Elke Schultz-Thater

Cancer/testis tumour-associated antigens: immunohistochemical detection with monoclonal antibodies

Cancer/testis tumour-associated antigens (C/T TAA) were the first human tumour-associated antigens to be characterised at the molecular level. Specific genes are expressed in the testis and in tumours of varying histological origin. The tissue expression pattern supports the notion that these antigens could be targets for active specific immunotherapy. Specific serological reagents have been developed and have helped to clarify biochemical characteristics of C/T TAA and to assess their distribution within clinical tumour samples. We review immunohistochemical evidence of the expression of C/T TAA known to be recognised by specific cytotoxic T lymphocytes. The emerging picture is consistent with a mostly heterogeneous expression in human cancers. These findings support the concept of multiantigenic tumour vaccine preparations. Moreover, the wide range of tumours in which C/T TAA have been detected urges further efforts to develop effective specific immunotherapeutic procedures.

Phase I/II clinical trial of a nonreplicative vaccinia virus expressing multiple HLA-A0201-restricted tumor-associated epitopes and costimulatory molecules in metastatic melanoma patients.

We performed a phase I/II clinical trial in metastatic melanoma patients with an ultraviolet (UV)-inactivated nonreplicating recombinant vaccinia virus enabling the expression, from a single construct, of endoplasmic reticulum-targeted HLA-A0201-restricted Melan-A/MART-1(27-35), gp100(280-288), and tyrosinase(1-9) epitopes, together with CD80 and CD86 costimulatory proteins. Corresponding soluble peptides were used to boost responses and granulocyte-macrophage colony-stimulating factor was used as systemic adjuvant. Safety and immunogenicity, as monitored with in vitro-restimulated peripheral blood mononuclear cells by cytotoxic T lymphocyte precursor (CTLp) frequency analysis and tetramer staining, were specifically addressed. Of 20 patients entering the protocol, 2 had to withdraw because of rapidly progressing disease. Immune responses were evaluated in 18 patients (stage III, n = 5; stage IV, n = 13) and increases in specific CTLp frequencies were observed in 15. In 16 patients responsiveness against all 3 antigens could be analyzed: 7 (43%), including all stage III cases, showed evidence of induction of CTLs specific for the three epitopes, and 2 (12%) and 4 (25%), respectively, showed reactivity against two or one tumor-associated antigen. In three stage IV patients no specific CTL reactivity could be induced. Increases in CTLp frequency were detected mostly after viral vaccine injections. However, in a majority of patients final CTLp levels were comparable to initial levels. Tetramer characterization of Melan-A/MART-1(27-35)-specific CTLs during the protocol also suggested preferential expansion after recombinant virus administration. Vector-specific humoral responses, frequently undetectable in stage IV patients, did not appear to prevent tumor-associated antigen-specific CTL induction. Aside from a single occurrence of transient grade 3 leukopenia, no major clinical toxicity was reported. Seventeen of 18 patients completed the 3-month trial (one patient died before the last delayed-type hypersensitivity test). Three displayed regression of individual metastases, seven had stable disease, and progressive disease was observed in seven patients. This is the first report on the administration of a UV-inactivated recombinant vaccinia virus coexpressing five transgenes in cancer patients. The results described here, in terms of safety and immunogenicity, support the use of this reagent in active specific immunotherapy.

Anti‐MAGE‐3 antibody 57b and anti‐MAGE‐1 antibody 6C1 can be used to study different proteins of the MAGE‐A family

MAGE-A genes are expressed by a variety of cancers but not by normal adult tissues, except for testis and placenta. They encode 12 highly homologous proteins of 309-369 aminoacids (De Plaenet al., 1994). Several peptide epitopes from MAGE-A proteins, including MAGE-1, 3, 6 and 10, recognized by specific cytolytic T lymphocytes derived from tumors or peripheral blood have been described (Van den Eynde and van der Bruggen, 1997; Huanget al., 1999; Zorn and Hercend, 1999). In addition, HLA class II restricted epitopes have been described for MAGE-3 (Chaux et al., 1999; Manici et al., 1999). MAGE-A genes are therefore regarded as attractive candidates for specific immunotherapy and clinical trials of cancer patients involving immunization with MAGE-A1 and 3 peptides are in progress (Marchandet al., 1999; Nestleet al., 1998). Monoclonal antibodies (MAbs) against MAGE-1, 3, 4 and 11 proteins have been developed using the respective recombinant E. coli proteins as immunogen (Chent al., 1994; Kocheret al., 1995; Shichijoet al., 1995; Carrel et al., 1996; Jurket al., 1998). The anti-MAGE-3 MAb 57B appeared to be particularly interesting as it was shown to be suitable for immunocytochemistry for both frozen and paraffin-embedded tumor samples (Hofbauer et al., 1997). However, the possibility that crossreactivity of this antibody with other members of the MAGE-A family may exist, due to the high degree of homology of these proteins (MAGE-3 and 6, in particular, differ only for 11 aminoacids), has not been formally ruled out. To this regard, we have demonstrated that a MAb raised against recombinant MAGE-1, named 6C1, strongly cross-reacts with MAGE-10 (Carrel et al., 1996; Rimoldiet al.,1999). To resolve this issue, we have investigated in detail the cross-reactivities of anti-MAGE-3 MAb 57B and anti-MAGE-1 MAb 6C1 using a transient transfection approach where human embryonic kidney 293T cells were used as recipient. Plasmids containing cDNAs encoding the more commonly expressedMAGE genes ( MAGE-1, 2, 3, 4, 6, 10and 12) were transfected individually and cell lysates analyzed by Western blotting. Figure 1a shows that both MAbs were able to recognize all the different MAGE proteins, except for MAGE-10 that was not recognized by MAb 57B. These results demonstrate the high degree of cross-reactivity of both MAbs and also show that all MAGE-A proteins have an apparent m.w. of 45–50 kDa, except for MAGE-10, which displays an unexpected MW of 72 kDa (Rimoldiet al., 1999). The extent of cross-reactivity towards the different proteins appeared to be variable. However, quantitative interpretations cannot be drawn from these experiments due to possible variations in transfection efficiencies. To confirm that cross-reactivity by the MAbs can take place whenMAGE genes are expressed at a normal physio logical level, such as found in tumors, a series of cell lines xpressing different combinations of MAGE genes, as determined by RT/PCR, was analyzed. Figure 1b confirms the ability of the 2 MAbs to recognize multiple members of the MAGE family. In particular, MAb 57B detected a band of approximately 50 kDa (similar to the size of MAGE-3, Kocher et al., 1995) in melanoma Me242 and Me244 cells (expressing only MAGE-2/6 and 1/2/6/10, respectively) and fibrosarcoma LB23-SAR cells (expressing only MAGE-4). In contrast, the 45 kDa/MAGE-A1 band was not detected by MAb 57B under these conditions. Interestingly, the pattern of proteins detected by MAb 6C1 closely resembled that of MAb 57B, with the addition of the 72 kDa/MAGE-10 and the 45 kDa/MAGE-1 bands. The ability of the 2 MAbs to detect multiple MAGE proteins was further assessed by immunocytochemistry on transiently transfected 293T and MAGE negative melanoma cells. The results paralleled those described above for the Western blot analyses (not shown). A summary of the cross-reactivities of MAbs 57B and 6C1 is given in Table I. Finally, to corroborate our findings, we sought to characterize the epitopes recognized by the MAbs. By screening several peptide sequences from MAGE-1 and 3 proteins in Elisa tests, we have identified 2 overlapping peptides from MAGE-1 (amino-acid 250-267 and 257-273) and 1 peptide from MAGE-3 (amino-acid 14-31) strongly recognized by MAb 6C1 and 57B, respectively, and capable of blocking the MAb activity against the respective recombinant protein (not shown). Figure 2 shows the MAGE-1 and 3 sequences containing the MAb epitopes aligned with the other MAGE-A sequences. In the case of MAb 6C1, the sequence recognized is identical for MAGE-1 and 10, confirming the high degree of cross-reactivity previously observed (Rimoldi et al., 1999), while the homologous MAGE-2, 3, 4, 6 and 12 sequences differ for one amino-acid. Elisa tests with recombinant MAGE-1 and 3 proteins indicated that the single aminoacid mismatch resulted in approximately 100-fold difference in affinity of 6C1

NY-ESO-1 tumour associated antigen is a cytoplasmic protein detectable by specific monoclonal antibodies in cell lines and clinical specimens

NY-ESO-1 gene encodes a novel member of the cancer/testis (CT) family of human tumour-associated antigens (TAA). Specific monoclonal antibodies (mAb) have identified the corresponding gene product in lysates of tumour cell lines as a 22 kDa protein but no data are available concerning its intracellular location or distribution within neoplastic tissues. We have generated NY-ESO-1 specific mAbs recognizing the target molecule in cytospin preparations and in sections from clinical tumour specimens. These reagents identify NY-ESO-1 TAA in melanoma cell lines expressing the specific gene as a cytoplasmic protein, sharing the intracellular location of most MAGE TAA. In a series of 12 melanoma specimens, specific staining, limited to neoplastic cells, was detectable in the five cases where NY-ESO-1 gene expression was observed. In two of them over 90% of tumour cells showed evidence of positive staining. Lower percentages of positive neoplastic cells ranging between single cells and 50% were observed in the remaining tumours. These data suggest that active specific immunotherapies targeting NY-ESO-1, alone or in combination with other TAA could be of high clinical relevance in sizeable subgroups of melanoma patients.

Prognostic relevance of MAGE-A4 tumor antigen expression in transitional cell carcinoma of the urinary bladder: A tissue microarray study

TAAs of the MAGE family are mostly studied as targets of specific immune responses. Their potential relevance as tumor markers has also been underlined. We used a MAb, 57B, recognizing MAGE‐A4 protein in paraffin‐embedded sections, to evaluate its expression in bladder cancers by employing TMA including 2,317 samples from 1,849 patients. In 2,090/2,317 cases (90.2%), immunostaining yielded interpretable results. Since for some patients more than 1 sample was available, only interpretable first biopsies (n = 1,628) were considered. MAGE‐A4 protein was expressed at significantly (p < 0.001) higher frequency in squamous (25/55, 45.5%) than in adeno (4/15, 26.7%), sarcomatoid (4/14, 28.6%), small cell (5/20, 25%) or transitional cell (281/1,522, 18.5%) carcinomas. In TCCs, overall MAGE‐A4 positivity was significantly correlated with invasive phenotype (p < 0.001) and high tumor grade (p < 0.0001). Clinical data from 908 TCC patients were retrospectively evaluated, revealing that strong 57B staining was highly significantly associated with decreased tumor‐specific survival (p < 0.0001). These data suggest that evaluation of MAGE‐A4 protein expression is useful in the identification of groups of TCCs characterized by severe prognosis, thus possibly providing indications for early MAGE TAA‐targeted immunotherapy.

NY-ESO-1/LAGE-1 coexpression with MAGE-A cancer/testis antigens: a tissue microarray study.

The characterization of the expression pattern of different families of cancer/testis (C/T) antigens in different tumors, at the protein level, might be of relevance in the development of multiantigen vaccine preparations for active specific immunotherapy. We have used tissue microarray (TMA) technology to explore in large numbers of tumor specimens the expression of NY‐ESO‐1/LAGE‐1 C/T antigens and its correlation with MAGE‐A expression by using D8.38 and 57B monoclonal antibodies (MAb). The epitopes recognized by these reagents in C/T antigens were identified by molecular mapping by using a bacterial expression system. Out of 2,052 samples, 119 (5.8%) scored positive upon staining with D8.38 NY‐ESO‐1/LAGE‐1‐specific MAb. Expression in >10% of cases was detectable in melanoma and basalioma (31.6 and 18.2%, respectively), large cell carcinomas and adenocarcinomas of the lung (17.8 and 10.5%, respectively), stomach adenocarcinomas of the intestinal type (13.2%), pT2‐4 bladder TCC (18.2%), nonseminomatous carcinomas of the testis (10.4%) and liposarcomas (15.4%). Simultaneous expression of NY‐ESO‐1/LAGE‐1 and MAGE‐A C/T antigens was then addressed in a TMA where 101/845 and 73/845 samples (12 and 8.6%, respectively) showed evidence of MAGE‐A or NY‐ESO‐1/LAGE‐1 specific staining, respectively. In 35/845 specimens (4.1%) concomitant expression of MAGE‐A and NY‐ESO‐1/LAGE‐1 was observed (p = 0.0002). Discrepancies in the expression of NY‐ESO‐1/LAGE‐1 and MAGE‐A were conspicuously detectable in squamous cell carcinomas of the skin (MAGE‐A positive but NY‐ESO‐1/LAGE‐1 negative) and in liposarcomas (NY‐ESO‐1/LAGE‐1 positive, but MAGE‐A negative). Taken together, these data suggest novel areas of application of C/T antigens targeted active specific immunotherapy possibly based on multiantigen vaccine preparations.

The role of 3D structure and protein conformation on the innate and adaptive immune responses to silk-based biomaterials

We have investigated monocyte and T cell responsiveness to silk based biomaterials of different physico-chemical characteristics. Here we report that untransformed CD14+ human monocytes respond to overnight exposure to silk fibroin-based biomaterials in tridimensional form by IL-1β and IL-6, but not IL-10 gene expression and protein production. In contrast, fibroin based materials in bidimensional form are unable to stimulate monocyte responsiveness. The elicitation of these effects critically requires contact between biomaterials and responding cells, is not sustained and becomes undetectable in longer term cultures. We also observed that NF-κβ and p38 MAP kinase play key roles in monocyte activation by silk-based biomaterials. On the other hand, fibroin based materials, irrespective of their physico-chemical characteristics appeared to be unable to induce the activation of peripheral blood T cells from healthy donors, as evaluated by the expression of activation markers and IFN-γ gene.

Expression of cancer/testis tumor associated antigens in cervical squamous cell carcinoma.

We investigated the expression of tumor-associated antigens (TAA) of the cancer/testis (C/T) gene family in cervical squamous cell carcinomas. First, we focused on the HeLa cervical cancer derived cell line, and we found that it expresses MAGE-A1, MAGE-A2, MAGE-A3, MAGE-A4, MAGE-A6, MAGE-A12, GAGE-3/6, LAGE-1, and PRAME genes, encoding defined C/T TAA. In contrast, no expression of MAGE-A10, BAGE, GAGE-1/2, or NY-ESO-1 genes was observed. Corresponding gene products could also be detected by immunoblotting and immunocytochemistry, taking advantage of monoclonal antibodies recognizing discrete TAA. Capitalizing on these data, a monoclonal antibody predominantly recognizing MAGE-A4 TAA in paraffin-embedded sections (57B) was used to investigate the C/T gene expression in clinical tumor samples. A group of 60 patients was studied, and 57B positivity was detectable to different extents in 33% of the cases (20/60). In 13 of them (21%), staining of over 50% of the tumor cells was evident, whereas healthy cells always scored negative. Remarkably, MAGE-A4 expression was significantly (p < 0.05) more frequently detectable in poorly differentiated tumors (8/13) than in well-differentiated or moderately differentiated cancers (3/15 and 9/32, respectively) and in stage FIGO II as compared with stage FIGO Ib tumors (12/23 and 5/24, respectively, p = 0.04). Interestingly, staining was mostly nuclear in well-differentiated tumors, but involved both nuclei and cytoplasm in less differentiated cancers. Positivities of comparable frequency were also detectable in a smaller series of specimens upon staining with MAGE-A1- or NY-ESO-1/LAGE-1-specific reagents. Considering the high tumor specificity of C/T TAA, our data provide the rationale for the design of immunotherapy procedures targeting these antigens in cervical cancers.

MAGE‐A10 is a nuclear protein frequently expressed in high percentages of tumor cells in lung, skin and urothelial malignancies

MAGE‐A10 is a highly immunogenic member of the MAGE‐A family of cancer/testis tumor‐associated antigens (C/T TAAs). Studies performed with broadly reactive antibodies have helped to initially characterize this TAA. However, no specific reagents have been developed so far, thus preventing a thorough analysis of its expression in healthy and tumoral tissues. We have produced MAGE‐A10 gene product in soluble recombinant form, and we have used it to generate specific monoclonal antibodies (mAbs). One of these reagents, recognizing an epitope located at the COOH terminus of the MAGE‐A10 gene product, was used to stain a multitumor tissue microarray comprising more than 2,500 paraffin‐embedded specimens including healthy tissues, benign tumors and malignancies of different histological origin. MAGE‐A10 protein was identified as an intranuclear protein of an apparent molecular weight of 70 kDa, expressed in normal spermatogonia and spermatocytes but in no other healthy tissue. Most importantly, this C/T TAA appears to be expressed in high (>50%) percentages of cancer cells from a number of malignancies, including lung, skin and urothelial tumors. Unexpectedly, high expression of MAGE‐A10 TAA at the protein level was also detectable in gynecological malignancies and stomach and gall bladder cancers. The characterization of MAGE‐A10‐specific reagents might set the stage for the development of targeted active immunotherapy by clarifying potential indications and by allowing the selection of patients eligible for treatment and the monitoring of its effectiveness.

MAGE‐A10 cancer/testis antigen is highly expressed in high‐grade non‐muscle‐invasive bladder carcinomas

Bladder cancer is a common urinary malignancy and a prevalent cause of cancer‐related death. Current therapies of early stage non‐muscle‐invasive bladder cancer (NMIBC) are frequently associated with undesirable toxicities and recurrence. Active antigen‐specific immunotherapy may provide a valid therapeutic option for patients with NMIBC. Cancer‐testis antigens (CTA) expressed in various tumour types and in a limited range of healthy tissues may represent potential targets for specific immunotherapy. MAGE‐A10 is probably the most immunogenic antigen of the MAGE‐A family. We evaluated the expression of MAGE‐A10 in NMIBC. Seventy‐nine patients undergoing surgical treatment for NMIBC were enrolled in the study. MAGE‐A10 gene expression was assessed by quantitative real‐time polymerase chain reaction. Immunohistochemistry was performed on paraffin‐embedded sections. MAGE‐A10 gene was specifically expressed in one‐third of NMIBC (n = 24: 32.43%). Gene expression was correlated with high tumour grade. MAGE‐A10 protein was exclusively detectable in nuclei of tumour cells. More importantly, MAGE‐A10 protein was also more frequently detectable in high‐grade tumours (p = 0.0001) and in stage T1 tumours invading subepithelial tissue or lamina propria (p = 0.01). A strong correlation between MAGE‐A10 staining score and tumour grade and stage could accordingly be observed. These data indicate that MAGE‐A10 expression is a feature of aggressive NMIBC and might be used as a novel target for specific immunotherapy of these cancers.