SUMMARYThe New Zealand black (NZB) mouse strain is genetically predisposed to develop, at approximately 6 months of age, a spontaneous and severe autoimmune anaemia caused by production of pathogenic anti-mouse erythrocyte autoantibodies. In order to investigate the molecular mechanisms which lead to anti-mouse erythrocyte autoantibody production we have generated eight anti-mouse erythrocyte MoAbs producing hybridomas from splenocytes of 9-and 12-month-old NZB with spontaneous autoimmune anaemia. IgG2a was the predominant isotype, while IgM, IgGl and IgG2b were each produced by one hybridoma cell line. All anti-mouse erythrocyte MoAbs were characterized for their antigen specificities. None of the MoAbs cross-reacted with ss-or dsDNA or with other species' erythrocytes, with the exception of one MoAb which cross-reacted with rat erythrocytes. None ofthe eight hybridomas was demonstrated to express surface or cytoplasmic CD5, suggesting that they derived from CD5" B lymphocytes. All hybridomas when implanted intraperitoneally into BALB/c mice caused anaemia. In order to define the genetic basis and investigate the molecular mechanisms resulting in pathogenic anti-mouse erythrocyte autoantibody production, the pattern of immunoglobulin variable region gene use has been studied. Five ofthe eight MoAbs whose IgVH genes were sequenced all have functionally rearranged genes from the VH J558 gene family. There is evidence for somatic point mutations in the complementarity-determining regions (CDR) ofthe IgVH genes in all of these five MoAbs when compared with the closest known germline gene. We suggest that these nucleotide sequence changes are likely to reflect selection by an antigen-driven mechanism. Furthermore, the data indicate that pathogenic anti-mouse erythrocytes are not derived from 'natural' autoantibodies. Keywords autoantibodies complementarity-determining region monoclonal antibody H genes B lymphocytes tNTRODUCTtONAntibodies have diverse antigen specificities resulting from the structure ofthe assembled V, D and J heavy chain and V and J light chain gene products. Antibodies generated during the rearrangement of germline V(D)J gene segments without the presence of foreign antigen represent the pre-immune repertoire. Due to the almost unlimited potential for V region diversity caused by the combinatorial events during VH-D-JH and VL-JL gene rearrangements and subsequent somatic mutations, antibodies with specificity for autoantigen can occur [1]. The immunoglobulin coding region is organized into largely non-overlapping Vn-gene family clusters, the most proximal to the D and J region loci being the 7183 and Q52 VH-gene families, and J558 and 3609 VH-gene families being most distal.

The NZB mouse is genetically predisposed to develop, at approximately 6 months of age, a spontaneous and severe autoimmune anaemia caused by production of pathogenic anti-mouse erythrocyte autoantibodies. Molecular analysis of a panel of five anti-erythrocyte monoclonal antibodies (MoAb) derived from splenocytes of unimmunized NZB mice revealed that these autoantibodies all had functionally rearranged genes from the VH J558 family of immunoglobulin genes with closest homology to germline genes H10 and H30. Owing to clustering of nucleotide differences within the CDRs, compared with the germline, it was concluded that these antibodies were most likely generated by an antigen-driven mechanism. We report here further molecular analysis of two (4.16.1 and B4.13.2) of the panel of five anti-mouse erythrocyte producing hybridomas which are apparently clonally related. Nucleotide analysis of the light chain cDNA indicated that both antibodies had closest homology to germline gene V kappa 24 and use J kappa 2 gene. Determination of the functional affinities of the MoAb reveal that B4.13.2 (IgG2a) has a > 10-fold higher affinity for mouse erythrocytes when compared to 4.16.1 (IgG1). This finding supports the view that these two autoantibodies are generated by an antigen-driven mechanism. The proposed mechanism would involve the selection and expansion of a small population of B-lymphocytes by antigen leading to isotype switch, somatic mutation and increased affinity. Our data also point to the possibility that some framework residues may be involved in the binding to antigen.

HLA-restricted T cell responses toward immunogenic peptides, whether mutated or non-mutated, can induce antitumor responses in patients with advanced cancer. The fact that potential non-mutated antigens are greater in number than mutated antigens by multiple orders of magnitude and the high polymorphism of HLA genes may have hampered comprehensive analyses of the specificity of antitumor T cell responses toward non-mutated antigens than mutated antigens. Unlike shared antigens, the vast majority of neoantigens are not shared and are unique to each patient. The elucidation of T cell epitopes derived from shared antigens may facilitate the robust development of an efficacious and safe adoptive T cell therapy that is readily available to a larger cohort of cancer patients. It is well established that melanoma tumor-infiltrating T lymphocytes (TILs) contain antitumor T cells that are specific for both non-mutated and mutated antigens. The adoptive transfer of TILs can induce sustained clinical responses in some patients with advanced melanoma. However, a precise and extensive understanding of the shared antigen targets of TILs has been lacking. In this study, TILs were isolated from 8 metastatic melanoma patients, polyclonally expanded in vitro, and their shared antigen specificities for all 45 (25 different) class I alleles were examined. The combination of structure-based analysis using human cell-based peptide/HLA (pHLA) multimers and functional analysis using artificial antigen-presenting cells (APCs) were used to determine antigen-specific T cell responses. We were able to determine the specificity of 12.2 ± 7.2% (mean ± SD, max 25.8%, min 4.6%) of CD8+ T cells in an expanded TIL culture toward 3.0 ± 1.8 (mean ± SD, max 6, min 1) previously known or novel peptides derived from shared antigens. Furthermore, we isolated a number of cognate TCR genes with potent tumor reactivity from the CD8+ T cells. The strategy employed in this study using a library of paired human cell-based pHLA multimers and artificial APCs has enabled us to decipher the antigen specificity of tumor-specific T cells for any given HLA class I allele, regardless of allele frequency and for an infinite number of peptides. It has also allowed us to build a large database of class I-restricted peptides and cognate tumor-reactive TCR genes at an unprecedented scale. This database will be a valuable tool in defining the immune response at the level of each individual cancer patient with precision as well as the identification and validation of biomarkers to aid in patient-based selection of a cancer immunotherapy regimen. Furthermore, this database will help the robust development of novel cancer vaccines and TCR gene therapies for patients with a low mutation burden. Citation Format: Kenji Murata, Kayoko Saso, Linh T. Nguyen, Douglas Millar, Marcus O. Butler, Pamela S. Ohashi, Naoto Hirano. Decoding shared antigenic epitopes and their cognate TCR genes in melanoma TILs using a library of paired human cell-based pHLA multimers and artificial APCs [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 568.