Hapten-specific cytotoxic T cell clones undergo somatic variation of their antigen recognition specificity

Two experimental systems have demonstrated somatic variation of antigen recognition specificity of long-term cytotoxic T cell (CTL) clones. System 1 used CTL clone BT7.4.1 with strict specificity for Kb/TNP, which had been continuously cultured for 15 months in the presence of H-2b/TNP stimulator cells and interleukin 2. Upon removal of the TNP antigen from the cultures, 99% of the clone cells within about 10 cell divisions lost their ability to grow in the presence of antigen and interleukin 2 (lethal variants). Of the surviving 1%, about 60% retained the ability to lyse target cells in the presence of lectins but only 12% could be considered as "wild type" BT7.4.1 cells, i.e. they still specifically lysed H-2b/TNP-bearing target cells. The majority of the growing cells, thus, had to be considered as specificity loss variants. Several specificity loss variants were established in culture and were shown to express membrane-bound T cell "receptor" heterodimer similar to their TNP-specific ancestor, BT7.4.1. Principally the same types of variants were generated in cultures growing in the presence of TNP antigen, although in quantitatively reduced numbers. Under these conditions the specific stimulator cells appeared to impose a significant selective advantage for "wild type" CTL since even after 15 months the cultures fully retained their specificity for the nominal antigen. In system 2, the development of cytolytic fine specificity of a panel of 42 individual Kb/TNP-specific CTL clones was followed over a period of 8 months of in vitro culture. At the beginning of the test, 37 of these clones exhibited significant cross-reactivity for lysis of H-2k/TNP target cells. This number of cross-reactive clones continuously diminished with time and dropped to only 4 clones after 8 months in culture. All 42 clones retained their original Kb/TNP specificity and after losing their reactivity for H-2k/TNP usually showed no decrease but rather an increase in their cytotoxic activity towards Kb/TNP target cells. Loss of H-2k/TNP cross-reactivity was not accompanied by loss of Lyt-2 or of LFA-1 surface antigens or by loss of sensitivity of the cytotoxicity to inhibition by anti-Lyt-2 or by anti-LFA-1 antibody. We conclude from these observations that in vitro cultivated CTL clones, at least those of C57BL/6 anti-TNP-C57BL/6 specificity, are not stable in terms of their antigen recognition specificity.(ABSTRACT TRUNCATED AT 400 WORDS)