Near-Infrared Optical Imaging of Ovarian Cancer Xenografts ... - Ucdavis

RESEARCH ARTICLE

Molecular Imaging . Vol. 4, No. 4, October 2005, pp. 439 – 447

439

Near-Infrared Optical Imaging of Ovarian Cancer Xenografts with Novel A3-Integrin Binding Peptide ‘‘OA02’’ Olulanu H. Aina, Jan Marik, Regina Gandour-Edwards, and Kit S. Lam University of California, Davis

Abstract Through screening of random one-bead one-compound (OBOC) libraries, we previously identified cyclic peptides with the cDGXGXXc motif that bind to A3 integrin subunit on ovarian adenocarcinoma cell lines ES-2, SKOV-3, and CaOV-3. We subsequently synthesized two secondary libraries based on this motif and identified new peptides that bound with a higher affinity to these cell lines. One of the peptides identified from the 20% ‘‘down-substituted’’ focused library was the cdGHCit-GPQc (‘‘OA02’’) peptide. The goal of this study was to determine whether this peptide labeled with near-infrared probes could be detected after intravenous injection in ovarian tumor-bearing mice and if it would selectively localize in the tumor. Three different forms of this peptide were synthesized, ‘‘OA02’’ – biotin (noncovalently linked to streptavidin – Cy5.5); ‘‘OA02’’ – Cy5.5 and ‘‘OA02’’ – AlexaFluo 680. Using a KODAK IS2000MM image station, these peptide probes were used at the near-infrared (NIR) spectra to image nude mice bearing ES-2 (A3 integrin positive) and Raji (A3 integrin negative) xenografts. The peptide probe displayed highly specific tumor uptake within 15 min, which lasted for 70 min for ‘‘OA02’’ – Cy5.5 and ‘‘OA02’’ – AlexaFluo 680 and for 24 hours for ‘‘OA02’’ – biotin – streptavidin – Cy5.5. Some kidney and bladder signal were noted. Prior injection with anti-A3 monoclonal antibody blocked the binding of this peptide to the ES-2 tumors. Mol Imaging (2005) 4, 439 – 447. Keywords: Peptides, ovarian cancer, combinatorial chemistry, optical imaging, integrin.

Introduction We had previously identified cDGXGXXc cyclic peptide motifs that bound preferentially to CaOV-3, SKOV-3 (ovarian adenocarcinoma cell line), and ES-2 (ovarian clear cell adenocarcinoma) through screening random OBOC libraries [1]. Using this original motif, we synthesized a secondary library cX1GX3GX5X6c with 41 amino acids in the X1 position and 42 amino acids in the X3, X5, X6 positions. These amino acids had both L- and D-isomers as well as unnaturally occurring amino acids. One of the peptides identified from this secondary library screening ‘‘cdG-HCit-GPQc’’ (‘‘OA02’’) was found to have a high affinity as well as specificity for a3 integrin [2]. This integrin is known to form heterodimers with b1 integrin [3,4]. Ligands to a3b1 include types I and VI collagen;

laminin-I, -5, -10, and -11; and thrombospondin-1 [5]. It is an integrin that is involved not only in disease but also in normal development. It plays an important role in normal lung, kidney, and epithelial development, as well as neuronal migration during cerebral cortical development [6– 9]. a3 integrin-deficient mice die around birth [10]. The a3b1 integrin is known to play a crucial role in tumor cell adhesion, migration, invasion, and metastasis. Its role in cell signaling and its association with regulation and degradation of extracellular matrix proteins are the subject of recent scientific studies [11 – 14]. Overexpression of integrin receptors on certain tumors makes targeted delivery an attractive option [11,15,16]. Both radiolabeled and optically labeled receptor-binding probes offer a noninvasive method of tumor diagnosis, studying tumor progression as well as tumor response to therapy. Each modality has its own unique application, however, optical imaging with fluorescent-labeled ligands can be advantageous in that it can reduce the risks of exposure of patient and personnel to radioactivity. It may reduce costs and can be a helpful way of generating preliminary data or optimizing a ligand prior to embarking on a more elegant study using radioactive probes. Other advantages of optical probes are that they are easy to synthesize and can be presynthesized and stored, meaning they do not have to be synthesized immediately prior to use because there is no ‘‘decay.’’ The time required for data acquisition in optical imaging is in seconds as the probe can be excited while its emission is collected at the same time, reducing the time required for the patient or mouse to be under anesthesia. For in vivo optical imaging, NIR optical reporters are superior to other fluorochromes because of their

Abbreviations: AU, arbitrary units; ELISA, enzyme linked immunosorbent assay; MALDI-TOF, matrix assisted laser desorption ionization-time of flight; NIR, near infrared; OBOC, ‘‘One-bead one-compound’’; RP-HPLC, reverse phase-high performance liquid chromatography. Corresponding author: Kit S. Lam, Department of Internal Medicine, Division of Hematology and Oncology, University of California Davis, 4501 ‘‘X’’ Street, Sacramento, CA 95817; e-mail: [email protected]. Received 16 May 2005; Received in revised form 15 June 2005; Accepted 20 June 2005. D 2005 Neoplasia Press, Inc.

440

Ovarian Tumor Optical Imaging with Novel Peptide Probe Aina et al.

enhanced tissue penetration and minimal light scatter. Although surface tissues or organs can be observed with lower wavelength dyes, dyes in the NIR range allow for deeper tissue observation in the 2- to 3-cm range. NIR optical imaging of epidermal growth factor receptor (EGFR) in breast cancer xenografts [17], overexpressed folate receptor in ovarian xenografts [18,19], endostatin in Lewis lung carcinoma [20], somatostatin and bombesin receptor expressing pancreatic tumor xenografts [21,22], as well as MMP-2 [23], have given credence to the fact that cell surface receptors can be exploited in targeted delivery of imaging agents to solid tumors. Using Cy5.5 and IRDye800 conjugated to c(KRGDf) peptide, Wang et al. [24] were able to image mice bearing human melanoma xenografts expressing avb3 integrin. This group also developed a dual-labeled RGD imaging probe composed of both radioisotope and NIR dye and demonstrated that detection of superficial tumors was superior using optical imaging [25]. NIR diffuse optical tomography, as well as contrastenhanced optical imaging, has already been used to detect breast tumors not only in humans but also in dogs [26 – 29]. Here we report for the first time the imaging of ovarian tumor xenografts using a novel a3integrin binding peptide (cdG-HCit-GPQc/‘‘OA02’’) that was identified from screening OBOC libraries [2].

Materials and Methods Cell Lines ES-2 (ovarian clear cell carcinoma), SKOV-3 (ovarian adenocarcinoma), and Raji (B-cell lymphoma) were purchased from American Type Culture Collection (ATCC, Rockville, MD). ES-2 and SKOV-3 cells were maintained in 90% McCoy’s 5a medium with 10% fetal calf serum, 1.5 mM L-glutamine, 50 units/mL penicillin, 50 mg/mL streptomycin. Raji cells were maintained in 90% RPMI 1640, 10% fetal calf serum, 50 units/mL penicillin, 50 mg/mL streptomycin. Cells were incubated at 37C in 5% CO2.

Nude Mice Four to 6-week-old female athymic nude (nu/nu) mice were purchased from Harlan Sprague Dawley (Indianapolis, IN). Mice were housed in the animal facility and fed ad lib with rodent pellets and water. Tumor xenografts were created by injecting ES-2 (1 106 cells in 200 mL incomplete McCoy’s media) subcutaneously on the left flank and Raji (1 107 cells in 200 mL incomplete RPMI media) subcutaneously on the right

flank while in the head-first prone position (sternal recumbency) under mild inhalation anesthesia (Halothane). Experiments commenced when tumors were about 10 mm in diameter. Typically, tumors arise within 2 – 3 weeks of inoculation. Animals were anesthetized intraperitoneally with Nembutal (1 mL/g body weight) of 6 mg/mL pentobarbitol sodium salt in diluent (50 mL 100% ethanol, 100 mL propylene glycol, 350 mL distilled water). Animals were imaged serially for tumor uptake and washout. Duration of scan was 30 sec for each acquired exposure. All procedures were conducted under an approved protocol according to guidelines specified by the National Institute of Health Guide for Animal Use and Care.

Peptide Synthesis and Labeling Protected amino acids were purchased from ChemImpex (Wood Dale, IL), D-biotin and all other chemicals were obtained from Aldrich (Milwaukee, WI). Hydrophilic linker Fmoc-Ebes-OH was synthesized in our laboratory according to the previously published procedure [30]. Solid-phase synthesis of cdG-HCit-GPQc-Ebes-EbesK(Alloc)-NH2 was carried out using the standard Fmoc chemistry on Rink MBHA amide resin (GL Biochem, China). The allyloxycarbonyl (Alloc) protecting group was removed from the C-terminal lysine side chain by reacting with 0.5 Eq tetrakistriphenylphospine palladium Pd[PPh3]4, 10 Eq phenylsilane PhSiH3 at RT twice for 30 min, and the resin was thoroughly washed with dichloromethane (DCM) 3 and N, N-dimethylformamide (DMF) 6. Cy5.5-NHS and Alexa 680-NHS 1 Eq was conjugated to the peptide in DMF at pH 9 adjusted with diisopropylethylamine (DIEA) overnight. The D-biotin (3 Eq) was coupled using 3 Eq of benzotriazol-1-yl-N,N,N0, N0tetramethyluronium hexafluorophosphate (HBTU) and 6 Eq of DIEA at room temperature overnight. Side-chain protecting groups were removed with mixture containing trifluoroacetic acid (TFA)/triisopropylsilane (TIS)/1,2ethanedithiol (EDT)/water (94:1:2.5:2.5, vol/vol/vol/vol). To cyclize the peptide via the flanking D-cysteines, the crude peptide was dissolved in 50 mL of aqueous 0.05 M ammonium acetate and stirred in an open flask overnight. The solvent was removed by lyophilization and the product was purified using semipreparative HPLC (C-18, Phenomenex Jupiter 250 10 mm) with water– acetonitrile gradient mobile phase and 0.1% TFA in water. The products were analyzed using HPLC (System A: Phenomenex Proteo; System B: Vydac C18 Protein and Peptide, 0.05% aqueous TFA – acetonitrile gradient 0 – 90% in 40 min) and MALDI-TOF mass spectrometer Bruker BiflexIII (Billerica, MA). Molecular Imaging . Vol. 4, No. 4, October 2005


Fluorochromes Cy5.5 Mono NHS ester (Ex 674 nm, Em 689 nm) was purchased from Amersham Biosciences (Piscataway, NJ). AlexaFluo 680 carboxylic acid, NHS ester (Ex 679 nm, Em 702 nm) was purchased from Molecular Probes (Eugene, OR). Streptavidin-conjugated Cy5.5 (Ex 678 nm, Em 703 nm) was purchased from Rockland Immunochemicals (Gilbertsville, PA). Cell Staining; Enzyme Linked Immunosorbent Assay SKOV-3 cells were grown in a monolayer on glass chamber slides. When cells were about 70% confluent, chambers were taken off. Single layer of Raji cells were pelleted unto a glass slide using a cytospin. All slides were fixed with 90% methanol in water at 80C overnight and then washed several times in phosphate buffered saline (PBS). Slides were blocked with 1 mL each of avidin D and biotin (‘‘Avidin –Biotin’’ blocking kit; Vector Labs). 10 mM of ‘‘OA02’’ – biotin peptide in 1 mL media was added to the slides and incubated for 1 hr at room temperature in a humid chamber. Subsequently, slides were washed and incubated with 1:1000 dilution of streptavidin phycoerythrin (Molecular Probes) in PBS for 15 min and then washed, mounted, and observed under a fluorescent microscope with an excitation wavelength of 555 nm and an emission wavelength of 605 nm. Antibody Blocking Anti a3 monoclonal antibody (ASC-1) was purchased from Chemicon International (Temecula, CA). Twenty micrograms of antibody diluted in 100 mL sterile saline solution was injected into the mouse via the tail vein 30 min prior to injection with 20 mg of either ‘‘OA02’’ – Cy5.5, or ‘‘OA02’’ – Alexa 680 in 100 mL of sterile PBS. NIR Detection Either 20 mg of ‘‘OA02’’ – Cy5.5 dissolved in 100 mL sterile saline solution, 20 mg of ‘‘OA02’’ – Alexa 680 dissolved in 100 mL sterile saline or 100 mL of 1 mg/mL streptavidin – Cy5.5 – PBS solution mixed (vortex) with 11 mg (10 mL of 1.1 mg/mL) of ‘‘OA02’’ –biotin overnight at 4C (or 1 hr at room temperature) was injected via the tail vein in an anesthetized mouse before imaging. Animals were placed on transparency in the head-first prone or supine position and immobilized by taping their limbs to the transparency. Images were acquired with a Kodak IS2000MM Image station (Rochester, NY), excitation filter 625/20 band pass, emission filter 700WA/ 35 band pass, light source was 150 W quartz halogen lamp set to 100 (max). Images were captured with a Molecular Imaging . Vol. 4, No. 4, October 2005

441

CCD camera set to F stop = 0, FOV = 150, FP = 0. Data were collected and analyzed using the Kodak ID 3.6 software by drawing regions of interest (ROIs) on the imaged mouse. Image data files were stored on a desktop computer in .bip format and later converted to .jpeg, whereas ROI data were stored and analyzed in Excel.

Results Peptide-Conjugate Synthesis and Purification We used our previously validated OBOC approach to identify ligands that bound preferentially to human ovarian adenocarcinoma cell lines (Figure 1). For this study, we chose (from secondary library screening) cdG-HCit-GPQc (‘‘OA02’’) as our candidate peptide based on (i) previous NMR data that suggested that cX1GX3GX5X6c peptides are stabilized at the X5 position by proline and (ii) X1 and X3 positions contained unnaturally occurring amino acids (D-Aspartic acid, LHomocitrulline) which could make the peptide less susceptible to peptidases [1]. Schematic of peptide synthesis for the three forms of this peptide on rink amide resin is shown in Figure 2A. The chemical structure of ‘‘OA02’’ – Cy5.5 as well as the Ebes linker used is shown in Figure 2B. ‘‘OA02’’ –Biotin: HPLC (System A) retention time 10.50 min, purity 95% (220 nm), MS MALDI m/z 1661.7 (M+H)+. ‘‘OA02’’ –Cy5.5: HPLC (System B) retention time 15.20 min, purity 85% (220 nm), 90% (254 nm); MS MALDI m/z 2457.2 (M+H)+. ‘‘OA02’’ – Alexa 680: HPLC (System A) retention time 10.20 min, purity 95% (220 nm), 97% (254 nm), 99% (600 nm); MS MALDI 2286.7 (M+H)+. The HPLC profiles are shown in Figure 2C. Cell Staining with ‘‘OA02’’ –Biotin We used ‘‘OA02’’ – biotin in a sandwich enzyme-linked immunosorbent assay (ELISA) with streptavidin – phycoerythrin as a secondary reagent to stain methanol-fixed ovarian adenocarcinoma cells (Figure 3). Our choice of fluorochrome was based on the filters available on our laboratory florescence microscope. SKOV-3 cells were used based on the high level of expression of a3 integrin on their cell surface. Staining was confined to the cell membrane and cytoplasm. This is consistent with the presence of integrin receptors on the cell surface. Integrin receptors have also been known to be present in the cytoplasm as a result of (a) endocytosis of the cell membrane and (b) recycling of the membrane through the endoplasmic reticulum [12]. The negative control cell line (Raji, B-cell lymphoma) did not show any staining because they are not known to express any appreciable levels of a3 integrin which is only known to

442


Figure 1. The scheme of experimental approach taken to identify, optimize, and use the peptide ligands identified. There is a continuous cycle of library screening, ligand motif identification, optimization and liquid-phase synthesis, and use of identified ligands.

be expressed on adherent cells (CD antigens and antibodies; DAKO, Carpinteria). Near-Infrared Optical Imaging with Peptide Conjugates To determine the specificity of ‘‘OA02’’-conjugates for ES-2 tumors, six groups of animals were evaluated. Each group had mice with both positive (ES-2) and negative (Raji) control tumors [1]. The first group received no injection, the second received ‘‘OA02’’ –biotin – streptavidin –Cy5.5, the third group received ‘‘OA02’’ – Cy5.5, the fourth group received ‘‘OA02’’ – Alexa 680, the fifth group received an injection of anti a3 monoclonal antibody prior to receiving either ‘‘OA02’’ – Cy5.5 or ‘‘OA02’’ –Alexa 680, and the sixth group received streptavidin– Cy5.5 alone. In-vivo and ex-vivo optical images and their quantitation are shown in Figures 4 and 5. In the first group of mice (n = 3), some light emission was observed in the gastrointestinal tract as a result of the animal’s diet and possible autofluorescence of GI bacteria flora. No signal was noted in either the positive or negative tumors (Figure 4A: I) Baseline signal intensity was (225 ± 15) arbitrary units (AU) for both tumors. Animals in Group 2 (n = 3) showed tumor uptake of ‘‘OA02’’ –biotin – streptavidin –Cy5.5 only in the ES-2 tumor. The average (average maximum intensity ± standard deviation) at 30 min for this group was (2229 ± 50) AU; at 60 min (1736 ± 1036) AU; at 6 hr (1809 ± 80) AU; and at 24 hr (1540 ± 556) AU. Fluorescence signal

was also noted in the kidneys and bladder which is the expected route of clearance for the peptide. The Raji tumor did not appear to bind to the conjugate above background levels (765 ± 53) AU (Figure 4A: II). Group 3 mice (n = 4) also showed specific uptake of the ‘‘OA02’’ – Cy5.5 conjugate in the ES-2 tumors at 15 min (6032 ± 4640) AU, at 70 min (5526 ± 3693) AU. Values for the Raji tumors at 15 min were (2685 ± 1103) AU; at 70 min (2786 ± 1583) AU. Fluorescence signal was observed within 5 min of injection and the intensity peaked at about 70 min (Figure 4A: III). The peptide probe was then rapidly cleared from the system via the kidneys and fluorescence intensity returned to 5 min postinjection time point around 5 hr. Some background signal was observed around the neck. Group 4 mice (n = 2) also showed specific uptake of ‘‘OA02’’ – Alexa 680 in the ES-2 tumors at 15 min (5015 ± 1443) AU, at 50 min (2999 ± 291) AU; at 90 min (1333 ± 301) AU. Maximum intensity values in the Raji tumors at 15, 50, and 90 mins were (2130 ± 907) AU, (1424 ± 888) AU, and (709 ± 471) AU. There seemed to be some light scattering and a rapid decline in signal intensity. The fifth group of mice (n = 2) injected with a3 monoclonal antibody prior to injection with the probe did not show any uptake of either the ‘‘OA02’’ – Cy5.5 or ‘‘OA02’’ – Alexa 680 probe into the ES-2 tumors (696 ± 24) AU or Raji tumors (757 ± 73) AU. Although there was no tumor uptake, the fluorescence signal intensity in the kidneys assured Molecular Imaging . Vol. 4, No. 4, October 2005


443

Figure 2. (A) Synthesis scheme of ‘‘OA02’’ – biotin, ‘‘OA02’’ – Cy5.5, and ‘‘OA02’’ – Alexa 680 on rink resin. (B) Structure of ‘‘OA02’’ – Cy5.5 and Ebes (linker). Structure of Alexa 680 is unavailable. (C) HPLC profile of ‘‘OA02’’ – biotin, ‘‘OA02’’ – Cy5.5, and ‘‘OA02’’ – Alexa 680 showing the purity of the compounds. Molecular Imaging . Vol. 4, No. 4, October 2005

444


Figure 3. Cell staining of monolayer of Raji (left panel) and SKOV-3 cells (right panel) on glass slides using a two-step incubation with ‘‘OA02’’ – biotin and streptavidin phycoerythrin. Fluorescence image was taken with an Olympus microscope ex 555, em 605 at 40 magnification.

us that the probe was in circulation and is been cleared (Figure 4A: IV). In one of the mice, 4 days after the antibody blocking study was done the mouse was injected again with 20 mg of ‘‘OA02’’ –Alexa 680 and we noted specific uptake of the dye in the ES-2 tumor (Figure 5D). Thus, by using this animal as its own control we were convinced that the a3 antibody was indeed inhibiting uptake of the peptide by the ES-2 tumor.

A representative 2-hr postinjection ex vivo image of a mouse injected with ‘‘OA02’’ – Cy5.5 showed that most of the probe signal intensity came from the ES-2 tumor, bladder, and kidneys (Figure 4B). Animals in Group 6 (injected with Streptavidin Cy5.5 alone) did not show any visible signal in the ES-2 tumors. At 15 min, maximum intensity values were (773 ± 465) AU, at 60 min (905 ± 295) AU, and at 2 hr (900 ± 140) AU. Maximum intensity values for the Raji tumors at 15 min were

Figure 4. (A) I: Representative negative control mouse not injected with any fluorescent probe, auto fluorescence is present in the GI tract and anus but absent in the ES-2 and the Raji tumor, II: Representative mouse injected with ‘‘OA02’’ – biotin – streptavidin – Cy5.5, 6 hr postinjection. Note specific uptake of peptide probe in the ES2 but not in the Raji tumor. Mouse kidney also shows fluorescence signal; III: Representative mouse injected with ‘‘OA02’’ – Cy5.5. Fluorescence signal observed in the tail (injection site), ES-2 tumor and kidneys 70 min postinjection, some nonspecific signal around the neck was observed; IV: Representative mouse injected with anti 3 integrin antibody 30 min prior to injection with ‘‘OA02’’ – Cy5.5. After 70 min no ES-2 tumor uptake of the probe was observed, however, fluorescence signal was seen in the kidneys and same nonspecific signal around the neck. Arrow represents ES-2 (positive control) tumor, star represents Raji (negative control) tumor. (B) Representative ex vivo image of mouse injected with ‘‘OA02’’ – Cy5.5, 2 hr postinjection. Fluorescence signal was observed in the ES-2 tumor, kidneys, bladder, ovaries and uterus. Minimal or no uptake was detected in lungs, liver, heart, spleen and the Raji tumor. Molecular Imaging . Vol. 4, No. 4, October 2005


445

Figure 5. (A) Comparison of maximum signal intensity values over time observed from injecting various ‘‘OA02’’ fluorescent probes into tumor-bearing mice. Highest ES-2 tumor intensity was observed with the ‘‘OA02’’ – Cy5.5 (cross), ‘‘OA02’’ – Alexa 680 showed steady decrease in signal from time of injection (square), ‘‘OA02’’ – biotin – streptavidin – Cy5.5 showed an increased uptake of the probe around 6 hr and a slow accumulation of the probe in the tumor over several hours (diamond). Negative control strepavidin – Cy5.5 did not show any appreciable accumulation in the ES-2 tumor within the first hour or thereafter (triangle). Mice that received no injection show background fluorescence around 200 AU (circle). (B) Comparison of ES-2 tumor uptake of ‘‘OA02’’ – biotin – streptavidin – Cy5.5, ‘‘OA02’’ – Alexa 680, and ‘‘OA02’’ – Cy5.5 using mean minus background intensity values. (C) Fluorescence time-course intensity of uptake of ‘‘OA02’’ – Cy5.5; a comparison between ES-2 (3 expressing) and Raji (non 3 expressing) tumors. (D) Comparison of fluorescence light intensity observed over time in an ES-2 and Raji tumor-bearing mouse that was first injected with 3 integrin antibody prior to injection with ‘‘OA02’’ – Alexa 680. The same mouse was used in the control experiment 4 days later with injection of ‘‘OA02’’ – Alexa 680 alone.

(538 ± 75) AU; at 60 min (1046 ± 113) AU; and at 2 hr 1123 ± 188) AU.

Discussion We had previously reported the identification of cDGXGXXc peptides that bind to ovarian adenocarcinoma cell lines CaOV-3, ES-2, SKOV-3, and OVCAR-3. Through screening of ‘‘focused’’ libraries, we identified cdG-HCit-GPQc (‘‘OA02’’) peptide which was found to bind to a3 integrin receptors on these cells. The purpose of this study was to determine whether ‘‘OA02’’labeled fluorescent probes can be used to image a3 integrin expression in ovarian tumor xenografts. We showed that we can synthesize various forms of Molecular Imaging . Vol. 4, No. 4, October 2005

‘‘OA02’’ fluorescent probe with high purity. Flow cytometry analysis of three of these cell lines (SKOV-3, CaOV-3, and OVCAR-3) showed that 82– 92% of the cells express a3 integrin, only second to b-1 integrin expressed by 89– 97% of the cells [31]. Although SKOV-3 cells seemed to express the highest level of this integrin among the other cell lines, we found that the xenografts were more challenging to grow than the less expressing ES-2 cell line. As previously stated, a3 integrin plays a very important role in normal and tumor development. It is therefore noteworthy that we can image this integrin expression by using a novel peptide identified from our validated ‘‘OBOC’’ technology. Using one of these probes, we were able to stain SKOV-3 cells that were grown on glass slides (Figure 3). The non-a3-expressing

446


Raji cell line did not show any cell staining with this peptide. Uptake of the ‘‘OA02’’ – biotin – streptavidin – Cy5.5 probe was much slower (1 hr) but florescence signal was sustained longer (24 hr) in the tumor. The lower molecular weight ‘‘OA02’’ – Cy5.5 and ‘‘OA02’’ – Alexa were observed in the tumor 15 min postinjection, fluorescent intensity peaked at 70 min and was completely washed out within 4 hr. Because there were 2 moles of Cy5.5 conjugated to streptavidin, the light intensity seemed to be higher than the other probes but in actuality AU values were lower. Kidney, bladder and urine excretion of the probe is most likely a result of the hydrophilic property of the peptide and the Cy5.5 dye. No fluorescence signal was observed in ES-2 tumors of mice injected with streptavidin– Cy5.5 alone. The fact that bulky (60 kDa) and tetravalent nature of biotin – peptide – streptavidin – complex was able to slow down the pharmacokinetics of the probe but not interfere with the specificity could be vital in modifying the peptide for therapeutic use. However, for imaging purposes, the smaller molecular weight (2.6 kDa) ‘‘OA02’’ –Cy5.5 is ideal because it is rapidly and specifically delivered to the tumor but also rapidly cleared within 3 to 4 hr. Because these probes are injected intravenously, it is important that the tumors have an adequate blood supply. The fact that preinjection with anti-a3 integrin antibody was able to block the uptake of the two peptide probes (‘‘OA02’’ –Cy5.5 and ‘‘OA02’’ – Alexa 680) was evidence that the peptide probes were specifically binding to the a3 integrin receptors in the tumors (Figure 4A: IV). Kidney uptake and intricate detail of background signal of the vertebrae is evidence that the peptide probe was in circulation but was not taken up by the tumor. Clinically, the use of this peptide as a targeting agent for imaging ovarian cancer and metastatic lesions as well as other a3 integrin expressing tumors is an attractive option. Although extensive toxicity studies have not been performed, the peptide conjugates seem to be well tolerated by the animals, they appear to be specific and rapidly cleared. In PET studies using F18-FDG to image ES-2 tumors, we observed that there was little or no uptake of the radioactive probe by this tumor (Aina O., unpublished data). It is therefore very interesting that we can image this tumor with the integrinbinding peptide ‘‘OA02.’’ Clinical use of the NIR peptide probe to image ovarian tumors is a possibility because the ovaries are somewhat close to surface of the abdomen. It is also important to note that SKOV-3 cell line is phenotypically identical to the epithelial component of

human ovarian cancer, but ES-2 cell lines are more fibroblastic in nature and could represent the stromal component of ovarian cancer which could be useful in tumor targeting using antineovascular therapies. Besides ovarian tumors, this a3 peptide probe could be useful in imaging other disease systems in which this integrin is overexpressed (e.g., astrocytomas, gliomas, melanoma, medulloblastoma, and gastric tumors [5]). Future studies will include optimization of this ligand for therapeutics as well as for other imaging modalities that may be more suited for clinical use such as PET and MRI.

Conclusion We have been able to demonstrate previously that using the one-bead one-peptide library technology we can identify novel ligands for tumor cells. Once these novel ligands are identified they can be further developed into tumor imaging agents. This study clearly validates that fact. Because we have been able to show the selectivity and specificity of this ligand, future investigation will include steps to optimize this ligand for therapy. We envision that we can conjugate a chemotherapeutic drug to this peptide and study tumor response.

Acknowledgments R33 CA89706 (NIH, NCI); 00-00764V-0133 (California Cancer Research Program); 5-T32-RR07038 (NIH Training Grant); Landgraf Cancer Research Award. We thank Drs Frederick Meyers, Julie Sutcliffe-Goulden, and Stephen Barthold for support; Drs Ruiwu Liu, Aimin Song, Huonggee Beak, and Xiaobing Wang for assistance.

References [1] Aina OH, Marik J, Liu R, Lau DH, Lam KS (2005). Identification of novel targeting peptides for human ovarian cancer cells using ‘‘one-bead one-compound’’ combinatorial libraries. Mol Cancer Ther. 4:806 – 813. [2] Aina OH, Marik J, Kukis D, et al. (in preparation). Optimization of cDGXGXXc peptide ligand for Ovarian PET imaging. [3] Yamada KM, Danen EHJ (2000). Integrin signaling. In: J. S. Gutkind (Ed.), Signaling Networks and Cell Cycle Control: The Molecular Basis of Cancer and Other Diseases. Humana, Totowa, NJ. [4] Hynes RO (2002). Integrins: Bidirectional, allosteric signaling machines. Cell. 110:673 – 687. [5] Kreidberg JA (2000). Functions of alpha3beta1 integrin. Curr Opin Cell Biol. 12:548 – 553. [6] Schmid RS, Shelton S, Stanco A, Yokota Y, Kreidberg JA, Anton ES (2004). a3b1 integrin modulates neuronal migration and placement during early stages of cerebral cortical development. Development. 131:6023 – 6031.

Molecular Imaging . Vol. 4, No. 4, October 2005

Ovarian Tumor Optical Imaging with Novel Peptide Probe Aina et al. [7] Kreidberg JA, Symons JM (2000). Integrins in kidney development, function, and disease. Am J Physiol Renal Physiol. 279: F233 – F242. [8] Wang Z, Symons JM, Goldstein SL, McDonald A, Miner JH, Kreidberg JA (1999). (Alpha)3(beta)1 integrin regulates epithelial cytoskeletal organization. J Cell Sci. 112:2925 – 2935. [9] DiPersio CM, Hodivala-Dilke KM, Jaenisch R, Kreidberg JA, Hynes RO (1997). Alpha3beta1 integrin is required for normal development of the epidermal basement membrane. J Cell Biol. 137: 729 – 742. [10] Kreidberg JA, Donovan MJ, Goldstein SL, Rennke H, Shepherd K, Jones RC, Jaenisch R (1996). Alpha 3 beta 1 integrin has a crucial role in kidney and lung organogenesis. Development. 122: 3537 – 3547. [11] Morini M, Mottolese M, Ferrari N, Ghiorzo F, Buglioni S, Mortarini R, Noonan DM, Natali PG, Albini A (2000). The alpha 3 beta 1 integrin is associated with mammary carcinoma cell metastasis, invasion, and gelatinase B (MMP-9) activity. Int J Cancer. 87:336 – 342. [12] Coopman PJ, Thomas DM, Gehlsen KR, Mueller SC (1996). Integrin alpha 3 beta 1 participates in the phagocytosis of extracellular matrix molecules by human breast cancer cells. Mol Biol Cell. 7:1789 – 1804. [13] van der Pluijm G, Vloedgraven H, Papapoulos S, Lowick C, Grzesik W, Kerr J, Robery PG (1997). Attachment characteristics and involvement of integrins in adhesion of breast cancer cell lines to extracellular bone matrix components. Lab Invest. 77: 665 – 675. [14] Kiefer JA, Farach-Carson MC (2001). Type I collagen-mediated proliferation of PC3 prostate carcinoma cell line: Implications for enhanced growth in the bone microenvironment. Matrix Biol. 20:429 – 437. [15] Curley GP, Blum H, Humphries MJ (1999). Integrin antagonists. Cell Mol Life Sci. 56:427 – 441. [16] Albelda SM, Mette SA, Elder DE, Stewart R, Damjanovich L, Herlyn M, Buck CA (1990). Integrin distribution in malignant melanoma: Association of the beta 3 subunit with tumor progression. Cancer Res. 50:6757 – 6764. [17] Ke S, Wen X, Gurfinkel M, Charnsangavej C, Wallace S, SevickMuraca EM, Li C (2003). Near-infrared optical imaging of epidermal growth factor receptor in breast cancer xenografts. Cancer Res. 63:7870 – 7875. [18] Moon WK, Lin YH, O’Loughlin T, Tang Y, Kim DE, Weissleder R, Tung CH (2003). Enhanced tumor detection using a folate receptor-targeted near-infrared fluorochrome conjugate. Bioconjug Chem. 14:539 – 545. [19] Tung CH, Lin YH, Moon WK, Weissleder R (2002). A receptor-

Molecular Imaging . Vol. 4, No. 4, October 2005

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

447

targeted near-infrared fluorescence probe for in vivo tumor imaging. ChemBioChem. 3:784 – 786. Citrin D, Scott T, Sproull M, Menard C, Tofilon PJ, Camphausen K (2004). In vivo tumor imaging using a near-infrared-labeled endostatin molecule. Int J Radiat Oncol Biol Phys. 58:536 – 541. Bugaj JE, Achilefu S, Dorshow RB, Rajagopalan R (2001). Novel fluorescent contrast agents for optical imaging of in vivo tumors based on a receptor-targeted dye – peptide conjugate platform. J Biomed Opt. 6:122 – 133. Achilefu S, Jimenez HN, Dorshow RB, Bugaj JE, Webb EG, Wilhelm RR, Rajagopalan R, Johler J, Erion JL (2002). Synthesis, in vitro receptor binding, and in vivo evaluation of fluorescein and carbocyanine peptide-based optical contrast agents. J Med Chem. 45:2003 – 2015. Bremer C, Bredow S, Mahmood U, Weissleder R, Tung CH (2001). Optical imaging of matrix metalloproteinase-2 activity in tumors: Feasibility study in a mouse model. Radiology. 221: 523 – 529. Wang W, Ke S, Wu Q, Charnsangavej C, Gurfinkel M, Gelovani G, Abbruzzese JL, Sevick-Muraca EM, Li C (2004). Near-infrared optical imaging of integrin alpha v beta 3 in human tumor xenografts. J Soc Mol Imaging. 3:343 – 351. Ke S, Wang W, Houston J, Sevick-Muraca EM, Gelovani JG, Li C (2004). Dual modality/nuclear imaging of integrin alpha v beta 3 in melanoma with a single imaging probe. Society of Molecular Imaging Annual Meeting, St. Louis, MO. Reynolds JS, Troy TL, Mayer RH, Thompson AB, Waters DJ, Cornell KK, Snyder PW, Sevick-Muraca EM (1999). Imaging of spontaneous canine mammary tumors using fluorescent contrast agents. Photochem Photobiol. 70:87 – 94. Godavarty A, Thompson AB, Roy R, Gurfinkel M, Eppstein MJ, Zhang C, Sevick-Muraca EM (2004). Diagnostic imaging of breast cancer using fluorescence-enhanced optical tomography: Phantom studies. J Biomed Opt. 9:488 – 496. Gu X, Zhang Q, Bartlett M, Schutz L, Fajardo LL, Jiang H (2004). Differentiation of cysts from solid tumors in the breast with diffuse optical tomography. Acad Radiol. 11:53 – 60. Hawrysz DJ, Sevick-Muraca EM (2000). Developments toward diagnostic breast cancer imaging using near-infrared optical measurements and fluorescent contrast agents. Neoplasia. 2:388 – 417. Song A, Wang X, Zhang J, Marik J, Lebrilla CB, Lam KS (2004). Synthesis of hydrophilic and flexible linkers for peptide derivatization in solid phase. Bioorg Med Chem Lett. 14:161 – 165. Cannistra SA, Ottensmeier C, Niloff J, Orta B, Dicarlo J (1995). Expression and function of beta-1 and alpha-v-beta-3 integrins in ovarian cancer. Gynecol Oncol. 58:216 – 225.