The relative contribution of the CD28 and gp39 costimulatory pathways in the clonal expansion and pathogenic acquisition of self-reactive T cells.

The zona pellucida (ZP), an ovarian extracellular structure, contains three major glycoproteins: ZP1, ZP2, and ZP3. A ZP3 peptide contains both an autoimmune oophoritis-inducing T cell epitope and a B cell epitope that induces autoantibody to ZP. This study investigates two major T cell costimulation pathways in this disease model. Herein we show that blockage of glycoprotein (gp)39 and CD40 interaction with gp39 monoclonal antibody (mAb) results in the failure to induce both autoimmune oophoritis and autoantibody production. Inhibition of ligand binding to the CD28 receptor with the fusion protein, murine CTLA4-immunoglobulin (Ig), also results in failure to generate antibody to ZP and significantly reduces disease severity and prevalence. Surprisingly, the frequencies of antigen-specific T cells in anti-gp39 mAb-treated mice, CTLA4-Ig treated mice, and in mice given control hamster IgG or control fusion protein L6, were equivalent as determined by limiting dilution analysis (approximately equals 1:5,000). These T cells, which produced comparable amounts of interleukin 4 and interferon gamma in vitro, were able to transfer oophoritis to normal recipients. When anti-gp39 mAb and CTLA4-Ig were given together, the effect was additive, leading to inhibition of T cell activation as determined by in vitro proliferation and limiting dilution analysis (approximately equals 1:190,000); disease and antibody responses were absent in these mice. By studying these two costimulatory pathways in parallel, we have shown that autoimmune disease and autoantibody production are inhibitable by blocking either the gp39 or the CD28 pathway, whereas inhibition of clonal expansion of the effector T cell population occurs only when both pathways are blocked.

Adoptive transfer of natural killer cell activity in B6D2F1 mice challenged with Salmonella typhimurium.

The purpose of this study was to extend our previous findings as to the role of murine NK cells in host protection to a challenge infection with virulent Salmonella typhimurium SR-11. B6D2F1 mice were depleted of NK cells with anti-asialo GM1 or a monoclonal antibody, anti-NK 1.1, followed by a salmonellae challenge. Significantly decreased numbers of splenic bacteria (P less than 0.005) in the NK cell-depleted mice were note at 12, 24, and 48 hr postchallenging, compared to the sham-injected control animals. When Percoll gradient-enriched large granular lymphocytes (NK cells) were adoptively transferred to NK cell-depleted mice followed by challenging, the splenic bacterial numbers were comparable to those present in NK cell-intact, control mice. These data indicate that large granular lymphocytes (NK cells) are responsible for the down-regulation of the protective host response in mice challenged with the facultative intracellular parasite. S. typhimurium.

Natural killer cell activity against uninfected and Salmonella typhimurium-infected murine fibroblast L929 cells.

Splenic and peritoneal cells from uninfected mice exhibited selective cytotoxic activity against Salmonella-infected L929 fibroblast cells. Although the level of killing was low, the results were significantly different from the killing of uninfected L929 cells (p < 0.01) by either effector cell population. Salmonella typhimurium-activated splenic and peritoneal exudate (PE) cells exhibited enhanced killing of Salmonella-infected L929 targets and YAC-1 cells compared to endogenous natural killer activity by splenic and peritoneal cells from uninfected mice. When Salmonella-infected (60 h postinfection) or uninfected mice received an i.p. injection of anti-asialo GM1 12 h prior to harvesting the splenic or peritoneal cells (72 h postinfection), target cell killing was < 2%. In all cases, activated PE cells exhibited the greatest killing activity against uninfected and Salmonella-infected L929 fibroblast cells and YAC-1 targets (p < 0.01).

Neonatal injection of an ovarian peptide induces autoimmune ovarian disease in female mice: requirement of endogenous neonatal ovaries.

Neonatal female mice injected with the self ZP3 peptide are not tolerant to the peptide; they develop autoimmune ovarian disease (AOD) and autoantibody response 5 weeks later. ZP3 challenge leads to severe AOD and ZP3-specific T cell and antibody responses. In contrast, neonatal tolerance to foreign ZP3 peptide is established in male mice: ZP3 peptide-specific T cell proliferative response is reduced and AOD is absent in ovarian grafts. Tolerance is associated with a Th2-dominant T cell cytokine and antibody isotype profiles. As controls, neonatal tolerance to foreign peptides, with Th2 deviation, was induced in both male and female mice. Endogenous ZP3 is important for the gender difference. Ablation of ovaries in female mice on days 2 and 5, but not on day 7 or 14, switches the ZP3 autoimmune response to a tolerogenic response with a concomitant change in cytokine profile. Thus, neonatal self ZP3 peptide, supported by endogenous ovaries within a neonatal time window, evokes a pathogenic autoimmune response.

Bacterial and retroviral superantigens share a common binding region on class II MHC antigens.

Staphylococcal enterotoxin A (SEA), one of the most potent T-cell mitogens known, has been classified as a bacterial superantigen on the basis of ability to stimulate V beta-specific T-cell subsets. SEA interacts with class II major histocompatibility complex (MHC) antigens on antigen-presenting cells and the T-cell antigen receptor (TCR) on T cells, resulting in a ternary complex of MHC-SEA-TCR. Mls antigens are known to be products of mouse mammary tumour virus (MMTV), and it has been reported that two exogenous strains of MMTV encode retroviral superantigens in the open reading frames of the 3' long terminal repeat of the viral genome; however, no binding of the putative MMTV superantigen to either MHC antigens or TCR has been demonstrated. Here we use synthetic peptides to identify a site on the MMTV-1 superantigen that binds to class II MHC antigens. The site is encompassed by amino-acid residues 76-119 of the MMTV-1 superantigen. Direct binding and competition experiments show that the MMTV superantigen and SEA bind to at least one common region on class II MHC antigens.

Agonist properties of a microbial superantigen peptide.

Staphylococcal enterotoxin A (SEA) binds to class II major histocompatibility complex (MHC) molecules and stimulates monocytes to produce tumor necrosis factor alpha (TNF alpha) and interleukin one (IL-1). We have examined the monocyte stimulatory activity of individual synthetic peptides encompassing the entire sequence of the SEA molecule. Only one peptide, SEA(121-149), induced both TNF alpha and IL-1 production at a concentration as low as 30 microM. Consistent with its effects on monocyte function, SEA(121-149) was shown to bind directly to class II MHC molecules on the surface of both monocytes and B cells, and its binding was inhibited specifically by native SEA. Further, polyclonal antibody to SEA(121-149) inhibited induction of TNF alpha by both SEA and toxic shock syndrome toxin one. Thus, we have identified SEA(121-149) as a peptide agonist of SEA monocyte stimulatory activity.

Dynamic regulation of gastric autoimmunity by thyroid hormone.

Neonatal thymectomy (NTX) of BALB/c mice between days 1 and 3 post-birth leads to a high incidence of autoimmune gastritis involving T cells and autoantibodies. Although progress has been made in understanding various immunological events associated with that disease process, much remains to be learned about the regulatory factors which control autoimmune gastritis. In this study we have examined the potential for neuroendocrine hormones of the hypothalamus-pituitary-thyroid (HPT) axis to alter the outcome of experimental autoimmune gastritis in NTX mice. As reported here, thyroxine administered to young adult NTX mice during an active phase of disease development dramatically reduced the incidence of gastritis and the overall severity of disease, and lowered the levels of anti-parietal cell antibodies. In contrast, treatment of young NTX mice with thyroxine or other HPT hormones prior to the onset of disease had no beneficial effect on gastritis, but instead resulted in significantly higher anti-parietal cell antibody levels compared to non-hormone-treated NTX mice or to NTX mice treated as adults. Reverse transcriptase spectratype analyses of 22 Vbeta gene junctional sites revealed pronounced oligoclonality and limited junctional diversity in stomach lymphocytes from untreated NTX mice compared to normal mice or thyroxine-treated NTX mice. These findings identify a set of dynamic immune-endocrine interactions, linked to critical development-dependent events, that are involved in the expression and regulation of peripheral autoimmunity.

Mapping of multiple binding domains of the superantigen staphylococcal enterotoxin A for HLA.

Multiple binding sites on the staphylococcal enterotoxin A (SEA) molecule which interact with class II MHC Ag have been suggested by previous studies comparing SEA binding with that of another superantigen, toxic shock syndrome toxin-1. Using the synthetic peptide approach we have identified multiple regions of the SEA molecule which are responsible for binding to HLA Ag on Raji cells. Overlapping peptides were synthesized corresponding to the complete amino acid sequence of SEA: SEA(1-45), SEA(39-66), SEA(62-86), SEA(83-104), SEA(102-124), SEA(121-149), SEA(146-173), SEA(166-193), SEA(187-217), and SEA(211-233). Like the native SEA molecule, all of the peptides exhibited relatively high beta-sheet and low alpha-helical structure as determined by circular dichroism spectroscopy. A direct competition assay was employed with peptide blockage of 125I-SEA binding to MHC Ag. SEA(1-45), SEA(39-66), SEA(62-86), and SEA(121-149) but none of the other peptides blocked binding to Raji cells. The relative potency of the peptides in blocking SEA binding was determined with SEA(39-66) much greater than SEA(1-45) = SEA(62-86) = SEA(121-149). Peptide competition was seen at concentrations as low as 55 microM. Further, antibodies were produced to all of the peptides and tested for their ability to bind to SEA and inhibit SEA binding to HLA. Consistent with the direct inhibition of binding, antisera to SEA(1-45), SEA(39-66), and SEA(62-86) reduced the ability of SEA to bind Raji cells, whereas, antisera to the remaining peptides failed to block binding. The data suggest that the binding of the superantigen SEA to MHC molecules involves several N-terminal regions on SEA as well as an additional internal domain. This allows for the presence of multiple binding sites in an extended N-terminal region of the SEA molecule or a discontinuous binding epitope.

Binding site on the murine IFN-gamma receptor for IFN-gamma has been identified using the synthetic peptide approach.

We have studied the structural parameters involved in the binding of murine IFN-gamma (MuIFN-gamma) to its receptor. Ten synthetic overlapping peptides corresponding to the extracellular domain of the MuIFN-gamma receptor (MuIFN-gamma R) were synthesized. In direct binding studies, biotinylated MuIFN-gamma bound specifically to receptor peptide (95-120). Further, the NH2-terminal IFN-gamma peptide, MuIFN-gamma (1-39), also specifically bound to receptor peptide (95-120). Binding of both labeled MuIFN-gamma and MuIFN-gamma (1-39) to MuIFN-gamma R peptide (95-120) was inhibited by either unlabeled molecule. The COOH-terminal receptor binding peptide, MuIFN-gamma (95-133), neither bound to any receptor peptides nor blocked the binding of intact MuIFN-gamma or MuIFN-gamma (1-39) to receptor peptide (95-120). Polyclonal antibodies to each of the peptides were then produced. Each of the anti-peptide antisera recognized its corresponding peptide and bound denatured cloned soluble receptor by Western blotting. Furthermore, the antisera to peptides representing the inner region of the extracellular domain of the receptor bound to nondenatured soluble MuIFN-gamma R. Specifically, antisera to the receptor peptides (73-97), (95-120), (118-143), (142-163), and (161-182) bound to soluble MuIFN-gamma R, whereas antisera to peptides (1-21), (20-49), (46-74), (178-203), and (202-227) did not bind. Most important, antisera to peptides (95-120) and (118-143) competed with [125I]MuIFN-gamma for binding to soluble receptor. These results show that the region of the MuIFN-gamma R encompassing amino acid residues (95-120) is a binding site on the receptor for the NH2-terminal of MuIFN-gamma by direct binding, and that the larger region (95-143) on the receptor may play a role in binding of intact MuIFN-gamma based on blocking of binding by site-specific antibodies.

The N-terminus and C-terminus of IFN-gamma are binding domains for cloned soluble IFN-gamma receptor.

The mechanism of binding of murine IFN-gamma to its receptor has not been determined. We have studied this mechanism by examining the binding of overlapping synthetic peptides of IFN-gamma to cloned soluble murine IFN-gamma R. IFN-gamma (1-39) and IFN-gamma (95-133) were able to compete with [125I]IFN-gamma for binding to cloned soluble receptor. Peptides corresponding to the inner region of IFN-gamma--IFN-gamma (36-60), IFN-gamma (54-91), and IFN-gamma (78-107)--showed a markedly reduced ability to compete with [125I]IFN-gamma for receptor binding relative to the N-terminal and C-terminal peptides. In direct binding studies, the binding of [125I]-IFN-gamma (1-39) to soluble receptor could only be competed by IFN-gamma (1-39) and IFN-gamma and not by any of the other peptides including IFN-gamma (95-133). This suggests that the N- and C-termini of IFN-gamma bind to different regions of the receptor. These data in conjunction with previous structure/function studies and x-ray crystallographic data have allowed us to formulate a "velcro-key" model of IFN-gamma binding to receptor that involves both the N- and C-terminal domains. The N-terminus binds in the classical "lock-and-key" manner characterized by specific ligand-receptor binding. The hydrophilic C-terminus binds to a region of the receptor distinct from the N-terminus likely through the polycationic region, which is conserved across species barriers. Binding of this type would exhibit high affinity and low specificity similar to a piece of velcro. This interaction becomes specific when the C-terminus is in the context of the whole IFN-gamma molecule and may act to increase the affinity of receptor binding and/or facilitate signal transduction.