Identification of an MHC class I-restricted autoantigen in type 1 diabetes by screening an organ-specific cDNA library.

Type 1 diabetes is an autoimmune disease in which the insulin-producing pancreatic beta cells are destroyed at an early age by an immune process that involves both CD4 and CD8 T lymphocytes. The identification of autoantigens in diabetes is very important for the design of antigen-specific immunotherapy. By screening a pancreatic islet cDNA library, we have identified the autoantigen recognized by highly pathogenic CD8 T cells in the non-obese diabetic mouse, one of the best animal models for human diabetes. This is the first identification, to our knowledge, of a CD8 T-cell epitope in an autoimmune disease. The peptide recognized by the cells is in the same region of the insulin B chain as the epitope recognized by previously isolated pathogenic CD4 T cells. This has very important implications for the potential use of insulin in preventative therapy.

Interferon gamma plays a critical role in induced cell death of effector T cell: a possible third mechanism of self-tolerance.

We have used anti-T cell monoclonal antibodies (mAbs) as mimic ligands to study the effects of T cell receptor (TCR) ligation of cloned T helper type 1 cells in the presence or absence of accessory cells. Our results demonstrate that ligation of the TCR in the absence of accessory cells rapidly induces cell death. Cell death can be prevented by addition of spleen adherent cells, leading to strong clonal expansion. Induced cell death is inhibited by cyclosporin A and by anti-interferon gamma (IFN-gamma), and is restored by adding exogenous recombinant IFN-gamma to cyclosporin A-treated cells. These results demonstrate that IFN-gamma plays a critical role in cell death induced by anti-TCR mAbs in the absence of costimulatory cells. We propose that induced cell death of active effector T cells provides a third mechanism of tolerance in addition to intrathymic deletion of developing autoreactive T cells and peripheral inactivation of mature, naive T cells.

Development of CD8alpha/alpha and CD8alpha/beta T cells in major histocompatibility complex class I-deficient mice.

Peripheral CD8(+) T cells mainly use CD8alpha/beta, and their development is mainly dependent on the major histocompatibility complex (MHC) class I proteins K(b) and D(b) in H-2(b) mice. In this report, we have shown that the development of CD8alpha/beta TCR-alpha/beta cells in lymphoid organs as well as in intestinal intraepithelial lymphocytes (iIELs) is dependent on the MHC class I K(b) and D(b) proteins. In contrast, TCR-alpha/beta CD8alpha/alpha cells are found mainly in iIELs, and their numbers are unaffected in K(b)D(b) double knockout mice. Most of the TCR-gamma/delta cells in the iIELs also bear CD8alpha/alpha, and they are also unaffected in K(b)D(b) -/- mice. In beta2-microglobulin (beta2m)-deficient mice, all of the TCR-alpha/beta CD8alpha/alpha and CD8alpha/beta T cells disappear, but TCR-gamma/delta cells are unaffected by the absence of beta2m.

The Fab fragment of a directly activating monoclonal antibody that precipitates a disulfide-linked heterodimer from a helper T cell clone blocks activation by either allogeneic Ia or antigen and self-Ia.

We characterize a monoclonal antibody directed against the antigen/Ia receptor of a cloned helper T cell line that induced T cell clone proliferation and T cell clone-dependent B cell proliferation at antibody concentrations as low as 10(-11) M. A Fab fragment of this antibody was not stimulatory, implicating cross-linking of antigen receptors as the primary signal for T cell activation. The Fab fragment inhibited activation of this clone by both allogeneic Ia and antigen plus self-Ia, but not by the nonspecific stimulators concanavalin A and rabbit anti-mouse brain serum. This strongly supports the hypothesis that a single molecule mediates both self-Ia plus antigen and non-self-Ia recognition. This molecule is presumably the disulfide-linked heterodimer comprised of 42,000 mol wt acidic and basic subunits precipitated by this monoclonal antibody. The cell surface and internal precursor forms of this protein are also identified. In addition, the response to allogeneic Ia stimulation was more readily inhibited by the Fab fragment than was the response to antigen plus self-Ia, suggesting that alloreactivity reflects a low affinity interaction with a ligand represented at high frequency on the stimulatory cell.

The specificity of cellular immune responses in guinea pigs. III. The precision of antigen recognition by T lymphocytes.

T lymphocytes from guinea pigs immunized with 2,4-dinitrophenyl (DNP) derivatives of mycobacteria respond to a variety of DNP conjugates. Preincubation of such cells with a given DNP conjugate under conditions which lead to the inactivation of responding cells causes a loss of the response to that conjugate, but has little effect on the response to DNP coupled to unrelated carriers. Thus, the responses of such cells to a variety of DNP conjugates can best be explained by the presence of a mixture of highly specific cells each responding to a different antigenic dterminant rather than by the presence of T cells with specificity limited to the hapten itself. Furthermore, the activity of T cells from DNP-mycobacteria-primed donors could not be blocked by a variety of nonstimulatory DNP conjugates. This suggests that while such T cells clearly recognize DNP with great precision, the receptor does not contain a very high affinity site for the hapten. A possible model for such a T-cell receptor is discussed.

Different superantigens interact with distinct sites in the Vbeta domain of a single T cell receptor.

CD4 T cell receptors (TCRs) recognize antigenic peptides presented by self major histocompatibility complex (MHC) class II molecules as well as non-self MHC class II molecules. The TCRs can also recognize endogenous retroviral gene products and bacterial toxins known collectively as superantigens (SAGs) that act mainly on the Vbeta gene segment-encoded portion of the Vbeta domain; most SAGs also require MHC II class for presentation. We have studied the interaction of the TCR from a well-characterized CD4 T cell line with SAGs by mutational analysis of its Vbeta domain. This appears to separate viral (v)SAG from bacterial (b)SAG recognition. T cells having a TCR with glycine to valine mutation in amino acid residue 51 (G51V) in complementarity determining region 2 of the TCR Vbeta domain fail to respond the bSAGs staphylococcal enterotoxin B (SEB), SEC1, SEC2, and SEC3, whereas they retain the ability to respond to non-self MHC class II molecules and to foreign peptides presented by self MHC class II molecules. It is interesting to note that T cells expressing mutations of both G51V and G53D of V beta regain the response to SEB and partially that to SEC1, but do not respond to SEC2, and SEC3, suggesting that different bacterial SAGs are viewed differently by the same TCR. These results are surprising, because it has been generally believed that SAG recognition by T cells is mediated exclusively by hypervariable region 4 on the exposed, lateral face of the TCR Vbeta domain. Response to the vSAG Mtv-7 was generated by mutation in Vbeta residue 24 (N24H), confirming previously published data. These data show that the vSAG Mtv-7 and bSAGs are recognized by different regions of the TCR Vbeta domain. In addition, various bSAGs are recognized differently by the same TCR. Thus, these mutational data, combined with the crystal structure of the TCR beta chain, provide evidence for distinct recognition sites for vSAG and bSAG.

Expression of an idiotype (Id-460) during in vivo anti-dinitrophenyl antibody responses. III. Detection of Id-460 in normal serum that does not bind dinitrophenyl.

Using an anti-idiotypic antibody previously characterized as specific for the hapten binding site of the 2,4-dinitrophenyl (DNP)-binding BALB/c myeloma protein MOPC-460, we have detected substantial amounts of this idiotype (Id-460) in the serum of normal mice. Whereas the idiotypic material in DNP-immune serum binds to DNP, the Id-460-positive material in normal mouse serum is not specific for DNP. The material in normal serum appears to be immunoglobulin. Furthermore, Id-460-positive, non-DNP-binding monoclonal immunoglobulins that completely inhibit our assay for Id-460 are repeatedly isolated when hybridomas are prepared from LPS-activated normal spleen cells. These data are interpreted in the context of Jerne's network hypothesis. It is our conclusion that the non-DNP-binding form of Id-460 is the inherited form and that this form establishes an idiotypic network favoring the production of anti-DNP bearing Id-460. Thus, the paradox of finding an inherited idiotype in the antibody response to the nonpathogen DNP may be resolved by proposing that the true form of Id-460 is specific for an environmental pathogen and that Id-460 dominance in the anti-DNP response is simply a consequence of idiotype-specific regulatory events preconditioned by Id-460-bearing immunoglobulin specific for antigenic determinants unrelated to DNP.

Exogenously provided peptides of a self-antigen can be processed into forms that are recognized by self-T cells.

Major histocompatibility complex (MHC) class II molecules can present peptides derived from two different sources. The predominant source of peptide in uninfected antigen presenting cells (APCs) is from self-proteins that are synthesized within the cell and traffic through the MHC class II compartment. The other source of antigen is endocytosed proteins, which includes both self- and foreign proteins. Foreign protein antigens generate adaptive immune responses, whereas self-peptides stabilize the MHC class II heterodimer on the cell surface, allowing positive and negative selection of thymocytes. Therefore, self-antigens play an important normal role in shaping the T cell receptor repertoire as well as a pathological role in autoimmunity. To determine whether processing and presentation of self-antigens by MHC class II molecules differs depending on whether the antigen is supplied through synthesis within the cell or by endocytosis, we used a T cell clone against an Ealpha peptide presented by I-Ab to show that processing through these two routes can differ. We also show that mice can be tolerant to the epitope formed through the endogenous route, but responsive to the epitope that can be formed through endocytosis. This suggests that negative selection occurs primarily against antigens that are synthesized within the APC, and that endocytosed self-antigens could serve as autoantigens. Finally, we also demonstrate that lipopolysaccharide-activated B cells are defective for uptake, processing, and presentation of this self-antigen, and that this correlates with the increased expression of the costimulatory molecules B7.1 and B7.2. This may provide a model for studying the onset of an autoimmune response.

Two distinct antigen-specific suppressor factors induced by the oral administration of antigen.

The feeding of sheep erythrocytes (SRBC) to mice leads to the production of two distinct T cell-derived suppressor factors by spleen cells. Each has been characterized for specificity, genetic restrictions, and cellular interactions. Fraction I has a 60,000-75,000 mol wt, is specific for antigen, and is suppressive of primary in vitro anti-SRBC responses at all times. It is not restricted by major histocompabitility complex (MHC)- or Igh-linked genes, but it fails to suppress spleen cells derived from any strain of mouse with a B10 background. It acts on an Lyt-2+ T cell to increase suppressive activity. An antiserum has been prepared against this factor that reacts with other, unrelated T cell suppressor factors. Fraction II has an approximately 30,000-40,000 mol wt, is specific for antigen, and has a dual effect on in vitro anti-SRBC responses. On day 3 of culture, it leads to augmentation of the response, whereas at day 5 it suppresses the response. It is not restricted by MHC genes, but it is restricted by Igh-linked genets. It acts by activating an Ly-1 t cell to both help and induce feedback suppression. These factors, and the antisera prepared against them, should allow more precise dissection of the molecular pathways by which immunoregulatory cells communicate with one another.

Experimental autoimmune encephalomyelitis induction in genetically B cell-deficient mice.

Experimental autoimmune encephalomyelitis (EAE) is an animal model for autoimmune central nervous system disease mediated by CD4 T cells. To examine the role of B cells in the induction of EAE, we used B10.PL (I-Au) mice rendered deficient in B cells by deletion of their mu chain transmembrane region (B10.PLmicroMT). By immunizing B10.PL and B10.PLmicroMT mice with the NH-terminal myelin basic protein encephalitogenic peptide Ac1-11, we observed no difference in the onset or severity of disease in the absence of mature B cells. There was, however, a greater variation in disease onset, severity, and especially of recovery in the B cell-deficient mice compared to controls. B10.PLmicroMT mice rarely returned to normal in the absence of B cells. Taken together, our data suggest that B cells do not play a role in the activation of encephalitogenic T cells, but may contribute to the immune modulation of acute EAE. The mechanisms to explain these effects are discussed.

T cells specific for hapten-modified self are precommitted for self major histocompatibility complex antigens before encounter with the hapten.

The technique of antigen-driven, 5-bromo-deoxyuridine and light suicide has been adapted to eliminate the precursors of cytotoxic effector cells both for alloantigen and for 2,4,6-trinitrophenyl(TNP)-modified stimulator and target cells. Using this technique, the following observations have been made. Precursors of killer cells specific for alloantigen can be suicided independently of precursors of killer cells specific for TNP-modified self cells. The loss of activity during this procedure is not due to either specific or nonspecific suppressor cells, as judged by mixing experiments. With responder cells from F1 animals, it has been possible to show that precursors specific for TNP-modified cells from one parent are suicided independently of precursors specific for TNP-modified cells of the other parent, but only if the parental strains differ in the K and D regions of the H-2 complex. Cells of F1 mice derived from K and D identical, I region different, parental strains were specifically suicided by TNP-modified stimulator cells from either parent. However, the cross-reactive killing of TNP-self targets induced by stimulation with allogeneic cells is not eliminated by first suiciding with TNP-parental cells, suggesting that the precursors of these two types of TNP-self killer cells are different. This is compatible with reported differences in their specificity, as confirmed in this report. Finally, deletion of alloreactive cells by this technique reveals little or no reactivity specific for TNP-modified allogeneic stimulator cells. In summary, these results strongly suggest that recognition of self MHC antigens is preprogrammed in peripheral T cells of normal animals, and is not acquired during the immunization process. They also suggest that cells specific for modified alloantigen are relatively rare in the strains of mice studied.

Both a monoclonal antibody and antisera specific for determinants unique to individual cloned helper T cell lines can substitute for antigen and antigen-presenting cells in the activation of T cells.

Two antisera and a monoclonal antibody raised in BALB.K mice against cloned, major histocompatibility complex (MHC)-restricted, antigen-specific helper T cell lines are described. These antibodies are specific for individual cloned T cell lines and are potent inducers of T cell proliferation. The induction of T cell proliferation by these antibodies requires the presence of an adherent accessory cell. There is no H-2 restriction between this accessory cell and the cloned T cell, nor is this antibody-induced proliferation blocked by a monoclonal anti-Fc receptor antibody. The requirement for an accessory cell, however, is eliminated in the presence of an IL-1- or IL-2-rich supernatant. Thus this system allows the analysis of helper T cell activation with only a single cell type present. Anti-T cell sera also induce T cell-dependent B cell proliferation and immunoglobulin secretion. The induction of T cell-dependent B cell activation by these sera does not require H-2-matched T cells and B cells. The specificity of these antibodies and their ability to stimulate cloned helper T cells in the absence of antigen and antigen-presenting cells strongly suggest that these antibodies are directed against antigen and/or Ia recognition sites on the T cell.

Expression of an idiotype (Id-460) during in vivo anti-dinitrophenyl antibody responses. II. Transient idiotypic dominance.

After immunization of mice with 2,4-dinitrophenyl-ovalbumin (DNP-OVA), it was shown previously that strains having Igh-Va genes and able to express light chains of the Vk1 group produce high levels of anti-DNP antibody bearing an idiotype (Id-460) associated with the combining site of the BALB/c DNP-binding myeloma protein MOPC 460. Expression of Id-460 in serum is transient; Id-460 levels peak early in the response and are regulated independently of total anti-DNP antibody. In this paper, the transient dominance of Id-460 expression has been confirmed at the cellular level by inhibition of splenic anti-DNP plaque-forming cells (PFC) with rabbit anti-Id-460 antiserum. Id-460+ PFC can account for 52-91% of anti-DNP PFC early after secondary challenge with DNP-OVA. Furthermore, Id-460 is represented at these high levels in IgM, IgG, and IgG1, and IgG2a, the three isotypes tested in the PFC assay, as well as in IgE, as tested by passive cutaneous anaphylaxis. Thus, there is no preferential association of Id-460 with a given isotype. We conclude from these studies that Id-460 is a dominant idiotype in the anti-DNP antibody response of BALB/c mice to DNP-OVA. This dominance is expressed transiently and is independent of isotype. A further conclusion from these studies is that regulation of isotype expression is independent of the regulation of idiotype expression in this system. We would suggest that regulation of Id-460 expression involves Ig-dependent helper T cells specific for Id-460 that induce Id-460+ B cells and also activate suppressor T cells, both events occurring via idiotype-anti-idiotype interactions.

The specificity of cellular immune responses II. The structure of antigenic determinants leading to T-lymphocyte stimulation.

T cells from guinea pigs immunized with the hapten 2,4-dinitrophenyl (DNP)-coupled directly to mycobacteria are of interest since they recognize and respond to DNP conjugated to many but not all carriers. The experiments reported here further analyze the structure of the complex, chemically defined antigenic determinants recognized by such T cells. These antigenic determinants can have DNP coupled either to the xi-amino group of lysyl residues or to the hydroxyl group of tyrosyl residues. Furthermore, essential contributions to the determinant recognized by such T cells are made by amino acid residues to which the hapten is not attached. Such residues are thought to be close to the hapten group itself, since introducing a small spacer between hapten and carrier prevents recognition. The hapten itself is also recognized and discriminated from other haptens with great precision by these T lymphocytes. The strain of guinea pig immunized affects the precise specificity characteristics of the responding T cells, in a way that may reflect the activity of histocompatibility-linked immune response genes. Finally, the characteristics of the immunogen have been studied. It is thought that the lipid content of the mycobacteria may be critical in inducing the hapten-reactive T cells, and this is supported by finding similar responses in T cells from guinea pigs immunized with DNP protein to which lipid has been covalently attached. Thus, the T-cell population being studied, while recognizing haptens with great precision, appears to require a larger determinant for activation than do hapten-specific B lymphocytes.

Expression of an idiotype (Id-460) during in vivo anti-dinitrophenyl antibody responses. I. Mapping of genes for Id-460 expression to the variable region of immunoglobulin heavy-chain locus and to the variable region of immunoglobulin kappa-light-chain locus.

The genetic contro of the expression of an idiotype (Id-460) associated with the 2,4-dinitrophenyl (DNP)-binding BALB/c myeloma protein MOPC 460 was studied using congenic strains of mice. It was shown that the expression of high levels of Id-460 during secondary in vivo anti-DNP-ovalbumin responses was determined by genes governing immunoglobulin heavy-chain variable and kappa-light chain variable regions (V kappa). Appropriate alleles at both loci were required for the expression of Id-460. Genes in the major histocompatability complex and the X-linked immune deficiency gene found in strain CBA/N did not greatly affect Id-460 expression. The V kappa gene controlling Id-460 expression can be differentiated from Lyt-3, and it is the first instance in which expression of an idiotype subdivides the V kappa genes associated with the Lyt-3a allele. Although it is likely that the V kappa gene(s) involved are structural, the involvememt of a regulatory gene linked to the structural gene can not be excluded.

Molecular characterization of antibodies bearing Id-460. II. Molecular basis for Id-460 expression.

Id-460+ immunoglobulins can be induced in vivo by immunization with dinitrophenyl (DNP) or P. pneumotropica and form two nonoverlapping groups of antibodies with respect to antigen binding specificity. In this study, using Id-460+ antibodies of differing antigen binding specificities, we compared on the molecular genetic level the five gene segment combinations (VH, DH, JH, VL, and JL) that encode the variable regions of these idiotype-positive immunoglobulins. The Id-460 determinant appears to be a conformational or combinatorial determinant encoded by VH460 and VK1 crosshybridizing genes. DH, JH, and JK gene segments appear to have no measurable effect upon expression of Id-460. Finally, antigen binding specificity does not appear to simply localize to any particular gene segment but may in part be the result of somatic mutation and/or VDJH junctional sequences, whose length correlates roughly with antigen binding specificity.

Surface expression of alpha 4 integrin by CD4 T cells is required for their entry into brain parenchyma.

Cloned CD4 T cell lines that recognize the Ac1-16 peptide of myelin basic protein bound to I-Au were isolated and used to analyze the immunopathogenesis of experimental autoimmune encephalomyelitis (EAE). T helper type 1 (Th1) clones induced disease, while Th2 clones did not. Using variants of a single cloned Th1 line, the surface expression of alpha 4 integrins (very late antigen 4 [VLA-4]) was identified as a major pathogenic factor. Encephalitogenic clones and nonencephalitogenic variants differ by 10-fold in their level of surface expression of alpha 4 integrin and in their ability to bind to endothelial cells and recombinant vascular cell adhesion molecule 1 (VCAM-1). The alpha 4 integrin-high, disease-inducing cloned Th1 T cells enter brain parenchyma in abundance, while alpha 4 integrin-low, nonencephalitogenic Th1 cells do not. Moreover, antibodies to alpha 4 integrin, its ligand VCAM-1, and intercellular adhesion molecule 1 all influence the pathogenicity of this encephalitogenic clone in vivo. The importance of the expression of VLA-4 for encephalitogenicity is not unique to cloned T cell lines, as similar results were obtained using myelin basic protein-primed lymph node T cells. alpha 4 integrin levels did not affect antigen responsiveness or production of the Th1 cytokines interleukin 2, interferon gamma, and lymphotoxin/tumor necrosis factor beta; and antibodies against alpha 4 integrin did not block antigen recognition in vitro. Thus, we conclude that surface expression of alpha 4 integrin is important in CD4 T cell entry into brain parenchyma. A general conclusion of these studies is that alpha 4 integrins may be crucial in allowing activated effector T cells to leave blood and enter the brain and other tissues to clear infections.

Induction of autoreactive B cells allows priming of autoreactive T cells.

A novel mechanism for breaking T cell self tolerance is described. B cells induced to make autoantibody by immunization of mice with the non-self protein human cytochrome c can present the self protein mouse cytochrome c to autoreactive T cells in immunogenic form. This mechanism of breaking T cell self tolerance could account for the role of foreign antigens in breaking not only B cell but also T cell self tolerance, leading to sustained autoantibody production in the absence of the foreign antigen.

Heat-stable antigen is a costimulatory molecule for CD4 T cell growth.

Optimal induction of clonal expansion by normal CD4 T cells requires a ligand that can engage the T cell receptor as well as functionally defined costimulatory activity on the same antigen-presenting cell surface. While the presence of effective costimulation induces proliferation, T cell receptor ligation in its absence renders T cells inactive or anergic. The molecular basis of this costimulatory activity remains to be defined. Here we describe a monoclonal antibody that can block the costimulatory activity of splenic accessory cells. Treatment with this antibody not only blocks the proliferation of CD4 T cells to a T cell receptor ligand, but also induces T cell nonresponsiveness to subsequent stimulation. Sequence analysis of the antigen recognized by this antibody indicates that it recognizes a protein that is identical to heat-stable antigen. Gene transfer experiments directly demonstrate that this protein has costimulatory activity. Thus, heat-stable antigen meets the criteria for a costimulator of T cell clonal expansion.

The mouse mammary tumor virus envelope gene product is required for superantigen presentation to T cells.

Transgenic mice expressing either the mouse mammary tumor virus (MMTV) superantigen gene (sag) alone or in combination with the viral envelope genes (env) (LEL), or all of the viral genes (gag, pol, env, and sag) (HYB PRO), deleted V beta 14+ T cells from their immune repertoire. However, only LEL or HYB PRO transgenic antigen-presenting cells were capable of stimulating a proliferative response from nontransgenic primary T cells or interleukin 2 production from a V beta 15-bearing T cell hybridoma. These T cell responses could be inhibited by a monospecific antibody directed against the MMTV gp52 cell surface glycoprotein. These results indicate that the MMTV gp52 gene product participates in the presentation of superantigen to T cells, resulting in their stimulation, a requisite step in the MMTV infection pathway. Thus, gp52 could play a role in the transfer of virus between different subsets of lymphocytes.