No abstract

Human melanoma line MZ2-MEL expresses several antigens recognized by autologous cytolytic T lymphocytes (CTL). As a first step towards the cloning of the gene coding for one of these antigens, we tried to obtain transfectants expressing the antigen. The DNA recipient cell was a variant of MZ2-MEL which had been selected with a CTL clone for the loss of antigen E. It was cotransfected with genomic DNA of the original melanoma line and with selective plasmid pSVtkneo beta. Geneticin-resistant transfectants were obtained at a frequency of 2 x 10(-4). These transfectants were then screened for their ability to stimulate the production of tumor necrosis factor by the anti-E CTL clone. One transfectant expressing antigen E was identified among 70,000 drug-resistant transfectants. Its sensitivity to lysis by the anti-E CTL was equal to that of the original melanoma cell line. When this transfectant was submitted to immunoselection with the anti-E CTL clone, the resulting antigen-loss variants were found to have lost several of the transfected pSVtkneo beta sequences. This indicated that the gene coding for the antigen had been integrated in the vicinity of pSVtkneo beta sequences, as expected for cotransfected DNA.

We have attempted to optimize the production of stable human cytolytic T lymphocyte clones directed against autologous melanoma cell lines. MLTC were restimulated every week with irradiated melanoma cells in medium containing human serum and IL-2. After 21 to 35 days, in 5 out of 6 patients, these cultures expressed a preferential cytolytic activity against the autologous melanoma cells, as compared to autologous EBV-B cells or NK target K562. Limiting dilution of MLTC responder cells was performed at times varying from days 7 to 28, in medium containing IL-2 and allogeneic EBV-B cells as feeders. Approximately 1% of these responder cells gave rise to CTL clones that lysed the autologous melanoma cells, but did not lyse K562 or autologous B cells. It was possible to maintain in culture for several months a large number of CTL clones that retained this specificity with high activity, and multiplied more than 5-fold every week. Some of these CTL clones were dependent on the presence of the autologous melanoma cells for their growth. With one melanoma, the use of autologous CTL clones made it possible to identify 3 different antigens on the tumor cells.

We derived from blood lymphocytes of a melanoma patient a large number of cytolytic T-cell clones directed against a cell line of the autologous tumor. Three distinct groups of antigens were recognized by these CTL on the autologous melanoma cells: group A consisted of stable antigens present on all sublines, whereas antigens B and C appeared unstable and were expressed by distinct sublines. In vitro immunoselections with various anti-A CTL clones were applied to the melanoma cells and variants resistant to 3 different CTL clones were obtained. These variants remained sensitive to other anti-A CTL clones, indicating that group A comprises at least 4 different antigens (D, E, F and A'). From a total of 76 CTL clones obtained from lymphocytes collected from the patient at various times, we found that 45 were anti-B, 17 were anti-C, 2 were anti-D, 9 were anti-E, 2 were anti-F and I was anti-A'. It is therefore likely that the 6 antigens identified by these CTL clones represent all or nearly all the transplantation antigens recognized by autologous CTL on this human melanoma.