No abstract

Stomach cancer is the major malignancy in Japan and one of the most common cancers worldwide. To establish the basis for an immunotherapeutic approach to stomach cancer, we have initiated an analysis of stomach cancer antigens recognized by human immunoglobulin G (IgG) antibodies using SE-REX, a powerful expression cloning method developed by Dr. M. Pfreundschuh's group. Five stomach cancer cDNA libraries have been screened with autologous patient sera: one moderately differentiated adenocarcinoma; two poorly differentiated adenocarcinomas; and two scirrhous-type poorly differentiated adenocarcinomas of Borrmann type 4, the most devastating form of stomach cancer. Based on the reactivities of clones with autologous IgG antibodies, an average of 50 independent clones from each library and a total of 297 clones were isolated. DNA sequencing revealed that these 297 clones were derived from 136 different genes. Comparison of the 136 genes to sequences in DNA databases showed that 95 are previously identified genes and 41 are newly identified in this study. The antigens are derived from various genes including a chimeric gene between E-cadherin and an unknown gene Y, AKT oncogene, genes overexpressed in stomach cancers, genes of which the transcripts are alternatively or aberrantly spliced, and genes known to be involved in autoimmune diseases. Thus stomach cancer patients can generate an immune response against a surprisingly diverse set of gene products. To identify antigens potentially useful in the diagnosis and therapy of gastric cancer, all 136 genes were tested for their reactivities with a panel of sera from 44 gastric cancer patients (17 women and 27 men, aged 35-81 years) and with a panel of sera from 100 control individuals with no previous history of cancer but some of whom had gastritis (55 women and 45 men, aged 30-69 years). Eleven antigens showed reactivity only with a certain proportion of cancer patient sera but not with any control sera. An additional 12 antigens elicited antibody production at a much higher frequency in cancer patients than in control individuals. To evaluate the clinical usefulness of these antigens we are now examining their expression in normal and malignant tissues.

Enormous strides in cancer immunology have been made during the past decade. This is largely due to the development of methodologies capable of defining the antigenic targets on cancer cells that elicit a host immune response. The molecular cloning of antigens recognized by cytotoxic T cells by Boon and his colleagues has provided a growing list of tumor antigens, particularly for melanoma, that allows detailed monitoring of T cell responses to these antigens and offers promising targets for cancer vaccine development. An alternative new method, SEREX, for the serological identification of cancer antigens has been developed by Pfreundschuh and his colleagues. SEREX can be applied to all types of cancer including breast cancer that have been unapproachable by using cytotoxic T cells and thus offers an opportunity to define a vast range of cancer antigens. Toward thedevelopment of a vaccine for breast cancer, we have begun using SEREX to study breast cancer and have identified a few promising cancer antigens. Each antigen is now being critically evaluated as a possible vaccine target.