Poly-guanosine motifs costimulate antigen-reactive CD8 T cells while bacterial CpG-DNA affect T-cell activation via antigen-presenting cell-derived cytokines.

Pathogen-derived pattern recognition ligands like lipopolysaccharide (LPS) and bacterial cytidine-guanosine (CpG)-DNA not only activate dendritic cells and macrophages but are also mitogenic for B cells. Less clear are the claimed effects of CpG-DNA on T cells, which range from direct activation, costimulation, or indirect transient activation via antigen-presenting cell (APC)-derived interferon type I (IFN type I). Here we demonstrate that CpG-DNA sequence specifically triggers macrophages to produce IFN type I, interleukin (IL)-12, IL-6 and tumour necrosis factor (TNF), but lacks the ability to directly costimulate T cells. Strikingly, poly-guanosine (poly-G) extensions to CpG-containing oligonucleotides (ODN) abolished the macrophage stimulatory potential yet generated T-cell costimulatory activities. In fact, independently of CpG-motifs, poly-G-ODN displayed the ability to costimulate T cells. Costimulation was operative on CD8 T cells but not CD4 T cells. Poly-G-mediated costimulation resulted in IL-2-driven T-cell proliferation and induced cytolytic T cells. Overall the data imply that poly-G motifs costimulate antigen reactive CD8 T cells, while CpG-DNA motifs fail to do so but may affect T-cell activation via APC derived cytokines such as IFN type I.

CpG-oligodeoxynucleotides co-stimulate primary T cells in the absence of antigen-presenting cells.

CpG-containing oligodeoxynucleotides (CpG-ODN) act as powerful adjuvant during in vivo induction of T cell responses. While CpG-ODN directly activate antigen-presenting cells (APC) and thus exert an extrinsic activity on T cells, it is unclear whether they directly affect T cells (intrinsic activity). Here we analyze the effects of CpG-ODN on T cells in an APC-free cell culture. We report that CpG-ODN co-stimulate T cells provided they were triggered via their TCR. CpG-ODN induced IL-2 production, IL-2 receptor expression and thus proliferation. Proliferation was blocked by cyclosporin A or anti-IL-2 monoclonal antibodies (mAb) but not by anti-IL-4 mAb. Moreover, CpG-co-stimulated T cells differentiated into cytolytic T lymphocytes in vitro. Of note, IL-2-driven growth of primed T cells was not affected by CpG-ODN. Co-stimulation was also operative in T cells from CD28-/- mice and in TCR-transgenic T cells stimulated with peptide. CpG-ODN-mediated co-stimulation of T cells in vitro may thus explain part of the potent adjuvant effects of CpG-ODN in vivo.

Superantigen and endotoxin synergize in the induction of lethal shock.

Endotoxin (lipopolysaccharide; LPS) and superantigens (exotoxins) have been identified as potent inducers of lethal shock. While endotoxin primarily interacts with CD14 receptors on macrophages, superantigens like the staphylococcal enterotoxin B (SEB) preferentially activate T cells. Both cell types are triggered to release pro-inflammatory cytokines that in turn induce lethal shock. We analyzed whether endotoxin and superantigen interact during the induction phase of lethal shock. We report that LPS and SEB operate synergistically. Lethal doses of both inducers were reduced 100-fold when given in combination. The induced serum levels of tumor necrosis factor, interleukin-6, and interferon-gamma (IFN-gamma) were elevated and remained high for a prolonged period. Moreover, synergistic action of LPS and SEB induced lethal toxic shock even without presensitization of mice with D-galactosamine (D-GalN). Opposed to D-GalN-pretreated mice, mice injected with LPS and SEB showed less liver damage, but rather apoptosis of epithelial cells in the bowel. Cyclosporin A and treatment with anti-IFN-gamma monoclonal antibody blocked the synergistic action of LPS and SEB, indicating that T cell-derived IFN-gamma is the mediator of the observed synergism. Concomitant injection of LPS and SEB had no influence on SEB-induced T cell deletion and anergy induction. Since Gram-positive and Gram-negative bacteria can be recovered from septic blood samples, the synergistic action of endotoxin and superantigens might be relevant during lethal septicemia.

Superantigen-reactive T cells that display an anergic phenotype in vitro appear functional in vivo.

Clonal deletion and/or inactivation establishes tolerance to self antigens. Endogenous and exogenous (bacterial) superantigens, like the staphylococcal enterotoxins, induce ligand-specific clonal anergy in vivo and thus are believed to mirror aspects of post-thymic tolerance mechanisms in mature peripheral T cells. Here we analyzed the level of anergy of ligand-responsive V beta 8+ T cells from staphylococcal enterotoxin B (SEB)-primed mice in vivo and in vitro. Upon in vitro restimulation with SEB, CD4+V beta 8+ and CD8+V beta 8+ T cells failed to produce IL-2. However, functional IL-2 receptors were triggered, since supplementation with IL-2 induced clonal growth in virtually all CD4+V beta 8+ and CD8+V beta 8+ T cells as determined by limiting dilution analyses. Thus in vitro unresponsiveness of lymphocytes from SEB-primed mice reflects the inability of SEB-reactive V beta 8+ T cells to produce IL-2. Surprisingly, anergy as defined in vitro was at variance with that in vivo. Following further challenge with SEB, systemic and acute lymphokine production (including IL-2 and tumor necrosis factor) occurred with almost identical peak values and kinetics to primary in vivo responses, and D-galactosamine-sensitized mice succumbed to lethal shock. Polymerase chain reaction analyses revealed that CD4+V beta 8+ expressed IL-2-specific mRNA in vivo upon restimulation with SEB. While lymphokine production and expression of the IL-2 receptor was similar to the response to in vivo primary stimulation, only CD8+V beta 8+ T cells expanded clonally upon reintroduction of SEB in vivo. Hence primed V beta 8+ T cells challenged with SEB display in vitro anergy yet in vivo responsiveness, at least in part. We conclude that the state of anergy is reversible, dependent upon the quality of activation signals provided in in vivo rather than in in vitro culture conditions.

Induction of unresponsiveness to the superantigen staphylococcal enterotoxin B: cyclosporin A resistant split unresponsiveness unfolds in vivo without preceding clonal expansion.

The continuous presence of antigen and powerful immune responses (exhaustive cell proliferation) of ligand reactive T cells are currently thought to condition clonal deletion and/or induction of unresponsiveness to endogenous or exogenous superantigens (SAg). Here we report that in vivo induction of unresponsiveness to the SAg staphylococcal enterotoxin B (SEB) can be an immediate process. Within ours a large portion of ligand reactive V beta 8+ T cells becomes clonally deleted by apoptosis. In parallel, the remaining V beta 8+ T cells are unresponsive to SEB, yet at the same time express functional IL-2 receptors (IL-2R) and thus are highly responsive to the growth promoting effects of IL-2. In a subsequent step refractory IL-2R+V beta 8+ T cells undergo a wave of cell proliferation for 48 h, presumably driven by IL-2. Thereafter a large proportion of V beta 8+ T cells succumb to apoptosis, the remaining cells display the hallmarks of split unresponsiveness, i.e. they display a selective failure to produce IL-2 upon SEB stimulation in vitro combined with a preserved capability to express functional IL-2R. Early deletion and induction of unresponsiveness to SEB are cyclosporin A (CsA) resistant, while clonal expansion with subsequent cell deletion is blocked by CsA, yet the development of split unresponsiveness is not impaired by CsA. The results suggest that IL-2 driven growth of refractory T cells may mimic powerful immune responses of ligand reactive V beta 8+ T cells. Since unresponsiveness to SEB precedes in vivo expansion, the results as such question the concept of 'exhaustive cell proliferation' as a prerequisite for induction of unresponsiveness.(ABSTRACT TRUNCATED AT 250 WORDS)

Murine CD4+ and CD8+ T cells are activated by the superantigen (SA) staphylococcal enterotoxin B (SEB) and exhibit MHC-unrestricted cytotoxicity.

Resting CD8+ and CD4+ murine T cells become efficiently activated by the superantigen (SA) staphylococcal enterotoxin B (SEB). Limiting dilution experiments reveal that roughly every third CD8+ or CD4+ T cell could be induced to proliferate. Surprisingly, besides CD8+ T cells, also CD4+ cells acquire a high cytolytic activity upon activation with SEB. Both subpopulations only lyse MHC class II positive target cells in the presence of SEB. However the recognition of SEB by CTL is independent of the haplotype of the MHC, i.e. is MHC-unrestricted. In addition the recognition of SEB is independent of the coreceptor molecules CD4 and CD8, respectively.

Cyclosporine A prevents the generation of single positive (Lyt2+ L3T4-, Lyt2- L3T4+) mature T cells, but not single positive (Lyt2+ T3-) Immature thymocytes, in newborn mice.

The influence of cyclosporine A (CsA) on T-cell maturation was investigated in newborn mice. CsA treatment during the pre- and postnatal periods resulted in a hypoplasia of peripheral lymphatic organs, and absence of mature T3+ T cells in lymph nodes and spleens; no functional T-cell reactivity was observed. In thymuses of CsA-treated mice, no T3+ single positive Lyt2+ or T3+L3T4+ thymocytes could be found, but double positive (DP) cells were readily detected. A thymocyte subset with the phenotype Lyt2+L3T4-T3- was still discernible; this population was non-functional in vitro. The data show that the maturation of single positive (SP) T cells is critically influenced by CsA; under the conditions used here we found no evidence that 'leaky' autoreactive SP T cells develop in CsA-treated newborn mice.

Cyclosporine A prevents the generation of single positive mature T cells in newborn mice.

The effect of Cyclosporin A (CsA) during T cell development was investigated in newborn mice. CsA treatment completely blocked the generation of peripheral single positive (SP) mature T cells: the lymphatic tissues were hypoplastic. However, double negative (DN) T3 expressing lymphocytes were still detectable. Thymuses from CsA-treated mice lacked the SP L3T4(CD4)+ subset, DN and double positive (DP) thymocytes were still present. We further defined a SP Lyt2(CD8)+ thymic subpopulation which lacked CD3 expression and displayed no functional activity in vitro. Thus, CsA critically interferes with the maturation of SP T lymphocytes; we found no evidence for 'leaky' autoreactive peripheral SP T cells in CsA-treated newborn mice.