Novel α2 β1 Integrin-Targeted Peptide Probes for Prostate Cancer Imaging

Abstract

Accumulating experimental evidence indicates that overexpression of α₂β₁ integrin may correlate with progression in human prostate cancer. The objective of this study was to design a novel imaging probe based on the Asp-Gly-Glu-Ala (DGEA) peptide for near-infrared-fluorescent (NIRF) imaging of α₂β₁, integrin expression in prostate cancer. The peptides were conjugated with appropriate fluorescent dyes, and the binding affinity of these probes was evaluated by flow cytometry in three human prostate cell lines (PC-3, CWR-22, and LNCaP). In vivo NIRF imaging of the α₂β₁-positive PC-3 xenograft model was performed to evaluate the α₂β₁ targeted probe. In vitro immunofluorescence staining was carried out to confirm the α₂β₁ integrin expression level. Flow cytometry analysis showed that PC-3 had the highest probe uptake, followed by CWR-22 and LNCaP tumor cells. In the subcutaneous PC-3 model, the tumor demonstrated prominent uptake with good tumor to background contrast. Immunohistochemistry staining also supported the in vivo optical imaging results. DGEA-based optical agents have been developed for specific imaging of α₂β₁, integrin expression. In vitro and in vivo localization demonstrated the potential of this agent to identify tumor subtypes amenable to anti-α₂β₁ integrin treatment and potentially provide prognostic information regarding tumor progression.

PROSTATE CANCER is the most common malignancy diagnosed in men and is the second most common cause of mortality for men in the United States and Europe. ¹ Despite the fact that prostate-specific antigen screening has greatly increased the number of patients detected with early-stage prostate cancer, about 40% of prostate cancers are first detected at an advanced stage, and half of these are found to be extracapsular at pathologic staging. ^{2 , 3} Therefore, development of an accurate noninvasive imaging technique to detect primary, recurrent, and residual prostate cancer is critical for the effective management of this group of patients.

The precise causes of prostate cancer remain poorly understood. Numerous growth factors and their receptors are overexpressed during the progression of this disease. Specific changes in protein expression in epithelial and stromal tumor cells during the different developmental stages of prostate cancer notably contribute to enhancing tumor cell growth, survival, migration, and invasiveness.

From a clinical perspective, metastatic bone disease is the major cause of mortality in prostate cancer patients. ⁴ Therefore, the high incidence of skeletal metastasis has been suggested to be a reflection of favorable reciprocal interactions between the bone microenvironment and disseminated prostate cancer cells.^5–7 This process involves tumor cell adhesion to other cells and extracellular matrix glycoproteins and eventual invasion through basement membranes. Such interactions could be mediated by a variety of cell surface biomolecules, including integrins. Integrins are essential for cell attachment and control cell migration, cell-cycle progression, and programmed cell death, which they regulate in synergy with other signal transduction pathways. The observation that integrins present on various tumor types are differentially expressed during tumor transformation, progression, and metastasis suggests that integrins may also be useful as prognostic markers. ⁸ In particular, the α₂β₁ integrin, a receptor mainly for type I collagens, laminins, E-cadherin, matrix metalloproteinase 1, and several viruses, ^{9 , 10} has been implicated in multiple aspects of tumor progression and metastasis. In many tumors, there are correlations between high expression of α₂β₁ and tumor progression.^11–16 For example, in human prostate cancer, the PC-3 cell line (androgen independent) has the highest expression of α₂β₁ integrin and is more aggressive in the mouse model compared to other cell lines, such as CWR-22 and LNCaP, which have lower α₂β₁ integrin expression (androgen dependent).^17–20 Moreover, recent studies indicate that α₂β₁ integrin is one of the specific markers that could be used to characterize prostatic adult stem cells, which could be associated with the initiation of prostate cancer. ²¹ Therefore, high expression of α₂β₁ integrin may also be used to demonstrate the signature of the prostate cancer stem cell. In summary, the α₂β₁ receptor is an important target for drug design and delivery. The development of α₂β₁ integrin cell expression profiles or “fingerprints” of individual tumors may have further potential in identifying a cell surface signature for a specific tumor type and/or stage, potentially leading to customized treatment options for patients.

It has been shown that α₂β₁ integrin binds to a site within the α₁(I)-CB3 fragment of type I collagen. ²² The minimal active recognition sequence for this integrin is a tetrapeptide of the sequence Asp-Gly-Glu-Ala (DGEA) corresponding to residues 435 to 438 of the type I collagen sequence. Although DGEA has been used for various purposes by different investigators, ^{23 , 24} there are no reports on the use of DGEA for in vivo imaging of α₂β₁ integrin expression to our knowledge. In the present study, we explored the possibility for specific in vivo imaging of α₂β₁ integrin expression in xenotransplanted prostate cancer models in mice with an optical system. We also performed in vitro and ex vivo experiments to confirm our in vivo imaging results.

Materials and Methods

General

All commercially available chemical reagents were used without further purification. 5(6)-Carboxyfluorescein (FAM), 9-fluorenylmethoxycarbonyl (Fmoc) amino acids, and Wang resin preloaded with the Fmoc-Ala amino acid were purchased from Novabiochem (San Diego, CA). Aspartic acid and glutamic acid were all protected as the tert-butyl ester. Cy5.5 monofunctional N-hydroxysuccinimide (NHS) ester (Cy5.5-NHS) was purchased from Amersham Biosciences (Piscataway, NJ). The purification of the crude product was carried out on a analytical reverse-phase high-performance liquid chromatography (HPLC) system equipped with a dual ulraviolet absorbance detector (Waters 2487, Waters, Milford, MA) using a Phenomenex (Torrance, CA) synergi 4μ Hydro-RP 80Å (150 × 4.6 mm, 4 microns). The flow was 1 mL/min, with the mobile phase starting from 98% solvent A (0.1% trifluoroacetic acid (TFA) in water) and 2% solvent B (0.1% TFA in acetonitrile) (0–2 minutes) to 50% solvent A and 50% solvent B at 40 minutes.

Synthesis of Peptides

All linear DGEA peptides were synthesized by a standard Fmoc solid-phase peptide synthesis method with fourfold excess amounts of benzotriazole-1-yl-oxy-tris-pyrrolidino-phosphonium-hexafluorophosphate (PyBOP), 1-hydroxybenzotriazole (HOBt) and an eightfold molar excess of diisopropylethylamine (DIPEA; Sigma, St. Louis, MO). Acylation was carried out for 60 minutes, and complete reaction was confirmed by the trinitrobenzene sulfonic acid test. ²⁵ Removal of the Fmoc protective group on the α-amino group was achieved with 20% piperidine in dimethylformamide (DMF) (v/v) (Sigma). DMF was used to wash the resin between each acylation and deprotection step.

FAM Conjugation

The N-terminal Fmoc groups were removed after coupling the last amino acid. FAM was then coupled onto the exposed amino group at threefold excess in the presence of equimolar amounts of PyBOP and a sixfold molar excess of DIPEA for 2 hours in the dark. Following the acylation, unbound FAM was removed by washing the resin with DMF, and then the conjugated peptide FAM-DGEA was cleaved from the resin by treating with cleavage solution (95% TFA and 5% water) for 3 hours. The desired conjugated peptides were purified and characterized by analytical HPLC.

Conjugation and Purification of Cy5.5-DGEA Conjugates

The synthesis of Cy5.5-DGEA conjugates was achieved through conjugation of Cy5.5-NHS ester with the N-terminal amino group of the aspartic acid residue of the DGEA peptides. The Cy5.5-NHS (1 mg) dissolved in DMF (77 µUL) was added to the fully protected DGEA peptide, which was still on the resin, followed by DIPEA (3.3 µUL). The reaction mixture was stirred overnight in the dark at room temperature. All conjugated peptides and side chain protecting groups were simultaneously removed by treating with cleavage solution (95% TFA and 5% water) for 3 hours. Peptide-containing supernatants were separated from the solid support by filtration and concentrated under a stream of nitrogen. Crude peptide was precipitated and washed twice with ice-cold diethyl-ether and dissolved in 10% acetic acid in water before lyophilization. The desired products were purified and characterized by analytical HPLC. The purity of Cy5.5-labeled peptides was over 95% from analytical HPLC analysis. The retention times on analytical HPLC for unlabeled and labeled DGEA peptides were 7 and 30 minutes, respectively. Fractions containing Cy5.5-DGEA conjugates were collected, lyophilized, and stored in the dark at −20°C until use. The purified Cy5.5-DGEA conjugates were characterized by LTQ Orbitrap (Thermo Scientific, West Palm Beach, FL) hybrid mass spectrometry.

Cell Lines

The human prostate cancer cell line PC-3 was obtained from American Type Culture Collection (Manassas, VA) and was maintained at 37°C in a humidified atmosphere containing 5% CO₂ in F-12K medium and 10% fetal bovine serum (Life Technologies, Inc., Grand Island, NY). CWR-22 and LNCaP cell lines were also from American Type Culture Collection and were grown in RPMI-1640 with 10% fetal bovine serum in 5% CO₂ at 37°C.

Flow Cytometry Analysis of the Cell Binding

Human prostate PC-3, CWR-22, and LNCaP cancer cells were used to assess the cell binding and internalization efficiencies of FAM-DGEA-targeted peptide probe. To minimize the nonspecific uptake of peptides by pinocytosis, incubations were performed on ice followed by flow cytometry analysis to allow rapid quantization of fluorescence. For the quantification of fluorescence by flow cytometry (FACScan, Becton, Dickinson, Franklin Lakes, NJ), 5,000 cells were counted and viable cells with similar size and granularity in the forward and sideways scatterplots were analyzed. The fluorescence profiles and the overall mean fluorescence intensities of the cells within this region were obtained and analyzed using CellQuest software (Becton, Dickinson).

Fluorescence Microscopy and Cell Uptake Studies of FAM-DGEA

For fluorescence microscopy studies, PC-3, CWR-22, and LNCaP prostate cells (1 × 10⁵) were cultured on BD Falcon four-chamber vessel culture slides (BD Biosciences, Bedford, MA). After 24 hours, the cells were washed twice with phosphate-buffered saline (PBS) and then incubated at 25°C in the presence of 1 μM FAM-DGEA for 30 minutes. After the incubation period, cells were washed three times with ice-cold PBS. For the blocking study, unconjugated DGEA peptide 20 μM was added to the culture medium before the addition of FAM-DGEA conjugates. Nuclear counterstain was performed with 4′,6-diamidino-2-phenlindole (DAPI) in PC-3 cells. The fluorescence signal from the cells was recorded using an fluorescent microscope (Axioskop 40, Carl Zeiss Micro-Imaging, Thornwood, NY) equipped with a fluorescein isothiocyanate filter set (exciter, HQ 475/20 nm; emitter, HQ 540/30 nm). An AttoArc HBO 100 W (Carl Zeiss/ AttoArc, Thornwood, NY) microscopic illuminator was used as a light source for fluorescence excitation. Images were taken using a thermoelectrically cooled charge-coupled device camera (Micromax, model RTE/CCD-576, Princeton Instruments, Trenton, NJ).

Tumor Xenografts

Animal procedures were performed according to a protocol approved by the University of Southern California Institutional Animal Care and Use Committee. Male athymic nude mice (BALB/c nu/nu), obtained from Harlan (Indianapolis, IN) at 4 to 6 weeks of age, were given injections subcutaneously in the right shoulder with 1 × 10⁶ of PC-3 human prostate cancer cells suspended in 100 µUL of PBS. When the tumors reached 0.4 to 0.6 cm in diameter (14–21 days after implantation), the tumor-bearing mice were subject to in vivo imaging studies.

In Vivo Near-Infrared Optical Imaging of Tumors

In vivo fluorescence imaging was performed with an IVIS 200 small-animal imaging system (Xenogen, Alameda, CA). A Cy5.5 filter set was used for acquiring the Cy5.5-conjugated DGEA peptide probes' fluorescence in vivo. Identical illumination settings (lamp voltage, filters, f/stop, field of views, binning) were used for acquiring all images, and fluorescence emission was normalized to photons per second per centimeter squared per steradian (p/s/cm²/sr). Images were acquired and analyzed using Living Image 2.5 software (Xenogen). For the control experiment, peptide probes were injected into three mice via tail veins with 1.5 nmol Cy5.5-DGEA and subjected to optical imaging at various time points postinjection. For the blocking experiment, the mice (n = 3) for each probe were also injected with a mixture of 10 mg/kg of unlabeled DGEA peptide and 1.5 nmol Cy5.5-conjugated DGEA peptide. All near-infrared fluorescence images were acquired using a 1-second exposure time (f/stop = 4). Mice of the experimental and blocking groups were euthanized at 28 hours postinjection. The tumor and major tissue and organs were dissected, and ex vivo fluorescence images were obtained.

Immunohistochemistry

For immunohistochemical examinations, the collected xenograft PC-3 tumor tissue samples were fixed in 4% freshly prepared buffered paraformaldehyde, embedded in paraffin according to routine histologic procedures, and sectioned at a thickness of 5 μm. Immunohistochemical analysis of paraffin-embedded PC-3 prostate carcinoma was done using a2/CD49b monoclonal antibody (R&D Systems, Minneapolis, MN) The antibody was applied by a two-step peroxidase method using the DakoEnVision +HP Mouse Kit (Dakopatts, Glosstrup, Denmark). Briefly, deparaffinized tissue sections were rinsed in 0.1 mol/L Tris-HCl buffer, pH 7.6, containing 0.15 mol/L NaCl (Tris-buffered saline [TBS]). Endogenous peroxidase was inactivated by incubating the sections with 1% (v/v) hydrogen peroxide in TBS for 20 minutes. The tissue sections were then rinsed thoroughly in TBS and incubated for a further 10 minutes with 1% bovine serum albumin (BSA) in TBS at room temperature before incubation with the primary antibody overnight at 4°C. The monoclonal antibody CD49b was diluted 1:200 in TBS containing 1% BSA. As negative controls, duplicate sections were incubated with 1% BSA instead of specific primary antibodies. The sections were washed three times for 5 minutes each time in TBS followed by 30 minutes' incubation with one drop of peroxidase-conjugated rabbit antimouse secondary antibody. After a washing step in TBS, peroxidase activity was visualized by incubation sections in TBS containing 0.06% (w/v) 3,3′-diaminobenzidine tetrahydrochloride (Sigma) and 0.034% (v/v) hydrogen peroxide for 8 minutes.

Data Processing and Statistics

All of the data are given as means ± SD of three independent measurements. Statistical analysis was performed with a Student t-test. Statistical significance was assigned for p values of .05. For determining tumor contrast, mean fluorescence intensities and the tumor area at the right shoulder of the animal and of the normal tissue at the surrounding tissue were calculated by the region-of-interest function of Living Image software. Mean fluorescence signals of each time point were plotted.

Results

Synthesis and Characterization of FAM-DGEA and Cy5.5-DGEA

The schematic molecule structures of FAM-DGEA and Cy5.5-DGEA conjugates are shown in Figure 1. The fluorophores are conjugated to the N-terminal amino group of the aspartic acid residue and purified by analytic HPLC with ultraviolet absorption at 500 nm (to purify FAM-DGEA conjugate) and 690 nm (to purify Cy5.5-DGEA conjugate), respectively. The retention time on analytical HPLC was 25 and 30 minutes, respectively. The mass analysis was performed by LTQ Orbitrap Hybrid Mass Spectrometer: m/z =1,287.3 for Cy5.5-DGEA [M + H]⁺ (calculated MW = 1,286.36 for C₅₅H₆₁N₆O₂₂S₄); m/z = 750.2 for FAM-DGEA [M + H]⁺ (calculated MW = 748.65 for C₃₅H₃₂N₄O₁₅).

Figure 1.

Schematic structures of FAM- and Cy5.5-conjugated DGEA peptide probes.

In Vitro FAM-DGEA Peptide Binding Specificity Studies

It has been demonstrated that the most aggressive PC-3 cells displayed the highest surface expression of α₂β₁ integrin compared to two other less aggressive prostate cancer cell lines (CWR-22 and LNCap). Figure 2 shows typical fluorescence histograms of FAM dye-conjugated DGEA peptide probes toward three prostate cancer cell lines. A comparison of the histograms resulting from this binding experiment suggests that there is little or no nonspecific binding of the LNCaP cell line (with low α₂β₁ integrin expression level) at the concentrations used. Given that the histograms are of a relatively normal distribution, the mean fluorescence intensity correlates well with the expression level of the α₂β₁ integrin in these cell lines. Although we cannot obtain an exact measure of the number of probes binding to the cell from these data, fluorescent peak shifts in these histogram profiles under a given set of conditions will reflect the binding affinity differences in the corresponding mean fluorescence intensity. Correlations between cell lines and individual experiments are dependent on maintaining the same average surface area for the cells. Fortunately, we have found through the observation of channel settings on the Coulter counter, forward and sideways scatter on the flow cytometer, and visual observation that all of the cell lines used here have very similar volumes and hence surface areas.

From the flow cytometry results, we can validate that the expression level detected with FAM-conjugated peptide probe was consistent with previous reports determined by the antibody labeling.^18–20 It was found that PC-3 cells incubated with the targeted DGEA peptide displayed the highest mean fluorescence intensity followed by CWR-22 and LNCaP (see Figure 2).

Figure 2.

Histograms of typical flow cytometry results. Each cell line was treated with a FAM-DGEA probe as described in Materials and Methods. Fluorescence histograms were produced and the results plotted. The binding percentages of each cell line are 99.7% (PC-3), 51.4% (CWR-22), and 15.6% (LNCaP), respectively.

In addition to the flow cytometry data, intracellular localization of the fluorescent constructs was also examined using fluorescence microscopy. In Figure 3, the labeled peptide bound distinctly to PC-3 cells, whereas the CWR-22 revealed substantially lower levels of cellular fluorescence signals and LNCaP cells showed almost no specific binding of the probes. In addition, binding of the FAM-DGEA could be blocked with the unlabeled DGEA peptide, which demonstrated that the binding is specific. Like many integrin-targeted probes, such as cyclic RGD peptide, internalization of the ligand after binding does occur, but to a limited extent. ^{26 , 27} Radiolabeled RGD peptides also demonstrate limited cell uptake (less than 1%). ^{28 , 29} However, it is also possible that the fluorescent dye motif in the conjugates may facilitate ligand internalization after binding owing to increased lipophilicity of the conjugate after attaching the fluorescent tags. ³⁰ Allosteric cooperation of the targeted motif and fluorochrome may further facilitate the internalization of the imaging probe, as we observed in this study. However, the mechanism of internalization of these ligand conjugates after receptor binding remains unclear and requires further investigation.

Figure 3.

Binding of FAM-DGEA to prostate tumor cell lines with different levels of α₂β₁ integrin expression. PC-3 cells (A), CWR-22 cells (E), and LNCaP cells (F) were incubated with 1 μM FAM-DGEA to assess the target selectivity of the probe. In accordance with the FACS analysis data, PC-3 cells overexpressing α₂β₁ revealed strong cellular fluorescence. (A, FAM-DGEA; B, DAPI; C, merge), which could be selectively blocked by coincubation with 20 μM unlabeled DGEA peptide (D).

In Vivo Fluorescence Imaging with Cy5.5-DGEA

Figure 4 shows typical NIRF images of athymic nude mice bearing a subcutaneous human prostate PC-3 tumor after intravenous injection of 1.5 nmol of Cy5.5-DGEA. The whole animal became fluorescent immediately after injection, and the subcutaneous PC-3 tumor could be clearly delineated from the surrounding background tissue from 30 minutes to 24 hours postinjection, with maximum contrast occurring around 2 hours postinjection. Significantly, the amount of fluorescence was still detectable in the tumor at 48 hours postinjection (data not shown).

Figure 4.

Top, In vivo near-infrared imaging of subcutaneous PC-3 prostate tumor-bearing nude mice after intravenous injection of 1.5 nmol of Cy5.5-DGEA. The position of the tumor is indicated by arrows. The fluorescence signal from probes is pseudocolored red. The tumor can be clearly visualized from 30 minutes to 24 hours postinjection. Bottom, The fluorescence intensity was recorded as per second per centimeter squared per steradian (P/s/cm²/sr). Tumor fluorescence was higher than that in the normal tissue (muscle) through 24 hours.

The fluorescence intensities defined as photons per second per centimeter squared per steradian (p/s/cm²/sr) in the tumor and the normal tissues as a function of time are depicted in Figure 4. The tumor uptake reached a maximum at 2 hours postinjection and slowly washed out over time. On the other hand, normal tissue had rapid uptake but overall lower uptake compared to tumor throughout the time period studied.

To validate the targeted specificity of the DGEA peptide probe, we performed a blocking experiment. The control mice were each given injections of 1.5 nmol of Cy5.5-DGEA, and those in the blocking experiment were each given coinjections of 1.5 nmol of Cy5.5-DGEA and 10 mg/kg unlabeled DGEA peptide (300 nmol). The typical NIRF images of PC-3 tumor-bearing mice of both groups are shown in Figure 5. The pseudocolored fluorescence images were acquired 4 hours after intravenous injection. At this time point, the contrast of tumor to normal tissue was maximal as the nonspecific binding had washed out. Unlabeled DGEA peptide successfully reduced tumor uptake compared to the unblocked imaging result. Furthermore, ex vivo evaluation of excised organs at 28 hours postinjection (see Figure 5) showed that the compound was predominantly taken up by the PC-3 tumor, which correlated well with our in vivo imaging results. We also noticed that the tumor fluorescence intensity and contrast on this ex vivo experiment were significantly higher than those obtained from in vivo imaging, which could be attributed to the difference in tissue penetrations. The immunohistochemistry staining results are shown in Figure 6. The xenograft PC-3 tumor tissue did have a strong positive staining in accordance with the in vivo and ex vivo imaging results.

Figure 5.

Left, Representative blocking experiment imaging (acquired 4 hours postinjection) of mice bearing subcutaneous PC-3 tumor on the right shoulder demonstrating blocking of Cy5.5-DGEA (1.5 nmol) uptake by coinjection with unlabeled DGEA (10 mg/kg). Right top row, Representative images of dissected organs of mice bearing PC-3 prostate tumor, A (experiment) and B (block), sacrificed 28 hours after intravenous injection of Cy-5.5-DGEA. (1 = muscle; 2 = heart; 3 = tumor; 4 = liver; 5 = lung; 6 = spleen; 7 = brain; 8 = pancreas; 9 = kidney). Right bottom row, Region of interest analysis of fluorescence intensity ex vivo of major organs with (block) and without (experiment) coinjection of a blocking dose of DGEA was plotted. Strong fluorescence signal could be detected in tumor and kidney tissue.

Figure 6.

Expression of α₂β₁ integrin in PC-3 prostate tumor xenografts. Tissue sections were preincubated without (negative control, left) or with the integrin antibody CD49b directed against α₂β₁ receptor (positive staining, right). Note the strong α₂β₁ integrin expression in the PC-3 tumor tissue (objective magnification X 20).

Discussion

The current treatments for prostate cancer, consisting of malignant prostate ablation by radical prostatectomy, radiotherapy, hormonal therapy, and/or neoadjuvant chemotherapy, ^{31 , 32} are generally curative for the majority of patients diagnosed with localized and androgen-dependent prostate cancer forms. However, progression to androgen-independent and metastatic disease states is often accompanied by a recurrence of prostate cancer.^33–35 The available chemotherapeutic treatment options for patients with hormone-refractory prostate cancer are rather palliative and remain mostly ineffective, with a poor prognosis. The prognosis is associated with a median survival rate of about 12 months after diagnosis. ¹ Therefore, development of an accurate noninvasive imaging technique to detect primary, recurrent, and residual prostate cancer is critical for the effective management of this group of patients. Given that the progression from the androgen-dependent prostate cancer into more aggressive and metastatic forms often leads to disease relapse, several novel therapeutic strategies have been investigated for improving treatments against metastatic prostate cancer.

The recent identification of distinct deregulated cellular targets in prostate cancer cells directly involved in prostatic carcinogenesis, such as the integrin family, may allow us to target several signaling elements in tumor cells to counteract prostate cancer progression. Integrins have been shown to mediate cellular adhesion as well as entry and withdrawal from the cell cycle. So far, 18 α and 8 β glycoprotein subunits of the integrin family have been discovered. ³⁶ They have been found to play a key role in a variety of processes, including cell migration, angiogenesis, cell invasion, and tumor growth and metastasis, all of which require integration of the cell into the extracellular environment. These subunits are expressed on cell surfaces in 24 different heterodimeric combinations, allowing them to respond to and modulate a broad array of extracellular matrix proteins.

The progression of prostate cancer primarily involves the formation of secondary metastatic lesions to bone, a process partially mediated by integrin cell adhesion proteins. Although molecular events responsible for the differences remain largely unclear, a number of studies have described changes in integrin expression in relation to the metastatic progression of prostate cancer. It has been demonstrated that the more aggressive PC-3 cells display increased surface expression of α₂β₁ integrin compared to two other less aggressive prostate cancer cell lines (CWR-22 and LNCap). Interactions of α₂β₁ integrin with type I collagen have also been implicated in the formation of bone metastasis. All of these results suggest that the overexpression of α₂β₁ integrin may facilitate migration and metastatic spread of prostate cancer cells. The possible correlation between α₂β₁ integrin expression level and tumor invasiveness makes α₂β₁ integrin a potential biomarker for cancer progression diagnosis and treatment monitoring.

In this report, we focused on the development of imaging agents that can probe the α₂β₁ integrin cell expression profiles of individual tumors. Furthermore, identification of a cell surface signature for a specific tumor type and/or stage may potentially lead to customized treatment options for patients.

After successful demonstration of the specificity of our novel α₂β₁ targeted peptide probe in three different human prostate cancer cell lines, including PC-3 and CWR-22 (both are α₂β₁ integrin positive) and LNCaP (α₂β₁ negative) in vitro, we further tested the PC-3 tumor model in vivo. Coupling of a NIRF dye to the DGEA peptide allowed optical imaging of the PC-3 human prostate cancer cell line model. The in vivo imaging demonstrated prominent uptake of the Cy5.5-DGEA with the receptor specificity confirmed in a blocking experiment. The expression of α₂β₁ integrin in PC-3 tumors was also confirmed with immunohistochemistry antibody staining. In the ex vivo biodistribution semiquantitative data (see Figure 5), we found that the fluorescence intensity of the dissected tumor or tissue was significantly higher than that measured by region of interest analysis of the noninvasive images, which may be explained by the more effective fluorescence detection of excised organs and tissues without attenuation of the excitation and emission light in and out of the skin, as well as the scattering caused by the skin. However, the residual tumor contrast in the blocking experiment was not completely blocked by unconjugated DGEA peptide as only a 40% reduction was observed. This may be due to the nonspecific binding, accumulation in extracellular space, or autofluorescence of the tissue itself. In addition, it has been reported that these cyanine dyes have tumor-targeting capability even without any specific targeting moiety conjugated, perhaps explaining why some fluorescence signals can still be observed in the blocking experiment. We also observed similar results for the optical imaging with dye-RGD conjugates.

It should be noted that some organs showed the signal reduction in the blocking experiment, such as liver, lung, spleen, and muscle. This phenomenon is most likely due to the normal expression of α₂β₁ integrin in these organs. In normal liver tissue, α₂β₁ integrin is expressed in vascular endothelia, bile duct epithelium, connective tissue stroma, and sinusoidal lining cells. ³⁷ α₂β₁ Integrin is diffusely expressed around surface airway epithelial cells. ³⁸ Decreased expression of α₂β₁ integrin has been reported in lung adenocarcinoma. ³⁹

Despite the success of this proof-of-principle optical imaging study in small-animal models, there are some barriers to overcome before the eventual clinical translation of this DGEA-based imaging agent for noninvasive imaging of tumor α₂β₁ integrin expression and in development of α₂β₁ integrin-targeted drugs. The major drawback of optical imaging is the poor tissue penetration and intense light scattering, which allows only for demarcation of superficial tumors and tissues accessible by endoscopy, as well as intraoperative imaging. Another concern is the use of different fluorochromes for in vitro and in vivo imaging studies. The chemical property of each dye may have not only a different impact on the binding affinity of the probe but also varying steric hindrance. Compared to the small peptide structure, the size of the fluorochrome is relatively large, which may also influence the binding affinity of the probe. To obtain quantitative tumor targeting and distribution patterns of the DGEA-based probes, radionuclide imaging modalities such as positron emission tomography and single-photon emission computed tomography will be required for further quantitative studies. Moreover, it is important to have high tumor to kidney ratios, as well as high absolute tumor uptake and longer retention, for both imaging and therapeutic applications. Based on our optical imaging results, this DGEA-based probe clears rapidly from the body. Thus, further modification is needed to improve the pharmacokinetics and binding affinity of these α₂β₁ integrin-targeted probes. Our future work will also focus on the structure-activity relationship to develop various high binding affinity ligands for α₂β₁ integrin, including the construction of DGEA multimers to enhance binding affinity through the multivalency effect.

Conclusions

It has been shown that the upregulation or overexpression of α₂β₁ integrin may correlate with tumor progression in human prostate cancer. Noninvasive imaging of α₂β₁ integrin expression may therefore play a key role in detecting prostate cancer progression, perhaps leading to treatment modification. The studies described above demonstrate that it is feasible to detect and semiquantify tumor α₂β₁ integrin status by noninvasive NIRF imaging with the DGEA-based optical agents. Despite the limited penetration of light through tissue, this proof-of-principle approach provides opportunities for rapid and cost-effective preclinical evaluation in animal models before the more costly radionuclide-based imaging techniques are applied.