Targeting epitopes in prostate-specific membrane antigen for ... - Nature

Prostate Cancer and Prostatic Diseases (2005) 8, 359–363 & 2005 Nature Publishing Group All rights reserved 1365-7852/05 $30.00 www.nature.com/pcan

Targeting epitopes in prostate-specific membrane antigen for antibody therapy of prostate cancer Y Kinoshita1, K Kuratsukuri1, N Newman2, PM Rovito1, Jr., PTP Kaumaya3, CY Wang1,4 & GP Haas1,4* 1

Department of Urology, SUNY Upstate Medical University, Syracuse, New York, USA; 2Department of Medicine, SUNY Upstate Medical University, Syracuse, New York, USA; 3Department of Obstetrics and Gynecology, Ohio State University, Columbus, Ohio, USA; and 4VA Medical Center, Syracuse, New York, USA

Prostate-specific membrane antigen (PSMA) is a target for immunotherapy of prostate cancer. It has been shown that antibodies against PSMA inhibited the in vivo growth of LNCaP tumor. In the present study, monoclonal antibodies against four epitopes in PSMA were raised. MAb 24.4E6 (IgG1), specific for the epitope (residues 638–657) in PSMA, significantly reduced the growth rate of established LNCaP tumor in SCID mice. Mouse IgG was detected in the tumor of mice treated with 24.4E6, but not with an unrelated MAb. These results suggest that this epitope may be the main target in PSMA for antibody therapy of prostate cancer. Prostate Cancer and Prostatic Diseases (2005) 8, 359–363. doi:10.1038/sj.pcan.4500835; published online 20 September 2005

Keywords: antibody; prostate-specific membrane antigen; LNCaP; prostate cancer; antitumor activity; glutamate carboxypeptidase II

Introduction Prostate specific membrane antigen (PSMA) is a folate hydrolase that hydrolyzes g-glutamyl linkages in methotrexate triglutamate,1 and it is synonymous to glutamate carboxypeptidase II (EC 3. 4.17.21).2 It also has a N-acetylated-a-linked acidic dipeptidase activity that hydrolyzes N-acetyl-L-aspartyl-L-glutamate to a strong neurotransmitter glutamic acid.3,4 PSMA is a type 2 transmembrane glycoprotein comprising 750 amino acids.5 The transmembrane region consists of about 43 amino-acid residues at the N-terminus. Between residue 418 and 567 is a transferring receptor-like domain. Alternative splice variants of PSMA that lack the transmembrane domain are present in the cytoplasm.6,7 Transmembrane form PSMA is upregulated in prostate cancer and it is expressed in neovasculature of a variety of cancers.8,9 It is therefore an attractive target for immunotherapy. The extracellular domain of PSMA may

*Correspondence: GP Haas, Department of Urology, SUNY Upstate Medical University, 750 East Adams Street, Syracuse, NY 13210, USA. E-mail: [email protected] Received 10 May 2005; revised 16 July 2005; accepted 20 July 2005; published online 20 September 2005

be accessible to antibodies and it may be a target for antibody-mediated cytotoxicity. We previously showed that vaccination with NIH3T3 cells that express PSMA or its C-terminal peptide elicited in vivo antitumor activity against Renca cells that express PSMA.10 The anti-tumor activity was also demonstrated following passive transfer of immune serum to LNCaP tumor-bearing athymic mice.10 The purpose of this study was to identify epitopes in the extracellular domain of PSMA that may be responsible for antibody-mediated cytotoxicity. In this study, the peptide at residues 567–750 that is the helical domain in PSMA as recently determined by crystallography11 was subjected to a computer-aided analysis and four potential epitopes were identified. Monoclonal antibodies against these epitopes were prepared. With these antibodies, we identified an epitope that was important for antibody-mediated anti-tumor activity.

Materials and methods Selection of B-cell epitopes of PSMA B-cell epitopes were determined by computer-aided analysis using various correlates of protein antigenicity and secondary structural predictions according to the

PSMA epitopes for antibody therapy Y Kinoshita et al

360

algorithm procedure described by Kaumaya et al.12,13 The peptide (residues 567–750) in PSMA that follows the transferrin receptor-like domain of PSMA was subjected to analyses. Four oligopeptides corresponding to the amino-acid residues of PSMA were selected and listed below from the highest to the lowest score: Residues Residues Residues Residues

638–657: 714–732: 614–628: 693–715:

NFTEIASKFSERLQDFDKSN-COOH, DIESKVDPSKAWGEVKRQI-COOH, ISMKHPQEMKTYSVS-COOH, APSSHNKYAGESF-COOH.

Searches in the NBCI website (http://ncbi.nlm.nih.gov/ BLAST/Blast.cgi) found that the first three peptides were unique for human PSMA. The underlined residues differed from rat or mouse glutamate carboxypeptidase. The last sequence was identical to mouse and rat glutamate carboxypeptidase. Linking these peptides to the C-terminal end of the ‘promiscuous’ measles virus fusion protein (MVF) T-cell epitope KLLSLIKGVIVHRLEGVE with a LSPG linker forms 35–42 amino acid-residue T- and B-cell epitope chimera. These oligopeptides were synthesized by the solid phase synthesis and purified by HPLC.12 Mass spectroscopic analysis gave the correct molecular weight.

Preparation of hybridomas The published procedure for hybridoma preparation14 was modified. Oligopeptide, 1.5 mg, was dissolved in 0.75 ml PBS and then emulsified with 0.75 ml complete Freund’s adjuvant (CFA) or incomplete Freund’s adjuvant (IFA) (Pierce Biotech. Inc., Rockford, IL, USA). Two week-old BALB/c mice (Charles River, Wilmington, MA, USA) were injected, s.c., at multiple sites with a total of 0.1 ml of the peptide in CFA. On week 3, the mouse was injected with 0.1 ml of the peptide emulsified in IFA. On week 6, 0.1 mg of the peptide dissolved in 0.1 ml PBS was injected, i.v., and the mouse was killed a week later. The spleen cells were fused with BALB/c myeloma P3C63 Ag8.653 cells (ATCC, Rockville, MD, USA) at a lymphocytes and myeloma cells ratio of 10 : 1 in the presence of polyethylene glycol 1500 (Roche Diagnostics Corp., Indianapolis, IN, USA). Culture media from hybridomas were screened with ELISA using the synthetic peptidecoated 96-well titer plates. Color development was performed using a Vectastain ABC kit (Vector, Burlingame, CA, USA). To prepare ELISA plates, synthetic peptides were dissolved in distilled water at 0.05 mg/ml and 0.05 ml of the solution was added to each well of the 96-well titer plate. The plate was stored at 41C until the well was dried. The plate was washed extensively with distilled water before being used. To exclude MAbs that were specific for the MVF, those crossreacted with different PSMA peptides were discarded. Isotyping of the antibody was performed using an isotyping kit (BioRad, Hercules, CA, USA).

Determination of affinity The affinity was determined with a conventional competition ELISA method.15 Appropriately diluted antibody solution was incubated for 1 h with targeted Prostate Cancer and Prostatic Diseases

peptide that had been diluted to as low as 0.1 nM. The mixture was then added to an ELISA plate and kept at room temperature for 10 min. The plate was washed and ELISA test was performed for the detection of bound MAb. The intensity of binding was determined using a BioRad 96-well plate reader.

Immunohistochemical staining Formalin-fixed and paraffin-embedded prostate tissues were sectioned to 5 mm thickness, deparaffinized and hydrated. The slide was immersed in 10 mM citric buffer (pH 6.0) and microwaved for 10 min to achieve antigen retrieval. A 5% bovine serum albumin solution was dropped on the section and incubated for 30 min for blocking non-specific binding. Undiluted media of hybridoma culture was used as the primary antibody and the section was incubated at room temperature for 3 h in a humidified chamber. Alkaline phosphatase conjugated anti-mouse IgG and anti-mouse IgM antibody (Sigma, St. Louis, MO, USA) were diluted 50 times with PBS and used as the secondary antibodies. After incubation at room temperature for an additional 2 h, color development was performed using FastRed color development substrate (Sigma). The slide was then counterstained with hematoxylin, dehydrated and mounted. Normal mouse IgG or IgM was used in place of the primary antibody as a negative control. Reaction of MAb with cultured LNCaP and PC-3 cells was carried out in the same manner.

Large-scale preparation of MAb To prepare large amount of MAb for experiments, SCID mice (Charles River) were injected i.p. with 0.2 ml Pristane (Sigma) 7 days prior to i.p. injection of 10 million hybridoma cells. Ascites fluid was harvested 10 days later and titrated. The maximum dilution with positive reaction was defined as the titer.

In vivo toxicity against LNCaP tumor LNCaP (ATCC) were cultured as described.16 SCID mice 5 to 6-week-old male were s.c. injected with 2 million LNCaP cells suspended in 0.1 ml of 50% Matrigel (BD Biosciences, Bedford, MA, USA). When tumors became palpable, they were randomly divided into 4 groups with 5–7 mice in each group and given weekly i.p. injection of 0.5 ml of MAb (ascites fluid) for 4 weeks. HB65 ascites served as a negative control. Tumor dimensions were measured and tumor volumes were approximated as (length width2)/2. The data were statistically analyzed by a least-square ANOVA analysis or the student’s t-test with a SAS program (SAS Institute, Cary, NC, USA).

Results Characterization of MAb More than 80 different hybridomas were expanded. Four were subcloned and reselected to ensure the purity. All


antibodies were specific to the peptides used for their preparation. All their light chains were k-type. Three were IgG1 and one was an IgM. Their characteristics are shown in Table 1. All the MAbs reacted with LNCaP cells but not with PC-3 cells as determined by immunohistochemical method (Figure 1). All the antibodies also reacted with prostate epithelial cells in formalin-fixed prostate tissue (Figure 1). Since mouse and rat kidneys express glutamate carboxypeptidase,17,18 this tissue was used to test the specificity of 24.4E6 and 14.3C8. Mouse and rat renal cortical tubule cells were stained with 14.3C8, but not with 24.4E6 (photos not shown).

contrasts between each of the treatment groups and HB65, Wilks’ Lambda yielded a significant difference in 24.4E6 only (F ¼ 4.95, Po0.01). There was no significant difference after week 4. When the comparison of tumor volume was made only at week 4, groups 24.4E6 and 58.3D3 were significantly different from group HB65 (Po0.05). The weekly tumor volume was significantly different from week 1 to 9 between HB65 and 24.4E6 groups (student’s t-test, Po0.05). At the end of the experiment, tumor tissues were subjected to microscopic examination. The infiltration of phagocytic macrophages into the tumor area and tumor necrosis were invariably observed in every group including the HB65 group.

361

In vivo activity against LNCaP tumor Since a preliminary experiment showed that 28.2E1 was not cytotoxic, this MAb was excluded from the experiment. To demonstrate the accumulation of MAb in the tumor area, the mice were injected, i.p., with 0.5 ml HB65 or 24.4E6 ascites when LNCaP tumor size reached approximately 5 5 mm in dimension. They were killed 2 days later and tumors were excised. Mouse IgG was detected immunohistochemically in the tumor of 24.4E6treated but not HB65-treated mice (Figure 2). To test anti-tumor activity, MAbs were injected weekly for 4 weeks to tumor-bearing mice. The result of antitumor activity was shown in Figure 3. Using a leastsquare ANOVA analysis, it was found that there was a significant difference in tumor growth rate between treatment groups over the first 4 weeks. In tests of Table 1 Characteristics of MAb MAb

B-cell epitope

Isotype

Titer

Affinity (nM)

14.3C8 24.4E6 58.3D3 28.2E1

Residues Residues Residues Residues

IgM/k IgG1/k IgG1/k IgG1/k

8000 128000 256000 16000

600 15 20 120

693–715 638–657 614–628 714–732

Figure 1 Immunohistochemical staining of LNCaP cells and prostate cancer epithelium with 24.4E6. Positive reaction of cultured LNCaP cells with 24.4D6 (a); negative reaction of LNCaP cells with normal mouse IgG (b); negative reaction of PC-3 cells with 24.4E6 (c). The slides were counterstained with hematoxylin and positive reaction showed red color. MAb 24.4E6 reacted with human prostate epithelial cells (d) but normal mouse IgG did not (e). The slides were counterstained with hematoxylin and positive reaction showed red color.

Discussion LNCaP cells express full-length PSMA as well as three alternative splice variants that lack the intracellular and transmembrane domains.6,7 MAb 7E11 is specific for an intracellular epitope and it binds to PSMA but not to

Figure 2 Accumulation of MAb in the tumor area. LNCap tumorbearing mice were injected, i.p., with either HB65 (b) or 24.4E6 (a) 2 days prior to being killed. Mouse IgG was detected in the tumor of mice treated with 24.4E6 (a) but not treated with HB65 (b). The slide was counterstained with hematoxylin. Positive stain shows red color.

Figure 3 Antitumor activities of MAb. MAb were injected, i.p., once a week for 4 weeks beginning on week 0. LNCaP tumor-bearing male SCID mice were injected with HB65 (K), 14.3C8 (’), 58.3D3 (&) and 24.4E6 (J). There were six mice for HB65 and 14.3C8 groups, seven mice for 24.4E6 group and five mice for 58.3D3 group. The bar equals to one s.d. Using a least-square ANOVA analysis, the growth rates over the first 4 weeks is significantly different between HB65 and 24.4E6 (F ¼ 4.95, Po0.01), but it is not significantly different between HB65 and 58.3D3 (F ¼ 2.75, P ¼ 0.06). The difference between HB65 and 24.4E6 over the entire period is not significant. At week 4, the tumor volume of HB65 group is significantly greater than that of the other groups (P ¼ 0.01). The weekly tumor volumes between HB65 and 24.4E6 were significantly different up to 9 week (Po0.05). Prostate Cancer and Prostatic Diseases


362

alternative splice forms.19 As a result, 7E11 does not bind viable prostate cancer cells.8,9 A number of other antiPSMA MAbs that bind epitopes that are distinct from that recognized by 7E11 have been reported.8,9 MAbs J591, J415, J533, E998,9 and PEQ226.520,21 bind to the extracellular PSMA domain. PM2J004.5 binds to an epitope of the intracellular domain that is distinct from that bound by 7E11.21 In the present study, we have produced four MAbs against four different epitopes in the extracellular domain of PSMA. These antibodies reacted with prostate epithelial cells as determined by immunohistochemical method. MAb 24.4E6 was also tested for mouse and rat kidney and it did not react with these tissues. Among the MAbs, 24.4E6 had the greatest in vivo anti-LNCaP activity. MAbs 58.3D3 and 24.4E6 were comparable in titer and affinity but may be due to a smaller number of animals in the 58.3D3 group, only 24.4E6 signficantly reduced the growth rate of LNCaP during the first 4 weeks of treatment. Tumors in 24.4E6-treated group were significantly smaller than in HB65-treated group up to 9 weeks after the treatment. The in vivo activity of 24.4E6 against LNCaP tumor was equivalent to that of the antiserum obtained from mice that have been vaccinated with PSMc-expressing NIH3T3 cells even though the treatment regimens of these two experiments were different.10 MAb 14.3C8 was an IgM subtype antibody and its titer was much lower than that of 24.4E6 and 58.3D3. This antibody was less active than 24.4E6 and 58.3D3 in anti-LNCaP activity. Our results suggest that the epitope (residues 638–657) in PSMA may be a prime target for antibody therapy. This is consistent with the results from computer-aided algorithm analysis in the determination of immunogenicity. Hormone ablation and anti-hormone agents remains the systemic treatment of choice for disseminated disease since they can achieve a high response rate. However, hormone independence and disease progression typically develop in less than 2 years, and survival benefits remain disappointing.22 Although cellular immunotherapies involving primed dendritic cells and activated T cells are highly attractive because of their potential for providing durable and specific cytotoxicity against established tumors, their efficacy against advanced prostate disease may be limited because of impairments in class I antigen presentation in most metastatic prostate cancers.23–25 Antibody treatment that is not dependent on antigen presentation by the target cells is a viable alternative. PSMA antibodies labeled with radioisotopes26,27 or toxins28 may increase the efficacy of the antibody. Radiolabeled monoclonal antibody J591, specific to the extracellular domain of PSMA, is currently used for clinical trials.26,27 Early results indicate that J591 has therapeutic potential in metastatic prostate cancer. In the present study, we have identified a prime epitope in PSMA for antibody therapy. Antibody against this epitope may be used for the treatment and imaging of prostate cancer.

References 1 Ghosh A, Heston WD. Tumor target prostate specific membrane antigen (PSMA) and its regulation in prostate cancer. J Cell Biochem 2004; 91: 528–539. Prostate Cancer and Prostatic Diseases

2 Barrett AJ. Nomenclature Committee of the International Union of Biochemistry and Molecular Biology (NC-IUBMB). Enzyme Nomenclature. Recommendations 1992. Supplement 4: corrections and additions (1997). Eur J Biochem 1997; 250: 1–6. 3 Carter RE, Feldman AR, Coyle JT. Prostate-specific membrane antigen is a hydrolase with substrate and pharmacologic characteristics of a neuropeptidase. Proc Natl Acad Sci USA 1996; 93: 749–753. 4 Luthi-Carter R et al. Molecular characterization of human brain N-acetylated alpha-linked acidic dipeptidase (NAALADase). J Pharmacol Exp Ther 1998; 286: 1020–1025. 5 Israeli RS et al. Molecular cloning of a complementary DNA encoding a prostate-specific membrane antigen. Cancer Res 1993; 53: 227–230. 6 Su SL et al. Alternatively spliced variants of prostate-specific membrane antigen RNA: ratio of expression as a potential measurement of progression. Cancer Res 1995; 55: 1441–1443. 7 Schmittgen TD et al. Expression of prostate specific membrane antigen and three alternatively spliced variants of PSMA in prostate cancer patients. Int J Cancer 2003; 107: 323–329. 8 Liu H et al. Monoclonal antibodies to the extracellular domain of prostate-specific membrane antigen also react with tumor vascular endothelium. Cancer Res 1997; 57: 3629–3634. 9 Chang SS et al. Five different anti-prostate-specific membrane antigen (PSMA) antibodies confirm PSMA expression in tumorassociated neovasculature. Cancer Res 1999; 59: 3192–3198. 10 Kuratsukuri K et al. Inhibition of prostate-specific membrane antigen (PSMA)-positive tumor growth by vaccination with either full-length or the C-terminal end of PSMA. Intl J Cancer 2002; 102: 244–249. 11 Davis MI et al. Crystal structure of prostate-specific membrane antigen, a tumor marker and peptidase. Pro Natl Acad Sci USA 2005; 102: 5981–5986. 12 Dakappagari NK et al. Prevention of mammary tumors with a chimeric HER-2 B-cell epitope peptide vaccine. Cancer Res 2000; 60: 3782–3789. 13 Kaumaya PTP et al. ‘De novo’ engineering of peptide immunogenic and antigenic determinants as potential vaccines. In: Basava C, Anantharamaiah GM (eds). Peptides: Design, Synthesis and Biological Activity. Birkhauser: Boston, 1994, pp 133–164. 14 Land SJ et al. Purification and characterization of a rat hepatic acetyltransferase that can metabolize aromatic amine derivatives. Carcinogenesis 1993; 14: 1441–1449. 15 Bumke MA, Neri D. Affinity measurements by band shift and competition ELISA. In: Kontermann R, Dubel S (eds). Antibody Engineering. Springer-Verlag: New York, 2001, pp 385–396. 16 Soos G et al. Comparative intraosseal growth of human prostate cancer cell lines LNCaP and PC-3 in the nude mouse. Anticancer Res 1997; 17: 4253–4258. 17 Slusher BS et al. Rat brain N-acetylated alpha-linked acidic dipeptidase activity. Purification and immunologic characterization. J Biol Chem 1990; 265: 21297–21301. 18 Bacich DJ et al. Cloning, expression, genomic localization, and enzymatic activities of the mouse homolog of prostate-specific membrane antigen/NAALADase/folate hydrolase. Mamm Genome 2001; 12: 117–123. 19 Horoszewicz JS, Kawinski E, Murphy GP. Monoclonal antibodies to a new antigenic marker in epithelial cells and serum of prostatic cancer patients. Anticancer Res 1987; 7: 927–936. 20 Sokoloff RL et al. A dual-monoclonal sandwich assay for prostate-specific membrane antigen: levels in tissues, seminal fluid and urine. Prostate 2000; 43: 150–157. 21 Grauer LS et al. Identification, purification, and subcellular localization of prostate-specific membrane antigen PSM’ protein in the LNCaP prostatic carcinoma cell line. Cancer Res 1998; 58: 4787–4789. 22 Crawford ED et al. Hormone refractory prostate cancer. Urol 1999; 54: S1–S7. 23 Sanda MG et al. Molecular characterization of defective antigen presentation in human prostate cancer. J Natl Cancer Inst 1995; 87: 280–285.

PSMA epitopes for antibody therapy Y Kinoshita et al 24 Blades RA et al. Loss of class I expression in prostate cancer: implications for immunotherapy. Urol 1995; 46: 681–687. 25 Bander N et al. MHC class I and II expression in prostate carcinoma and modulation by interferon-alpha and -gamma. Prostate 1997; 33: 233–239. 26 Bander NH et al. Targeted systemic therapy of prostate cancer with a monoclonal antibody to prostate-specific membrane antigen. Seminars Oncol 2003; 30: 667–676.

27 Bander NH et al. Targeting metastatic prostate cancer with radiolabeled monoclonal antibody J591 to the extracellular domain of prostate specific membrane antigen. J Urol 2003; 170: 1717–1721. 28 Henry MD et al. A prostate-specific membrane antigen-targeted monoclonal antibody-chemotherapeutic conjugate designed for the treatment of prostate cancer. Cancer Res 2004; 64: 7995–8001.

363

Prostate Cancer and Prostatic Diseases