Abstract
Abstract ID: 680 Poster board space: 6
Abstract ID: 681 Poster board space: 7
We have designed and manufactured a lipid-encapsulated contrast microbubble which is ultrasonically echogenic at intravascular frequencies (30-40 MHz). Moreover, by incorporating specific ligands into the surface of the microbubbles, we have shown that the microbubbles can be targeted to specific cellular and tissue sites within the vasculature using antibodies to CD31 and CD54. We describe the acoustic characterisation of these agents and provide images and analysis of attachment of targeted microbubbles to cells, to agar surfaces and to ex vivo tissues. The 3-D structure of the microbubble agents was investigated by incorporation, post-bubble-synthesis, of the styryl lipophilic fluorescent dye, N-(3-triethy-lammoniumpropyl)-4-(4(dibutylamino)styryl) pyridinium dibromide (FM1-43) and obtaining z-stack images using confocal laser scanning microscopy (CLSM). The CLSM image data were deconvolved and reconstructed using 3-D modelling software. The data obtained demonstrate the heterogeneity of shapes and sizes of the contrast agents immediately after synthesis. Transmission electron microscopy of the microbubble constructs demonstrate the multi-lamellar shell. Interactions of the bubbles with cells under the influence of ultrasound and the influence of ultrasound, in the presence and absence of microbubbles, on the cellular stress response (induction of hsp70) are also described.
Abstract ID: 682 Poster board space: 8
The application of targeted diagnostic and therapeutic agents has emerged as an area of intense interest in oncology. Specific delivery and monitoring of these agents to cancer cells promise improved efficacy. Here, we reported the synthesis and preliminary biological evaluation of a new optical imaging-guided multimodality therapeutic agent comprising a tumor homing molecule, a chemotherapeutic agent and a photodynamic therapy (PDT) agent. This folate receptor (FR)-targeted agent is capable of the selective and simultaneous delivery of paclitaxel (PTX, a leading anticancer drug) and pyropheophorbide (Pyro, a well-known PDT agent and a fluorescent probe) to human cancer cells overexpressing FR. The structure of this molecule is represented as Pyro-FA-PTX. This agent has the following features: 1) FR is overexpressed in epithelial malignancies (e.g., ovarian, colorectal, and breast cancer) and many folate-conjugated agents were known to maintain high binding affinity to FR; 2) Potential synergistic effect generated by PTX (dark cytotoxicity) and Pyro-PDT treatments (light cytotoxicity) is highly desirable; 3) The near-infrared fluorescence capability of Pyro is a built-in optical guide for in vivo real time monitoring the drug accumulation, thus providing a seamless “see and treat” approach for multimodality cancer therapy.
To validate the Pyro-FA-PTX accumulation in cancer cells via FR-mediated endocytosis, confocal microscopy was performed. The following results were obtained: 1) there is strong accumulation of Pyro-FA-PTX throughout the whole KB cells (overexpressing FR) except for nucleus; 2) there is minimal accumulation of Pyro-FA-PTX in HT1080 cells (lack of FR expression); 3) An excess of free folic acid effectively blocked the uptake of Pyro-FA-PTX in KB cells. Taken together, these data indicated that Pyro-FA-PTX uptake in KB cells is FR-specific. The multimodality therapeutic function of Pyro-FA-PTX was supported by a preliminary MTT study, in which PTX-induced dark toxicity and Pyro-induced light toxicity were both observed in KB cells incubated with Pyro-FA-PTX.
Abstract ID: 683 Poster board space: 9
Abstract ID: 684 Poster board space: 10
A simple x-ray offers the possibility of an inexpensive, clinically-relevant technique for monitoring gene expression assuming a suitable enzyme and substrate can be found. The human sodium-iodide symporter which has been used successfully as a SPECT reporter gene, sequesters iodide intrazcellularly. Non-radioactive iodide is a widely used radiographic contrast agent, but it is not clear at present that enough iodide can be sequestered at a tolerable systemic dose to allow for monitoring with conventional radiographic equipment. We tested the feasibility of using radiographic contrast from non-radioactive iodide sequestered by the human sodium-iodide symporter, hNIS, in human tumor xenografts grown in mice and in the prostates of normal dogs made to express the hNIS. We measured radiographic contrast in mice using micro CT, clinical X-ray equipment, specimen radiography, and in dogs using a clinical spiral CT. Radiographic images were obtained one to three days after a local injection of a replication competent adenovirus expressing hNIS under control of the CMV promoter. Radiographic contrast due to hNIS gene expression, confirmed by the presence of fluorescence-tagged hNIS-antibody, was observed in tumors grown in mice and less convincingly in the prostates of dogs. Future studies to test the potential of using iodide sequestered as a result of NIS expression to render cells radiopaque is warranted since an inexpensive clinically-relevant tool to measure gene expression would facilitate the testing of improvements to gene therapy and speed the translation of promising technology to clinical use. Complicating issues such as existing microcalcifications and possible solutions with this technique will be discussed.
Abstract ID: 685 Poster board space: 11
Abstract ID: 686 Poster board space: 12
The Notch signaling pathway is important in cell proliferation, differentiation and apoptosis. Deregulated Notch signaling is also involved in tumorigenesis (breast cancer, lung cancer, pancreatic cancer and brain tumor). Binding of the ligands (Jagged and Delta) to the Notch receptor results in the transactivation of target genes such as Hes1 and Hes5. To monitor Notch activity by both bioluminescence and fluorescence imaging, we used the mouse Hes1 promoter to drive expression of a dual-reporter system (GFP fused to firefly luciferase) and verified its expression in culture. We then created a transgenic mouse line with this construct to monitor endogenous Notch activity postnatally. Many organs showed elevated signaling near birth with a decrease in reporter activity with increasing postnatal age consistent with the known reduction in Notch signaling during postnatal development. The expression patterns of GFP and firefly luciferase were correlative (Figure 1A). Primary cultured neurospheres from cerebrums of newborn transgenic pups were established. Cells in the center of the neurosphere expressed reporter, and expression was lost as cells differentiated and migrated from the center to the periphery (Figure 1B). The primary cell cultures were also used to investigate gamma-secretase-mediated Notch signaling. Notch reporter activity was suppressed following treatment with gamma-secretase inhibitors JC-18 and JC-34 (Figure 1C). We are crossing the nestin tv-a transgenic mouse with our reporter line, because Hes1 is overexpressed in RCAS induced gliomas and medulloblastomas, and we will quantify the Notch signaling in these RCAS/tv-a-induced tumors.
Abstract ID: 687 Poster board space: 13
Epidermal growth factor receptor (EGFR), a member of the EGF superfamily of receptor tyrosine kinases, is a critical regulator of cell growth and proliferation and an important target for anti-cancer therapeutics. To further investigate the dynamics of EGFR regulation, we have developed a firefly luciferase (FLuc)-based reporter system to image ligand-induced processing of EGFR in real time.
The reporter was created by fusing EGFR to FLuc downstream of the CMV promoter. A flexible glycine-serine linker was engineered between the two proteins to minimize steric hindrance and allow for full function of both components; this fusion reporter was then used to create a stable HeLa cell line. Photon output from these cells was continuously monitored for as short as 1 minute intervals after treatment with 100 ng/mL EGF, which results in autophosphorylation and internalization of EGFR. A decrease relative to photon output of untreated cells showed a reproducible minimum approximately 30 minutes post addition of EGF. This time course correlated with the ligand-induced response of endogenous EGFR as determined by Western blot. Treatment of the cells with varying concentrations of EGF indicated that the EC50 at 30 minutes was approximately 10 ng/mL. The initial decrease in photon output was contingent upon phosphorylation of EGFR, as it was abrogated by treatment of the cells with tyrphostin AG1478, a specific inhibitor of EGFR kinase function. Treatment with wortmannin, which inhibits the formation of intraluminal vesicles in the multivesicular endosome, significantly attenuated the decrease in photon output, indicating a requirement for proper intracellular sorting of EGFR. Thus, a novel FLuc fusion reporter allows continuous readout of EGFR activation in cellulo for temporal-based investigations of EGF-mediated signaling.
Abstract ID: 688 Poster board space: 14
Ras proteins play important roles in cell signal transduction including the cell division cycle, programmed cell death and differentiation. In addition, Ras proteins are also responsible for about 30% of all human carcinomas, including 90% of pancreatic, 50% of colon, 30% of lung and breast cancers. The proteins must be localized to the inner surface of the plasma membrane to be biologically active in all of these events. This first and most critical step in the post-translational modification is mediated by farnesyl protein transferase (FPT) where a lipid farnesyl group from farnesyl pyrophosphate (FPP) is transferred to the cysteine sulfhydryl of Ras which ends with a CA1A2X motif at its carboxyl-terminal. Positions A1 and A2 are occupied by aliphatic amino acids, while X is held by either methionine or serine. Since the relocation of Ras proteins from the cytoplasm to the plasma membrane upon farnesylation is crucial for tumor growth and survival, detection of FPT activity using molecular imaging probes based on a signal transduction platform is of considerable biomedical significance.
We report on the development and application of a fluorescent reporter probe to detect Ras-related FPT. We showed that upon farnesylation by FPT, the probe was brought close to a hydrophobic milieu and as a consequence emitting enhanced fluorescent light. A FPT activity assay confirmed the specificity of the probe with IC50= 1.2 μM. Using linear regression analysis fitting to a quadratic equation, we found that the probe has remarkable binding constant of Kd=26 nM. The specificity of the enzyme activation was proved in pure enzyme assays as well as in cell-based assays.
Abstract ID: 689 Poster board space: 15
HIF-1α is a key subunit of the HIF-1 transcription factor and plays an important role in the response to hypoxia. Proteasome-degradation of HIF-1α protein is regulated by oxygen-dependent prolyl-hydroxylation and mediated by pVHL. We have developed reporter constructs to image and study the regulation of HIF-1α protein stabilization by fusing the full length cDNA of HIF-1α, and a mutated HIF-1α (deleted oxygen-dependent degradation domain, ODDD), with FLuciferase in a bicistronic IRES-GFP cassette. Mice were injected subcutaneously with U87HIF1-α/FLuc or U87HIF-1α (ODDD-)/FLuc cells. After 1 week of tumor growth, D-Luciferin loaded micro-osmotic pumps delivered a constant infusion of D-Luciferin. BLI was performed before and sequentially after intraperitoneal injection of the hypoxia-mimetic CoCl2 (60mg/kg i.p.) or CoCl2+Rapamycine (1mg/kg i.p. at the same time). U87HIF-1α/FLuc xenografts showed a time-dependent increase in bioluminescence, peaking over pretreatment values (6-fold) 1.5 hours after CoCl2 injection, followed by a return to background levels at 5 hours. Rapamycine results in downregulation CoCl2-induced HIF-1α/Fluc bioluminescence by ≈ 50%. The mutated oxygen-insensitive U87HIF-1α (ODDD-)/Fluc xenografts demonstrated little or no response to CoCl2. We have shown that upregulation of the full-length chimeric HIF-1α/FLuc protein, as a marker of HIF-1α stabilization, can be visualized sequentially over time using BLI. Our Rapamycine inhibition data indicate that regulation of HIF-1α stabilization occurs in part through the mTOR pathway. Non-invasive imaging with these paired reporter systems can be used for high-throughput in vitro and in vivo screening of inhibitors affecting oxygen- and nonoxygen-dependent HIF-1α stabilization/degradation.
Abstract ID: 690 Poster board space: 16
A novel ionizing radiation (IR)-inducible system was developed by cloning five tandem repeats of IR-activated nuclear factor-κB (5NF-κB, 14 bp each) responsive elements to drive the expression of Gaussia luciferase (Gluc). Different spacer sizes ranging from zero to 12 bp were inserted between each element. These constructs were directly compared to well established IR-inducible promoters also driving the Gluc expression which includes the wild-type p53-activated fragment 1 (WAF1, 2.4 kb); early growth response factor (Egr-1, 550 bp); nine tandem repeats of the CArG elements of Egr-1 promoter (E9ns2); and four repeats of NF-κB binding sites (4NFB, 400 bp) from the cytoplasmic inhibitor of apoptosis (c-IAP2) promoter. Upon transfection of 293T human fibroblast cells with these constructs Gaussia luciferase bioluminescence was monitored without and after γ-ray irradiation (3–10 Gy). 5NF-κB without any spacers resulted in the highest induction levels, three-fold higher than the other IR-inducible promoters tested. The 5NF-κB responsive elements were also induced with bleomycin, a radiomimetic drug known to cause DNA strand breaks and cellular response similar to those induced by radiation. This system provides a tool to image radiation induced tissue changes and therapeutic gene expression in culture and potentially in vivo.
Abstract ID: 691 Poster board space: 17
Thomas Wurdinger, Christian Badr, Ralph Weissleder, Xandra Breakefield,
Nuclear factor-κB (NF-κB) transcription factor plays a key role in cellular stress response, in innate and adaptive immune response, and in the induction of anti-apoptotic factors. Here we analyze a novel construct to image NF-κB activation in endothelial cells by bioluminescence and fluorescence. Primary human brain microvascular endothelial cells (HBMVEC) stably expressing mCherry were transfected with a construct containing four tandem repeats of NF-κB responsive elements driving the expression of the naturally secreted Gaussia luciferase (NF-κB-Gluc). These cells were then cultured with either EBM-2 (basal medium) or EGM-2 (basal medium containing serum and growth factors. After 24 hrs, a 30-fold growth factor-induced expression of Gluc was observed. Further, when these cells were co-cultured with U87-tumor cells stably expressing Cerulean, a 10-fold induction of Gluc expression was detected. In addition, increased endothelial tube formation was observed by fluorescent microscopy upon co-culturing of the HBMVEC-mCherry and U87-Cerulean cells on matrigel. These results indicate that NF-κB-Gluc construct can be used to efficiently image growth factor and glioma induced NF-κB activation in HBMVEC cells. Since the Gluc reporter is naturally secreted, it allows easy monitoring of NF-κB activation by sampling the conditioned medium and could be applied to high throughput anti-angiogenesis drug screening in combination with morphological fluorescent microscopic analysis
Abstract ID: 692 Poster board space: 18
For spatial and quantitative analysis of cellular signal transductions in living mammalian cells, a novel method was developed, in which split Renilla luciferase complementation triggers spontaneous emission of luminescence, with a cell membrane permeable substrate, coelenterazine. This split Renilla luciferase complementation readout was shown to work for locating a protein-protein interaction. The split position of Renilla luciferase was elaborated for efficient complementation, which was found to be between 91 and 92 residues. Protein-protein interactions were simultaneously converted into the changes of the luminescence intensities in living cells via Renilla luciferase inter-molecular and intra-molecular complementation. The intra-molecular complementation readout of split Renilla luciferase is capable of detecting cellular signal transductions with expression of one protein in living cells.
Unlike conventional complement enzymes that lead to stable diffusive products, this split Renilla luciferase reporter is capable of detecting protein interactions with emission of bioluminescence only at the sites and time of their occurrence in living cells.
Abstract ID: 693 Poster board space: 19
Kristen Adams, Sunkuk Kwon, Feng Liang, George Mercier, Michael Barry, Eva Sevick,
Gene therapy promises new treatments for a wide range of common conditions in individualized molecular medicine, but is plagued by inefficient transport to diseased sites. Recent developments in the modification of the virus could allow more targeted approaches and herald the advent of systemic delivery of therapeutic viruses. Encouraging clinical trial results bodes well for the future of a targeted viral or non-viral approach, yet the safety of viral vectors underscores the importance to monitor their immediate and long-term delivery to diseased sites and their exclusion from normal tissue sites.
Herein, we report using IRDye 800 to dynamically image virus trafficking in vivo using 780 nm excitation light and emission collected at 830 nm. Adenoviral particles were labeled with IRDye800 particle number of 1times109 or 1 nanogram (or 1 picomolar) dye equivalent was administered i.v. (Figure 1). Particle numbers of 1times1010, or 10 nanogram (or 10 picomolar) dye equivalent was delivered to mice through oral gavage (Figure 2). Near, real-time virus imaging data were acquired at 800 milliseconds per frame and showed selective accumulation of stained virions in the mouse, dependent upon administration route in the early stage. Re-distribution of labeled virons to the other organs was followed over a period of time up to 24 hours after administration.
Our imaging data demonstrate that oral and intravenous administrated viruses can be dynamically imaged in mouse model with low background when near-infrared fluorophores are used.
Images of intravenously injected virus. Pictures show the times that after injection in second (s), minute (m) or hour (h).
Images of orally gavaged virus. Pictures show the virus in the GI system and then In the feces.
Abstract ID: 694 Poster board space: 20
The Fas-Associated Death Domain (FADD) protein was initially identified as a pro-apoptotic adapter that is essential for the induction of apoptosis in response to death receptor activation. Recent reports have demonstrated that on phosphorylation at Ser194, Phosphorylated FADD (FADD-P) instead of being cytosolic and pro-apoptotic translocates to the nucleus and has anti-apoptotic activity. We have revealed using quantitative 2D-PAGE and TMA analyses that levels of P-FADD correlate with aggressive phenotype and poor clinical outcome in cancer. P-FADD was shown to act as an important stimulus to promote the activation of NF-κB, which plays a pivotal role in tumor development, progression, and therapy. Although a number of Ser/Thr Kinases have been implicated as FADD kinases, their role in cancer and their regulation is not well understood. To elucidate the molecular mechanism of FADD phosphorylation in tumor progression and uncover the upstream signaling events that lead to expression of the FADD Kinase, we have developed a genetically engineered reporter named FADD Phosphorylation Reporter (FPR) that consists of the amino- and carboxyl-domains of luciferase coupled to the FHA2 phos-phor-aminoacid-binding domain and a 66 amino-acid sequence from FADD harboring Ser194. Phosphorylation of the FPR in response to FADD-Kinase activity results in a conformational change such that the reporter molecule lacks luciferase activity. However, treatment of cells with inhibitors of upstream signaling cascade resulted in loss of FADD-Kinase activity and activation of the FPR reporter with a concomitant decrease in P-FADD levels. Thus, FPR allows non-invasive imaging of FADD phosphorylation both in live cells and animals. Furthermore, it would aid in interrogating the regulation of FADD-Kinase activity, signal transduction pathways and in understanding how FADD phosphorylation balances the processes of apoptosis and proliferation in normal and cancer cells. This reporter is also being used to identify novel anticancer drugs using cell-based high throughput screening.
Abstract ID: 695 Poster board space: 21
Abstract ID: 696 Poster board space: 22
Abstract ID: 697 Poster board space: 23
Dual non-invasive imaging, bioluminescence and PET imaging for apoptosis using DEVD-aminoluciferin and 124I-Annexin V could be useful for the evaluation of apoptosis in cancer therapy monitoring and new drug development.
Abstract ID: 698 Poster board space: 24
Apoptosis or programmed cell death, in multicellular organism, plays a crucial role in development and in maintenance of cellular homeostasis. Dysregulated apoptosis is implicated in an array of clinical pathologies including cancer, autoimmune and neurodegenerative diseases. During programmed cell death a series of proteases known as caspases play critical roles in the initiation and propagation of apoptotic signal. Caspase-3, a protease with a key role in terminal phase of apoptosis, recognizes and cleaves cellular targets harboring the amino acid sequence DEVD. Based on this target recognition and cleavage we developed a number of novel imaging platforms for sensing caspase-3 proteolytic activity. First strategy included development of a recombinant luciferase reporter molecule which when expressed in mammalian cells has attenuated levels of reporter activity. In cells undergoing apoptosis, a caspase-3-specific cleavage of the recombinant product resulted in the restoration of luciferase activity that can be detected in living animals. In an effort to enhance signal to noise ratio, we utilized β-galactosidase, which functions as a tetramer, to generate a second generation of caspase-3 sensing recombinant reporter. Finally, we developed a cell surface single chain antibody (sFv) based highly sensitive and specific caspase-3 imaging reporter which may be adapted for multimodal molecular imaging. This reporter, when expressed in cells localized to Golgi bodies. However, in cells undergoing apoptosis, caspase-3 dependent cleavage of the reporter resulted in translocation of the sFv to cell surface where it is detected by its specific interaction with hapten moiety coupled to imaging agents. These imaging strategies provide a platform to dynamically and non-invasively evaluate efficacy of anticancer treatment regimens, screening for novel experimental therapeutics and target validation both in vivo and in cell lines.
Abstract ID: 699 Poster board space: 25
We present two methodologies for mapping histology to in vivo imaging. One uses many histological sections and the other few. Both make few a priori assumptions regarding morphological deformations, and neither requires insertion of extrinsic fiducials. The cornerstone of both methods is the digital photographs taken of the sample's tissue block face before the histology section is cut. This block face photograph is used to unwarp in 2 dimensions the deformations induced during sectioning of the histology slide. Ex vivo geometry is established either by the presence of many cut sections of the tissue block, or by a high resolution MR imaging acquisition of the whole ex vivo specimen. In vivo geometry is established by a typical MR acquisition of the specific animal, or in the case of stereotaxic animals such as mice and rats, by atlases such as those available from LONI at UCLA. Individual steps required in each of these methods will be covered, as well as a list of pitfalls commonly seen in current literature.
Abstract ID: 702 Poster board space: 26
Abstract ID: 703 Poster board space: 27
Alterations in myocardial substrate utilization have been implicated in several cardiac diseases, including diabetes. Recent publication of the mitochondrial metabolic network offers an opportunity to explore and profile imaging probes, in silico, for characterization of mitochondrial abnormalities and elucidation of molecular imaging probes (MIPs). In this work, we use concepts embedded in metabolic flux analysis (MFA) to profile metabolic differences in normal and diabetic rat hearts. The mitochondrial metabolic network consists of 189 reactions and 230 metabolites. The biochemical reactions governing the metabolic network are represented by a mass-balanced stoichiometric matrix, S, and a flux vector, v such that at steady state S·v=0. A linear programming (LP) formulation of the model was coded in MATLAB for constrained MFA. Flux constraints were derived from the literature, when available, and from tracer kinetic modeling of [15O]-Water, [11C]-Glucose, [11C]-Acetate, and [11C]-Palmitate microPET imaging of Zucker Lean (ZL) and Zucker Diabetic Fatty (ZDF) rat models through evaluation of myocardial oxygen utilization (MVO2), fatty acid oxidation (MFAO), glucose utilization (MGU), and blood flow (MBF). In both the ZL and ZDF rat models, the objective function of the constrained LP optimization was to maximize ATP demand. Following optimization, the metabolic profiles of ZL and ZDF models were compared and differences in metabolic flux were represented as percent difference from ZL rats. Of the 189 reactions in the mitochondrial metabolic network, 39 registered no change in net flux between ZL and ZDF models. Of the remaining 150 reactions, the greatest metabolic flux change was attributed to reactions involved in fatty acid metabolism, transport and activation, offering targets for imaging alterations in myocardial fatty acid utilization. MFA offers the opportunity to explore and profile MIPs in silico under varying metabolic states and correlate such results with biomarkers of disease.
Abstract ID: 704 Poster board space: 28
Colorectal cancer is one of the most prevailing and lethal forms of cancer in USA. Early diagnosis and treatment are crucial for disease fatality. Traditional screening approaches are either inaccurate or unpleasant and therefore are avoided by many patients. Recent developments in virtual colonoscopy to detect colon polyps by CT have made screening less uncomfortable. However, CT can only detect anatomical abnormalities.
Emerging biotechnologies in biomarker detection and quantification, enriched by promising genomics and proteomics techniques, are leading to new noninvasive tests to accurately detect colorectal cancer and estimate disease progression. The merging of biomarker and imaging approaches will lead to a more precise noninvasive screening method.
Markov model simulating a screening protocol combining in vitro diagnostics (IVD) with CT colonography.
Simulation results of different screening methodologies.
We propose an integrative screening method for colorectal cancer screening that merges in vitro molecular diagnosis with clinical and imaging information for clinical trials. It aims to address the two principal concerns: maximizing effectiveness and reducing patient discomfort for higher rate of patient compliance. Moreover, it also has the potential to be less costly than a standard colonoscopy.
Markov modeling simulations were carried out. The results indicate that our proposed non-invasive methods may increase the number of life-years saved by over 50% and improve the cost efficacy when compared to a screening protocol based on colonoscopy or CT colonography.
Abstract ID: 705 Poster board space: 29
Although there was great success in identifying disease gene in simple, monogenic Mendelian traits, the understanding of genetic mechanisms in most complex diseases remains challenging. A central goal is to identify single nucleotide polymorphisms (SNPs) and their interactions that confer the susceptibility of the disease. Traditional methods, such as multiple dimensional reduction method and combinatorial partitioning method, etc., provide good tools to decipher such interactions in the absence of genetic heterogeneity among population. However, these traditional methods have not managed to solve genetic heterogeneity problem common to diseases. As prior knowledge regarding the causes of genetic heterogeneity is rarely known, these methods based on estimation over the entire population, are unlikely to succeed in tackling the genetic causes of disease.
Directed graphs from Bayesian learners in Boosted Generative Modeling illustrate the high-order interactions
Thus, we are motivated to propose a novel boosted generative modeling approach for structure-modeling the interactions leading to diseases in the presence of genetic heterogeneity. This method innovatively bridges the ensemble method and generative modeling in the genetic association study. Generative modeling is to model interaction network configuration and the causal relationship, while boosting is used to address the genetic heterogeneity problem, a common problem in genetic epidemiology study. We perform our method on simulation data of complex diseases. The results indicate that our method is capable of structure-modeling of interaction networks among disease-susceptible loci and addressing genetic heterogeneity issues concurrently, where the traditional methods, such as multiple dimensional reduction method, fail to apply. It provides an exploratory tool for potential SNPs that are likely to contribute to the diseases.
Abstract ID: 706 Poster board space: 30
Belinda Beck1, Steve Bryant1, Kenneth Cheng1, William Eckelman1,
MICAD (http://micad.nih.gov) is a freely accessible online source of scientific information on in vivo molecular imaging agents. It has been developed as a key component of the Molecular Libraries and Imaging (MLI) program of the National Institutes of Health (NIH) Roadmap for Medical Research in the 21st century. The MLI program is a component of the NIH Roadmap that addresses the critical issue of finding new pathways to discovery by supporting the emerging field of molecular imaging. As a part of the MLI program, MICAD focuses on providing critical and timely scientific information to enhance the discovery and availability of molecular imaging agents. The MICAD team is composed of NIH staff who have been working with the guidance of a trans-NIH panel of experts in the field of molecular imaging. MICAD is designed to provide concise, up-to-date, and the most relevant information on molecular imaging agents to the basic and clinical research community. The database includes, but is not limited to, agents developed for positron emission tomography, single photon emission computed tomography, magnetic resonance imaging, ultrasound, computed tomography, optical imaging, planar radiography, and planar gamma imaging. The information is summarized in five major sections (background, synthesis, in vitro studies, animal studies, and human studies) that describe the specificities, biological activities, and medical applications of imaging probes for a wide range of diseases and biological functions. The database also features literature references with links to MEDLINE/PubMed, chemical structures with links to PubChem, and additional related resources at NCBI. The MICAD website was officially launched September 2005. It currently contains more than 120 agents and is fully searchable. All molecular imaging agents that are published in peer-reviewed literature will eventually be included.
Abstract ID: 707 Poster board space: 31
Ruping Shao, Chiyi Xiong, Xiaoxia Wen, Renata Pasquilini, Wadih Arap, Juri Gelovani,
The ability to evaluate apoptosis by a noninvasive method should facilitate the clinical care of cancer patients by abbreviating the time required to assess the efficacy of anticancer therapies. One of the markers of apoptosis is phosphatidylserine (PS) exposed on the surface of apoptotic cells. To identify PS-binding ligands for apoptosis imaging, we performed panning of a random 12-mer peptides phage library on PS coated on 96-well plate. A phage clone with 10% appearance frequency on the selections was isolated after 6 rounds panning. Both colony-forming assay and ELISA assay showed that particles from this phage clone (PS-3-10), but not a control phage clone (4-2-2), bound to aged red blood cells (RBC) that had PS exposed on their outer membrane leaflet. Moreover, binding of PS-3-10 phage particles to RBC was blocked by pre-incubation of the phage particles with PS, suggesting this binding is PS-specific. The binding of PS-3-10 phage to PS coated surface was 2–4 fold higher than binding to surfaces coated with phosphocholine (PC) or bovine serum albumin. To further test whether PS-3-10 phage particles bind to apoptotic cells, we incubated PS-3-10 with human colon cancer cells DLD1 that had been treated with tumor necrosis factor-related apoptosis-inducing ligand (TRAIL). TRAIL induced greater than 80% apoptosis in DLD1 cells. Binding of PS-3-10 phage particles to TRAIL-treated cells was four times higher than their binding to untreated viable cells, while control 4-2-2 phage did not bind to apoptotic DLD1 cells. Binding of PS-3-10 particles and the corresponding biotinylated 12-mer peptide to apoptotic DLD1 cells was also visualized by immunohistochemistry staining. Taken together, our data suggest that both PS-3-10 phage particles and PS-3-10 peptide targeted to PS on cell surface. Studies addressing the imaging potential of radiolabled PS-3-10 particles and the corresponding peptides are in progress (supported by John S. Dunn Foundation).
Abstract ID: 708 Poster board space: 32
Pre-clinical and clinical studies demonstrate that tumors respond to a variety of anti-cancer therapies by increasing the MRI-measured apparent diffusion coefficient of water (ADCw) early in the treatment regimen, usually within 1 to 4 days after initiation of therapy [1–3]. It is proposed that these ADCw increases result from increased tumor extracellular volume fraction, which allows for an increase in unrestricted water motion. This increase in extracellular volume has often been ascribed to apoptotic cell shrinkage that can occur early in the therapeutic timeline. However, other mechanisms of death can occur soon after the initiation of therapy, including mitotic catastrophe and lytic necrosis [4]. These other cell death modalities may also alter the fraction of water with unrestricted motion. This study examined the regional increases in the ADCw that occurred in the context of multiple cell-death modalities in breast cancer xenografts responding to the anti-microtubule drug docetaxel (Taxotere®). Regions of MR images were manually registered to histological sections and each region was scored for type of cell death using hematoxylin/eosin, cleaved caspase, and cytokeratin-specific dyes. Results indicated that early and significant changes in ADCw occurred concomitantly with mitotic catastrophe or necrotic cell condensation, in the absence of apoptosis. These data were also observed in vitro, where mitotic catastrophe was associated with increased plasmalemmal permeability, which would lead to increased ADCw [5]. Hence, ADCw may be a general measure of cytotoxic therapy response and mechanisms other than apoptosis can mediate these MR-measured changes.
Abstract ID: 709 Poster board space: 33
Seung-Cheol Lee, Ming Huang, David Nelson, Stephen Pickup, Harish Poptani,
Magnetic resonance spectroscopy (MRS) has great potential for prediction and detection of therapeutic response of cancer. To determine if MRS can detect such treatment response in non-Hodgkin's lymphoma and to identify the most reliable and earliest biomarker of response, we performed an in vivo longitudinal MRS study of a mouse xenograft model of human non-Hodgkin's lymphoma. A CHOP (Cyclophosphamide, Hydroxydoxorubicin, Oncovin and Prednisone) chemotherapy regimen was performed on SCID mice subcutaneously implanted with diffuse large B-cell lymphoma, the most common form of non-Hodgkin's lymphoma. We noninvasively measured lactate, total choline, total lipid by 1H MRS, and the phosphomonoester (PME)/β-nucleoside triphospate (βNTP) ratio by 31P MRS before and during the three week chemotherapy period. A Selective Multiple Quantum Coherence (SelMQC) sequence was used to detect lactate, a STimulated Echo Acquisition Mode (STEAM) sequence to detect total choline and total lipid, and an Image-Selected In vivo Spectroscopy (ISIS) sequence to detect 31P metabolites. With ongoing treatment, lactate decreased, choline did not change, and lipid and PME/βNTP increased. Among these four MRS biomarkers, significant changes occurred for lactate (p< 0.05) as early as one week post-treatment and for PME/βNTP (p< 0.05) at three weeks post-treatment. These results demonstrate that MRS detects therapeutic response in non-Hodgkin's lymphoma xenografts, and lactate is the most reliable and earliest MRS response biomarker.
Abstract ID: 710 Poster board space: 34
Abstract ID: 711 Poster board space: 35
The herpes simplex virus type 1 thymidine kinase (TK) enzyme has been well studied as both a reporter and a suicide gene for cancer gene therapy. However, one problem associated with cancer gene therapy has been non-target expression of the transgene. In this regard, we have constructed an adenovirus (AdhspmTK) encoding a mutant TK (mTK) under control of the full-length human heat shock 70 promoter (hsp). The mutant TK (mNLS-sr39TK-EGFP) has been fused to enhanced green fluorescent protein (EGFP) and previously been shown to be an improved reporter and suicide gene compared to wild-type TK. Since local heating can be applied to head and neck squamous cell carcinoma (HNSCC), we evaluated AdhspmTK in the human HNSCC cell line, SCC-9. The cells were infected at 0, 50, 100, or 500 plaque forming units (pfu)/cell and then heated for 0.5, 1, 2, and 4 h at 41°C at 1, 2, and 3 days after infection. Transgene expression was then evaluated one day after heating using 3H-penciclovir assays and determining EGFP fluorescence by flow cytometry. This showed that a 1 h heat time, 1 day after infection resulted in the greatest uptake of 3H-penciclovir at 3450 DPM/mg. This correlated with the highest expression of EGFP. At these time points, infection at 500 pfu/cell and heating resulted in a 20-fold increase in transgene expression compared to an 1.3-fold increase at 500 pfu/cell without heating. Another study in which the cells were infected, heated one day after infection for 1 h, and then assayed for transgene expression 2 and 3 days after heating showed that transgene expression did not increase 2 or 3 days after heating compared to one day after heating. Future studies will evaluate this vector as a PET reporter gene and a suicide gene in mice bearing SCC-9 xenografts.
Abstract ID: 712 Poster board space: 36
Abstract ID: 713 Poster board space: 37
Imaging the delivery of a drug activating enzyme to a tumor creates a unique opportunity for optimizing the delivery of cytotoxic therapy while minimizing normal tissue toxicity. Here we report a novel bioconjugate in which bacterial cytosine deaminase (bCD) was covalently conjugated to poly-
We would like to acknowledge the support from P50 CA103175 and helpful discussions with Prof. M. Neeman and Dr. V. Raman.
Schematic structure of bCO-PLL conjugate
Abstract ID: 714 Poster board space: 38
Abstract ID: 715 Poster board space: 39
In vivo functional and molecular MR imaging can potentially be used as biomarkers to monitor and/or predict response to anti-cancer therapy. Response prediction is especially important as it can be used for patient segmentation that would hence make clinical trials more efficient. This project will test the hypothesis that cancer cells with different molecular phenotypes can be discriminated by non-invasive imaging. Hence, pre-therapy images may predict response to targeted therapies, such as those inhibiting HIF, PI3 kinase or EGF receptors. Primary breast cancer cell lines (3199, 3171, 2150) that differentially express HIF-1α, c-myc and pAkt have been developed and characterized in vitro. In this study, these cells have been grown as orthotopic xenografts in 6-week-old female SCID mice. At tumor volumes of 300–800 mm3, mice were imaged by diffusion and dynamic contrast enhanced MRI. Diffusion MRI is sensitive to tissue cellularity and reflects cytotoxic or vasogenic edema. Preliminary data indicate that the 3199 cells (HIF+ / Akt− / myc–) grew without a lag period and had increased ADCw values with increasing tumor volumes. In contrast, 3171 (HIF+ / Akt+ / myc−) and 2150 (HIF− / Akt− / myc+) tumors had slower tumor growth rates and elevated ADCw values even at small tumor volumes, indicating the presence of edema (Figures 1 and 2).
Abstract ID: 716 Poster board space: 40
The peripheral cannabis (CB2-) receptor has been found in atherosclerotic plaques of both mice and man and stimulation has proven to ameliorate plaque phenotype in mice. Incorporation of the highly selective CB2 agonist HU-308, which has a lipohilic tail and a polar head, into a micelle has both diagnostic and therapeutic potential.
The diameter of the micelles was 8–14 nm (dynamic light scattering).
Chinese hamster ovary cells (CHO cells) with and without CB2 receptor were grown on cover glasses and incubated with micelles for 1.5 hours.
MRI measurements were performed on a 9.4 T magnet. T1 of a micellar solution (2.5 micromole/L Gd-BSA) was 2050 ms as compared to 3600 ms for the alfa-MEM without micelles (Fig. 1A/B). Also cover glasses were imaged with inversion recovery (Fig. 1C).
Binding of micelles to the CB2 receptor was verified after measurement of the expression of phosphorylated and total MAP kinases with Western blot analysis. Also micelles without HU-308 were tested (Fig. 2).
MRI: Inversion Recovery
Western Blot
Abstract ID: 717 Poster board space: 41
We describe a novel in vivo biomarker for assessing the location and quantity of immune cells participating in an immune response. This approach labels primary immune cells ex vivo with fluorocarbon-based nanoparticles, introduces the labeled cells into a subject, and then tracks the cells in vivo using 19F MRI [1]. The key advantage of this method is that the 19F images have superior selectivity for only the labeled cells, with no background signal from unlabeled cells. The Larmor frequencies of 1H and 19F differ by approximately 6% and the relative sensitivity is 0.83. The fluorocarbon agents are biologically inert and have no apparent cytotoxicity. Furthermore, conventional 1H anatomical images acquired in the same imaging session can be used to place 19F-labeled cells into their anatomical context. Importantly, the absolute number of labeled cells in a region of interest (ROI) can be calculated directly from the in vivo 19F images, as it is simply proportional to the integrated 19F MRI signal within an ROI. This approach offers an alternative to metal-ion based 1H contrast agents. In this presentation we show examples of how these 19F MRI methodologies can be applied to a variety of cell types and be used to follow immune cell trafficking due to innate immunological cues. The PFPE labeling and 19F MRI methodologies are readily extended to other cell types, such as stem cells, and represent a unique biomarker for emerging cellular therapeutics [1].
Abstract ID: 718 Poster board space: 42
Abstract ID: 719 Poster board space: 43
Functional diffusion mapping (fDM) is an MRI biomarker for mapping heterogeneous changes in tumor cell density during therapy. A potential application of fDM is in tailoring induction therapy for organ sparing in the treatment of head and neck tumors. A comprehensive head and neck MR protocol was performed on 5 patients at baseline and three weeks into a clinical chemo/radiation therapy protocol. The images (T2, T1 and T1 gadolinium enhanced and diffusion weighted) were acquired on a 3T MRI system using an eight channel head and neck coil. The diffusion images were trace weighted and acquired with a single-shot SENSE-EPI sequence, from which the ADC images were calculated and spatially co-registered for fDM calculation. Five patients with malignant squamous cell carcinoma of the head and neck were recruited to this MRI feasibility trial. Images had minimal image distortion and excellent signal to noise in the gray matter, brain stem and tumor. This resulted in excellent ADC images at baseline and three weeks into therapy. Both primary and metastatic parotid gland lesions were clearly seen on the anatomical T2, T1 enhanced and ADC images. The excellent signal to noise allowed all diffusion images and ADC maps to be co-registered to the baseline T2 weighted images. The fDM results revealed significant regions where the cell density had clearly decreased. Analysis of the tumor ADC histograms and tumor size revealed that there was a significant decrease in both tumor cell density and tumor size as early as three weeks into therapy. The results show the feasibility of obtaining ADC images in a single shot with minimal motion and susceptibility artifact. The fDM quality will allow for a rigorous clinical trial to determine whether diffusion imaging during radiation/chemotherapy of head and neck tumors is capable of predicting patient outcome and tumor response to therapy.
Abstract ID: 720 Poster board space: 44
The field of cancer molecular imaging would benefit from advancement in two major areas of research, i.e. the development of tumor-targeted imaging strategies and the acquisition of multi-modality capabilities. Recently, we have described an imaging probe, CLIO-EPPT, which addresses both of these issues. CLIO-EPPT targets the tumor-specific underglycosylated MUC-1 (uMUC-1) antigen, characteristic of over 50% of all human cancers[1]. In addition, CLIO-EPPT can be detected both by magnetic resonance imaging (MRI), by virtue of its superparamagnetic iron oxide core, and by near-infrared fluorescence imaging (NIRF), since it carries Cy5.5 dye, attached to the nanoparticle.
In our more recent studies, we explored the utility of CLIO-EPPT for tumor delineation and monitoring in an orthotopic model of pancreatic cancer[2]. Furthermore, we investigated the potential of the multimodal imaging strategy for the monitoring of tumor response to chemotherapy. Tumor-bearing animals were treated with the standard chemotherapeutic agent, 5-fluorouracil, and imaged by MRI and NIRF imaging before and after treatment. Quantitative assessment of change in tumor size over time revealed a remarkable agreement between MRI and NIRF imaging (Fig. 1). Thus, by combining both imaging modalities one can derive comprehensive dimensional information about tumor evolution over time with high spatial (MRI) and temporal (NIRF) resolution, unlimited depth penetration (MRI) and high sensitivity (NIRF).
While not clinically relevant at this stage, due to the limited applicability of fluorescence imaging in humans, the described studies have profound research implications. They validate the feasibility of employing similar multimodal imaging strategies with the ultimate goal of defining the dynamics of tumor response to a variety of factors, including, but not limited to, therapeutic intervention.
Abstract ID: 721 Poster board space: 45
The aim of this study was to compare the response of a glioma tumor to chemotherapy using 2-[18F]fluoro-2-deoxy-
Abstract ID: 722 Poster board space: 46
ZD6126 (AstraZeneca®) is an antivascular agent that selectively targets the tumor vasculature. We measured the agent's antivascular effects with vascular volume and permeability MRI measurements.
106 DU-145 prostate tumor cells were inoculated subcutaneously in SCID mice. Volume-matched tumors were imaged 24 h post i.v. treatment with ZD6126 on a 4.7T Bruker MRI system. Multislice relaxation rate (1/T1) maps were obtained by saturation recovery combined with fast T1 SNAPSHOT-FLASH imaging. An Mo map was acquired (recovery delay of 7s). Then, images for three relaxation delays (100ms, 500ms, and 1s) were acquired on 1 mm slices with in-plane resolution of 125 μm prior to i.v. administration of 0.2 ml of 60 mg/ml albumin-Gd-DTPA in saline (dose of 500 mg/kg). Starting 3min after injection of albumin-Gd-DTPA, imaging was repeated 6 times every 3min. Then the animals were sacrificed, T1 of blood withdrawn from the inferior vena cava or tail vein was measured, and tumors were fixed in 10% buffered formalin for histological analysis. A two-sided Wilcoxon exact test was used; p< 0.05 was considered significant.
Central tumor slices were analyzed. Histology showed significantly increased necrosis in the treated group (61.3% ± 19.8 SD, n=8) compared to controls (14.4% ± 7.2 SD, n=7). Vascular volume for the treated group (29.8% ± 18.5 SD) was significantly lower than that of controls (58.3% ± 14.1 SD). The permeability-surface-area-product (PSP) for the treated group (32.8% ± 18.2 SD) was significantly lower than that of controls (64.7% ± 13.5 SD).
A significant reduction in the % area of detectable vascular volume and PSP was apparent within 24 h post treatment with ZD6126, compared to controls. Vascular MRI parameters were found to be useful noninvasive endpoints of early response to antivascular therapy with ZD6126 in our prostate cancer model.
Abstract ID: 723 Poster board space: 47
Abstract ID: 724 Poster board space: 48
Abstract ID: 725 Poster board space: 49
Ryuchi Nishii, Asutosh Pal, Suren Soghomonyan, Julius Balatoni, Ioseb Mushkudiani, Hsin Hsien Yeh, Uday Mukophadhay, Andrei Volgin, Aleksander Shavrin, David Maxwell, William Tong, Mian Alauddin, William Bornmann,
Abstract ID: 726 Poster board space: 50
Breast cancer remains a prevalent and serious form of neoplastic disease. Assessment of treatment response is traditionally accomplished by physical exam and volumetric-based measurements using mammography or MRI. Early evaluation of treatment response using an imaging biomarker could provide an opportunity to provide patients with improved clinical care. Diffusion MRI is emerging as an imaging biomarker for treatment assessment as tumor water diffusion values are related to the cellularity of the tissue. Treatment-induced loss of tumor cells results in detection of increased tumor diffusion values. Mice with orthotopically implanted MX-1 human breast cancer cells were divided into control (n=7) and cyclophosphamide (n=7) treated groups. Treated animals were observed to have a significant reduction in tumor volumes with recurrence beginning at day 15 post-treatment. The mean apparent diffusion change for treated tumors increased by 44 and 94% at days 4 and 7, respectively which was statistically greater than control tumors. This study reveals that changes in diffusion MRI reflect early alterations in tumor microenvironment which predict tumor response and improved outcome. Thus, diffusion MRI as a noninvasive imaging biomarker has significant potential for evaluation of neoadjuvant breast cancer therapy in the clinical setting.
Abstract ID: 727 Poster board space: 51
Overhauser-enhanced magnetic resonance imaging (OMRI) is a novel functional MRI technique, in which a paramagnetic contrast agent such as the trityl radical is introduced to enhance the proton signal based image intensity through dynamic nuclear polarization. The enhancement mainly depends on local concentration of contrast agent, oxygen concentration, and proton T1 in vivo. The concentration of contrast agent depends on the efficacy of vascular delivery and transcapillary transfer. Thus OMRI image intensity well reflects the tumor vascularization. In the OMRI-based oximetry, a narrow-line paramagnetic contrast agent based on the trityl radical was injected into the subject of study, causing oxygen dependent enhancement of MR signal. The change in the MR intensity, obtained at two different EPR irradiation power levels, is used to compute the EPR spectral line width of the contrast agent, which in turn is correlated to partial pressure of oxygen, one of the most important factors in radio-sensitization. The pO2 mapping of three different kinds of solid tumors implanted into mice was performed, resulting in the visualization of hypoxic region. The OMRI sequence used for oximetry simultaneously permits calculation of the concentration of contrast agent. On account of non-membrane permeable property of used contrast media, the inverse of local concentration of contrast agent could be estimated to correlate with the increased interstitial fluid pressure in tumor region that has been correlated to resistance to anti-tumor drugs delivery. This study presents the feasibility of noninvasive Overhauser-enhanced MR imaging to simultaneously obtain three key factors in cancer therapy: oxygenation, interstitial fluid pressure, and perfusion in tumors.
Signal intensity (A). Interstitial fluid pressure (B). and pO2 mappings of murine SCC tumor by OMRI
Abstract ID: 728 Poster board space: 52
Prostate cancer is the leading non-cutaneous malignancy among American men and accounts for the second highest number of cancer related deaths within that group. In cases of advanced prostate cancer, high incidences of skeletal metastases are commonly observed to which there is currently no cure. Furthermore, treatment of oncogenic bone lesions is hampered by difficulties in diagnosis and assessment of treatment response where bone disease is considered “nonmeasurable” by RECIST clinical response criteria. As such, treatment of bone cancer remains a major hurdle due to the inability to reliably and efficiently assess therapeutic response in vivo.
In the present study, we investigated the use of diffusion magnetic resonance imaging (MRI) as a sensitive approach for monitoring treatment response of bone metastases. As our model for metastatic disease, we employed a previously described PC3 based murine model, which mimicked clinical manifestations of prostate cancer that allowed for rapid assessment of disease progression using bioluminescence imaging.
Diffusion MRI was employed to monitor response of bone metastases to docetaxel therapy throughout the course of the experiment. Histogram analysis of tumor ADC throughout the experiment demonstrated that docetaxel therapy induced increases in ADC values among treated tumors whereas ADC values remained unchanged in control tumors. Subsequently, we employed a previously described functional diffusion mapping (fDM) technique to quantify observed changes in tumor ADC, which revealed that approximately 30 percent of the tumor had significantly increased in ADC 10 days post initiation of therapy. These results demonstrate, for the first time, that diffusion MRI can serve as an early imaging biomarker for treatment response in bone metastases. Finally, given the clinical translatability of diffusion MRI, this study signals for the potential use of this approach in clinical care where current methods fail in providing reliable assessment of therapy response in patients with metastatic bone disease.
Abstract ID: 729 Poster board space: 53
Geoffrey Wang,
In 2005, we have been awarded NCI grant to establish Emory-Georgia Tech Center of Cancer Nanotechnology Excellence (CCNE). The goal of this center is to combine cancer biology with nanotechnology and informatics for personalized oncology. That is, to deliver novel molecular imaging probes, nano-therapeutics, and computing tools to treat cancer.
To conduct this translational research, we design the following phases: (1) to identify biomarkers using knowledge-based data mining; (2) to synthesize quantum dot molecular imaging probes targeting these biomarkers; (3) to perform in vitro and in vivo imaging of biomarkers using these new molecular probes; (4) to analyze molecular imaging data; and (5) to correlate analysis with clinical information.
As the first step towards this Bio+Nano+Info fusion for personalized medicine, we have researched and developed a database to archive information involved in the whole process of study. We quantify the molecular imaging data to create feature descriptors. Then we investigate to correlate these with clinical tissue sample and biomarker information. Currently, we have been able to archive the data and we are in the process of developing correlation score.
Abstract ID: 732 Poster board space: 54
Both glioma and areas of radiation necrosis are typically permeable to contrast agents and show enhancement on CT or MRI. However, there is little active angiogenesis at the site of radiation necrosis. The purpose of this study was to differentiate glioma from radiation necrosis based on differential active angiogenesis using magnetically labeled endothelial progenitor cells (EPCs) and MRI. Double labeled human EPCs (SPIO and DiI) were injected IV in rats bearing an intracranial human glioma or in rats with radiation injury. Multi-echo T2W, T2*W and 3D-GRE MRI were obtained on day 7 after the injection of cells. Following in vivo MRI, rats were euthanized, perfused and the brains were collected. Randomly selected brains also underwent ex vivo MRI. Collected brains were snapped frozen for sectioning. Sections were stained for blood vessels using FITC labeled lectin, and the presence of endothelial markers was detected by antibodies against human CD31 and vWF. Prussian blue staining was performed to detect the iron positive cells. Distributions of blood vessels and patterns of neovascularization at the site of radiation were compared with that of implanted tumor and surrounding brain. Both in vivo and ex vivo MRI showed low signal intensity areas at the margin of the tumors in rats injected with labeled cells. Iron positive cells were present at the corresponding sites of low signal intensities seen on MRI. There were also CD31 and vWF positive cells at the corresponding site of iron positive cells in tumors. Fluorescent microscopy showed DiI positive cells forming new vessels. There was no definite low signal intensity area indicating accumulated iron positive cells seen at the site of radiation injury. Prussian blue staining also did not show any iron positive cells. By targeting angiogenesis, it is possible to differentiate glioma from radiation necrosis using magnetically labeled EPCs and MRI.
Abstract ID: 733 Poster board space: 55
Circulating bone marrow-derived stem cells have a tropism for tumors and thus could be used as delivery vehicles for tumor therapy. Previous studies have demonstrated the migration of endothelial precursor cells into the neovasculature in implanted glioma [1,2]. Of the various stem cells that exist within bone marrow, human mesenchymal stem cells (MSC) are particularly attractive because they can be easily isolated, expanded in culture, and genetically manipulated. Quantitative tracking of MSCs within the tumor will further investigate the incorporation of these cells into the tumor [3]. Eight mice were implanted subcutaneously with 1×106 glioma cells into the flank. Four mice (Control) were injected with unlabeled 1×106 MSCs intravenously through the tail vein on the same day as tumor implantation. Four mice were injected three times with 1×106 Ferumoxides protamine sulfated complex labeled MSCs every other day after the implantation (TD). T2* map were acquired approximately two weeks after the inoculation. The tumor volume for the two groups was similar at scan time (87.3±30.7 mm3 vs. 88.8±24.3 mm3, p=NS). Tumor T2* of TD mice was significantly lower than that of the control mice (27.4±5.2 ms vs. 36.0±3.5 ms, p< 0.05). Figure 1A is the scattered plot of tumor T2* of the two groups. The T2* histograms of a TD mouse and a control mouse are illustrated in Figure 1B and 1C, respectively. The results imply that T2* mapping could serve as a potential method for quantitative tracking of magnetically labeled stem cells in vivo.
Abstract ID: 734 Poster board space: 56
Abstract ID: 735 Poster board space: 57
The lack of tumor-specificity and consequent toxicities to the normal brain tissue compromise the therapeutic efficacy of conventional radiation and chemotherapy in patients with malignant gliomas (MGs). We have recently found that high molecular weight melanoma associated antigen (HMW-MAA), a melanoma surface antigen, is expressed in high frequency in human MG surgical specimens. Because of the potential of radioimmunotherapy for selectively irradiating tumor cells, targeting HMW-MAA utilizing radiolabeled monoclonal antibodies (mAb) might be a novel approach of treatment in MGs. The current study investigated the specificity of VT 68.2, a HMW-MAA specific mAb, in targeting this antigen in vitro and in a clinically relevant mouse model of cerebral glioma in vivo.
VT68.2, and an isotype matched (negative) control mAb, MF 11-30, were labeled with I-124, a positron emitter with half-life of 4.2 days, using standard iodogen method. Flow cytometric analysis of GL261 glioma cells with stained mAbs showed equivalent immunoreactivity to radiolabeled and unlabeled VT68.2, while there was no reactivity with MF-11-30. Sequential microPET imaging of C57BL/6 mice with GL-261 cerebral glioma (15 d post-implant) were imaged with a Focus 120® scanner at 24h, 48h and 72 h following a single injection of radiolabeled mAbs. The results showed high-intensity accumulation and prolonged retention of radiolabeled VT68.2 in cerebral glioma lesions. The specificity of this accumulation was demonstrated by i) the absence of uptake of radiolabeled MF-11-30 by the MG lesions, ii) significant decrease in glioma uptake of I-124 labeled VT68.2 in mice pretreated with high dose of unlabeled VT68.2 than in untreated mice, and iii) no visible uptake of I-124 VT68.2 in extra-tumoral brain tissues. Thus, glioma-selective uptake of HMW-MAA specific VT68.2 suggests the feasibility of targeting HMW-MAA in MGs with VT68.2 mAb for radioimmunolocalization and radioimmunotherapy.
Abstract ID: 736 Poster board space: 58
Dendritic cell (DC) based vaccination therapy against glioblastoma multiforme (GBM) is on clinical trial. Reports showed migration of cytotoxic T-cells as well as DC at the site of recurrent GBM in animal models. The purpose of this study was to determine whether magnetically labeled DC can migrate to the site of glioma and whether migrated DC can be detected by MRI. CD34+ hematopoietic stem cells were collected from umbilical cord blood and differentiated into dendritic cells (DC) using GM-CSF, IL-4 and TNF-α. Differentiation of DC was confirmed by the expression of CD86, CD83 and HLA-DR. Differentiated DC were magnetically labeled using ferumoxides-protamine sulfate complexes. Five million magnetically labeled dendritic cells were injected intravenously in rats bearing intracranial glioma (U251) on day 10 after tumor implantation. Rats were imaged by a 3 Tesla clinical MRI system before and 3 days after the intravenous administration of labeled DC. Rats were also imaged by a 7 Tesla MRI system 3 days after administration of labeled DC. Following MRI, rats were euthanized, perfused with saline and paraformaldehyde, and the brains were collected for histology examination. Pre and post gadolinium enhanced FIESTA images (on 3 Tesla) clearly delineated the tumor and its margin. None of the tumors showed low signal intensity areas around the tumor before the administration of magnetically labeled DC. MR images following administration of DC showed low signal intensity areas around the tumors. Low signal intensity areas were seen on images obtained by both 3 and 7 Tesla MRI systems. Prussian blue staining with or without DAB enhancement showed iron positive cells at the margin of the tumors. If DCs are thought to be used against GBM, or as delivery vehicles for gene, mixture of magnetically labeled DC will enable the investigators to track the migration and homing by in vivo MRI.
Abstract ID: 737 Poster board space: 59
Differentiating between the hypointensities caused by the SPIO nanoparticles in the cells and native low signals in tissues may be problematic. To improve the detection sensitivity, alternative approaches that result in positive contrast have been proposed. Stuber et al [1] utilized spectrally selective RF pulses to pre-saturate on-resonant water (Frequency WM) and Seppenwoolde et al [2] achieved positive contrast by dephasing the background signal with a slice gradient (Gradient WM). Recently, a 3D technique [3] was proposed to give optimum positive contrast without prior knowledge (3D WM). This study was to investigate the efficacy of these techniques in vivo using experimental models. Nude rats (n = 6) were inoculated with ferumoxides-protamine sulfate labeled C6 glioma cells [4]. Figure 1A-1D were acquired with a tumor size approximately 0.5mm representing highly concentrated SPIO labeled cells. All three techniques generate positive contrast in surrounding tissues. However, Gradient WM and Frequency WM are highly geometry dependent. The 3D positive contrast technique clearly outlines the tumor border although there are more susceptibility artifacts. As the tumor cells proliferate, the SPIO particles become dilute resulting in a lower concentration. Figure 1F-1G illustrate the positive contrast from Gradient WM and 3D WM corresponds well to the dark regions on the T2* image (1E). The Frequency WM was unable to generate any positive contrast image (image not shown) at the lower SPIO concentration. In conclusion, in vivo MRI using positive contrast techniques have demonstrated the advantages and potential limitations of the positive contrast techniques.
Abstract ID: 738 Poster board space: 60
Dominique Mrugala1, Danielle Noël1, Christian Jorgensen1,
Regenerative medicine suffers from a lack of a high-resolution imaging technique for cells tracking in vivo. This study shows that the Cellvizio can perform tracking of labeled cells both for in vitro evaluation of grafts as well as during longitudinal studies in vivo.
The purpose of the experiment was to detect mesenchymal stem cells (MSC) in sheep knee joints, following an ex vivo exogenous labeling. Prior to this feasibility in vivo study, a comparison of ex vivo results between a standard confocal fluorescent microscope and the Cellvizio was performed.
Abstract ID: 739 Poster board space: 61
Elizabeth J. Akins, Miranda L. Moore,
Cellular adoptive immunotherapy is studied in animal models and clinical trials as a treatment for solid tumors. In a small number of cases when the tumor is rejected it is clear that effector cells were active at the site. When the tumor is not rejected, it is unclear whether functional effector cells reached the tumor.
To detect modulation of effector cells during an immune response to tumors or other pathogens, it is necessary to separately track this cell population in vivo. We developed a system to visualize the movement and function of effector CD8+ T cells within the context of an intact immune system. We constructed lentiviral vectors to express Firefly luciferase behind the murine CD8 promoter and enhancer. Bone marrow stem/progenitor cells were transduced in vitro, and adoptively transferred into lethally irradiated syngeneic recipients. As a result, the CD8+ T cell subset of reconstituted mice was specifically marked with the luciferase gene. We challenged marked animals with a sarcoma tumor that is rejected in immunocompetent mice, and detected an expansion of CD8+ T cells at the tumor site during tumor rejection. We also challenged marked animals with a variant that grows progressively in immunocompetent mice. We detected CD8+ T cells at the growing tumor, indicating that T cells can reach the tumor site.
To determine functional status of CD8+ T cells at the progressively growing tumor, we developed a lentiviral construct where Firefly luciferase expression in CD8+ T cells is controlled by the NF-AT enhancer element, making luciferase expression dependent on signaling through the T cell receptor (TCR). In vitro, luciferase is expressed in CD8+ T cells in a time-dependent manner after TCR ligation. Consequently, we can determine the functional status of CD8+ T cells in vivo, and the efficacy of immunotherapies designed to potentiate the anti-tumor immune response.
Abstract ID: 740 Poster board space: 62
Abstract ID: 741 Poster board space: 63
Noninvasive assessment of therapy efficacy during the early stages would be valuable for facilitating quick therapy modification, evaluating clinical outcome, and developing more effective drugs. There is growing evidence that synovial macrophages play a central role in rheumatoid arthritis (RA). USPIO-enhanced MRI can visualize macrophage infiltrations has been confirmed in animals and patients in several applications: carotid atherosclerotic lesions, stroke, brain tumors, multiple sclerosis, transplanted graft rejections, and RA. In the present study, we examined whether methotrexate (MTX) treatment would alter synovial macrophage content, and whether longitude cellular MRI after in situ cellular labeling with USPIO could be used as an early surrogate marker of treatment response in an adjuvant-induced arthritis. Arthritis was induced in the right knee joint of each rabbits by means of intraarticular injection of methylated bovine serum albumin in 20 knees of 20 rabbits that had been presensitized to the same antigen as described previously. All these 20 rabbits were distributed randomly into 2 groups. One group received intravenous injections of saline (arthritis, untreated; n=10). Another 10 mice were treated with MTX (intravenous injection of 35 mg/kg body weight; arthritis, MTX-treated; n=10). Four nonimmunized rabbits were injected intravenously with saline (no arthritis, untreated; n=4) served as controls. Dextran-coated USPIO particles were administered by direct intravenous injection 48 hours after MTX treatment. Serial USPIO-enhanced MR imaging was performed 24 hours after administration of USPIO at a clinical 1.5T MR scanner. On USPIO-enhanced images obtained 24 hours after contrast agent administration, a significantly higher T2* signal was observed in the synovium of all 10 MTX treated arthritic knees compared to untreated- and control groups (P< 0.05). The results suggested that cellular MRI with USPIO might provide an early surrogate marker for monitoring of treatment response in RA
Abstract ID: 742 Poster board space: 64
Dendritic cells (DC) are a type of professional antigen presenting cell of the immune system. Upon encountering antigens in peripheral tissues, DC migrate to draining lymph nodes (DLN) where they stimulate naïve T cells and initiate the immune response. Cellular based cancer vaccines employ autologous ex vivo generated DC loaded with tumor antigens. Although the optimal route for administering DC vaccines is debatable, intradermal (ID) delivery is being investigated in a number of clinical trials. Here, we utilized microneedle-based delivery devices for ID administration of autologous DC and describe a novel method to monitor the migration of the injected cells from the skin to the DLN in a swine model. Mature DC were generated from porcine blood and characterized for expression of various cell surface markers by flow cytometry. DC were then labeled with the lipophilic membrane dye, DiR, and monitored by near infrared imaging. We observed in vivo migration of labeled DC in real time with DiR signal appearing in the DLN within 20 minutes of the injections. The signal in the DLN continued to increase over a 24 hr period. At 24 hr post injection, DLN were excised and single cell suspensions prepared to quantify migration efficiency by flow cytometry. Cells co-expressing DiR and MHC class II were identified, thus allowing for determination of migration rates of DC from injection sites to the DLN. Migration rates varied from 1–7% and are comparable to published rates for radiolabeled migration studies in humans. The infrared imaging combined with flow cytometry establishes a new method for evaluating parameters to improve delivery of DC cancer vaccines in real time in a large animal model.
Abstract ID: 743 Poster board space: 65
The development of novel imaging technologies for cellular tracking is a major research goal of modern biological research. Fluorescence-based imaging with near infrared (NIR) light is potentially well suited for this application. Noninvasive optical imaging enables quantitative in vivo detection of fluorophore-laden cells, at centimeter tissue depths, without disturbing cellular function. NIR-emissive polymersomes are self-assembled, biocompatible polymeric vesicles incorporating porphyrin-based NIR fluorophores (NIRFs) within their hydrophobic bilayered membranes. Here, we conjugate a cell permeable peptide, Tat, to the polymersome surface to enable efficient intracellular uptake by both immature and mature dendritic cells (DCs). DCs play a pivotal role in immune tolerance and the initiation of immunological responses. The ability to track DCs in vivo is imperative for the development of DC-based cellular therapies and to advance our understanding of DC function and pathophysiology. Using flow cytometry and confocal microscopy, we demonstrate that Tat-mediated NIR-emissive polymersome delivery to DCs is concentration and time dependent resulting in punctate intracellular localization. Further, loading cells with Tat-NIR-emissive polymersomes does not interfere with cytokine induced DC maturation and has modest effects on DC viability and DC-induced activation of naïve T cells. Significant uptake of NIR-emissive polymersomes is observed when conjugated to the Tat peptide, with a lower detection limit of 5000 labeled DCs in a single well or approximately ≈ 150 DC/mm2. The extent of polymersome delivery is estimated as 11,000 ± 2,000 polymersomes/DC, equivalent to 0.1 ± 0.02 fmol of NIR fluorophore. Our studies highlight the potential for utilizing NIR-emissive polymersomes to track DCs, noninvasively, by fluorescence-based in vivo imaging.
Abstract ID: 744 Poster board space: 66
There is evidence of an accumulation of primed dendritic and sensitized T-cells at the site of glioma, suggesting an initiation of immunity. If intracranial glioma initiates an immunogenic reaction, there should be differential migration and incorporation of magnetically labeled sensitized splenocytes (T-cells) into glioma, but not in radiation necrotic areas. The purpose of this study was to differentiate glioma (9L) from radiation necrosis using ferumoxides-protamine sulfate (FePro) labeled sensitized splenocytes. Labeled rat splenocytes were injected intravenously in rats bearing an intracranial gliosarcoma (9L) or radiation necrosis. Cells were injected 11 days after tumor implantation or irradiation and MRI was obtained 3 days after the injection of cells (day 14). Multi-echo T2-W, T2*-W and 3D gradient echo MRI were obtained by a 7 Tesla MR system. Following in vivo MRI, rats were euthanized, perfused, and the brains were collected. Randomly selected rat brains also underwent ex vivo high resolution MRI. Collected whole brains were further fixed and snapped frozen for sectioning. Sections were stained for blood vessels using FITC labeled lectin. Prussian blue staining was performed to detect the iron positive cells. Distributions of blood vessels and patterns of neovascularization at the site of radiation were compared with that of implanted tumor and surrounding brain. Both in vivo and ex vivo MRI showed low signal intensity areas at the margin of the tumors in rats injected with labeled cells. Iron positive cells were present at the corresponding sites of low signal intensities seen on MRI. There was no definite low signal intensity area indicating accumulated iron positive cells seen at the site of radiation injury and Prussian blue staining also did not show any iron positive cells. By targeting immunogenic reaction, it is possible to differentiate glioma from radiation necrosis using magnetically labeled sensitized splenocytes (or T-cells) and MRI.
Abstract ID: 745 Poster board space: 67
Abstract ID: 746 Poster board space: 68
The therapeutic use of microglial cells has recently gained attention for the treatment of Alzheimer disease (AD), however, few non-invasive techniques have been applied for investigating the cells after administration. We have exploited the applicability of MR tracking of magnetically labeled microglia following arterial infusion for Alzheimer model rat. Solution of β-amyloid 42 (Aβ42), which was preincubated to make aggregations, was injected into the left hippocampus and saline into the right hippocampus of rats. A mouse microglial cell line expressing EGFP was used in this study. For the paramagnetic labeling of the cells, we employed the hemagglutinating virus of Japan envelope (HVJ-E, GenomONE, Ishihara Sangyo, Japan) to transport Resovist (Schering) into the cells. The labeled cells were injected through the internal carotid artery into anesthetized male Wistar rats one day (n=5) or 3 days (n=5) after the injection of Aβ42. All experimental procedures were approved by the Committee on Animal Care of the Shiga University of Medical Science. One day after the microglial injection, MR images were obtained with a Varian 7T Unity Inova. MR images detected accumulation of administered labeled microglial cells in the Aβ-injected sites within 24 hours, in contrast, no significant change in MRI was detected in the saline-injected right side. After the MR measurements, the brains were fixed and processed for the counting of fluorescent cells, and Prussian blue and 4G8 double staining. Cell counting revealed that significantly more EGFP-positive microglial cells were found on the Aβ-injected side than on the saline-injected side. In addition, microglial cells expressing EGFP were also observed in the Aβ mass in the hippocampus. This study demonstrates the usefulness of MRI for non-invasive monitoring of dynamic behavior of exogenous microglia, and suggests a promising future for microglia/macrophages as therapeutic tools for AD.
Abstract ID: 747 Poster board space: 69
To obtain positive contrast from transplanted cells in the potentially hypo-intense contrast region of the infarcted tissue, we have generated two different Gd-bearing contrast agents (T1 = approximately 150 ms). Various incubation times and CA concentrations were used to label murine myoblasts.
The two different CA's showed significant differences in labeling efficiency of cells and in MR relaxivity. Under non-toxic labeling conditions, intracellular concentrations of Gd of 10–90 pg per cell were achieved. These amounts of intra-cellular Gd allowed for visualization of the cells in vitro and in vivo resulting in positive contrast.
Abstract ID: 748 Poster board space: 70
Abstract ID: 749 Poster board space: 71
Endothelial progenitor cells, which are derived from adult hematopoietic stem cells, are postulated to localize to areas of injury or ischemia to facilitate vasculogenesis. Techniques for precise, non-invasive stem/progenitor cell tracking with MRI would enable further investigation of their regenerative capacity as they migrate to and within living tissues. While most current cell labeling strategies for MRI use iron oxide particles to produce contrast on standard 1H images, we have developed a liquid perfluorocarbon (perfluoro-15-crown-5 ether) nanoparticle that is readily endocytosed by stem/progenitor cells and permits 19F imaging and spectroscopy for tracking with no endogenous background signal. To demonstrate this concept, stem/progenitor cells were derived from mononuclear cells isolated from human umbilical cord blood. After 7–14 days in endothelial medium, cells were incubated for 12 hours with rhodamine-labeled crown ether nanoparticles created in our lab as previously described. Flow cytometry and confocal microscopy revealed high labeling efficiency and abundant internalization of nanoparticles. Labeled cells (≈ 2times106) were then injected into the hindlimb of an anesthetized balb/c mouse. Proton scanning was conducted for localization and identification of the injection site (3D TSE sequence) with a 1.5 T Philips Intera CV scanner. Fluorine images were then acquired for definitive localization of the perfluorocarbon-labeled cells using a custom-built 2.5 cm diameter surface coil (sagittal projection image, balanced SSFP sequence, 2.3times2.3times15 mm resolution, ≈ 7 min scan time). As shown in the figure, the 19F signal, generated in a relatively short scan time, allowed precise in-plane localization of the cells at a clinical field strength. Therefore, this technique, which combines high-resolution 1H anatomical imaging with the unique 19F signature, may be useful in studies designed to track cell migration in vivo.
(Lift) 1H image of a balb/c mouse, where the injection tract for labeled cell injection can be identified in the leg (arrow) (Middle) 19F image from the same mouse. The signal indicates the presence of the labeled cells. (Right) Overlay of the 19F signal onto the 1H image.
Abstract ID: 750 Poster board space: 72
Adoptive transfer of antigen-specific cytotoxic T lymphocytes (CTL) has shown to be an effective therapy for the treatment of cancer and viral infection. Infusion of Epstein-Barr virus (EBV)-specific CTL has been particularly effective for the treatment of post-transplant EBV-associated lymphomas. However, use of EBV-specific CTL for the treatment of EBV-positive Hodgkin's disease (HD) and nasopharyngeal carcinoma (NPC) has been less successful. One reason may be the inefficiency in which CTL migrate to tumor sites. While there is evidence that CTL reach tumor sites and kill tumor cells, many of these cells may become trapped in other organs (e.g. lung, liver) reducing the efficacy of the treatment. Consequently, in order to anticipate response to immunotherapy, it is important to determine the in vivo distribution of CTLs both immediately and in the long term using noninvasive imaging methods.
We labeled CTL with ICG (100 ng/ml for 15 minutes at 37°C) and tested cell function and for NIR fluorescence. ICG labeling of EBV-specific CTL did not affect phenotype, viability or cell function (e.g. cytotoxicity, IFN-γ production by ELISPOT). Additionally, specific uptake of ICG by activated CTL was calculated to be 1.5times10−7 ng per cell, making it possible to calculate the low ICG dose that patients would receive during CTL therapy.
In preclinical imaging of tumor-free animals, we injected 10 million labeled CTLs i.v. and conducted NIR imaging showed selective accumulation of stained cells in the mouse circulation system at early stages and redistribution to the other organs over 24 hours (Figure 1). NIR imaging of tumor bearing animals is currently underway. Strategies to image CTL distribution over the long term after administration will be discussed.
Approximately one hr after i.v. administration of 10 million ICG labeled CTLs in a non-tumor bearing animals, CTLS tissue distribution is within the gut. Free CardioGreen dye has a different distribution than exhibited by these studies.
Abstract ID: 751 Poster board space: 73
We addressed three FDA requirements for preclinical data to support an IND application for a phase I clinical trial of magnetically labeled PBPC (CD34+/AC133+) for magnetic resonance imaging (MRI) in patients with malignant gliomas. FDA's requirements were to 1) develop a clinical scale method for magnetic cell labeling with ferumoxides-protamine sulfate (FEPro) using reagents and methods suitable for manufacturing in our CGMP facility; 2) perform survival analysis and MRI of escalating cell doses in mice with intracranial gliomas; and 3) evaluate donor-origin cell biodistribution by histopathology. RT2 gliomas were stereotaxically implanted in three groups (G) of 30 SCID mice each, receiving 3times105 (G1), 3times106 (G2), or 3times107 (G3) PBPC respectively. PBPC enriched for CD34+/AC133+ cells using immunomagnetic selection were prepared for labeling and half were labeled with FEPro. On day 2 after tumor implant, labeled cells (n=10), unlabeled cells (n=10) or saline (n=10) were administered i.v. to each of the 3 cell dose groups. Seven mice per group were then imaged on day 9–12 after PBPC infusion (day 10–13 of tumor growth) at 7T with the rest used for survival endpoint. Mice were necropsied for histological exams. There were no significant differences in survival when comparing saline treated animals to those receiving labeled or unlabeled cells for each group (Figure 1). MRI showed a similar effect for labeled animals over all dose groups with linear hypointense features in and around the tumor compared to sham or unlabeled EPC transplanted animals. These data, which show comparable outcomes for imaging and mortality in the labeled, unlabeled control, and saline treatments, support the safety of administering FEPro labeled PBPC in human clinical trials.
Abstract ID: 752 Poster board space: 74
Abstract ID: 753 Poster board space: 75
A) Prussian Blue and B) FITC Anti-Dextran staining of GM-CSF secreting cells labeled with SPIO (40X). C) T2* image of vaccinated mouse leg. SPIO labeled cells induce dark contrast in the footpad (indicated with arrow). D) Control leg.
Abstract ID: 754 Poster board space: 76
In vivo tracking by SPECT of stem cells labeled in vitro with 111In-tropolone is a promising technique, but requires connecting the presence of an 111In-signal to the presence of viable stem cells. This in turn requires knowing the fate of the 111In label following cell death. In an in vitro experiment, two sets (n=5) of mesenchymal stem cells (MSC) were labeled with 111In-tropolone and then killed by using an ultrasound probe to rupture the cell membrane or by freezing the cells at −80°C for 60 min. The resulting cell debris and supernatant were used to label two new sets of MSC. The labeling efficiency was compared to a control labeled with 111In-tropolone. In an in vivo experiment, 107 MSC labeled with 1.28MBq 111In-tropolone were killed using an ultrasound probe and then injected into the healthy myocardium of a canine. Ten SPECT images were acquired over 18 hrs post-injection. The animal was then sacrificed and the 111In remaining in the heart-tissue samples was measured in a HPGe-well-counter. In the in vitro experiment, the labeling efficiencies were found to be (0.2±0.5)% and (0.3±0.5)% for cells labeled with ultrasound-killed and frozen-killed cells respectively, while the labeling efficiency of the control cells was found to be (53.0±2.0)% for a 30 min cell culture. In the canine experiment, the tissue samples indicated 1% of the injected activity remained in the myocardium 18 hrs after injection. The 111In SPECT images showed a biological half-life of 2.6 hrs, much shorter than the 6-day half-life we have previously observed for the injection of labeled live MSC. We conclude that 111In leaked from dead MSC does not label other viable MSC and will be rapidly cleared from the myocardium. This implies that the 111In signal seen when tracking transplanted stem cells is associated with viable cells.
Abstract ID: 755 Poster board space: 77
Abstract ID: 756 Poster board space: 78
The relative MR signal intensity of solutions with iron oxide labeled macrophages was compared with that of PBS containing the same amount of SPIO.
Abstract ID: 757 Poster board space: 79
Abstract ID: 758 Poster board space: 80
Abstract ID: 759 Poster board space: 81
Monocytes play an important role in the pathogenesis of atherosclerosis but their spatial and temporal involvement is largely unknown. This study used scintigraphic and autoradiographic techniques to quantify monocyte accumulation in atherosclerotic lesions. Peripheral blood monocytes from C57BL/6 mice were isolated by negative selection, labeled with [111In]-oxine and injected IV to syngeneic 10 and 20 wk old Apolipoprotein E-deficient (ApoE–/–) mice. Negative selection of monocytes yielded a pure population (>90%) that, upon labeling, was viable (as assessed by MTS assay) and in a resting state (as assessed by flow cytometry). Scintigraphic and autoradiographic imaging of excised aortas revealed that 5 days following adoptive transfer monocytes had been recruited largely to the root of the aorta, an area with abundant atherosclerotic lesions. Accumulation tended to be highest in 20 wk old mice on high cholesterol diet (HCD), compared to mice on regular chow diet. In addition, 10 wk old chow-fed ApoE–/– mice without mature lesions displayed greater signal than age matched chow-fed C57BL/6 controls. To corroborate these findings, immunohistochemical analysis was performed and demonstrated accumulation of GFP+ monocytes in 20 wk old ApoE–/– mice on HCD. These data demonstrate that peripheral blood monocytes rapidly accumulate in the aortic root of atherosclerotic mice, and that their accumulation precedes the development of histologically identifiable lesions.
Abstract ID: 760 Poster board space: 82
A. SPECTICT imaging on day 6 after CTL-TK GFP trasnfer B. Bioluminesence imaging of TC-1 tumor-bearing mouse (left) and the naïve mouse (right) on day 3 post-injection of CTL-FLUC
Abstract ID: 761 Poster board space: 83
Abstract ID: 762 Poster board space: 84
(A)Flow cytometry of labeled cells Gate A is the less granular T cells, gat E as the more granular T cells, Due to the particles, well labeled cells become more granular. (B) Representative heavily labeled cell from gate E. MPIOs are green nucleus is stained blue- with DAPI. (C) Representative lightly labeled cell from gate A Each T cell measures approximately 10 microns in diameter. (D) MRI of lymphocytes Individual well labeled lymphocytes treats a hypointense contrast region 300 micron across, even though they themselves are only 10 microns in diameter. Signal reductiot in the center pixel was 84–91% for the Largest dark spots. The scale bar is mm.
Abstract ID: 763 Poster board space: 85
Non-invasive imaging of T cell trafficking is an important tool for preclinical studies to gain a better understanding of the effector phase of T cell mediated autoimmune diseases. In comparison, T cell labelling was performed by small iron particles for visualization by magnetic resonance imaging, by a fluoroprobe for optical imaging, by radio labelling with [Cu-64]PTSM and [In-111]oxine to detect them by positron emission tomography (PET) or single photon emission computed tomography, respectively. We compared all four different labelling methods by using interferon-γ (IFN-γ) producing ovalbumin (OVA) specific Th1 cells and assessed viability (Trypan blue staining), OVA-specific T cells functioning (specific IFN-γ expression by ELISA) and double strand brakes (phosphorylated histone H2AX-γ) of the cells at different time points after labelling. For each experiment 10^6 Th1 cells were incubated for 6 hours with iron particles, for 1 up to 5 minutes with a Cy5 dye, for 15 minutes up to 4.5 hours with [Cu-64]PTSM (740–2260 kBq), and for 15 minutes with [In-111]oxine. Labelling of the Th1 cells with iron particles (SPIOS) did not significantly influence viability or functionality of the cells. However, an internalization of the SPIOS into the cells could not be reached, even by using JetPEI as transfecting agent. Cy5 labelling did not influence specific IFN-γ secretion or viability. 24 hours after [Cu-64]PTSM and [In-111]oxine (740 kBq) labelling specific IFN-γ expression was reduced to 20%. A significant increase (factor of 10 compared to unlabeled cells) of double strand brakes was measured for the radio-labeled cells. Although radiolabelling of Th1 cells influences specific cell functions, in vivo high sensitivity nuclear imaging is still possible for the first 48 hours after cell labelling. MicroPET studies using Cu-64 labelled cells were performed in different mouse models showing the dynamic of cell migration into single lymph nodes, in vivo.
Abstract ID: 764 Poster board space: 86
Novel strategies that enable long term monitoring of mesenchymal stem cells (MSCs) are needed to dynamically track them in vivo for long periods of time. We presently label MSCs in vitro with iron oxide particles, using established techniques. Two prostate tumor models with significantly different growth rates were generated in mice, and labeled MSCs were administered intravenously. Incorporation of MSCs into the vasculature of mouse prostate cancer models was tracked with MRI. Although the in vitro iron oxide labeling protocol resulted in efficient cell labeling for MRI, such labeling is transient because of dilution of the label during subsequent cell divisions. In order to overcome this problem, we are developing a novel genetic approach, which will result in engineered MSCs stably expressing a unique cell surface receptor. We constructed a recombinant plasmid pEF-1α-GFP-IRES-δCD4, which allows for the simultaneous expression of GFP and the cell surface receptor δCD4. GFP facilitates fluorescence microscopy while the receptor can be imaged with an antibody-targeted MRI contrast agent. At this time we have demonstrated stable expression of GFP in transfected cells. Our next step is to track these cells in vivo with intravascular targeted MR contrast agent. This approach will be validated against an established MRI detection strategy that uses precursor cells pre-labeled with iron-oxide. Circulating MSCs localize in the tumor vasculature where they are accessible to highly efficient MR targeted contrast agents such as iron-oxide nanoparticles. We believe that this new approach can provide significantly improved detection sensitivity at later time points when other labeling techniques might fail due to dilution of the label.
Transection strategy to generate MSC with double GFP/δCD4 expression
Abstract ID: 765 Poster board space: 87
Hui Qiao,
Stem cell based cellular cardiomyoplasty represents a promising therapy for myocardial infarction. Labeling stem cells with superparamagnetic iron oxide (SPIO) particles represents a simple and potentially clinical applicable method for tracking these cells by MRI. The effect of iron loading on proliferation and cardiac differentiation of embryonic stem cells (ESCs) is unknown. To clarify this issue, murine ESCs were incubated with different size and concentration of SPIO: nanometer-sized Feridex and micron-sized Bangs particles; two concentrations (12.5 or 50 μG Fe/ml) were used for each type of SPIO particles. Cardiac differentiation was induced by formation of embryonic bodies (EB); cardiac differentiation became visible when patches of cells within an EB were undergoing synchronized contractions (beating) and was further confirmed by immunohistological staining of cardiac markers. The number of beating regions in EBs was counted for each SPIO concentration. Viability and proliferation of labeled ESCs were measured by MTT assay. Cellular iron content was assessed by ICP-MASS: 1.2 and 5.6 pg Fe/cell was obtained for Feridex labeling corresponding to the above concentrations; 2.1 and 5.4 pg Fe/cell resulted from labeling with Bangs particles. The relatively low labeling efficiency is the result of small size and nonphagocytic characteristics of ESCs. Feridex labeling did not seem to alter viability, proliferation and cardiac differentiation of ESCs compared with the unlabeled controls. For ESCs labeled with Bangs particles, there was a marked inhibition of proliferation and viability especially at 50 μg Fe/ml. However, cardiac differentiation was not affected: percent of beating EBs was slightly higher in Bangs labeled ESCs compared with controls. This is the first report that iron overloading (≤ 5.4 pg/cell) does not affect cardiac differentiation of ESCs. These results suggest that iron overload may affect ESCs proliferation and differentiation through different pathways. Grant support: NIH R21EB002473 and R01HL081185.
Abstract ID: 766 Poster board space: 88
Left panel: X-ray CT showing location of transplanted cells. Right panel: Multi-spectral SPECT showing transfected cells, radiolabeled in vitro with 111In (green), transplanted into myocardium. Blue scale shows distribution of 1311 reporter probe, radiolabeling transplanted cells in vivo, but accumulating non-specifically in surrounding tissue. 99mTc-Sestamibi - a marker of perfusion - was injected to identify myocardium (label M, red scale).
Abstract ID: 767 Poster board space: 89
Abstract ID: 768 Poster board space: 90
Hui Qiao1, Datta Ponde1, Seok Rye Choi1, Paul Acton2, Hank Kung1, Victor Ferrari1,
Recent study showed that stem cell therapy is a promising method for regeneration of the injured tissue. However, monitoring stem cell in vivo remains problematic due to limitations and invasion of conventional histological assays. In order to explore non- invasive imaging to track the survival status and location of ESCs that are transplanted in the heart, we labeled murine embryonic stem cells (ESCs) by stably transducing them with a reporter gene: HSV1-tk (herpes simplex virus type 1 thymidine kinase) or its mutant form, HSV1-sr39tk, detectable by PET imaging. In vitro cell uptake study showed that PET tracer [F-18]-FHBG was preferably accumulated in ESCs expressing sr39tk whereas [I-125]-FIAU was preferably taken up by ESCs expressing tk. Seven million of sr39tk expressing cells were grafted in the rat heart; a perfusion tracer, 1.0 mCi [N-13]ammonia, was injected intravenously (i.v.) followed by data acquisition using a micro-PET system (A-PET); N-13 signal is shown in grey scale and serves to delineate the myocardium in long (LA) and short axis (SA) views; 1.5 mCi [F-18]-FHBG was then injected i.v. and cardiac PET images were acquired one hour later; strong tracer uptake by the cells grafted in the heart was shown in color scale. We have further performed autoradiographic studies that confirmed the imaging signals. The technique opens up the possibility of in vivo tracking the survival status of the cells over time since nonviable cells will not express the reporter gene, therefore, will not be detected. Grant support: NIH R21EB002473 and R01HL081185.
Abstract ID: 769 Poster board space: 91
Abstract ID: 770 Poster board space: 92
Suren Soghomonyan, Uday Mukophadhay, Frank Marini III, Steven Kornblau, Mian Alauddin,
Abstract ID: 771 Poster board space: 93
With increasing concern for insertional mutagenesis by transducing stem cells with retroviral vectors or even lentiviral vectors, low multiplicity of infection could decrease chances of unwanted chromosome insertions. However, low copy-number of transgene in stem cells may subject to viral silencing, resulting in low or none transgene expression. We propose that overcoming viral silencing with sustained transgene expression can be achieved and even enhanced with selection pressure. We have generated a bicistronic lentiviral vector containing MGMT-P140K gene and firefly luciferase gene linked by 2A self-cleavage sequence of FMDV. MGMT-P140K has been shown to provide protection against chemotherapeutic drugs, including Temozolomide (TMZ). In vitro transduction of human and murine cell lines with MOI of 1 did not show any loss of transgene expression for over 13 weeks. Transduction of primary murine whole bone marrow cells and lin− cells with MOI of 0.5–1 showed sustained transgene expression of 4–5% of cell population after two weeks of ex vivo expansion. However, in murine bone marrow transplant model, we have observed dynamic transgene expression patterns from transduced bone marrow cells in both non-myeloablated and lethally irradiated recipients by bioluminescence imaging. Clonal expansion and engraftment can be detected at day 6 with strong bioluminescence signals in bone marrows. At day 12-14, transgene expression can be detected in spleen in most of the animals. By day 17-20, transgene expression in many of the recipients faded away. CFU analysis of bone marrow cells from those mice indicated a decrease of transgene positive progenitor cells. However, when recipient mice received BG+TMZ treatment 3 weeks after transplantation, and subsequent treatments every 3 weeks, transgene expression is sustained and even enhanced over 3 months. This drug selection approach can abate potential viral silencing in stem cell transduction and maintain transgene expression.
Abstract ID: 772 Poster board space: 94
Abstract ID: 773 Poster board space: 95
Peta O'Connell,
The exquisite antigen processing and presenting capacity of dendritic cells (DC), together with their affinity for trafficking into secondary lymphoid tissues, imbues DC with a powerful therapeutic potential. DC vaccine-based anti-tumour immunotherapy in the clinic is feasible and non-toxic. Despite promising trials, significant barriers remain. In general, fewer than 5% of ex vivo generated DC reach draining lymphoid tissues. At present it is not possible to assess this critical aspect of DC based immunotherapy without direct sampling of the target tissue. Sensitive and specific non-invasive imaging techniques are key to gaining important information about the fate of DC vaccines in vivo and will facilitate the development of effective DC-based immunotherapies.
We are developing in vivo techniques to monitor migration of ex vivo generated DC vaccines following injection in a mouse model, using a clinical grade 1.5T MR scanner with custom built gradient insert and RF coils. Using EGFP+ bone marrow-derived DC, the effect of USPIO, SPIO and MPIO uptake by DC was examined. DC exhibit differential uptake of SPIO (Feridex) and MPIO (Bangs), but USPIO were not ingested. More importantly, MPIO uptake altered DC morphology as assessed by microscopy and flow cytometry. In addition, MPIO uptake was associated with an enhanced state of DC activation/maturation, while DC-labeled with SPIO were similar to untreated DC with respect to DC activation/maturation markers as determined by flow cytometry. Initial ex vivo studies detected SPIO-labeled DC migration from the footpad to the draining popliteal lymph node. By assessing the consequences of SPIO/MPIO labeling on DC, we hope to develop MR techniques that will allow assessment of DC migration in humans, and lead to more effective cancer vaccine therapies.
Abstract ID: 775 Poster board space: 133
Jed E. Rose1, Sudha Garg2, Stephen J. Lokitz1, Timothy G. Turkington1, Holly C. Smith2, Joseph E. Herskovic1, Frederique M. Behm1, Richard C. Minton2, Sarah Wildrick1,
Cigarette smoking is highly addictive and it has been suggested that the rapid absorption of nicotine inhaled into the lungs provides a reinforcing dose to the brain within 7–10 seconds. Previous research measuring arterial nicotine concentrations has indicated that the lungs may serve as an initial depot of distribution, slowing entry of nicotine into the arterial circulation. Therefore, characterizing the time course of inhaled nicotine may have important implications for understanding nicotine addiction and for the design of effective nicotine replacement therapies to aid smoking cessation. The aim of this study is to characterize the time course of nicotine delivery into the lungs, arterial circulation and its subsequent accumulation in the brain when inhaled in cigarette smoke. Cigarette smokers were presented with a single puff of smoke from a cigarette that had 5 mCi of C-11 nicotine deposited into the tobacco rod. PET scanning commenced prior to inhalation and continued for 9 minutes. Results from 12 subjects showed that nicotine remained in the lung compartment for several minutes with a mean t 1/2 of elimination = 247 ± 51.7s. Time to reach 90% of peak brain levels averaged 126 ± 52.5 s, longer than the 7–10 s frequently cited. The time to peak nicotine levels in the brain is also longer than the typical inter-puff interval exhibited by the smokers (30-60 s), suggesting that brain nicotine levels continue to increase over the course of smoking a cigarette. These results imply that some nicotine delivery systems (e.g. nasal spray) could deliver nicotine as rapidly as cigarettes, and that factors other than pharmacokinetics (e.g. irritation) may be important in influencing their acceptability and efficacy. Our methodology could also aid in studying the inter-individual differences among smokers and in correlating nicotine pharmacokinetics to differing degrees of dependence.
Abstract ID: 776 Poster board space: 134
Image reconstruction algorithm based on statistical method such as the maximum-likelihood expectation-maximization (MLEM) algorithm is being commonly used in positron emission tomography (PET). In clinical use, the two-dimensional Ordered Subsets Expectation Maximization (2D-OSEM) method is most commonly used not only for acquired in 2D mode, but also in 3D mode. To apply 2D-OSEM, 3D acquired data are necessary to convert into 2D sinogram by an approximation method such as Fourier Rebinning (FORE) algorithm.
Recently, a fully 3D Image Reconstruction method using maximum a posteriori (MAP) estimation was developed and released as a part of standard software of a microPET system. This 3D MAP reconstruction algorithm is using 3D Gibbs prior, blur kernels model of Point Spread Function (PSF) and the system geometrical effects.
MicroPET P4 (CTI Molecular Imaging, Inc., Knoxville, TN., USA) has 4 rings of 42 detectors in each, producing a 22-cm diameter by 7.8-cm axial extent cylindrical detector geometry. Data are acquired in list mode. The system operates only in full 3D mode, permitting coincidences between any two detector rings. The system is composed of 168 detector modules, each with an 8 × 8 array of 2.2 × 2.2 × 10 mm lutetium oxyorthosilicate (LSO) crystals, arranged as 32 crystal rings 26 cm in diameter.
To perform 3D MAP reconstruction, parameters must be determined, number of iterations and beta value. The beta is the hyper-parameter of the Gibbs prior, and works as a smoothing function. The purpose of this study is to determine the relationship between the reconstruction parameters (the number of iterations and beta value) and the reconstructed image quality. Image power spectrum was used to evaluate the image noise and correlation of pixel values between 3D MAP reconstructed images and those reconstructed by using filtered back projection algorithm as a gold standard for quantitative evaluation.
Abstract ID: 777 Poster board space: 135
Ultrasound contrast agents consist of air-filled microparticles (MPs). Due to the high sensitivity of ultrasound for contrast agents even single MPs can be detected in vivo. Recently, experimental polymer-stabilized microparticles with a streptavidin linker system for antibody conjugation have been applied for molecular imaging. The aim of this study is to evaluate the pharmacodynamics of those experimental MPs.
For this purpose, MPs were labelled radioactively with 111indium and their pharmacodynamics were investigated in nude mice bearing subcutaneous squamous cell carcinoma xenografts. Distribution of radioactively labelled MPs was observed with a SPECT camera in vivo (whole animal) and with a γ-counter in vitro (selected organs). Additionally, dispersion of fluorescent labelled MPs was analysed using immunofluorescence staining of selected organs.
Highest accumulation of MPs was found in liver and lung followed by the spleen. Surprisingly, blood levels were much lower, pointing to an early retention of MPs in these organs. While in liver and spleen there was a constant plateau of MP concentrations over the entire examination period of 48h, concentrations in the lung rapidly decreased. The decreasing MP concentrations in the lung, however, were accompanied by a short increase of MP concentration in the blood within 1h. Subsequently, MPs were rapidly cleared from the blood. Gastrointestinal tract and tumour showed almost no accumulation of the MPs.
In contrast to previous assumptions there is no uniform and exclusive distribution of MPs in the blood, but an early retention in liver, spleen and lung. While the retention in liver and spleen can be explained by deposition of MPs in the interstitial space and in cells by internalisation, the pooling in the lung is most probably related to temporary retention in the pulmonary-arterial system. Slow redistribution of the particles from the lung may be responsible for the late increase in blood concentration at 1 hour.
Abstract ID: 778 Poster board space: 136
The objective of this study was to develop performance characterization and quality control protocols for a commercial small animal SPECT/CT system based on pixellated crystals (XSPECT®, Gamma Medica-Ideas).
The XSPECT® system consists of two 12.5 cm × 12.5 cm modular cameras each consisting of an 85×85, 6 mm thick pixellated NaI(Tl) crystal array in combination with a 5×5 array of small position sensitive photomultiplier tubes (PSPMT) for readout. The cameras can be fitted with interchangeable parallel hole or pinhole collimators.
The NEMA NU 1–2001 recommendations for conventional scintillation cameras were modified to adapt to the higher resolution of the modular cameras with pixellated crystals. By scanning a narrow collimated photon beam from a 99mTc point source, we determined the size of the pixellated crystal and width of the gaps between adjacent crystals to be 1.4 mm and 0.2 mm, respectively. A test pattern of parallel line sources was used to determine the intrinsic spatial resolution and energy resolution. The system intrinsic flood field uniformity was determined using a 99mTc source.
We utilized 57Co point sources to test the system alignment, on-axis resolution and sensitivity for SPECT imaging. The uniformity in the SPECT images was measured using a cylindrically shaped phantom. The count rate capability was evaluated using known increments of 99mTc activity. The extrinsic resolution of the pinhole SPECT system was determined from the FWHM of the reconstructed images of a 99mTc line source in air and in a scatter medium. Finally, we designed daily, weekly, monthly and semi-annual measurements for quality control purposes.
In conclusion, we demonstrate that specially designed measurements and quality control protocols are required to fully characterize the performance of a commercial SPECT/CT system consisting of modular cameras based on pixellated crystals and to ensure consistent performance of small animal SPECT imaging.
Abstract ID: 779 Poster board space: 137
To investigate the role of Toll-like receptor (TLR) signaling in acute and adaptive immunity to vaccinia infection, we measured induction of innate immune response and CD8+ T-cell activation in response to a virulent strain of vaccinia virus (Western Reserve) and a corresponding knockout of viral gene A46R, a potent inhibitor of TLR signaling. For these studies we constructed a novel tri-functional reporter for in vivo imaging and immunologic analysis, GFP-SIINFEKL-luciferase (GSL), which combines bright fluorescence and bioluminescence signal with a model antigen, SIINFEKL, useful for monitoring CD8+ T-cell activation in response to viral infection. The GFP reporter facilitates rapid identification of recombinant virus and permits microscopic and flow cytometric analysis to identify infected cells in tissues from infected animals. The bioluminescence reporter allows us to monitor viral replication and clearance as well as anatomic localization of virus throughout the course of infection. We confirmed efficient presentation and recognition of the antigen in the context of the tri-functional reporter. We then incorporated this trifunctional reporter into either a privileged site in strain WR to retain wild-type virulence or replaced the full open reading frame of A46R with the reporter. We report data for cytokine induction in cell culture and lung in response to intranasal infection, and correlate these data with viral replication and CD8+ T-cell activation.
Abstract ID: 780 Poster board space: 138
We have designed a new adenoviral PET ligand-reporter gene system utilizing [18F] 16α-fluoroestradiol (FES) and human estrogen receptor α ligand binding domain (hERL) which can work in multiple tissues including brain, with efficient expression and safe human use [1]. Now, we report its potential application to different species.
An in vitro FES binding assay performed on infected Cos-7 cells showed more than 50-fold uptake with the test adenovirus than the control. An in vitro expression assay was performed in single cell phenotype of diverse species, using lung fibroblast cell lines of mouse, rat, guinea pig, rabbit and human to evaluate species-based differential expression. The highest expression was observed in human, with moderate levels in rat, rabbit and guinea pig and minimal expression in mouse cells, suggesting variable infection efficiency of our recombinant adenovirus. In vivo studies were performed by injecting the test and control viruses into opposite limbs of mouse calf muscle. After 48hrs, 400 μCi of FES/mouse was injected via tail vein and autoradiography (30 μm slice thickness) was performed from 6 mice samples. The side injected with test adenovirus showed higher accumulation of FES than the other side. The averaged peak radioactivity (PSL) ratio, quantified from test and control slices of the 6 mice, was 1.6±0.2 with a p-value of < 0.01 as determined by paired Student's t-test. We expect in future, the small animal PET or clinical PET imaging trials using our system in higher animal models will yield better expression, since we could get promising results in spite of low coxsackie and adenovirus receptor (CAR) expression in the murine tissues [2]. Our study demonstrates the prospect of using hERL as an effective PET reporter gene for monitoring of gene therapy.
Abstract ID: 781 Poster board space: 139
Polymeric solid nanoparticles that can selectively bind to cell surface molecules and generate contrast enhancement for ultrasonic imaging are being developed. Previous studies have shown that submicron perfluorocarbon (PFC) emulsions created acoustic enhancement in vitro and in vivo upon tissue targeting. We hypothesize that polymeric nanoparticles may exhibit greater contrast enhancement than PFCs due to greater acoustic impedance mismatch, if they can be sufficiently targeted at tissue sites. Herceptin molecules (anti-HER2/neu monoclonal antibodies) were covalently linked to polylactide acid (PLA) nanoparticle surfaces using a carbodiimide technique. The surface modification was characterized by microBCA assays and FTIR measurements. Sk-Br-3, a human breast cancer cell line that overexpresses HER2/neu surface receptors, was used to test the targeting capability and specificity of the conjugated nanoparticles. MDA-MB-231 cells, which express HER2/neu receptors at a low level, were used as controls. The antibody-conjugated nanoparticles were incubated with cell suspensions for 30 min at room temperature. The cells were washed and incubated with a fluorescent secondary antibody for confocal imaging to verify the presence and specificity of nanoparticle targeting. The cells were pelleted and then scanned by a Vevo660 ultrasonic imaging system. Both Sk-Br-3 and MDA-MB-231 cells incubated with either Herceptin or Herceptin-conjugated nanoparticles were scanned. The normalized reflectivities of the cell layers were measured and compared. MicroBCA assays and FTIR results demonstrated effective covalent binding of Herceptin onto PLA nanoparticle surfaces. Cell membrane staining by conjugated nanoparticles on Sk-Br-3s but not control cells was confirmed by confocal imaging. An increase in ultrasonic reflectivity was observed in SK-Br-3 cells targeted with PLA nanoparticles. This study suggests a potential method of using targeted nanoparticles for enhanced ultrasound imaging of breast cancer.
Abstract ID: 782 Poster board space: 140
Dynamic tracer biodistribution images can be analyzed by reconstructing list-mode event data with explicit time dependence of tracer activity for arterial input, tissue uptake and washout included in the iterative reconstruction model.
A CT imager is located adjacent to the SPECT on the same gantry axis. The CT images are used for attenuation and scatter correction of the SPECT images, as well as anatomical reference.
Abstract ID: 783 Poster board space: 141
By modifying the Förster equation, we deduced an equation to predict the relative bioluminescence resonance energy transfer (BRET) efficiency of any possible luciferase/fluorophore combination. The predictive power of the equation was tested in bacteria by constructing five lux A-fluorescent protein fusion pairs: luxA-eGfp, luxA-mRfp, luxA-mCherry, luxA-tdTomato, and luxA-mPlum. The generality of this approach was further validated in mammalian cells by constructing and testing three click beetle green (CBG) luciferase/fluorophore fusion constructs: CBG-mRfp, CBG-tdTomato, and CBG-mPlum. Each of these was conjugated with five lengths of interposed Gly-Ser linkers. The equation fitted the data in both the bacterial and mammalian cells (R2 > 0.95; p< 0.05). As predicted, the best red fluorophore acceptor for CBG was tdTomato. A DEVD sequence was then added in the linker domain to quantitatively image executioner caspase activity and cell death by disruption of BRET and decrease in overall bioluminescence, respectively. Since CBG activity depends on ATP availability, the change in BRET efficiency measures the executioner caspase activity only in living and dying cells, while decreases in overall photon output reflect cell death. To test the utility of these constructs in cellulo, apoptosis was induced in HeLa cells by treatment with increasing doses, from 0.01 μM to 10 μM, of apoptosis-inducing agents, either doxorubicin or C6-ceramide, over 24 hours. At peak reduction of BRET, 12 hours of treatment with 10 μM doxorubicin, the signal to noise ratio was 43.5 and Z' = 0.87, indicating the precision of the assay. This demonstrates a novel probe for rapid real-time measurement of executioner caspase activity. Furthermore utilizing the BRET-modified equation, one can rationally choose optimal luciferase/fluorophore pairs as new luciferases and fluorophores become available.
Abstract ID: 784 Poster board space: 142
Extravasation of radiopharmaceutical and residual activity within the angiocath are typically ignored as significant are potential sources of error in quantifying SUV. Although it is rare that these residual activities are substantial enough to have a significant impact on SUV, it is important to determine which patients are affected and correct SUV for these activities. This study describes a method for screening patients for significant non-IV injected activity and for correcting injected dose for accurate SUV calculations.
Fifty patients were evaluated by measuring FDG activity residual a) in the injection syringe (via dose calibrator), b) in the angiocath (via dose calibrator and gamma probe), c) as extravasated dose in the subcutaneous tissues of the arm (via gamma probe and PET/CT scan). Measured activities were decay corrected back to time of injection and the injected dose was corrected for these errors. SUV measurements were then compared with corrected versus uncorrected injection dose. Gamma counts in the injected arm versus the contralateral arm were compared to each other and correlated with the activity measured in the injection site by PET/CT.
Although only a small number of patients were found to have significant residual activity in the injection site (≈ 2%), this activity resulted in a significant error in SUV calculation in these patients (>3%). Gamma counts measured at the injection site accurately predicted which patients had significant extravasation and required scan of the injection site. Residual activity within the syringe and angiocath was negligible.
The gamma probe is useful in screening patient injection sites for significant extravasation of radiopharmaceutical, guiding decision making for obtaining an extra bed position of the injection site in those patients requiring injected dose correction for accurate calculation of SUV. Residual syringe and angiocath activities are negligible but can also be factored in the injected dose correction equation.
Abstract ID: 785 Poster board space: 143
Renilla luciferase (RLuc) is commonly used as a reporter gene either on its own or in conjunction with firefly luciferase. Its use in bioluminescence imaging, however, has been hampered by the emission spectrum generated when it catalyzes its substrate coelenterazine (481 nm peak, 497 nm mean), as blue wavelength photons are strongly attenuated in biological tissues. To overcome this difficulty, we have explored red-shifting the emission spectrum of RLuc in order to increase its utility for small animal imaging applications.
Through the use of a homology model of RLuc and an estimate of the substrate orientation in the active pocket, residues were selected for site-directed mutagenesis. Starting from a stabilized and more active variant of RLuc (RLuc8) as the parental enzyme, a range of emission shifts were created (475-513 nm peak, 491–536 nm mean) by these single residue alterations. As these variants with active pocket mutations invariably exhibited significantly reduced enzymatic activity, further random mutagenesis and site-specific saturation mutagenesis studies were undertaken. These experiments resulted in additional RLuc variants with up to 60 nm red-shifted emission spectra (541 nm peak, 557 nm mean). More importantly, some of the identified variants showed only minimal changes in enzymatic activity. The most promising for in vivo use is a 4 mutation variant of RLuc8 that exhibits a green emission spectrum (532 nm peak, 545 nm mean) yet retains 85% of RLuc8's enzymatic activity.
It is estimated that at a depth of 5 mm of liver tissue, this 4 mutant variant of RLuc8 will result in 23-fold more light escaping the tissue than RLuc, with 70% of this improved performance arising from the green peaked emission spectrum. This is a significant improvement for imaging over previous RLuc variants, and should greatly aid the use of Renilla luciferase as an in vivo reporter gene.
Abstract ID: 786 Poster board space: 144
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 out of tissues such as tumors, contrast agents should be ‘nano-sized’ to be delivered. However, such small agents are an order of magnitude too small to produce high contrast with ultrasound used for ordinal medical purposes. We propose the use of nano-sized emulsion systems which change their phase from liquid to gas, generating microbubbles, upon physical energy such as ultrasound pulses. They would 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.
As a preliminary study for applying our nano-sized emulsion for diagnosis and drug delivery, we investigated if contrast imaging can be performed by modified diagnostic pulse sequences with a medical ultrasound scanner (Hitachi Medical Corp. EUB-8500 [modified for this study]). By using ultrasound pulses with 2–4 waves, it was found that the nano-sized emulsions selectively changed their phases from liquid to gas and produced microbubbles both in vitro and in vivo. The microbubbles formed were clearly visualized ultrasonically in gels and mice livers and tumors. Also, the echo signal intensity increased on the microbubble generation roughly proportionally to the chromatographically determined concentration of the emulsion in the tissues.
Moreover, repeated exposure of pulsed ultrasound resulted in irreversible changes of tissues in ultrasonic images, which indicate the collapse of microbubbles generated from nano-sized emulsions by the phase shift. Such collapse of bubbles would be useful for selective therapy by incorporation appropriate drugs into our emulsion systems.
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: 787 Poster board space: 145
Abstract ID: 788 Poster board space: 146
A new class of genetic-based MRI reporters for monitoring gene expression in real-time has recently been reported utilizing ferritin (FT) as a probe-less metalloprotein reporter [1,2]. FT transgene expression results in accumulation of ferric iron in situ, converting naturally present labile iron into ferrihydride nanocrystals within the FT's polymeric cores. These crystallites, known to exhibit superparamagnetic properties, induce a local increase in the water transverse relaxation rate. Thus, cellular accumulation of iron-loaded holoferritin provides contrast enhancement of reporter gene-targeted tissues in T2- or T2*-weighted images.
In an effort to develop second generation FT-based reporters exhibiting higher relaxivity, we have engineered constructs encoding single-chain ferritin chimeras (scFT). The scFT molecules are composed of both the H and L subunits of FT expressed as a single polypeptide. This design provides a fixed 1:1 ratio of H and L subunits in the assembled polymer and ensures that the essential functions of FT are present, including ferroxidase activity provided by the H subunit and Fe storage stability conferred by the L subunit [3]. The scFT molecules are efficiently expressed in cells by transduction of an adenovirus vector encoding the reporter. Importantly, scFT molecules in transduced cells are functional and scavenge significantly more Fe than cells co-expressing the H and L subunits together (i.e., H+L). Furthermore, we show that cells co-expressing scFT along with a transferrin receptor (TfR-1) transgene sequesters approximately three times the amount of iron as compared to H+L expressing cells. Inoculating these vectors into mouse brain parenchyma shows that scFT expression provides detectable MRI contrast in vivo. These results demonstrate different molecular genetic strategies for the engineering of improved FT-based MRI reporters.
Abstract ID: 789 Poster board space: 147
Lipid-coated perfluorocarbon nanodroplets are sub-micron diameter, liquid-filled droplets with proposed applications in molecularly-targeted therapeutics and ultrasound imaging. Ultrasonic molecular imaging is unique in that the optimal application of these agents depends not only on the surface chemistry, but also on the applied ultrasound field, which can increase receptor-ligand binding and membrane fusion. We demonstrate that ultrasound radiation force and acoustic streaming displace these nanodroplets in the direction of ultrasound propagation and increase the adhesion of targeted droplets in a dynamic system. A traveling ultrasound wave with a peak pressure on the order of MPa and frequency in the MHz range produces a particle translational velocity that is proportional to acoustic intensity and increases with increasing center frequency. When applied perpendicularly to a vessel, ultrasound can spatially localize targeted nanodroplets against a vessel wall, increasing retention of targeted droplets by over 100 fold in-vitro. Additionally, we demonstrate that ultrasound promotes cellular internalization of nanoparticle contents. Cell monolayers exposed to a solution of targeted nanodroplets and exposed to ultrasound show increased internalization of nanodroplet contents in the area of acoustic focus. Lastly, we show that acoustic parameters utilized to produce spatial localization and internalization of nanoparticles do not significantly affect cell viability.
Enhanced delivery of fluorescent nano droplets to PC3 cells at center of acoustic focus (dashed circle).
Abstract ID: 790 Poster board space: 148
We are evaluating TCT imaging in conjunction with NIR-dye labeled Herceptin antibody for noninvasive assessment of HER-2 expression in tumors. A NIR absorbing molecular probe has been designed by coupling the organic dye (Alexa-fluor 750 or Li-Cor 778) to the Herceptin antibody. Increasing the dye to antibody molar ratio in the conjugation reaction over the range 4-36:1 resulted in increasing actual dye to antibody ratios of approximately 2-12:1. Labeling purity after size-exclusion chromatography was greater than 98%. Labeling efficiency of Alexa-fluor and the stability of its conjugates with Herceptin were comparatively better than Li-Cor.
Breast cancer cell lines of HER-2 positive (BT-474, MCF-7 HER-2) and HER-2 negative (MDA-MB-231, MCF-7-PC-DMA) were incubated separately with probe at 37°C for 1 hr. ≈ 100 micro gm of probe was used for the labeling of 1 million of cells. After washout removal of unbound probe, fluorescence intensities were measured (both qualitative and quantitative) by flow-cytometry to determine the binding to cells. Significant cellular binding of labeled Herceptin probe was observed only in the HER-2 positive cell lines, indicating binding specificity. Labeling of Herceptin with NIR dyes allows for detection of HER-2 expressing tumor cells.
This work is supported by NIH: R01-CA108620 and R44-CA 114839.
Abstract ID: 791 Poster board space: 149
We are developing a series of assays for multiplex detection and identification of microbial and viral pathogens using molecular inversion probes (MIP). The potentially massive parallel detection capability of the pathogen MIP-assay will provide valuable information about the association of different microbial and viral pathogens in disease formation. Specific microbial and viral typing and antimicrobial drug susceptibility measurement will be possible with high sensitivity and specificity. PCR biases and non-specific amplifications from genomic DNA are avoided as amplification is performed with universal primers, pre-designed on MIPs. Microarray-based techniques, mass spectrometry, Taqman assay and Pyrosequencing technology are all applicable methods for obtaining relevant data. We have applied this model system on human papillomaviruses (HPV), known cervical cancer-causative agent. The HPV-MIP-assay has been developed for the detection and identification of clinically relevant HPVs in a single reaction mixture. We will eventually expand the assay to cover 96 HPV genotypes. A second model MIP-assay currently under investigation is the profiling of Neisseria gonorrhoeae antimicrobial resistance gene mutation and linkage studies of susceptibility to various antimicrobial agents.
Abstract ID: 792 Poster board space: 150
We have developed a new delivery vehicle which combines the echogenicity of microbubbles and fusogenicity of liposomes by attaching liposomes to the surface of microbubbles. As a model system, we used the high affinity interaction between avidin and biotin molecules to conjugate biotinylated liposomes to biotinylated microbubbles via avidin molecules. To optimize and evaluate the binding between the liposomes and microbubbles, NBD-cholesterol was used as a fluorophore in the liposome formulation. Under optimized concentrations of avidin and liposomes, we measured and calculated values as high as 105 liposomes attached to each microbubble. Interaction of this vehicle with human prostate cancer cell lines (PC-3) was further examined by incubation of liposome-attached microbubbles with a monolayer of PC-3 cells for two hours. Liposome-attached microbubbles immediately adhered to the monolayer, followed by internalization of NBD-cholesterol molecules within minutes. In a second study, we employed ultrasound radiation force to displace the liposome-attached microbubbles initially in solution to the cells. As shown in Fig. 1, the application of ultrasound radiation force successfully localized the liposome-attached microbubbles on the cells at the focal point of the beam. Further incubation of the cells resulted in internalization of the NBD-cholesterol by cells associated with microbubbles at the beam center, as shown in Fig. 2. Under similar conditions, no significant interaction was observed between microbubbles and cells without attached liposomes.
US-assisted localization of liposome-attached microbubble on PC-3 cells (5x)
After 2h incubation (20x)
The support of NIH CA 103828 is acknowledged.
Abstract ID: 793 Poster board space: 151
The herpes simplex virus type 1 thymidine kinase (HSV1-TK) is a widely used reporter for in vivo noninvasive monitoring of therapeutic gene expression, immune cell trafficking and protein-protein interactions in various living animals. HSV1-TK is a stable protein; this property allows its accumulation and easy detection in cells. However, this stability also limits its application in studies that require rapid reporter turnover. We created a destabilized HSV1-TK for use in such studies by fusing the degradation domain of mouse ornithine decaroxylase (MODC) to the C-terminal end and targeting inactivating mutations in the nuclear localization signal (NLS) of HSV1-TK. This destabilized HSV1-TK, unlike wild-type HSV1-TK, was unstable in the presence of cycloheximide and had a short half-life of protein and enzyme activity using Western blot analysis and HSV1-TK enzyme activity assay. The proteasome assay also confirmed the destabilized HSV1-TK was degraded through 26S ubiquitin/proteasome pathway. The suitability of destabilized HSV1-TK as a transcription reporter was tested by linking it to tetracycline-turn-off-expressing system. The dynamic transcriptional events mediating a series of doxycycline (Dox) inductions were monitored by destabilized HSV1-TKor by wild-type HSV1-TK. The results of dynamic gene expression studies showed that the destabilized HSV1-TK is an optimal reporter gene for monitoring short-time scale, dynamic transcriptional events mediating a series of Dox inductions, whereas the wild-type HSV1-TK is not optimal to achieve this purpose. Therefore, use of destabilized HSV1-TK as a transcription reporter allows monitoring of transcription induction more directly and its coincidence with other biochemical change easier.
Abstract ID: 794 Poster board space: 152
C.C. Ke1,
Gene expression imaging was carried out while the tumor reaching 0.5 cm in diameter. Imaging was first acquired with a gamma camera equipped with a pin hole collimator with a 4 mm aperture at 30 min after injection of 300 μCi of Tc-99m pertechnetate through the tail vein. The animal was then imaged with HRMGC, which focused on the tumor and obtained a 100-sec acquisition. Regions of interest (ROIs) were then selected on the hNIS reporter gene transfected ARO tumor and wild type ARO tumor for calculating the total and mean counts from the tumor.
Abstract ID: 795 Poster board space: 153
Nuclear medicine is used to study tissue functions in vivo by imaging tracer concentrations of radiopharmaceuticals. The finite spatial resolution of nuclear imaging, such as PET and SPECT, impacts the activity at the edge of an object blurring it out into the surrounding region, even for large objects, and blurring the surrounding activity into the edge region. One of the methods which can be employed to correct for this inherent blurring is restoration. In this work, we apply three directed restoration methods to deblur the micro-PET and micro-SPECT images. Based on the mathematical model,
Abstract ID: 796 Poster board space: 154
FDG uptake patterns in mouse tissues can be drastically altered by temperature, anesthetic choices and glucose levels. Typically in vivo animal imaging systems require heat removal via substantial cool airflow, which is not optimal for rodent physiology. Room temperature FDG uptake can result in high brown fat uptake in neck and shoulder regions, potentially masking signals in the upper torso and reducing blood flow to peripheral tumors. Anesthetized animals rapidly become hypothermic and will die without proper heating. Fasting and insulin modify endogenous glucose concentrations and can switch heart metabolism between glucose and fatty acid usage; thus heart uptake with FDG can be turned off or on based on animal treatment. Anesthesia modifies heart FDG uptake as well, with isoflurane dramatically increasing both FDG uptake and the variability in SUV measurements.
To reduce these effects, we have developed methods, equipment and protocols to make measurements simple and physiological conditions reproducible. One example is the use of conscious uptake for FDG cardiac studies. Five mice imaged with unconscious uptake under 1–2% isoflurane had an average whole heart SUV of 5.24, with a standard deviation of 63%. The same mice using conscious uptake averaged SUVs of 0.59, with SD of only 31%. This was a large significant change in SUV, and more importantly a large change in standard deviation, thus ability to see changes in SUV measurements. Additional protocol optimizations include glucose values measured using a tail sampling method that minimizes damage by a small needle stick and temperature control by heating plates and imaging chambers to eliminate brown fat uptake. The choice of injection method, animal handling, fasting state and when to administer anesthesia and which anesthetic to use should be chosen to optimize the particular imaging experimental situation, since all of these factors can alter the FDG distribution in mice.
Abstract ID: 797 Poster board space: 155
Jon Marsh, Kathryn Partlow, Michael Scott, Dana Abendschein, Michael Hughes, Gregory Lanza,
Targeted, liquid perfluorocarbon nanoparticles have been shown to be effective for acoustic contrast enhancement of abundant cellular epitopes (e.g. fibrin in thrombi) as well as in physiologic settings having lower concentrations of binding sites, e.g. when targeting angiogenic factors associated with tumor neovasculature. In this study we sought to more clearly delineate the relationship between contrast enhancement of targeted surfaces and the density of bound perfluorocarbon (PFC) nanoparticles. Two substrates were utilized for targeting. The surfaces of avidin-coated agar disks were exposed to biotinylated nanoparticles to yield a thin (submicron) layer of targeted contrast. Additionally, we used an antibody for “tissue factor” to target PFC nanoparticles to cultured smooth muscle cells expressing this protein at the cell surface, yielding somewhat thicker (several micron) layers of targeted contrast. Non-echogenic nanoparticles having oil-based cores were used to compete against PFC nanoparticles for binding, in order to define the PFC concentration dependence of contrast enhancement. An acoustic microscope was used to characterize backscatter as a function of bound PFC (as determined with gas chromatography). Agar surfaces and cells targeted with either PFC nanoparticles alone or a mixture of PFC- and oil-based nanoparticles exhibited substantially increased reflectivity relative to untreated samples or those exposed only to targeted oil-based nanoparticles. The relationship between reflectivity enhancement and bound PFC content for targeted cells was curvilinear, and exhibited an apparent asymptote at the maximum concentration. This behavior compared well with predictions of reflectivity from a “transmission-line” model. These data indicate that the utility of targeted PFC nanoparticles may extend beyond the detection of abundant molecular epitopes to pathology with more sparse concentrations of molecules such as constitutive or inducible cellular tissue factor.
Abstract ID: 798 Poster board space: 156
Melvin Sims, John Viator, Raina Cepel,
The plasmon resonance of metallic nanoparticles generally falls within the optical wavelengths of light. This has prompted work towards development of metallic nanoparticles as photoacoustic contrast agents. Theoretical models exist for the determination of absorption of light by metallic nanoparticles; however, to the best of our knowledge, careful measurements of the photoacoustic signal under well-controlled conditions have not yet been published. Absorption efficiency (see Figure 1) and nanoparticle density are two factors which will play a role in the suitability of these particles as photoacoustic agents. In addition, the response of the acoustic transducer will also be an important factor.
This work presented here uses a tunable laser and a controlled photoacoustic measurement apparatus. Measurements are made to verify that the relative strength of the photoacoustic signal is related to the absorption curve for the size and composition of the nanoparticle.
A summary of the study will be presented showing acoustic signal versus laser wavelength, corrected for energy density and nanoparticle concentration. This summary will include several sizes, shapes, and compositions of nanoparticles. In addition, the acoustic signal amplitude will be related to the absorption curve using the acoustic wave propagation concepts. This enables the development of a relationship between the laser power and wavelength, nanoparticle size, composition and concentration, and the resultant photoacoustic signal.
The result of this work will be useful in understanding the potential for generating sufficient photoacoustic signal for imaging applications. This will be expressed in terms of uptake and binding limitations for targeted nanoparticles.
Figure 1 is a plot of the absorption efficiency versus wavelength for gold nanoparticles of three sizes.
Abstract ID: 799 Poster board space: 157
Gallium-68 is a 68-minute half-life positron emitter produced from 271-day germanium-68 that is suitable for synthesizing cyclotron-independent PET radiopharmaceuticals using a suitable generator system. Unfortunately, previous generator systems based on solvent extraction or alumina, tin dioxide, or titanium dioxide supports have either been highly inconvenient or produced Ga-68 contaminated with trace metals which limited its suitability for labeling receptor agent bioconjugates.
A Ge-68/Ga-68 generator based on TiO2 is available from commercial sources. A comparison has been made at the multi-mCi level of Ga-68 elution yields, Ge-68 breakthrough, trace metal contamination, and DOTA labeling yields from this generator and one based on selective Fajan's adsorption of Ga-68 on various hydroxyl containing supports such as glass microspheres, followed by desorption with HCl. The latter generator holds the possibility of Ga-68 eluants containing substantially fewer DOTA-binding extraneous metals as well as a much lower eluant volume.
Both generators exhibit high Ga-68 yields of 80–90% and low Ge-68 breakthrough. Trace metals in the eluant have been analyzed using inductively coupled plasma-optical emission spectroscopy (ICP-OES) and ICP-mass spectroscopy. The TiO2 generator exhibits significant contamination with Ti, Fe, Zn, and other metals that increases with successive elutions. Various methods of rapidly purifying Ga-68 from interfering metals in the eluants of both generators have been investigated.
Labeling yields of diminishing DOTA levels with Ga-68 from both generators have been measured. These data provide a useful means of assessing the utility of each generator system in providing Ga-68-labeled receptor agent bioconjugates.
Abstract ID: 800 Poster board space: 158
We are developing a small animal SPECT system to obtain images of the biodistribution of radiopharmaceuticals in unrestrained, unanesthetized mice. This system uses an infrared based animal position tracking system while acquiring SPECT projections. The awake mouse is confined within an infrared-transparent burrow at the center of rotation of the gantry. The mouse will have three infrared reflecting spheres attached to its head that are imaged by a pair of CMOS based infrared cameras. The tracking apparatus will measure the spatial position or pose of the mouse's head with 6 degrees of freedom at a rate of 10–15 frames per second with sub-millimeter accuracy. A prototype system with two gamma cameras with active areas of 10 cm × 20 cm and a 2.25 mm intrinsic resolution has been constructed and used for phantom tests at Oak Ridge National Laboratory. SPECT reconstruction algorithms have been developed and tested on moving phantoms demonstrating our ability to remove the effects of movement. A second imaging system using a SPECT gantry with servos instead of stepper motors and with X-ray CT capabilities has been installed at Johns Hopkins University. Two high resolution gamma cameras 10 cm × 20 cm in size with a 1.2 mm intrinsic resolution are under construction for installation on this SPECT gantry in preparation for animal studies. The gamma cameras are based on pixellated NaI(Tl) crystal scintillator arrays and arrays of compact flat panel position-sensitive photomultiplier tubes to enable high spatial resolution in a compact low, profile device. Our system will be operational in the near future to image pharmocokinetics in the brains of live mice without the confounding effects of anesthesia or physical restraint.
Abstract ID: 801 Poster board space: 159
In vivo imaging of a specific mRNA target has remained a challenge due to the small copy number of mRNA targets per cell. We present here a new approach to imaging mRNA in vivo based on group I introns Tetrahymena thermophila ribozyme. Tetrahymena ribozymes constitute a class of catalytic RNA molecules, capable of catalyzing cis- and trans-splicing reactions. We designed plasmid constructs containing the trans-splicing ribozyme, an engineered reporter mRNA, and an antisense sequence against an mRNA target. These constructs were able to bind the mRNA target and produce a fusion mRNA that could be translated into functional reporter enzymes. Two different reporters have been exploited to image the mRNA of a dominantly negative p53 (p53DN) gene: beta-lactamase for single living cell imaging and firefly luciferase for whole living mice. Analyses of the splicing reaction products confirmed that these constructs were able to target and image mRNA targets in vivo. For a transiently transfected construct, spliced products could be seen with the RT-PCR assay as early as 24 hrs after transfection, and splicing-dependent luciferase activity began to increase at 48 hrs, maximizing around 72 hrs. A 9.5±1.9 fold difference was observed between tumors with and without the p53DN mRNA 24hrs following the tumor implant (72 hrs after transfection). In summary, we have shown here the first example of imaging ribozymemediated trans-splicing activity in living animals. The design is highly generalizable for different mRNA targets and genetically encoded. This splicing-dependent reporter assay offers high contrast and may be applied to directly image endogenous mRNAs, especially over-expressed tumor-specific mRNAs in living subjects.
Abstract ID: 802 Poster board space: 160
Contrast-enhanced ultrasound has shown utility in the field of molecular imaging. This technique requires the adhesion of ultrasound contrast agents (UCA) to molecular markers of disease. This is typically accomplished by coating the surface of the contrast agent with a ligand that binds the intended molecular marker. A micron-scale UCA, such as a microbubble, bound to a molecular target on the vascular endothelium is subject to the dislodging force from incident blood flow, which counteracts retention. Additionally, contrast agent adhesion to a target molecule requires rapid binding kinetics, especially in vigorous blood flow. The ability of a ligand:target bond complex to mediate rapid adhesion and withstand dislodging forces is critical for efficient ultrasound-based molecular imaging. In the current study, we describe a flow-based adhesion assay which, combined with a novel automated tracking algorithm, enables rapid determination of the ability of a targeting ligand to mediate effective contrast agent adhesion. This system was used to explore the adhesion of UCA targeted to the pro-inflammatory endothelial protein P-selectin using four targeting ligands, which revealed several interesting adhesive behaviors. Contrast agents targeted with glycoconjugate ligands modeled on P-selectin glycoprotein ligand-1 exhibited primarily unstable or transient adhesion, while UCA targeted with an anti-P-selectin monoclonal antibody exhibited primarily firm adhesion, although the efficiency with which these agents were recruited to the target surface was relatively low. Dual targeting with two ligands, one of which enabled rapid, although unstable, adhesion, and one which enabled stable, although low-efficiency, adhesion, exhibited optimal retention under most conditions. This points to the necessity of screening potential ligands under various flow conditions, and suggests a role for synergistic dual-ligand UCA retention to molecular targets.
Abstract ID: 803 Poster board space: 161
Abstract ID: 805 Poster board space: 163
Accurate intraoperative determination of negative tumor margins is a major pitfall in surgical oncology. Frozen section is currently the standard for evaluating tumor margins intraoperatively but is subject to sampling error. Better mechanisms for detecting residual tumor are clearly needed and could have a significant impact on patient outcome. We present here a method of improving evaluation of tumor margins, intraoperatively, utilizing PET/CT.
Five patients with known colorectal cancer received a diagnostic PET/CT scan for staging of disease. Patients were injected with ≈ 12 mCi 18F-FDG just prior to their surgical procedure. Intraoperatively, tumors were localized, resected, marked with sutures and glass beads for tumor and normal tissue and scanned PET/CT. Regions of interest were drawn around metabolically active foci and normal surrounding tissue. Specimens were then sent to pathology for processing. Representative samples of the specimen were measured in a well counter and compared to pathologic findings and measured activity from PET.
Tumor counts were significantly higher than background counts for all specimens (p< 0.05) in vivo. Activity (Bq/ml) within tumor tissue was significantly greater than the activity in surrounding tissues (p< 0.05) with PET quantification and well counter measurements. Pathology findings (benign versus malignant) correlated with measured activity of tissue in the well counter as well as ROI activities from the specimen PET/CT scan. Tumor to background ratios accurately predicted negative tumor margins when compared to pathologic findings.
Pathologic findings correlate well with measured activity from intraoperative PET/CT specimen scanning as well as specimen activity measured in a well counter. This method may prove useful in intraoperatively evaluating tumor margins in the entire specimen and has the potential for guiding pathology sampling and improving the accuracy of pathology findings. The combined benefit of this technique could significantly improve patient outcome by overcoming the pitfalls of frozen section determination of tumor margins.
Abstract ID: 806 Poster board space: 164
Abstract ID: 807 Poster board space: 165
Abstract ID: 808 Poster board space: 166
Funded by NCI contract N01-CO-12400.
Abstract ID: 809 Poster board space: 167
Single-photon nuclear imaging of the whole body and the myocardium is now exclusively by dual-headed systems. The only rationale for the use of two heads is a doubling of counts acquired or a halving of imaging time. Preliminary clinical results with a dual-head breast scanner demonstrate that several advantages far beyond a simple doubling of counts can be exploited. Most importantly, the fall-off of spatial resolution with distance allows low-contrast lesions located opposite a single detector to be lost in noise, whereas a second detector will find this lost lesion due to its proximity. Several cases of increased detection sensitivity are presented to accentuate this elegant finding. Secondly, two opposing views offer two samplings of spatial resolution and thereby a method to localize the lesion in depth as well as to estimate the blur-free lesion size (partial volume correction) through resolution recovery. Knowledge of depth location and lesion size allows attenuation correction and quantitation of tracer uptake to be computed. Thus, a simple opposing view hardware design can provide quantitative evaluation of hot lesions in the breast for any of the variety of new radiopharmaceuticals that are under development for molecular imaging of breast cancer. This “Molecular Breast Imaging” tool's performance is further enhanced through the use of semiconductor CZT, which has superior energy resolution performance resulting in scatter rejection and uniform-appearing images from chest wall to nipple. Clinical examples of CZT's superior scatter rejection are shown. Finally, the data acquisition method is upright, mammography-type to provide patient familiarity and comfort since the compression is mild compared with that of x-ray. Examples are shown of cranio-caudal and medio-lateral views that are rendered in a format that is easily compared with their corresponding digital mammography images.
Abstract ID: 810 Poster board space: 168
Ali Bahadur1, Mark Hansen2, Mary Blandford2, Mario Capecchi2,
Bioluminescent reporters are an increasingly utilized, sensitive but semi-quantitative tool for following disease progression in preclinical models. Like many optical reporters, signal from luminescent proteins such as luciferase is confounded by scatter of and absorption of light in tissue, which is a function of depth. Using gene targeting, we have generated a bicistronic Cre/LoxP reporter mouse line that pairs semi-quantitative expression of firefly luciferase with quantitative expression of a human placental alkaline phosphatase that is secreted into the serum (SeAP). The two reporter genes were targeted to the Rosa26 locus, and driven by cis-elements consisting of the native Rosa26 promoter, the CMV immediate-early promoter/enhancer, and the SV40 late viral protein gene 16S/19S splice donor and acceptor signal sites. Injecting a single 150 mg/kg dose of luciferin into a dual reporter mouse with ubiquitous Cre expression, luminescent signal could be observed for greater than 20 hours. To determine the range of sensitivity for the secreted alkaline phosphatase, the dual reporters were activated focally by injection of an adeno-associated virus encoding Cre into thigh muscle, or across all skeletal muscle by breeding to a herculin-Cre mouse line. SeAP activity measured by a microtiter plate assay using 12.5 μl of serum demonstrated detectable signal above background for the focal reporter activation, and an almost 2-log increase in signal for ubiquitous skeletal muscle activation. SeAP quantitation was comparable in fold change to luciferase signal, matching or exceeding reproducibility. In summary, this dual-modality reporter allele can be bred into any Cre/LoxP conditional mouse model of disease, thereby permitting dual modality biomarker measurements of disease progression.
Abstract ID: 811 Poster board space: 169
Abstract ID: 812 Poster board space: 170
Tl-201 has high first-pass extraction and is ideal for absolute quantitation of myocardial perfusion (MBF) and coronary flow reserve (CFR) in vivo. This study was intended to investigate feasibility of quantitation of MBF and CFR in small animals in a conscious condition.
Arterial plasma input function (IF) was determined from 8 rats (under anesthesia) by frequently sampling, and the accuracy of using a standardized IF was tested. Catheters were placed into the femoral artery (for monitoring the hemodynamics) and the femoral vein (for Tl-201 injection) of Wister rats one day before SPECT experiment. Rats were softly fixed at a conscious condition during SPECT in a specially designed acryl container. After determining the scanning position, sequential images were obtained at every 30 sec following i.v. Tl-201 using a dedicated pinhole SPECT system that consists of 2 pixelized NaI detectors combined with position-sensitive PMT. Images were reconstructed according to a previously validated technique (Zeniya 2004), and data were fit to a two compartment model (2CM) as proposed previously[1].
IF's obtained were reproducible among 8 rats, and the area under the curve varied only ±7.8% and ±9.7%, respectively, suggesting feasibility of using the standardized IF. Simulation study also showed errors in MBF < 10%. The heart rate and the systolic blood pressure varied only from 396 to 441 bpm and from 138 to 125 mmHg, respectively, causing variation of RPP minimal (< 2%). Sequential myocardial images obtained were clear, to which 2CM can be applied, yielding MBF of 2.99±0.23 mL/min/g at rest and the distribution volume (DV) was 107±12.4 mL/mL, after correcting for the partial effect.
Quantitation of MBF in small animals at a conscious condition was feasible using our pinhole SPECT system.
Abstract ID: 813 Poster board space: 171
Cadmium zinc telluride, or CZT, is a room-temperature semiconductor radiation detector that offers superior energy resolution for nuclear medicine imaging. The energy resolution advantage can be used to reject scattered events. This is important for dual-isotope applications in which the photons are relatively close in energy. The present work tested CZT in dual-isotope imaging of nude mice carrying a neuroblastoma xenograph. Two combinations of radiopharmaceuticals were used: 1) 99mTc-labeled MDP (bone agent) and 111In-DTPA (tumor agent); and 2) 99mTc-labeled MDP and 123I-MIBG. In the first case injections of MDP were 2 hours (bone uptake and blood pool clearance) and DTPA 20 minutes (seepage into extra-cellular space) prior to the start of image acquisition. In the second case, the MIBG was injected the evening prior (blood pool clearance) and MDP 2 hours prior to imaging. The CZT detector has FOV of 12.7 × 12.7 cm2 with 1.59 mm pixels forming a matrix of 80times80 imaging elements. The energy resolution averages less than 5% at 140 keV. A single pinhole of 1.0 mm diameter was used with a 360-degree orbit and 120 second projections were acquired at 64 angular positions. Data were corrected for pinhole geometry and misalignments, a technique that has been shown to give 0.7 mm FWHM resolution for the 1.0 mm pinhole. The radius-of-rotation was 2.28 cm with a 3.24 cm field-of-view. In vivo spectra of the two combinations show the peak separation that CZT can achieve. Tomography of the upper thorax and neck found that the tumor was visible with DTPA but not MIBG; however 123I accumulation in the thyroid was clearly seen in the context of the bone scan. CZT has the energy resolution needed for dual-isotope SPECT. We gratefully acknowledge Drs. Moats and Karapetyan of Children's Hospital of Los Angeles for their assistance in these studies.
Abstract ID: 814 Poster board space: 172
A
Magnetosomes are intracellular magnetic structures synthesized by naturally occurring magnetotactic bacteria, a morphologically diverse set of gram-negative bacteria which produce magnetosomes and exhibit motility thought to be directed by the earth's magnetic field. These structures are usually comprised of a magnetic iron oxide crystal core, surrounded by a phospholipid membrane. Although many of the genes, that were previously shown to be regulated by iron, are likely involved in normal magnetosome production within the bacterium, the specific gene(s) sufficient for magnetosome formation remains unknown. Here we show that a single gene in naturally occurring magnetotatic bacteria is sufficient to produce magnetosomes in a mammalian cell line. Because the magnetosomes are formed with a single gene and because all iron oxide nanoparticles alter the magnetic resonance imaging (MRI) signal in a similar manner, the possibility of utilizing this gene as a marker for gene expression with MRI is raised. Our results show that magnetosome producing cells have a transverse relaxivity (R2) up to four times that of normal cells (see Figure), making them clearly visible by MRI. Cells expressing this gene can also generate contrast sufficient to be seen in vivo by MRI as can be seen in the figure below of a mouse brain. From a microbiology perspective, these results suggest that the vast majority of magnetosome related genes support functions aside from actual crystallization of iron-oxide. For biotechnology applications, these results are a significant step forward for the use of MRI in molecular imaging.
Abstract ID: 815 Poster board space: 173
Joonyoung Kim,
In the quantification of oncology studies in mice, it is important to know how the recovery coefficient and spillover vary with tumor volume.
The purpose of this study is to measure the partial volume and spillover effects as a function of object volume and reconstruction algorithm. A 4-layer phantom was made in-house to measure the partial volume and spillover effects. Capsules (cylinders with hemispherical end-caps) were milled out with six different diameter holes (5, 4, 3, 2, 1.5, and 1 mm). This phantom was assembled with 2, 3, or 4 layers to create 18 capsule volumes from 2 to 389 μL. The phantom was imaged with a calibrated Siemens microPET™ F120 with a known activity concentration of 18F filled in each volume. Images were reconstructed using either filtered backprojection (FBP) or maximum a posteriori (MAP, β=0.01). Recovery coefficients were calculated from selected profiles in each capsule by two methods: (1) ratio of peak measured and true activity concentrations RCpeak and (2) ratio of area under the measured and true distributions RCarea. The RCs were fitted with bi-exponential functions of object volume and extrapolated to 4000 μL. The spillover was calculated as the fraction of measured distribution outside the capsule. RCpeak was 0.858 in FBP and 1.252 in MAP at 389 μL. RCarea was 0.506 in FBP and 0.704 in MAP at 389 μL. According to extrapolation, RCpeak and RCarea reached unity for volumes over 1150 μL and 2250 μL respectively in FBP images. In MAP images, the object volumes needed to be over 550 μL for RCpeak, and 1400 μL for RCarea. In addition, the spillover fractions were 0.387 in FBP and 0.315 in MAP at 2 μL. This method can be used to correct for partial volume and spillover effects in preclinical tumor studies.
Abstract ID: 816 Poster board space: 174
Marites Melancon, Wei Wang, Xiaojun Ji, Chusilp Charnsangavej, Juri Gelovani,
Poly(
Abstract ID: 817 Poster board space: 175
Currently, the main limitation of 19F MR imaging is the paucity of available fluorine-containing compounds that can be administered in sufficient quantities for in vivo imaging while remaining non-toxic. To fill this void, we have designed and synthesized silica nanoparticles containing an abundance of 19F nuclei to be utilized as contrast agents for MR acquisitions. 19F MR spectra of silica based nanoparticles with mean diameters of 10–20 nm yielded ≈ 30 fold increase in signal-to-noise compared to 1M NaF. Measured MR relaxation times of neat solutions at 4.7T were T1 = 482.9 ms and T2 = 14.7ms. Additionally, HPPH (2-[1-hexyloxyethyl]-2-devinylpyropheophorbide-a), a sensitizer used in photodynamic therapy (PDT), was encapsulated within these nanoparticles and characterized for its potential use as a bi-functional agent for diagnostic imaging and therapeutic assessment. Furthermore, these nanoparticles could also serve as single probes for multi-modality acquisitions using fluorescence and MR imaging. Preliminary experiments have demonstrated the validity of this approach in preclinical model systems. In vitro fluorescence studies revealed tumor uptake of HPPH doped silica nanoparticles in quantities sufficient for visualization. In vivo studies have demonstrated that high resolution structural (1H MR) and functional (19F) information (targeted delivery with no background signal) could be obtained using these methods. Finally, suspensions of 19F nanoparticles were concentrated to form semisolid crystalline media visible by MR for potential application as coatings for medical devices. In conclusion, these results show that high resolution anatomical and physiologic information could be obtained with a single MR acquisition. Nanoparticles incorporating 19F nuclei could facilitate the development of new multi-modality based imaging approaches applicable to cancer research.
Abstract ID: 820 Poster board space: 97
Abstract ID: 821 Poster board space: 98
Merging of tumor targeting and molecular-genetic imaging into an integrated platform has been thus far limited by lack of an effective strategy to enable systemic yet ligand-directed delivery and image of specific transgenes. Many eukaryotic viruses have potential for targeted delivery of transgenes but require elimination of native tropism for mammalian cells; in contrast, prokaryotic viruses can be adapted to bind to mammalian receptors but are otherwise poor vehicles. Here we introduce a genetic targeting and imaging system containing cis-elements from adeno-associated virus (AAV) and single-stranded bacteriophage. Our AAV/phage (AAVP) chimera is designed to target an integrin receptor within tumors. First, we show that targeted AAVP provides superior tumor transduction over standard phage constructs and that incorporation of AAV inverted terminal repeats is functionally associated with an improved fate of the delivered transgene. Moreover, we show that the temporal dynamics and spatial heterogeneity of gene expression mediated by targeted AAVP can be monitored using clinically applicable positron emission tomography with [18F]-FEAU in preclinical settings. Finally, a single systemic dose of a prototype delivered a suicide/reporter, enabled transgene imaging, caused marked apoptosis and tumor growth suppression. Given that this new class of hybrid particles home to tumors, mediate systemic delivery, and reporter imaging based on the selective expression of a specific receptor, it might become useful for targeting and molecular imaging applications in human disease.
Abstract ID: 823 Poster board space: 99
The tumor growth is dependent on angiogenesis, which has led to the development of new approaches to treatment directed at tumor vasculature. Xenografts bearing human tumors are predominantly used to characterize human tumor vasculature for the purpose of generating preclinical data for clinical trials. However, even though tumor cell lines may be the same, the tumor vasculatures with xenografts show different characteristics caused by mouse-host human tumor cell interactions as opposed to human-host human tumor cell interactions. Therefore, a more clinically relevant model of tumor growth and angiogenesis is necessary in order to mimic cell-stroma interactions.
In this study, human neonatal foreskin was transplanted into mouse skin to provide a human stroma background for human cell inoculation. When the grafts were completely healed, human Kaposi's sarcoma (KS SLK) cells were inoculated into the human skin. Once the tumor was palpable, dynamic fluorescence imaging was conducted from human skin grafted mice (chimeric mice) and xenografts bearing a subcutaneous KS SLK tumor immediately following the i.v injection of an integrin αvβ3 targeting 111-In-DTPA-IRDye800-c(KRGDf). Imaging was conducted for a period of 40 min and every 24 hrs thereafter for up to 48 hrs. In addition, scintigrams were acquired. The results showed that the KS SLK in chimeric mice exhibited early and rapid uptake of 111-In-DTPA-IRDye800-c(KRGDf). The fluorescence signal in human skin region in chimeric mice was not uniform. In addition, the whole body imaging showed little fluorescence around wound region indicating signal was due primarily to tumor angiogenesis.
In summary, our preliminary results demonstrated that imaging of angiogenesis with vascular targeting is dependent upon cell-stromal signaling. Investigators employing preclinical imaging of vascular therapeutic and diagnostic agents should consider the extent to which cell-stromal interactions impact results and their translation to human clinical trial.
Abstract ID: 824 Poster board space: 100
A. Volgin, Y. Ying, Z. Peng, David Schellingerhout, W. Bornmann, G. Powis,
The goal of this investigation was to assess which of the novel drugs with different mechanisms of action may inhibit or potentiate the inhibition of HIF-1α expression and activity in tumor cell spheroids under hypoxia.
Abstract ID: 825 Poster board space: 101
RNA interference is a powerful tool to inhibit gene expression and its efficiency is well documented in cultured cells. In vivo, this approach is limited by the bioavailability of RNA. We prepared siRNA in native RNA or modified 2′-ribose chemistry and (i) evaluated their RNAi efficiency, (ii) labeled them with fluorine-18 and imaged their pharmacokinetics and in vivo biodistribution and (iii) analyzed their plasmatic metabolism.
2′-Fluoro and 2′-O-methyl chemical modifications were incorporated into siRNA directed against the luciferase gene transcript, and their ability to inhibit gene in cells stably expressing luciferase was measured. The antisense strand was labeled with [18F]-fluoropyridine-bromoacetamide, the two strands of siRNA were hybridized and the maintenance of RNAi effect was measured at the protein- and mRNA-level. Biodistribution of radiolabeled siRNA was followed in mice and rats by PET imaging. Plasmatic metabolism was studied by RP-HPLC.
2′-Fluoro siRNA showed the same RNAi efficiency than unmodified siRNA. The modification by 2 ′-O-methyl nucleotides of both strands, but not of the sense strand only, hampered RNAi activity. All the radiolabeled siRNAs were obtained in more than 92.5% purity and showed the same interference efficiency as non-conjugated oligonucleotides. The main route of siRNA elimination was the renal system followed by the hepato-enteric route. The kinetics of radioactivity was similar for non-modified siRNA and 2′-O-methyl-modified siRNA. For 2'fluoro-siRNA, the radioactivity peak was reached later and eliminated more slowly from the liver, heart and muscles. The initial peak in muscles was also higher than for the two other siRNAs. The biodistribution of 2 ′fluoro-modified single-strand oligonucleotide showed weaker organ uptake than that of the double-stranded oligonucleotide.
In itself, the double-stranded nature of siRNA improves their bioavailability. 2'Fluoro-modified siRNA on the two strands has a half-life in plasma increased by three compared with non modified siRNA, shows better uptake in all organs and slower elimination.
Abstract ID: 826 Poster board space: 102
This study was designed to monitor a novel selective cancer treatment in vivo by analysing the binding characteristics of tumor selective antibodies with near-infrared imaging (NIR) technique in combination with flat-panel based volume computed tomography (fpVCT). This new therapeutic approach is the antibody directed enzyme prodrug therapy (ADEPT) in which conjugates of enzymes and tumor selective antibodies are applied for a transformation of non-toxic prodrugs into highly potent anticancer drugs at the tumor site. For this purpose human breast cancer MDA-MB-231 cells were orthotopically implanted into female SCID mice or syngeneic A20 lymphoma cells were injected intravenously into BALB/c mice. Prior to scanning with the eXplore Optix system Cy5.5 labeled monoclonal antibodies were injected into the tail veins of mice. These mice were also imaged in combination with contrast media by fpVCT (both images devices from General Electrics) with tube settings at 80 kVp and 100 mA resulting in an isotropic high-resolution volume data set. In these tumor bearing mice binding of tumor selective antibodies to breast tumors or lymphoma cells within the liver could be clearly determined over time with the eXplore Optix system by measurement of fluorophore concentration, fluorescence lifetime and location of fluorescent material within the mice. Tumor growth rates by measuring tumor volumes as well as morphological alterations of tumors could be clearly depicted by fpVCT imaging. Fusion of data sets generated by both imaging techniques in the same mice enables us to correlate antibody binding with tumor sites and morphology. Monitoring in vivo the binding efficacy, the binding kinetics as well as the binding sites of tumor selective antibodies that will be used in the ADEPT concept allows a targeted and effective evaluation and management of this novel cancer therapy during the course of the disease.
Abstract ID: 827 Poster board space: 103
Abstract ID: 828 Poster board space: 104
Richard Halberg, Ben Durkee, Raj Ambay,
Abstract ID: 829 Poster board space: 105
Single photon-emission tomography (SPECT) is a molecular imaging technique that offers opportunity to visualize directly and non-invasively the pharmacokinetics and tissue disposition of new drug in human or experimental animals. In this field detailed analysis of uptake of regions of interest (ROIs) from images is important; thus quantification is needed. The aim of this study is to achieve semi-quantification of both the absolute activity concentrations and the relative values of 131I labeled vaccine from the reconstructed micro-SPECT/CT images.
The 131I labeled peptide vaccine (p607E) was used to monitor the retention of the vaccine in rats. Male SD rats were intramuscularly injected with 131I-p607E. Micro-SPECT/CT images were acquired every seven days until 28 days after injection. Medium-energy parallel hole collimators were used for scans. In-house software was applied for drawing the ROIs, and the activity concentration analysis. Both absolute and relative quantification were analyzed. The absolute quantification “mean value in region of interest” results at the day 0, 7, 14, 21, and 28 were 100%, 128±0.11%, 137±0.50%, 127%, 123%, respectively. And the relative quantification “mean value in region of interest” results at the day 0, 7, 14, 21, and 28 day were 100%, 86±0.22%, 68±0.04%, 63%, 58%, respectively. The longitudinal micro-SPECT/CT imaging reveals that the most 131I-p607E retained in injected muscle site until 28 days post injection, and it met with the data from bio-distribution. The quantification data assessment showed the relative data are credible.
Quantitative imaging is difficult to achieve because of several effects that can lead to errors in activity estimates, 131I as a source is more apparent. Nevertheless, this work shows the credibility of quantitative 131I labeled agent of micro-SPECT.
Abstract ID: 830 Poster board space: 106
Abstract ID: 831 Poster board space: 107
The development of imaging/therapeutic agents against neuronal targets is hampered by the limited access of probes to the CNS across the blood brain barrier (BBB). To discover new transport systems that can be exploited for noninvasive brain delivery of therapeutics and imaging probes, we isolated a single-domain antibody (FC5) which selectively recognized human cerebral endothelial cells and transmigrated across the BBB. A free cysteine was engineered in FC5 to enable its conjugation with PEG and biologics. To increase the antibody avidity, FC5 pentamerization on VT1B subunit was performed.
To assess FC5 penetration across the BBB, Cy5.5-labeled FC5 was injected in mice and its pharmacokinetics and biodistribution were followed using time-domain optical imaging. The fluorescence signal intensity analysis demonstrated higher accumulation of FC5 in the brain compared to negative control antibody, NC11. Site-specific PEGylation of FC5 resulted in increased plasma half life, reduced uptake by the hepatic reticulo-endothelial system and higher accumulation in the brain. Pentameric FC5 also exhibited improved half-life and superior brain accumulation. To confirm the imaging data and to determine the fate of FC5 in the brain, immunohistochemistry on brain sections from animals undergone optical imaging evaluation was performed using selective markers of brain vessels, astroglia and neurons. Cy5.5 signal of Cy5.5-FC5 co-localized with brain vessels and neurons, and not with astrocytes, suggesting that, after crossing the BBB, FC5 predominantly targets neurons. To unambiguously demonstrate the presence of FC5 protein in the brain, laser-capture microdissected brain vessels and vessel-free brain parenchyma were analyzed by LC-MS. Unique peptide signatures of FC5/P5, but not control antibody, were detected in brain parenchyma and confirmed by LC-MS/MS sequencing. These alternative detection methods have confirmed the accuracy of results obtained by optical imaging. We conclude that time-domain optical molecular imaging is potentially suited for discovery and development of drug candidates targeting neurological diseases.
Abstract ID: 832 Poster board space: 108
We are interested in monitoring osteoarthritis (OA) progression and its response to therapeutic treatment in pre-clinical animal models. With this aim, we have implemented T1p-weighted imaging on a Bruker Biospec 9.4T/30cm system. In preparation for in vivo experiments, we have performed ex vivo experiments on excised bovine patella cartilage plugs to explore the relationship between T1ρ and proteoglycan (PG) content of the tissue at high field. A 90°x - SLy −90° -x spin-lock pulse cluster was prepended to a standard fast spin-echo image acquisition scheme and T1ρ parameter maps of the cartilage plugs were generated from an image series acquired with five spin-lock durations between 5 and 50 msec [1]. This acquisition regimen was performed with various spin-lock pulse amplitudes between 500 and 3000 Hz to characterize T1ρ dispersion in the tissue. T1ρ parameter maps were calculated from the reconstructed data sets on a pixel-by-pixel basis for each spin-lock amplitude, an example of which is shown in Fig. 1A for four individual cartilage plugs acquired with a 3000 Hz spin-lock pulse. The points in the T1ρ dispersion curves of Fig. 1B represent average T1ρ values from ROIs that represent the superficial and transitional layers of the articular cartilage. We note that the linear dispersion model of Chaumette et al [2] fits these data better than a previously described bi-Lorentzian model [3]. The distinction between slopes of the dispersion curves from the two cartilage regions suggests a novel means of generating contrast in models of cartilage degradation.
Abstract ID: 833 Poster board space: 109
Multivalent ligands can provide high binding avidity and unique specificity for early detection and treatment of cancer. We are developing multivalent ligands that target
Abstract ID: 834 Poster board space: 110
We have developed a model of carcinogenesis that describes the impact of tumor cell growth on the physiological microenvironment and the effect of this altered environment on cancer progression. Much of the insight and testing of this model has used in vivo and in vitro molecular imaging. In this model, hyperplasia in early lesions leads to hypoxia due to a physical separation of the proliferating cells from the blood supply by the basement membrane. Intermittent hypoxia selects for a glycolytic phenotype through transcription factors such as HIF-1α. A glycolytic phenotype is common to metastatic cancers, as observed by FDG-PET imaging. Increased acid production by glycolysis lowers tumor pH and selects for cell invasion. Low pH is commonly observed in experimental tumors with 31P-MRS, 1H-MRS or 4AmP-MRI.
To test these predictions, normal, tumorigenic, and metastatic cell lines have been acutely, chronically, or cyclically treated with low pH and hypoxia. Surviving, resistant cells were selected and observed for phenotypic changes. A consistent finding of these in vitro evolution experiments was an increased acid production and glycolysis by hypoxia-selected clones (Table 1). Cell lines selected by exposure to low pH, showed elevated in vitro transwell invasion, altered motility, and altered gene expression patterns (Table 2). The effect of pH on formation of metastases in vivo is being investigated with luciferase, GFP and RFP transfected breast cancer cell lines. These data are consistent with a model that predicts that low pH and hypoxia of early tumors selects for a more aggressive phenotype.
Effect of periodic hypoxia (4 months @ 1 cycle per week) on the glycolytic phenotype of MCF-10T cells
Invasion (relative to parental cells) time at low pH
Abstract ID: 835 Poster board space: 111
Molecular recognitions between RGD peptides and integrin receptors have been an important basis for in vivo tumor imaging and targeted drug delivery. Desferrioxamine (DFO), originally derived from Streptomyces pilosus, has been used in chelation therapy to treat iron-overload diseases. Its potentials in oncology including antiproliferative activity, radiosensitization, and molecular imaging have also been reported. Over the past years, we have explored novel, diverse DFO-bearing RGD compounds for targeting tumor via optical imaging and/or nuclear imaging. In this presentation, we will report some DFO-bearing compounds based on the tripeptide RGD unit and its cyclic pentapeptide analog, i.e. cyclo(RGDfD). Generally, DFO can be introduced into the molecular system by conjugation with a carboxylic acid group in the presence of PyBOP/HOBT/DIEA in solution or on a solid support. The synthetic strategies we have developed permit us to decorate the system with the multiple RGD peptide units for improving molecular recognitions and label such compounds with a NIR fluorescent probe for optical imaging as well. The related metal binding, cytotoxicity, αvβ3 integrin receptor binding, cellular internalization, apoptosis, caspase-3 activity, radiolabeling, and in vivo biodistribution in tumor-bearing nude mice will also be presented. The results should provide some insights into the potentials of these DFO-bearing RGD compounds in oncology.
Abstract ID: 836 Poster board space: 112
Robyn Hickerson1, Qian Wang2, Robert Reeves2, Christopher Contag2, Frances Smith3, Irwin McLean3, Devin Leake4,
RNAi offers a novel approach for treating genetic disorders including the rare monogenic skin disorder pachyonychia congenita (PC). PC is caused by mutations in keratin (K) 6a, K6b, K16 and K17 genes, including single nt changes, which result in dystrophic nails, plantar keratoderma, and painful blisters on the feet. Transfection of a K6a/YFP fusion gene resulted in normal keratin filament formation as assayed by fluorescence microscopy. Similar constructs containing a single nt change (N171K) showed keratin aggregates. A sequence walk of all possible N171K mutant-specific siRNAs identified inhibitors that showed single nt specificity. To test whether lead siRNAs can discriminate mutant mRNA in the presence of both wt and mutant forms, a dominant negative tissue culture model was developed. As predicted for a dominant negative disease, expression of both wt and mutant K6a forms resulted in defective keratin filament formation (keratin aggregates observed in 80–90% of cells). Addition of the lead K6a mutant-specific siRNA allowed normal filament structure formation (only 21% of cells contained keratin aggregates compared to 84% in the presence of a non-specific control siRNA).
In order to test the effectiveness of the lead inhibitors in an animal model, a bicistronic reporter construct was made consisting of firefly luciferase linked to either wt or N171K mutant K6a. Co-delivery of these constructs with perfectly matched siRNAs (e.g. mutant siRNA targeting the K6a N171K mutation or wt siRNA targeting wt K6a) resulted in potent inhibition as assayed by the Xenogen IVIS in vivo imaging system. In contrast, a single nt mismatch (e.g. mutant siRNA targeting the wt K6a or vice versa) resulted in little or no inhibition. Our results indicate that siRNAs can discriminate single nt mutations in animals and further suggest that “designer siRNAs” may allow effective treatment of a host of genetic skin disorders including PC.
Abstract ID: 837 Poster board space: 113
Apoptosis, an essential cellular process for maintaining normal physiology, has garnered intense attention in neoplastic development and response to therapy. Though many of the intracellular networks that govern this pathway have been well elucidated in vitro, the ability to examine this pathway in vivo is severely lacking. In a previous report, a recombinant luciferase reporter, activated during apoptosis via caspase-3 cleavage, was developed for imaging of apoptosis using bioluminescence (BLI). The ability to non-invasively image apoptosis in living animals allows for unique observations in vivo, which could benefit the preclinical development of therapeutics targeting the apoptotic pathway.
5-Fluorouracil (5-FU), a common anti-metabolite, has been reported to sensitize certain cell lines to TRAIL/Apo2L induced programmed cell death. In this study, we examined the effects of 5-FU on sensitizing subcutaneous human glioblastoma xenografts to TRAIL therapy by non-invasively observing the activation of apoptosis using our BLI reporter system. Using different regimens of 5-FU, TRAIL, and the combination of the two, we demonstrated that concomitant 5-FU and TRAIL therapy did indeed enhance apoptotic activity in vivo. Moreover, this technique shed light on the synergy of 5-FU and TRAIL as evidenced by differences in the temporal activation of caspase-3 resulting from the different therapeutic regimens. These findings were corroborated by an enhanced tumor growth delay in response to concomitant 5-FU and TRAIL. Finally, diffusion MRI, a clinically applicable biomarker of tumor response, was employed and demonstrated that combination of 5-FU and TRAIL was more efficacious than TRAIL alone.
Abstract ID: 838 Poster board space: 114
Micro computed tomography (micro-CT) for morphological and functional analysis of mouse and rat models of human disease has evolved dramatically in the past few years as a powerful tool for evaluating the efficacy of novel therapeutics. However, due to the insignificant difference in X-ray attenuation within soft tissues and blood vessels, non-contrast imaging is of a limited value unless intended for the imaging of bony structures. Intravascular injection of water-soluble iodinated contrast media is the commonly used clinical method to opacify blood vessels. Following injection, contrast is distributed rapidly between circulating blood and other extracellular fluids, which allows visualization of anatomical structures of the perfused organs. These agents are excreted through the kidneys following intravascular administration, and in the absence of renal dysfunction, have very short elimination half-lives. While such transient changes in the levels of contrast within body compartments are not critical in clinical settings due to the short imaging times of clinical CT systems, these changes can significantly impair data quality acquired with slow micro-CT scanners. In order to maintain consistent levels of contrast throughout the entire imaging period, several research groups have used a continuous intravenous infusion of clinical radiographic agents up to one hour. However, this approach results in the administration of a maximum-tolerated infusible volume (25 ml/kg, or 0.75 ml per 30-g mouse), with doses of iodinated agents high enough to cause renal hemodynamic changes and to induce adverse events such as acute renal failure. Here we present the effect of high-, low-, and isoosmolar contrast media administration in normal C57BL/6 mice on urine creatinine levels as a measure of radiocontrast induced nephropathy. We speculate that these agents, in addition to being nephrotoxic, may also significantly alter the pharmacokinetics, toxicity profiles, and maximum-tolerated dose of drugs tested in mouse and rat animal models.
Abstract ID: 839 Poster board space: 115
We have combined two imaging probes and used PET as a means to provide image-based validation for a novel targeted drug delivery system. The first probe was a direct labeling of the drug [18F]paclitaxel, which was inserted into various carrier vehicle formulations. The second probe, [18F]-1,2-dipalmitoylglycerol, ie. [18F]DP involved radiolabeling a component of the vehicle itself. Free paclitaxel has a biphasic clearance from plasma, with alpha and beta half-lives on the order of minutes and 1 hour, and therefore altered pharmacokinetics caused by encapsulation of paclitaxel can be explored with 18F labeling. Paclitaxel, which is poorly soluble in aqueous media, also has limited solubility and stability in lipophilic environments such as liposomes. Its stable association with the lipid bilayer is greatly affected by a variety of physicochemical factors such as temperature and liposome composition. Here we provide dynamic in vivo image sets providing biodistribution and time activity curves of free [18F]paclitaxel and liposomal [18F]paclitaxel as well as free [18F]DP, liposomal [18F]DP, and [18F]DP in an ultrasound contrast agent. Serial studies were performed within the same three rats, minimizing inter-animal variability. The two labeled molecules have different biodistributions: paclitaxel is rapidly taken up in the liver, intestines and kidneys (right image) while the labeled lipid stays in circulation with minimal uptake in organs other than spleen (left image). A variety of vehicles are under development to carry paclitaxel to a target site and have demonstrated success in the in vitro setting. Here, we have developed a quantitative method to follow paclitaxel and lipid vehicles to their destination in vivo in order to improve targeted paclitaxel delivery.
Abstract ID: 840 Poster board space: 116
Pre-clinical imaging continues to play an increasingly significant role in the development of new therapeutic compounds. As is typical of most in vivo studies, biological variation inherent to the response measurement requires perhaps dozens of animals to produce data with sufficient statistical power. With the development and evolution of new screening applications of imaging technologies, a challenging constraint that imaging sites will need to overcome is the relatively low throughput of small animal scanners. To address this issue, we have initiated efforts to develop technologies and processes to increase the throughput of pre-clinical imaging studies. We report here current accomplishments in multiple-animal, parallel imaging with micro-PET and MRI.
For our micro-PET experiments, we have established animal handling and experimental work flow routines that enable preparation, radiotracer injection and scanning of two mice simultaneously whereby proper physiologic conditions are monitored and maintained over the course of the experiment. This approach is now being routinely used in our micro-PET laboratory to i) monitor efficacy of compounds in pre-clinical models of cancer; and ii) measure gene expression levels in transgenic mouse models.
In the MR realm, we have implemented simultaneous four mouse scanning, with appropriate physiological monitoring and gating capabilities, through the use of a multi-animal RF coil assembly (Rapid Biomedical) comprised of four independent, quadrature driven high pass birdcage resonators arranged in a square grid pattern. The coil assembly is interfaced to a Bruker Biospec 9.4T/310AS system equipped with a single RF transmitter, four receive channels and GRAPPA acquisition/reconstruction protocols.
Simultaneous, multi-animal scanning not only shortens the overall study duration but also allows one to compare treatment and control animals under identical scan conditions and achieve temporal resolution in certain time-course studies that otherwise may not be possible. Details of our high-throughput imaging set-up and their integration in drug discovery research will be presented.
Abstract ID: 841 Poster board space: 117
The development of new imaging and therapeutic agents with capabilities for precise targeting of bones is vital in the care and management of bone cancer. Due to their antiresorptive and bone protecting activity, Geminal bisphosphonates (Gbp) are extensively used in the treatment of bone related diseases such as osteoporosis and bone cancer. In order to investigate the potential of increasing x-ray contrast in computed tomography (CT) we have studied gold nanoparticles labeled with bone seeking Gbp. In the present work, we synthesize and characterize gold nanoparticles labeled with Gbp (GbpAuNPs) and study their stability. GbpAuNPs are stable over a 7-day period, as confirmed by monitoring the plasmon absorption and through transmission electron microscopy (TEM) measurements. For imaging applications, it is critical that GbpAuNPs are stable under in vivo conditions. An in vivo environment is mimicked by placing GbpAuNPs in saline and cysteine solutions at various concentrations. The stability of GbpAuNPs in saline and cysteine solutions are monitored by width of the plasmon absorption band and through TEM measurements. The GbpAuNPs were found to be stable over a period of 48hrs, suggesting their potential applicability under in vivo conditions for molecular imaging via CT technique. Biodistribution studies of GbpAuNPs and investigations of their potential use in CT imaging in small animal models are under way.
Abstract ID: 842 Poster board space: 118
Osteoarthritis (OA) is a debilitating disease characterized by articular cartilage degradation (including disruption of the collagen network and proteoglycan erosion). Proteolytic enzymes, such as those in the cathepsin and MMP families, are thought to contribute to the pathogenesis of OA. This presentation reports preliminary results from an optical imaging study using a cathepsin activatable fluorophore probe in an experimental mouse model of OA involving rupture of the anterior cruciate ligament (ACL) of the knee.
Swiss-Webster mice were serially scanned at baseline (pre-ACL rupture), 7-, 14- and 21-days post ACL rupture after receiving a tail vein injection (2 nmol/150 ml 1X PBS) of the cathepsin activatable probe (ProSense750, VisEn Medical, Woburn, MA) 24-hours prior to the imaging session. The resulting image time-course shows an increase in fluorescence signal in a localized region encompassing the articular joint relative to the baseline time point. The signal intensity peaked at the 7-day point and then became less intense and increasingly diffuse at the 14- and 21-day points. The fluorescence intensity peak at the 7-day point corresponded with cartilage degradation from histopathology analysis. In contrast, the fluorescence signal from naïve control mice did not show localization to the knee region and remained consistently lower than the experimental group over the course of the four week study.
These results suggest that the onset of proteolytic activity in this model likely occurs within a matter of days following the ACL rupture procedure and the duration of activity persists over an approximate two-week period. We are currently evaluating the utility of this technology as an early marker of drug efficacy in pre-clinical models of OA.
Abstract ID: 843 Poster board space: 119
Cellular proteases are crucial to proper functioning of a number of physiological processes including development, digestion, immune response, blood clotting, and skin and bone formation. The necessary step in maturation of proproteins, proteolytic processing, occurs mainly in the Trans-Golgi Network (TGN) where prodomains are cleaved resulting in mature peptides. The TGN contains many proprotein converting enzymes, such as carboxypeptidases, members of the prohormone convertase family, and the B-site amyloid precursor protein (BACE) family. Cell based assays used to report on cellular protease activity allow monitoring of enzyme activity within its natural environment. We have constructed a versatile and sensitive assay system whereby the activity of TGN associated proteases can be monitored using cells containing an engineered fusion protein that is retained in the TGN, contains a proteolytic cleavage site, and upon cleavage, secretes alkaline phosphatase, whose activity can be assayed using the media of undisturbed cells. Levels of alkaline phosphatase present in the media are indicative of protease activity and can be assayed to measure activity and identify inhibitors. We have used this system to monitor the activity of and identify novel inhibitors of two TGN proteases, BACE and Furin and have adapted this assay for use in high-throughput screening where we have identified novel small molecule inhibitors of Furin. We show that these molecules inhibit Furin both in vitro and in vivo. We propose that this system can be used to monitor and report on the activity of other golgi-associated proteases.
Abstract ID: 844 Poster board space: 120
Newer and more diverse nanomaterials are being developed at an unprecedented pace, and one challenge is to identify properties which govern nanoparticle behavior in living systems. In an ongoing project, small animal PET imaging of radioactive nanoparticles directs the cellular and ultrastructural localization of systemically administered particles. Parallel molecular analyses of the same tissue, also guided by the PET images, gives the expression levels of genes involved in host/pathogen interactions. Specifically, the kinetics and distribution of carbon-11 labeled nanoparticles are observed in vivo and the tissue then analyzed using transmission electron microscopy (TEM), X-ray microanalysis and finally with DNA microarray. Thus far, 4.0 and 9.0 nm particles made of different core materials (cadmium/selenium with a zinc shell, CdSe/ZnS; or gold, Au) have been successfully labeled with carbon-11 using an end-chain radiolabeling approach. Initial experiments employed the evaporative loss method to confirm that the radiolabel remains with the nanoparticle in systemic circulation. Where possible (when fluorescent probes are radiolabeled), PET radioactivity distribution has been verified using fluorescent microscopy. TEM analysis of labeled nanoparticle solutions indicates that they are monodisperse at injection. Labeled nanoparticle distribution varied slightly between the two sizes, with the larger particles retained in the cortex of the kidney while smaller particles passed readily through. When coated with surfactant, [11C]-CdSe/ZnS readily penetrated the blood brain barrier (BBB), with peak brain concentrations (0.1% of injected dose) occurring within 3 min. A similar effect was observed with [11C]-Au, where surfactant coating significantly altered the distribution and accumulation of both smaller and larger particles. Thus, the immediate in vivo distribution of these particles appears to be dictated by size and surface chemistry rather than by core material. Image-guided nanoparticle localization is a novel approach that has the potential to give important new information about nanoparticle interactions with living systems.
Abstract ID: 845 Poster board space: 121
Koon Yin Kong1, Adam Marcus2,
Advance in fluorescence microscopy technology allows microtubules dynamics to be captured and to be used to study the effectiveness of microtubule-targeted anticancer drug. We propose an automatic microtubule quantification algorithm that can replace current manual quantification system, which is time-consuming and error-prone. The algorithm will automatically enhance, segment, and track the tip of microtubules using open active contours. We find that open active contour is better suited to represent narrow, curvilinear microtubules structures than traditional close contour. Microtubules are enhanced in the microscopy images using directional matched filters and Hessian matrix. Hough transform is used to locate candidate microtubule segments. We redefine new internal energy terms specifically for open contour, and redesign new external energy terms, based on our previous work, so that the contour converges to both a shrinking and growing microtubule. Simulated images, untreated MCF-7 breast cancer cell lines, and cell line treated with the microtubule-targeting chemotherapeutic agent, Taxol, are used to validate our algorithm. Quantification result can be visualized by highlighting detected microtubules throughout the image sequence, or converted to metadata in preparation for large-scale statistical analyses. Automated quantification system makes analysis of large image data possible and accelerates drug development process.
Poster Session III: P16: Imaging in Neuroscience
Abstract ID: 848 Poster board space: 176
Abstract ID: 849 Poster board space: 177
The thalamus is the central relay station for the brain, and there is evidence of its change with aging and laterality. The nuclei of the thalamus respectively comprise fibers connecting associated cortical regions, and it is known that abnormalities of the thalamus are correlated with abnormalities in cognition and behavior. Yet, little is known about the differences in the tissue characteristics of thalamic nuclei. We assessed regional characteristics in five major subregions [1,2] of both the left and right thalamus using diffusion-tensor imaging. We evaluated 16 healthy right-handed males aged under 40 years for the lateralities and regional differences of the thalamic subregions. The significant differences among the ractional anisotropy (FA) and mean diffusivity (MD) of the divided regions of the thalamus were obtained (F=44.1, df=9, 150. p< 0.001. F=48.1, df=9, 150, P< 0.001, ANOVA, respectively). The MD value of each subregion was lower in the left side than in the right side except for the posterior (Po) subregions. FA values of anterior lateral (AL) and posterior (Po) subregions were higher in the left side than in the right side. There were significant differences in FA and MD values among the subregions, respectively. These results suggest that both sides of thalamic nuclei behave independently, and that we should evaluate the respective microstructures of the thalamic nuclei.
Abstract ID: 850 Poster board space: 178
Abstract ID: 851 Poster board space: 179
Abstract ID: 852 Poster board space: 180
Inflammation is a double-edged sword. After an ischemic insult, inflammation contributes to the delayed lesion growth in the late stage of damage. But some of the inflammatory processes seem to take part in neuroprotection, tissue repair and recovery of function. At present, it is unclear whether detrimental effects of inflammation outweigh neuroprotective mechanism or vice versa. To further understand the role of brain inflammation in stroke, non-invasive imaging strategies to visualize these processes are highly desirable.
In the present study, we evaluated stroke-induced brain inflammation in a mouse stroke model using non-invasive planar near-infrared fluorescence (NIRF) imaging and a monoclonal antibody against CD40 labeled with the NIRF dye Cy5.5. The fluorescent reporter agent was injected intravenously 96 hours after transient middle cerebral artery occlusion (MCAO) in C57Bl6 mice. The imaging was performed 16 hours after intravenous injection of the compound. In MCAO-mice, significantly increased fluorescence was detected through the skull and skin over the affected hemisphere (p < 0.007). Corresponding ex vivo NIRF images of the brain and brain sections confirmed localization of the fluorescence in the ischemic territory. MCAO-mice receiving Cy5.5-labeled IgG as a control for non-specific dye extravasation, binding, and accumulation did not show fluorescence enhancement over the affected hemisphere in vivo. Fluorescence microscopy and immunohistochemistry revealed the specific delivery of the reporter agent to activated endothelium and to resident and blood-derived cells of the mononuclear lineage. In conclusion, the results show that non-invasive NIRF imaging can be used to visualize stroke-induced brain inflammation in mice with high sensitivity and specificity. To our knowledge, this is the first study demonstrating ligand-based optical molecular imaging of inflammation in the mouse brain. Since the reporter agent can be labeled with a radionuclide, CD40-targeted antibodies might also be used for human brain imaging purposes.
Abstract ID: 853 Poster board space: 181
Abstract ID: 854 Poster baord space: 182
The aim of the study was to examine the effects of smoking on brain metabolism of healthy persons using 1H NMR in vivo. The NMR results were compared with those of the metabolomics approach.
The metabolic disturbances were confirmed using Principal Component Analysis. The data clustering is seen in the PC1 vs. PC2 and PC1 vs. PC3 plots (Figure 1). As revealed from the loadings' plots, the spectral segment of the largest differences falls between 0.8 and 1.8 ppm, which corresponds to the lipids and alanine chemical shifts range. Thus, these compounds would be the main “biomarkers” of brain harm due to smoking.
A metabolomics approach that simultaneously observes multiple metabolites and requires no spectral resolving seems to be helpful in classifying the smokers and non-smokers spectra and identifying the spectral regions of importance.
Abstract ID: 855 Poster board space: 183
Brain diseases are characterized by neuronal cell death implicating an active process which is called programmed cell death or apoptosis. During this process, specific enzymes, especially caspase-3, are activated in order to disintegrate cell components. Targeting this enzyme enables to discriminate the cells dead by apoptosis from the one dead by necrosis. One of the aims of our project is to develop molecular probes able to detect in vivo and with high specificity apoptotic cell death during neurodegenerative diseases using optical imaging. The probes currently under development designed to measure caspase-3 activity are labeled with fluorochromes emitting in the near-infrared spectrum range and use fluorescence resonance energy transfer (FRET) technology.
Tests performed on cultured cerebellar granule cells reveal that the probe is penetrating into the cells where it gets activated and does not interfere with neuronal survival. 100% efficiency of quenching is observed on native state and an increase of the probe donor emission is measured after its cleavage by cells. Induction of apoptosis in cultured neurones by an oxidative stress increase dramatically the cleavage rate of the probe when compared to control cells.
Preliminary in vivo tests on mice using a custom optical imaging system with high sensitivity and resolution indicate that this probe injected into animals after ischemia is able to highlight the infarct zone. Detection of the probe in the hippocampus and striatum is possible in the concentrations of micromolar order. The probe's ability to penetrate the brain through the blood brain barrier after peripheral injection is under evaluation.
Abstract ID: 856 Poster board space: 184
Abstract ID: 857 Poster board space: 185
Abstract ID: 858 Poster board space: 186
Abstract ID: 859 Poster board space: 187
Chronic or acute neuroinflammation has been identified in several degenerative disorders (e.g. Alzheimer and Parkinson diseases) and in stroke. Peripheral benzodiazepine receptor (PBR) is mainly expressed by activated microglia in the brain, and is therefore a marker of neuroinflammation.
This work aimed at evaluating by positron emission tomography (PET) two new PBR radioligands, namely [11C]DPA-713 and [11C]CIINME, and at comparing them to the reference radiotracer [11C]PK11, 195, which use is hampered by a high non specific signal giving rise to a difficult quantification of its pharmacological parameters. The three compounds were tested in unilaterally striatum-lesioned rats.
Imaging of the lesioned rats shows a higher contrast between the lesioned area and the symmetrical area in the intact contralateral hemisphere with DPA-713 (Figure 1) and CIINME (Figure 2), compared to PK11,195 (ratio ipsi/contro 20 min post-injection: DPA-713, 2.7 ± 0.6, n = 4; CIINME, 2.2 ± 0.3, n = 4; PK11,195, 1.7 ± 0.1, n = 5). DPA-713 is totally displaced by an excess (1 mg/kg) of PK11,195 or DPA-713. These displacements also occur with CIINME, with slower pharmacokinetics compared to DPA-713. Immunohistochemistry correlates with PET imaging by showing strong activation of microglia in and around the lesion in rats.
This preliminary study in a rodent model of neuroinflammation demonstrates that those two new tracers give rise to a better signal/noise ratio than PK11,195. They thus have the potential to improve the efficiency of neuroinflammation imaging.
Supported by EMIL (European Molecular Imaging Laboratories) EU contract LSH-2004-503569.
Abstract ID: 860 Poster board space: 188
Stroke therapy can only be effective if viable and potentially salvageable brain tissue (ischaemic penumbra) is present. However, current brain imaging techniques have limited ability to image the penumbra.
The magnetic properties of hemoglobin, oxyhemoglobin and deoxyhemoglobin (the basis of the BOLD signal) were demonstrated in 1936 making it possible to follow oxygen metabolism in vivo using MRI. We propose to combine oxygen challenge with BOLD imaging to improve identification of the penumbra ischaemic penumbra.
Abstract ID: 861 Poster board space: 189
Transplantation of genetically manipulated cells to the central nervous system offers immense potential for the treatment of several neurological disorders including brain tumors. The success of this strategy relies on sufficient levels of transgene expression and understanding the fate of NSCs after transplantation. Using in vivo marking studies, we show that both human NSCs and glioma cells can be efficiently engineered with lentiviral vectors (LVs) expressing fusions between in vitro and in vivo marker genes (GFP-Rluc and DsRed2-Fluc). We have used a malignant DsRed2-Rluc expressing human-glioma model and implanted human NSCs expressing GFP-Fluc intracranially and shown that these cells survive better in the presence of gliomas and migrate extensively towards and into glioma tumors. The migration of NSCs into the tumors and the tumor mass could be visualized in real time by in vivo bioluminescence imaging and intravital microscopy. These studies demonstrate the application bimodal LVs in evaluating the fate of NSCs and changes in glioma-burden in vivo and provide a platform for accelerating cell based therapies for neurological disorders.
Abstract ID: 862 Poster board space: 190
Norman Kock, Randa Kasmieh, Ralph Weissleder,
We have previously shown that the secreted form of TRAIL (S-TRAIL) has enhanced apoptosis inducing ability when delivered to tumors via neural precursor cells or by viral means. In this study we explored the cytokine death-signal transduction pathways in glioma cells and investigated the inhibitory effects of Bcl-2 on S-TRAIL induced apoptosis in human gliomas. We have employed a lentiviral-delivery system for the expression of short-hairpin RNAs (shRNA) to specifically down regulate Bcl-2 expression. Using a highly malignant human glioma model expressing δEGFR and firefly luciferase (Fluc), we show in real time that glioma cells expressing Bcl-2-shRNA and S-TRAIL have enhanced cell killing ability both in vitro and in vivo. Furthermore we demonstrate that the down regulation of Bcl-2 itself does not lead to altered morphology or glioma cell death; however expression of activated Bid (tBID) was increased, thus resulting in accelerated S-TRAIL mediated apoptosis. These results reveal that simultaneous triggering of TRAIL mediated death receptor and mitochondrial apoptotic pathways leads to enhanced eradication of gliomas that can be visualized in real time in vivo. This study may have clinical implications for the treatment of resistant cancers.
Abstract ID: 863 Poster board space: 191
One of the main limitations to the development of therapy for tumors has been the lack of noninvasive imaging methods to monitor gene delivery, specificity and duration, as well as to assess tumor cell death. The aims of this study are 1) to develop non-invasive optical imaging of gene delivery; and 2) to asses therapeutic properties of S-TRAIL (secretable tumor necrosis factor related apoptosis inducing ligand) delivered viral vectors on gliomas. We have developed lentiviral (LV) and adeno-associated viral (AAV) delivery systems that simultaneously allow imaging of two different molecular events both by bioluminescence imaging and by intravital microscopy in real time. Human glioma cells infected in culture with the LVs and AAVs bearing S-TRAIL cloned upstream of both bioluminescent (Fluc, Rluc and Gluc) and fluorescent (GFP, DsRed2 and Tdtomato) markers showed robust expression in glioma cells that could be followed both in culture and in vivo by real time imaging. Furthermore, we show that expression of the S-TRAIL via these vectors induces caspase-3 mediated apoptotic cell death that can be imaged both in culture by luciferase based cell viability assays and in vivo by following the glioma volumes by real time bioluminescence imaging. This study demonstrates the feasibility of using viral vectors to simultaneously express therapeutic genes for tumor specific cell killing and dual imaging markers for monitoring therapeutic gene delivery and glioma fate in vivo.
Abstract ID: 864 Poster board space: 192
PET scanners specifically designed for small animal imaging have been developed with high sensitivity and satisfactory spatial resolution. The drawback of PET imaging technology is the lack of morphological information which makes it hard to localize exactly the regions of interest. In an effort to obtain both the functional and structural information, a microPET/CT system was developed in our Institute, in which an in-house CT is combined with a microPET R4 scanner from Siemens Concorde. To demonstrate the feasibility of this integrated microPET/CT modality, we co-registered the data from 18F-FDG/PET and CT into fused images to examine the infarct volumes in ischemic stroke rat.
Male Sprague-Dawley rat was subjected to both common carotid arteries (CCAs) and right middle cerebral artery (MCA) occlusion for 60 min followed by reperfusion. PET/CT scans were performed at 24 hours,72 hours and 96 hours after surgery. The rat was then administrated 85 MBq of 18F-FDG via tail vein and image was acquired with a Concord R4 microPET scanner. An x-ray computed tomography was acquired after the PET scan. Isoflurane anesthesia was maintained during data acquisition. The co-registered PET/CT image was used to help determining anatomic localization of radiotracer uptake. Infarct area was evaluated by comparing %ID/mL in the right side infarct region to the opposite side intact region.
Images of 18F-FDG-microPET co-registered with CT allowed depicting the brain area under skull and the infarct regions were found at right side frontal cortex as expected. Uptakes of 18F-FDG in lesion areas were decreased up to 15% when compared with the opposite intact side. Recovery from the lesions can be detected and quantified at 72 hours and 96 hours post ischemic procedure.
Our results revealed the 18F-FDG MicroPET/CT technique provides a sensitive and reliable way in terms of localizing the cerebral ischemic lesions in rodents.
Abstract ID: 865 Poster board space: 193
Abstract ID: 866 Poster board space: 194
Abstract ID: 867 Poster board space: 195
Abstract ID: 868 Poster board space: 196
Ependymal cells are derived from radial glial cells during brain development and adult ependymal cells are postmitotic and highly differentiated. Radial glial cells are neurogenic precursors. We tested the hypothesis that stroke induces adult ependymal cells to transform into radial glial cells. Adult rats were subjected to embolic middle cerebral artery occlusion and ependymal cells were analyzed by means of the patch clamp, time-lapse microscopy and laser scanning confocal microscopy (LSCM). We found that ischemic stroke acutely stimulated ependymal cell proliferation, and dividing ependymal cells of the lateral ventricle had genotype, phenotype and morphology of radial glial cells. The majority of radial glial cells exhibited symmetrical division with respect to the cell cleavage plane, and a vimentin-positive radial fiber was maintained throughout each stage of cell mitosis. Radial glial cells were surrounded by clusters of highly proliferative transit-amplifying cells (type C cells) and neuroblasts (type A cells) in the ipsilateral subventricular zone (SVZ), underlining their critical role in stroke-induced neurogenesis. LSCM revealed that stroke substantially induced radial glial fibers in the ipsilateral SVZ and striatum, which were associated with leading processes of migrating neuroblasts oriented to the ischemic boundary. Time-lapse microscopy shows that radial glial cells derived from stroke SVZ supported neuron migration. These results indicate that adult ependymal cells divide and transform into radial glial cells after stroke, which could function as neural progenitor cells to generate new neurons after stroke and act as scaffolds to support neuroblast migration towards the ischemic boundary region.
Abstract ID: 869 Poster board space: 197
DTI and fiber tracking is a promising method for displaying MS-related pathological changes of white matter tracts. This may lead to an improved monitoring of neurological deficits in MS.
Abstract ID: 870 Poster board space: 198
We developed novel PET ligands for CNS-type prostacyclin receptor, (15R)-16-tolyl-isocarbacyclin (15R-TIC) and 15-deoxy-16-tolyl-isocarba-cyclin (15D-TIC) [1,2], and succeeded in PET molecular imaging in animals and human brain by using their methyl esters (15R-TIC-Me and 15D-TIC-Me) as a pro-drug. 15R-TIC and 15D-TIC are potent neuroprotective agents in primary cultured neurons [2], gerbil transient ischemic model [2], rat focal ischemic model [3], rat beta-amyloid model, and monkey MCAO model. In the course of development of the efficient derivatives for 15R-TIC and 15D-TIC in terms of penetration to the blood-brain barrier (BBB), we found species-specific nature of uptake between two ligands. The methyl esters of 15R-TIC and 15D-TIC were highly penetrable than the carboxylic acid form of both compounds in every species, mice, rats, and monkeys. However, 15R-TIC-Me was more efficiently penetrable to BBB than 15D-TIC-Me in monkeys, but 15D-TIC-Me was a few times efficient for penetration in rodents. The half life of the ester moieties of both TIC compounds in the blood circulation was of course different among species, but there was no difference between the de-esterification (the possibly sole metabolism in the time window of PET) rates of the two compounds in the same species. The trends were also confirmed by dose-dependency of neuroprotective effects by 15R-TIC-Me and 15D-TIC-Me in rats and monkeys. This information well illustrates the significance and potential of molecular imaging for better drug development for human.
Abstract ID: 872 Poster board space: 200
Altered expression of endogenous genes in the brain often accompanies acute or chronic neurological disorders. Some changes in expression are linked to neuronal activities and others to death or repair processes. In this study, we utilized a novel MR contrast probe to detect the effects of an acute dose of amphetamine (4mg/kg, i.p.) on cerebral gene transcripts in live mouse brains. Adult C57b6 mice were infused intracerebroventricularly (ICV) with an iron-based MR T2 susceptibility agent (SPION)-labeled phosphorothioate-modified oligodeoxynucleotide (oligo) four hours prior to amphetamine or saline (vehicle) injection. Two types of probes were used: a c-fos mRNA targeting probe (SPION-cfos) and a non-targeting probe with randomized sequence (SPION-Ran). We analyzed R2* maps of live mouse forebrain at different time points after amphetamine administration. While we did not observe significant elevations in R2* values between the amphetamine- and vehicle-treated animals previously infused with SPION-Ran in all time points, we found that amphetamine-treated animals previously infused with SPION-cfos exhibited significantly elevated R2* in the retrosplenial cortex (RS), nuclear accumbens (NAc) and caudoputamen (CPu), compared to the vehicle-treated animals at three hours. Preliminary data indicated a more pronounced difference at 24 hours post infusion in brain regions including the cingular (cing), motor and sensory cortices (M/SC). Brain regions exhibiting elevated R2* are consistent with reports by other investigators using 35S-riboprobes on postmortem rodent brain samples for detection of elevated c-fos mRNA levels after a similar stimulation paradigm, which included but were not limited to CPu, NAc, cing and SC [1–3]. These data imply that the SPION-labeled oligo can facilitate in vivo MR detection of endogenous gene up-regulation due to acute amphetamine stimulation. (NSR01045845, P41RR14075 and the MIND Inst).
Abstract ID: 873 Poster board space: 201
This study quantitatively analyzes the regionally dependent clearance characteristics of Gd-chelates in rat brain. The blood brain barrier precludes delivery by vascular routes so agents were infused directly into ventricular cerebrospinal fluid yielding useful quantities of agent throughout the brain parenchyma for many hours after infusion. Identical quantities (30âμL,250mM) of each respective agent, GdDTPA-BMA, GdDTPA2-, or GdDOTP5-, were delivered by direct infusion at a constant rate (6âμL/hr) over 5 hours into the left lateral ventricle. Experiments were performed with Sprague-Dawley rats and agent distributions in the brain were obtained from MR T1 maps (Turboflash) acquired at 9.4T. Qualitative results indicate that all 3 agents readily distribute throughout brain regardless of charge-state, that equivalent agent doses result in similar decreases in T1 throughout the brain for GdDTPA2- and GdDTPA-BMA, while the decrease in T1 is as much as 2 to 3-fold greater with GdDOTP5-, that differences in in vivo T1 change cannot be accounted for by differences in measured in vitro relaxivities, and that it is possible to deliver Gd-chelates in a dose-dependent manner with a linear response in brain T1. Rate constants of contrast agent clearance from specific regions of interest throughout the brain were obtained from exponential fits to time-course curves for each ROI (below, left). While agent charge does not preclude agent exchange across the negatively charged ependymal membrane, it does influence the clearance rate, which was slowest for the more negatively charged GdDOTP5- (z-profile of the clearance rate constants on the right).
Abstract ID: 874 Poster board space: 202
We investigated effects of a single nucleotide polymorphism (SNP) genotype at the alpha7 nicotinic cholinergic receptor (CHRNA7) locus on visuo-spatial memory, brain morphology, and fMRI brain activation patterns. Thirty healthy adults were genotyped at a cytosine-to-thymine (rs868437) SNP at the CHRNA7 locus and underwent cognitive testing, structural and functional MRI (fMRI) scanning. Morphological analysis revealed SNP-related gray-matter volumetric differences in specific posterior regions (i.e. middle/inferior occipital gyrus, precuneus and lingual gyrus). SNP rs868437 differentiated scores on an in-scanner figural memory fMRI paradigm (FIGMEM) where stimuli were abstract shapes designed to have low verbal encodability. CHRNA7 genotype effect was specific to FIGMEM in that it did not discriminate out-of-scanner IQ scores or memory scores from the Mini Mental Status Examination, or the Wechsler Memory Scale-Revised. Subjects who scored poorly on FIGMEM generally had greater regional brain activation than did the high scoring group during encoding and recognition. Principal component analysis followed by hierarchical regression found that genotype and fMRI activation during the encoding period of FIGMEM accounted for as much as 50% of the variance in the prediction of FIGMEM recognition scores. Regional brain activation associated with FIGMEM scores was generally consistent with the cortical differences, showing relationships in both posterior (i.e., fusiform gyrus, inferior occipital gyrus, and the precuneus) and anterior (medial frontal gyrus, body of the caudate, and paracentral lobule) regions. Interestingly, when an expanded sample (sixty six healthy adults) was examined for genotypic-related gray-matter patterns, volumetric differences were observed in the posterior regions identified above, plus the parahippocampal gyrus and hippocampus. This pattern is consistent with regions implicated by fMRI activation. Overall, these results, using both structural and functional MRI, suggest that the CHRNA7 plays an important role in brain areas associated with encoding and recognition of nonverbal, complex visual-spatial figures in healthy adults.
Abstract ID: 875 Poster board space: 203
Poster Session III: P17: Mouse Models in Human Diseases
Abstract ID: 878 Poster board space: 204
Upregulation of the heme oxygenase-1 (HO-1) pathway has been identified as one of the critical cytoprotective mechanisms activated during cellular stress. In human liver transplantation, it has been shown that HO-1 expression in liver grafts before transplantation correlates with graft injury and function post-transplantation. In this study the feasibility of in vivo bio-luminescence imaging (BLI) has been investigated to non-invasively image and quantify the spatiotemporal expression of endogenous HO-1 in an established animal model of warm hepatic ischemia.
Representative serial in vivo expression of HO-1 in the model of partial hepatic (left lobe) ischemia/reperfusion in FVB/N-Tg (Hol-luc) mice.
Localization of signals from reporter gene activation using image-guided dissection, confirming that the postischemic left liver lobe is indeed the principal source of the bioluminescent signal. In vivo bioluminescent imaging of HO-1 expression in FVB/N-TG (Hol-luc) mice in left lateral, right lateral and supine position (panel A) and with open abdomen (panel B) at 9 hr after reperfusion. Exposure integration time for images was 10 seconds. Arrow identifies the postischemic left lobe.
Luciferase activity was measured by BLI in transgenic reporter mice (Ho1-luc) at standardized time points after 60 min of warm hepatic ischemia. Besides, HO-1 mRNA levels were measured in post-ischemic livers of wild-type FVB/N mice at the same time points. Bioluminescent signals from post-ischemic liver lobes were first detected at 3 hr post reperfusion. Peak levels were reached at 9 hr and returned to baseline values at 48 hr. Images in the left and right lateral position, supine position and with open abdomen at 9 hr after reperfusion, confirmed that the post-ischemic hepatic lobe was the principal source of bioluminescence. Changes in HO-1 expression detected by BLI were confirmed by enhanced immunohistochemical staining of hepatocytes. Expression of HO-1 mRNA in wild-type FVB/N mice increased immediately after ischemia/reperfusion, reaching a peak at 3 hr after reperfusion and 6 hr before the peak in BLI signal was detected. In conclusion, in vivo BLI allows a quantitative assessment of HO-1 expression after hepatic ischemia in living animals. The capability of whole-body temporal imaging of HO-1 expression provides a novel tool in the development of new strategies to modulate HO-1 expression in liver transplantation.
Abstract ID: 879 Poster board space: 205
Abstract ID: 880 Poster board space: 206
This model also allows to study the inhibitory effects of sclerostin and to develop and evaluate new bone formation modulating drugs.
Abstract ID: 881 Poster board space: 207
Daniel Eefting1,
Abstract ID: 882 Poster board space: 208
Molecular imaging provides a sensitive and minimally invasive technology for interrogating molecular events that impact disease over the lifetime of a single animal. We have developed a model for using bioluminescence imaging to visualize spontaneous tumorigenesis in vivo.
Transgenic mice expressing human T-cell leukemia virus type 1 transcriptional transactivator (HTLV-1 Tax) under the control of the human granzyme B promoter (GzBTax) develop large granular lymphocytic leukemia/lymphoma, splenomegaly, bone destruction, and tumors of the ears, legs, nose, and tail. Tax, through physical interaction with components of transcription pathways mediated by CREB and NFkB, transforms cells by dysregulating genes involved in proliferation, apoptosis, and cell-cycle control.
While transformation by HTLV-1 is relatively rare, the mechanism by which Tax transforms cells through NFκB is analogous to that of a wide range of hematopoietic malignancies including Hodgkin's and non-Hodgkin's lymphomas. In order to utilize Tax activity as a biomarker for tumorigenesis, we generated transgenic mice with firefly luciferase driven by the HTLV-1 long terminal repeat (LTRLUC) and bred them with the GzBTax animals to produce double transgenic LTRLUC/GzBTax (LLT) mice.
We have shown that the bioluminescence of the LLT mice i) serves as a reporter for spontaneous tumorigenesis in vivo, ii) is transferable as a useful tumor marker in cell culture and allograft models derived from primary tumors, iii) is inducible by immunomodulatory agents, and iv) is a sensitive indicator for the earliest events in tumor development.
Abstract ID: 883 Poster board space: 209
Cathepsin K is a cysteine proteinase that is selectively and highly expressed by osteoclasts and has been proposed to play a key role in bone resorption. The inhibition of cathepsin K is expected to suppress osteoclast-mediated bone resorption in post-menopausal osteoporosis. We have demonstrated that high-resolution micro-CT may serve as a valuable tool to evaluate the effects of anti-osteoporotic agents at an early stage of ovariectomy (OVX)-induced osteoporosis in mice. In this study, the inhibitory effect of SB-553484, a potent human and mouse cathepsin K inhibitor, on bone resorption in this model was evaluated by micro-CT bone morphometric analysis.
Female BALB/c mice were OVX- or sham-operated at 6 weeks of age. Each OVX group was then treated for 2 weeks with SB-553484 (30mg/kg, sc) or Rolipram (a PDE4 inhibitor which promote bone formation) (10mg/kg, po) twice a day. On day 14 after operation, the left femur bones were removed and then the distal metaphyseal bone was scanned by microCT at 8 μ resolution. The three-dimensional (3D) microstructural images of trabecular and cortical bone were reconstructed and analyzed.
Micro-CT analysis showed that SB-553484 significantly prevented the decrease of trabecular and cortical bone volume as well as the deterioration of trabecular architecture similar to Rolipram. Interestingly, SB-553484 showed more remarkable effect in improvement of trabecular separation, number and connectivity than Rolipram. On the other hand, SB-553484 had a weaker effect than Rolipram on improvement of trabecular thickness. These differences in the changes of micro-CT 3D bone parameters between SB-553484 and Rolipram indicate an inhibitory effect of SB-553484 on bone resorption.
Our data suggest that targeting cathepsin K may prove therapeutically beneficial in the treatment of diseases with bone loss such as post-menopausal osteoporosis.
Abstract ID: 884 Poster board space: 210
Circadian rhythms enable organisms to adapt to daily environmental changes such as light, temperature and social communication and serve to synchronize multiple molecular, biochemical, physiological and behavioural processes. The clock mechanism mainly involves an integrated network of interacting self-sustained transcriptional-translational feedback loops, composed of positive and negative regulators. By molecular biology techniques, it was shown that rev-erbα gene plays a major role in the biological clock as it is a clock controlled gene driving in turn clock gene expression. It is also suggested that rev-erbα gene can integrate several levels of regulation both at the circadian and physiological levels. Nevertheless, these experiments require the sacrifice of numerous animals and do not allow a longitudinal follow-up of the same subject. Subsequently adapting atraumatic imaging techniques to explore the circadian system in the entire animal is required. We are now developing a mouse model using a mutated dopamine receptor as reporter to follow-up the circadian expression of rev-erbα by PET. This should help us to define the real contribution of the rev-erbα gene in connecting circadian biology to physiological response and to better understand the regulation of this gene under different stimuli.
This work was supported by Génopole Rhône Alpes and Fondation Rhône Alpes Futur.
Circadian accumulation of rev-erba mRNAs in murine liver and fenofibrate response of rev-erba gene.
Strategy used to follow-up the circadian expression of the rev-erbα gene in mouse by PET.
Abstract ID: 885 Poster board space: 211
Joo Ho Tai, Binh Nguyen, Rebecca McGirr, R. Glenn Wells, Micheal Kovacs, Jane Sykes, Lela Dorrington, Edith Arany,
We are developing novel methods of engineering pancreatic islet cells for non-invasive imaging of gene expression. Single photon emission computed tomography (SPECT) is a highly sensitive, radionuclide-based imaging technique that can image multiple molecular events simultaneously. We have engineered insulin- and glucagon-secreting cells for the imaging of location and gene expression by SPECT both in vitro and after transplantation. INS-1 832/13 and aTC1-6 cells were stably transfected with an HSV1-thymidine kinase-GFP fusion construct. After clonal selection, cells were incubated with 5μCi/ml [131I]FIAU for 3 h. High-expressing aTC1-6 clones showed a 15-fold uptake (p< 0.001 compared with low-expressing clones) of [131I]FIAU, while INS-1 clones showed a 2-fold uptake. Signal intensity was quantified with gamma camera imaging, and photon counts correlated well with uptake of [131I]FIAU. For in vivo experiments, aTC1-6 cells were labelled with both [131I]FIAU and 111In-tropolone and transplanted under the left kidney capsule of CD1 mice. Mice were anesthetised and imaged for 3.75h using a single-head clinical gamma camera equipped with a 1mm pinhole collimator. Data for both 131I and 111In were acquired simultaneously. Anatomical imaging was conducted using computed tomography (CT). SPECT images were reconstructed with ordered subset expectation maximization (OSEM) and coregistered to CT using fiducial markers. After imaging, mice were sacrificed and kidneys were excised and counted in a gamma well counter. Cells expressing tkgfp could be visualized in vivo using dual-isotope SPECT. Kidneys bearing labelled cells showed an 491-fold uptake of 131I/g tissue and 7.7-fold uptake of 111In/g tissue compared with the right kidneys (p< 0.001 for both). Immunohistochemistry confirmed the presence of glucagon-positive cells at the site of transplantation. We conclude that pancreatic islet cells can be engineered for in vivo imaging of location and gene expression using dual-isotope SPECT. This work will enable us to develop an in vivo model of imaging islet function.
Abstract ID: 886 Poster board space: 212
Protein-protein interactions play an essential role in all biological processes, from replication and transcription, to cell division and cell proliferation. One standard technology to study protein-protein interactions is the two-hybrid technique. When fused in frame to respective domains of a split transcriptional activator, two proteins of interest can reconstitute a functional activator, resulting in expression of a reporter gene.
To enable non-invasive molecular imaging of protein-protein interactions within specific tissues of living animals, we created a transgenic reporter mouse that expresses firefly luciferase (Fluc) under the regulatory control of a concatenated Gal4 promoter (5x Gal4). Luciferase expression was transcriptionally dependent on interaction of two proteins wherein one was fused to the Gal4 binding domain (Gal4-BD) and the other to the VP16 activator domain. To demonstrate that the transgene was pandemically expressed, we used an adenovirus that constitutively expressed a fused binding-domain-activator (Gal4-BD-VP16). Intravenous, intramuscular, inhalational, and intracranial delivery of the adenovirus achieved signal induction in tissues of up to 4 logs over control. Expression of the transgene reporter was specifically shown in adult skin fibroblasts, liver, muscle, respiratory epithelia, and brain.
To demonstrate protein-protein interactions in vivo, we used the well characterized interaction between the tumor suppressor p53 (fused to Gal4-BD) and large T antigen (TAg) from SV40 (fused to the activator VP16). As a negative, non-interacting partner of p53 we used the polyoma virus coat protein (CP) fused to VP16 as a control. Using hydrodynamic somatic gene transfer into Gal4→Fluc reporter mice, we transfected hepatocytes in vivo with various protein pairs (p53/TAg; p53/CP) as well as each protein by itself. Renilla luciferase (Rluc) was used to control for efficiency of plasmid delivery. The interacting pair demonstrated 48-fold greater induction of Fluc expression in the liver compared to the non-interacting pair or proteins by themselves.
Abstract ID: 887 Poster board space: 213
The development of successful therapies for restoration or preservation of viable pancreatic beta-cell mass (BCM) in treating diabetes is hampered by the lack of non-invasive methodologies for quantitative imaging of BCM. The aim of this study is to assess the effectiveness of fluorescently tagged Exendin-3 and −4 (Sigma) as non-invasive molecular imaging probes targeted to the glucagon-like peptide-1 (GLP-1) receptor of beta-cells. Exendin-3 and −4 were conjugated to the monofunctional dye Cy5.5 (Amersham) yielding dye to protein ≈ 0.4. Saturation binding was assessed using the INS-1 rat insulinoma cell line. Cells were grown to confluence in 96-well plates and incubated with varying concentrations of exendin (3 or 4)-Cy5.5 (± GLP-1 to establish specificity) for 1-hr in Krebs-Ringer Buffer (2.5 mM glucose) and washed with PBS (2x). Fluorescence was measured at 700 nm (excitation: 625 nm) using the Kodak Image Station In-Vivo FX multi-model imaging system. The amount of binding was quantified from the fluorescence intensity in a defined region of interest using the Kodak Molecular Imaging® program. Minimal differences of binding exendin-3 and −4 to INS-1 cells were observed. We then demonstrated significant and specific binding of exendin-4-Cy5.5 to isolated rat pancreatic islets as described above for INS-1 cells. Finally, the efficacy of targeted imaging of BCM in vivo was evaluated in hairless SKH1 mice within a customized heated chamber. Baseline and sequential 1-min images were collected in the anesthetized (isoflurane) mouse following a bolus injection of approximately 40 nmoles of exendin-4-Cy5.5 via an indwelling jugular catheter. In vivo fluorescence was localized primarily to the region of the pancreas as determined by overlay of the baseline X-ray image and was confirmed by a relative fluorescence intensity of pancreas to liver of ~5 to 1. In conclusion, these results demonstrate the potential of using fluorescently-labeled exendin to determine BCM in vivo.
Abstract ID: 888 Poster board space: 214
Accurate non-invasive methods for monitoring changes in skeletal metabolism are important for understanding diseases of bone remodeling. In post-menopausal osteoporosis, altered estrogen status induces increased osteoclastic bone resorption unmatched by subsequent increases in formation, resulting in net bone loss. Tools to monitor these changes are needed to provide insight into osteoporosis as well as a range of other pathologies where bone metabolism may be altered.
Female Balb/c mice were subjected to ovariectomy (OVX), stimulating estrogen-induced bone loss, while control mice either received daily pamidronate (OVX-Pam) to block the OVX effects, or sham surgery alone. After 7 days, mice were administered a cathepsin K-cleavable near infrared imaging probe to monitor changes in osteoclastic bone resorption (n=5/group). Cathepsin K (CatK) is the primary cysteine protease produced by osteoclasts and is sufficient for degradation of collagen in bone. 2 mice per group received an additional dose of Osteosense (VisEn Medical), a fluorochrome-labeled bisphosphonate which binds to newly forming bone, in a separate near-infrared optical wavelength. 3 additional OVX mice received a non-cleavable d-form of the CatK probe to demonstrate substrate specificity.
Day 8 imaging by fluorescence molecular tomography demonstrates enhanced cleavage of the CatK probe in the proximal tibia of OVX mice compared to sham (p< 0.05). Daily pamidronate prevented this (p< 0.001), while mice receiving the control form of the probe showed 6-fold reduction in fluorescence (p< 0.001). Osteosense fluorescence correlated well with CatK activity (R2=0.88), demonstrating upregulation of both bone formation and resorption in the OVX model. Experiments on osteoclasts generated from mouse marrow cultures demonstrate intracellular CatK probe accumulation when cells are induced to resorb bone in vitro.
Optical imaging of bone formation and resorption allows for non-invasive assessment of altered osteoblast and osteoclast activity.
Abstract ID: 889 Poster board space: 215
Abstract ID: 890 Poster board space: 216
The use of catecholamines is a potential strategy to improve impaired insulin stimulated glucose uptake in diabetic patients. While this effect has primarily been studied in adipose tissue, similar effects may be present in skeletal muscle. In this study, we investigated the characteristics of catecholamine mediated facilitation of FDG uptake in cultured skeletal muscle cells.
Abstract ID: 891 Poster board space: 217
A transgenic mouse model of Clim transcriptional co-activators of LIM domain proteins is used to establish the sensitivity of intrinsic reflectance multi-photon optical microscopy (MPM) to cellular and extracellular matrix (ECM) changes in connective tissues due to gene alterations, as compared to those observed during normal tissue physiology. The dominant negative Clim is expressed under the K14 promoter in the epithelial tissues of mice and is involved in organ development and cancer.
We non-destructively mapped the 3D structural arrangement of papillary and reticular layers in mice dermis. In healthy animals, the collagen fibers in the reticular layer located underneath the upper papillary layer were three times thicker and were arranged parallel to the surface of the skin. The terminal phenotype of the transgenic mice revealed a highly remodeled dermis with no distinct papillary or reticular layers. A significant loss of fibrillar collagen was detected and quantified by reflected second harmonic intensity. Cellular autofluorescence was co-localized with the most profound alterations in the ECM. We are investigating the biochemical origins and biomechanical consequences of this remodeling.
Inside the hair follicles, in situ multi-photon optical tomography revealed intricate spatial interactions of cellular components with the surrounding connective tissue. In the hair follicle of a transgenic mouse with mild phenotype, the structural ECM alterations were observed. In the terminal phenotype, cellular fluorescence was absent from most hair follicles studied. We are correlating imaging with the molecular events involved in hair cycling.
Resolution of MPM imaging is comparable to that of histology; however, the absence of chemical tissue processing and staining reduces the number of artifacts observed in samples prepared and analyzed by histology or electron microscopy. This feature is extremely important in documenting structural pathologies of connective tissues with aberrant mechanical responses and aids in correlating them with gene and protein expression profiles.
Abstract ID: 892 Poster board space: 218
To investigate the metabolite profiles from livers of toxin-treated rats using 1H high resolution magic angle spinning (MAS)-NMR spectroscopy. Rats treated with a model of liver fibrosis were randomly chosen at various time-points in 12 weeks and (MAS)-NMR spectra of excised intact liver (15mg) were studied using principal components analysis (PCA) to extract novel toxicity biomarker information, compared with histology. The MAS 1H NMR spectra of controls showed signals from a range of organic acids and bases, a range of saturated and unsaturated triglycerides, phospholipids and cholesterol, amino acids, sugars and glycogen. Following complexes (CCL4, fat, alcohol) treatment, the 1H MAS-NMR spectra of livers indicated metabolic characterization of the lesion, which included steatosis, bile duct obstruction and altered glucose/glycogen metabolism; the 1H MAS-NMR-PCA analysis showed there was difference between control and treated groups, which was mainly determined by the signals at 1.24ppm-1.36ppm from elevated triglyceride levels, but there was no separation between different grades of fibrotic livers. The results suggest that 1H MAS-NMR spectroscopy requires minimal sample preparation and allows direct detection of tissue perturbations; however, it is not sensitive to demonstrate the difference in various degree of liver fibrosis induced by the complexes (CCL4, fat, alcohol).
Abstract ID: 893 Poster board space: 219
The study objectives were to assess proton magnetic resonance spectroscopy (1H-MRS) as means to distinguish among mice with dissimilar intra-abdominal body fat conformations, and to evaluate intra-abdominal fat burden for changes during weight loss and regain. Intra-abdominal fat burden was analyzed as a ratio of integrated fat to water 1H-MRS signals collected from a region of interest standardized across obese, trim and skim mice, the latter having little body fat. This was compared to intra-abdominal fat volume and surface area calculated from segmented magnetic resonance images. Similar measurements were made from obese mice before, during and after they were induced to lose weight. Significant differences in intra-abdominal fat burden were detected among mouse strains using 1H-MRS. Body weight, intra-abdominal fat volume and fat surface area measurements were different among strains as well. Spectrographic assessment of fat correlated well with other measures across 3 mouse strains. Correlations were not as strong within certain strains, however. During weight loss and regain, there was a significant overall pattern of changes in intra-abdominal fat quantity that occurred, which was reflected by 1H-MRS. Localized 1H-MRS was useful in distinguishing amounts of intraabdominal fat when comparing among obese, trim, or skim body fat conformations. This coupled with utility of 1H-MRS for tracking pattern of weight loss and regain supports the use of 1H-MRS for evaluating intraabdominal fat burden. Voxel characteristics, mouse numbers and genotype, and instrument variation are important design considerations for strengthening measures correlations within certain body conformations.
Abstract ID: 894 Poster board space: 220
The increasing use of non-invasive imaging modalities has become critical toward evaluation and validation of pre-clinical model systems and subsequent development of novel therapeutics. In particular, the application of optical imaging modalities employing the use of fluorescent and/or luminescent reporters in addition to NIR contrast reagents has been important in pre-clinical development. Fluorescent optical contrast is however subject to limitations due to background generated by endogenous auto-fluorescence or non-specific reagent uptake. Furthermore, the impact of tissue optical properties such as absorption and scattering on optical signal quality limits sensitivity in large parts of the visible spectrum. Thus, imaging of GFP reporters is generally restricted to sizable tumours subcutaneously with deeper lesions necessitating the use of skin-flaps.
Despite advances in the characterisation and treatment of acute myeloid leukaemia (AML) the majority of patients die from their disease. Accordingly, the development of more targeted therapeutics demands relevant xenograft models of this disease. However, imaging AML; diffuse disease states with non-palpable, deep tissue infiltrates, remains challenging. Here we present the generation of GFP labelled human leukaemic cell-lines in the development of imageable xenograft models of AML through the use of a raster scanned, time domain (TD) imager, eXplore Optix (ART/GE). Employing this TD imager, background auto-fluorescence lifetimes could be discriminated from GFP signals as early as one week following inoculation of GFP-labelled cells, at depths >1 mm. Disease development could be followed longitudinally, in real time, with excellent size and depth correlation of leukaemic infiltrates in and ex vivo. Additionally, functional analysis of apoptotic chemotherapeutic response in this model system using far-red (Cy5.5, Amersham Biosciences, GE Healthcare) labelled Annexin A5 (PharmaTarget) could be determined, by fluorescence lifetime, from spectrally identical unspecific probe. Thus, results will validate the use of optical imaging in both longitudinal monitoring of AML and in quantifying response to chemotherapeutic intervention.
Abstract ID: 895 Poster board space: 221
Current treatment of the human pediatric tumour retinoblastoma includes a combination of chemotherapy and laser coagulation of associated blood vessels. Antiangiogenic agents are being investigated as possible therapies with reduced toxicity and improved accessibility. High frequency ultrasound microimaging (UBM) is well suited for the study of angiogenic tumour vascular patterning and hemodynamic function noninvasively. We have longitudinally assessed the SV40 LHbTAg mouse model of retinoblastoma, with administration of the antiangiogenic agent DC101 (Imclone Systems Inc.), an anti-mouse VEGF receptor 2 antibody, at 4.5 months of age. Ultrasound imaging was performed with a Vevo 770 (VisualSonics Inc.) mouse scanner every two weeks from 1 month to 5 months of age. Contrast enhanced MicroCT (MS8 scanner, GE Medical Systems) was used to image vessel structure at these timepoints. Intraocular tumours were visible at 2 months and by 3.5 months highly vascularized tumours had filled the vitreous cavity, with angiogenically induced vessels branching through the vitreous to the lens. UBM Speckle variance flow processing (SFP) tracks intensity changes of ultrasound image pixels over time to isolate areas of active blood flow within tumours. Retinoblastoma tumours exhibited an increase in perfused area from negligible at 1 month to 1.2±0.2% (2.5 months), 3.0±0.2% at 3.5 months, and 8.5±2.4% at 4.5 months. DC101-treated tumors (i.p. (40mg/kg) at 4.5 months of age) exhibited a significant decrease in perfusion, compared with saline-injected controls, after 24 and 48 hours post DC101 administration. Percent perfused tumour area significantly declined from 6.9±2.1% prior to DC101 injection (0h) to 1.2±0.2% at 24 hours (p< 0.0001) and 0.7±0.5% at 48 hours, suggesting DC101 mediates an early effect on tumour blood flow. UBM enables non-invasive evaluation of vascular parameters, which provide useful functional information when testing antiangiogenic therapies. DC101 is a promising agent for antiangiogenic treatment of the pediatric tumour retinoblastoma.
Abstract ID: 896 Poster board space: 222
While inflammatory arthritis (IA) is a common chronic joint disease, its pathogenesis has not been completely elucidated. Neutrophils are believed to play an important early role in the process of joint destruction and persistence of inflammation, but their recruitment has not been completely characterized. Here, the development of a model for imaging neutrophil response in IA, utilizing isolation and ex vivo labeling techniques along with multichannel intra-vital fluorescent microscopy, allows examination of key elements involved in the neutrophil response.
IA was induced in C57BL/6 mice using the established KBN model. Following development of arthritic joints in these mice (10 days post serum injection), neutrophils isolated from bone marrow of wild type donor mice were fluorescently labeled and injected into the circulation of IA and control mice. Neutrophil response to the inflamed ankles and ventral paws was imaged using laser-scanning microscopy and a fluorescent blood-pool agent that delineated vasculature. IA mice showed significantly greater neutrophil response than control mice, both in terms of the stationary (5.3-fold increase) and rolling (5.0-fold increase) neutrophils, which dynamically interacted with vascular endothelium. Furthermore, images taken 8 hours post-KBN-serum injection revealed substantial numbers of rolling neutrophils relative to controls (6-fold increase). To examine the role of the C5a receptor, neutrophils isolated from C5aR-/- mice were imaged simultaneously with WT neutrophils fluorescently labeled at a second wavelength. Deficient neutrophils showed significantly decreased response compared with WT neutrophils in terms of rolling and stationery cells in IA mice (5-fold and 11-fold decrease, respectively) in established disease, suggesting the chemotactic effect of this receptor in neutrophil recruitment.
This study, which successfully imaged differential neutrophil response to arthritic joints, showed early cellular-level inflammatory response, before clinical signs of IA evolved. Continued work with this model will elucidate pathogenic mechanisms of IA through imaging of neutrophil response in genetically modified mice.
Abstract ID: 897 Poster board space: 223
Considerable evidence suggests a role for platelets in the metastatic process; specifically, activated platelets are believed to promote tumor cell survival and homing. The platelet defective β3 integrin null mice are protected from bone metastasis and a specific inhibitor of αIIbβ3 decreases bone and visceral metastasis. Because αIIbβ3 inhibitors are currently clinically available for short-term use in potent intravenous forms, we evaluated upstream inhibitors of αIIbβ3 activation in pre-clinical metastasis models. Aspirin and APT102 (a novel apyrase) block the platelet activators thromboxane and ADP, respectively. Murine melanoma B16-F10 and murine breast cancer 4T1 cell lines stably expressing firefly luciferase were used to study metastasis by real time in vivo bioluminescence imaging. Mice were pre-treated with vehicle or the combination of aspirin and APT102 and intracardially injected with B16luc or 4T1luc cells. Serial bioluminescence imaging demonstrated significant reduction in luciferase activity in the bones and lungs in inhibitor treated animals, which was confirmed by histology and X-ray imaging. However, the rate of tumor growth in each treatment group was similar. Thus, to evaluate tumor cell localization, organs were harvested one hour after 4T1-luc intracardiac inoculation and ex vivo quantification of luciferase activity was performed. No significant differences in initial end-organ localization to lung, liver or bone was observed in aspirin and APT 102 treated mice or in the β3-/- mice despite a marked difference in bone and lung tumor burden 14 days later. Interestingly, administration of aspirin/APT102 one hour after tumor cell inoculation failed to decrease metastatic burden. Based on our bioluminescence imaging, we conclude that blocking the activation of platelets pharmacologically leads to decreased metastasis independent of initial tumor cell dissemination if administered early in the metastatic process.
Abstract ID: 898 Poster board space: 224
It is widely held that the binding of VEGF to the VEGFR2 receptor is required to ensure effective vascularization of tumors growing in vivo. The expression of VEGFR2 is predominantly restricted to endothelial cells and is known to be up-regulated in blood vessels during angiogenesis in the context of wound healing, inflammation and tumorigenic processes.
In an attempt to non-invasively monitor dynamic changes in VEGFR2 expression during tumor angiogenesis with in vivo bioluminescence imaging (BLI), we have developed a genetically modified mouse that has a single targeted copy of firefly luciferase inserted into the VEGFR2 locus. The resulting mice appear to develop and grow normally. As all endogenous transcriptional regulatory elements have been preserved intact, however, the mice express firefly luciferase in a manner identical to that of native VEGFR2.
Prior to using this reporter mouse to visualize tumor angiogenesis in vivo, we first crossed the targeted luciferase allele onto a nu/nu genetic background. A range of different non-bioluminescent xenograft tumor cell lines were then implanted subcutaneously and in vivo bioluminescence was measured and quantified using an IVIS® Imaging System 100 Series and Living Image® software (Xenogen Corporation, Alameda). Our data analysis to date suggests that different tumor xenograft lines induce VEGFR2 expression very differently over the course of tumor development. We show that certain tumor lines exhibit a robust induction of bioluminescence as they develop in vivo whereas other lines do not, even though the tumors develop to a comparable size.
Taken together our data suggest that there may be a differential reliance on the VEGF/VEGFR2 pathway for successful vascularization between various tumor types. Further, this reporter 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: 899 Poster board space: 225
Abstract ID: 900 Poster board space: 226
Rodent models of respiratory disease have been used to examine the changes in lung 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 measured lung volume and CT density in free-breathing mice using retrospectively respiratory-gated, dynamic micro-computed tomography.
Two wild-type and four TIMP-3 deficient C57/B16 mice were anaesthetized with a mixture of ketamine and xylazine. The free-breathing animals were positioned prone 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. Throughout the scan, the respiratory signal was monitored and recorded, along with a signal indicating when the x-rays were on. Micro-CT images were acquired throughout the respiratory cycle over 10 gantry rotations at 80 kVp, 50 mA, with a scan time of 50 seconds and 0.28 Gy entrance dose to the animal. Volumetric images were reconstructed with an isotropic voxel spacing of 0.15 mm, using only the views from the same phase of the respiratory cycle (peak inspiration and end expiration).
For each image, an intensity-based seeded region-growing algorithm was used to extract the entire lung and calculate the lung volume. The mean lung density was measured at each respiratory phase. At end expiration, the TIMP-3 deficient mice exhibited increased air content in the lungs (−391±6 HU for wild-type and −479±4 HU for TIMP-3 deficient, p< 0.001).
Retrospectively 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. Since this dynamic micro-CT technique is non-invasive and the images are acquired in less than one minute, it promises to become a valuable tool for studying lung disease in mouse models.
Abstract ID: 901 Poster board space: 227
Wenbing Yun,
Using a unique in-line phase-contrast imaging scheme, an x-ray volumetric micro-tomography system with high resolution and high image contrast has been demonstrated. This imaging scheme greatly enhances the image contrast for soft tissues and overcomes the chief limitations of the existing commercial x-ray imaging tools based on absorption contrast. The proposed system introduces several key new capabilities of critical importance to biomedical research with small animal models: (1) direct observations of soft-tissue structures at high (20 μ) resolution with very low radiation dosage and without the need of contrast agents, thus enabling extensive longitudinal studies of the same animal, particularly its soft-tissue organs; (2) real-time 2D imaging allowing direct observation of dynamic biological processes of soft tissue organs and rapid 3D imaging that eliminates the need of extended sedations; (3) easy integration with other modalities, such as PET and SPECT to allow precise co-localization of functional or compositional features directly with the soft tissue structures.
Abstract ID: 902 Poster board space: 228
Guangbin Luo1, X. Lu1, Hong Guo1, Joseph Molter2, Stan Majewski3, Andrew Weisenberger3, Bingcheng Wang1,
Recent progresses in transgenic and gene targeting technologies in mice have demonstrated the great power of using reporter-gene-based imaging techniques for studying biological processes. We have created a mouse model for studying naturally occurring hepatocellular carcinoma (HCC) using a knock-in gene targeting approach. In this mouse model, a dual reporter expression construct was introduced into the endogenous Afp locus so that the expression of these two reporter genes is under the control of an endogenous Afp promoter. The mouse Afp promoter is active during embryonic development in several tissues including the fetal liver, but it becomes inactive in normal adult mice. However, it is re-activated in the cells of HCC. Therefore, this set of mouse models provides a potentially very useful tool for studying liver cancer in vivo by monitoring the expression of the two Afp-driven reporters through in vivo imaging. Our dual reporter construct encodes firefly luciferase (luc) and herpes viral thymidine kinase (tk). Therefore, both bioluminescent imaging (BLI) and PET imaging may be used to monitor the status of a specific HCC, depending on the specific circumstances.
Our studies have shown that the expression profiles of tk and luc in these mice were identical to that of endogenous Afp gene. Furthermore, the expression of these exogenous proteins did not result in any significant adverse effects to these mice. Most importantly, when HCC were induced in these mice, the tumors could be detected by both BLI and microPET with virtually no background. Therefore, we demonstrated the use of dual reporter genes driven by a well-characterized tumor-specific promoter in conjunction with in vivo imaging as a paradigm in studying naturally occurring HCC in live mice. This model may also be used for early identification of cancer-bearing animals, for monitoring changes of the disease in response to therapeutic interventions.
Abstract ID: 903 Poster board space: 229
Micro computed tomography (micro-CT) for in vivo lung imaging has been somewhat limited by the long acquisition time and inherent motion blurring artifacts. We present here a novel breath hold technique, adaptable to current commercial micro-CT systems, that significantly increases the effective resolution by reducing motion artifacts. This technique, coined Intermittent Iso-pressure Breath Hold (IIBH), utilizes a special breathing pattern consisting of 3 phases. During the “breath-hold” phase the scanner is triggered and multiple angles of view are captured. Figure 1 represents an in vivo no gating axial slice; Figure 2 is from the same mouse utilizing the IIBH techniqe.
We utilize this technique in a longitudinal study to track the progression of emphysema in a genetic strain of mice (tight-skin), which are known to exhibit emphysematous-like lung structure over time. We analyze changes in total lung volume, regional lung density and total lung compliance all of which are related to the progression of emphysema. A significant increase in lung volume and compliance was found between the tight-skin and normal mice. Reduction of regional changes in lung density was also visualized in 3D.
Pathology at the final time point confirms the presence of emphysema in the tight-skin mice.
Through our new IIBH technique we showed progressive changes in lung volume, compliance and regional lung density for tight-skin vs normal mice in vivo over a period of five months. Using this technique we now have the ability to track the early stages of this destructive disease and its response to therapies.
Abstract ID: 904 Poster board space: 230
Bacteria cause severe infections of the bone that are difficult to treat, yet few animal models of such diseases are available, and these models depend on using invasive procedures. Here, we describe the use of in vivo bioluminescence imaging (BLI) as a method of studying bone and joint infection following infection by the bacterial pathogen Listeria monocytogenes. Osteomyelitis has been reported to be a rare complication of infection by this organism in both humans and livestock, the true incidence of which is unknown. Intracellular growth is required for virulence of L. monocytogenes; if this ability is compromised the bacterium is five orders of magnitude less virulent. L. monocytogenes displays intracellular replication in a wide variety of cell types, and is thought to be distributed throughout the body in infected macrophages. We have previously used BLI to reveal tissue sites of extracellular replication, and found that the gallbladder lumen may serve as a reservoir of bacteria in the body. Here we use BLI to follow L. monocytogenes in bone, and demonstrate that mutants of this pathogen that are defective in intracellular growth persist for many weeks in the leg bones of mice, and that the bacteria can be cultured from the bone marrow. This result is surprising given the expected disease course of this organism, and indicates that L. monocytogenes is capable of colonizing certain tissues in the absence of intracellular growth. Following oral and intravenous administration, virulent L. monocytogenes can also lead to infection of the bone marrow, and signals from a specific location within the femur are often observed. The sensitivity and versatility of BLI offer significant opportunities for the study of bacterial infection and this method continues to reveal broad spectra of infections from both well-studied and emerging pathogens.
Abstract ID: 905 Poster board space: 231
Metastasis is the primary cause of death in breast cancer patients and it is therefore necessary to identify and study factors that control metastatic potential. Our recent studies have shown that increased levels of early placenta insulin like growth factor (EPIL) in HER2 positive breast cancer cells lead to an invasive phenotype in an in vitro extravasation assay. Furthermore clinical data indicate that EPIL is highly abundant at the invasive tumor border compared to the rest of the tumor. In addition, EPIL is an independent prognosticator of poor outcome and indicates a poor prognosis subgroup of HER2 positive tumors. To evaluate the role of differential EPIL expression in tumor cell invasion and metastasis in vivo, we designed a mouse xenograft model. This model is based on the breast cancer cell lines SKBR3 (high endogenous level of HER2) and MDA-MB-468-HER2 (artificially overexpressing HER2) which were demonstrated to show increased invasivity upon EPIL stimulation in vitro.
To track the cells by whole body in vivo imaging and also identify them microscopically in immunohistochemical analyses we labeled these cell lines with a dual reporter encoding Firefly-Luciferase (fLuc) and enhanced green fluorescent protein (eGFP) via retroviral transduction. Control of EPIL expression was conferred on the cells by transfecting them with an EPIL expression plasmid and a vector encoding an EPIL specific shRNA.
Tumor cells were introduced into the animals at different anatomic sites including inoculation into the mammary fat pads, tail vein and left ventricle. Whole body imaging using bioluminescence was used to follow primary tumor growth and track metastatic foci. Immunohistochemical analysis is used to characterize EPIL expression in the primary tumor and metastases. The in vivo data will be used to support or refute the hypothesis that EPIL plays a role in metastatic disease progression.
Abstract ID: 906 Poster board space: 232
Abstract ID: 907 Poster board space: 233
A model for the evolution of autoimmunity in type 1 diabetes is the NOD mouse. In this model, the autoimmune response is progressive and age-dependent, beginning at around 3-wks of age and leading to overt diabetes by 12-14-wks of age. A significant contributor to the progression from inflammation to overt diabetes in NOD mice is the “avidity maturation” of a prevailing, pancreatic beta-cell-specific CD8+ T lymphocyte population. This population recognizes a Kd-restricted peptide ligand (NRPV7) with high avidity[1]. Previously, we have developed a probe (MN-NRPV7) specific for this T cell population[2]. MN-NRPV7 consists of superparamagnetic iron oxide (for MRI) coated with NRPV7/ Kd peptide/MHC ligands.
Here, we utilized MN-NRPV7 to define the evolution of the NRPV7-specific autoimmune response by non-invasive MRI. Eight-, fifteen- and twenty-four-wk old NOD mice were imaged by MRI before and after i.v. injection of MN-NRPV7. By comparing our findings with controls injected with non-modified nanoparticles, we established that MN-NRPV7 accumulation was antigen-specific. Furthermore, semi-quantitative analysis of pancreas-associated R2 relaxivity suggested a trend toward age-dependence. Remarkably this trend mirrored the relative recruitment of NRPV7-reactive CD8+ T cells to the pancreata of experimental mice (Fig). Histological studies suggest that MN-NRPV7 accumulation reflects the time-course of autoreactive T cell priming in the pancreatic lymph nodes and the subsequent recruitment of these cells to pancreatic islets.
Our imaging strategy allows us to non-invasively characterize the evolution of the antigen-specific autoimmune response in the progression of type 1 diabetes. This approach would lead to a better understanding of the mechanisms behind the initiation and progression of the autoimmune process and ultimately in designing immunomodulatory therapeutic schemes.
Abstract ID: 908 Poster board space: 235
Auto-antibodies against glucose-6-phosphate isomerase (GPI) induce arthritis in mice, that closely resembles rheumatoid arthritis. In this model injection of GPI-serum induced severe arthritis within 6 days. As mast cells are important amplifiers of T cell-dependent and T cell-independent inflammation, and can produce mediators involved in angiogenesis, we studied whether mast cells influence neoangiogenesis, a hallmark of rheumatoid arthritis. Comparing wild type mice, mast cell-deficient kitW/W-v mice and mast cell-reconstituted kitW/W-v mice, we found that mast cells were essential for pannus formation and neoangiogenesis during GPI-arthritis. This was confirmed by in vivo quantification of activated αvβ3 integrin using [18F]Galacto-RGD and positron emission tomography (PET). One hour prior to imaging we injected animals 5550 kBq with [18F]Galacto-RGD. When compared to mice receiving control serum, [18F]Galacto-RGD uptake increased markedly in joints of GPI-serum-injected wild-type mice, while [18F]Galacto-RGD uptake remained at background levels in mast cell-deficient KitW/KitW-v mice. As these data suggested that preformed mediators released by mast cells may cause inflammation and tissue destruction we induced GPI arthritis in mice treated with mast cell stabilizers such as cromolyn or salbutamol. Joint swelling was reduced to 16% in salbutamol-treated and to 30% in cromolyn-treated mice. In vivo quantification of activated αvβ3 integrin by [18F]Galacto-RGD and PET showed significantly reduced [18F]Galacto-RGD uptake in joints of salbutamol or cromolyn treated mice compared to placebo treated mice; [18F]Galacto-RGD uptake was suppressed to the levels of mice injected with control serum. Thus, in vivo examination of angiogenesis by [18F]Galacto-RGD in rheumatoid arthritis might be a new tool to imaging the successful therapeutic suppression of angiogenesis. Inhibition of mast cell-degranulation may not only attenuate symptoms of allergic diseases but also the damages leading to rheumatoid arthritis including neovascularization and scar formation.
Abstract ID: 909 Poster board space: 237
Ganapathy Krishnamurthi1, Vincent Gattone2, Keith Stantz1, Minsong Cao1, Gary Hutchins2,
Abstract ID: 910 Poster board space: 239
Radiolabeled MIBG (meta-iodobenzylguanidine)I123 is used clinically for localizing and monitoring neuroblastoma (NB) progression and response to treatment and also for therapy. MIBG imaging in NB mouse models could provide a valuable tool for preclinical therapeutic studies. Using a recently acquired Gamma Medica Dual headed small animal SPECT (single photon emission computer tomography)/CT scanner, we have developed methods for imaging MIBG-avid human NB xenografts in mice. We selected a subset of cell lines from a panel of 37 NB cell lines which had been characterized for MIBG uptake in vitro for in vivo imaging. Low, medium and high uptake cell lines were implanted subcutaneously in nude mide. CHLA-140, a line with moderate MIBG uptake and moderate growth rate, was imaged repeatedly. We have obtained consistent imaging results in CHLA-140 tumors over a six month period. CHLA-140 is a multi-drug resistant NB used for pre-clinical drug testing in xenografts. The purpose of this study was to optimize protocols for MIBG-based imaging in mouse models and determine the practical spatial and detection limits.
Abstract ID: 912 Poster board space: 241
Mouse models are becoming increasingly important for pulmonary disease research. Current techniques use 2D histology to explain structural changes seen in disease conditions. However, the correlation of 2D histology to 3D anatomy, which is currently lacking, would enhance our understanding. Using a combination of micro computed tomography (micro-CT), a novel 3D pathology imaging system and standard histology, we investigate the airway geometry of three common mouse strains used in pulmonary research.
In vivo micro-CT scans were performed on three mice per strain. The lungs of each were then excised, fixed. Each lung was embedded in agarose, then using the custom Large Image Microscope Array (LIMA) 250μm sections of tissue were sequentially removed for histological processing, while between each section the remaining tissue block was imaged with a high magnification microscope. Using this combination, histology slides were accurately correlated back to the LIMA dataset. This is important since the process of histology on tissue can introduce distortion artifacts and its correlation back to imaging modalities such as the micro-CT is challenging.
Following image acquisition, the airways were segmented from both the micro-CT and LIMA datasets using an in house automated segmentation program. A 3D iso-surface was generated for each airway tree and used to compile distinctive geometry information including airway diameter, branching length and angle, as well as airway wall thickness. A comparison between the mice strains was made in order to characterize each strains normal airway anatomy.
Significant differences in airway geometries were found between the strains and should be considered when studying the effects of pulmonary diseases on mice.
Abstract ID: 913 Poster board space: 243
Molecular imaging of atherosclerosis is one of the current challenges in modern medicine. Its goal is to predict individual cardiovascular risk based on the identification of vulnerable atherosclerotic plaques. Imaging modalities such as MRI and CT are focussing on the morphological characteristics of plaque progression and vulnerability.
We have examined lesions of atherosclerosis in ApoE-deficient mice using a micro-CT (SkyScan 1072, resolution< 1.8 μm) (Fig. 1) and have compared CT-data on plaque morphology and volume with quantitative histomorphology at identical anatomically defined sites throughout the aorta and brachiocephalic artery in the same animals. We have observed an excellent correlation between plaque/lumen volume as measured by CT (calculated by the Analyze 6.0 software) and the morphometrically measured volumes as determined by the Cavalieri's mathematical method of volume estimation by serial cross-sections (≈ 250–300 histological sections at 5 μ from the entire brachiocephalic artery from the aorta to the bifurcation). In addition, histopathological plaque characteristics such as lipid-rich areas, fibrous cap and calcifications were readily identified by the Micro-CT.
Our study allows a reliable quantitative evaluation of atherosclerotic plaque volume, burden and morphology in the mouse, enabling a rapid evaluation of experimental models of atherosclerosis as well as monitoring of pharmacological therapy. Our next objective is to combine CT-generated morphological data with PET-derived molecular data on glucose metabolism, metalloproteinase activity and apoptosis (caspase imaging and annexin V) in the same animals. Such multi-modality imaging of atherosclerosis will provide valuable information on plaque progression and vulnerability.
From Concurrent 5: Session 13: Imaging in Inflammation, Infectious and Immune Diseases
Abstract ID: 076
Gene therapy promises new treatments for a wide range of common conditions in individualized molecular medicine, but is plagued by inefficient transport to diseased sites. Recent developments in the modification of the virus could allow more targeted approaches and herald the advent of systemic delivery of therapeutic viruses. Encouraging clinical trial results bode well for the future of a targeted viral or non-viral approach, yet the safety of viral vectors underscores the importance to monitor their immediate and long-term delivery to diseased sites and their exclusion from normal tissue sites.
Images of intravenously injected virus. Pictures show the times that after injection in second (s), minute (m) or hour (h).
Images of orally savaged virus. Pictures show the virus in the Gl system and then In the feces.
Herein, we report using IRDye 800 to dynamically image virus trafficking in vivo using 780 nm excitation light and emission collected at 830 nm. Adenoviral particles were labeled with IRDye800 particle number of 1×109 or 1 nanogram (or 1 picomolar) dye equivalent was administered i.v. (Figure 1). Particle numbers of 1×1010, or 10 nanogram (or 10 picomolar) dye equivalent was delivered to mice through oral gavage (Figure 2). Near, real-time virus imaging data wereas acquired at 800 milliseconds per frame and showed selective accumulation of stained virions in the mouse, dependent upon administration route in the early stage. Re-distribution of labeled virons to the other organs was followed over a period time up to 24 hours after administration.
Our imaging data demonstrate that oral and intravenous administrated viruses can be dynamically imaged in mouse model with low background when near-infrared fluorophores are used.