Tumor infiltrating lymphocytes (TIL) obtained from patients with melanoma have been used to clone the genes encoding the antigens recognized by these TIL. TIL have been identified that can recognize unique cancer antigens on murine and human cancers including melanoma, breast cancer, colon cancer and lymphoma. The MHC restricted recognition of human cancer antigens was detected by assaying panels of HLA-typed target cells and by transfection into target cells of genes encoding the appropriate HLA specificities. In clinical trials of TIL administration, 36% of patients with metastatic melanoma underwent objective cancer remission. TIL trafficked to and accumulated in cancer deposits. In an attempt to isolate antigens specifically involved in vivo cancer rejection in humans with melanoma we have used an approach involving the following four steps. TIL were identified from patients with metastatic melanoma that selectively recognized autologous cancer antigens in vitro based on assays of either tumor cell lysis or cytokine secretion. These TIL were administered to the autologous cancer patient and those TIL that were capable of mediating the regression of cancer in vivo were identified. TIL that could mediate in vivo cancer rejection were used to clone the genes that encoded the antigens recognized by these TIL. cDNA libraries were transfected into target cell lines bearing the appropriate MHC antigen and the resulting transfectants were tested for the ability to mediate specific cytokine release when cocultivated with the TIL. To determine whether the identified genes actually encoded cancer regression antigens, patients can be immunized with these genes or gene products to determine whether the regression of growing cancers would be induced. Alternatively, in vitro sensitized lymphocytes against the putative cancer regression antigens can be generated and adoptively transferred into patients to determine whether they could mediate cancer regression. Utilizing TIL capable of mediating in vivo regression, eight genes encoding tumor antigens recognized in the context of Class I and two genes recognized in the context of Class II have been identified. The MART-1 and gp100 antigens, restricted by HLA-A2, and the tyrosinase antigen, restricted by HLA-A24 are normal nonmutated differentiation antigens present in melanomas and normal melanocytes. The TRP-1 and TRP-2 antigens, restricted by HLA-A31, were encoded by normal melanoma differentiation antigen genes but were translated from an open reading frame different from that of the normal protein. The p15 antigen, restricted by HLA-A24, was derived from a normal gene of unknown function which was transcribed in a variety of cells but was only expressed on the cell surface of melanomas. The beta-catenin tumor antigen,restricted by HLA-A24, was derived from a normal gene containing a single base mutation that resulted in a single amino acid change. Beta-catenin binds cell surface adhesion molecules as well as the APC tumor suppressor gene product and this mutation also may have been involved in the malignant phenotype of this tumor. The NY-ESO-1 gene shared by melanomas, breast cancer and prostate cancer has also recently been identified. The immunodominant peptides present on each of these antigens have been identified. Of 18 HLA-A2 restricted TIL administered to cancer patients, 15 recognized the MART-1 antigen and all 15 recognized the same nine amino acid peptide (AAGIGILTV.) Eleven of the 18 HLA-A2 restricted TIL recognized gp100 and five different immunogenic epitopes were identified. The nine amino acid immunodominant epitopes from the TRP-1 and tyrosinase antigens have also been identified. Interestingly, the peptide epitope in the beta-catenin molecule was generated by a single amino acid mutation which led to a dominant binding motif for the HLA-A24 MHC molecule. MART-1, gp100, TRP-1 and tyrosinase are all normal melanocyte differentiation antigens that are the products of normal genes. Evidence that these normal melanocyte differentiation antigens can serve as tumor rejection antigens came from studies of vitiligo in patients in the Surgery Branch, NCI treated with IL-2 based immunotherapy. No cases of vitiligo were seen in 104 patients with metastatic renal cancer compared to 12 cases of vitiligo in 69 patients with melanoma (p < 10-4). Thus the presence of a growing melanoma appeared to sensitize patients to antigens present on melanocytes that could result in melanocyte destruction. Further, no vitiligo was seen in 27 melanoma patients unresponsive to therapy with IL-2 although 12 of 42 patients who did experience IL-2 induced cancer regression developed vitiligo (p = 0.002). This association between the presence of vitiligo and cancer regression provides strong presumptive evidence that the immunotherapy targets involved in cancer regression involve these normal differentiation antigens. It thus appears that a growing melanoma is capable of overcoming tolerance to normal tissue antigens although the mechanism by which this occurs is unclear. Interestingly, the immunodominant peptides from MART-1 and gp100 have an intermediate rather than a high binding affinity to HLA-A2 due to the absence of optimal residues at either the dominant second or ninth anchor positions for binding to HLA-A2. These observations suggest that self peptides that bind to MHC molecules with high affinity and which therefore are expressed at high levels on the cell surface may induce tolerance to T cells which recognize these epitopes. Thus, only lower binding epitopes or ?cryptic? self determinants may be capable of stimulating residual precursor cells. A series of clinical studies have been initiated utilizing these genes and gene products for the immunotherapy of patients with metastatic melanoma. Twenty-three patients have been immunized with the immunodominant MART-1 peptide in incomplete Freunds adjuvant (IFA) and 27 patients have been immunized with one of three gp100 immunodominant peptides in IFA. In studies of peptides containing individual amino acid substitutions designed to increase MHC binding two modified peptides from the gp100 molecule were identified that had far greater in vitro immunogenicity than the native peptides. A clinical trial using immunization with these immunodominant peptides in IFA revealed that the gp100: 209-217(210M) peptide is highly immunogenic in humans and gives rise to high levels of antipeptide and antitumor precursors. Immunization with these peptides plus IL-2 resulted in a 42% objective response rate in patients with metastatic melanoma. Recently new techniques have been developed to clone the genes encoding tumor antigens presented on Class II and recognized by CD4 positive T cells. Three new Class II restricted melanoma antigens have been identified and many more are under study. Recombinant vaccinia, adenovirus and fowlpox viruses have been constructed encoding either the MART-1 or gp100 gene. Clinical trials utilizing recombinant adenovirus encoding MART-1 have recently been initiated in the Surgery Branch, NCI. Successful immunization appears to be limited by the generation of neutralizing antibodies to the envelope proteins of these viruses. Other ongoing trials include immunization with multiple peptides or peptides modified to increase presentation on the cell surface. Clinical protocols in which adoptive transfer of cloned peripheral blood lymphocytes sensitized in vitro to the immunodominant peptides are in progress. These studies represent our initial efforts to identify the genes encoding cancer regression antigens and their use in cancer treatment protocols. TIL have now been identified that can identify antigens on breast cancer, colon cancer and lymphomas and attempts to clone the genes encoding these antigens are in progress.
Showing the most recent 10 out of 127 publications
