Altered CD40 ligand induction in tolerant T lymphocytes.

CD40 ligand (CD40L) is a member of the tumor necrosis factor superfamily and is expressed on the surface of activated T lymphocytes. The interaction of CD40L with CD40 on B cells results in B cell activation, immunoglobulin (Ig) secretion and Ig class switching. To study anergy as a mechanism of murine CD4 T cell tolerance, we determined both in vivo and in vitro that CD3-activated anergic cells are deficient in the ability to stimulate B cell proliferation, and that anergic cells are defective for the T cell receptor/CD3-mediated induction of CD40L expression. These results have implications for the recruitment of B cell responses by anergic T cells in vivo.

Induction of interleukin 12 responsiveness is impaired in anergic T lymphocytes.

The cytokine, interleukin 12 (IL-12), stimulates both natural killer cells and T cells to proliferate and to secrete interferon gamma (IFN-gamma). The T cell proliferative response to IL-12 must be induced and is evident after T cell receptor-mediated stimulation. As reported here, tolerant CD4+ T cells and clones, that are anergic for IL-2 production, are also anergic for induction of the proliferative response to IL-12. Murine T helper 1 clones tolerized in vitro, as well as anergic CD4+ T cells isolated from mice tolerized to the Mls-1a antigen (Ag) in vivo, demonstrated defective induction of proliferation to IL-12 upon restimulation with Ag. IL-12-enhanced production of IFN-gamma was observed in both control and anergic cells after Ag/antigen-presenting cell (APC) activation, although total IFN-gamma secretion by anergic cells was less than that produced by control cells, even in the presence of IL-12. These data indicate that T cell clonal anergy results in profound inhibition of proliferative responses, since the autocrine growth factor, IL-2, is not produced, and the APC-derived cytokine, IL-12, is not an effective stimulus for anergic T cell proliferation.

Reduced CD3-mediated protein tyrosine phosphorylation in anergic CD4+ and CD8+ T cells.

Mice inoculated i.v. with superantigens exhibit long lived Ag-specific T cell tolerance. An in vitro model for this phenomenon is the ensuing unresponsiveness of Th1 T cell clones activated via the TCR/CD3 complex in the absence of co-stimulation. We have previously demonstrated alterations in TCR-mediated early protein tyrosine phosphorylation events in Th1 clones anergic for IL-2 production. In this study, we demonstrate unresponsiveness in CD4+ and CD8+ T cells from V beta 8.1 transgenic mice inoculated i.v. with the superantigen Mls-1a. The unresponsiveness of both CD4+ and CD8+ T cells involves defective IL-2 production upon restimulation, with CD4+ T cells exhibiting an additional defect in IL-2 utilization. The transgenic model allowed study of T cell signaling in a relatively homogeneous population of unresponsive cells without elaborate purification of Ag-reactive populations. Both CD4+ and CD8+ T cells exhibit altered tyrosine phosphorylation of two protein substrates upon CD3-mediated restimulation. The substrates involved, p38 and p75, are of identical size to substrates similarly affected in anergic Th1 clones. Altered tyrosine phosphorylation is therefore closely associated with defective IL-2 production in these three anergic T cell types, and may play a role in the maintenance of anergy.

Altered protein tyrosine phosphorylation in anergic Th1 cells.

Clonal anergy as a mechanism for tolerance in T lymphocytes can be studied using an in vitro culture system, in which cloned CD4+ Th1-type murine T cells are rendered anergic for IL-2 transcription. The long-lasting molecular changes in anergic cells that prevent the response to Ag restimulation are not yet known. To determine whether the TCR might be uncoupled from normal intracellular signaling pathways, we investigated the response of anergic T cells to Ag, to anti-CD3 antibodies, or to anti-CD4 antibody restimulation in terms of early protein tyrosine phosphorylation events. Tyrosine phosphorylation of the CD3 zeta chain was apparently normal. In contrast, defects in the induction of tyrosine phosphorylation of three major T cell protein substrates were demonstrated. Altered phosphorylation correlated with functional nonresponsiveness for proliferation and reversal of anergy by growth in exogenous IL-2 resulted in reversal of the phosphorylation defects as well as in recovery of Ag responsiveness. These results suggest that specific defects in tyrosine phosphorylation pathways required for the induction of IL-2 synthesis may help to explain nonresponsiveness to Ag in tolerant T cells.

Anergic Th1 cells express altered levels of the protein tyrosine kinases p56lck and p59fyn.

Tolerance in T lymphocytes can result from clonal anergy, or paralysis, of Ag-specific T cells. To investigate the molecular mechanisms responsible for anergy, a system in which tolerance can be induced in vitro was employed. Anergy, as defined by long-lived nonresponsiveness to normal antigenic stimulation for IL-2 production, was produced in cloned murine CD4+ Th1 cells. Here we report that such anergic Th1 cells express constitutively reduced amounts of the protein tyrosine kinase p56lck and constitutively elevated levels of the protein tyrosine kinase p59fyn. Because protein tyrosine phosphorylation is known to be important for the normal induction of IL-2 synthesis, these results suggest that T cell anergy may be maintained, at least in part, by alterations in tyrosine phosphorylation signaling events.

A recessive defect in lymphocyte or granulocyte function caused by an integrated transgene.

A line of transgenic mice has been identified with a recessive defect in lymphocyte or granulocyte function, presumably as a result of insertional mutagenesis by the integrated transgene. Transgenic mice homozygous for the transgene integrant showed nearly complete absence of lymphocytes in peripheral lymph nodes and Peyer's patches, a severely diminished thymus medulla, and a greatly enlarged spleen. These animals also developed a syndrome characterized by granulocyte and mononuclear infiltrates in numerous tissues, including skin, liver, and lung, and immunoglobulin deposits in kidney glomeruli. Lung infiltrates were specifically localized around large blood vessels and bronchi, accompanied in some cases by destruction of arterial walls. The light scatter profile of spleen lymphocytes suggested an extremely high percentage of blast cells. Because tissue development and morphology appears to be normal in all other tissues observed, the genetic lesion appears to specifically affect the regulation of lymphocyte or granulocyte activation.

Suppression of antibody synthesis by CD4+ T cell clones and normal T cells stimulated with monoclonal anti-CD3 antibody.

CD4+ve Th1 clones, as well as normal splenic T cells, were found to suppress LPS-driven antibody secretion in a non-Ag-specific and non-MHC-restricted manner when the T cells were activated with the anti-CD3 mAb, 145-2C11. Suppression was observed with both primed and naive B cells, as well as with purified hapten-specific B cells, a result that suggests a direct effect of anti-CD3-activated T cells on B cell differentiation. Th1 clones activated by cognate Ag also suppressed LPS-driven antibody secretion. Furthermore, suppression of LPS-driven antibody secretion could be achieved across a cell-impermeable porous membrane when T cells were activated with anti-CD3. Suppression by Th1 clones and by normal T cells could not be attributed to a concomitant decrease in B cell proliferation or to a shift in the kinetics or isotype of the antibody response. These data demonstrate that CD4+ve Th1 clones, as well as normal T cells, can effect suppression of polyclonal antibody formation.

Synergistic activation of granulocyte-macrophage colony-stimulating factor production by IL-1 and IL-2 in murine Th1 cells.

Recent reports indicate that murine CD4+ Th1-type cloned T cells are insensitive to IL-1 because specific IL-1R are not detected on these cells and IL-1 does not modulate proliferative responses. However, we have determined that Th1 clones can respond to IL-1, because they function synergistically with IL-2 to induce granulocyte-macrophage-CSF secretion. This response to IL-1 plus IL-2 could be induced by IL-1 alpha or IL-1 beta and by membrane-bound IL-1 on macrophages. However, IL-1R could not be detected, and Th1 cells did not respond to IL-4 in the presence or absence of IL-1, as measured by either proliferation or granulocyte-macrophage-CSF production. Therefore, IL-1 functioned as a cofactor in Th1 cells stimulated with IL-2, but not with IL-4. A possible mechanism whereby IL-1 activates Th1 cells is discussed.

Characterization of two major populations of lung fibroblasts: distinguishing morphology and discordant display of Thy 1 and class II MHC.

We have determined that murine lung fibroblasts are divisible into two major subpopulations based on expression of Thy 1. Twenty-four to fifty-three percent of freshly isolated lung cells displayed Thy 1 and were separated using FACS into Thy 1+ and Thy 1- fractions for morphologic examination. Analysis by electron microscopy revealed that both the Thy 1+ and Thy 1- fractions contained fibroblasts. Freshly isolated lung cells cultured for 2 wk consisted of greater than 95% fibroblasts, with 28 to 49% displaying Thy 1. These cells were sorted by FACS into Thy 1+ and Thy 1- lines that maintained a stable phenotype over many weeks and that were used as a source to obtain stable fibroblast clones. Adherent pulmonary fibroblasts are not phagocytic and lack the markers of macrophages, dendritic cells, B lymphocytes, and T lymphocytes (with the exception of Thy 1). Interestingly, the Thy 1- fibroblasts spread more and contained a more extensive microfilament and microtubule network than did the spindly and often lipid-containing Thy 1+ population. Both populations of fibroblasts synthesized collagen. Class I MHC expression was very low on Thy 1+ and Thy 1- fibroblasts, but high levels were displayed after gamma-IFN treatment. Most exciting was the unexpected finding that only the Thy 1- lines and clones displayed class II MHC (Ia) in response to treatment with gamma-IFN. Moreover, only the Thy 1- fraction (gamma-IFN-treated) presented antigen to T lymphocyte clones, an observation that suggests that this subset of cells may be involved primarily in promoting chronic lung inflammation, which is associated with developing fibrosis. Thus, two populations of pulmonary fibroblasts exist, defined by the expression of Thy 1, distinguishing morphology, inducibility for Ia expression, and antigen-presenting function. It should now be possible, using these characteristics, to ascertain the role of pulmonary fibroblast subpopulations in developing fibrosis.

Prostaglandin E2-dependent induction of granulocyte-macrophage colony-stimulating factor secretion by cloned murine helper T cells.

PG are known to inhibit T cell proliferation, at least in part by suppressing IL-2 production, but effects of PG on the production of other lymphokines have not been well studied. We have found that PGE2 and PGE1, but not PGF2 alpha, inhibit both proliferation and production of granulocyte-macrophage (GM)-CSF by murine TH clones stimulated with Ag or anti-CD3 antibody. Thus, signals generated via the Ag receptor:CD3 complex were inhibited by PGE. Most interesting, however, was the finding that PGE2 and PGE1 could act synergistically with IL-2 for the induction of GM-CSF in some TH1 clones. Dependence on PGE2 for this response was not found in all clones, as some TH1 cells could produce GM-CSF after IL-2 alone, and some cells did not produce GM-CSF even in the presence of PGE2 and IL-2. These observations indicate that there is a subset of TH1 cells receptive to a stimulating activity of PGE2 in the presence of IL-2. PGE2 is known to elevate cAMP levels in T cells. Therefore, we tested whether other agents known to increase cAMP, such as forskolin and cholera toxin, could act in conjunction with IL-2 to induce GM-CSF secretion. As was found with PGE2, these compounds also induced GM-CSF activity in the presence of IL-2, suggesting a critical role for cAMP in this process. Overall these data indicate that the requirements for activation of GM-CSF secretion vary among individual T cells. Most importantly they provide the first evidence that E-series PG are positive signals for lymphokine induction in certain T cells, whereas simultaneously acting as negative signals limiting proliferation. This result also suggests that treatment with anti-inflammatory drugs that decrease PGE2 concentrations may inhibit lymphokine secretion normally stimulated by this pathway.

Stimulation of normal inducer T cell clones with antigen presented by purified Ia molecules in planar lipid membranes: specific induction of a long-lived state of proliferative nonresponsiveness.

Culture of normal inducer T cell clones with antigen and purified Ek beta:Ek alpha incorporated into planar lipid membranes resulted in specific T cell activation as determined by cell volume increase and IL 3 production. However, in contrast to results obtained with T cell hybridomas, antigen presentation by planar membranes did not induce measurable IL 2 production, and proliferative responses were not detected. Rather, recognition of only Ek beta:Ek alpha and antigen resulted in the specific induction of a long-lived state of proliferative nonresponsiveness to subsequent stimulation by conventional APC and antigen. Induction of nonresponsiveness required protein synthesis, and was not simply due to the absence of IL 2. The antigen-nonresponsive cells could respond to either PMA plus ionomycin or IL 2, and they expressed normal levels of surface antigen-receptor molecules. These results demonstrate that recognition by normal T cell clones of antigen and Ia molecules in the absence of other accessory cell molecules and signals results in a prolonged state of proliferative nonresponsiveness, possibly similar to a state of T cell tolerance in vivo.

Quantitative analysis of the T cell response to antigen and planar membranes containing purified Ia molecules.

Planar lipid membranes containing the purified Ia molecule E beta k:E alpha k can present a peptide antigen derived from cytochrome c to the T cell hybridoma 2B4.11. The incorporation of E beta k:E alpha k into planar membranes was linear over a 120-fold range of Ia molecule concentrations, permitting the dependence of the T cell response on the Ia molecule concentration to be examined. As the Ia molecule concentration was increased in the planar membranes, two parameters changed: less antigen was needed to stimulate the T cells, and the plateau response seen at functionally saturating concentrations of antigen increased. The antigen sensitivity was analyzed by plotting the antigen concentration (log2) required to stimulate the release of 10 U of IL 2 from the T cells as a function of the Ia molecule concentration (log2). If the T cell recognized a simple unit of one antigen molecule and one Ia molecule, this plot should have generated a straight line with a slope of -1. Surprisingly, a line with a slope of -2.04 X/divided by 1.12 was observed, suggesting that the T cell might recognize one antigen molecule and two Ia molecules. This complexity, however, resulted from changes in the maximal response achieved at different Ia molecule concentrations. A similar phenomenon was observed when the Ia molecule concentration was decreased in cultures containing splenic antigen-presenting cells (APC) by the addition of an anti-E beta k:E alpha k monoclonal antibody, or the use of [B10.A(4R) X B10.PL]F1APC. The Ia molecule concentration can therefore be limiting for T cell hybridomas in cultures containing normal APC and functionally saturating amounts of antigen. When the planar membrane data were normalized to the maximal response to eliminate the effect of the changing plateau response, the resulting plot generated a line with a slope of -1.17 X/divided by 1.11. These results suggest that the sole stimulatory signal for this T cell hybridoma consisted of a 1:1 ratio of antigen and Ia molecules.

Optimization of antigen presentation to T cell hybridomas by purified Ia molecules in planar membranes. Ia molecule polymorphism determines the antigenic fine specificity of the response to cytochrome c peptides.

Ia molecule (Ek,b beta:Ek alpha or Ak beta:Ak alpha)-containing planar membranes were constructed with cholesterol and a 9:1 molar ratio of the phospholipids dipalmitoyl phosphatidylcholine and dilinoleoyl phosphatidylcholine. This lipid composition was found to be optimal for the stimulation of T cell hybridomas of different specificities. Use of this system allowed the detection of weak responses not measurable when other artificial membranes were used. Activation of the cytochrome c, Ek,b beta:Ek alpha-reactive hybridoma 2B4.11 using such membranes resulted in responses comparable to those found using antigen-presenting cells (APC); that is, similar amounts of IL-2 were produced at the same concentrations of antigenic peptides. Presentation of moth and pigeon cytochrome c peptides by Ek beta:Ek alpha- or Eb beta:Ek alpha-reconstituted membranes resulted in 2B4.11 response patterns similar to those previously described using B10.A or B10.A(5R) APC. These data conclusively demonstrate that differential stimulation by moth and pigeon cytochrome c peptides depends solely on structural differences in the E beta:E alpha molecules used for antigen presentation.

Ek beta mutant antigen-presenting cell lines expressing altered Ak alpha molecules.

Antigen-presenting cells (APC) expressing mutant Ek beta and Ak alpha proteins were isolated after chemical mutagenesis of TA3 cells and negative immunoselection for altered Ek beta molecules. Mutant clones were analyzed for biosynthesis, assembly, and cell surface expression of altered Ia molecules, and were assayed for antigen-presenting function by using a variety of T cell clones. Three types of mutants were detected: type 1, which had lost expression of the Ek beta chain and produced altered Ak alpha chains; type 2, which also expressed altered Ak alpha chains, and which expressed Ek beta proteins that had lost reactivity to the 17.3.3 and 74D monoclonal antibodies (mAb), but retained reactivity to other anti-Ek beta mAb; and type 3, which had lost expression of both Ek beta and Ak beta: Ak alpha surface molecules. Thus, all of the mutant clones that produced modified Ak alpha proteins also displayed either total loss or serologic modification of the Ek beta molecule. Ek beta:E alpha-reactive T cell clones were not stimulated when type 1 or type 3 cells were used as APC, but all such T cells were fully reactive with type 2 mutant APC. Most Ak beta:Ak alpha-reactive T cell clones could respond to type 1 and 2 APC, and none were responsive to type 3 APC. However, two autoreactive Ak beta:Ak alpha-specific T cell hybridomas were stimulated only very weakly by type 1 and type 2 cells expressing modified Ak alpha proteins. These results demonstrate that Ia mutations can have highly selective effects on antigen presentation to T cells as well as on mAb binding, and thus suggest that individual Ia molecules may be composed of many different functional subsites.

Unexpected expression of a unique mixed-isotype class II MHC molecule by transfected L-cells.

Class II (Ia) major histocompatibility complex (MHC) molecules are heterodimeric integral membrane proteins composed of non-covalently linked alpha and beta glycoprotein chains. Studies of both normal cells and L-cell transfectants have shown that neither alpha- nor beta-chains are found on the cell surface alone, and that alpha beta dimers are required for membrane expression. In both mouse and man, several distinct non-allelic alpha and beta genes exist. Analysis of Ia molecules by immunoprecipitation and two-dimensional gel electrophoresis has demonstrated apparently selective association of particular pairs of the various alpha- and beta-chains to form the expressed class II isotypes I-A and I-E (mouse) or DQ, DP and DR (human). Because the various alpha- or beta-chains encoded by distinct loci exist in many allelic forms within a species, such specific pairing suggests a special role for isotypically conserved regions of each chain in the association process. In attempting to localize such putative assembly-controlling regions using the technique of DNA-mediated gene transfer, various combinations of murine alpha and beta genes were introduced into L-cells. Here we report the unexpected observation, following transfection, of mixed-isotype (Ad beta Ea/k alpha) molecules on the L-cell membrane and document that the formation of this pair is strongly influenced by allelic polymorphism of the A beta chain.

Influence of allelic polymorphism on the assembly and surface expression of class II MHC (Ia) molecules.

Cell surface Ia expression was examined following transfection of murine alpha and beta class II major histocompatibility complex (MHC) genes into L cells. Although haplotype-matched (e.g., Ak beta Ak alpha) gene pairs yielded high expression in primary transfectants, haplotype-mismatched pairs (e.g., Ak beta Ad alpha) gave unexpectedly low expression. RNA analysis revealed a requirement for greater A beta, and particularly A alpha transcript levels in haplotype-mismatched vs. haplotype-matched transfectants with similar levels of membrane Ia. A beta allelic control of this assembly/expression process was mapped to the NH2-terminal (beta 1) domain, the locus of high intraspecies polymorphism. These data on the effects of allelic variation on Ia levels provide a possible explanation for the strong linkage disequilibrium of A alpha and A beta genes and may account for the current molecular organization of the I region of the MHC.