During the past decade the mechanism of CTL-mediated killing has been resolved into 3 steps, and its cation requirements, and general nature have been well defined. However, biochemical understanding of the CTL-target interaction has made little progress. Recently, we have developed a monoclonal antibody (MAb) which blocks killing by binding to a previously undescribed molecule on the CTL membrane, a molecule which we therefore have termed lymphocyte function-associated antigen one (LFA-1). LFA-1 and Lyt-2,3 are the only presently identified sites for such blocking; antibodies to over a dozen other molecules expressed on the CTL do not block killing. Present evidence suggests that LFA-1 is crucial in the adhesive interaction of T cells with other cells (e.g., targets, macrophages, perhaps B cells) The continuing search for blocking MAbs provides a systematic way to link specific molecules with CTL function.

Mouse monoclonal antibodies to rat IgG were obtained by fusion of immune SJL mouse spleen cells to NSI myeloma cells. Seven monoclonal antibodies have been labeled with 125I and studied as to specificity and avidity by using a panel of rat monoclonal antibodies both as inhibitors and target antigens in soft well plate and indirect cell binding assays. All MAb were selected for high avidity of 4 X 10(7) to greater than or equal to 2 X 10(9) M-1. Four MAb were subclass-specific. RG11/39, RG7/1, and RG7/11 were absolutely specific for the Fc' region of IgG1, IgG2a, and IgG2b, respectively. RG9/6 showed specificity for the Fab' region of IgG2a but crossreacted with lower avidity with IgG2c. Three MAb reacted with rat kappa chains. RG7/9 defined a monotypic (common) kappa chain determinant. RG11/15 and RG7/7 were specific for allelic kappa 1a and kappa 1b determinants, respectively. The monotypic and kappa 1a allotypic determinants are topographically separated. The antibodies can be used as screening reagents in indirect cell binding assays. They have sensitivity similar to affinity-purified rabbit anti-rat IgG and more defined specificity. They do not crossreact with mouse or human IgG, making them particularly suitable companion reagents for rat anti-mouse or anti-human MAb. One Mab, RG7/7, strongly crossreacts with Syrian hamster IgG.

LFA-1, an antigen associated with antigen-specific T lymphocyte-mediated killing, and Mac-1, a macrophage differentiation antigen associated with type three complement receptor function, contain alpha chains of Mr = 180,000 and 170,000, respectively, and beta chains of Mr = 95,000. The monoclonal antibodies defining these antigens do not cross-react. The LFA-1 and Mac-1 beta chains are highly homologous or identical, whereas the alpha chains are highly different by tyrosyl tryptic peptide mapping (Kürzinger, K., Ho, M. K., and Springer, T. A. (1982) Nature (Lond.) 296, 668-670). T lymphoma cell lines express LFA-1 but not Mac-1 as shown by immunofluorescence and immunoprecipitation. Conversely, some macrophage-like lines express Mac-1 but not LFA-1. Other macrophage-like lines co-express Mac-1 and small amounts of LFA-1. Mac-1 and LFA-1 are present as separate molecules in these cells. [35S]Methionine and [[3H]glucosamine are incorporated into both alpha and beta chains of Mac-1 and LFA-1, showing both chains are endogenously synthesized and are glycoproteins. Cross-linking and two-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis experiments show that in both Mac-1 and LFA-1 the alpha and beta chains are noncovalently associated in alpha 1 beta 1 quaternary structures. By quantitative immunofluorescence flow cytometry, the EL-4 T lymphoma and P388D1 macrophage-like lines were estimated to express 10(5) LFA-1 and 1.6 x 10(5) Mac-1 molecules/cell, respectively. From these sources the antigens have been purified to homogeneity in 200-400-micrograms quantities by monoclonal antibody affinity chromatography. The purified antigens contain only the alpha and beta subunits.

Under conditions in which antigen dose is suboptimal, the recognition of Ia determinants is a necessary component of the generation of allospecific CTL responses. With monoclonal anti-Ia antibodies used as blocking reagents, it is demonstrated that the recognition of alloantigens may proceed via two pathways. Alloantigens can be recognized in the context of syngeneic Ia determinants in a similar fashion to conventional antigens. If, however, the Ia+ cells are removed from the responder population, the generation of such responses involves the recognition of allogeneic Ia determinants directly. A non-T, non-B, adherent accessory cell was identified as the critical syngeneic Ia+ cell required for CTL responses in these cultures.

A xenogeneic rat anti-mouse Ia monoclonal antibody, M5/114 (gamma 2b, kappa), was studied for its effects in vitro on T cell proliferative responses. Strain distribution studies revealed that M5/114 could inhibit I-A subregion-restricted T cell responses of the H-2b,d,q,u but not the H-2f,k,s haplotypes, indicating that this xenoantibody recognizes a polymorphic determinant on mouse Ia molecules. This same monoclonal antibody was found to inhibit BALB/c (H-2d) T cell proliferation to both G60A30T10 and G58L38 phi 4. The Ir genes regulating responses to these antigens map to either the I-A subregion (GAT), or the I-A and I-E subregions (GL phi), raising the possibility that M5/114 recognizes both I-A and I-E subregion-encoded Ia glycoproteins. It could be shown, using appropriate F1 responding cells, that M5/114 does in fact affect GAT and GL phi responses by interaction with both the I-A and the I-E subregion products, and not by any nonspecific effect resulting from binding to the I-A subregion product alone. These results are consistent with genetic and biochemical studies directly demonstrating that M5/114 recognizes A alpha A beta and E alpha E beta molecular complexes. The existence of a shared epitope on I-A and I-E subregion products suggests the possibility that these molecules arose by gene duplication. Finally, the precise correlation between the Ia molecules recognized by M5/114 and the ability of this antibody to block T cell responses under Ir gene control strengthens the hypothesis that Ia antigens are Ir gene products.

Monoclonal antibodies were prepared to anti-HLA-DR cytolytic T lymphocytes (CTLs) and screened for inhibition of CTL-mediated killing. Binding of monoclonal antibodies to four types of molecules, LFA-1, LFA-2, LFA-3, and HLA-DR, inhibited killing, suggesting that these molecules participate in the CTL-target cell interaction. The antigens were characterized by immunoprecipitation, crosslinking, NaDodSO4/polyacrylamide gel electrophoresis, and immunofluorescence flow cytometry. The LFA-1 antigen contains alpha and beta polypeptide chains of Mr 177,000 and 95,000 that are noncovalently associated in an alpha 1 beta 1 structure. It is present on both B and T lymphocytes and marks subpopulations that differ in quantitative expression. Human LFA-1 appears to be the homologue of mouse LFA-1. Human LFA-2 is of Mr 49,000 with a minor component of Mr 36,000. It is expressed on CTL lines but not on a B-cell line and in peripheral blood preferentially on T lymphocytes. Human LFA-3 is of Mr 60,000 and is expressed on both B and T lymphocytes.

A rat anti-mouse monoclonal antibody (MAb), M1/42, has been found to react with H-2 antigens from cells of the a, b, d, j, k, s, and u haplotypes (all haplotypes tested). This antibody, when bound to cells and reacted with FITC-conjugated anti-rat Ig, could be used to quantitate H-2 expression on several cell types. The antibody was also useful in comparing the H-2 products precipitated from a variety of haplotypes. M1/42-coupled Sepharose-4B beads were used to purify H-2d antigens by affinity chromatography. Pure H-2 molecules eluted from the column in 0.5% DOC, 0.65 M NaCl, 20 mM Tris, pH 8.0, yielding 110 to 180 micrograms H-2d/10(10) P815 tumor cells. This antibody, when used in series with H-2Kk-specific MAb 11-4.1, allowed purification of Dk and Dd from RDM-4 and YAC cells, respectively. H-2d purified by column chromatography on M1/42 was found to be serologically and biologically active, as determined by MAb rebinding, inhibition of cell lysis by alloantisera plus complement and ability to stimulate alloreactive CTL. This antibody and the described protocols should be useful in the preparation of relatively large quantities of a number of H-2 antigens.

A monoclonal antibody produced by a hybridoma cell line that has previously been shown to recognize an antigen present on murine macrophages and granulocytes (Mac-1) has now been shown to bind to human monocytes and polymorphonuclear leukocytes. Human monocytes bind about 40,000 M1/70 (anti-Mac-1) F(ab')2 or IgG molecules per cell in saturating conditions. In addition, M1/70 antibody recognizes a small population (less than 10%) of human blood lymphocytes. These cells express approximately 3-fold fewer Mac-1 antigen determinants than monocytes. Separation of this lymphocyte subset on a fluorescence-activated cell sorter has shown that all the natural killing activity in human blood can be found among these cells. Similarly, separation of the natural killer cells by an independent method based on their surface Fc receptor has shown that nearly all of them can be labeled by the hybridoma antibody. The same results are obtained when an F(ab')2 fragment of the M1/70 hybridoma antibody is used. The anti-Mac-1 antibody does not interfere with binding to the Fc receptor, nor does it interfere with either natural killing or antibody-dependent cellular cytotoxicity mediated by these cells. We conclude that there is a similar antigenic structure on the surface of murine and human monocytes and granulocytes and that this structure is also found on human natural killer cells.

Monoclonal antibodies (MAb) have been used to probe the relationship of cytolytic T lymphocyte (CTL) surface molecules to CTL function. Rat MAb to mouse CTL were generated. Twelve MAb so obtained gave preferential binding to T cells as compared to B cells, and three of these recognized previously undescribed surface polypeptides. These Mab and more broadly reactive and previously obtained MAb were tested for their ability to block CTL-mediated killing in the absence of complement. To ensure that any observed blocking was due to binding of MAb to the effector cell rather than the target cell, a xenogeneic mouse CTL anti-rat BN lymphoma target cell system was utilized (MAb and target cells both of rat origin). Of 24 MAb tested here, 21 had little or no effect on CTL function, including those to H-2, Thy-1, Lyt-1, Ly 5, Ly 6, Lgp 100, and at least six other defined antigens. We confirmed inhibition of killing with two MAb to Lyt-2,3. Another MAb, M7/14, gave profound and consistent blockade of CTL function. It was confirmed that M7/14 MAb blocks killing by binding to the mouse CTL and does not bind to the rat lymphoma target cells used for the CTL assay. The findings suggest that the antigen defined by M7/14, termed a lymphocyte function-associated antigen, LFA-1, participates in or is closely associated with the mechanism of CTL-mediated killing. LFA-1 contains two polypeptide chains of 180,000 and 95,000 Mr and is distinct from other described lymphocyte glycoproteins. LFA-1 thus represents both a previously undescribed lymphocyte surface antigen and molecular site for blockade of CTL-mediated killing.

Exudates induced by i.p. injection of five listeria monocytogenes (LM) constituted a rich source of CBA/J murine natural killer (NK) cells. Maximum expression of NK activity was seen from day 2 through day 6 after initial exposure to LM. When nylon wool nonadherent peritoneal exudate cells were examined by a single-cell cytotoxicity assay, the number of cells binding to YAC-1 target cells increased after infection as did their individual lytic capacity. A monoclonal rat anti-murine macrophage antibody (M1/70), previously shown by our group to recognize human NK cells, can also be used as a marker for murine NK cells. Utilizing M1/70 and the fluorescence-activated cell sorter, selection of M1/70-labeled mononuclear cells led to the enrichment of both NK and antibody-dependent cellular cytotoxicity. These M1/70-positive cells had a distinctive morphology and contained granules on Wright-Giemsa staining. They were not phagocytic, did not contain nonspecific esterase, and lacked surface I-Ak, IgM determinants, complement receptors, and high levels of Thy 1.2.

We have previously described a monoclonal antibody (MAb), M7/14, which blocks a variety of T cell functions, including CTL-mediated killing, the mixed lymphocyte response, and antigen-specific proliferation. The antigen defined by M7/14 has been designated lymphocyte function-associated antigen one (LFA-1). In this report, LFA-1 has been studied as to cell distribution, surface abundance, structure, and in comparison to other CTL surface antigens, LFA-1 is expressed on lymphoid cells of both the T and the B lineages and on a large fraction of bone marrow cells, but not on exudate macrophages or nonlymphoid tissues. T cells express more LFA-1 than B cells, both in the unstimulated and stimulated states. Compared with unstimulated spleen cells, cytolytic T lymphocyte cell preparations (CTLP) and Con A blasts, but not LPS blasts, show increased LFA-1 expression relative to H-2, and for T cell-containing populations, Lyt-2. M7/14 MAb binds to about 1.5 X 10(4) and 7 X 10(4) LFA-1 sites per average spleen cell or CTLP cell, respectively. M7/14 Mab binds to cTLP in quantitites of 2.5-fold and ies 10.4-fold less than H-2 and Thy-1 Mab, respectively; since the latter have little or no effect on CTL function, inhibition of killing by M7/14 MAb is specific for the LFA-1 surface site. M7/14 MAb and a blocking Lyt-2 MAb are bound in similar quantities of CTLP. LFA-1 is a glycoprotein and consists of 2 noncovalently linked polypeptide chains of 180,000 and 95,000 Mr. The same molecular species as on CTL is present on other T cells and on B cells. The molecular structure and cell distribution of LFA-1 clearly distinguishes it from Lyt-2,3, Ly-5, T145, and T11, which were previously suggested to be either associated with the function of and/or present on the surface of CTL.

Two monoclonal antibodies to mouse Ia antigens were produced by fusion of xenoimmune rat spleen cells with the NSI myeloma. These monoclonal antibodies detect polymorphic determinants present on B cells and activated T lymphocytes from mice carrying the H-2b, H-2d, H-2k, H-2r, and H-2q haplotypes but not from mice carrying the H-2s or H-2r haplotypes. Antigenic site number determinations showed the positive haplotypes can be divided into 2 groups. Mice bearing the H-2b, H-2d, and H-2q haplotypes express a high number--40,000 to 80,000--of antigenic sites per B lymphocyte, and monoclonal antibody plus complement can lyse B cells from these mice. In contrast, mice bearing the H-2k and H-2r haplotypes express a low number of antigenic sites--about 5000 per cell. Spleen cells from mice carrying the latter haplotypes are not lysed with monoclonal antibody and complement. Genetic mapping demonstrated that high and low expression map to the I-A and I-E subregions, respectively. The monoclonal antibodies detect an Ia specificity on I-Ab, I-Ad, I-Ed, and I-Ek molecules. These observations were confirmed using several different experimental approaches, i.e., cytotoxicity, fluorescent staining, competitive inhibition of monoclonal antibody binding, and 2-dimensional gel electrophoresis of immunoprecipitates. The avidity for A alpha b A beta b and E alpha k E beta k is 5 to 7 x 10(-9) M-1. The antigenic determinant is heat labile, which suggests that it is not carbohydrate. The results imply that Ia antigens encoded by distinct subregions share sequence homology, which may be a consequence of ancestral gene duplication.