論文

Using cDNA expression cloning, a cDNA encoding a novel human melanoma Ag, MART-2 (melanoma Ag recognized by T cells-2), recognized by HLA-A1-restricted CD8+ T cells from tumor-infiltrating lymphocytes (TIL1362) was isolated from an autologous melanoma cell line, 1362 mel. Homologous sequences to the cDNA had been registered in the EST database. This gene encoded an uncharacterized protein expressed ubiquitously in most normal and cancer cells. A mutation (A to G transition) was found in the cDNA obtained from the1362 mel melanoma cell line in the sequences encoding the phosphate binding loop (P-loop) that resulted in loss of the ability to bind GTP. Transfection of NIH-3T3 with the mutated MART-2 did not result in the development of significant foci. By screening 36 various cancer cell lines using single-strand conformation polymorphism, a possible mutation in the P-loop of MART-2 was found in one squamous cell lung cancer cell line, EBC1. The T cell epitope for TIL1362, FLEG-NEVGKTY, was identified to be encoded by the mutated sequence of the MART-2 Ag. The mutation substituted glycine in the normal peptide with glutamic acid at the third amino acid of the epitope, which is an important primary anchor amino acid for HLA-A1 peptide binding. The normal peptide, FLGGNEVGKTY, was not recognized by TIL1362, suggesting that this T cell response was specific for the autologous tumor. Although transforming activity was not detected in the NIH-3T3 assay, MART-2 with the mutation in the P-loop may be involved in the generation of melanoma through a loss of GTP binding activity.

A variety of molecules involved in the immune system have recently been isolated. These molecules have allowed the design of immunological interventions of immune responses to human cancer, or so called immuno-gene therapy. Clinical trials have been performed in Europe and the U.S. using various protocols, including gene-modified T cells; highly immunogenic cancer cells transfected with genes encoding cytokines, co-stimulatory molecules, and foreign antigens; immunization with tumor antigen genes as forms of plasmids and recombinant viruses; and immunization with dendritic cells transduced with tumor antigen genes. However, the anti-tumor effects observed in these reports have been limited. Improvement of various points will be necessary to establish immuno-gene therapy for cancer on the basis of fundamental research on immunoregulation against cancer, including isolation of tumor antigens and analysis of tumor escape mechanisms.

T cell responses specific for melanoma cells and melanocytes appear to be involved in the rejection of melanoma tumors, as well as in the development of autoimmune reactions in patients with Vogt-Koyanagi-Harada disease (VKH), sympathetic ophthalmia, or autoimmune vitiligo. Some of the target antigens for those T cells have been isolated using cDNA expression cloning with melanoma reactive T cells derived from tumor infiltrating lymphocytes (TIL) of patients with melanoma. These include melanocyte specific proteins, such as tyrosinase, TRP1, TRP2, gp100, and MART-1, cancer-testis antigens, and mutated peptidcs derived from genetic alterations in melanoma cells. Some of the melanoma reactive T cells appear to respond to cryptic or subdominant self epitopes in melanosomal proteins. Modification of those epitopes to increase their immunogenicity by replacement of amino acids at primary anchor residues for peptide/MHC binding, allowed an improvement in immunotherapy for patients with melanoma. Targets for autoreactive T cells against melanocytes in those autoimmune disorders remain to be identified. Isolation of novel target antigens is important for understanding these pathological T cell responses, as well as for developing new diagnostic and treatment methods for these diseases. A variety of techniques, including cDNA expression cloning with T cells, serological analysis of recombinant cDNA expression libraries (SEREX), cDNA subtraction with representational differential analysis (RDA), and serial analysis of gene expression (SAGE) are now being applied to identify novel melanoma/melanocyte antigens recognized by T cells and antibodies.

A total of 123 tumor-infiltrating T lymphocyte (TIL) cultures established from patients with HLA-A1, -A2, -A3, -A24, or -A31 metastatic melanoma in the Surgery Branch, National Cancer Institute, were screened for recognition of shared melanoma antigens including five melanosomal proteins (tyrosinase, MART-1/melan-A, gp100, TRP1, TRP2) as well as peptides derived from MAGE-1 and MAGE-3. Examination of the specificity of these T cells indicated that 16% of HLA-A1 TIL, 57% of HLA-A2 TIL, 7% of HLA-A3 TIL, 13% of HLA-A24 TIL, and 27% of HLA-A31 TIL recognized shared melanoma antigens restricted by major histocompatibility complex class I. Melanosomal proteins were frequently recognized by these TIL, and MART- 127-35, gp100154- 162, gp100209-217, and gp100280-288 represent highly immunogenic epitopes that were recognized by a high percentage of HLA-A2 restricted melanoma reactive TIL. Recognition of gp100 by HLA-A2 restricted TIL significantly correlated with clinical response to adoptive immunotherapy with TIL in 21 HLA-A2 melanoma patients (p = 0.024). Four HLA-A1, two HLA-A2, two HLA-A3, one HLA-A24, and two HLA-A31 restricted shared antigen-specific TIL did not recognize the previously identified antigens tested in this study, and may be useful for the identification of new melanoma antigens. The observation that TILs isolated from patients with metastatic melanoma recognized melanosomal proteins in the context of predominant HLA-A alleles implies that it may be possible to develop immunotherapies for patients with melanoma expressing diverse HLA types.

HLA class II alleles and immunoglobulin allotypes were determined in 83 Japanese patients with immune thrombocytopenic purpura (ITP) and 114 race-matched healthy controls. Distribution of DRB and DQB1 alleles as well as G1M(f, z), G2M(n+, n-), and KM(1, (1,2), 3) were not different between ITP patients and healthy controls, while DPB1*0201 was marginally increased in ITP patients vs. healthy controls (51% vs. 28%, Pc = 0.04, OR = 2.6 [1.4-4.8]). In contrast, strong associations between anti-glycoprotein autoantibodies and HLA class II genes were found as follows: Anti-GPIIb-IIIa antibody with DRB1*0405 and DQB1*0401; and anti-GPIb-IX antibody with DRB1*0803 and DQB1*0601. When factors influencing therapeutic responses to splenectomy were examined, a poor response was correlated with the presence of DRB1*0405, DQB1*0401 and anti-GPIIb-IIIa antibody (P = 0.01, 0.002, and 0.03, respectively). Our results indicate that HLA class II genes influence the production of anti-glycoprotein antibody specificities rather than the development of ITP. In addition, HLA class II genotyping could be useful in predicting therapeutic responses to splenectomy in Japanese patients with ITP.

Dendritic cells engineered to express tumor-associated antigens can be used to prime and potentiate cytotoxic T cell responses against cancer cells. Numerous delivery systems for introducing genes to dendritic cells have been examined, which include (I) Viral vector systems such as Moloney leukemia virus vector (MLV), human immunodeficiency virus (HIV) vector, and adenovirus vector, vaccinia virus vector, canarypox virus vector and herpes simplex virus (HSV) vector, and (II) Non-viral vector systems such as cationic liposome, particle bombardment (gene gun) and electroporation methods. In this review, we adress several issues regarding these gene delivery systems in terms of genetic modification of dendritic cells. Since every method has advantages and/or disadvantages, the optimal methods for presenting antigen epitopes to dendritic cells need to be clarified. Several animal models by using dendritic cells transduced with model tumor antigens have shown in vivo antitumor effects, suggesting the potential clinical usefulness of this strategy.

We report a treatment for spinal cord injury involving implantation of dendritic cells (DCs), which act as antigen-presenting cells in the immune system. The novel mechanisms underlying this treatment produce functional recovery. Among the immune cells tested, DCs showed the strongest activity inducing proliferation and survival of neural stem/progenitor cells (NSPCs) in vitro. Furthermore, in DC-implanted adult mice, endogenous NSPCs in the injured spinal cord were activated for mitotic de novo neurogenesis. These DCs produced neurotrophin-3 and activated endogenous microglia in the injured spinal cord. Behavioral analysis revealed the locomotor functions of DC-implanted mice to have recovered significantly as compared to those of control mice. Our results suggest that DC-implantation exerts trophic effects, including activation of endogenous NSPCs, leading to repair of the injured adult spinal cord.

Human tumor antigens recognized by T cells have been recently identified in various cancers, including pancreatic cancer. With the identified antigens, new immunotherapies can be developed using more efficient immunologic intervention (due to sufficient amounts of antigens in a more immunogenic form), as well as more quantitative and qualitative immunomonitoring. Various immunotherapies for patients with various cancers, including pancreatic cancer, are currently under evaluation in clinical trials. These include adoptive transfer of tumor reactive T cells and LAK cells; nonmyeloablative stem cell transplantation; active immunization with the identified tumor antigens, various tumor-derived products, dendritic cells pulsed with tumor antigens, and gene-modified tumor cells. Although these efforts in the realm of pancreatic cancer are still limited, various groups in Japan continue to be actively involved in this field of research.

Single nucleotide polymorphisms (SNPs) of inflammatory cytokine genes were examined in 84 adult Japanese patients with chronic immune thrombocytopenic purpura (ITP) and 56 race-matched healthy controls. The SNPs examined were within the genes encoding tumour necrosis factor (TNF)-α (-238 G/A and -308 G/A), TNF-β (+252 G/A), and interleukin (IL)-1β (-511 C/T and +3953 T/C). Of these SNPs, the frequency of the TNF-β (+252) G/G phenotype was significantly higher in ITP patients than in healthy controls (21% vs. 7%, P = 0.04, odds ratio = 3.6, 95% confidence interval 1.1-11.1), while no significant association was detected for the other SNPs. The distribution of the TNF-β (+252) phenotype was not associated with human leucocyte antigen class II alleles or the therapeutic response in ITP patients. The frequency of circulating anti-glycoprotein IIb/IIIa antibody-producing B cells was significantly higher in ITP patients with the TNF-β (+252) G/G phenotype than in those with the G/A or A/A phenotype (11.9 ± 4.9 vs. 6.8 ± 4.9 and 3.7 ± 2.8 per 105 peripheral blood mononuclear cells; P = 0.02 and P < 0.001, respectively). These findings suggest that the SNP located at TNF-β (+252) contributes to susceptibility to chronic ITP by controlling the autoreactive B-cell responses to platelet membrane glycoproteins.

Intratumoral inoculation with a herpes simplex virus (HSV) mutant, G207, as an in situ cancer vaccine has been shown to inhibit tumor growth by inducing tumor-specific immune responses. Here, as a step toward the clinical application of this therapeutic approach, we evaluated different protocols for enhancing the antitumor effect. First, in a bilaterally established tumor model with CT26 colon carcinoma, we demonstrated that multiple intratumoral inoculations with G207 induced a greater antitumor effect on both the inoculated and distant tumors than did 1 or 2 inoculations. Second, to boost this antitumor effect, we developed 2 strategies: multiple in situ cancer vaccines with G207 in combination with systemic administration of recombinant interleukin-12 (rIL-12) (G207/systemic rIL-12) or local administration of rIL-12 (G207/local rIL-12). The antitumor effects in both the inoculated and distant tumors by the combined treatments were significantly greater than by either G207 or rIL- 12 treatment alone. G207/systemic rIL-12 and G207/local rIL-12 mediated the complete regression of both the inoculated and distant tumors in 67% and 79% of the animals, respectively. These results indicate that multiple intratumoral inoculations of G207 and systemic or local rIL- 12 administration work synergistically to facilitate tumor regression and that this combination of treatments may have potential for treating cancer metastasis.

To identify tumor antigens useful for the diagnosis and treatment of patients with bladder cancer, a lambda phage cDNA library constructed from a high-grade bladder cancer cell line was screened with autologous serum from a patient with metastatic bladder cancer. Forty-eight distinct antigens were isolated. By evaluating the immunogenicity and the tissue-specific expression, KU-BL-1 and KU-BL-2 were identified as immunogenic antigens with restricted tissue expression. KU-BL-1 was found to be a putative human lipoic acid synthetase with a metal-binding site, CXXXCXXC, that was expressed in bladder cancer cell lines and most bladder cancer tissues, as well as normal bladder mucosa and testis tissues. Immunoglobulin (Ig)G antibody to KU-BL-1 was detected in 2 of 28 patients with bladder cancer, but not in 30 healthy individuals. KU-BL-2 was found to be a putative human kelch-like protein that was homologous to Drosophila kelch, with a BTB/POZ domain and kelch repeats. KU-BL-2 was expressed in bladder cancer cell lines, most bladder cancer tissues, testis and heart, but not in normal bladder mucosa. IgG antibody to KU-BL-2 was detected in 8 of 28 patients with bladder cancer, but not in 16 healthy individuals. Tumor reactive T cells were induced from peripheral blood mononuclear cells (PBMC) by stimulation with one of the HLA-A24 binding KU-BL-2 peptides. Therefore, KU-BL-1 and KU-BL-2, which showed preferential expression in bladder cancer with restricted expression in normal tissues, as well as immunogenicity in multiple patients with bladder cancer, may be useful for the development of diagnostic and therapeutic methods for patients with bladder cancer.