A peptide encoded by human gene MAGE-3 and presented by HLA-A2 induces cytolytic T lymphocytes that recognize tumor cells expressing MAGE-3

HLA-A locus-restricted and tumor-specific CTLs in tumor-infiltrating lymphocytes of patients with non-small cell lung cancer

HLA class I restriction and tumor specificity of cytotoxicity in the IL-2-activated tumor-infiltrating lymphocytes from 16 patients with non-small cell lung cancer were investigated. Six HLA class I-restricted and tumor-specific CTL lines were established: (i) HLA A2-restricted and adenocarcinoma-specific CTLs in three (two A0201+ and one A0206+) patients with adenocarcinoma, (ii) HLA A3101- and A3302-restricted and adenocarcinoma-specific CTLs in an HLA A3101/3302+ patient with adenocarcinoma, and (iii) HLA A3302-restricted CTLs and (iv) HLA A2402-restricted CTLs recognizing tumors with different types of histology in an HLA A3302+ patient with adenocarcinoma and an HLA A2402+ patient with squamous cell carcinoma (SCC), respectively. The three HLA A2-restricted CTL lines recognized 4, 4, or 6 of 15 HLA A2+ adenocarcinoma cell lines that originated from lung, stomach, colon, and breast with different subtypes (HLA A0201, A0206, and A0207), respectively. Furthermore, the CTLs of an HLA A0206+ patient recognized five different fractions of peptides eluted from an HLA A0201+ adenocarcinoma cell line. These results showed evidence of the existence of HLA class I-restricted and tumor-specific CTLs recognizing peptide antigens on HLA-A alleles of adenocarcinoma or SCC in tumor sites of a substantial number of patients with non-small cell lung cancer.

Characterization of an antigen that is recognized on a melanoma showing partial HLA loss by CTL expressing an NK inhibitory receptor

Melanoma lines MEL.A and MEL.B were derived from metastases removed from patient LB33 in 1988 and 1993, respectively. The MEL.A cells express several antigens recognized by autologous cytolytic T lymphocytes (CTL) on HLA class I molecules. The MEL.B cells have lost expression of all class I molecules except for HLA-A24. By stimulating autologous lymphocytes with MEL.B, we obtained an HLA-A24-restricted CTL clone that lysed these cells. A novel gene, PRAME, encodes the antigen. It is expressed in a large proportion of tumors and also in some normal tissues, albeit at a lower level. Surprisingly, the CTL failed to lyse MEL.A, even though these cells expressed the gene PRAME. The CTL expresses an NK inhibitory receptor that inhibits its lytic activity upon interaction with HLA-Cw7 molecules, which are present on MEL.A cells and not on MEL.B. Such CTL, active against tumor cells showing partial HLA loss, may constitute an intermediate line of anti-tumor defense between the CTL, which recognize highly specific tumor antigens, and the NK cells, which recognize HLA loss variants.

Detection of MAGE-4 protein in sera of patients with head-and-neck squamous-cell carcinoma

The MAGE-4 gene, a member of the MAGE gene family, is expressed in various cancers, including head-and-neck squamous-cell carcinomas (HN-SCC), but is not expressed in any normal tissues except for the testis and placenta. The aim of this study was to determine whether serum MAGE-4 protein is a useful tumor marker for detection of HN-SCC. An enzyme-linked immunosorbent assay was used to measure serum level of MAGE-4 protein. The serum level of MAGE-4 in pre-operative HN-SCC patients was significantly higher than that in patients with non-malignant diseases (NMD) of the head and neck, volunteers undergoing cancer screening (VOL), or healthy donors (HD). When the cut-off level was determined at 1.15 ng/ml (mean plus 3 SD of HD), sera from 28 of 96 patients with HN-SCC (p < 0.0001 vs. the other groups), 7 of 82 patients with NMD, 2 of 92 with VOL, and 0 of 68 HD were positive for MAGE-4. Serum levels of MAGE-4 protein in all 7 HN-SCC patients whose sera were positive for MAGE-4 before operation decreased after operation, and, in one patient, a renewed rise in serum level was followed by recurrence. These results indicate that MAGE-4 protein is detectable in sera of a significant number of HN-SCC patients, and that serum MAGE-4 protein might be a useful tumor marker to monitor the recurrence of MAGE-4-positive HN-SCC.

Antigen organization regulates cluster formation and induction of cytotoxic T lymphocytes by helper T cell subsets

Generation of effector cytotoxic T lymphocytes (CTLs) is a process tightly governed by regulatory helper T (Th) cells. The nature of cellular interactions as well as the precise role of distinct Th cell subsets involved in efficient CTL activation remains elusive. Employing in vitro cultures for primary induction of human, peptide-specific CTL, a strict requirement for Th cells and linkage of epitopes for helper and CTLs on the surface of antigen presenting cells was found, suggesting a three cell type cluster as minimal immune regulatory entity. Cognate and antigen-driven interactions of T cells were neither essential nor sufficient to override the need for linked epitopes. Within the three cell type cluster complex, keyhole limpit hemocyanin or tetanus toxoid-reactive Th cells promoted generation of MAGE-3- or HIV-gag-specific CTL. Both type 1 and type 2 Th cells were recruited and induced by CTL. Interleukin 2 and interferon gamma were essential in early stages, and interleukin 4 was utilized in later stages, of CTL maturation. Synergistic effects of CD45RA+ and CD45RO+ Th cells were found. The data reported here suggest a critical link between the innate and adaptive immune system in the initiation process of cytolytic immune responses and offers the basis for efficient vaccine strategies.

Interleukin-2 gene transduction into freshly isolated lung adenocarcinoma cells with adenoviral vectors

We evaluated the efficiency of gene transduction and of gene expression by adenoviral vectors in human lung adenocarcinoma cells. Freshly isolated cancer cells were collected from pleural effusions in adenocarcinoma patients by centrifugation with a Percoll gradient. Adenoviral vectors resulted in effective gene transduction into human lung cancer cell lines and into freshly isolated lung adenocarcinoma cells. In an experiment using the beta-galactosidase (LacZ) gene, the Adex1CA vector with a regulatory sequence of chicken beta-actin as promoter and an enhancer derived from cytomegalovirus produced a higher transduction ratio and greater expression levels than adenoviral vectors with other promoter systems. Transduction with Adex1CA vectors containing the human interleukin-2 (IL-2) gene (Adex1CAhIL-2) resulted in enhanced secretion of IL-2 from gene-modified lung cancer cells. Treatment with normal human serum inhibited gene transduction by Adex1CAhIL-2 but did not inhibit gene expression after transduction by Adex1CAhIL-2. The secretion of IL-2 from the gene-modified cells, which were irradiated at 100 Gy before transduction, continued for 8 days. In a mouse model, the intrapleural injection of IL-2 gene-modified 3LL cells transduced by Adex1CAhIL-2 could cure the pre-existing lung tumours with malignant pleural effusions to induce tumor-specific immunity. But these therapies did not show any therapeutic benefit on the pre-existing tumor in subcutaneous region. These data suggest a potentially useful but limited clinical role of Adex1CAhIL-2 in gene therapy for lung cancer patients.

Immunobiology of human melanoma antigens MART-1 and gp100 and their use for immuno-gene therapy

Two genes encoding human melanoma antigens MART-1 and gp100 recognized by HLA-A2 restricted melanoma reactive CTL derived from tumor infiltrating lymphocytes (TIL) were isolated by cDNA expression cloning methods. Multiple unmutated self peptides were identified as T cell epitopes in these melanocyte/melanoma specific proteins (2 from MART-1 and 5 from gp100). Most of these melanoma epitopes contain non-dominant anchor amino acids at the primary anchor positions and have intermediate binding affinity to HLA-A2.1. Melanoma reactive CTL were efficiently induced from PBL and TIL of patients by in vitro stimulation with PBMC pulsed with these epitopes. There is a significant correlation between vitiligo development and clinical response to IL2 based immunotherapy, suggesting that autoreactive T cells are involved in melanoma regression in vivo. These results have implications for understanding the nature of tumor antigens recognized by T cells and for the development of new cancer immunotherapies.

Expression of MAGE genes in osteosarcoma

Expression of MAGE genes that encode tumor-rejection antigens recognized by cytotoxic T lymphocytes with major histocompatibility complex class-I antigens was investigated in human osteosarcomas (20 cell lines and eight fresh tumor tissues). MAGE-1, 2, 3, 4, and 6 genes were expressed at the mRNA level in 11 (52.4%), 10 (47.6%), 10 (47.6%) one (4.8%), and 10 (47.6%) of 21 tumor cell lines, respectively, and in five (62.5%), six (75%), five (62.5%), one (12.5%), and five (62.5%) of eight fresh tumor tissues as determined by the reverse transcription-polymerase chain reaction method. MAGE-1 or 4 protein was detected by immunoblot analysis in eight of 11 or one of one tumor cell lines, respectively, where it was expressed at the mRNA level. Major histocompatibility complex class-I antigens were expressed in 19 of 21 tumor cell lines. These results suggest that MAGE tumor-rejection antigens are expressed in substantial numbers of osteosarcomas in a major histocompatibility class-I-restricted manner.

Development of cancer vaccine by tumor rejection antigens

Identification of the MAGE genes allowed us the molecular approach to identify genes encoding tumor rejection antigens expressed on human cancer cells. MAGE-1 proteins are normal tissue antigens compartmentalized in the particular testicular cells playing an important role in the early phase of the spermatogenesis. The MAGE-1, -2, -3, -4 and -6 genes are preferentially expressed in many different cancers at both the mRNA and protein levels. The MAGE genes, particularly MAGE-1, became positive in relatively advanced stages of cancers and recurrent cancers. Approximately one-third to half of human cancers except for myelo-monocytic leukemia expressed at least one of these MAGE genes. The MAGE gene products shall be appropriate target molecules for development of new cancer vaccine.

Tumor antigens recognized by T lymphocytes

In the last five years, knowledge of human tumor antigens recognized by autologous cytolytic T lymphocytes (CTL) has increased considerably. So far, genetic and biochemical approaches have led to the molecular identification of three classes of antigens. Most of these antigens consist of peptides that are presented to T cells by HLA molecules. The first class comprises antigens encoded by genes such as MAGE, BAGE, and GAGE, which are expressed in various tumors of different histological origins, but not in normal tissues other than testis. The second class represents differentiation antigens encoded by genes that are only expressed in melanoma and normal melanocytes like tyrosinase, Melan-A/MART-1, gp100 and gp75. The third class includes antigens produced by unique point mutations in genes that are ubiquitously expressed. In most cases, the antigenic peptide is encoded by the mutated region of the gene. A number of these antigens provide promising targets for new protocols of specific cancer immunotherapy.

Characterization of antigenic peptides presented by HLA-B44 molecules on tumor cells expressing the gene MAGE-3

The amino acid sequence of the protein encoded by the gene MAGE-3 was screened for peptides containing the binding motif for HLA-B44. Nine peptides were synthesized, and their binding affinity for HLA-B*4402 and -B*4403 was analyzed in an HLA class I alpha-chain refolding assay. Four peptides with binding affinity for HLA-B*4403 were chosen for in vitro cytotoxic T-lymphocyte induction assays using as antigen-presenting cells peptide-pulsed, autologous activated B lymphoblasts from a healthy, B*4403+ donor. Peptide-specific effectors could be raised only against one peptide, M3-167. Cytotoxic T lymphocytes specific for this peptide were also able to recognize melanoma cell lines expressing HLA-B44 and the gene MAGE-3, strongly suggesting that M3-167 is a naturally processed MAGE-3-encoded epitope presented by HLA-B44. M3-167 is a I amino acid N-terminal extension of M3-168, a naturally processed epitope MAGE-3-encoded epitope presented by HLA-A1 that has been previously described. TAP binding studies of these 2 peptides revealed that the TAP affinity of M3-167 is about 9-fold higher than that of M3-168. M3-167 or a longer precursor could be transported into the endoplasmatic reticulum, where it could be trimmed for presentation by HLA-A1 or -B44 molecules. Taken together, our data suggest that M3-167 could be an immunodominant peptide encoded by the gene MAGE-3.

A peptide encoded by the human MAGE3 gene and presented by HLA-B44 induces cytolytic T lymphocytes that recognize tumor cells expressing MAGE3

The human MAGE3 gene is expressed in a significant proportion of tumors of various histological types, but is silent in normal adult tissues other than testis and placenta. Antigens encoded by MAGE3 may therefore be useful targets for specific antitumor immunization. Two antigenic peptides encoded by the MAGE3 gene have been reported previously. One is presented to cytolytic T lymphocytes (CTL) by HLA-A1, the other by HLA-A2 molecules. Here we show that MAGE3 also codes for a peptide that is presented to CTL by HLA-B44. MAGE3 peptides containing the HLA-B44 peptide binding motif were synthesized. Peptide MEVDPIGHLY, which showed the strongest binding to HLA-B44, was used to stimulate blood T lymphocytes from normal HLA-B44 donors. CTL clones were obtained that recognized not only HLA-B44 cells sensitized with the peptide, but also HLA-B44 tumor cell lines expressing MAGE3. The proportion of metastatic melanomas expressing the MAGE3/HLA-B44 antigen should amount to approximately 17% in the Caucasian population, since 24% of individuals carry the HLA-B44 allele and 76% of these tumors express MAGE3.

Tumor antigens and tumor vaccines: peptides as immunogens

Tumor antigens recognized by human cytotoxic T lymphocytes (CTL) have been identified for multiple types of solid tumors. These include both shared and unique antigens. Unique antigens are those expressed uniquely by one patient's tumor, and shared antigens are those present on tumor cells from many different patients. Many of the shared antigens are derived from tissue-specific differentiation antigens, oncogenes, or a set of antigens expressed only in tumors or in testis. In addition to advances in understanding tumor antigens that stimulate CTL and T-helper cell responses, there have been advances in understanding immunity in general, including the characterization of cytokines, the recognition of the dendritic cell as an optimal antigen-presenting cell (APC), and the characterization of costimulatory molecules as critical components of antigen presentation. Together, these developments have breathed new life into tumor immunology, and they promise to lead to a new generation of peptide- and cell-based tumor vaccines.

Tumor escape from immune recognition: lethal recurrent melanoma in a patient associated with downregulation of the peptide transporter protein TAP-1 and loss of expression of the immunodominant MART-1/Melan-A antigen

In the last few years, mutiple protein target antigens for immunorecognition by T cells have been identified on human melanoma. How melanoma lesions escape from functional antigen-specific immune recognition remains poorly understood. We have identified the concomitant loss of the immunodominant T cell-defined MART-1/Melan-A antigen and downregulation of the TAP-1 gene in a recurrent metastatic melanoma that was resected in 1993. This phenotype was not observed for an earlier autologous melanoma lesion resected in 1987. The "antigen loss" could be restored in the variant tumor cell line by simultaneously providing both the MART-1/Melan-A gene (by retroviral transfer) and the TAP-1 gene (by a bioballistic approach) resulting in tumor cell sensitivity to MART-1/Melan-A-specific cytotoxic T lymphocytes. This suggests that tumor escape from immune surveillance may have occurred in vivo as a sequential result of (a) antigen loss, and (b) downregulation of the peptide-transporter protein TAP-1 expression by this patient's tumor over a 6-yr period from 1987 to 1993. These results suggest that the characterization of the T cell response to melanoma in individual patients and definition of the immunologically relevant genetic defects in tumors may be required to select the most effective therapeutic strategies for a given patient.

A new gene coding for an antigen recognized by autologous cytolytic T lymphocytes on a human renal carcinoma

Previous reports have described antigens that are recognized on human melanoma cells by autologous cytolytic T lymphocytes (CTL). The genes coding for a number of these antigens have been identified. Here we report the cloning of a gene that codes for an antigen recognized by autologous CTL on a human renal carcinoma cell line. This antigen is presented by HLA-B7 and is encoded by a new gene that we have named RAGE1. No expression of RAGE1 was found in normal tissues other than retina. RAGE1 expression was found in only one of 57 renal cell carcinoma samples, and also in some sarcomas, infiltrating bladder carcinomas, and melanomas. This represents the first identification of an antigen recognized by autologous CTL on a renal tumor.

T-cell recognition of self peptides as tumor rejection antigens

Human melanoma antigens and their epitopes recognized by T cells have recently been identified. HLA-A2 binding epitopes of melanoma antigens MART-1 and gp 100 were characterized and suspected to be subdominant/cryptic self determinants. Together with other findings of tumor-specific mutated self peptides as tumor antigens recognized by T cells, the nature of the antitumor immune response to human melanoma has been revealed at a molecular level. These findings have implications not only for understanding of the immune response to self peptides in normal and pathologic conditions, but also for the development of immunotherapies for cancer and autoimmune diseases.

Expression of MAGE genes in human colorectal carcinoma

OBJECTIVE

The human genes MAGE-1 and -3 encode tumor-specific peptide antigens, which are recognized by autologous cytotoxic T lymphocytes. The antigens coded by those genes may be useful for cancer immunotherapy. There is, however, little information on the expression of these genes in human colorectal carcinomas.

METHOD

The expression of MAGE-1, -2, and -3 genes in 54 pairs of tumor and corresponding normal tissue specimens of the colorectum was determined by means of reverse transcription polymerase chain reaction. The induction of MAGE-1, -2, -3, and -4 gene expression in eight colorectal carcinoma cell lines also was examined by use of a demethylating agent, 5-Aza-2'-deoxycytidine (DAC).

RESULTS

The expression of MAGE genes was not recognized in normal colorectal tissues at all. In tumor tissue specimens, the expression of MAGE-1, -2, and -3 was recognized in 16 (30%), 15 (28%), and 11 (20%) patients, respectively. The expression was seen frequently in patients with liver metastasis (p < 0.01). Although MAGE-1 or -3 genes were not induced by DAC, MAGE-2 or -4 genes were induced in three of four MAGE-2 negative cell lines or three of seven MAGE-4 negative cell lines, respectively.

CONCLUSIONS

The MAGE genes were expressed exclusively in tumor tissues of one third of patients with colorectal carcinoma. The identification of such tumor rejection antigens is considered to uncover a new possibility for the specific immunotherapy of colorectal carcinoma. The demethylating agent may increase the number of patients who might be candidates for MAGE-specific immunotherapy.

Peptide-pulsed dendritic cells induce tumoricidal cytotoxic T lymphocytes from healthy donors against stably HLA-A*0201-binding peptides from the Melan-A/MART-1 self antigen

The melanoma antigen Melan-A/MART-1 was screened for the presence of potential HLA-A*0201-binding cytotoxic T lymphocytes (CTL) epitopes. The immunodominant nonamer epitope AAGIGILTV demonstrated weak binding to T2 but a significant half-life of binding to HLA-A*0201 in contrast to the decamer EAAGIGILTV. In addition to the immunodominant CTL epitope, we describe two peptides, GILTVILGV and ALMDKSLHV, that display stable binding to HLA-A*0201. Using cultured autologous dendritic cells pulsed with these peptides, CTL lines were induced from peripheral blood lymphocytes that displayed reactivity with HLA-A2+, Melan-A/MART-1+ melanoma cells. CTL reactivity against the immunodominant epitope could be induced with the nonamer epitope alone, but not with the decamer variant. CTL clones generated from an (EAAGIGILTV + AAGIGILTV)-induced CTL line recognize the appropriate melanoma cells and normal melanocytes. Upon further characterization of one of these CTL clones, it was found to be of surprisingly high affinity considering that it is directed against a self antigen. This study demonstrates that immunogenic peptides can be selected based on stability (half-life) of peptide/HLA binding. In addition, cultured DC were found to efficiently induce CTL responses in vitro against such selected peptides, and some of these CTL were capable of recognizing endogenously processed antigen.

Infrequent expression of the MAGE gene family in uveal melanomas

It has previously been reported that MAGE-1, -2, -3 and -4 genes are expressed in human cancers including cutaneous melanoma. MAGE-1 and MAGE-3 represent targets for specific immunotherapy because they encode peptide antigens which are recognised by cytotoxic T lymphocytes (CTL) when presented by HLA class I molecules, and pilot clinical trials with these peptides are currently in progress. It is likely that other members of the MAGE gene family may also encode antigens recognised by CTL. Uveal melanomas, like cutaneous melanomas, arise from melanocytes that are derived from the neural crest. To determine if uveal melanoma patients would be suitable for MAGE-peptide immunotherapy, the expression of MAGE-1, -2, -3 and -4 genes was assessed by reverse transcription followed by polymerase chain reaction (RT-PCR) amplification and ethidium bromide staining. Expression of MAGE genes was not detected in any of 27 primary tumours. Either MAGE-1 or MAGE-4 was expressed in only 2 of 26 metastatic samples, but expression of MAGE-2 or -3 was not detected. Our data suggest that, unlike cutaneous melanomas, uveal melanomas may not be suitable candidates for MAGE-peptide immunotherapy.

Expression of MAGE-1, MAGE-2, MAGE-3/-6 and MAGE-4a/-4b genes in ovarian tumors

MAGE genes encoding tumor-rejection antigens recognized by cytotoxic T lymphocytes are expressed at the mRNA level in various malignant tumors. We have investigated the expression of genes MAGE-1, -2, -3/-6 and -4a/-4b at the mRNA level in malignant and non-malignant ovarian tumors as well as in normal ovaries by reverse transcription-polymerase chain reaction. MAGE-1, -2, -3/-6 and -4a/-4b were expressed in 12, 5, 11 and 4 of 58 malignant tumors, respectively. The majority of these MAGE-mRNA-positive tumors were histologically surface-epithelial-stromal tumors, in particular serous adenocarcinomas. They mostly consisted of either advanced-stage or recurrent tumors. In contrast, neither benign tumors nor normal ovaries expressed any of the MAGE genes investigated. A 46-kDa MAGE-1 protein was identified in MAGE-1-mRNA-positive serous adenocarcinomas by immunoblot analysis with polyclonal anti-MAGE-1 antibody. These results provide important information for specific immunotherapy of ovarian serous adenocarcinomas with MAGE gene products.

Epitope selection and development of peptide based vaccines to treat cancer

Cytotoxic T lymphocytes recognize peptides that associate with class I major histocompatibility complex molecules. Since cytotoxic T cells have the capacity to recognize and destroy tumor cells, identification of epitopes recognized by these cells in tumor-associated antigens would allow the production of compounds for the treatment of cancer. Here we review some of the approaches being explored to identify tumor-associated antigens and to develop peptide-based vaccines that induce cytotoxic T lymphocytes against specific tumors.

Human esophageal carcinomas frequently express the tumor-rejection antigens of MAGE genes

The human genes MAGE-1 and -3 encode melanoma peptide antigens that are recognized by autologous cytotoxic T lymphocytes. Tumors expressing MAGE genes are potential targets for cancer immunotherapy, because MAGE genes are expressed only in tumor tissue and not in any normal tissue except testis and placenta. However, little is known about MAGE gene expression in human esophageal carcinoma. The purpose of this study was therefore to analyze MAGE gene status in human esophageal carcinoma. We studied the expression status of these genes in 42 surgical samples and in 12 cell lines of human esophageal carcinoma using the reverse transcription polymerase chain reaction (RT-PCR). Various clinicopathological factors were also analyzed. No MAGE gene expression was seen in any of the 42 normal esophageal tissue specimens. In contrast, tumor tissue expressed MAGE-1, -2, and -3 in 26, 18 and 24 specimens, respectively. Thirty-three of 42 tumors expressed at least one MAGE gene. Significant clinicopathologic differences between the tumors were not observed, regardless of the presence or absence of MAGE gene expression. In cell lines, MAGE-1, -2, and -3 gene expression was recognized in 5, 4 and 4 cell lines, respectively. This study demonstrates that MAGE genes are frequently expressed in clinical samples as well as in cell lines of esophageal carcinoma. The identification of MAGE genes, therefore, may open up a new modality of treatment, namely specific immunotherapy, for patients with esophageal carcinoma.

Expression of MAGE genes in primary and metastatic cutaneous melanoma

Human genes MAGE-1 and MAGE-3 code for antigens that are recognized on melanoma cells by autologous cytolytic T lymphocytes. These antigens may constitute useful targets for specific anti-tumor immunization of cancer patients, since genes MAGE-1 and MAGE-3 are expressed in a number of tumors of different histological types, but are not expressed in normal adult tissues other than testis. This also applies to genes MAGE-2 and MAGE-4, which are closely related to MAGE-1 and MAGE-3. We have analyzed the expression of these 4 MAGE genes in cutaneous melanoma. Sixteen of 100 primary tumors vs. 69 (48%) of 145 metastases from individual patients expressed MAGE-1. Similar differences in the frequency of gene expression between primary and metastatic tumor samples were observed for MAGE-2, MAGE-3, and MAGE-4. MAGE expression in primary tumors was correlated with tumor thickness: there was a significantly increased frequency in the expression of MAGE-1, -2 and -3 in tumors of greater thickness. Benign and dysplastic nevi, as well as in situ melanomas, did not express any of the 4 MAGE genes.

The expression of tumor-rejection antigen "MAGE" genes in human gastric carcinoma

BACKGROUND & AIMS

The genes MAGE-1 and MAGE-3 both encode melanoma peptide antigens recognized by major histocompatibility complex-restricted cytotoxic T lymphocytes. The antigens may be a target for immunotherapy. There is, however, little information on the expression of these genes in gastric carcinomas. Therefore, the expression of MAGE genes in gastric carcinomas was evaluated.

METHODS

The expression of MAGE-1, MAGE-2, and MAGE-3 genes in tumors and corresponding normal tissue specimens was studied using a reverse-transcription polymerase chain reaction. The results were analyzed according to clinicopathologic factors of the tumor.

RESULTS

In the 68 gastric carcinomas studied, MAGE-1, MAGE-2, and MAGE-3 messenger RNA were detected in 41%, 31%, and 38%, respectively. Fifty percent of the gastric carcinomas expressed at least one of the MAGE genes. Messenger RNA for the three MAGE proteins was not detected in normal gastric tissue. MAGE gene expression in gastric carcinomas was not associated with a significant clincopathology of the tumor. However, gene expression was lower in mucinous carcinomas (3 of 10).

CONCLUSIONS

MAGE-1, MAGE-2, and MAGE-3 are expressed in a high percentage of gastric carcinomas. These tumor rejection antigens may provide tumor-specific targets for immunotherapy.

Expression of the MAGE gene family in human head-and-neck squamous-cell carcinomas

The MAGE genes encode certain tumor-associated antigens recognized by cytotoxic T lymphocytes. We investigated the expression of the MAGE-1, -2, -3, -4, -41, and -6 genes in 88 head-and-neck squamous-cell carcinomas (83 fresh tumor samples and 5 cell lines), using a reverse-transcription-polymerase-chain-reaction assay, followed by dot-blot hybridization with sequence-specific oligonucleotides and/or restriction enzyme-pattern analysis. The MAGE-1, -2, -3, -4, -41 and -6 genes were expressed at the mRNA level in 27, 34, 36, 22, 16 and 35, respectively, of 83 fresh tumor samples. At least one of these genes was expressed in 59 of the 83 samples. Neither non-tumor inflammatory cells nor normal tissues were positive for these genes. The MAGE-1 gene was expressed relatively frequently in SCC of the oropharynx, hypopharynx and maxillary sinus, but at lower rates in SCC of the larynx and of the tongue and oral cavity. MAGE-1 was frequently expressed in poorly differentiated SCC, somewhat less frequently in moderately differentiated SCC, and only infrequently in well-differentiated SCC. The expression levels of the other MAGE genes also varied with the anatomic site as well as the degree of differentiation. Our results suggest that specific immunotherapy against MAGE gene products may be useful for patients with head-and-neck carcinomas.

Establishment of an enzyme-linked immunosorbent assay (ELISA) for measuring cellular MAGE-4 protein on human cancers

The MAGE genes encoding tumor-rejection antigens are expressed on various human cancers. An enzyme-linked immunosorbent assay (ELISA) was established for measuring cellular MAGE-4 protein (MAGE-4a and/or -4b) expressed on human tumor cells using a monoclonal antibody (mAb) and polyclonal Ab to recombinant MAGE-4b protein. Both the R5 mAb (IgG1) and the polyclonal Ab recognized a 45 kDa protein in extracts of MAGE-4 mRNA positive cancers, and showed no apparent cross-reactivity to the other MAGE gene products (MAGE-1, -2, -3, -6, and -12) by the immunoblot analyses. The R5 mAb and the polyclonal Ab primarily recognized one (the position 119-133) and two oligopeptides (the positions 119-133 and 259-273), respectively, among a series of 31 different MAGE-4b oligopeptides. The amino acid sequences of these two peptides were identical to those of MAGE-4a and -4b, but differed from those of all the other MAGE proteins (MAGE-1, -2, -3, -6, and -12). Substitution of glycine for amino acid in position 123 (arginine, R), 124 (lysine, K), 126 (R) or 128 (K) in a MAGE-4b oligopeptide of the position 119-132 severely decreased the reactivity of the R5 mAb to the oligopeptide. This ELISA also showed no apparent cross-reactivity with the other MAGE gene products (MAGE-1, -2, -3, -6, and -12). The minimum detectable level of MAGE-4 protein was determined to be 10 pg/well (100 pg/ml). The results suggest that this ELISA is a reliable and quantitative method to measure cellular MAGE-4 protein that is a potential target molecule for specific immunotherapy of human cancers.

Loss of HLA class I antigens by melanoma cells: molecular mechanisms, functional significance and clinical relevance

Malignant transformation of melanocytes may be associated with changes in the expression of major histocompatibility complex (MHC) HLA class I antigens. Interest in the characterization of abnormalities in the expression of MHC class I by melanoma cells has been rekindled by the current emphasis on the application of T-cell-based immunotherapy to melanoma. Here, Soldano Ferrone and Francesco Marincola review defects in class I expression as described in melanoma cells, as well as the molecular mechanisms, functional significance and clinical implications of such defects.

New tumor antigens recognized by T cells

A series of tumor cell antigens that are recognized by cytolytic T lymphocytes has been characterized this year. Besides the antigens derived from proteins specifically expressed in tumors, many melanoma antigens derive from melanocytic differentiation proteins. In addition, antigens unique to individual tumors result from mutations in ubiquitously expressed genes.

Anti-tumor activity of cytotoxic T lymphocytes elicited with recombinant and synthetic forms of a model tumor-associated antigen

The recent cloning of tumor-associated antigens (TAAs) recognized by CD8+ T lymphocytes (TCD8+) has made it possible to use recombinant and synthetic forms of TAAs to generate TCD8+ with anti-tumor activity. To explore new therapeutic strategies in a mouse model, we retrovirally transduced the experimental murine tumor CT26(H-2d), with the lacZ gene encoding our model TAA, beta-galactosidase (beta-gal). The transduced cell line, CT26.CL25, grew as rapidly and as lethally as the parental cell line in normal, immunocompetent animals. In an attempt to elicit TCD8+ directed against our model TAA by using purely recombinant and synthetic forms of our model TAA, we synthesized a nine-amino-acid long immunodominant peptide of beta-gal (TPH-PARIGL), corresponding to amino acid residues 876-884, which was known to be presented by the Ld major histocompatibility complex (MHC) class I molecule, and a recombinant vaccinia virus encoding the full-length beta-gal protein (VJS6). Splenocytes obtained from naïve mice and co-cultured with beta-gal peptide could not be expanded in primary ex vivo cultures. However, mice immunized with VJS6, but not with a control recombinant vaccinia virus, yielded splenocytes that were capable of specifically lysing CT26.CL25 in vitro after co-culture with beta-gal peptide. Most significantly, adoptive transfer of these cells could effectively treat mice bearing 3-day-old established pulmonary metastases. These observations show that therapeutic TCD8+ directed against a model TAA could be generated by using purely recombinant and synthetic forms of this antigen. These findings point the way to a potentially useful immunotherapeutic strategy, which has been made possible by the recent cloning of immunogenic TAAs that are expressed by human malignancies.

The failure of current immunotherapy for malignant glioma. Tumor-derived TGF-beta, T-cell apoptosis, and the immune privilege of the brain

Human malignant gliomas are rather resistant to all current therapeutic approaches including surgery, radiotherapy and chemotherapy as well as antibody-guided or cellular immunotherapy. The immunotherapy of malignant glioma has attracted interest because of the immunosuppressed state of malignant glioma patients which resides mainly in the T-cell compartment. This T-cell suppression has been attributed to the release by the glioma cells of immunosuppressive factors like transforming growth factor-beta (TGF-beta) and prostaglandins. TGF-beta has multiple effects in the immune system, most of which are inhibitory. TGF-beta appears to control downstream elements of various cellular activation cascades and regulates the expression of genes that are essential for cell cycle progression and mitosis. Since TGF-beta-mediated growth arrest of T-cell lines results in their apoptosis in vitro, glioma-derived TGF-beta may prevent immune-mediated glioma cell elimination by inducing apoptosis of tumor-infiltrating lymphocytes in vivo. T-cell apoptosis in the brain may be augmented by the absence of professional antigen-presenting cells and of appropriate costimulating signals. Numerous in vitro studies predict that tumor-derived TGF-beta will incapacitate in vitro-expanded and locally administered lymphokine-activated killer cells (LAK-cells) or tumor-infiltrating lymphocytes. Thus, TGF-beta may be partly responsible for the failure of current adoptive cellular immunotherapy of malignant glioma. Recent experimental in vivo studies on non-glial tumors have corroborated that neutralization of tumor-derived TGF-beta activity may facilitate immune-mediated tumor rejection. Current efforts to improve the efficacy of immunotherapy for malignant glioma include various strategies to enhance the immunogenicity of glioma cells and the cytotoxic activity of immune effector cells, e.g., by cytokine gene transfer. Future strategies of cellular immunotherapy for malignant glioma will have to focus on rendering glioma cell-targeting immune cells resistent to local inactivation and apoptosis which may be induced by TGF-beta and other immunosuppressive molecules at the site of neoplastic growth. Cytotoxic effectors targeting Fas/APO-1, the receptor protein for perforin-independent cytotoxic T-cell killing, might be promising, since Fas/APO-1 is expressed by glioma cells but not by untransformed brain cells, and since Fas/APO-1-mediated killing in vitro is not inhibited by TGF-beta.

A new family of genes coding for an antigen recognized by autologous cytolytic T lymphocytes on a human melanoma

Human melanoma MZ2-MEL expresses several distinct antigens that are recognized by autologous cytolytic T lymphocytes (CTL). Some of these antigens are encoded by genes MAGE-1, MAGE-3, and BAGE, which are expressed in a large fraction of tumors of various histological types but are silent in normal adult tissues with the exception of testis. We report here the identification of the gene coding for MZ2-F, another antigen recognized by autologous CTL on MZ2-MEL cells. This gene, which was named GAGE-1, is not related to any presently known gene. It belongs to a family of genes that are expressed in a variety of tumors but not in normal tissues, except for the testis. Antigenic peptide YRPRPRRY, which is encoded by GAGE-1, is recognized by anti-MZ2-F CTL on class I molecule HLA-Cw6. The two genes of the GAGE family that code for this peptide, namely GAGE-1 and GAGE-2, are expressed in a significant proportion of melanomas (24%), sarcomas (25%), non-small cell lung cancers (19%), head and neck tumors (19%), and bladder tumors (12%). About 50% of melanoma patients carry on their tumor at least one of the presently defined antigens encoded by the MAGE, BAGE, and GAGE genes.

A mutated intron sequence codes for an antigenic peptide recognized by cytolytic T lymphocytes on a human melanoma

We have identified an antigen recognized on a human melanoma by autologous cytolytic T lymphocytes. It is encoded by a gene that is expressed in many normal tissues. Remarkably, the sequence coding for the antigenic peptide is located across an exon-intron junction. A point mutation is present in the intron that generates an amino acid change that is essential for the recognition of the peptide by the anti-tumor cytotoxic T lymphocytes. This observation suggests that the T-cell-mediated surveillance of the integrity of the genome may extend to some intronic regions.

Expression of MAGE-1 gene by esophageal carcinomas

Expression of the MAGE genes encoding tumor-rejection antigens on HLA-A1 and -Cw1601 recognized by cytotoxic T lymphocytes was investigated in esophageal carcinomas at the mRNA level by the semiquantitative reverse transcription-polymerase chain reaction method. MAGE-1 and -2 genes, but not MAGE-3, -3/-6 and -4a/-4b genes, were expressed in substantial proportions of the primary esophageal carcinomas and their metastatic lymph nodes. The proportion of MAGE-positive samples in the primary esophageal carcinomas correlated with the T factor of the TNM classification (pT1: 2 of 12 tumors, pT2: 1 of 6, pT3: 12 of 29, and pT4: 7 of 18). These results have important implications for specific immunotherapy of esophageal carcinomas using MAGE-1 gene product.

Generation of human-melanoma-specific T lymphocyte clones defining novel cytolytic targets with panels of newly established melanoma cell lines

Melanoma is a cancer where the immune system is believed to play an important role in the control of malignant cell growth. To study the variability of the immune response in melanoma patients, we derived melanoma cell lines from several HLA-A2+ and HLA-A2- patients. The melanoma cell lines studied were designated FM3, FM6, FM9, FM28, FM37, FM45, FM55p, FM55M1 and FM55M2 and were established from eight metastatic tumors as well as from one primary tumor from a total of seven different patients. On the basis of the ability of tumor cells to induce specific cytotoxic T lymphocytes (CTL) from peripheral blood lymphocytes (PBL) in mixed lymphocyte/tumor culture with HLA-A2+ melanoma cells, the FM3 cell line was characterized as highly immunogenic. To investigate the expression of different melanoma-associated antigens recognized by CTL on different melanoma cell lines, we selected the cell line FM3 for restimulation and further T cell cloning experiments. The lytic activity of CTL clones with good proliferative activity was examined using a panel of HLA-A2+ and HLA-A2- melanoma cell lines. None of the tested HLA-A2- melanoma cell lines were susceptible to lysis by the CTL clones, whereas allogeneic HLA-A2+ melanoma cell lines were lysed only by a few CTL clones. On the basis of their reactivity with different melanoma cell lines, it was possible to divide the present CTL clones into at least four groups suggesting the recognition of at least four different antigens. Three of these target structures probably are different from already-described HLA-A2-restricted melanoma-associated antigens, because their expression in the different melanoma cell lines do not correlate with the recognition of melanoma cells by these CTL. The results first indicate that poorly immunogenic melanoma cells may express melanoma-associated antigens, and also suggest that, by using CTL clones obtained against different HLA-class-I-matched melanoma cells, it is possible to define such antigens.

Antigens recognized on human tumors by cytolytic T lymphocytes: towards vaccination?

It has been known for many years that cytolytic T lymphocytes that specifically recognize the tumor cells of the same patient can be derived from the blood of melanoma patients. Several of the antigens recognized by these antitumor T lymphocytes have now been completely identified. Some of them are sufficiently tumor-specific to envision their use as antitumor vaccines in selected cancer patients.

Expression of the MAGE gene family in primary and metastatic human breast cancer: implications for tumor antigen-specific immunotherapy

Some human tumors express known antigens that can be utilized as targets for specific immunotherapy. An absolute requirement for the efficacy of this therapeutic strategy is an adequate expression of the candidate antigen by all cells of the primary and metastatic tumor. To examine the presence and distribution of tumor-associated antigens in metastatic breast cancer, we used PCR analysis and ethidium bromide staining to test the expression of genes of the MAGE family in 28 primary tumors and related metastatic samples. Overall, samples obtained from 7 of 28 patients revealed positive. However, 2 of 3 primary tumors positive for MAGE-1 and/or MAGE-3 had corresponding negative metastatic lesions. On the contrary, 4 of the 25 MAGE-negative primary tumors gave rise to positive metastatic nodes. Our results confirm in vivo, at the molecular level, the tumor-antigen heterogeneity previously observed at the cellular level by in vitro analysis. Our data strongly suggest that, at least in patients with breast cancer, multiple different antigens would be required to optimize the recognition of neoplastic cells in immunotherapeutic protocols using MAGE products as target antigens.