection. Two days after vaccination patients were monitored by scintigraphy and MRI at 3 Tesla. We show that in vivo MR tracking of magnetically labeled cells is feasible in humans for detecting low numbers of DCs together with detailed anatomical information. In contrast to scintigraphic imaging, MRI allowed assessment of accurate DC delivery and migration patterns. SPIO-labeled cells were detected in the injected LN, but also in nearby LNs (Figure 1A). The LNs were then resected and imaged by MRI at 7 Tesla (Figure 1B). The presence of SPIO-labeled cells in the LNs was confirmed by histology (Figure 1C). Moreover, histochemistry revealed that SPIO-labeled DCs had migrated into the T cell areas (Figure D) of both the injected and nearby LN. In conclusion, MRI cell tracking using iron oxides appears clinically safe and extremely well suited to monitor cellular therapy in humans.
3T-MRI of patient injected with SPIO-labeled DCs. Arrows indicate SPIO-positive LNs.
Abstract ID: 439
The purpose of this study was to elucidate whether ferumoxides-protamine sulfate (FePro) labeled splenocytes from tumor (glioma) bearing rat would enable us to delineate the margin of the same tumor in a second recipient rat by MRI. Gliosarcoma (9L) cells were implanted intracranially in syngenic Fisher 344 rats. Fourteen days after inoculation spleens from tumor bearing as well as from control rats were acquired, homogenized and splenocytes were collected, labeled with FePro and re-injected in tumor bearing rats. In vivo multiecho T2-weighted, multiecho T2*-weighted and 3D gradient echo images were obtained by a 7 Tesla magnetic resonance (MR) system at 3, 24 and 72 hours after injection. After 72 hours images, rats were euthanized, and the brain with tumors were sectioned, and processed for Prussian blue and immunohistochemistry. Migration and homing of labeled splenocytes (splenocytes from tumor bearing rats) were detected as low signal intensity along the margin of the tumor on MRI as early as 24 hours and the low signal intensity rim was more obvious at 72 hours on in vivo MRI and the tumor could be delineated from the surrounding brain tissues. In contrast, labeled splenocytes from control rats without tumors did not migrate or home along the margin of the tumor and tumor could not be delineated from the surrounding normal brain tissues on MRI. Prussian blue staining showed iron positive cells along the margin of the tumor that received FePro labeled splenocytes from tumor bearing rats but not in the tumor that received labeled splenocytes from control rats. Immunohistochemistry showed accumulation of cytotoxic T-cells at the site of tumor. In conclusion tumor cell-lysate sensitized cells (such as T-cells, APC or others) can be used as cellular probe to detect the margin of the tumor or early detection of metastatic or recurrent tumor.
Abstract ID: 440
Cancer patients can harbor high numbers of activated tumor-specific CD8 T cells that exhibit tumoricidal activity in vitro but fail to eradicate the tumor in vivo. Recent studies indicate that regulatory CD4 T cells (Treg) may contribute to immune evasion in cancer by containing tumor-specific T cell responses, however the immunosuppressive mechanisms utilized by Treg have been mostly investigated in vitro. Here we employ fluorescent protein tomography (FPT), a novel noninvasive whole-body imaging technique, for spatiotemporal kinetic mapping of GFP+ tumor-specific CD8 and Treg cells during anti-tumor immune responses. Photons emitted by GFP+ cells were collected from combined measurements obtained after stimulation from different sources and three-dimensional reconstruction of cell distribution was achieved with specific algorithms and algebraic reconstruction techniques. We have used CD8 cells and Treg with specificity for different peptides of the same protein (hemagglutinin, HA) and that constitutively express GFP. Both CD8 and Treg were initially present at a very low frequency and subsequently challenged with s.c. HA+ tumors. We find that Treg abrogate CD8 T-cell-mediated tumor rejection, however they do not interfere with CD8 T-cell expansion in tumor-draining lymph node. Preliminary observations also indicate that Treg do not control CD8 T-cell accumulation in the tumor bed. However, Treg proliferation upon recognition with antigen in tumor-draining lymph node and quick accumulation in the tumor bed may be responsible for sustained local suppression of the CD8 cells. The unprecedented performance in terms of spatial resolution and quantification accuracy of FPT show the potential of noninvasive whole-body imaging to further study how immune tolerance in controlled in vivo and to develop more effective cancer vaccines.
Abstract ID: 441
Effector CD8 T cells contribute to the clearance of abnormal cells. In vitro studies have shown that specific killing is assured by focused release of specialized cytolytic proteins upon recognition of cognate MHC/peptide complexes on the surface of target cells. However, the interactive and cytolytic activities of effector cells have never been analyzed in complex in vivo environments. Here we have employed intravital multiphoton microscopy in the deep cortex of popliteal lymph nodes coupled with a novel assay for concomitant spatiotemporal visualization of effector CD8 T cells and their targets (intact or lysed). Immunocompetent mice were transferred with naive CD8 T cells that carry a transgenic T-cell receptor directed against a Kd-restricted peptide from hemagglutinin (HA) and that constitutively express GFP. Mice were challenged with HA+ tumors, which generated high numbers of effector HA-specific GFP+ CD8 T cells in tumor draining lymph node. As targets, we used HA peptide-pulsed target B cells previously labeled with blue nuclear (Hoechst 33342) and red cytoplasmic (CellTracker Orange) dyes. Longitudinal monitoring of red/blue fluorescence intensity of single target cells revealed that lysis can be identified in vivo in real-time. Lysis is accompanied by a selective decrease of red fluorescence, which is likely due to diffusion of CellTracker Orange to the extracellular milieu as a consequence of the disruption of membrane integrity. Preliminary studies of cytolytic activity reveal that effector CD8 T cells are highly motile, migrate without apparent directional bias and selectively form stable conjugates upon encounter with cells presenting cognate antigen. CD8 T cells kill their targets within minutes in a two-step process, characterized by initial loss of target cell motility and followed by loss of structural integrity. These experiments show the potential of intravital multiphoton microscopy for intimately dissecting the functionality of immune responses.
Abstract ID: 442
Abstract ID: 443
MRI cell tracking aims to obtain a better understanding of the dynamics of cell-tissue interactions, as well as to aid in guiding the development of effective stem cell therapies.
We have investigated the applicability of MR tracking of a magnetically labeled (magnetofected) neural stem cell (NSC) line following transplantation into the dysmyelinated shiverer mouse brain. LacZ-transfected C17.2 NSCs were magnetofected with Feridex and poly-L-lysine Neonatal (P1-P3) shiverer mice (n = 20) were cryo-anesthetized and cells were stereotaxically injected into the lateral ventricle. In vivo MR imaging was performed at 1, 4, 7 and 14 days after cell transplantation using a Bruker 11.7T Avance spectrometer. To validate MR data and confirm the distribution of transplanted cells, the brains were processed for b-galactosidase, dextran immunohistochemistry, and Prussian blue staining for iron. New myelination was assessed with antibody against myelin basic protein (MBP). MR imaging at early time points (1–4 days) revealed iron-induced MR signal loss throughout the entire ventricular system reaching the central canal of the spinal cord. These findings were in good agreement with conventional histological detection. At later time points (beyond 1 week) the migration of transplanted cells was extensive, covering the majority of the brain; however, this pattern only correlated partly with the MRI findings. This discordance was found to a result from multiple cell divisions, significantly diluting the contrast agent. The shaking phenotype of the transplanted shiverer mice, apparent at P12, was not altered throughout the course of 60 days as compared to non-transplanted controls, in agreement with the histological absence of oligodendrocytic differentiation and new myelination formation. We conclude that MR imaging is useful for early non-invasive monitoring of transplanted stem cells, but should be considered unreliable for rapidly proliferating (stem) cells at later time points.
Abstract ID: 444
We attempted to develop an MRI-based method to track the migration of bone marrow (BM)-derived progenitor cells to the site of injured arteries. The study was divided into two phases: (1) histologically confirming the migration of BM progenitor cells to the injured arteries; and (2) monitoring the migration of BM progenitor cells using MRI. In the first phase, BM cells from donor LacZ or GFP mice were transplanted into six nude mice. The left femoral arteries of recipient mice were injured using a cuff-constriction approach, while the right femoral arteries were uninjured to serve as controls. In the second phase, BM cells from donor LacZ mice were labeled with the superparamagnetic MR contrast agent Feridex, and then transplanted into eight nude mice with left femoral artery injury. In-vivo detection of Feridex-BM cell migration to the injured arteries was studied using a 4.7T MR scanner (FSE, TR/TE = 2s/13.4ms, 1mm slice thickness, and +0.08 mm2resolution). Subsequently, the mice were perfused with 4% paraformaldehyde, and ex-vivo high resolution MR imaging of the bilateral femoral arteries was performed using a 9.4T or 11.7T scanner (3D FSE, TR/TE = 1500/30ms, 512 × 256 × 256 matrix, approximately 0.08 mm3 resolution). Bilateral arterial segments were then harvested for histological validation of LacZ+or GFP+cells at the injury sites. In the first group, histochemistry detected LacZ-positive cells and fluorescence microscopy revealed an increase in green signal intensity at the injured arteries compared to the control arteries (5/6). In the second group, in-vivo and ex-vivo MR imaging demonstrated regions of hypointensity at the sites of the injured arteries (7/8), while no such signal reduction was visualized in the control arteries. This study provides initial evidence that MRI can potentially be used to track the migration of magnetically-labeled BM progenitor cells to the site of injured arteries.
Abstract ID: 445
Abstract ID: 446
Abstract ID: 447
We determined the lower detection limits for 111In-tropolone labeled bone marrow derived stem cells from the maximum activity per cell which did not affect viability, proliferation and differentiation, and the Minimum Detectable Activity (MDA) of 111In by SPECT. Canine bone marrow mesenchymal cells were isolated, cultured and expanded. A number of samples, each containing 5 × 106cells, were labeled with 111In-tropolone from 0.1 to 18 MBq, and cell viability, proliferation and differentiation were assessed afterwards for each sample. To determine the MDA, the anthropomorphic torso phantom (DataSpectrum Corporation, Hillsborough, NC) was used. A point source of 202 kBq 111In was placed on the surface of the heart compartment, and the phantom and all compartments were then filled with water. Three 111In SPECT scans (16, 32 and 64 minutes; 128 × 128 matrix with 128 projections over 360°) were acquired every three days until the 111In radioactivity decayed to undetectable quantities. 111In SPECT images were reconstructed using OSEM with and without background, scatter or attenuation corrections. Contrast to Noise Ratio (CNR) in the reconstructed image was calculated, and MDA was set equal to the 111In activity corresponding to a CNR of 4. The cells had 100% viability when incubated with no more than 0.9 MBq of 111In (80% labeling efficiency), which corresponded to 0.14 Bq per cell. Background correction improved the detection limits for 111In-tropolone labeled cells. The MDA for 16, 32 and 64 minute scans with background correction were observed to be 1.4 kBq, 700 Bq and 400 Bq, which implies that, in the case where the location of the transplantation is known and fixed, as low as 10,000, 5,000 and 2,900 cells respectively can be detected.
The detection limits for 111In-tropolone labeled bone marrow derived stem cells in 16, 32 and 64 minute scans with and without background, scatter or attenuation corrections.
Abstract ID: 448
Adoptive immunotherapy of tumors using dendritic cells (DCs) has gathered much interest due to its capability to induce therapeutic antitumor immunity. In this technique autologous DCs are injected into a patient, which traffic into the tumor and trigger an immune response. Despite some successful experimental clinical applications, the trafficking of DCs to tumors is poorly understood. We have initiated a project to enhance the ability to monitor the trafficking of autologous DCs into tumors using magnetic resonance imaging (MRI).
Using a modification of the method developed by Crich et al.i, we chose a clinically approved formulation of a contrast agent (CA): Gd(III)-HP-DO3A in the form of Prohance™, to label DCs. The cellular uptake of the CA was dependent on the extracellular concentration of the CA and resulted in a significant uptake of label by DCs at the higher CA concentration. Gelatin suspended DCs labeled with CA showed significant differences in the signal intensity in both T1 and T2 weighted images (Figure 1) relative to unlabeled DCs. The use of the clinically approved version of Gd(III)-HP-DO3A in this technique and its uptake in detectable levels by DCs, signifies the potential for its rapid translation into a clinically application means of monitoring DC trafficking into tumors.
Abstract ID: 449
Optical imaging using near-infrared fluorescent contrast media promises to be one of the most successful developments in molecular imaging techniques. Peptide dye conjugates combine easy of synthesis with high affinity and specificity of the conjugate. In contrast to antibody conjugates, they are much smaller, have a higher diffusion in tissues and can be easily modified. We have made use of several peptide dye conjugates with emission in the visible and near infrared spectrum to visualize binding, internalization and trafficking of somatostatin and other neuropeptide hormones in neuroendocrine tumor cells and sections of human neuroendocrine tumors.
Using the 633 nm emission line available in many confocal laser scanning microscopy systems we have been able to excite fluorescence of near-infrared carbocyanine dyes and visualize binding and internalization of octreotate in neuroendocrine tumor cells in vivo and after fixation. This endocytosis was compared with endocytosis of the receptor labelled with GFP. Similarly, we were able to monitor binding of these conjugates to fresh frozen sections of tumor tissue and to compare this binding pattern with binding of radiolabelled conjugates in autoradiography and receptor staining. Although most conventional laser scanning setups are not equipped with lasers in the near infrared spectrum, due to their excitability at 633 nm near-infrared dyes and conjugates thereof can easily be visualized, even in living cells. One of the great challenges will be to enhance sensitivity to allow for ex vivo tissue monitoring of dye conjugates after systemic application.
Abstract ID: 450
Exocytosis is important to protein transport, neurotransmission and hormone secretion. Because there are hundreds to thousands secretory vesicles in each single cell, it needs lots of labor-loading to quantitate dynamic properties of secretory vesicles. We therefore develop software to measure dynamic properties of secretory vesicle using dual-threshold method. The object of the next time frame, which has shortest distance to the interested object of the current time frame, is chosen as the new position of the interested object. Then the traces, intensity and areas of tracked objects were measured. Three secretory vesicles markers (EGFP-Rab3A, synaptopHluorin and NPY-EGFP) were used as fluorescence probes to visualize exocytosis of living PC12 cells under either total internal reflection fluorescence microscopy (TIRFM) or epi-fluorescence microscopy. Since EGFP-Rab3A dissociated from secretory vesicles during exocytosis, the fluorescence dispersed after stimulation. Fluorescence of NPY-EGFP will increase temporally and decrease. Time-lapsed videos of cells expressing these four fluorescent markers individually are used as source to test performance of software. Software can track all distinguishable peroxisomes correctly, and show their intensity and sizes were constant during stimulation. From the results of measurements of either EGFP-Rab3A or NPY-EGFP vesicles with static fluorescence during exocytosis are highly accurate, about 146/163 (17 missed) and 223/223 (60 missed). When software measures EGFP-Rab3A or NPY-EGFP vesicles with various fluorescence intensities, accuracy of analysis is lower (EGFP-Rab3A: 12/17; NPY-EGFP: 15/20). The area measurements are only correct when tracked objects have high fluorescence intensity. When starting intensity of the tracked object is above 80 grayscale, system can track accurately even decrease of intensity is over 20 grayscale/frame. When S/N is below 1.2 or intensity is below 25 grayscale, vesicles cannot be tracked by software. In future, we will base on the results of stimulations to improve accuracy of our tracking system.
Abstract ID: 451
The non-invasive detection of macrophage infiltration into inflammatory sites using MRI techniques has been demonstrated in several models of diseases, by labeling macrophages with iron-containing nanoparticles of mean size between 30 and 150 nm, administered intravenously. These nanoparticles are not immediately trapped by the mononuclear phagocytic system and may thus be internalized by macrophages through absorptive endocytosis. Iron-labeled macrophages are then tracked by MRI by monitoring contrast changes (usually signal attenuations) induced in inflammatory regions infiltrated by the cells. This approach has been applied here to detect macrophage infiltration into the lung following allergen (ovalbumin, OVA) or endotoxin (lipopolysaccharide, LPS) instillation. The difficulties in this application are on the one hand due to the faint parenchymal signal detected by MRI, and on the other related to the different macrophage populations present in the lung even before any inflammatory stimulus is applied. The intensity of the edematous signals detected 24 h after intra-tracheal (i.t.) administration of OVA or LPS were reduced depending on the dose of the contrast agent. The signal reductions suggested that macrophage labeling by USPIO was more effective than that achieved by SPIO particles. Histology confirmed this fact and indicated that all iron administered i.v. was entrapped by macrophages. However, some macrophage-entrapped iron was also found in saline-challenged animals, a result that is consistent with the migration of monocytes in pulmonary capillaries into alveolar spaces even in the absence of lung inflammation. Thus, in order to maximize the labeling of inflammatory macrophages in the lung, it is important to well choose the time point of administration of the contrast agent. The basic challenge consists in appropriately dosing the contrast agent, as well as in finding the appropriate time point for its administration, for differentiating the contributions from blood borne and tissue own macrophages.
Abstract ID: 452
Inflammation in multiple sclerosis (MS) and experimental autoimmune encephalomyelitis (EAE) is driven by T-cells directed against myelin specific protein. So far, in-vivo imaging focused on downstream processes such as the infiltration of macrophages or structural alterations linked to inflammation and tissue damage. In this work the infiltration of Cy5.5-labeled T-cells into the brains of EAE rats was studied by NIRF-imaging. For this purpose T-cells, which were extracted from the spleens and thymus of naïve or immunized rats, respectively were cultured in presence of Cy5.5 (200μg/mL) and administered i.v. to Lewis rats. Three groups were studied: EAE or naïve rats received activated T-cells 7 days after immunization. In addition, another group of EAE rats received naive T -cells at 7 dpi. Measurements were carried out from day 8 to 22 on a daily basis using a fluorescence tomography system (eXplore Optix, GE Medical System). The highest fluorescence signal was observed for EAE animals, which received myelin-activated T-cells during the acute phase of EAE. The temporal profile of T-cell accumulation in this group resembled the pattern of neurological impairment consisting of acute phase and first relapse. No disease specific fluorescence pattern was observed for EAE animals, which received naïve T-cells. Inoculation of myelin-activated T-cells in naïve animals led to a peak of fluorescence signal on day 13, which was less pronounced as in the first group. These results demonstrate that TAT-Cy5.5 can be used for labeling of T-cells, which can then be tracked in-vivo by repetitive NIRF imaging. This approach may particularly be useful for evaluation of novel anti-inflammatory therapies.
Abstract ID: 453
Owing to its superb spatial resolution (<100 μm) MRI is the technique of choose for cellular tracking. Currently most of the work in cellular labelling is carried out mainly with iron oxide particles but its use may have some drawback. An alternative approach is based on the use of paramagnetic metal complexes (T1 agents) In the latter class a particularly safe agent is represented by GdHPDO3A. It is an highly stable neutral, well soluble and tolerated agent currently used in the clinical practice. The incubation (16–24h) of cells in media added of GdHPDO3A (at 25–50 mM concentration) leads to the entrapment of high amounts of the agents into endosomic vescicles by the pinocytotic route. Alternatively, internalization into cell cytoplasm may be pursued by electroporation (an electric pulse is applied to a cell suspension containing the Gd complex in order to induce the membrane permeation). The obtained results show that the relaxation enhancing ability of GdHPDO3A internalized into cells is dependent on the peculiar internalization route. When the probe is confined into endosomic vescicles, its efficiency can be markedly quenched by the slow exchange of water molecules between endosome and cytosol compartments. Conversely, upon entrapment in the cytosolic compartment, the presence of only one membrane to be crossed makes the internalized agent more efficient for the MRI visualization. Labeling with GdHPDO3A has been carried out on different tumor cells (HTC, C6, B16), on stem cells (Endotelial Progenitor Cells), mouse and human islets. The intracellular localization of the contrast agent can be assessed by confocal microscopy using Eu-HPDO3A, a fluorescent analogue of GdHPDO3A.
Abstract ID: 454
Abstract ID: 455
Our aim was to monitor proliferative cells of the SVZ by MRI after labeling with iron oxide nanoparticles, and to follow their migration towards the olfactory bulb (OB).
Abstract ID: 456
The capacity to track specific cells and cell populations non-invasively over time in animals and humans will have a tremendous impact on our ability to study, diagnose and treat disease. Over the past few years, MRI has emerged as a leading technology for tracking cells. One of the major criticisms levelled at MRI, however, has been that it is intrinsically a low sensitivity imaging modality in the context of detecting labelled cells. In the current work we show for the first time, that single cells can be detected in vivo, in mouse brain using MRI. Macrophages were labelled with superparamagnetic iron oxide (SPIO) nanoparticles (Feridex, 58 pgFe/cell) and the fluorescent dye DiI and injected into the circulation of mice. Individual cells trapped within the microcirculation of the brain could be visualized in vivo as discrete areas of signal loss using MRI (Figure 1). Direct optical validation of single cell detection was provided by co-registering signal voids on MRI with low and high magnification confocal microscopy (Figure 2). This work demonstrates that MRI has the sensitivity for detecting single cells in vivo at SPIO loading levels that can be achieved in a number of cell types such as stem cells and cancer cells and opens up the possibility of non-invasively tracking these cells in animal models of disease at the single cell level.
In vivo (A) and ex vivo (B) MRI of mouse brain. In vivo MRI was performed at 100 × 100 × 200 μm3 resolution and shows three signal voids (white arrows) corresponding to SPIO labeled cells. Ex vivo MRI was performed on the same brain at 100 μm isotropic resolution. Two signal voids are visualized because of the smaller slice thickness. The third signal void is located on an adjacent MR slice. (cc. corpus callosum, ac anterior commissure, CPu caudate putamen).
Correlation of signal voids with confocal micrscopy of single cells. (A) Magnified view of MRI (Figure 1B) showing two signal voids (black arrows) (B) Corresponding low magnification confocal image (overlay of differential interference contrast (DIC) and DiI fluorescence image) from a histological section corresponding to MRI in (A) (bar=200 μm). There is good correlation between the signal void in (A) and the area of DiI fluorescence in (B) (white arrow), as well as anatomical landmarks between the MR and optical image. (C) Magnification of the area of DiI fluorescence indicated by the white box in (B). Optical z-sectioning corresponding to the DiI fluorescence from one of the cells in (C) showing a solitary nucleus (n) and confirming that the signal void in (A) is the result of a single cell (bar= 10 μm).
Abstract ID: 457
There is increasing interest in using superparamagnetic iron oxide nanoparticles to label and track stem cells in vivo by MRI. However, ex vivo cultivation strategies for transduction and loading of nano-particles into haematopotic stem cells (HSC) have not been well defined. Using two FDA approved agents, Feridex and protamine sulfate, we present data on the in vitro uptake of iron nanoparticles in defined human stem cell subsets. Protamine sulfate complexed Feridex nanoparticles conjugated to Alexa 647 dye (FE-PRO-647) were loaded into cells and monitored at different time points. Most cell lines reported in the literature show uniform 60–99% uptake of nanoparticles within the first 24 h. In contrast, uptake in primary human cord blood linage depleted (lin-) cells shows that the overall particles load declines from 61% after 24 h to 40% after 72 h in culture. Based on this data a number of stem cell subsets were studied to determine their ability to uptake and retain FE-PRO-647. One subset, CD34+, showed 17% uptake, with only 15% of these cells expressing p27kip1, indicating that the majority of these cells were actively cycling ex vivo. Subsequent transplantation of cord blood CD34+ into beta2-NOD/SCID mice for in vivo tracking by flow cytometry showed that CD34+ cells labeled with FE-PRO-647 reside in the bone marrow of the recipients three weeks post transplantation. Correlating the ratio of the mean fluorescent activity of the initial FE-PRO-647 CD34+ cells in cord blood to the FE-PRO-647+ CD34+ cells found in the bone marrow 3 weeks post transplantation, indicates that these CD34+ cells divided a total of 5 times over the 3 week period. Overall, these data show that iron oxide nanoparticles can be used to transiently transduce repopulating human HSC subsets and may provide a clinically feasible technique for MRI imaging of stem cells transplanted in vivo.
Abstract ID: 458
Abstract ID: 459
Superparamagnetic iron oxide(SPIO) is widely used as T2 contrast agent for hepatic imaging and cell trafficking. The reticuloendothelial system (RES) accounts for most of the SPIO uptake in vivo. Recent study revealed the uptake of the SPIO doesn't affect the viability of labeled cells. However, the phagocytic activity of the RES, after SPIO uptake has not been evaluated.
Mouse macrophage cell line (Raw 264.7) was chosen in this in vitro study because its phagocytic activity. Ferucarbotran (Resovist) at the concentration of 100ug Fe/ml, 10ug Fe/ml (plasma concentration), and 1ug Fe/ml were incubated with Raw 264.7 for 12 hours. The SPIO uptake was proved by Prussian blue staining. The maximum iron uptake was found at the concentration of 100ug Fe/ml which was determined by magnetic resonance imaging. The phagocytic activity of the macrophage cell line was determined by the uptake of red fluorescent particles. More than one particle uptake was counted as positive cell and the total cell number was counted under fluorescent microscope. The ratio of positive cells at different SPIO concentration was calculated. Decreased phagocytic activity of the Raw264.7 was found when it is co-cultured with SPIO at the concentration of 100ug Fe/ml. It is concluded that SPIO uptake will affect phagocytic activity when there is strong SPIO uptake. The significance of the influence on macrophage phagocytic activity should be further investigated in vivo.
Abstract ID: 460
Non invasive imaging of cell trafficking would be an important tool for pre-clinical studies to get a better understanding of diseases like carcinogenesis or autoimmune diseases. While fluorescence markers have a limited penetration depth, radiolabelling of cells allows non invasive imaging by positron emission tomography (PET) or single photon emission tomography (SPECT) without any restrictions in resolving deeper organs. Our work focuses on tracking specific interferon- γ (IFN-γ) producing CD4+ (Th1) cells. Th1 cells mediate delayed type hypersensitivity reactions such as rheumatoid arthritis, psoriasis, or experimental allergic encephalitis. Hapten and ovalbumin (OVA) specific Th1 cells were labelled in vitro with [Cu64]PTSM. The labelling process was done by incubating 1 × 106 Th1 cells in 250–1000μL medium with 20–60 μCi [Cu64]PTSM for a total of three hours. Extensive in vitro studies showed a viability of 70–80% up to 48 hours after labelling. Specific T cell responses were analyzed by ELISA and proliferation assay and compared to non labelled T cells. The achieved labelling efficiency was around 25%. Ear challenge with OVA peptide or hapten induced specific Th1 cell trafficking and homing into ear lymph nodes. Trafficking was dynamically and quantitatively imaged by microPET. The findings were confirmed by autoradiography and biodistribution studies with a gamma counter. Exact morphological information of Th1 homing sites was achieved by H&E staining of the autoradiographic sections. Specific homing of transgenic OVA Th1 cells into draining ear lymph nodes was confirmed by FACS analysis.
Although Cu-64 with a half life of 12 hours can only be scanned by PET for up to three days, it provides an important tool for non invasive cell trafficking, especially in early time points after cell transfer to gain basic biological information.
Abstract ID: 461
The conventional method for the analysis of stem cell transplantation depends on postmortem histology. Here, we have sought to demonstrate the feasibility of a longitudinal monitoring of transplanted cell survival in living animals, by employing optical imaging techniques. Mouse embryonic stem cells (ESC) were obtained from American Type Culture Collection (ES-E14TG2a). Mouse ES cells were cultured in the DMEM (Gibco-BRL, Gaithersburg, MD) supplemented with 3.7 g/L sodium bicarbonate, 1% penicillin and streptomycin, 1.7 mM L-glutamine, 0.1mM β-mercaptoethanol, 5 ng/mL mouse leukemia inhibitory factor (LIF), and 15% fetal bovine serum (FBS) with or without a feeder layer and cultured for five days in standard medium plus LIF. ESCs were then transfected (MOI=100) overnight with Ad-CMV-Fluc. Our experimental Sprague-Dawley rats (n=7) were given with different numbers of ESCs (105, 106, 5 × 106) expressing Fluc into corpus cavernosum. Sham-operated rats were controls (n=6). Cell survival was assessed histologically and/or by optical bioluminescence imaging which was conducted using a cooled charged-coupled device camera (Xenogen), beginning on the day after the transplantation. In cell cultures, firefly luciferase activity correlated linearly with cell numbers from 105 to 5 × 106 (r2=0.95). In living animal imaging, imaging signal activity correlated linearly with cell numbers injected from 105 to 5 × 106 at each time point (r2=0.62 ~0.98). In all three groups of rats, imaging signal was detected in rat genital area from the 2nd day to the 47th day after cellular injection. Adenovirus mediated transient expression of firefly luciferase reporter gene in ESCs was feasible to monitor cell survival over a month after transplantation. The locations, magnitude, and survival duration of the ESCs were non-invasively monitored with a bioluminescence optical imaging system.
Abstract ID: 462
Abstract ID: 463
In many attempts to use stem cell for the tissue engineering, quantification of the engraftment of stem cells on scaffold is mandatory in real time. In this study, we measured luciferase activity in engrafted neural stem cells by luciferase activity.
F3 human neural stem cells were isolated from human telencephalon (14–16 weeks), immortalized with v-myc by Dr. SU Kim. 0.75 MOI (multiplicity of infection) of luciferase-expressing adenovirus (Ad-Luci) was infected into F3 cells. 24 hours later, Ad-Luci infected F3 cells were seeded on chitosan scaffold (9 × 2 × 2 mm) and were cultured for 48 hours. Luciferase activity in Ad-Luci infected F3 on chitosan scaffold was measured using the caged luciferin (Molecular Probe Co.).
When we measured the luciferase activity in Ad-Luci infected F3 cells on chitosan scaffold with Luminometer, the luciferase activity was increased according to Ad-Luci infected F3 cell numbers. And we measured the luciferase activity in Ad-Luci infected F3 cells on chitosan scaffold with CCD camera, the luciferase activity was also increased according to Ad-Luci infected F3 cell numbers.
The neural stem cells seeded on chitosan scaffold was successfully monitored by luciferase activity in vitro. Further studies are in progress on in vivo imaging of mice transplanted with luciferase-expressing neural stem cells on chitosan scaffold.
Abstract ID: 464
The goal of the current study was to follow the migration activated dendritic cells (DC) in an in vivo mouse model. The DCs were labelled by the MR contrast agent Endorem. After injection into the food pad the labelled DCs migrate from the paw into the popliteal lymph node via the lymph vessels. Using a BRUKER BioSpec 47/40 and a dedicated mouse resonator for transmit and receive we imaged the popliteal lymph nodes of the DC injected and control side using a T2 weighted MSME sequence (Te: 16 echoes, 10 ms to 160 ms, TR: 1500 ms, 256 × 256 resolution, 16 slices, 0.35 mm slice thickness, in vivo 8 averages, in situ 126 averages). One day after injection a MR signal loss could be detected in the lymph node corresponding to the side where labelled DCs were injected. This signal loss could be detected injecting 2 × 1*107 down to 1*105 labelled cells. Injection of Endorem in PBS into the food pad also led to a signal loss in the lymph node but at a spatially different substructure of the lympnode. Whereas the labelled DCs accumulate inside the T-cell area of the lymph node Endorem alone accumulates in the surrounding cortical region of the lymph node and is not able to enter the T-cell area. This is so because the transport into the T-cell area is an active, receptor mediated (CCR7) mechanism. With lower SNR the same imaging procedure could be performed in living mices making this approach a useful tool to study migration of iron labelled DCs in vivo.
Poster Session: Molecular and Functional Imaging in Cancer, Part 4
Abstract ID: 465
Abstract ID: 466
The new developments in systemic radionuclide therapy based on chelated peptides now strongly call for metallic radioisotope preparations with new characteristics.
The presently used methods in production of the corresponding isotopes have reached their technical limitations and the progress in the systemic radionuclide therapy is limited by the availability of radio nuclides with high quality characteristics. These radionuclides are either only or best produced in high-energy spallation and fragmentation reactions. An opportunity for industrial scale production of such new radioisotopes for future medical use in a fully parasitic or prime user mode is described.
The target and ion-source techniques developed for making the high purity mass-separated radioactive ion-beams (RIB) at CERN ISOLDE is now a well-documented new type of rapid, efficient, continuous and automatic radiochemical separations. Its key element, the electromagnetic Isotope Separator On Line (ISOL), allows producing efficiently very pure samples of almost all radioactive isotopes with the highest possible specific activity, i.e. the carrier free form. At present they permit producing “exotic” radioisotopes that hitherto were not available on the market for medical use but only in amounts that has supplied the R&D work and validation of the methods.
We believe that the time has come to prepare for more rational and large-scale industrial radioisotope production methods using the ISOL target techniques in particular in conjunction with electromagnetic mass separation.
The future production sites should be put in synergy with either the planned upgrade of the CERN ISOLDE and/or one of the major new physics-research facilities planned for using GeV proton beams of MW power like EURISOL, neutrino factories or spallation neutron sources where the production rates will be orders of magnitude higher than at present facilities.
Abstract ID: 467
(i) anion-exchange resin AG 1 × 8, 200–400 mesh, in 0.01 N HNO3 / 80% ethanol systems, containing different amounts of nitrate salts; (ii) cation-exchange resin Dowex 50 × 8, 200–400 mesh, in high concentrations HClO4/water solutions; (iii) cation-exchange resin Dowex 50 × 8, 200–400 mesh in α-hydroxy-isobutyrate (α-HIBA) solutions, pH = 4.75.
Abstract ID: 469
Gefitinib is a selective epidermal growth factor receptor tyrosine kinase inhibitor that is currently evaluated alone or in combination with other drugs in numerous clinical trials for breast, colorectal, neck, ovarian and prostate cancer treatment. Our aim was to evaluate the effect of gefitinib treatment on 99mTc-MIBI uptake in Bcl-2 over-expressing breast cancer cells. MCF-7 and T47D breast cancer cells stably transfected with cDNA of human Bcl-2 gene and mock transfected cells were incubated with increasing concentration of gefitinib ranging between 0.5 and 20 μM for different time intervals. 99mTc-MIBI uptake was assessed in treated and untreated Bcl-2 overexpressing and control cells by 60 min incubation. A dramatic increase of 99mTc-MIBI uptake was observed in Bcl-2 overexpressing cells after treatment with gefitinib whereas mock transfected cells did not show any significant changes of 99mTc-MIBI uptake in response to treatment. This effect on 99mTc-MIBI uptake was both time and dose-dependent. To elucidate the molecular mechanism underlying the changes of 99mTc-MIBI uptake in Bcl-2 overexpressing cells upon exposure to gefitinib, we tested the involvement of Bcl-2 phosphorylation and Bax translocation to mitochondria as well as the effect of Bad transient cotransfection. Neither Bax nor Bad appeared to be involved in the mechanism underlying changes of tracer uptake whereas a reduction of Bcl-2 phosphorylation was observed in Bcl-2 overexpressing clones upon exposure to gefitinib. In conclusion, these findings indicate that an increase of 99mTc-MIBI uptake in malignant tumor lesions of patients exposed to gefitinib treatment may reveal overexpression of Bcl-2 anti-apoptotic protein thus predicting Bcl-2 mediated resistance to therapy.
This work was supported by EU grant EMIL contract N. 503569.
Abstract ID: 470
Aberrant HGF/SF-Met signaling is important in the pathogenesis of many types of human tumors and is often associated with poor clinical outcome. To follow primary events of tumorigenesis and metastasis, we generated transgenic mice expressing functionally active green fluorescent protein (GFP)-Met chimeric receptor. In parallel we developed high resolution confocal molecular imaging (HRMI) methods to allow intravital analysis at single cell resolution. HRMI analysis of GFP-Met mice show enhanced fluorescence in cells of sebaceous glands. Within several months, male mice spontaneously develop sebaceous gland tumors that are metastatic to the lung, kidney, and liver. Using intravital HRMI, we observe proliferating single cells in sebaceous glands preceding tumor formation. These cells show high GFP-Met levels relative to adjacent normal tissue. We also observe single cells spreading from the tumor mass. These cells express high GFP-Met levels of and appear to be single cell precursors of metastases.
We have developed a novel, real time imaging modality, HRMI, for intravital imaging at single cell resolution. Using HRMI we can follow and quantitatively evaluate functional activity and localization of a member of the tyrosine kinase growth factor receptor family, Met, in pre-cancerous lesions and possibly pre-metastatic cells, in live animals. This technology should be valuable for pre-clinical evaluation of therapeutic strategies targeting oncogenes as the Met receptor.
This work was supported in part by a research grant from the NIH (P50CA93990) and by the Michigan Economic Development Corporation and the Michigan Technology Tri-Corridor (MAMC 085P1000815).
Abstract ID: 471
Abstract ID: 472
A new prototype of drug delivery system, the nanoconjugate Polycefin, was tested for its ability to show tumor accumulation due to enhanced permeability and retention (EPR). Polycefin was synthesized for targeted delivery of Morpholino oligonucleotides and antibodies into certain tumors, and consists of modules active in endosomal uptake, disruption of endosomal membranes, oligonucleotide release in the cytoplasm, and protection against enzymatic degradation in the vascular system. These are covalently conjugated with polymalic acid (Mw 50000, Mw/Mn 1.3) highly purified from cultures of myxomycete Physarum polycephalum. The polymer is biodegradable, non-immunogenic and non-toxic. Alexa Fluor 680 C2-maleimide was coupled to the polymer and the tissue-targeting antibody was omitted to study the EPR-effect. MDA-MB 468 human breast cancer cells were injected subcutaneously into the left posterior mid-dorsum of nude mice [Tac:Cr:(MCr)-Foxn nu ], and two months later, 100 microliter of 2.5 mg/kg of the drug were intraveneously injected via the tail vein. In vivo fluorescence tumor imaging was performed at 0 min and 2–180 min after the injection. The macro-illumination and detection system consisted of a Xenon300 Storz fiber-optically coupled light source and a cooled, scientific-grade monochrome digital camera (Hamamatsu ORCA-ER 1024/1024 pixels) coupled to a Computer TV zoom lens. Wavelengths were 660 nm / 20 nm band width for excitation and 700 nm / 30 nm bandwidth. The distance lens - specimen was 12 cm. The fluorescence signal in large blood vessels and in the tumor increased rapidly until 60 min and remained at a level 6 times higher in the tumor, than in sourrounding tissue (180 min) (p < 0.003). The results of fluorescence imaging were corroborated by fluorescence microscopy, examining 7 micrometer thick sections of various organs, only the tumor, kidney and liver giving fluorescent signals.
Abstract ID: 473
The first use of fluorescent proteins in animals was by Chishima et al., to visualize cancer cells in fresh tissue. Fluorescent proteins can be used to image essentially any type of cancer process including primary tumor growth, tumor cell motility, tumor cell invasion, intravasation, extravasation, metastatic seeding, metastatic colonization, angiogenesis, cell deformation, drug sensitivity, effects of molecular alterations, and tumor-host interaction. The first use of green fluorescent protein (GFP) for whole-body imaging was by Yang et al., who visualized primary and metastatic tumor growth in real time. The intrinsic brightness of fluorescent proteins enables cells to be visualized inside intact animals by whole-body imaging. GFP-expressing primary and metastatic tumor growth could be visualized by whole-body imaging in numerous organs. Many colors of fluorescent proteins are now known and can be used to color code cancer cells of a specific genotype or phenotype. Upon co-implantation in a single animal, the interaction between the variants can be visualized. For example, a highly metastatic variant can be labeled with GFP and its low metastatic variant can be labeled with red fluorescent protein (RFP). The host and the tumor can be differentially labeled with fluorescent proteins. The host mouse can be a transgenic organism expressing GFP in all of its cells or specifically in cells of interest such as endothelial cells. The GFP-expressing mice can be transplanted with cells expressing RFP. Interaction of the tumor cells with the host cells can then be imaged. Cells themselves can be double-labeled with GFP in the nucleus and RFP in the cytoplasm. With such double-labeled cells, nuclear-cytoplasmic dynamics can be imaged. In addition to cancer, many other processes can be visualized in vivo with fluorescent proteins including gene expression, infection, and behavior of stem cells.
Abstract ID: 474
This study was funded by NIH EB002168, HL042950, and CO-27031. The research was carried out at the Washington University School of Medicine.
Abstract ID: 475
With F-18 labeled Fluoroazomycin arabinoside (FAZA) a specific tracer for hypoxia has been developed. Longitudinal measurements enable the monitoring of tumor hypoxia as a function of treatment with hypoxia-directed chemotherapy, radiation or combined radio-chemotherapy.
14mg/kg of hypoxia-directed tirapazamine i.p. plus sham radiation (CHT),
4.5Gy plus NaCl 0.9% i.p. (RT),
tirapazamine combined with radiation (RCT) or
0.5ml NaCl 0.9% i.p. plus sham radiation (controls).
Two PET measurements within 3d were performed on the controls to determine the reproducibility of the derived oxygenation tension.
The reproducibility measurements yielded a Pearson correlation coefficient of 0.79.
The results highly suggest that tirapazamine has a hypoxygenic effect. Furthermore, oxygenation tension stays stable under radiation in hypoxic tumors but is affected in normoxic tumors.
Abstract ID: 476
PET scanning is increasingly been used to stage various primary malignancies. The purpose of this study was to compare established size criterion, used to diagnose metastatic nodal disease, to FDG uptake, within nodes in patients undergoing combined PET/CT with 18F-Fluorodeoxyglucose for staging of various primary malignancies.
32 patients (M: F; 10:22) with a mean age of 55 years (range; 20–82 yrs) undergoing combined PETCT studies were included in the study. Pathological examination of the detected nodes where performed after surgical dissection or image guided biopsy. Of the sampled nodes, 104 could be unequivocally correlated to pathology and were included in the study. Primary efficacy parameters of sensitivity, specificity and accuracy were calculated for size criteria and FDG uptake. A McNemar Test was performed to evaluate these criterions.
54 of the 104 included nodes were benign and 50 nodes were malignant on histological examination. The mean short axis diameter of malignant nodes was 10.5mm (5–41mm). The mean short axis diameter of malignant nodes showing increased FDG uptake was 14.1 mm (6–41mm). Size criterion demonstrated a sensitivity of 64.7%, a specificity of 83% and accuracy of 74%. Quantitative evaluation of FDG uptake within the same nodes demonstrated a sensitivity of 76%, specificity of 61% and accuracy of 68%. All false negative nodes on PET were less than 10mm in short axis. The McNemar test showed statistical significance (p = 0.014).
PET imaging demonstrates a lower specificity for detecting metastatic nodal disease than size criteria on CT, probably as multiple non-malignant conditions can also cause an increase in nodal FDG uptake. However it has a higher sensitivity for nodal metastasis detection compared to size criteria, which has a minimally higher accuracy. This can be attributed to the decreased resolution of PET imaging compromising characterization of nodes less than 10mm in size.
Abstract ID: 477
Abstract ID: 478
Abstract ID: 479
The more heat resistant cells are the longer it takes until HSP70 gets expressed at increasing temperatures.
Abstract ID: 480
Abstract ID: 481
Abstract ID: 482
Abstract ID: 483
Abstract ID: 484
Abstract ID: 485
Multidrug resistance (MDR) desensitizes the tumor to a variety of structurally and functionally different chemotherapeutic agents and is linked to the expression of ATP-dependent drug efflux transporters. In vivo imaging of MDR was demonstrated with radioactive probes using 99mTc-sestamibi, an MDR substrate, for SPECT and 11C-labeled drug analogs for PET imaging. Major limitations of nuclear imaging are low spatial resolution of SPECT and short lifetime of PET probes. We developed a novel specific imaging agent that enables MR imaging of MDR. The imaging agent consists of three functional domains as shown in the figure.
The first domain, amphiphilic HIV1 Tat membrane translocation peptide, facilitates internalization of the agent; the second domain, GdDOTA complex, generates T1 MR contrast, and the third domain, rhodamine, an MDR substrate, provides specific efflux of the agent from the drug resistant cell and also can be used for fluorescence detection. The agent was synthesized using a standard solid-state peptide synthesis. Uptake of the compound was studied in wild type MCF-7wt and drug resistant MCF-7adr breast cancer cells using fluorescent microscopy and FACS. A significantly reduced accumulation of the agent was detected in the drug resistant cells. Treatment of MCF-7adr with 100 uM dimethyl-beta-cyclodextrin (DM-CyD), inhibitor of P-glycoprotein and MRP2, resulted in reversal of the effect.
Initial MRIstudies of isolated cells demonstrated positive signal enhancement in both MCF-7wt and MCF-7adr cells and significantly increased contrast uptake in the wild type cells.
This study was supported by NIH RO1 CA097310 grant.
Abstract ID: 486
Histone deacetylase (HDAC) inhibitors (HDACi) are emerging as a new and exciting class of anti-neoplastic agents. Our goal is to develop a non-invasive magnetic resonance spectroscopy (MRS) method to determine inhibition of HDAC in vivo. We used a fluorinated HDAC substrate, Boc-Lys(Tfa)-OH, and show that it permeates cells and its levels, as observed by MRS, are significantly higher in cells treated with an HDAC inhibitor compared to controls.
In vitro cleavage of Boc-Lys(Tfa)-OH by HDAC was assessed by incubating with recombinant HDAC-8 and using 19F MRS to monitor Boc-Lys(Tfa)-OH consumption and trifloroacetic acid (TFA) formation. This confirmed that Boc-Lys(Tfa)-OH is a substrate of HDAC, and that its metabolism can be monitored by 19F MRS. To test the applicability of this method to tumor cells, we investigated PC3 cells in the presence of 1mM Boc-Lys(Tfa)-OH following treatment with 10μM of the clinically relevant HDAC inhibitor SAHA. 1mM Boc-Lys(Tfa)-OH did not significantly affect cell proliferation or HDAC activity. Treatment with SAHA for 24 h lead to a significant drop in HDAC activity down to 41% as well as inhibition in cell proliferation by 38% relative to control. The 19F MR spectra of PC3 cells demonstrated that Boc-Lys(Tfa)-OH is detectable in cells using 19F MRS. The spectra further indicated that following a 24 h incubation with SAHA a significant increase in Boc-Lys(Tfa)-OH levels can be detected from 8 fmol/cell in controls to 35 fmol/cell in HDACi-treated cells. No TFA could be detected in the cells, probably indicating that the acid is removed from the intracellular compartment, as reported by others. This study indicates that 19F MRS of fluorinated HDAC substrates could be used to non-invasively monitor HDAC activity in cells, and possibly in vivo.
Abstract ID: 487
In-vivo FMT of a mouse with a 7 mm tumor in the upper left mammary tissue. A: Tomographic slice depicting probe biodistribution. B: Tomographic slice of active cathepsin concentration. Slice depth = 0.5 mm.
Abstract ID: 488
The zebrafish has been shown to be a valuable vertebrate model to study cancer development. During maturation, zebrafish become progressively less transparent, thus identification of internal organs in adult zebrafish depends largely on the in vivo detection of fluorescent signals in transgenic fish. However, bright tissue autofluorescence and limited depth penetration of visible light can hamper accurate organ analysis by brightfield or fluorescence microscopy and prevent non-invasive detection of tumors. Furthermore, real-time 3-dimensional imaging has not been available to allow volumetric tumor measurements and longitudinal in vivo studies of tumor development in zebrafish. We have adapted a high-resolution microscopic ultrasound for use in zebrafish that has been successfully utilized to image murine tumors. We are using the VisualSonics Vevo 770 system with 40 and 55 MHz ultrasound probes. This imaging system allows spatial resolution to 30 micrometers in B-mode as well as rapid acquisition of 3-dimensional images. Normal fish ultrasound anatomy was compared with histopathological sections of the fish to provide a framework to compare diseased fish to. We imaged both genetic and carcinogen-induced models of various cancers to detect tumors and describe their invasion, vascular status, and progression over time. We examined in B-mode both zebrafish with obvious tumors as well as fish that ppeared normal, yet had a predisposition for tumor development. We were able to accurately detect and image small (<2mm) hepatic and testicular tumors in fish that showed no external abnormalities. B-mode image cine loops obtained at 40 Hz could also give a general impression of vascularization of tumors, while 3-dimensional imaging enabled volumetric measurements of tumors. Use of ultrasound biomicroscopy allows longitudinal studies of tumor development and screening of fish lines with a high propensity for tumors, and will facilitate the measurement of therapeutic effects in zebrafish cancer models.
Abstract ID: 489
It is well known that the integrin αvβ3 plays a critical role in tumor-induced angiogenesis and metastasis. This receptor is highly expressed on tumor cells and newly formed tumor microvessels. Imaging of integrin expression would be useful to characterize tumor progression and to monitor responses to anti-integrin therapeutics. Our previous studies with Cy7-labeled monomeric RGD peptide probe Cy7-c(RGDyK) showed αvβ3-selective tumor uptake but also fast tumor washout rate, limiting its further application for optical imaging. Herein we report Cy7 conjugated dimeric and tetrameric RGD peptides for near-infrared fluorescent imaging of integrin expression in subcutaneous brain tumor models. We hypothesized that multimeric RGD peptides with more than repeating cyclic RGD units will offer the enhanced the receptor-ligand interactions through polyvalency effect. Receptor binding assay on U87MG cells showed that Cy7-labeled RGD tetramer had higher binding affinity for cell surface αvβ3 integrin receptor than the dimeric and monomeric counterparts, with IC50 values for Cy7-c(RGDyK),Cy7-E[c(RGDfK)]2, and Cy7-E{E[c(RGDfK)]2}2 being 70.1 nM, 32.5 nM, and 23.2 nM, respectively. In vivo fluorescent imaging was performed on mice with s.c. glioma U87MG tumors at a dose of 50 nmol/kg. Whole-body fluorescence images were obtained for 10 s at 0.5, 1, 2, 4, and 24 h postinjection. The tetramer gave the highest tumor-to-background contrast at all time points tested (e.g. 4.75 ± 0.3 for tetramer, 2.75 ± 0.2 for dimer, and 2.5? ± 0.1 for monomer at 2 h p.i.) (Fig. 1). Most of the probes have been cleared from the tumor and the rest of body at 24 h p.i. This study suggests that Cy7-labeled tetrameric RGD peptide is suitable for near-infrared fluorescence imaging of tumor integrin expression in vivo.
Abstract ID: 490
It is of great clinical significance to predict tumor metastatic potential by noninvasive imaging methods. Here we report that MR T1ρ-weighted imaging can differentiate two typical human melanoma xenografted in SCID nude mice, the aggressive C8161 and the indolent A375P. T1ρ relaxation, which is sensitive to macromolecular concentration and macromolecule-water interactions, has been found to be a sensitive index of ischemia and an early marker of tumor response to therapies. We employed a T1ρ-weighted imaging sequence (a spin-locking pulse with B1 ranging from 500~3000Hz followed a FSE acquisition, TR~1–2s, esp 6ms, etl 16, matrix 128 × 128, FOV 2.5–4cm, B0=4.7T) to image a single slice through the mid-portion of subcutaneously implanted tumors (size 6~15 mm). We found that the indolent tumors have significantly higher T1ρ values than the aggressive tumors (Table 1). In contrast, T1 and T2 measurement did not show statistically significant difference between the two lines. Studies on more melanoma cell lines are needed to test if T1ρ can be a new biomarker of melanoma metastatic potential.
We also employed a 3D low temperature fluorescence imager to measure the relative fluorescence intensities of NADH and flavoproteins (Fp) in mitochondria in these xenografts. As indicated by the relative redox ratio Fp/(Fp+NADH) for a typical tissue slice (slice thickness 10μm, in plane resolution 80~100μm), an aggressive C8161 xenograft is at least two times more oxidized, which probably reflects much higher mitochondria respiration, greater hypoxia, higher metastatic potential (due to VEGF) and lower radiosensitivity than an indolent A375P xenograft. Higher metabolism seems to correlate with more macromolecule synthesis implied by the T1ρ data.
Comparison of mean relaxation times T1, T2, T1ρ between two cell lines.
Poster Session: Molecular imaging in the Drug Discovery Process
Abstract ID: 491
Abstract ID: 492
Non-invasive bioluminescence imaging (BLI) of luciferase reporters in small animal models offers an attractive approach to monitor transcriptional/translational regulation of gene activity, protein-protein interactions as well as following tumor progression, cell engraftment, pathogen infection and drug action. Conventional BLI can be repeated within the same animal after bolus re-injections of a bioluminescent substrate, thus, partially resolving the temporal dynamics of these processes. However, intervals between image acquisitions are governed by substrate pharmacokinetics and excretion, therefore restricting temporal resolution of re-injection protocols to the order of hours, limiting analyses of processes with short time constants. Furthermore, bioluminescence signals can be correlated to reporter activity only after normalizing for transduction efficiency and for nonspecific effects on transcription and perfusion.
To eliminate these constrains, we propose to use implanted micro-osmotic pumps for persistent and constant delivery of bioluminescent substrates, enabling multi-hour/multi-day experiments. Such a setup provides a stable baseline for each animal to serve as its own control, allowing rapid, temporally-resolved analysis of molecular and cellular processes. For example, in mice bearing tumors stably expressing ubiquitin-firefly luciferase (Ub-FLuc) or Iκα-firefly luciferase (IκBα-FLuc) fusion reporters, subcutaneously-implanted D-luciferin-loaded pumps were applied to analyze the pharmacodynamics of the proteasome-inhibiting drug Bortezomib. Continuous acquisition of bioluminescence signals before and up to six hours post I.V. drug administration now enabled fully-resolved temporal analysis of drug action in two previously partially-resolved models for imaging total- and substrate-specific proteasomal activity, respectively (Nat Med. (2003), 9:969–73 and Mol Imaging (2004),
We suggest that this innovation should be generalizable to a variety of other imaging strategies in vivo.
Abstract ID: 493
The diagnostic and therapeutic efficiency of drug depends highly on the drug distribution in target tissues (tumor for example) both specifically and accumulatively. We report here a powerful approach in design of tissue targeting leads with the assistant of radio-combinatorial screening technique developed in our laboratory.
Abstract ID: 494
The aim of this work was to design a method for automatic classification of voxels into pharmaco-organs based on multi-compartmental modeling of whole body tracer kinetics and on time activity curves (TACs) analysis.
According to multi-compartments models, tracer concentration may be separated into blood and tissue kinetics, homogeneous inside organs and weighted by the varying vascular fraction. Thus, physiological kinetics inside an organ belongs to a line (the organ kinetics model) in the kinetics' space.
Segmentation was performed in 3 steps : 1) computation in each voxel neighborhood of a criterion testing whether the kinetics belong to the same organ, 2) around each local minimum, definition of a sample of TACs representative of the voxel cluster (i.e. organ), and fusion of similar samples, 3) aggregation of TACs to the closest kinetics model.
Fourteen tracers derived from Peptid Nucleic Acid were marked by 18F, injected to four rats each. Whole body rat images were acquired with Siemens HR+ PET imager (Figure 1). Mean organ kinetics (Figure 3), in images segmented with the method (Figure 2), showed a 0.94 correlation with manually delineated organs mean kinetics.
Validation of the classification method steps using a realistic phantom based on the MOBY dynamic phantom (bladder filling has been added) will be presented. Images will be simulated with an analytic simulator including a realistic description of noise.
Abstract ID: 495
Bioluminescent imaging is a tool that can be used in transgenic (tg) animals to evaluate the efficacy of drug treatment. We have created a stable of tg mice containing luciferase reporters driven by promoters of genes activated during inflammation. The reporter constructs include: iNOS, VEGFR2, GFAP, SAA, EPX, NFκB-response elements, COX2, iκB and Il2. These transgenic mice have been used to study the progression of rheumatoid arthritis (RA), neuro-inflammatory disease, sepsis, inflammatory bowel disease (IBD), asthma, hypersensitive skin reactions, bacterial infections, wound healing and T-cell mediated responses. For example the NFκB-response element tg mouse reports on the final step of the NFκB signaling pathway and it has been used to characterize the temporal and spatial inflammatory response during RA, sepsis, and wound healing. The response to drug treatment provides clues as to the bioavailability and distribution of therapeutic agents. New instrumentation developed by Xenogen provides finer anatomical assessment of the tissues involved in the inflammatory response. These new tools include diffuse luminescent imaging tomography (DLIT) that provides a pseudo 3-dimensional representation of the luminescent source, as well as the IVIS® imaging system 3D. The 3D system requires multiple views and calculation of the 3-dimensional shape and size of the luminescent source inside of the animal.
Abstract ID: 496
We have upgraded a Mediso XRing camera with a dedicated multi-pinhole collimator for high-resolution and high-sensitivity single-photonimaging of mice and rats. In addition to the high-resolution multi-pinhole module, we have outfitted the system with an automated 3D-translation stage enabling us to perform helical scanning, Translation-SPECT, automatic positioning, etc (Figure 1). The system is currently outfitted with a single 2 mm intrinsic high-resolution detector (for 99mTc), though it is capable of housing up to four detector heads. With four heads, the system can achieve averagesensitivities as high as 2000 cps/MBq. Mouse and rat imaging is performed at radii between 25 and 40mm with corresponding magnifications of 6 to 3.5. We currently have five different aperture designs for small-animal imaging. In general, the locations, tilts and acceptance angle of each pinhole are chosen according to the relevant imaging task. Helical scanning is made possible by stepping the object through the field of view along the axis of rotation of the gantry. There construction is performed with a self-developed ray-tracing based MLEM algorithm with optional ordered subset implementation. In this work we present a description of our system along with a performance analysis and numerous small-animal studies (Figure 2). Additionally, our presentation will focus on the implementation of helical scanning; a feature that allows us to perform very high-resolution SPECT over a large field of view.
Abstract ID: 497
DCE-MRI: Vascular permeability was measured by DCE inversion recovery (IR) TrueFISP at 4.7 T. CA's (Dotarem 0.1 mmol Gd/kg BW; Vistarem 0.028 mmol Gd/kg BW) were infused by a spectrometer controlled pump.
Post-processing: Vascular permeability (Ktrans) was quantified using a two compartment kinetic model [1].
Ktrans was lower for Vistarem than for Dotarem owing to its larger hydrodynamic diameter (Vistarem: 5 nm; Dotarem: 1 nm). Decrease of Ktrans after treatment with PTK/ZK was more pronounced for Vistarem. This study has shown that large MW CA's are better suited for assessment of drug induced Ktrans changes in small animals.
Ktrans measured in BN472 tumors and B16 metastases
Abstract ID: 498
We report here the in vitro and in vivo characteristics of a new molecular conjugate vector for targeting and imaging of tumors. Its core is a cyclo-decapeptide plateform named RAFT, onto which two spatially independent functional domains can be covalently and stereo-specifically linked : a cell targeting domain for tumor targeting and a labeling domain able to carry two drugs and/or labeling agent. In order to prove the interest of this carrier, we used a well known cRGD cyclopeptide, a ligand for the αvβ3 integrin. We demonstrate that this vector presenting 4 cRGD motifs very efficiently prevents αvβ3-mediated cell adhesion to vitronectin. Furthermore, it is actively endocytosed because of the multivalent cRGD presentation, a major advantage for drug delivery. In vivo experiments in nude mice reveal that repeated intratumoral injections of low doses of RAFT(cRGD)4 reduce tumor growth by reducing the number of tumor blood vessels. Furthermore, RAFT(cRGD)4 significantly improves the targeting specificity of subcutaneous tumor masses as well as disseminated metastasis after intravenous injection (figure).
Thus, RAFT(cRGD)4 is specific, internalized, perfectly controlled and can carry multiple biological functions on a single, spatially defined backbone, making of it a powerful and versatile synthetic vector for drug delivery, molecular imaging or both.
This versatility is a major advantage since we can load it with multiple functions. This was demonstrated by the addition of a cytotoxic KL peptide on the cargo domain of the RAFT. RAFT-RGD-KL shows a very promissing specific antitumoral activity when injected intravenously in tumors bearing mice.
Finally, recent developments in the optical system allowed us to follow these optical probes using a new 3D optical tomographic system in lung metastasis. This allowed us to detect lung metastasis in 3D for the first time to our knowledge using optical NIR probes.
Abstract ID: 499
Abstract ID: 500
In solid organ transplantation, recipient actuarial survival has dramatically increased since the introduction of adequate immunosuppression such as cyclosporine. Despite the fact that acute rejection can in most cases be managed, chronic rejection continues to be a major challenge to achieve longterm graft survival.
Abstract ID: 501
Detection of bone formation rates by histomorphometric evaluation of fluorochrome markers in bone is very slow. It takes months from the actual in vivo experiment to the final readout. We tested whether we can monitor local bone forming activity in vivo by near-infrared fluorescence imaging (NIRF), using a near-infrared fluorescent bisphosphonate derivative homing to bone and binding to hydroxyapatite, as suggested previously [1].
In a first experiment we treated 3-month-old OF1/IC mice with 0.1 or 1 mg/kg of Cy5.5 labeled pamidronate or vehicle intravenously for 1 to 5 days. NIRF images of the calvaria of the anaesthetized mice were recorded 0.5, 1, 2, 4 and 24 hours after the last administration of the near-infrared fluorescent bisphosphonate derivative. Fluorescent signals could be detected in the calvaria of all mice treated with the labeled bisphosphonate in a dose- and application number-dependent manner but not in vehicle treated controls. Histological examination confirmed the presence of Cy5.5 signal on calvaria microtome sections of animals treated with the labeled bisphosphonate. In a second experiment 4-month-old OF1/IC mice were administered hPTH, BMP-2 or vehicle (PBS + 0.1 BSA) twice daily subcutaneously onto the calvaria for 5 or 7 days. The mice were administered intravenously 0.1 mg/kg of the Cy5.5 labeled bisphosphonate or control on day 7 and 8. A dose-dependent increase of NIRF signal was observed upon hPTH and BMP-2 treatment 4 and 24 hours after the last application of the labeled bisphosphonate.
Taken together these results confirm that homing of fluorescent bisphosphonate into skeletal structures can be monitored by NIRF. Furthermore they demonstrate that bone formation responses can be quantified rapidly in vivo by the use of this technology.
Abstract ID: 502
Fluorescence imaging of cells with conventional fluorophores or genetically encoded proteins suffers from the contribution of background noise, both out of focus emission as well as autofluorescence, to the detected signal. This can be alleviated by using a confocal microscope to reduce out of focus signal. Alternatively, the short lived autofluorescent component could be reduced by using long lived fluorophores. We have developed time resolved fluorescence resonance energy transfer (TR-FRET) assays with europium cryptates as the donor for dosing protein levels or to reveal protein-protein interactions. Europium cryptates are highly fluorescent and have a large Stokes shift with sharp emission peaks and long decay times (10 to 1000μsec). These characteristics make them ideal for immunolabelling since the short decay time of autofluorescence (less than 100ns) can be gated out by the detector and only the specific signal is measured. Background originating from the biological sample, the buffer and sample holder material is virtually eliminated.
Europium cryptates have not been used for cellular imaging so far. We have set up a time resolved imaging platform with a wide field microscope, a pulsed nitogen laser and an intensified camera synchronised with the laser. Time gating by the camera allows us to acquire images with a specified delay, thus filtering out autofluorescent background noise.
We have optimised time resolved fluorescence detection with different quantities of antibody-dye conjugates coupled to microbeads. Cells expressing tagged receptors have been labelled with europium cryptate conjugated antibodies and visualised. To the best of our knowledge this is the first time that cells have been imaged with long lived fluorophores at this high resolution. We are currently optimising the detection of time resolved fluorescence in living cells labelled with europium cryptates. This imaging modality will open up new possibilities for investigating molecular events inside living cells.
Abstract ID: 503
Abstract ID: 504
Optical molecular imaging is one of the fastest developing imaging modalities for preclinical testing of anticancer agents. Early xenograft tumor models have focused mostly on the use of cancer cell lines expressing fluorophores such as green (GFP) and red (DsRed) fluorescent proteins and different forms of luciferase. These approaches have now been supplemented by novel “smart” probes that release contrast fluorophores to report tumor-specific physiology.
In the current studies we have evaluated a protease sensing probe, ProSense™, and blood-pool probe for detection of tumor vasculature, AngioSense™ (VisEn Medical). These novel probes were studied in subcutaneous, metastatic and orthotopic xenograft tumor models utilizing cancer cell lines expressing fluorescence proteins and firefly luciferase (example in Fig.1). The correlation between tumor size measured by caliper, tumor volume detected by concentration of the smart probes and fluorescence-bioluminescence of the cancer cells from primary tumors and metastases were compared. Probe concentrations were determined using the FMT™ tomographic system (VisEn Medical) and compared with data from other imaging systems. These studies provide cross validation between established and novel optical imaging agents for in vivo cancer imaging research.
Tumor developed after subcutaneous implantation of HeLa-luc cells.
Abstract ID: 505
Abstract ID: 506
RNA interference offers the potential of a novel therapeutic approach for treating skin disorders. As a model to evaluate inhibition we delivered plasmids encoding a biscitronic message comprised of the reporter genes firefly luciferase (fLuc) and eGFP separated by a 2A ribosome slippage site, to mouse skin by intradermal injection into the footpads and evaluated expression patterns using in vivo bioluminescent imaging. Time course and dose response experiments showed peak gene expression at 24–48 hours post-delivery using 2 μg of reporter plasmid. Reporter expression was confined to skin keratinocytes as determined by sectioning of the skin and microscopy. Electroporation following injection resulted in a 10–100 fold increase in reporter expression. To evaluate inhibition of expression by RNAi, specific (directed at eGFP) or non-specific siRNAs (provided by Dharmacon) were co-injected with the reporter gene plasmids and the effects measured by imaging. The results revealed that specific unmodified and modified siRNAs directed against eGFP (but not non-specific matched NSC4 siRNA controls), strongly inhibited reporter gene expression in mice with 94 and 97% inhibition for unmodified and modified eGFP inhibitors, respectively (following normalization with footpads receiving NSC4 siRNA; n=3–5 mice). The next step was to evaluate the effects of injected siRNA inhibitors on pre-existing reporter expression. Using a transgeneic reporter mouse that constitutively express eGFP in skin keratinocytes we observed a specific reduction of greater than 90% with siRNA inhibitors. Non-specific siRNAs had little or no effect on pre-exisiting eGFP levels. These results indicate that siRNA (or small hairpin RNAs), delivered locally as RNA directly may be effective agents for treating skin disorders.
Abstract ID: 507
We have developed a technique for ultrasound-enhanced drug delivery, using molecular imaging to monitor resulting biodistribution. Engineered delivery vehicles consist of a gas core, a thick oil shell capable of carrying imaging probes and hydrophobic drugs, and an outer lipid membrane that incorporates targeting ligands. We have previously demonstrated that the application of a specific ultrasound pulse sequence in the presence of these vehicles can cause local build-up of a fluorescent model drug on a cell monolayer, with delivery limited to the focal area of ultrasound application. Here we demonstrate that (1) in vivo, drugs can be successfully delivered to or beyond the vascular endothelium, (2) chemotherapeutic-containing vehicles and ultrasound cause significant cytotoxic effects, and (3) enhanced drug delivery is achievable with the use of targeting ligands incorporated into the vehicles' outer lipid membranes. In vivo studies in a chorioallantoic membrane model show that fluorescent dye can be deposited on vessel endothelium or delivered to the interstitium, depending on the ultrasound parameters. In vitro, two cytotoxicity assays indicate that paclitaxel is successfully delivered to the surface of human melanoma cells. Viable cell area following treatment with ultrasound and paclitaxel-containing vehicles was reduced by more than an order of magnitude over control treatments using saline alone, ultrasound alone, or drug-free vehicles with ultrasound. Adhesion assays indicate that the addition of RGD targeting ligands to the vehicles' lipid shell improves binding to alpha-V-beta-3-expressing cells 26.5-fold over non-targeted agents; and the use of targeted vehicles and ultrasound increased cytotoxicity by another 3–5 fold over the same ultrasound treatment with non-targeted vehicles. These results collectively demonstrate that in vivo treatments with our technique are likely to be successful. Currently, studies with [F-18] paclitaxel have defined the biodistribution of free paclitaxel in rats, and PET studies of delivery vehicle biodistribution are underway.
Abstract ID: 508
PX-478 is an investigational drug that suppresses both constitutive and hypoxia -induced levels of the transcription factor, HIF-1α, in cancer cells and tumor xenografts[1]. This study was undertaken to identify and characterize non-invasive MR-visible biomarkers for PX-478 activity that may be used to monitor therapy response in clinical trials of this drug. Dynamic contrast enhanced (DCE)-MRI, diffusion-MRI and 1-H magnetic resonance spectroscopy (1H MRS) were used to measure, respectively, hemodynamics, cellularity and metabolism of HT-29 human colon xenografts at times following therapy with PX-478 [2]. PX-478 induced a substantial reduction in tumor blood vessel permeability within 2 hours (P=0.012), which lasted for an additional 22h (P=0.003) before returning to baseline at 48 hours. Immunocytochemistry and showed decreases in VEGF, but with a different time course from permeability. The apparent diffusion coefficient, ADC, increased 24h after treatment (P=0.005), and returned to control values by 48h. 1H MRS showed significant changes in the total choline (tCho) at 12 and 24 hr. Although there were decreases in the lactate+lipid resonance, these were not significant, and this was confirmed in extracts. In conclusion, pre-clinical trials have provided sufficient evidence to justify incorporating DCE and DWI imaging endpoints into upcoming clinical trials, and have also uncovered interesting tumor biology that occurs in response to drug.
Abstract ID: 509
The ever expanding applications of nanoparticles in diagnostic and therapeutic drug design for detection and treatment of diseases under in vivo conditions require a thorough understanding of the interaction of nanoparticles within the biological media. Pigs are excellent animal models because of their similar physiology and organ sizes to human beings. Pharmacokinetic studies of gold nanoparticles (AuNPs) were carried out in five pigs weighing 11–15 kg via intravenous (I.V) administration of 300–420 μg of starch stabilized AuNPs of 20 nm diameter per dose per animal. The results of bi-distribution studies of AuNPs were estimated from NAA and atomic absorption analysis. It is highly surprising that the %ID of accumulation of AuNPs in the liver is <3%. This finding is in sharp contrast to literature precedents where significantly higher uptake of AuNPs in the liver is prominent. Accumulation of AuNPs in lungs ranged from 7–14%ID. This presentation will include detailed pharmacokinetic and biolocalization studies of AuNPs stabilized by various chemical vectors using pig as the animal model. Pharmacokinetic and in vivo profiles relating to the controlled release of AuNPs on cells and tissue and their movement across the blood stream and ultimately efforts toward site specific localization of nanoparticles for potential applications in imaging/therapy will be discussed.
Abstract ID: 510
Poster Session: Optical Technologies
Abstract ID: 511
Bioluminescent imaging is a novel imaging modality in vivo for revealing the molecular and cellular activity and monitoring the growth and regression of tumor noninvasively. Improving the accuracy of numerical solution for diffusive photon propagation in complicated biological tissue is helpful to achieve better resolution in bioluminescent imaging. Adaptive finite element methods (FEMs) require lower computational complexity than general FEM at certain precise. In this work, based on diffusion equation we present adaptive FEMs to improve the accuracy of numerical simulation for diffusive photon propagation, and compare the numerical solution with the analytical solution in a homogeneous medium. Figure-1(A) depicts the initial finite element mesh and 1(B) indicates the solution of global precision at 0.01 is much closer to analytical result than at 0.1.
Furthermore, the experiment for physical phantom is conducted to compare simulation with experimental results. Figure-2(A) describes the experimental setup including the cylindrical phantom, the red luminescent source located in the undersurface center of the phantom, and the CCD camera parallel to the undersurface; Figure-2(B) shows that there is a good agreement between simulation and measured data of a row of pixels on the top surface. Figure-3(A, B, C) illustrates the adaptive process of finite element mesh. The results show that Adaptive FEMs effectively simulate diffusive photon propagation in bioluminescent imaging.
Numerical vs. analytic solution. (A) Initial finite element mesh of modeling ideal source and detection sphere, and (B) Comparison between the adaptive finite element solution and the analytic solution.
Experimental vs. simulative results. (A) Bioluminescent experiment for phantom, and (B) Comparison between experimental results and adaptive FEMs results.
Refinement process of finite element mesh. (A) Initial finite element mesh, (B) The second refined mesh, and (C) The third refined mesh.
Abstract ID: 512
Near-infrared time-resolved methods have proven to be an efficient diagnostic tool for determining optical properties of tissues. Specific fluorescent labelling is a way to provide an interesting contrast to detect malignant lesions. The aim of this work is to establish and investigate different methods for quantification and localization of a fluorophore embedded in a turbid medium and to test them on an experimental set-up developed in our laboratory.
Light propagation is assumed to satisfy time-dependent diffusion equation. Analytical solutions describing the fluorophore response are expressed under several approximations. Because they connect localization and concentration of the inclusion with the source-detector positions and optical properties of the medium, we can recover these data under few assumptions.
Two different ways to process the rich information contained in the temporal point spread functions are explored using the theoretical solutions. In the first method the different moments of the measured temporal point spread function have been calculated and compared to the theoretical ones to recover the position of the fluorophore. The second method involves a fitting algorithm based on frequency-domain calculation from at least two different measurements.
The instrumentation consists of a TCSPC system. Pulsed laser light is delivered into and collected from the medium by optical fibres. The collection fibre is coupled to a fast photomultiplier. According to the fluorophore, excitation wavelength and high pass filters are chosen to optimize the collected fluorescent signal. Experiments are performed with fluorophores embedded in tissue-like phantoms, with Intralipid® as the scattering medium and ink as the absorber.
Results are presented in a simple geometry of an infinite homogeneous medium and a point-like fluorophore. The two methods enable good localization of the inclusion. In the final paper, we will present measurement configurations representative of actual in vivo conditions by taking into account surface measurements and inhomogeneous media.
Abstract ID: 513
Bioluminescent probes are commonly used as reporters of gene expression in living cells for the study of biological activity and the development of therapeutic agents. Although the current trend for in vivo measurements is to monitor the relative emission levels, in some applications there is a need for more quantitative information regarding the spatial localization and relative power of the photon source distribution. In the work presented here, a high sensitive optical imaging system that detects the steady state emission from bioluminescent reporters on the surface of the animal is described and a diffuse tomographic imaging technique known as DLIT™ is demonstrated.
The reconstruction technique is based upon single-view two-dimensional images of the surface topography and the photon emission from the animal. The surface topography is determined from structured light data and the photon emission is measured at multiple wavelengths using bandpass filters. Assuming a homogenous tissue approximation, the bioluminescent source distributions in three-dimensions can be determined since light is attenuated differentially with respect to wavelength as it travels through tissues. DLIT gives the researcher spatial localization for an absolute quantification of the bioluminescent light sources inside the animal. The reconstructed data can then be overlaid onto a digital mouse atlas which references the source location with respect to the internal mouse anatomy.
Experimental results will be presented which validate the DLIT method. Data will be shown based on measurements made using an experimental mouse phantom that has optical properties similar to living tissue. Then, in vivo measurements from mice implanted with calibrated luminescent sources at several different locations show how well this technique determines the 3D source location and flux. Finally, reconstructions of B16-F10 melanoma metastasis in mice will be shown and compared to macroscopic cryosections.
Abstract ID: 514
Bioluminescent markers have been valuable tools for targeted protein research in small animals. Estimates of internal source distribution and intensity provide improved quantification of cells labeled by the luciferase gene. Source distributions and intensities were reconstructed using DLIT™, a tomographic algorithm for bioluminescent sources in optically diffuse media. CCD images of radiance emitted from animal surface are acquired in multiple views transaxially. The multi-view system makes manually flipping the animal to detect all signals unnecessary, as demanded by a single-perspective system. The animal surface topography, required to define the air-tissue boundary, is reconstructed from simultaneously acquired structured light images. For each view, diffuse photon propagation in the tissue is modeled with the assumption of homogeneity in a semi-infinite plane under the partial-current boundary condition. Details of the DLIT algorithm will be discussed.
The tomographic potential of our algorithm is investigated with experimental data. For validation with known homogeneous optical properties and source intensities, a mouse-shaped tissue phantom embedded with LED-sourced optical fibers was imaged spectrally in the multiple-view camera system. In vivo benchmarking with independently measured source fluxes was made by surgically implanting tritium-filled, phosphor-coated glass beads into living animals at various locations. In vivo studies with bioluminescent sources involved mice, which were administered B16-F10 melanoma luciferase-labeled cells intravenously. In addition, in vivo multiple-view data from bioluminescent transgenic animals have been collected. Animals were selected for post-mortem freezing and cryosectioning, or necropsy. DLIT analysis for bioluminescent sources is presented here, and solutions are compared with the post-mortem data. A digital mouse atlas scaled to the animal boundary is employed for validation studies.
The tomographic advantage of the multiple-view camera will be demonstrated with a comparison between the multiple-view system and a single-perspective camera system.
Abstract ID: 515
Band-pass and band-rejection interference filters have been employed extensively in small animal fluorescence enhanced imaging studies in order to reject excitation light and to enable efficient collection of weak fluorescent signals emanating from nano- to picomolar concentrations of fluorophore. However, elimination of the back-reflected excitation light remains a major challenge for further reducing the noise floor and increasing the sensitivity.
Herein, we present evidence to show that the combination of three-cavity interference and holographic super notch filters reject collimated excitation light while efficiently passing fluorescence signals. To assess excitation leakage, the “out-of-band (S (Λx))” to “in-band (S (Λm) - S (Λx))” signal ratio from phantom studies and the target to background ratio (TBR) and signal to noise ratio (SNR) from in vivo animal imaging were acquired with and without collimating optics. The addition of collimating optics resulted in a 51 to 75% reduction in the ratio of (S (Λx))/ (S (Λm) - S (Λx)) for the phantom studies and an increase of TBR from 11% to 31% and signal to noise (SNR) from 11% to 142% for an integrin-targeting conjugate in human glioma xenografts. The results suggest that sensitivity of fluorescence imaging can be improved by the simple addition of appropriate filters and collimation lenses.
Abstract ID: 516
This study is to compare three cameras to determine their spectral response sensitivity to bioluminescent and fluorescent light sources. The three cameras are the Xenogen IVIS 200®, Roper Scientific Versarray 1024, and Princeton Instruments PIXIS 1024B. Both the Roper and Princeton cameras have 1024 × 1024 image pixels and the Xenogen camera has 2048 × 2048 pixels. Testing various cameras with chemiluminescent kits would not be consistent since the half life of most chemiluminescent material is short and remixing with fresh chemicals would not be exact for each test. In contrast, an electronic phantom offers reproducibility. Two such phantoms were used in this study: the commercially available Xenogen XLS-4, and the second built in-house and consisted of four white LEDs and four color LEDs, with additional neutral density filters placed in front of white LEDs to extend its dynamic range.
Both phantoms were imaged for 20 seconds on all three cameras. The in-house phantom was imaged a second time on all three cameras with the addition of extra neutral density filters over the four white LEDs with reduction factors of 0.5, 0.1, 0.01, 0.001. The default imaging distance and lens f/stop for each camera were used as follows: f2.0 at 13.5 inches for the IVIS 200®; f1.2 at 18.5 inches for the Roper camera; f1.4 at 10 inches for the PIXIS. The default pixel binning was 4 for IVIS 200®, and 2 for the other two cameras. Overall as expected the IVIS 200® camera had the best performance. It has much higher photon counts for the same exposure time than the other two cameras, the best color response, and the best sensitivity to detect small differences in light outputs. The performances of the other two cameras are comparable with the PIXIS better than the Roper camera due mainly to the difference in imaging distance.
Abstract ID: 517
Molecular imaging (MI) aims at the detection of molecular processes in-vivo and promises to dramatically improve the future of healthcare. Therefore, molecular imaging is since a couple of years a rapidly expanding field. MI requires the specific targeting of a detectable unit to a target. For this, recombinant targeting proteins like antibodies or genetically engineered fragments thereof are of great interest. Here we developed novel conjugates consisting of fluorescent semiconductor nanocrystals, quantum dots (QD) and genetically engineered human single chain antibody fragments (scFv). The advantages of QDs are their small size, high quantum yield, broad absorption, narrow emission peak, which in addition is tuneable by size and high photostability. One conjugate we report here targets MUC-1. This is a tumour-associated transmembrane protein, suitable for diagnosis of e.g. breast cancer. The scFv was conjugated to red fluorescent QDs with the right orientation through site-specific binding. The optical properties of the M12QD remained sufficient for optical detection. This conjugate, M12QD, proved stable for at least 21 days in human serum. The M12QD conjugates showed specific binding to human cell lines expressing the MUC-1 antigen (see Figure) whereas non-expressing cell lines showed no binding. Non-conjugated QDs showed no binding.
Abstract ID: 518
Abstract ID: 519
The overall goal of this study was to validate 3D fluorescence mediated tomography (FMT) for the quantitative measurements of fluorochrome-based affinity tags in live xenograft mouse models. The imaging probes used in this study consisted of a) a long circulating magnetofluorescent nanoparticle (CLIO-Cy5.5), b) a long circulating polymer (Angiosense-750, VisEn Medical, Inc, Woburn, MA) and c) fluorescently labeled herceptin (Herceptin-Cy5.5; Trastuzumab, Genentech, San Francisco, CA). The FMT imaging system was based on a prototype, similar as previously described (Proc Natl Acad Sci USA; 2004: 101:12294–12299) but with improved reconstruction and display options and multiwavelength capability. Breast cancer (MDA-MB-468) and glioma (9LGFP) xenografts were used with or without avastin (bevacizumab, Genentech, San Francisco, CA) treatment as model systems. We show that a) FMT measurements of angiogenesis closely correspond to histopathological gold standards and MRI/SPECT measurements, b) that angiogenesis inhibition can be reliably measured within the same animal (Fig 1) and c) that it is feasible to perform multichannel measurements of distinct biological phenomena within short periods of time (Fig 1). FMT should be valuable for rapidly profiling biological phenomena in vivo.
Abstract ID: 520
The statistical evaluation of molecular imaging approaches for detecting, diagnosis, and monitoring molecular response to treatment are required prior to their adoption. The assessment of fluorescence optical imaging is particularly challenging since neither instrument or agent has been established. Small animal imaging does not adequately address depth of penetration issues and the risk of administering all molecular optical imaging agents into patients remains unknown. In order to generate the data for statistical analysis, we employ a model of our imaging system and the “lumpy object” model developed by Rolland and Barrett to simulate natural anatomical tissue structure as well as represent the non-specific distribution of fluorescent contrast agents.
We predict the time-dependent light propagation within a 300 cm3 breast interrogated with 25 points of excitation illumination and 128 points of fluorescent light collection. We simulate the fluorescent generation from Cardio-Green at tissue target concentrations of 1 micromolar with backgrounds containing one hundred fold less. Actual experimental measurements in a phantom system match our predictions for uniform background properties (Godavarty et al., Radiology, 2005). Predictions for targets simulated within a background of (i) a heterogeneous endogenous optical properties (absorption and scattering) with strength equivalent to 200% the average background properties and (ii) heterogeneous exogenous fluorophore cross section with strength equivalent to 5% of the background values were used as input into the tomographic algorithm (Roy et al., IEEE Transactions on Medical Imaging, 2005). Our results show that image recovery is robust to heterogeneity of endogenous optical properties indicating that objective assessment of image quality (OAIQ) through statistical methods is possible. In addition, we find that the image recovery is limited by the strength and uptake of fluorophore cross section. OAIQ may therefore provide insight to the degree of molecular specificity required for developing optical imaging agents.
Abstract ID: 521
A small animal multimodality tomographer dedicated to co-registration of fluorescence optical signal and X-rays measurements is under development in our laboratory. The purpose of such a system is to offer the possibility to get in vivo anatomical and functional information at once. Moreover, anatomical measurements will be used as a regularization factor in order to get the reconstructions of the biodistribution of fluorochromes more accurate and to speed up the treatment.
The best configuration for X-rays imaging is obtained when the animal is illuminated perpendicularly and around its vertical axis. Furthermore, analytical solutions of the optical problem can be derived for simple geometries. The cylindrical geometry has been naturally chosen as a compromise: i) this geometry is suitable with X-rays cone beam CT of the whole animal body; ii) by immersing the animal in an index matching fluid, an analytical solution to the optical forward problem can be derived and introduced in a reconstruction scheme.
An exact analytical solution to the Diffusion Approximation to the Radiative Transport equation used for modelling the optical forward problem has been established. For this geometry, we proceeded by analogy with the classical solutions to the equations of heat conduction in solids.
The optical system is basically composed with a laser beam for excitation of the fluorochromes, and a CCD camera coupled with a chromatic filter for the fluorescence detection.
Experimental measurements on cylindrical phantoms with different fluorescent inclusions geometries have been performed as a validation. The cylindrical analytical approach has been compared to a Finite Element Method approach: analytical approach brings a significant gain of computing time.
Abstract ID: 523
Optical imaging and tomography in tissues immerged as a valuable modality for the quantitative study of several important processes in biology and medicine. In recent years Fluorescence-mediated Molecular Tomography has been used in the study of several important functions and molecular pathways in-vivo in whole animals. When applied in reflection geometry it can facilitate the detection of fluorescent signals from superficial organs and targets in a highly sensitive and quantitative fashion, by far improving imaging compared to the photographic methods used to date. However, skin autofluorescence can be a limiting factor for the contrast and sensitivity of the technique, significantly impairing imaging quality.
We have performed in-vivo autofluorescence measurements on the skin of transgenic mice covering the whole body in a geometry based on reflection tomography acquisitions. The system employed the combined use of a spectrograph, a highly sensitive CCD camera and an argon ion laser as the excitation source covering a spectral range of about 200nm in the visible, as well as the use of a back illumination setup enabling the highly accurate three dimensional surface reconstruction of bodies with arbitrary shapes. We were thus, able to accurately render the autofluorescence signals on the reconstructed surface of the mice creating a skin autofluorescence map. Furthermore the tomographic system can facilitate the detection of fluorescence in many spectral regions using different filters and applying simple spectral deconvolution algorithms for distinguishing different fluorophores with overlapping spectral signatures, as well as isolating and subtracting the autofluorescence signal. Thus, highly contrast tomographic images can be obtained and better quantification accuracy can be achieved, improving our ability for in vivo detection and monitoring of fluorescent signals in whole animals.
Abstract ID: 524
Imaging of fluorescence proteins has gained significant attention in recent years both at the microscopic level using a variety of advanced microscopic techniques, such as in-vivo, two-photon and multi-photon microscopy as well as macroscopically using photographic methods in whole animals. The latter has been shown to be useful in detecting and following tumor progression in-vivo but is suited for imaging superficial activity or imaging and quantification is otherwise significantly complicated due to the non-linear dependence of signal intensity to lesion depth and tissue optical properties.
Here we present an imaging modality developed for tomographic imaging and quantization of FP expressing cells in vivo in whole animals. The technique employed a multi-angle, multi-projection transillumination scheme and a highly sensitive CCD camera for detection and can be used in a wide wavelength range spanning the visible and near infrared. Light propagation was modeled using a normalized Born field and an absorption dependent diffusion coefficient that accounts for the high absorption of visible light into tissue, extending the validity of the diffusion equation.
We have been able to detect FP expression in deep tissues quantitatively (current detection threshold of 50 × 103GFP cells in the center of a mouse) and have used the method for in-vivo real time imaging of GFP expressing lung tumors and contrasted results with X-ray CT images. Other tomographic imaging and multicolor detection of viral transfection and gene delivery in-vivo in a tumor model was also possible. The technique can offer unprecedented capacity in quantitatively detecting deep-seated FP activity and has shown insensitive to tissue auto-fluorescence. This technology can serve as a significantly more accurate detection method for macroscopic fluorescent protein imaging in whole animals.
Abstract ID: 525
At present, tomographic optical imaging of fluorescent/bioluminescent marker distributions in vivo has been accomplished through fiber-based detectors in contact with the object or by means of lens mounted CCD cameras. While the former is limited by insufficient spatial sampling and object-fiber coupling issues the latter is restricted by the dynamic range of the CCD. To overcome these shortcomings we propose a non-contact imaging system concept using a cylindrical arrangement of micro lens array based detector blocks, and demonstrate its performance by means of Monte Carlo simulation studies. For the purpose of this study, a single detector unit consists of a 10mm square size lattice of 10 × 10 lenses of 1mm diameter/pitch each, yielding a 1mm intrinsic spatial detector resolution. A network of optical fibers is mounted on a multi-hole plate such that the focal points of the individual micro lenses correspond locally to single fiber ending points. Depending on user-selectable operation mode each lens-fiber unit takes on two purposes: either light collected by the lens will be directed through the fiber to a photo diode (array) for detection or light from a laser diode (array) will be directed to the imaged object for fluorochrome excitation. In order to block ambient light a parallel beam collimator with matching hole diameter/pitch is attached to the object-facing side of the micro lens array. For tomographic imaging a cylindrical ring incorporating ten such detector blocks is formed yielding a transaxial diameter of 30mm; an axial field of view of 80mm is defined by axially mounting eight rings. One advantage of using micro lens arrays, as compared to CCD camera(s), lies in the dynamic range of the overall optical system which is greatly improved allowing for fast (parallel) fully tomographic list-mode data acquisition independent of laser excitation position and pattern.
Abstract ID: 526
Abstract ID: 527
Accurate in vivo assessment of bone remodeling is critical for proper management of pathologic skeletal events. This study demonstrates the utility of in vivo fluorescence molecular tomography (FMT) in monitoring bone turnover using near-infrared fluorescent probes in two animal models of bone repair.
A tibial defect model was created to generate focal regions of new bone formation similar to that observed in fracture. Mice received multi-wavelength doses of Osteosense, a near-infrared bisphosphonate-based probe that incorporates into new bone, at early (3d), late (7d), or combined (3 and 7d) time points to assess temporal healing. FMT imaging on day 4 revealed significant probe uptake in the growth plate, however no new bone formation was evident local to the defect, consistent with early fracture healing events. On day 8, FMT revealed focal regions of probe uptake distinct from the growth plate. Mice were sacrificed and tibiae were dissected and cleaned of soft tissue for validation of FMT by ex vivo fluorescence reflectance imaging (FRI). High probe uptake was evident both at the defect site as well as in regions distal to the defect, suggesting an associated adaptation response.
In a model for prostate cancer metastasis to bone, nude mice received intra-tibial injections of PC-3 cells. Four weeks post-injection, FMT and FRI imaging followed a single Osteosense dose. PC-3 cells induced local osteolysis, progressing to fracture in several mice. Robust probe signal suggested new bone formation in metastasis-associated fracture calluses, while purely osteolytic non-fractured tibiae were indistinguishable from saline-injected controls, consistent with probe mechanism of action. Following sacrifice, ex vivo FRI imaging and histologic observation confirmed these findings. These results demonstrate the utility of FMT in assessing bone turnover associated with a variety of skeletal pathologies.
Abstract ID: 528
Tomographic reconstructions of attenuating and fluorescing objects buried deep in diffusing tissue and tissue phantoms are presented through the use of time-gated selection of early photons. Early photons travel preferentially along or around a straight path from source to detector, and as such represent less-diffuse sampling of a medium than traditional continuous-wave measurements. Here we present the first such measurements to be combined with high angular sampling and large numbers of source-detector pairs, and demonstrate the reconstructed image enhancement in comparison with continuous-wave measurements.
Abstract ID: 529
Abstract ID: 530
Multiphoton fluorescence microscopy has been demonstrated to be of great value for deep tissue penetration, particularly when used with exogenous or with genetically expressed fluorophores (e.g. GFP). However, many interesting molecular targets are invisible to this method because the quantum yield is low.
We have demonstrated two-photon absorption imaging, both in live mice (with B16 tumors) and in excised human melanomas. In the cases we have demonstrated (with 800 nm excitation) contrast comes primarily from hemoglobin and melanin, but other water windows permit access to a wide spectral range. Melanin absorption in deep tissue is an interesting target, as spatial distribution of melanin (and lesion thickness measurement without excision) are of great interest to the dermatology community. However, melanin has extremely low quantum yield for fluorescence. In both cases, the two-photon absorption imaging produces vastly different images than does conventional microscopy or two-photon fluorescence.
Figure 1 (left to right) shows conventional, two-photon fluorescence, and two different two-photon absorption images (surface and depth-resolved) of an excised human melanoma. These images show a 266 micron field of view, with substantial signal to about 200 microns in the absorption case, but far less in the other two images. We can also measure quantum yield, and (with a higher sensitivity variant we have demonstrated) self phase modulation as complementary signatures from lower concentration species.
This work is supported by the NIH.
Abstract ID: 531
Many of the current approaches to small animal optical tomography are based on the diffusion equation. While the diffusion equation has been validated for near-infrared light in large, clinically relevant volumes, its applicability is uncertain in small volumes. We are developing a forward model for predicting time-dependent photon migration using the radiative transport equation (RTE) for the 3-D, in vivo reconstruction of the absorption cross section due to fluorophores emitting in the visible range.
Our radiative transport model uses a lumped linear discontinuous finite element scheme to predict the time-dependent propagation of photons through small volumes. This algorithm is unique in that its solution (i) is independent of differencing step size and (ii) approaches the diffusion limit when the light propagation and optical properties are appropriate. To validate the model's time-dependent predictions, a cylindrical holding tube was constructed with source and detector fibers placed circumferentially about the tube as illustrated in Figure 1. The tube was filled with a tissue mimicking fluid and heterogeneities (fluorescent, absorptive, and void-like) were suspended therein. Time-dependent data was obtained at 100 MHz using frequency-domain homodyne techniques.
Our results show the ability to accurately predict the time-dependent propagation of excitation and fluorescent light through a wide range of optical properties and time-of-flights that are encountered in small animal imaging. The predictions from the RTE are compared to those from the diffusion equation and demonstrate the accuracy of RTE for small volumes typical of murine models. Our results further demonstrate that the additional information obtained from time-dependent techniques provides more accurately reconstructions when compared to that from time-invariant, or continue wave measurements.
This work is supported by NIH EB0003132.
Schematic of instrumentation used to validate the time-dependent predictions of our RTE reconstruction algorithm.
Abstract ID: 532
Real time, imaging guided surgery is critical in defining tumor margin during surgical procedures, especially for brain tumor surgery. The extent of tumor resection is the survival predicting factor in patients with glioblastoma multiforme or brain metastasis. However, accurate determination of tumor margin intra-operationally remains a challenge. The gold-standard imaging modality, intra-operative magnetic resonance imaging (MRI), requires expensive equipment and interruption of surgical procedure and is not target specific. Therefore, an accurate, inexpensive, noninvasive, target-specific, and real-time imaging modality is needed. Near-infrared (NIR) optical imaging offers unique advantages due to its high sensitivity, versatile fluorescent probe design, and the ability to provide real-time data. Herein we report the real-time NIR optical imaging guided surgery of brain tumors in a xenograft models. A NIR dye Cy5.5 (excitation/emission, 675/694 nm) peptide conjugate targeted to MMP-2/9 was used in the study. Human melanoma tumor cells M21 were inoculated intracranially. Real-time imaging guided surgery was performed on day 8 after the inoculation. The imaging properties were assessed using continuous wave fluorescence imaging accomplished via an intensified charge coupled device camera. Data was acquired at 800 milliseconds per frame. Images were obtained 24 hours following the tail vein injection of contrast agent at an equivalent dose of 15 nmol of Cy5.5. NIR imaging showed selective accumulation of the contrast agent in the tumor and helped to define the tumor margin in real-time. Pathology validated the real-time NIR imaging guided surgical result that no tumor cell was found at the surgical margin. Our data demonstrate that NIR optical imaging can be used to guide surgical procedure in real-time. Such an imaging technique may be used in defining intra-operational tumor margin. (Supported by NIH grants R01 EB00174, Prostate Cancer SPORE Career Development Award CA90271, and Texas Advanced Technology Program).
Abstract ID: 533
Abstract ID: 534
We have developed a three-dimensional (3D) image reconstruction algorithm for whole-body optical molecular imaging of fluorescent probes using non-contact fluorescence measurements. This algorithm is based on a radiative transfer model for visible and near-infrared light and calculates the 3D spatial distribution of fluorescent molecular probes in small animal models of human disease. CCD camera images of the fluorescent light distribution on the tissue surface can be taken upon excitation by an external light source at different positions. The camera images become input to the tomographic image reconstruction code. This code comprises a nonlinear optimization method that iteratively employs a light propagation model based on the equation of radiative transfer and an updating scheme for the unknown fluorescent source concentration. The numerical light propagation model predicts the fluorescence light distribution on the tissue surface. It uses a structured 3D cartesian grid and a blocking-off region approach that takes the curved geometry of the tissue surface into account. An objective function compares the predicted fluorescent light intensities on the numerical mouse model surface with the CCD camera images. The optimization method minimizes the objective function and subsequently reconstructs the 3D spatial distribution of fluorescent source concentration inside the tissue. We have tested the performance of the reconstruction code using synthetic measurement data and a numerical mouse model consisting of 80,000 grid points as shown in Figure A). The mouse model contained a fluorescent source with size of 1.8 mm3 and dye concentration of 200nM close to the liver area. The reconstructed fluorophore concentration distribution is shown in Figure B).
Abstract ID: 535
Abstract ID: 536
The development of novel Indocyanine derivatives, targeted dyes and smart molecular beacon-like agents have opened exciting pathways for research and disease detection based on functional and molecular contrast. Fluorescence Molecular Tomography (FMT) is an imaging technique developed to three-dimensionally image and quantifies fluorescence bio-distribution in-vivo.
An important aspect of tomographic performance is the ability to produce accurate images in highly heterogeneous backgrounds and in the presence of background fluorescence due to intrinsic auto-fluorescence of tissues and imperfect probe targeting or specificity. In this study, we have investigated the performance of the reconstruction achieved when strong background heterogeneity and fluorescence is present. We also evaluated the performance of two analytical subtraction schemes applied either on the raw data or on the three-dimensional reconstructed fluorescence. The subtraction scheme on the raw data uses an analytical model, based on the diffusion approximation, of the fluorescence produced by an homogeneous fluorescent medium.
Experimental measurements were acquired using a previously described FMT system consisting of a parallel plate-imaging chamber and a lens coupled CCD camera, which enables conventional planar imaging as well as fluorescence tomography. Tissue-mimicking phantom studies were conducted using tubes containing Cy5.5 dye with different concentrations immersed in optically homogeneous and heterogeneous media containing background fluorescence. In vivo studies in tumor bearing mice in the presence of non-specific background fluorescence also demonstrated significant improvement in image quality and detection sensitivity.
Reconstruction obtained with the FMT system illustrating the background subtraction procedure on a phantom made of a plastic tube containing 80nM of Cy5.5 immersed in an intralipid solution.
Reconstructed fluorescence from a tumor in a Her-2/neu transgenic mouse model developing spontaneous mammary tumors and imaged with a cathepsinactivable fluorescent probe.
Abstract ID: 537
An adaptive finite element based reconstruction algorithm for high resolution, fluorescence tomography is developed and validated with non-contact plane wave frequency domain fluorescence measurements on a tissue phantom. Frequency domain data collected at the illumination surface was used to reconstruct the fluorescence yield distribution in a 512cm3 cubical phantom filled with 1% Liposyn solution. Fluorescent targets contained 1 micro-molar Indocyanine Green in 1% Liposyn and were suspended at depths of up to 2cm from the illumination surface. The tissue phantom was illuminated at the top surface by an expanded 785nm laser modulated at 100 MHz. Fluorescence measurements at the illumination surface were acquired by a gain-modulated image intensified CCD camera system outfitted with holographic band rejection and optical band pass filters. Excitation light at the phantom surface was characterized by utilizing cross polarizers.
The image reconstruction problem was posed as an optimization problem in which the fluorescence yield map minimizing the difference between the experimentally observed boundary fluorescence and that predicted from the diffusion model was sought. A regularized Gauss-Newton algorithm and dual adaptive meshes were employed for solution of coupled photon diffusion equations and for updating the fluorescence yield map in the tissue phantom. Three dimensional images of fluorescence yield distribution in the interior of the tissue phantom are presented.
Figure 1 illustrates a typical reconstructed image and adapted mesh for a fluorescent target at the depth of 1 cm. The mesh automatically refined itself around the target boundary, providing high resolution images. Analysis of reconstructed fluorescent target depth and size showed the ability to recover targets as deep as 2cm. This work demonstrates the feasibility of high resolution, non-contact, optical molecular imaging.
Abstract ID: 538
Poster Session: Multifunctional Probes
Abstract ID: 539
Lanthanide ions are widely used in a variety of molecular imaging applications. In recent years the solution state luminescence properties of lanthanide ions have found relevance in areas such as bioassay, imaging and sensor development [1]. We are interested in the syntheses of multinuclear complexes which incorporate lanthanide ions and their photophysical properties. In particular, complexes incorporating near-IR emitting lanthanide ions (ytterbium, neodymium and erbium) are of interest due to the optical characteristics of biological tissue (which absorbs far less near-IR radiation potentially allowing a greater imaging depth). Our synthetic strategy is dictated by the need for kinetic stability and high molar absorption characteristics.
The former is addressed by using macrocyclic or DTPA-like ligand derivatives to encapsulate the lanthanide ions, whilst the latter is achieved by the incorporation of either aromatic chromophores or d-metal components.
This presentation will describe multinuclear lanthanide complexes based on cyclen derivatives [2,3] and the synthetic methodologies used to access mixed f-f species [4]. Routes to covalently linked, kinetically stable d-f hybrids [5] will also be discussed together with the use of self-assembly experiments to effectively screen the photophysical components [6]. In all cases the near-IR emissive properties of the species will be described including the use of time-resolved spectroscopy to interpret multi-component luminescence.
Abstract ID: 540
There are a wide variety of bioluminescent reporters available for use in transcriptional or biochemical reporter assays. However, they cannot be monitored simultaneously. In particular, a method for spectrally deconvoluting multiple luciferase signals in cell lysates for high throughput screening is sought. Furthermore, it would be desirable to determine whether this strategy could be extended to applications in cellulo or in vivo. To implement this strategy, images were acquired sequentially using a CCD camera equipped with optical filters. However, there was still significant spectral overlap in the images. To address this problem, a Java plug-in for ImageJ, “Biolumunmixing v. 1.0,” was written to deconvolute bioluminescent images composed of signals from multiple luciferases using a linear unmixing algorithm. In order to validate this methodology, “Biolumunmixing v. 1.0” was tested using both simulated and real luciferase images. HeLa cells were transfected with either click beetle red luciferase (CBR), click beetle green luciferase (CBG99), or humanized Renilla luciferase (Rluc). Lysates from these cells were then mixed in varying proportions in a 96 well plate and imaged on an IVIS 100 using a <510 nm short pass filter, a 540 nm band pass filter, and a >650 nm long pass filter. After spectral deconvolution, one could accurately (within 10% error) recover the predicted pure luciferase signal from a well containing a mixture of all three luciferases. This work was further extended to counting mixtures of live cells by transiently expressing CBR or CBG Finally, transcriptional and proteosomal reporters for studying the NF?B pathway were also successfully tested in cellulo. These results demonstrate that reporter activity, transfection efficiency, and cell viability can all be determined in a single well during time resolved bioluminescence reporter assays in cellulo. Thus, “Biolumunmixing v. 1.0” provided a rapid,
Abstract ID: 541
Multimodality imaging is increasingly important in the field of MR molecular imaging. The validation of MRI findings with optical methods at the subcellular level may greatly enhance the understanding of the mechanism and location of contrast agent accumulation. Quantum dots are colloidal semiconductor crystals, which are a few nanometers in diameter and have excellent fluorescent properties. The emission wavelength is dependent on the quantum dot composition and size. Quantum dots can be solubilized and functionalized by coating them with a biocompatible and hydrophilic coating.
In this study, we prepared quantum dots with a paramagnetic lipid and hydrophilic poly(ethylene glycol) (PEG) lipid coating. The relaxivity of these quantum dots at 60 MHz is 3 mM-1s-1 when expressed in Gadolinium concentration. Per particle approximately 150 amphiphilic Gd-chelates were incorporated in the lipid shell. This means that the relaxivity per particle enhances dramatically to 450 mM-1s-1. Furthermore, the shell of these quantum dots can be used to functionalize them by covalently attaching targeting ligands like antibodies and peptides to the distal ends of the PEG-chains. Our results suggest that this nanoparticle may serve as a useful molecular imaging contrast agent to investigate disease processes at the molecular level with both MRI and fluorescence microscopy.
Abstract ID: 542
Recent developments in MR imaging have enabled in vivo imaging at near microscopic resolution. In order to visualize and track cells by MR imaging, it is necessary to tag cells magnetically. Cell-penetrating peptides (CPPs) have been used as an efficient way of internalizing a number of marker proteins into cells. Here we describe the synthesis and testing of a series of bi-labeled (magnetic and fluorescent) Gd(III)-based MR contrast agents conjugated to fluorescent dye and CPPs, including l-Tat49–57, d-Tat57–49, PTD-4 and NLS (Figure). The CPP fragments were synthesized by solid phase with the Fmoc (9-fluorenylmethoxycarbonyl) mediated scheme. FITC (fluorescein isothiocyanate) was coupled to Fmoc-lysine at first. Then FITC-Fmoc-lysine and diethylenetriaminepenta-acetic dianhydride (DTPA dianhydride) were coupled to CPPs, respectively. Finally the conjugates were chelated with Gd3+. The products were purified by reversed-phase HPLC and characterized by ESI-MS. Cellular uptake of these agents were confirmed by fluorescent microscopy and spectroscopy, as well as by T1 and T2 MR analysis of the Gd(III) agents in NIH 3T3 fibroblasts. Further optical and MR evaluation is under progress. The comparison of these different CPP conjugates can provide helpful data for the design of new intracellular MR contrast agents for in vivo tracking of cells.
Abstract ID: 543
Reporter genes are widely applied to study gene expression and regulation in biological systems. Bacterial LacZ gene, which encodes enzyme beta-galactosidase, represents one of the most commonly used reporter genes, among others such as green fluorescent protein and luciferase. Together with fluorogenic and chromogenic substrates of the enzyme beta-galactosidase, LacZ has been utilized as a standard method of assaying clonal insertion, transcriptional activation, protein expression and protein interaction as well. The use of MR detectable substrates provides a new non-invasive tool for detection of gene expression [1].
beta-Galactosidase catalyzes the hydrolysis of the glycosidic bond between the anomeric carbon at position C1 of the beta-D-galactopyranose and an aglycone part. The replacement of hydroxyl groups at positions C2 - C6 of galactopyranose may induce the loss of enzyme activity. However, it offers a possibility to develop new marker molecules for bi-modal detection by MR/optical imaging.
Here we report the syntheses of a series of compounds containing a MR detectable part (Gd(III)-DO3A) attached to beta-galactopyranose moiety by different types of linkers, and a chromophore (p-nitrophenol or 7-hydroxy-9H-(1,3-dichloro-9,9-dimethylacridin-2-one) (DDAO) [2]) at the anomeric carbon for testing enzyme activity. The tests are based on colorimetric detection of released yellow p-nitrophenol or the absorbance/fluorescence shift of released DDAO upon beta-galactosidase hydrolysis.
These compounds can prove helpful for the development of new gene expression markers offering the possibility of detection by MR and optical imaging. Coupling to molecules which facilitate internalization into cells can provide a new class of intracellular contrast agents.
Abstract ID: 544
Noninvasive imaging techniques like MRI possess the prospective to observe molecular-genetic and cellular processes. But the exogenously administered molecular imaging agents are often unable to reach their molecular and cellular targets as the lipid bilayer of the cell poses a formidable natural barrier. However, a unique class of peptides known as cell penetrating peptides (CPPs) has the ability to traverse this barrier and convey cargo molecules attached to it across the cell membrane. Amongst a variety of natural and chimeric CPPs, HIV-1 tat protein derived Tat peptide (Tat49–57) has received much attention mainly because of its high efficiency to deliver a large variety of cargo molecules across the membrane.
Considering the potential of Tat as a molecular transporter, we coupled its derivatives with fluorescence imaging agent FITC as well as with MR agent Gd-DTPA, thus obtaining bimodal cell internalizing agents. We aimed at comparing the effect of chirality on the internalization efficiency and thus synthesized l-Tat49–57 and its retro-inverso isomer d-Tat57–49. The effect of the two isomers on the vitality of cells and induction of metabolic changes was also studied. Fluorescence microscopy and spectroscopy as well as MRI were used on cell cultures for these studies.
Although both the peptides showed concentration and time dependent cellular delivery, the unnatural d-Tat coupled agent exhibited better cellular internalization compared to the natural form. This could be attributed to the increased stability of the d-form of the peptide to enzymatic cleavage. Also the d-form affected the cell vitality at concentrations above 9μM and induced slight changes in metabolic activities.
Abstract ID: 545
In recent years, many contrast agents have been developed and are being used as non-specific extracellular MR contrast agents. Intracellular contrast agents are another class whose prerequisite is to cross the plasma membrane by different mechanisms like endocytosis. For therapeutic applications, internalization is an absolute precondition, as the desired moiety has to reach its target within the cell. Cell pene-rating peptides (CPPs) have been used to facilitate the delivery of macromolecules into the cells both in vitro and in vivo. The transfection ability of D-octaarginines, L-octaarginines and stearyloctaarginines has already been studied; stearyl-D-octaarginines showed improved transfection efficiency [1]. This versatile approach can be used for the internalization of contrast agents into the cell for bimodal optical and MR imaging.
As a part of our ongoing research on synthesis of intracellular contrast agents, we have synthesized D-octaarginines, L-octaarginines and stearyloctaarginines covalently attached to the well established contrast agent (Gd-DTPA) and a fluorescent moiety, useful for MR imaging as well as optical imaging. Synthesis was performed on polystyrene based Wang resin, containing Fmoc-protected arginine. The synthesized peptide was coupled with lysine linked to FITC by its ɛ amino group and further conjugated with DTPA. In case of lipid modified octaarginines stearyl-lysine was additionally introduced in between polyarginine and contrast agent. Finally complexation with Gd (III) was performed at pH 5–6 in room temperature. Comparative studies on the uptake mechanism were done by both optical (fluorescent microscopy and spectroscopy) and MR imaging techniques. The optimization of the cellular uptake can prove useful by coupling to the arginines and its derivative (stearyl) for the development of new intracellular targeted contrast agents.
Abstract ID: 546
Apoptosis plays an important role in the etiology, pathology and treatment of a variety of diseases, including myocardial infarction and cancer. Early detection of apoptosis in-vivo would allow disease staging and evaluation of disease treatment at an early stage. To that aim, we propose the use of targeted lipid-based bimodal contrast agents to allow parallel high-resolution MRI and optical imaging to study apoptosis in-vivo and in-vitro. Superparamagnetic micelles (~10nm, relaxivity per mM Fe : r2 ≈ 160 mM−1s−1 at 20 MHz) and paramagnetic liposomes (~100nm, relaxivity per mM Gd: r1 ≈ 7 mM−1s−1 at 20 MHz) were developed as potent T2- and T1- reducing agents, respectively. In both systems fluorescein lipids are incorporated for optical detection. The protein annexinV, which binds to phosphatidylserine on the outer layer of the apoptotic cell membrane, is covalently coupled for targeting. Apoptotic Jurkat cells that were treated with annexinV-functionalized superparamagnetic micelles (5 μg/ml iron) showed a 64% T2 decrease compared to untreated apoptotic control cells, whereas annexinV-functionalized paramagnetic liposomes (0.2mM total lipid) led to 89% T1 decrease compared to control cells. Decreased relaxation times were confirmed to result from association of the contrast agents to apoptotic cells by confocal microscopy. In conclusion two types of annexinV-functionalized contrast agents were presented that can be used for in-vivo imaging of apoptotic sites in multiple models exploiting either their T1- or T2-based contrast properties.
Abstract ID: 547 “
Abstract ID: 548
Currently, multimodality is at the focus of attention in biomedical imaging, whereas at the same time efforts are made to improve the affinity of target-specific contrast agents for their marker through multivalency [1]. Dendrimers with their well-defined, highly branched structure are uniquely qualified to address these issues. Recently, we reported a general strategy to introduce multimodality to target-specific oligopeptides, using a sequence of highly efficient, chemoselective reactions, such as native chemical ligation [2,3].
Here, we demonstrate the general applicability of this strategy for the synthesis of dendritic target-specific contrast agents. Both poly(propylene imine) dendrimers and poly(lysine) dendritic wedges are functionalized with multiple oligopeptides and a tunable number of labels (Figure 1).
In our research we focus on the synthesis of target-specific contrast agents for the imaging of cardiovascular disease. The in vitro and in vivo efficacies of these contrast agents are currently under investigation.
Schematic representation of a multivalent, target-specific contrast agent.
Abstract ID: 549
Efficient routes to accumulate imaging probes at tumor cells may be found by exploiting the up-regulation of trans-membrane transporting systems. [1,2] In fact, rapidly growing tumors require an increased and continuous supply of aminoacids and other nutrients. In this contest, glutamine appears an interesting candidate as it is considered the main source of nitrogen for tumor cells. Several DTPA and DOTA ligands conjugated to glutamine residues have been synthesized and their complexes with Gd(III) and 111-In(III) have been evaluated for MRI and SPECT, respectively. In spite of the reduction of the electric charge some of the investigated systems showed a good uptake by different tumor cell lines. The systems that showed the best performance in “in vitro” cellular uptake experiments have been tested on tumor bearing mice by MRI and SPECT modalities.
Abstract ID: 550
Semiconductor quantum dots (QDs) represent a new class of fluorescent labels with unique advantages, including high fluorescence intensity and photo-stability, color tunability, and simultaneous excitation of all QDs with a single wavelength, allowing for multiplexed optical imaging. Here we report the delivery of QD bioconjugates as both optical and electron-dense markers for coordinated light and electron microscopy imaging of biomolecules in microbial cells. With the recent advent in proteomics, there is a pressing need to characterize multi-protein complexes in microbial cells, and to visualize and track their assembly and disassembly. The first step in developing QD probes for optical imaging of protein complexes in live cells is to deliver the probes into microbial cells, a task that is distinct from probe delivery into mammalian cells due to the extra barrier provided by the cell wall of yeast and the outer membrane of bacteria. In this study, we have compared several delivery approaches in S. cerevisiae and E. coli, including peptide-mediated delivery, heat shock and toxin-based membrane permeation. Specifically, the TAT peptide, a polyarginine peptide and the sequence YKKSNNPSFD, shown to deliver GFP into both S. cerevisiae and E. coli, were tested as delivery peptides. QDs with streptavidin conjugated to the surface were used to attach biotinylated peptides. These QD probes were delivered into S. cerevisiae with or without heat shock. Further, for delivery of QDs into E. coli, the outer membrane-permeating toxin Polymyxin B nonapeptide was investigated. We found that heat shock improved the uptake of the QDs into both cell types. The effect of heat shock on the uptake of Streptavidin-coated QDs with and without peptide and the utility of various concentrations of Polymyxin B nonapeptide were examined.
Abstract ID: 551
In the past few years, a plethora of “ 99m Tc biocomplexes” have been investigated as radioactive probes for monitoring in vivo biochemical alterations underlying different diseases. It will be highly desirable to have the ability of following the trafficking and fate of the compounds at the cellular and subcellular level. Unfortunately, this is not achievable with the radioactive probes by themselves, and requires the use of other imaging modalities, like fluorescence microscopy. Due to the similarities between the chemistry of Re and Tc, fluorescent rhenium complexes emerged very recently as good candidates to bridge the gap between in vitro fluorescence imaging studies and in vivo radioimaging with 99mTc.[1]
Within our interest on Re and Tc complexes for biomedical applications, we have introduced bifunctional pyrazole-diamine ligands that act as powerful tridentate chelators towards the fac-[M(CO)3]+ (M=Re, 99mTc) moieties. These chelators lead to tricarbonyl complexes displaying excellent in vitro and in vivo stabilities, and are easily derivatiseable with bioactive vectors and/or fluorescent chromophores.[2] In this communication, we will report on the synthesis and characterization of novel Re(I) and 99mTc(I) tricarbonyl complexes anchored by pyrazole-diamine chelators bearing anthracenyl chromophores, as part of our ongoing efforts to introduce isostructural fluorescence (Re) and radioactive (99mTc) probes with interest for in vitro or in vivo imaging applications.
Abstract ID: 553
The development of a high relaxivity Gd(III) MRI contrast agent involves the fine-tuning of several determining parameters such as an optimal water exchange rate and a slow overall rotation of the thermodynamically stable Gd(III) complex. With this criteria in mind we have designed a novel heterotritopic ligand comprising a 2,2′-bipyridine moiety for specific Fe(II) binding and two poly(aminocarboxylate) groups for Gd(III) coordination [1]. The ligand self-assembles into a rigid metallostar bearing 6 paramagnetic Gd(III) around its core. The stability constants of the Gd(III) complexes formed with the ligand and the metallostar, an important factor for in vivo safety, as well as the protonation constants of the ligand, were determined by potentiometry. 17O NMR and NMRD studies performed show that a water exchange rate is close to the optimal values and that the presence of two inner-sphere water molecules double the inner-sphere contribution to the relaxivity. Analysis of the longitudinal 17O and 1H relaxation rate data according to the Lipari-Szabo approach evidences the presence of internal flexibility.
Abstract ID: 554
The electronic spin relaxation of Gd(III)-based MRI contrast agents is one of the most important factors which determines the enhancement of water proton relaxation (relaxivity) and is currently the subject of both experimental and theoretical studies.
In the aim determining the influence of the electron relaxation on the relaxivity, multiple frequency and temperature EPR measurements on several Gd-complexes in solution have been performed.
A multifrequency EPR approach allows more accurate determination of the field-dependent (Zeeman contribution) and the field-independent components (the Static and the Transient Zero Field Splitting) of the electron spin Hamiltonian. In liquid solution we can only observe the effect of the ZFS parameters on relaxation. Thanks to the lack of molecular motions in frozen solutions, it is now possible to directly measure both the static and the transient ZFS.
Here we report the study by High Field EPR (240GHz) of glasses of Gd(III) complexes at several temperatures.
For the first time with that type of compounds, we are able to determine from these EPR data the average ZFS parameters and their distributions. Since the ZFS is related to the crystal field, our results constitute a first step towards establishing a relationship between molecular structure and the ZFS.
Abstract ID: 555
Phosphorus analogues of DOTA are known to have interesting properties for the usage as potential MRI contrast agents. Their residence time of water in the inner sphere ( T M) of Gd(III) complexes is much smaller than for [Gd(DOTA)(H2O)]- and the presence of second sphere water molecules enhances the overall relaxivity.[1] We report the syntheses of two new phosphorus containing ligands and their lanthanide(III) complexes. Multinuclear NMR studies followed by simultaneous fitting show that T M is smaller than that of [Gd(DOTA)(H2O)]-. The relaxivities of complexes [Gd(L1)(H2O)] and [Gd(L2)(H2O)]- are equal to 3.4 and 3.5 mM-1s-1, respectively (20 MHz, 37 °C, pH 6.5). The phosphonic function allows further functionalization of the ligands and the synthesis of targeting MRI contrast agents.
Abstract ID: 556
2. Reviewing current PDT regarding its primary targets and clinical applications.
3. Deducing targetability mechanisms shared by NACAs and PDT for new opportunities in molecular imaging and medicine.
2. PDT evolves with these trends [2–4]: 1) from 1 day to less than 1 hour of drug-to-light interval for higher drug concentrations; 2) from parenchyma to stroma for intratumoral distribution; 3) from systemic to local drug administration; and 4) from malignancies to benign disorders for therapeutic indications, all suggesting a lack of real “tumor” selectivity and supporting the concept of “NACAs” [1].
3. All NACAs and most PDT agents appear to share the same targets, i.e. nonviable components including denatured tissue debris, collagen, fibrin, proteoglycans, elastin, etc. Stromal cells including macrophages, fibroblasts, monocytes, mast cells and neutrophils likely participate in the secondary intracellular uptake of bound products, suggesting natural cell labeling for potential diagnostic and therapeutic utilities.
Abstract ID: 557
The rate of water exchange is one of the most significant parameters to determine proton relaxivity for GdIII complexes. Eight coordinate poly(amino carboxylate) ligands are particularly preferred for complexation of GdIII, since they are capable of ensuring sufficient thermodynamic and kinetic stability, primordial for the in vivo non-toxicity of the contrast agent. The GdIII complexes of such ligands are generally monohydrated, and undergo a dissociative water exchange process.
In order to attain maximum proton relaxivities - provided the rotation is also optimized, the value of the water exchange rate has to be optimal, i.e. around 5 × 107 s−1. We have shown recently that by increasing the steric crowding around the bound water site, one can remarkably increase the water exchange rate on both acyclic, DTPA-type, and macrocyclic, DOTA-type GdIII complexes [1,2]. Here we present examples of modulation of the steric compression in the inner sphere of the complex by minor structural changes of the ligand which induces then small variation of the water exchange rate.
Abstract ID: 558
The phosphorus analogue of DTPA, promising fast water exchange, was conjugated with PAMAM G5 dendrimer via benzyl-thiourea linkage.
The proton relaxivity of gadolinium(III) complex G5-[Gd(L)]58 measured at 37 °C, 20 MHz and pH 6 was 26.8 mM-1s-1. The relaxivity decreases with increasing temperature, therefore slow water exchange is not limiting the relaxivity. [1] The phosphinic function allows second sphere water molecules in the system which enhance the overall relaxivity. The 17O and 1H relaxation rates have been analysed using Lipari-Szabo model, modified for the presence of second sphere water molecules. Free NH2 groups give the possibility to introduce a targeting group into the molecule, which would increase the specificity of this potential MRI contrast agent.
Abstract ID: 559
Random peptide bacteriophage (phage) display libraries can be selected for clones that bind specific tissues, tumors, and antigens. However, once the targeting-peptide is synthetically produced, it often displays lower affinity compared to the original phage clone due to lack of avidity effects and removal from the virion surface. Using the selected phage directly for imaging would be advantageous due to the increased avidity for the target.
Abstract ID: 560
The purpose of this study is to use αvβ3 targeted molecular imaging to determine the optimum therapeutic window for vascular targeted non-viral gene therapy using the same integrin targeted nanoparticles (NP) platform.
Male Fisher Rats were divided into 3 groups. Treated and control groups (Groups 1 and 2) received RT2 tumor cells implanted intracranially into the striatum. Normal group (Group 3) received no tumor cells or treatment. Group 1 received an αvβ3 targeted NP that carried a mutated Raf gene (NP-ATPμ-Raf) complex. This gene disrupts the VEGF and FGF mediated signaling pathway.
The optimum treatment window was determined by molecular imaging using MRI and gamma imaging. Additionally, T2-weighted MR imaging was used to determine the size of tumors. Both the treatment group and the control group demonstrated 3 mm tumors prior to the start of treatment. The tumor in the controls continued to increase in size and signal intensity. Most of the controls died between days 10 and 21. The tumors in the treated group showed significant reduction in tumor size by day 24. These tumors demonstrated signs of encephalomalacia and significant reduction in vessel density by histologic analysis. 80% of all treated animals are living, while 100% of untreated tumor-bearing animals died. Dual modality targeted probes are a powerful way of separating patient population based on treatment responders vs. non-responders through molecular imaging.
Abstract ID: 561
In the second study, similar infusions of Gd3-N@C80 and Gd-DTPA were performed in two additional groups of five rats. Using MR, T1 maps were acquired immediately following infusion, and at post-infusion days 3, 6, 11, 17 and 25. Using changes in T1 relaxation times to determine infusate concentration, the distributions of each agent were analyzed over time.
Abstract ID: 562
By non-invasive 19F MRI we selectively imaged spatial and temporal distribution biodistribution patterns of the molecule fluor contained in perfluorocarbon (PFC). Aerosolized PFC was used as a marker to study pulmonary gas exchange and lung mechanics in a surfactant depleted piglet model. Sixty minutes after induction of lung injury by bronchoalveolar lavage, the piglets received aerosolized PFC (10 ml/kg/h for 2h). After PFC-treatment ventilation was continued for 6 hours. From previous PFC measurements it is known that sixty minutes after therapy onset, PaO2 was significantly higher and PaCO2 was significantly lower. Using 19F MRI we imaged the distribution of PFC in the lung before, during, and after the treatment. The temporal activity changes over time were analysed. First 19F signal was measurable after 30 min therapy onset. Initially the signal was detectable in the apical part of the lung where the concentration of PFC increased fast to near maximal levels and remained at this high value as long as the therapy lasted. As could be confirmed in all cutting directions, the PFC appeared later in the medial and after that in the distal parts of the lung. The time profiles show a slower increase in the signal intensity and the more distal the less maximal end concentration was reached. After end of therapy the decrease of PFC concentration was very fast but nearly similar in all regions and reached baseline level after 30 min. After an extrapolation of PFC amount below the level given by MRI sensitivity, we can predict much longer persistence of PFC in the lungs. The kinetics of the lung filling by PFC, taking together the results from all cutting directions is from apical dorsal to distal dorsal over the whole anterior-posterior extend of the lung. Afterwards, the lungs are filled from dorsal to ventral directions.
Abstract ID: 563
Multiple myeloma is a differentiated B-cell neoplasm characterized by the clonal proliferation of plasma cells that accumulate in the bone marrow, leading to severe osteoporosis and bone destruction. In order to evaluate new therapies, several preclinical models have been described using hind leg paralysis as a read-out for disease progression. Since quantification of hind leg paralysis is rather subjective, we evaluated NIRF for monitoring the homing of myeloma cells to the target tissue. For this, we injected 5 × 10e6 H929 cells i.v. into SCID beige mice and cells were identified by using a Cy5.5 labeled monoclonal antibody against the human IGF1 receptor. Four weeks and 3 months after cell application, 50 microgramm of antibody was injected i.v. and at different time points thereafter (5 hrs, 24 hrs and 48 hrs) NIRF was measured using the BonSAI imaging system from Siemens Medizintechnik, Germany equipped with an inhalation mask for anesthesia. Maximum acquisition time of 5 seconds was sufficient to identify the myeloma cells. NIRF signal was located in hind legs and after 3 months NIRF signal intensity was dramatically increased, whereas hind leg paralysis was not manifest at this time point. Injection of the labeled antibody in tumor free mice gave no NIRF signal. Human DNA was quantified by Alu-PCR of explanted hind leg tissue and a strong correlation of NIRF signal and of Alu signal was observed. Thus, NIRF is a reliable method to monitor tumor progression in vivo. Since manifestation of osteolytic lesions in this preclinical model is variable, staging of animals can be accomplished before starting of treatment making this method an excellent tool to evaluate the therapeutic efficacy of new drugs against multiple myeloma.
Abstract ID: 564
Abstract ID: 565
Abstract ID: 566
Dendrimers are well-defined, hyperbranched polymers with a high density of functional groups that have been applied as contrast agents and inhibitors of cell-cell interactions. Multivalent conjugation of peptides or recombinant proteins to a dendritic scaffold would yield ligands that bind to multivalent targets with increased affinity and specificity. However, no method is currently available that allows conjugation of dendrimers with both oligopeptides and recombinant proteins in a chemoselective manner that would not interfere with their biological activity. Here we report such a general synthetic strategy by using native chemical ligation and cysteine-functionalized dendrimers. Native chemical ligation, the reaction between a thioester-terminated peptide or protein and an N-terminal cysteine, is highly specific and affords a native peptide bond in high yields (Figure 1). Poly (propylene imine) dendrimers of generations 1–3 were functionalized with cysteine residues. Reaction of the first generation cysteine dendrimer with excess thioester-terminated green fluorescent protein (GFP) yielded a branched protein tetramer (Figure 2). The same methodology was used to generate peptide dendrimers of generation 1–3 (4–16 peptides) and dendrimers displaying a combination of peptides and GFP. Dendrimers can also be functionalized with exactly one copy of a targetting protein, while using the remaining cysteines to conjugate spectroscopic probes. The modular approach presented here provides access to a wide variety of well-defined multivalent peptides and proteins that are attractive both for understanding the fundamental mechanisms of multivalency in biological interactions and for biomedical applications in molecular imaging and targeted drug delivery.
Native chemical ligation of multiple peptides and proteins to cysteine dendrimers.
A multivalent protein-dendrimer conjugate obtained via native chemical ligation of GFP to a 1st generation cysteine dendrimer.
Abstract ID: 567
Abstract ID: 568
Luminescent lanthanide containing compounds find niche applications as fluorescent labels in biological imaging and dopants for fiber amplifiers, lasers and optoelectronic devices [1]. The longevity and internal quantum efficiencies of the excited states can be enhanced by the employment of multidentate chelating ligands that satisfy the coordination preference of a given ion, exclude inner sphere solvent molecules and limit detrimental vibrational quenching mechanisms. Here, we describe the synthesis of a series of lanthanide complexes of the heptadentate ligand HXTA (N,N′-(2-hydroxy-1,3-xyrylene)-bis-(N-carboxymethylglycine)) [2]. Structural studies show that HXTA forms 2:2 complexes with all the lanthanide ions; no solvent molecules are coordinated in both solution and the solid-state yielding complexes that exhibit long luminescent lifetimes and high quantum yields. Additionally, we will describe the synthesis of two related ligands incorporating the HXTA building block N,N′N″,N″N-(2-hydroxy-1,3-4,4′-biphenol)-tetra-methylene(N-carboxymethylglycine) (HBTA) and N,N′N″,N″-(2-hydroxy-1,3-bisphenol A)-tetra-methylene(N-carboxymethylglycine) (HSTA). The potential of these ligands to form extended metal containing luminescent arrays with controllable size and shape for possible functional materials will be investigated.
Abstract ID: 569
Abstract ID: 570
Evaluation of radiation and chemotherapy efficacy for brain cancer is traditionally accomplished by measuring changes in tumor size over time, after therapy has been administered. Brain tumor therapy could be more efficiently administered if a single parameter or a combination of parameters could be identified that would indicate early on whether a particular therapy is efficacious. Diffusion MRI, FDG-PET and methionine-PET (met-PET) have all proven their potential to individually act as imaging biomarkers of tumor therapeutic efficacy long before any apparent volumetric decline, thereby identifying particular tumors as responders to therapy. While each parameter has its own strengths and weaknesses depending on tumor type, the combined evaluation of these parameters has the potential to illustrate more clearly the underlying tumor physiology at any given point in time allowing for a more precise assessment of therapeutic efficacy. Using an orthotopic rat 9L glioma model, we have co-registered diffusion MRI, FDG-PET and met-PET image acquisitions at baseline, day 3 and day 7 post 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) therapy against the anatomical T2-weighted MR image, thereby allowing us to more fully and reliably appreciate the early therapeutic-induced changes in solid tumor.
Abstract ID: 571
A productive immune response requires the contribution of various cell types of the immune system, including T cells, B cells and antigen-presenting cells. We are developing a system to visualize the role of CD8+ and CD4+ T cells separately during an anti-tumor immune response, in the context of an intact immune system. We have generated lentivirus constructs where either Firefly or Renilla luciferase is expressed behind the global human ubiquitin promoter or the murine CD8a promoter and enhancer, which are expressed in murine CD8+ T lymphocytes. Bone marrow is harvested from 5-fluorouracil treated C57BL/6 mice, infected in vitro with lentiviruses, and injected into lethally irradiated syngeneic C57BL/6 recipients. Optical imaging of mice transduced with bone marrow expressing Firefly luciferase behind the ubiquitin promoter shows that the luciferase signal is detected in the skull and spine beginning at 6 days after bone marrow transfer. The signal is amplified by day 14 and is also detectible in the femur, humerus and ribcage. By day 28, luciferase activity is detected in areas corresponding to the thymus and spleen. In contrast, animals transduced with CD8α-driven Renilla luciferase display a signal beginning only at day 17 in the area of the spleen. Reconstitution of the intact immune system where the CD8+ T cells are marked with the luciferase gene will allow us to track the localization of cytolytic T cells during a primary and secondary immune response in vivo using bioluminescent optical imaging techniques.
Abstract ID: 572
An activating mutation in the protein kinase B-Raf is known to occur in more than 60% of melanomas. In order to better understand the role of B-Raf in tumor initiation and progression, we have developed and characterized human melanoma cell lines in which endogenous B-Raf expression can be regulated in vivo using a tetracycline-inducible short-hairpin RNA (shRNA) system. We utilized A375M cells, selected in vivo for high metastatic potential to evaluate the role of B-Raf signaling in metastatic melanoma. Furthermore, these A375M cells were transfected with firefly luciferase to enable noninvasive monitoring of tumor progression in an appropriate mouse model using in vivo bioluminescence imaging (BLI). This same cell line was then further engineered to stably express B-Raf-shRNA to enable doxycycline-regulatable knockdown of B-Raf protein and downstream signaling. To evaluate the efficacy of B-Raf knockdown in this systemic model, cohorts of mice with similarly increasing tumor burden, as determined by BLI, were divided into treatment groups. Mice were monitored longitudinally via BLI for tumor progression and response to B-Raf knockdown in addition to monitoring survival. In this melanoma metastasis model, shRNA-mediated B-Raf inhibition significantly slowed tumor growth and prolonged survival. When induction of knockdown was delayed until systemic tumors were established, progression of disease was still inhibited as shown by bioluminescence quantification of pulmonary tumor burden. In summary, this work demonstrates the utility of BLI imaging and the possibility of targeting B-Raf in order to inhibit certain metastatic tumors. In addition, we demonstrate that the A375M-luc/shRNA model can be used to evaluate the efficacy of novel B-Raf inhibiting agents in development as cancer therapeutics.
Abstract ID: 573
Microtubules are an important element of the cellular cytoskeletal structure. In the case of neurons, microtubules interact with several microtubule-associated proteins (MAPs) to guide the formation and growth of processes. Tau is a MAP that normally is enriched in axons where it is thought to regulate microtubule dynamics. In order to measure tau dynamics in living cells, a fusion construct between human tau and photoactivatable green fluorescent protein (PA-GFP) (Patterson & Lippincott-Schwartz, 2002) was prepared. The construct was transfected into neural cells (PC12) and clonal lines stably expressing the fusion protein isolated. Tubulin-tau colocalization experiments confirmed binding of the fusion protein to microtubules in the cells. Photoactivation was carried out by irradiating a 7–9 μm long segment of individual processes of NGF-differentiated cells with a 405 nm blue diode using a laser scanning microscope equipped with an incubation chamber. GFP-fluorescence was recorded in 3 second-delay time series with a 488 nm laser over a period of 15–20 minutes. Pixel analysis of individual frames was performed and intensity versus distance profiles over time determined. We observed that in the middle portion of processes a subpopulation of PA-GFP-tau quickly redistributes after activation (< 60s) while a significant portion remains stationary over a long time (up to 10 min). At the tip of processes, PA-GFP-tau had a higher mobility compared to tau located more proximally and redistributed to the shaft. The results indicate that tau dynamics are substantially different in distal and proximal processes and suggest that different subpopulations of tau exist that differ in their intracellular interactions. Photoactivation of cytoskeletal proteins may provide a useful tool to analyze cytoskeletal dynamics in living cells.
Abstract ID: 574
The homing and outgrowth of luciferase gene-transduced hematopoietic cells can be visualized in live animals on sequential time points by bioluminescent imaging (BLI), using a highly sensitive liquid nitrogen cooled charge-coupled camera (CCCD).
A safe transduction of bone marrow (BM) hematopoietic cells was optimised for the retroviral vector encoding the Green Fluorescent Protein-luciferase (GFP-Luc) fusion gene. To validate the signal of the BLI in relation to the number of transduced hematopoietic cells, different cell doses were transplanted. There is a linear correlation of the number of cells and the bioluminescence signal in the BM compartment during the first 5 weeks after transplantation. However, after a longer period of time the variation increased between the individual mice.
Studying the fate of different transduced murine HSC populations after transplantation into lethally irradiated mice, not-treated BM was compared to Sca-1 positive cells from 5Flurouracil-BM, a technique to enrich for the primitive hematopoietic stem cell. After transplantation of the total cell population with 20% transduced cell, different foci in the BM showed luciferase activity, predominantly in the femurs and sternum. Luciferase activity in mice transplanted with transduced BM cells decreases below detection level after 6 weeks, suggesting that only committed progenitors were transduced in this cell sample. Mice transplanted with transduced Sca-1 positive cells reached a maximum level of luciferase expression at week 4–5 and thereafter a consistent signal during the 7 months, indicative for the activity of the transduced primitive stem cells. The transduction of primitive stem cells was confirmed by a secondary transplantation in which long-term expression of luciferase was observed.
These results show that the BLI might be of value to study different populations of hematopoietic stem cells or for monitoring and quantitate the proliferation of locally active hematopoietic cells.
Abstract ID: 575
There is increasing interest in a reliable in-vivo non-invasive method to monitor migration of endogenous neural stem cells with the intention to monitor (experimentally) upregulated neuronal recruitment as a therapeutic strategy for neurodegenerative diseases. Shapiro et al1 labeled subventricular neuronal precursor cells in situ in the rat brain with micron-sized superparamagnetic iron-oxide particles (SPIOs) and visualized their recruitment along the rostral migration stream (RMS) up to the olfactory bulb (OB) with magnetic resonance imaging (MRI).
The goal of our study was to test different SPIOs in a related protocol to visualize neuronal recruitment in healthy adult mice aiming at future applications in mouse models for neurodegenerative diseases.
Five different SPIOs - a commercial μm-SPIO product and four newly developed nano-sized magnetoliposomes (MLs) with different surface coating (neutral, anionic, cationic and PEGylated)2 but equal Fe-concentration (0.67 mgFe/ml) - were stereotactically injected (1.5μl) in the left RMS of C57BL mice (n=10). At four different time points: 5, 13, 21 and 60 days post injection, the distribution of the contrast agents in the mice brain was monitored in vivo with T2* weighted gradient-echo MRI. Except for the cationic MLs (0/2) and the μm-SPIOs (0/2), the different MLs revealed a clear migration into the OB (6/6) from 21 days onwards (Figure 1A). Two months post injection the mice were perfused with formaldehyde for high resolution in-vitro 3D-FLASH neuroMRI (Figure 1B) and histology. Appropriate histology is in progress to confirm uptake of the SPIO's and recruitment of the contrast loaded neuronal progenitor cells. The obtained results suggest that particle size and surface coating of the contrast agents have an influence on their in situ cell labeling capacity.
Olfactory bulb of an adult mouse with anionic MLs injected in the RMS: A) in vivo MRI 21 days post injection (156 μm in plane resolution, 1mm slice thickness). B) in vitro MRI 60 days after injection (66 micron isotropic resolution). Arrows indicate the contrast visibility in the RMS, the dark vertical line in A is due to the injection.
Abstract ID: 576
Osteoarthritis is a degenerative joint disease characterized by destruction of articular cartilage, subchondral bone erosion, and associated joint pain. The mono-iodoacetate (MIA) model promotes loss of articular cartilage similar to that seen in humans. In this study, 1 mg of MIA in 50 μl of saline was injected into the left knee joint of 8-week-old male Lewis rats (n=6) while 50 μl of saline was injected into the right knee as a contralateral control. A control group (n=6) had saline injected in both knees. Both knees were imaged by micro-computed tomography μCT, 47 μm isotropic voxels) at baseline, and at 1, 2, 3, and 4 weeks post injection. The entire subchondral bone region of the tibia was segmented for structural analysis. A comprehensive analysis including bone-mineral density (BMD), structure model index (SMI), anisotropy, connectivity by Euler number calculation, structural parameters from the plate model, and a topological analysis was performed. Highly significant changes in all parameters were observed due to MIA injection relative both to the contralateral control and to the control group as early as 1 week post injection. This model combined with high-resolution μCT has the potential to provide a facile method for the in-vivo screening of potential OA disease-modifying drugs that act through mechanisms influencing subchondral bone turnover.
Abstract ID: 577
The neural substrate for song behaviour in songbirds, the song control system, is the best documented brain circuit to study neuroplasticity. Not only the volume of the key song control nuclei (SCN) HVC, RA and X change in size, but also the density of the connections between them changes as a function of seasonal and hormonal influences. This study uses repeated in vivo Diffusion-Tensor MRI (DTI) and the resulting parameters (FA=Fractional Anisotroptry, MD=Mean Diffusion, Λr=Radial diffusity, Λ//=Axial diffusity) to quantify seasonal changes in the connections between different SCN in starling brains.
Nine male starlings (Sturnus vulgaris) were measured repeatedly in spring and in summer. The starlings were sedated with an FM injection of xylazine and ketamine. DTI was performed on a 7T-system (MRRS) using multislice DW-SE and diffusion sensitizing gradients along 7 directions. Sagittal slices with image resolution of 100*100*400μm3 were obtained covering one hemisphere of the brain.
FA (0.17±0.02 vs 0.22±0.03, p-value:<0.001) and Λ// (6.67±0.29 vs 7.04±0.23, p-value:0.004) values in the HVC-to-RA-Tract (1) were significantly increased in spring and reflect the neuronal sprouting from RA-projecting HVC-neurons well known from literature. The laminae harbouring other SCN connections such as LMAN-RA (2) (0.21±0.02 vs 0.24±0.02, p-value:0.02) and HVC-X (3) (0.19±0.009 vs 0.20±0.01, p-value:<0.05) also showed a significant seasonal difference in FA-values while no seasonal changes were discerned in the lamina which doesn't harbour SCN connections (4).
The different DTI-parameters FA, Radial (Λr) and Axial (Λ//) diffusity clearly allowed discriminating seasonal changes in the amount of axonal projections in the songbird brain. The songbird brain offers the exclusive opportunity to validate new (MRI) imaging tools for their capacity to study different features accompanying neuronal plasticity and neurogenesis.
Figure shows two sagittal FA maps of the same starling brains obtained during spring (A) and summer (B). The arcopallium (enlarged in the inset) reveals the Robust Nucleus of the Arcopallium (RA) as the dark core surrounded by white fiber layers and the tract between HVC and RA (1). Other white fiber layers are Lamina Mesopallialis (2), Lamina Palio-Subpallialis (3), Lamina Frotalis Superior (4).
Abstract ID: 578
Abstract ID: 579
Thrombosis is the unwanted formation of a clot within the blood vessels or heart, which is associated with a number of diseases such as pulmonary embolism, stroke, and myocardial infarction. There is a need for a high resolution, non-invasive bright spot imaging modality to identify clot and assess thrombosis. Our thrombus-imaging program has focused on development of a fibrin-targeted MRI contrast agent.
For this purpose, a class of fibrin-binding peptides (Tn7: X2CXYYGTCX2) was discovered via phage display. An initial medicinal chemistry screen led to the identification of two lead peptides within the Tn7 family with sufficient fibrin affinity and selectivity to fibrin vs. fibrinogen. This paper describes the design and synthesis of these novel peptides conjugated in different ways to gadolinium chelates. The conjugates were evaluated with respect to fibrin binding, relaxivity, and in vitro stability. The output of these studies was a prototype fibrin-targeted MRI contrast agent with excellent relaxivity and in vitro stability.
Abstract ID: 580
[11C]DAA1106 (
Both methods were implemented on automated systems yielding 2.6 GBq (method A), 6.3 GBq (method B) both at high purities (>98%) and good specific radioactivity of 225 GBq/μlmol.
For a preliminary evaluation of [11C]DAA1106 a microPET study was then performed on a SHR rat to compare it's brain kinetics with the widely used PBR radioligand [11C]PK11195. Time-activity curves (Figure 2) shows a rapid uptake of
Abstract ID: 581
The intercellular adhesion molecular-1 (ICAM-1) expressed by endothelial cells plays an important role in inflammatory processes. We designed a molecular MR contrast agent for inflammation-specific imaging in vivo. The anti-ICAM-1 antibody purified from the cultured hybridoma was conjugated with Gd-DTPA. The Gd-DTPA-anti-ICAM-1 antibody displayed increased T1 relaxivity that is four times higher than that of Gd-DTPA and showed predominant binding to cultured endothelial cells where ICAM-1 expression was stimulated. The Gd-DTPA-anti-ICAM-1 antibody (dose of 10μmol/kg of Gd) was injected intravenously in 10 mice where brain abscess was induced with the staphylococcus aureus. Standard T1 weighted spin echo images were acquired with 1.5T clinical MR unit at twenty-four hours after injection of contrast agent. Serial MRI showed that on 1~2 days after inoculation, diffuse increased signal intensity was observed in the inoculated hemisphere and high signal intensity was found in all around of abscess core from 4 days after inoculation. No significant increase in signal intensity was found in the control group injected with the control rat IgG-conjugated Gd-DTPA. The immunohistochemical analysis confirmed that enhancing area in MRI correlate with ICAM-1 expression. These results suggest that the Gd-DTPA-anti-ICAM-1 antibody can be used as MRI contrast agents to visualize ICAM-1 expressed in brain inflammatory disease.
Abstract ID: 582
About 70 million Americans suffer from one or more types of cardiovascular disease, which are the major, primary or contributing causes of death in the United States (58% of all deaths). Current methods for imaging disease are geared towards identification of advanced disease with significant stenosis and indirectly image the lumen, rather than the lesion. We describe development of targeted contrast agents that will enable in vivo, combined PET (Positron Emission Tomography) and MR (Magnetic Resonance) imaging of early atherosclerosis and restenosis. We propose to use the high sensitivity of PET to first locate regions of potential blockages of the blood vessel, which then can be probed at higher resolution by using MRI. Interest in multimodality imaging has surged in recent years and multimodality probes will play a pivotal role in clinical molecular imaging of the future.
The contrast agent is targeted to macrophages, one of the earliest cellular components of developing plaques. The agent is based on a maleylated bovine serum albumin (mal-BSA) with a conjugated 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) ligand system, that can tether both paramagnetic Gd3+and a positron emitter, e.g. 64Cu2+. Conjugation of 18 Gd-DOTA groups per molecule resulting in relaxivity of 31 mM−1s−1which is comparable to literature values for similarly substituted proteins. Incorporation of ≥0.1 μCi/ml 64Cu2+ shows sufficient activity for imaging by PET. Experiments in cultured cells demonstrate receptor-mediated uptake of the probes by macrophages, and show measurable contrast in MR image. The agent is rapidly cleared from the blood. Biodistribution studies in the rat show clearance of the contrast agent by kidney and liver which decreases to low levels by 24 hours. This novel MRI contrast agent has been combined with features for plaque lesion detection using PET, allowing the application of only one contrast agent for imaging using both modalities.
Late Breaking Poster Session, Part 2.
Abstract ID: 583
In vivo imaging provides a means for non-invasive longitudinal monitoring of biological processes in live animals. The application of these new technologies on engineered animal models offers potential opportunities for studying cancer as they develop in a natural microenvironment and to provide rapid assessments on their responses to various therapeutic interventions in live animals. We have generated a mouse model by gene targeting in the mouse embryonic stem cells in which the expression of a HSV1-tk-IRES-Luc chimeric gene was dirven by an endogenous mouse alpha fetoprotein (Afp) promoter. Since this Afp promoter is highly specific for the hepatocytes in developing fetal liver and hepatoma, the expressions of the gene products, tk and Luc, respectively, of this chimeric gene provide excellent indicators for hepatomas in adult mice. Because tk and Luc can be monitored by in vivo imaging through PET and BLI, respectively, these mice could provide a model for studying hepatoma by live imaging. Preliminary data confirm that the engineered chimeric gene exhibit a typical Afp expression profile as expected. BLI imaging of one-day-old mice revealed a highly liver-specific signal, confirming that this novel mouse model can indeed be used for in vivo imaging. Experiments are under way to investigate the feasibility of using this model for studying hepatoma development by Luc-based BLI and HSV1-tk-based PET imaging. This model was designed for demonstrating the principle of using reporter genes driven by well-characterized cancer-specific transcriptional units in conjunction with in vivo imaging as a paradigm in studying naturally occurring cancer in live animals. It may also be used for early identification of cancer-bearing animals, for monitoring changes of the disease in response to specific interventions, etc. Furthermore, this paradigm, once established, should be applicable to the studies of other types of cancer in which highly cancer-specific transcriptional control units are available.
Abstract ID: 584
Two APD-based, PET modules have been evaluated for use in a combined PET/MR system for simultaneous imaging. Each module consists of 4 independent, optically isolated detectors. Each detector consists of an 8 × 8 array of 2 mm × 2 mm × 20 mm LSO crystals read out by a 2 × 2 array of 5 mm × 5 mm Hamamatsu S8664–55 APDs. The APD signals were amplified by two 10-channel, charge-sensitive preamplifiers (per module) and processed by Pico3D electronics.
The modules were positioned 36 cm apart in the bore of a 1.5 Tesla Siemens Symphony MR scanner. Simultaneous PET/MR images were acquired using a small Derenzo phantom filled with water and 1.25 g NiSO4/ litre and approximately 50 MBq of FDG. 3D images were acquired by rotating the phantom in a “step-and-shoot” technique. There were no artefacts observed in either the PET or MR images. The reconstructed PET images showed a spatial resolution of less than 2 mm. The PET detectors decreased the MR signal-to-noise ratio by 15% and the MR gradient and RF sequences decreased the PET sensitivity by 3–10% depending on the MR sequence. These results show that APD-based PET detectors can be used for a high-resolution and cost-effective integrated PET/MR system.
Abstract ID: 585
Abstract ID: 586
Abstract ID: 587
Early detection and differential diagnosis of tumors are always challenging in clinic. In this study, a cationic immunoliposome system that includes an anti-transferrin receptor single chain antibody fragment (TfRscFv) as the targeting molecule was used to encapsulate the MR contrast agent gadolinium (TfRscFv-Lip-GAD-d) for specific targeting to cancer cells and MR contrast enhancement. This system was evaluated for in vivo MR imaging in MDA-MB-435 breast cancer cells growing as solid tumor xenografts in athymic nude mice. The TfRscFv-Lip-GAD-d complex was administered intravenously at an encapsulated contrast agent dose of 0.2 mM/kg body weight. Following this, the T1-weighted MR images were acquired at a 12 min interval over 96 min. The signal intensity of tumors was significantly enhanced compared to free contrast agent and the enhancement was closely related to the pathology of the tumors. The proliferating tumor cells at the peripheral region showed the strongest enhancement and the necrotic region showed lesser enhancement. Furthermore, the liquefied amorphous necrotic area showed stronger signal than the coagulation necrotic area. Calcification showed no enhancement. The image intensity of the tumor returned to the pre-injection baseline level within 24 hours. Contrast enhancement with gadolinium alone was much lower with less resolution than that achieved using the targeted immunoliposome system. These results indicate that the newly developed TfRscFv-Lip-GAD-d system significantly enhances the image contrast in solid tumors and is much superior to the contrast agent alone for identifying the tumor pathological features. This targeted immunoliposome system may serve as a powerful MR imaging probe for early detection and differential diagnosis of tumors.
Abstract ID: 588
Bone is the most frequent site of metastasis of prostate cancer in men resulting in significant morbidity and mortality. Hormone refractory prostate cancer that has metastasised to bone is incurable. Development and validation of strategies for the prevention and treatment of cancer bone metastasis requires appropriate animal models. Currently, there are no adequate models of prostate cancer metastasis to bone. We have used the intracardiac injection technique combined with digital fluorescent imaging of GFP expressing tumours to select a bone seeking sub-line of the mouse prostate carcinoma cell line, RM-1. Plasmid transfection was used to create an RM-1 cell line expressing GFP and the hygromycin resistance gene as a selectable marker. These cells were injected into the left ventricle of anaesthetised mice, which were monitored daily for loss of condition and X-rayed weekly. All mice were killed when they began to show signs of distress. As intraosseous RM-1 tumours are largely non-lytic, they were difficult to detect by X-ray, but GFP fluorescence was easily visualised. Mice were carefully scanned with GFsPectacles and a digital fluorescent imaging system. Rare intra-osseous tumours were collected for in vitro culture and transfected RM-1 cells were selected by treatment with the selectable antibiotic, hygromycin. Successive passages through mice have resulted in bone-seeking RM-1 sub-lines with high affinity for bony sites and few non- bony metastases. These RM-1 sub-lines are generally less adherent in in vitro culture than the parental population. The resulting tumours have both lytic and sclerotic components, similar to human prostate cancer metastasis. The sub-lines provide an excellent model for developing therapies for treatment and prevention of prostate cancer bone metastasis and the continued expression of GFP allows for real-time monitoring of tumour response to treatments using in vivo imaging techniques.
Abstract ID: 589
We introduce a novel and efficient algorithm for reconstructing the 3D locations of tumor sites from a set of 2D bioluminescence images taken by the same camera but after continually rotating the object by a small angle. Our experimental set up is simpler and more flexible than those using multiple cameras, since we only need a single CCD camera that is readily available in commercial BLI imaging systems, and we can acquire any number of images by adjusting the rotation angle.
The image-based reconstruction is rooted in the stereoscopy algorithm in Computer Vision. The algorithmic steps include: (1) alignment of the BLI images so that the projections of the rotating axis on all images overlap, (2) finding of feature correspondences between consecutive images, (3) mapping of all BLI images onto their respective projection planes in the 3D space using an orthographic projection model, and (4) recovery of 3D depths of feature points and tumor sites by computing the intersection of 3D rays passing perpendicularly through corresponding points on the BLI images in 3D. The algorithmic steps in our framework are efficient and robust enough to facilitate its use in analyzing the repeated imaging of a same animal transplanted with gene marked cells.
In order to visualize in 3D the structure of the tumor, we also co-register the structure reconstructed from BLI with detailed anatomical structure extracted from high-resolution microCT on a single platform.
Experimental results on both phantom studies and real animal experiments show excellent reconstruction accuracy using our method.
3D depth recovery by 3D ray intersection. (A) determining 3D locations of the tube center and two feature points on the tube surface. (B) 3D location of the tumor center.
Object surface and tumor 3D location reconstruction result, (A) reconstructed 3D tube (gray surface) and tumor center (red sphere) superimposed on a projected BLI image. (B) Another view of the reconstruction result.
Abstract ID: 590
In order to acquire simultaneously nuclear medicine image and optical image, we expressed fusion protein containing sodium/iodide symporter (NIS) in N-terminal region and luciferase in C-terminal end and characterized NIS-luciferase fusion protein. A transgene vector comprising human NIS gene without termination codon and luciferase gene without start codon was constructed in pcDNA3.1 and named as pNIS-fLuc. pNIS-fLuc was transfected to human hepato-celluar carcinoma cell lines, SK-Hep1 using Lipofectamin and stably expressed cells were selected with G418 for 2 weeks (SK-Hep1/NIS-fLuc). To verify the expression of NIS and luciferase gene, I-125 uptake and luminescence of cell were measured. RT-PCR and Western blot with luciferase specific antibody were also performed. For in vivo imaging studies, tumor xenograft were generated in nude mice and imaged with an in vivo imaging system. Luciferase activity of SK-Hep1/NIS-fLuc increased to 2,000 times higher than that of SK-Hep1, but NIS activity did not increase. To confirm the NIS and luciferase gene expression, RT-PCR and Western blot were performed. NIS as well as luciferase of SK-Hep1/NIS-fLuc was expressed by RT-PCR and Western blot. Luciferase domain of fusion protein might be localized in cytosol by confocal microscopy with luciferase specific antibody. The bioluminescence images of SK-Hep1/NIS-fLuc tumor sites were clearly visible whereas the control tumor was not. In this study, luciferase activity of NIS-fLuc fusion protein was still active but NIS was not. This result suggests that NIS function in NIS-Luc fusion protein might be perturbed by luciferase domain in c-terminal region. We successfully acquired bioluminescence images using NIS-fLuc fusion protein.
Abstract ID: 591
The field of molecular imaging calls for imaging systems with both high resolution and high sensitivity. In response to such a demand, we developed a high-resolution high-sensitivity multiplexing multi-pinhole SPECT upgrade for clinical cameras. To date we have successfully upgraded several gamma cameras for both clinical and pre-clinical applications. In this work we focus on advanced applications of our multiplexing multi-pinhole imaging module. Specifically, we will present a comprehensive description of the multi-pinhole apertures used for full-body (A), focusing (C) and dynamic studies of mice and rats. In this contribution we report on the performance of three different clinical cameras upgraded with our multi-pinhole module: two dual-headed systems (Picker PRISM 2000 located at the University of Duesseldorf, Siemens ECAM located at the University of Marburg) and a triple-headed system (Trionix TRIAD located in the Research Center Juelich). We will present the results of high-resolution semiquantitative single-isotope mouse and rat studies imaged with a wide range of apertures according to application. We will also present the results of dual-isotope studies (B), absolute tracer quantification and the fusion of our multi-pinhole SPECT images with MR and CT data of the same animal. In addition to static imaging, we will display our first dynamic SPECT acquisitions using the three-headed system in Juelich. We designed 12-pinhole apertures dedicated to dynamic brain studies in rats with a time resolution of 1 minute and reconstructed resolutions as low as 1.7mm.
Results of three studies performed with our multiplexing multi-pinhole upgrade: (A) full-body tumor study in a mouse; (B) Tc99m-MDP and In111-Octreoscan full-body dual-isotope mouse study; (C) D2 receptor study performed with a focusing aperture.
Abstract ID: 592
Gene therapy offers new perspectives for treatment of renal diseases. In this study, RGD-modified adenoviral (RGD-Adv) renal gene therapy was compared between healthy and nephrotic rats by in vivo bioluminescence imaging (BLI).
Nephrotic syndrome was induced by adriamycin injection 6 weeks prior to BLI. Both healthy (n=10) and nephrotic rats (n=10) underwent selective RGD-Adv renal (left kidney) gene therapy. Upon cannulation of the renal artery after laparotomy, the kidney was flushed with saline, injected with RGD-adenovirus encoding GFP and luciferase (4.10E11 VP) and incubated in ice-cold saline for 1 h before reperfusion. In both groups, the right kidney served as an internal control. Renal damage and immunogenicity were quantified by the number of ED-1 positive cells (macrophages) and smooth muscle actin levels (SMA, fibroblasts). Furthermore, levels and localization of transgene luciferase/GFP expression were quantified. In vivo expression of luciferase was evaluated by imaging with an IVIS 100 system (Xenogen).
Markers of renal damage were unchanged in both transfected as non-transfected kidneys in both groups. Selective transfection resulted in a 30-fold increase in luciferase expression, but did not differ in level/localization between both groups. Double staining (GFP/SMA) showed mainly infection in interstitial fibroblasts. In nephrotic rats, BLI detected a much higher luciferase expression in the transfected kidney compared to the non-transfected kidney. Liver tropism of adenovirus, did not interfere with the observations.
The results of this study show that RGD-Adv is an efficient vector for gene therapy both in healthy and in nephrotic kidneys, predominantly targeting interstitial fibroblasts. In nephrotic kidneys this specificity becomes even more apparent as detected by both in vitro and in vivo BLI. The apparent specificity for fibroblasts may offer new perspectives for treatment of renal fibrosis.
Abstract ID: 593
Effective and easily prepared radioligands for imaging brain mGluR5 with PET would enable the involvement of these receptors in neuropsychiatric disorders to be explored in vivo. So far there is no such 18F-labeled radioligand, since the only effective radioligand ([18F]
A series of such ligands was synthesized from which
After [18F]
[18F]
Abstract ID: 594
Specific genotypes of human papillomavirus (HPV) are well correlated with cervical oncogenesis. The major transforming and immortalizing protein in high risk HPVs, namely HPV16, is E7 protein. E7 protein functions by deregulating the cell cycle and promoting S-phase reentry in differentiated keratinocytes. Currently, clinical diagnosis of cervical cancer is based on phenotypic changes observed in a screening Papanicolaou smear. Although screening has been effective in reducing the occurrence of cervical cancer, the low specificity of the Pap smear results in resources wasted on the evaluation of low-grade lesions not likely to progress to cervical cancer. Molecular characterization of active HPV infections using molecular specific contrast agents combined with in-vivo optical imaging is proposed to be a cost-effective, non-invasive technique for the detection of cervical pre-cancers. Contrast is achieved by exploiting the peak absorbance and scattering shift in aggregated gold nanoparticles over isolated ones and molecular specificity is achieved via recognition moieties with high affinities for E7. Conjugates of gold nanoparticles and HPV16 anti-E7 antibodies are delivered into the nucleus of living cells and imaged with reflectance confocal microscopy. These contrast agents have been used to successfully enhance contrast in HPV16+ cervical cancer cells over HPV- cells by a factor of 2.5. Further characterization and development of these contrast agents will provide a robust, low cost screening tool for the detection of cervical pre-cancers.
Abstract ID: 595
The purpose was to characterize a macromolecular contrast agent based on Gadolinium complexed in N-(2-hydroxypropyl)methacrylamide copolymers (Gd-HPMA). The dependency of the T1-relaxivity on molecular weight (MW) of the polymer and the stability of the gadolinium complex were investigated. Furthermore, micro-MR-angiographies of rat chest, brain and subcutaneous tumors were performed at 1.5T.
Gd-HPMA copolymers with 16–46kD MW were synthesized. In the polymers gadolinium was complexed in side chains formed by three aspartic acid residues extending four carboxylic groups. In a copolymer with 26kD MW tyrosinamide in side chains (0.3%) enabled radiolabeling. This was used for pharmacodynamic studies. Thermic stability of Gd-HPMA was investigated by incubating the copolymers in PBS for 1–25h at 60°C. Released gadolinium content was determined by mass spectrometry. Cell toxicity was assessed by incubating fibroblasts and hepatoma cells with Gd-HPMA (0.05/0.5mmol Gd/ml) for 24/48h. T1-relaxation rates (r1) of Gd-HPMA copolymers were determined with a turbo FLASH-sequence. MR-angiographies of tumor bearing rats were performed using a 3D-FLASH-sequence. Pharmacodynamic behaviour of 25kD Gd-HPMA labeled with 131Iodine was studied after intraveneous injection in 4 rats.
The Gd-complex proved to be stable. No release of gadolinium was observed after incubation of Gd-HPMA for 1–25h at 60°C. No increased death rates were observed in cells after incubation with Gd-HPMA. T1-relaxation rates of Gd-HPMA with different MW varied between 21–36mM−1s−1without a dependency of r1on MW. Pharmacodynamic studies showed a rapid elimination from blood (<50%ID after 15min) but delayed retention occurred in the reticuloendothelial system. In MR-angiography Gd-HPMA (25kD) was capable to visualize vessels down to <100μm in the rat brain, chest and tumors.
Gd-HPMA showed potential as biocompatible macromolecular MR-contrast agent. It is suited for MR-angiography and the possibility to vary MW without changing r1recommends it also for DCE MRI studies on vessel permeability and as matrix for molecular imaging probes.
Abstract ID: 596
Low expression levels of Vascular Endothelial Growth Factor (VEGF) are linked to motor neuron disease or Amyotrophic Lateral Sclerosis (ALS). VEGF is an angiogenic factor but recently its neuroprotective function is established in-vitro1. To unravel the underlying mechanism of VEGF participation in the pathogenesis of ALS, the group of Carmeliet P.2intercrossed VEGF knock-out mice with mice expressing the SOD1G93A transgene, a well known model for ALS3, and generated 4 genotypes of VEGF/SOD: Wt/Wt-Wt/He-Ho/Wt and Ho/He.
Basal Cerebral Blood Volume (CBV) and Flow (CBF)-explored with Bolus tracking (BT) MRI-was coupled to regional differences in cerebrovascular reactivity-measured with Arterial Spin labeling (ASL) MRI and the BOLD (Blood Oxygenation Level Dependent) response upon a hypoxic challenge (208s, 8% O2)-in all genotypes (ngenotype=8). These measurements were performed on a 7T MRRS system under isoflurane anaesthesia and accurate monitoring.
CBV was unaffected while CBF significantly decreased (50%) in the double transgenicsas compared to Wt/Wt. CBF increase upon hypoxia was 26.3% in Wt/Wt mice while absent in Ho/He and significantly reduced in Ho/Wt vs Wt/SOD and Wt/Wt. Increasing amounts of deoxyhemoglobin during hypoxia result in a BOLD drop, in proportion with the regional vascular density. These proportional changes are deficient in Ho/Wt, Wt/He and Ho/He resulting in a higher BOLD drop in regions with low vascular density. This is because less fresh blood is provided as the CBF increase upon hypoxia is low (or even absent). We believe that this phenomenon is also a sign of low basal CBF (as shown by our BT-MRI data) which might indicate impaired neuronal activity.
Based on neuropathological susceptibility to hypoxia we conclude that vascular deficits in these models are a result of a molecular interaction between the SOD and VEGF mutations.
Abstract ID: 597
The purpose of our study is to determine whether changes in tumor perfusion following antivascular treatment can be measured using contrast enhanced dynamic scanning with a micro-CT scanner capable of fast rotation speed(Explore Locus, General Electric Healthcare). Tumor cells from a human fibrosarcoma line were injected into a hind limb of four syngeneic C3H mice. At 2 weeks post, tumor perfusion (BF), blood volume (BV) and vascular mean transit time (MTT) (a marker for interstitial pressure) were measured before and at 1 hour after intratumoral injection of an antivascular drug that targets the tubulin cytoskeleton in endothelial cells. For these measurements, 0.2 ml of contrast (150 mgI/ml) was injected via a tail vein over 15 s and the tumor was scanned at 1 s intervals for 40 s. The other scanning parameters were 80 kVp, 60 mA, 150 micron resolution and 0.9 mm slice thickness. The scanner has a 4 cm axial FOV enabling the tumor and the heart to be imaged together. The arterial input function was obtained from a region of interest placed in the left ventricle. CT Perfusion software (General Electric Healthcare) was then used to measure BF, BV and MTT in the tumor as well as contralateral muscle (CM). BF and BV in tumor and CM decreased after antivascular treatment (P < 0.05) and the respective mean decreases in tumor 30 ± 11% and 43 ± 18% were different from those in CM 9 ± 14 % and 13 ± 13% (P < 0.05). There was a larger mean decrease in MTT (interstitial pressure) in tumor (20 ± 13%) than in CM (2 ± 26%), however, the difference was not significant. These results suggest that contrast-enhanced dynamic CT scanning holds promise for the monitoring of antivascular and by inference antiangiogenesis treatment of tumors in rodents.
Abstract ID: 598
We have reported a novel generic method to detect gene expression and transfer in vitro using MRI based magnetic-antigen-gene imaging contrast (M.A.G.I.C., [1]). Basically, exogenous gene expression is imaged by using a surface antigen of choice to which a specific antibody binds, the antibody being conjugated to a monocrystalline iron oxide nanoparticle (MION). To evaluate M.A.G.I.C. in vivo, nude mice were injected intravenously with either 50ug of plasmid encoding H2Kk (pMACSKK.II, n=2) or CD4 (pMACS4.1, n=1) with a transfection agent. A day later, animals were given anti-H2Kk-MION (150ul) intravenously following a baseline MRI scan. MRI was performed at 2, 24, 48 and 72h post-anti-H2Kk-MION injection. Negative enhancement (~30–50%) was observed in the liver, spleen and muscle at 48h post-anti-H2Kk-MION injection in the pMACSKK. II treated animals. A skin graft model was also employed to evaluate M.A.G.I.C. in vivo. Tail skin from CBA (H2K positive) and C6 (H2B positive) mice were grafted on either side of a nude mouse (H2D positive, n=1), effectively performing gene transfer ‘mechanically’. A week later, the mouse was given anti-H2Kk-MION (150ul) intravenously. Following sacrifice 24h later, the grafts were scanned in situ before being dissected and fixed in 10% formalin for MRI. Signal intensity was generally decreased in the H2K compared to the H2B graft in situ but particularly evident in the high resolution MRI images of the grafts ex vivo in which a 20% decrease in the T2 value of the H2K graft was observed. The attenuated signal intensity in the H2K graft arises from specific accumulation of the MION conjugated antibodies recognising the H2K antigen. In these preliminary studies using different gene transfer strategies, we have shown the possibility of M.A.G.I.C. to image gene transfer in vivo.
Abstract ID: 599
Abstract ID: 600
19F MRI allows selective imaging of the unique perfluorocarbon core of two types of fibrin-bound nanoparticles. As the concentration of the CE-nanoparticles decreases on the clots, the signal to noise ratio on the CE-selective images decreaeses (solid bars). Likewise, the SNR increases on the PFOB-selective images (open bars) as the concentraction of the PFOB nanoparticles increases.
Abstract ID: 601
Abstract ID: 602
Ultrasound is regarded as promising tool for molecular imaging because of inherent merits of ultrasound system as well as high sensitivity. We have been developing a starting microbubble ultrasound contrast agent for the molecular targeting and determining the combinations of shell materials for appropriate size, videointensity and stability of microbubbles.
We prepared the microbubbles with combinations of three parts of 4%, 10%, 20% or 40% dextran (molecular weight; 500K, 10K) and one part of 1% or 4% human serum albumin. The reaction mixtures were sonicated for about 120 seconds, with perfluorohexane infused.
The size of microbubbles was ideal with products using 10K dextran and the videointensity was high in the products with 1% albumin. The stability of microbubbles in all those products was good for as least 1 hour, in terms of bubble shape, size and videointensity.
The microbubbles with 10% dextran (M.W;10K) and 1% albumin, which showed highest videointensity, was injected to mouse for in-vivo study. Ultrasonogram using a 30-MHz transducer showed highest videointensity at 25min after injection in the liver and persistently increased intensity for at least 1 hour.
Abstract ID: 603
Abstract ID: 604