Search

To detect novel molecules involved in immune functions, a subtracted cDNA library between closely related murine lymphoid cells was prepared using improved technology. Differential screening of this library yielded several clones with a very restricted tissue specificity, including one that we named CTLA-8. CTLA-8 transcripts could be detected only in T cell hybridoma clones related to the one used to prepare the library. Southern blots showed that the CTLA-8 gene was single copy in mice, rats, and humans. By radioactive in situ hybridization, the CTLA-8 gene was mapped at a single site on mouse chromosome 1A and human chromosome 2q31, in a known interspecific syntenic region. The CTLA-8 cDNA sequence indicated the presence, in the 3'-untranslated region of the mRNA, of AU-rich repeats previously found in the mRNA of various cytokines, growth factors, and oncogenes. The CTLA-8 cDNA contained an open reading frame encoding a putative protein of 150 amino acids. This protein was 57% homologous to the putative protein encoded by the ORF13 gene of herpesvirus Saimiri, a T lymphotropic virus. These findings are discussed in the context of other genes of this herpesvirus homologous to known immunologically active molecules. More generally, CTLA-8 may belong to the growing set of virus-captured functionally important cellular genes related to the immune system or to cell death and cell survival.

CD28, initially detected on human T lymphocytes with the help of antibodies, and CTLA-4, obtained by reverse genetics through its preferential expression in mouse activated T cells, are both single-V domain members of the Ig superfamily. Early work showed a relationship between these two molecules, which we wished to further document, in particular because of the growing realization of the functional importance of CD28 in some T cell activation pathways. Isolation and analysis of the mouse CTLA-4 gene and further analysis of the human CTLA-4 gene showed that both of these and the human CD28 gene share the same overall intron/exon organization. The nucleic acid sequence homology of the exons was found to extend across both molecules and species, whereas the 5' and 3' flanking regions exhibited homology across species but not between molecules. Message expression of human CTLA-4 was only detected in activated T cells and, thus, shares with that of mouse CTLA-4 and of mouse and human CD28 a lymphoid tissue distribution, although apparently broader for the latter. Two main human CTLA-4 transcripts of about 1.8 and 0.8 kb were detected, the smaller of which may derive, as reported for human CD28, from the use of an alternate degenerated polyadenylation signal sequence. The nucleic acid sequence data allowed a direct comparison of the four putative complete protein sequences of CD28 and CTLA-4 in the mouse and the human, showing striking homologies, especially in some stretches (such as a MYPPPY hexamer in the hinge region) conserved across molecules and across species. The mouse CD28 gene was localized to chromosome 1 band C by in situ hybridization with three different radioactive probes, indicating, together with previous data, that the CD28 and CTLA-4 genes map to the same chromosomal region in both the mouse and the human. Thus, CD28 and CTLA-4 were found to be strikingly similar in most respects, in terms of structure, sequence, expression, and gene location, furthermore in two species, strongly suggesting that their genes are the direct products of a duplication event and raising the possibility of functional homologies between the corresponding proteins.

Murine LMTK- cells were unexpectedly found to cross-react with a murine anti-human beta 2-microglobulin (beta 2-m)monoclonal antibody (m.Ab) after transformation with cosmid clones containing different purified HLA class I genes. The same cross-reactivity was observed with CTP 34 B4 (murine x human) somatic hybrid cells, which express class I molecules constituted of human HLA heavy chains and murine beta 2-m. Inhibition studies of the complement-dependent cytolysis mediated by the cross-reacting m.Ab indicated that isolated murine beta 2-m does not express the cross-reacting determinant, suggesting that its expression by the transformed cells reflects conformational modification of murine beta 2-m upon its association with HLA heavy chains. These results illustrate one of the possible post-translational mechanisms through which the antigenicity of a polypeptide chain can be modified. They might provide a serologic marker of the third domain of HLA class I heavy chains. Finally, because quantitative differences of reactivity with the anti-human beta 2-m m.Ab were observed, depending on the HLA class I genes used for transformation, these results individualize two families of HLA class I heavy chains responsible for different conformational modifications of murine beta 2-m.

It was previously established that a fetal calf serum-induced C57BL/6 T cell line that induces T and B cell differentiation could be kept proliferating in vitro only if cultured in the presence of irradiated syngeneic spleen cells and FCS. The present experiments were performed in order to investigate a) whether this cell line was a pure T cell line, b) whether the cells in this cell line (called line 12) were homogeneous with regard to Lyt phenotype, c) whether its growth was H-2 restricted, and d) whether line 12 cells reacted with our anti-idiotype (5936) and anti-T cell receptor allotype/isotype (6036) antisera. The results showed that line 12 consisted of T cells of Lyt 1+, 2.3- and phenotype. Its growth and proliferation was restricted to Kb and/or IAb alloantigen, and this phenomenon was observed with isolated Lyt 1+, 2.3- T cells. Line 12 cells reacted with both 5936 and 6036 antisera, and the positive cells were of Lyt 1+, 2.3- phenotype. Thus, our data indicate that Lyt 1+, 2.3- line 12 T cells interact with FCS and Kb/IAb alloantigen via receptors, which may bear 5936 and 6036 antisera-defined determinants. However, because these antisera only react with a subpopulation of Lyt 1+, 2.3- cells, proof that the same T cell has both MHC specificity and B cell idiotypic determinants will require further experimentation. 5936 and 6036 antisera-reactive molecules could be isolated from the supernatants of line 12 cells. Such molecules had characteristics similar to the 50,000 m.w. form of receptor molecules isolated from B6 anti-CBA T cell supernatants: a single chain polypeptide carrying both 5936 and 6036 antisera-defined determinants.

Previous studies indicated that Ca++ ionophores and phorbol esters in synergy could substitute for the initial activation step of normal T lymphocytes or T cell clones leading to increased expression of receptors for the growth factor interleukin 2 (IL 2) and secretion of interleukins, with the mitogenic signal for T cell proliferation being dependent on the presence of IL 2. In this study, the question was addressed as to whether T lymphocytes activated through the Ca++ ionophore ionomycin and the phorbol ester 12-o-tetradecanoyl phorbol 3-acetate (TPA) also acquired the competence to kill relevant target cells. The results indicate that T lymphocytes from primed mice proliferate and lyse the relevant allogeneic target cells after in vitro stimulation with ionomycin plus TPA, and that T lymphocyte preparations enriched for a subpopulation bearing the Lyt-2 marker are dependent on exogeneous sources of IL 2 to proliferate and become competent killer cells, whereas preparations enriched for subpopulations bearing the L3T4 marker grow independently of exogenous IL 2.

The present experiments showed 1) that it was possible to produce mouse T cell lines against MHC determinants with a relatively high success rate by stimulation of purified T cells with allogeneic cells in the presence of irradiated syngeneic spleen cells; 2) that these lines could be led to react against selected H-2 specificities; 3) that only T cell lines established from Ig-1b allotype mice contained 5936-Id+ T cells (5936-Idiotypes are defined by an antiserum against B6 anti-CBA IgG produced in rabbit no 5936, which was tolerant to mouse gamma-globulin); and 4) that antigenic determinants coded by IAk genes induce the 5936-Idiotype(s). The latter data are in accordance with the 5936-idiotype characteristics of primary MLC T blasts. All T cell lines contained both specific MLC-responding cells and cytolytic cells. However, studies on the functional capacity of 5936-Id+ T cells from both primary MLC and the T cell lines showed that neither MLC-responding cells nor cytolytic cells directed against H-2Kk, IAk, or H-2Dk were 5936-Id+. Thus, 5936-Id+ T cells may be regulator cells induced by IAk antigens.

"Stimulus-secretion" processes, and notably those in which the stimulus is provided by the formation of an antigen-antibody complex at the cell surface, require the presence of external Ca++. To test the possiblity that effector cells, in non-T cell mediated cytolysis, might be triggered to release a cytolytic product by a similar mechanism, we investigated the requirements of this phenomenon for divalent cations. We found that Ca++ is not sufficient, not necessary, and not inhibitory, whereas Mg++ is necessary and sufficient. Other cations were tested, and we found that Mn++ behaved like Mg++. The Ca++ data suggest that no conventional "stimulus-secretion" process is operative in non-T cell mediated cytolysis of antibody-coated sheep red blood cells.

Most studies of recognition events in cell-mediated cytolysis have analyzed the conditions that are required to establish mechanically stable binding between effector cells and target cells. This approach has shown that cytolytic T cells require a temperature above 20 degrees C and free divalent cations to bind specifically to target cells. K cells on the other hand were shown to bind to IgG on target cells in the absence of divalent cations at 4 degrees C. In this paper we show that both K cell- and T cell-mediated cytolysis are inhibited by cytochalasin A, if this drug is added to cultures before the effector and target cells have interacted. However, the drug does not block the calcium-dependent lethal hit stage of cytolysis in either system. We then show that in both T cell- and K cell-mediated cytolysis, the cytochalasin A-inhibitable stage can proceed in the absence of divalent cations. We conclude from these data that: 1) both K cell- and T cell-mediated cytolysis have a divalent cation-independent stage, which is inhibitable by cytochalasin A; 2) firm mechanical binding of T effectors to target cells requires divalent cations. It is uncertain if the latter is necessary for cytolysis of whether K cells also have to go through a divalent cation phase before the lethal hit stage. The implications of these findings are discussed, especially as to the possibility that strong binding may not be necessary for T cell-mediated cytolysis.

We and other investigators previously reported the cloning of CTLA-1 (or CCP-1) and CTLA-3 (or H Factor) serine esterase-related transcripts preferentially expressed in cytolytic T lymphocytes. We extended the survey of the tissue specificity of these molecules. Two main sets of results were obtained. First, both CTLA-1 and CTLA-3 transcripts could be found in the various cytolytic T cells tested, although in widely different amounts, and in some cases just at the threshold of detection. Secondly, these transcripts were not found in most of the other cells tested, including in some natural cytotoxic cells and in activated cytotoxic macrophages; however, they could be detected in mast cells for CTLA-1 and in some noncytotoxic lymphocytes for CTLA-3. Thus, the CTLA-1 and CTLA-3 serine esterase products are most probably not required for macrophage or natural cytotoxicity; their presence cannot be taken as characteristic of cytotoxic T cells; and a discussion about their relevance to T cell-mediated cytotoxicity should take into account their widely different amounts from one cytotoxic T cell to another.