Abridged from Cancer Research 57, 3629-3634, September 1, 1997.
Monoclonal Antibodies to the Extracellular Domain of Prostate Specific Membrane Antigen Also React With Tumor Vascular Endothelium1
He Liu, Peggy Moy, Sae Kim, Yan Xia, Ayyappan Rajasekaran, Vincent Navarro, Beatrice Knudsen and Neil H Bander2
From the Departments of Urology [H.L., P.M., S.K., Y.X., V.N., N.H.B.], Cell Biology [A.R.] and Pathology [B.K.], New York Hospital-Cornell Medical Center, New York, New York 10021, and the Ludwig Institute for Cancer Research, New York Branch [N.H.B.], New York, New York 10021
Prostate specific membrane antigen (PSMA), initially defined by monoclonal antibody (mAb) 7E11, is a now well-characterized type 2 integral membrane glycoprotein expressed in a highly restricted manner by prostate epithelial cells. 7E11 has been shown to bind an intracellular epitope of PSMA which, in viable cells, is not available for binding. We herein report the initial characterization of the first 4 reported IgG mAbs which bind the external domain of PSMA. Competitive binding studies indicate these antibodies define 2 distinct, non-competing epitopes on the extracellular domain of PSMA. In contrast to 7E11, these mAbs bind to viable LNCaP cells in vitro. In addition, they show strong immunohistochemical reactivity to tissue sections of prostate epithelia, including prostate cancer. These mAbs were also strongly reactive with vascular endothelium within a wide variety of carcinomas (including lung, colon, breast and others), but not with normal vascular endothelium. These antibodies should prove useful for in vivo targeting to prostate cancer as well as to the vascular compartment of a wide variety of carcinomas.
Prostate specific membrane antigen (PSMA) is a highly restricted prostate epithelial cell membrane glycoprotein of approximately 100 kD (1,2). The PSMA gene has been cloned, sequenced (2), and mapped to chromosome 11q14 (3). In contrast to other highly restricted prostate-related antigens such as prostate specific antigen (PSA), prostatic acid phosphatase (PAP) and prostate secretory protein (PSP), which are secretory proteins, PSMA is an integral membrane protein. Among reasons for significant interest in PSMA is that it is ideal for in vivo prostate-specific targeting strategies. In addition to its prostate specificity (1,2,4,5), PSMA is expressed by a very high proportion of prostate cancers (PCa)(6), expression is further increased in higher grade cancers and in metastatic disease (6) and in hormone-refractory PCa (5,6,7).
Initial validation of PSMA as an in vivo target has been borne out by imaging trials with mAb 7E11/CYT-356 (8-10). Epitope mapping, however, indicates that 7E11/CYT-356 targets an intracellular epitope (12,13). In viable cells, this binding site is not accessible to antibody (1,13). Successful imaging with 7E11/CYT-356 probably relates to targeting of dead/dying cells within tumor sites (6,12,13). It has been noted (2,12-14) that a mAb to the extracellular domain would provide benefits including improved in vivo localization and enhanced imaging and therapy. In this study, we report the development of 4 IgG mAbs to the external domain of PSMA. These mAbs also have been found reactive to vascular endothelium within a wide range of carcinomas, but not with normal endothelial cells.
Immunoprecipitation/Immunoblot. In western blot analysis, mAbs J591 (IgG1), J533 (IgG1), J415 (IgG1) and E99 (IgG3), as well as 7E11, identified a 100 kD band from LNCaP lysate but not from the PSMA-negative PC3 lysate (data not shown). To confirm that mAbs J591, J533, J415 and E99 detected the same antigen as 7E11, a cross immunoprecipitation experiment was performed. Fig. 1 illustrates that the 100 kD band which was immunoprecipitated by mAbs J591, J533, J415, E99 or 7E11 were detectable by immunoblot using either J591 or 7E11 as a probe (Fig. 1A and B, respectively). Sequential immunoprecipitation studies (data not shown) also demonstrated that 7E11 and the 4 new mAbs can pre-clear reactivity to one another.
Immunohistochemical reactivity. The reactivity of mAbs J591, J533, J415 and E99 with normal human tissues and cancers, with rare exception (vide infra), were similar to 7E11. Normal tissues with similar immunohistochemical reactivity included prostate (normal and hyperplastic glands demonstrated heterogeneous, weak to moderate staining intensity), kidney (subset of proximal tubules) and duodenum (weakly reactive). The only normal tissue in which we found any difference in reactivity was striated muscle. While 7E11 was strongly reactive to striated muscle, mAbs J591, J533, J415 and E99 demonstrated no reactivity. In neoplastic tissues, findings were again similar when comparing 7E11 to mAbs J591, J533, J415 and E99. All 21 prostate cancers studied were strongly reactive with mAbs J591, J533, J415 and E99, being somewhat more intense and more homogeneous than 7E11. As previously reported (16), we found 7E11 reacted with vascular endothelium in a subset of tumors. However, mAbs J591, J533, J415 and E99 reacted more strongly with vascular endothelium in all 23 carcinomas studied (Fig. 2) including 9/9 renal, 5/5 urothelial, 6/6 colon, 1/1 lung, 1/1 breast and 1/1 metastatic adenocarcinoma to the liver.
Immunofluorescence staining of LNCaP cells. We compared, by indirect immunoflourescence, mAbs J591, J533, J415 and E99 to mAb 7E11 on viable or fixed, permeabilized or non-permeabilized LNCaP cells (Fig. 3). LNCaP cells with intact plasma membrane (i.e., either viable [data not shown] or fixed without permeabilization) demonstrated cell surface reactivity with mAbs J591, J533, J415 and E99 (Fig. 3A,C,E,G), but not with mAb 7E11 (Fig. 3I). Only after LNCaP cells were permeabilized could 7E11 reactivity be demonstrated (Fig. 3J). Once permeabilized, reactivity of all mAbs appeared both in the cytoplasm and on the plasma membrane.
Immunoelectron microscopy. IEM similarly demonstrated immunoreactivity of mAb J591 (Fig. 4A), but not 7E11 (Fig. 4B), with viable LNCaP cells. Furthermore, the IEM photomicrographs of mAb J591 show the gold particles localized to the extracellular face of the plasma membrane confirming reactivity with the extracellular domain of PSMA.
Competitive bindingassay. A double antibody sandwich competition ELISA was used to determine whether the 4 mAbs recognize the same or different epitopes (Fig. 5). Each unlabeled mAb was able to block its biotinylated counterpart serving as a positive control. An unrelated IgG1 antibody (I56) did not block any of the mAbs to PSMA. J591, J533, and E99 were each able to block each other, but were not blocked by J415. Conversely, J415 was blocked only by its unlabeled counterpart but not by any of the other 3 mAbs. These results indicate that J591, J533, and E99 recognize the same epitope which is distinct and non-cross-reactive with the epitope recognized by J415.
This manuscript defines 4 new IgG mAbs which detect 2 distinct extracellular epitopes of PSMA (PSMAext1 and PSMAext2). The reactivity of these mAbs with PSMA has been defined by immunoprecipitation and immunoblotting studies and reactivity against cell lines (data not shown) and tissue sections using the 7E11 mAb as a reference. Immunoprecipitation and immunoblotting studies demonstrate identical reactivity to that seen with 7E11. Reactivity in vitro (data not shown) and on tissue sections of normal and neoplastic specimens demonstrates nearly identical results. The exceptions in immunohistochemical reactivity were limited to striated muscle and tumor vascular endothelium. Striated muscle is reactive with 7E11 but not with mAbs J591, J415, J533 or E99. 7E11 reactivity with striated muscle had been previously reported by Lopes et al (4) who, like the present study, utilized frozen sections but reported as negative by Silver et al (17) who studied paraffin sections. This discrepancy is most likely explained by some loss of 7E11/PSMA immunoreactivity in the fixation/embedding process. The difference in reactivity of 7E11 and the present mAbs to striated muscle, both herein studied on frozen sections, may represent differences in post-translational processing of PSMA (the external domain of which is heavily glycosylated) occurring in prostate as compared to muscle.
Reactivity of 7E11 with tumor but not normal vascular endothelium also was previously noted by Silver et al (17), although 7E11 reactivity was reported in only half of their renal and urothelial cancers (15 of 30) and 3 of 19 colon carcinomas. In the present study, mAbs J591, J415, J533 and E99 demonstrate reactivity with tumor vasculature in all 23 non-prostate carcinomas tested. Some of the increased reactivity seen herein may represent the benefit of studying frozen as compared to paraffin sections. Within this study, when comparing mAbs J591, J415, J533 and E99 to 7E11 using a constant tissue preparation (frozen sections), we found stronger reactivity with mAbs J591, J415, J533 and E99 than with 7E11. The most likely explanation for the generally stronger reactivity seen with these new mAbs is that they were selected for, amongst other features, strong immunohistochemical reactivity. We have not studied the immunohistochemical reactivity of mAbs J591, J415, J533 and E99 on paraffin sections.
The initial study with 7E11 (1) indicated reactivity to fixed, but not viable, LNCaP cells later explained by epitope mapping studies indicating the 7E11 epitope to be intracellular (12). A more recent study by Troyer et al (13) studying ultrathin sections by IEM demonstrated 7E11 reactivity on the cytoplasmic aspect of LNCaP cells plasma membrane. Troyer et al also confirmed 7E11 reactivity with permeabilized but not with non-permeabilized LNCaP cells (13). Our studies comparing 7E11 with the present mAbs by immunoflourescence assays on viable and fixed, permeabilized and non-permeabilized LNCaP cells confirmed the previously published data that 7E11 detects an intracellular epitope not available for mAb binding unless the cell membrane is disrupted. A recent report by Barren et al (18) represents the sole study indicating that 7E11 can react with viable LNCaP cells. The report by Barren et al is inconsistent with other published work (1,12,13), as well as the results reported here, and may be due to a technical point. Barren et al , after incubating 7E11 with viable LNCaP cells, harvested LNCaP for flow cytometry by scraping the cells in the presence of 7E11. As scraping can disrupt cell membranes, this would have provided 7E11 access to its intracellular epitope likely accounting for the reactivity reported. Importantly, mAbs J591, J415, J533 and E99, unlike 7E11, can bind to either viable or non-permeabilized cells consistent with targeting accessible epitopes on the extracellular domain of PSMA. Our IEM finding of mAb J591 localization on the extracellular aspect of the plasma membrane (Fig. 4A), in contrast to the intracellular localization of 7E11 on IEM reported by Troyer et al (13), provides further evidence of reactivity of the present mAbs to the extracellular domain of PSMA.
Epitope mapping of the 4 present mAbs demonstrates that J591, J533 and E99 each bind to a single epitope (PSMAext1), while J415 binds to a different, non-competing site (PSMAext2). This will allow development of a "sandwich" assay to determine the presence and measure the level of PSMA in serum -- an area of some current controversy (14,16).
By allowing study of viable cells, these mAbs will be useful for studies of PSMA function and prostate cancer cell biology. Recent work indicates that PSMA has glutaminase (19,20) activity. Studies are underway to determine whether mAbs to PSMAext1 and/or PSMAext2 can block this enzymatic activity and, if so, the effect of such blockade on normal and neoplastic prostate physiology.
Given prior understanding of PSMA specificity and expression and the established ability of 7E11/CYT-356 to localize in vivo to a substantially less available epitope, one would anticipate the likelihood that these new mAbs might demonstrate significantly improved in vivo targeting for imaging and therapy. The immunoreactivity of these mAbs to vascular endothelium of a wide variety of cancers may significantly broaden their in vivo utility.
1This work was supported by grants from CaP CURE, the David H. Koch Charitable Foundation, The Ronald P. and Susan E. Lynch Foundation, the Lawrence and Carol Zicklin Philanthropic Fund, the Willam T. Morris Foundation, the John E. Wilson Research Fund, the Alissa Beth Bander Memorial Foundation, the Cornell Medical College Urological Oncology Research Fund and from BZL, Inc. NHB is a consultant to BZL, Inc. NHBs agreement with BZL is managed by Cornell University in accordance with its conflict of interest policies.
2To whom requests for reprints should be addressed, at New York Hospital-Cornell Medical Center, Box 23, 525 East 68th St., NY, NY 10021. e-mail: firstname.lastname@example.org
3Abbreviations: BSA, bovine serum albumin; FITC, fluorescein isothiocyanate; IEM, immunoelectron microscopy; IF, immunofluorescence; mAb, monoclonal antibody; PAP, prostatic acid phosphatase; PBS, phosphate buffered saline, ph 7.4; PCa, prostate cancer; PSA, prostate specific antigen; PSMA, prostate specific membrane antigen; PSP, prostate secretory protein; RT, room temperature; TBST (Tris-buffered saline-Tween 20).
1. Horoszewicz, J.S., Kawinski, E. and Murphy G.P.: Monoclonal antibodies to a new antigenic marker in epithelial cells and serum of prostatic cancer patients. Anticancer Res, 7: 927-936, 1987.
2. Israeli, R.S., Powell, C.T., Fair, W.R. and Heston, W.D.W.: Molecular cloning of a complementary DNA encoding a prostate-specific membrane antigen. Cancer Res., 53: 227-230, 1993.
3. Rinker-Schaeffer, C.W., Hawkins, A.L., Su, S.L., Israeli, R.S., Griffin, C.A., Isaacs, J.T. and Heston, W.D.W.: Localization and physical mapping of the prostate-specific membrane antigen (PSM) gene to human chromosome 11. Genomics, 30:105-108, 1995.
4. Lopes, D., Davis, Wendy L., Rosenstraus, M.J., Uveges, A.J. and Gilman, S.C.: Immunohistochemical and pharmacokinetic characterization of the site-specific immunoconjugate CYT-356 derived from antiprostate monoclonal antibody 7E11-C5. Cancer Res., 50: 6423-6429, 1990.
5. Israeli, R.S., Powell, C.T., Corr, J.G., Fair, W.R. and Heston, W.D.W.: Expression of the prostate-specific membrane antigen. Cancer Res., 54: 1807-1811, 1994.
6. Wright, G.L. Jr., Haley, C., Beckett, M.L. and Schellhammer, P.F.: Expression of prostate-specific membrane antigen in normal, benign, and malignant prostate tissues. Urol. Oncol., 1: 18-28, 1995.
7. Diamond, S.M., Fair, W.R. and Heston, W.D.W.: Modulation of prostate specific membrane antigen (PSM) expression in vitro by cytokines and growth factors. Proc. Am. Assoc. Cancer Res., 36: 643 (abstract 3826), 1995.
8. Wynant, G.E., Murphy, G.P., Horoszewicz, J.S. Neal, C.E., Collier, B.D., Mitchell, E., Purnell, G., Tyson, I., Heal, A., Abdel-Nabi, H. and Winzelberg, G.: Immunoscintigraphy of prostatic cancer: preliminary results with 111In-labeled monoclonal antibody 7E11-C5.3 (CYT-356). Prostate, 18: 229-241, 1991.
9. Murphy, G.P.: Radioscintiscanning of prostate cancer. Cancer (Suppl), 75: 1819-1833, 1995.
10. Babaian, R.J., Sayer, J., Podoloff, D.A., Steelhammer, L.C., Bhadkamkar, V.A. and Gulfo, J.V.: Radioimmunoscintigraphy of pelvic lymph nodes with 111indium -labeled monoclonal antibody CYT-356. J. Urol., 152: 1952-1955, 1994.
11. Kahn, D., Williams, R.D., Seldin, D.W., Libertino, J.A., Hirschhorn, M., Dreicer, R., Weiner, G.J., Bushnell, D. and Gulfo, J.: Radioimmunoscintigraphy with 111indium-labeled CYT-356 for the detection of occult prostate cancer recurrence. J. Urol., 152: 1490-1495, 1994.
12. Troyer, J.K., Feng, Q., Beckett, M.L. and Wright, G.L., Jr.: Biochemical characterization and mapping of the 7E11-C5.3 epitope of the prostate-specific membrane antigen. Urol. Oncol., 1: 29-37, 1995.
13. Troyer, J.K., Beckett, M.L. and Wright, G.L., Jr.: Location of prostate-specific membrane antigen in the LNCaP prostate carcinoma cell line. Prostate, 30:232-242, 1997.
14. Rochon, Y.P., Horoszewicz, J.S., Boynton, A.L., Holmes, E.H., Barren, R.J. III, Erickson, S.J., Kenny, G.M. and Murphy, G.P.: Western Blot assay for prostate-specific membrane antigen in serum of prostate cancer patients. Prostate 25: 219-223, 1994.
15. Rajasekaran, A.K., Hojo, M., Huima, T. And Rodriguez-Boulon, E.: Catenins and Zonula Occludens-1 form a complex during early stages in the assembly of tight junctions. J. Cell Biol., 132:451-463, 1996.
16. Troyer, J.K., Beckett, M.L. and Wright, G.L., Jr.: Detection and characterization of the prostate-specific membrane antigen (PSMA) in tissue extracts and body fluids. Int. J. Cancer, 62: 552-558, 1995.
17. Silver, D.A., Pellicer, I., Fair, W.R., Heston, W.D.W. and Cordon-Cardo, C.: Prostate-specific membrane antigen expression in normal and malignant human tissues. Clin. Cancer Res., 3: 81-85, 1997.
18. Barren, R.J. III, Holmes, E.H., Boynton, A.L., Misrock, S.L. and Murphy, G.P.: Monoclonal antibody 7E11.C5 staining of viable LNCaP cells. Prostate, 30:65-68, 1997.
19. Pinto, J.T., Suffoletto, B.P., Berzin, T.M., Qiao, C.H., Lin, S., Tong, W.P., May, F., Mukherjee, B. And Heston, W.D.W.: Prostate-specific membrane antigen: A novel folate hydrolase in human prostatic carcinoma cells. Clin. Cancer Res., 2:1445-1451, 1996.
20. Carter, R.E., Feldman, A.R. and Coyle, J.T.: Prostate-specific membrane antigen is a hydrolase with substrate and pharmacologic characteristics of a neuropeptidase. Proc. Natl. Acad. Sci., U.S.A., 93:749-753, 1996.