Lipopolysaccharide-dependent down-regulation of CD27 expression on T cells activated with superantigen

Abstract

To investigate the mechanisms underlying T-cell responses during superantigen (SAg) stimulation, we analysed the effects of SAg on CD27 expression with or without lipopolysaccharide (LPS) as a novel regulator of T-cell function. CD27 is expressed on the majority of resting peripheral blood T cells (CD27^low). Activation of T cells by SAg induces high levels of CD27 surface expression (CD27^high) accompanied with simultaneous CD30 receptor expression. After prolonged activation in vitro, the level of CD27 expression became intermediate. The effects of LPS on down-regulation of CD27^high expression on CD30⁺ T cells were dose-dependent. Separating LPS-stimulated monocytes from T cells by mechanical dispersion abolished its inhibitory effect, indicating the requirement for direct interactions between monocytes and T cells. We also found that SAg up-regulated CD80 expression on CD14⁺ monocytes and LPS inhibited SAg-induced CD80 expression after 24 hr of stimulation. Up-regulation of CD152 (CTLA-4) was selective, since it was found to be preferentially expressed on the CD30⁺ population. Competitive experiments using soluble blocking peptides showed that addition of CD28 or CD80 peptide recovered LPS-induced down-regulation of CD27^high expression on CD30⁺ T cells. These observations suggested that the presence of low levels of CD80 on monocytes may partially inhibit CD27 expression due to inefficient delivery of positive signals via CD28/CD80 interaction, and that the increased levels of CD80 enhance the inhibition through interactions with CD152 which is expressed at the highest levels after 48 hr of activation.

INTRODUCTION

Staphylococcus aureus and Streptococcus pyogenes produce a large family of exotoxins,¹ which have been implicated in the pathogenesis of a number of diseases.^2,3 Included within this group of proteins are the staphylococcal enterotoxins (SEA, SEB) and streptococcal pyrogenic exotoxins (SPEA, SPEC). These toxins are prototypic superantigens (SAg), which stimulate large populations of T cells expressing particular T-cell receptor (TCR) β-chain variable gene segments (Vβ) and activate cytokine release from T cells and monocytes/macrophages in a major histocompatibility complex (MHC)-dependent but unrestricted manner.² Recently, we identified three novel SAg (CAP, SPM and SPM-2) from S. pyogenes products, and studies of the immunological reactivities of these SAg are currently in progress.^4–8 On the other hand, bacterial lipopolysaccharide (LPS) constitutes a component of the outer membrane of Gram-negative bacteria and activates monocytes/macrophages. Recent studies have shown that induction of proliferation and cytokine production of human T cells can be stimulated by LPS or LPS partial structure.^9,10 Although the unique modulation of T-cell functions thought to be triggered by these bacterial products is quite informative, the actual events are poorly understood. One element controlling T-cell functions that has not been studied with regard to these modulations is surface changes in the activation molecules, expression of which is induced or up-regulated after T-cell activation.

One interesting class of activation molecule is the newly defined tumour necrosis factor (TNF) receptor family that includes TNF receptors I and II, OX40, CD30, CD40, Fas, 4-1BB and CD27 (reviewed in refs ¹¹ and ¹²). CD27 is a disulphide-linked 120 000 MW type I transmembrane glycoprotein that is expressed on both CD4⁺ and CD8⁺ resting peripheral T lymphocytes, on mature thymocytes, and on subsets of natural killer (NK) and B cells.^13–18 As CD27 expression is markedly up-regulated after activation with anti-CD3 antibody or mitogenic lectins, CD27 has been classified as a T-cell activation antigen that amplifies proliferative responses of activated T cells.^13,14,19 In addition to CD27, CD30 is also an inducible costimulatory receptor of T cells, and is preferentially expressed by human CD4⁺ and CD8⁺ clones with T helper type 2 (Th2) cytokine profile.²⁰ These members of the TNF receptor family are likely candidates as regulators of T-cell activation.

Since regulation of costimulatory receptors on human peripheral T cells by SAg and LPS is a less well-known phenomenon and may be of relevance during bacterial infection, we compared CD27 and CD30 expression on freshly isolated and SAg-activated T cells. We report here evidence that the SAg up-regulates CD27 expression on T cells, and addition of LPS with SAg down-regulates CD27 expression on CD30⁺ T cells. Furthermore, we showed that costimulatory signals through CD28 and/or CD152 are required for down-regulation of CD27 expression by LPS.

MATERIALS AND METHODS

Monoclonal antibodies and reagents

Anti-CD27 (M-T271, Ancell Co., Bayport, MN) and anti-CD30 (Ki-1, Ancell Co.) monoclonal antibodies (mAb) were purchased from the sources shown. Anti-CD8 (NU-Ts/c) and anti-CD28 (KOLT-2) mAb were purchased from NICHIREI Co. (Tokyo, Japan). Anti-CD3 (Leu-4), anti-CD4 (Leu-3a) and anti-CD14 (Leu-M3) mAb were purchased from Becton Dickinson Immunocytometry Systems (Mountain View, CA). Anti-CD25 (33B3.1) mAb was obtained from Immunotech (Marseille, France). The mAb against CD80 (BB1), CD86 (IT2.2) and CD152 (CTLA-4, BNI3.1) were obtained from PharMingen (San Diego, CA). Fluorochrome-conjugated goat anti-mouse immunoglobulin G (IgG) was obtained from Southern Biotech (Birmingham, AL). SEB, phytohaemagglutinin (PHA), concanavalin A (Con A) and LPS (from Escherichia coli 055:B5) were purchased from Sigma Chemical Co. (St. Louis, MO). Recombinant interleukin-2 (rIL-2) was a generous gift from Shionogi Pharmaceutical Co. (Osaka, Japan), and rIL-12 and neutralizing anti-IL-12 mAb were gifts from Genetics Institute (Cambridge, MA). Recombinant (r) TNF-α, rIL-1α, rIL-4 and rIL-6 were purchased from R & D systems (Minneapolis, MN). Natural interferon-γ (IFN-γ) was supplied by Hayashibara Biochem. Labs. Inc. (Okayama, Japan).

Preparation of SPM-2

Purified SPM-2 was obtained from the culture supernatant of S. pyogenes strain T12 as previously described.⁷ Briefly, the supernatant was precipitated with 60% ammonium sulphate. The desalted sample was applied to a diethylaminoethyl (DEAE)–cellulose column. The active fractions in passing through the column, were purified by preparative isoelectric focusing (mature protein: 223 amino acids, manuscript in preparation).

Cell isolation and activation

Peripheral blood mononuclear cells (PBMC) obtained from normal donors were isolated from heparinized venous blood by density gradient sedimentation over Lympholyte-H (Cedarlane Labs, Hornby, Ontario, Canada). Cells were then washed three times in phosphate-buffered saline and resuspended in medium, as appropriate, for either cell culture or immunofluorescence staining of freshly isolated cells. For cell culture studies, PBMC (1×10⁶ cells/ml) were cultured in RPMI-1640 supplemented with 10% heat-inactivated fetal calf serum and 2 mm l-glutamine. Cells were cultured in 96-well culture plates (Falcon, Becton Dickinson) at a final volume of 200 μl. Cells were activated in the presence of either Con A (10 μg/ml), SEB (1 μg/ml) or SPM-2 (1 μg/ml) for various periods (usually 3 days), washed, and resuspended in staining buffer for immunofluorescence staining. In the CD27 induction studies, IL-1α (5 ng/ml), IL-2 (100 U/ml), IL-4 (1000 U/ml), IL-6 (5 ng/ml), TNF-α (5 ng/ml), IFN-γ (500 U/ml), or IL-12 (5 ng/ml) and occasionally anti-IL-12 mAb (10 μg/ml) were added.

Assay for proliferative response

In proliferation assays, the capacity of PBMC (1×10⁶/ml) cultured in RPMI-1640 medium supplemented with 10% of autologous or pooled human sera to respond to the given stimulus was determined by adding 0·2°μCi/well of [³H]thymidine during the last 8 hr of culture.

Effects of LPS on SPM-2 stimulation

PBMC (1×10⁶/ml) were treated with SPM-2 in the presence of LPS from E. coli in concentrations ranging from 0·001 to 1000 ng/ml, or were treated with LPS for 24 hr, followed by washing, and then treated with SPM-2. For studies of LPS receptors, PBMC were preincubated with anti-CD14 mAb (MY4, Coulter, Miami, FL) or control mouse IgG at 1:500 dilution for 1 hr and then stimulated with SPM-2 plus LPS. Cells were removed and scored for viability by trypan blue exclusion assay. Viable T-cell blasts were washed and analysed for expression of CD27 and related molecules (CD25, CD28, CD30 and CD152) by flow cytometry. CD14⁺ cells were stained with anti-CD80/CD86 mAb.

Blocking studies with cell interaction

In experiments designed to block the interaction between monocytes and T cells, PBMC were stimulated with SPM-2 plus LPS and the clustered cells were dispersed 10 times with a pipette tip, followed by additional culture. Cells were also stimulated with SPM-2 plus LPS in the presence of synthetic blocking peptides (CD28_21–40 and B7-1_40–59 of precursors) or CD21_1067–1086 peptide of precursor as a negative control at 5 μg/ml (Santa Cruz Biotechnology, Santa Cruz, CA). After culture for 72 hr, CD3⁺ T cells were analysed by staining with anti-CD27 and anti-CD30 mAb.

Flow cytometry

The methods used for direct and indirect immunofluorescence staining in conjunction with single- or two-colour flow cytometry and for counting of fluorescein-positive cells were described previously.²¹ For triple staining, cells were incubated with fluorescein isothiocyanate (FITC)-, phycoerythrin (PE)- and peridinin chlorophyll protein (PerCP)-conjugated mAb at saturating concentrations at 4° for 30 min, and subsequently washed twice. CD27 expression was then analysed on a FACScalibur (Becton Dickinson). Appropriately labelled, isotype-matched immunoglobulins were used as controls for non-specific fluorescence. Gates were set around the lymphocyte population, and a total of 20 000 cells was analysed. Blast cells exhibiting a larger size were determined by an arbitrary gate that distinguished resting from activated cells.

RESULTS

Increase of CD27 expression on SAg-activated T cells

To determine whether SAg induced CD27 receptor expression, we performed flow cytometric analysis for the expression of CD27 on CD3⁺ T cells within freshly isolated PBMC preparations from healthy donors and on T cells after 72 hr of in vitro incubation with Con A or SAg (SEB, SPM-2). As shown in Fig. 1(a), the majority (90%) of resting peripheral blood T cells weakly expressed CD27 (CD27^low), whereas activated T cells with high levels of CD25 (Fig. 1b) appeared to express high levels of CD27 (CD27^high) within a few days. Percentages of SPM-2-induced CD27^high T cells ranged from 25 to 50% of the CD3⁺ population after 3–4 days of stimulation. The percentage of CD27^high T cells did not substantially change after 4 days of in vitro stimulation between Con A and SAg. Kinetic analysis demonstrated that the peak of CD27^high in SPM-2-stimulated T cells occurred from day 3–4, and the level of CD27 expression shifted to intermediate between high and low by day 12 (Fig. 2). The intensity of expression in the CD27^low population also shifted to intermediate by day 12. The percentage of CD27⁻ T cells was low throughout the entire culture period (Fig. 2). Both the CD4⁺ and CD8⁺ T-cell subsets participated in this induction of CD27^high, showing significant increases in percentage of CD27^high in each subset at day 3 as compared to day 0 (data not shown).

Figure 1.

Flow cytometric analysis of CD27 expression. PBMC (1×10⁶/ml) were cultured with Con A (10 μg/ml), SEB (1 μg/ml), or SPM-2 (1 μg/ml) for 4 days and stained with FITC-conjugated anti-CD27 and PE-conjugated anti-CD3 mAb (a), or FITC-conjugated anti-CD25 and PE-conjugated anti-CD27 mAb (b). The percentages of respective cell types in the CD3⁺ lymphocytes (a) or in total lymphocytes (b) are shown. Profiles in (a) illustrate results of the highest frequency of CD27^high T cells after stimulation.

Figure 2.

Kinetics of enhanced CD27 expression after T-cell activation. PBMC (1×10⁶/ml) were cultured with Con A (10 μg/ml), SPM-2 (1 μg/ml), or medium alone. Cells were harvested at various time-points, and level of CD27 expression of the CD3⁺ lymphocytes was measured.

CD27^high expression on SAg-induced CD30⁺ T cells

Very low expression of CD30 was observed on resting lymphocytes, while CD30⁺ T-cell number increased after SAg activation (approximately 20% of lymphoblasts after 72 hr of culture) and decreased from day 4 after activation (H. Rikiishi et al. unpublished observations). Two-colour immunofluorescence analysis (Fig. 3) showed simultaneous expression of CD27 and CD30 receptors on cells stimulated with SAg compared with those stimulated with Con A; a number of CD30⁺ cells were CD27^high and, conversely a proportion of CD27^high cells expressed CD30, although a small subset expressing CD30⁺ and CD27^low was also detected. Taken together, these results indicated that CD27 and CD30 receptors may in some instances be expressed on the same cell population, although certain populations showed a mutually exclusive pattern of CD27 and CD30 expression.

Figure 3.

Flow cytometric analysis of coexpression of CD27^high and CD30. PBMC (1×10⁶/ml) were cultured with Con A (10 μg/ml), SEB (1 μg/ml), SPM-2 (1 μg/ml), or medium alone for 72 hr and stained with mouse anti-CD30 mAb, followed by addition of FITC-conjugated goat anti-mouse IgG, and PE-conjugated anti-CD27 mAb. Numbers represent the percentages of cells expressing both CD27^high and CD30 of the total lymphocyte population.

Effects of LPS on SAg-induced CD27 expression

Since LPS could regulate the functions of T cells through primed monocytes,¹⁰ we tested the effects of LPS on the expression of CD27^high by SAg-activated T cells (Table 1). LPS in combination with SPM-2, did not inhibit proliferation, as measured by [³H]thymidine uptake on day 4 of culture. When LPS (1 μg/ml) was added to SPM-2-stimulated cultures, the percentage of CD27^high T cells decreased in the CD30⁺ T-cell subpopulation. However, the percentage of CD27^high T cells in the CD30⁻ subpopulation was similar to that in cultures treated with SPM-2 alone. We added graded doses of LPS, and the percentages of CD27^high T cells with CD30⁺ phenotype decreased in a dose-dependent manner, with a significant effect at 1 ng/ml of LPS (data not shown). PBMC preincubated for 24 hr at 37° with LPS and then washed are still able to induce inhibition of T-cell activation in the absence of additional LPS (data not shown). Pretreatment of PBMC with anti-CD14 mAb (MY4) slightly recovered the decrease in number of CD27^high T cells by LPS (Fig. 4). Next, we investigated the roles of monokines in substituting for LPS-primed monocytes in induction of CD27^high T cells by SAg. PBMC were stimulated with SPM-2 or SEB in the presence of IL-1, IL-6, IL-12, or TNF-α. None of the monokines examined showed any significant enhancement or inhibition of SAg-induced CD27^high expression. Nor were any effects observed following addition of IL-2, IL-4, or IFN-γ (data not shown).

Table 1.

Effects of LPS on induction of CD27^high T cells by SPM-2

Figure 4.

Effects of blocking of cell interactions on induction of CD27^high CD30⁺ T cells. PBMC (1×10⁶/ml) were stimulated with SPM-2 (1 μg/ml) plus LPS (1 μg/ml) for 48 hr, and the clustered cells were dispersed with a pipette tip (P), followed by culture for a further 24 hr. Cells were also preincubated with anti-CD14 mAb (MY4) or control mouse IgG for 1 hr and then stimulated with SPM-2 plus LPS for 72 hr. Cells were stained with mouse anti-CD30 mAb, followed by addition of FITC-conjugated goat anti-mouse IgG, PE-conjugated anti-CD27 mAb and PerCP-conjugated anti-CD3 mAb. The SPM-2-induced CD27^high CD30⁺ T-cell population is expressed as 100%, and the effects of treatments are shown as percentages of the CD27^high CD30⁺ population. SPM-2-induced CD27^high CD30⁺ T-cell populations ranges from 8·2% to 10·5%. A representative result from three independent experiments is shown.

Effects of LPS on CD80 expression

The failure of monokines to substitute for LPS indicated that direct interactions between T cells and accessory monocytes are necessary for inhibition of CD27^high expression by LPS. To determine the direct interactions, these reaction mixtures were subjected to mechanical dispersion after stimulation. Figure 4 shows that mechanical dissociation after 48 hr of stimulation completely blocked the inhibitory effect of LPS on induction of CD27^high CD30⁺ T cells, but not after 24 hr of stimulation. This finding prompted us to compare the expression of CD80 on monocytes with or without LPS treatment. As shown in Fig. 5(a), CD80 was not expressed, or was only expressed at low levels, on resting or monocytes (CD14⁺) treated with LPS alone. After culture of PBMC for 24 hr in SPM-2 alone, significant induction of CD80 expression was observed. In LPS-treated cultures, SPM-2-induced expression of CD80 was markedly decreased at 24 hr and then increased up to 72 hr. In contrast to CD80, the majority of resting monocytes expressed CD86, and no significant changes were observed in CD86 expression on monocytes treated with a combination of SPM-2 and LPS (data not shown).

Figure 5.

Expression of costimulatory molecules on responsive cells. PBMC (1×10⁶/ml) were cultured in medium alone or in that containing SPM-2 (1 μg/ml) with or without LPS (1 μg/ml). Each day, cells were stained with FITC-conjugated anti-CD80 and PE-conjugated anti-CD14 mAb (a), FITC-conjugated anti-CD3 and PE-conjugated anti-CD152 mAb (b), or PE-conjugated anti-CD28 and PerCP-conjugated anti-CD3 mAb (c). Cells were gated for CD14⁺ monocytes (a), CD3⁺ blasts (b), or CD3⁺ cells (c).

Preferential expression of CD152 on SAg-activated CD30⁺ T cells

CD152, a close relative of CD28, on T cells with its ligands CD80 and/or CD86 on monocytes is considered to be a key molecule for delivering the negative signals necessary for inhibition. Anti-CTLA-4 mAb was used to assess CD152 expression on freshly isolated and activated T cells (Fig. 5b). Although CD152 was not detectable on freshly isolated T cells, it was induced on T blasts (48 hr after stimulation) by addition of SPM-2 to PBMC, and its expression declined to half of the maximal level 72 hr after stimulation. CD28 expression was not markedly altered at significant levels by stimulation (Fig. 5c). When stimulated with SPM-2 for 48 hr, CD152 was selectively expressed at significant levels by CD30⁺ T cells (Fig. 6). Furthermore, to investigate the roles of CD28/CD152 and CD80 during LPS inhibition on SAg-induced CD27^high expression, competitive experiments were performed with soluble CD28 or CD80 peptide. As shown in Fig. 7, SPM-2-induced CD27^high CD30⁺ T cells were inhibited by addition of CD28 or CD80 peptide. LPS-induced inhibition was blocked by these peptides, especially CD28 peptide, while no effect was observed by CD21 peptide as a negative control (data not shown).

Figure 6.

Expression of CD152 on activated CD30⁺ T cells. PBMC (1×10⁶/ml) were cultured with Con A (10 μg/ml), or SPM-2 (1 μg/ml) for 48 hr and stained with FITC-conjugated anti-CD30 and PE-conjugated anti-CD152 mAb. Data represent the percentages of CD152⁺ T cells in CD30⁺ or CD30⁻ blasts. A representative result from three independent experiments is shown.

Figure 7.

Effects of blocking peptides on induction of CD27^high CD30⁺ T cells. PBMC (1×10⁶/ml) were cultured with SPM-2 (1 μg/ml) or SPM-2 plus LPS (1 μg/ml) in the presence or absence of blocking peptides (CD80, CD28; 5 μg/ml) for 72 hr. Cells were stained for the presence of CD27^high CD30⁺ T cells. The SPM-2-induced CD27^high CD30⁺ T-cell population is expressed as 100%, and the effects of treatments with LPS and/or blocking peptides are shown as percentages of the CD27^high CD30⁺ T-cell population. SPM-2-induced CD27^high CD30⁺ T-cell populations ranged from 9·9% to 15·5%. A representative result from three independent experiments is shown.

DISCUSSION

Staphylococcal and streptococcal toxins engage class II MHC antigens on mononuclear phagocytes and act as SAg that powerfully stimulate T cells expressing specific TCR Vβ gene segments and induce the release of cytokines.^2,22 Here, we provided evidence for yet another mechanism of SAg-mediated modulation of the immune response; surface changes on SAg-activated T cells. Our results indicated that SAg (SEB and SPM-2) stimulate T cells and up-regulate expression of CD27 and CD30 receptors on activated T cells. Furthermore, we showed that surface expression of CD27 is in part correlated with the expression of CD30 on SAg-activated T cells. Interestingly, addition of LPS with SPM-2 down-regulated expression of CD27 on CD30⁺ T cells in a dose-dependent manner in contrast with the up-regulation observed with SPM-2 alone. Thus, the available data suggest that LPS-mediated signals regulate overactivation of the responsive T cells by signalling through the TCR by SAg.

We examined the regulation of CD27 membrane expression subsequent to T-cell activation. CD27 becomes constitutively expressed on CD3⁺ T cells (CD27^low), and its expression is enhanced upon activation. After activation with SAg, a number of T cells converted to the CD27^high phenotype, although prolonged incubation of CD27^high T cells resulted in down-regulation of CD27 expression (Fig. 2). Activation of T cells via the TCR/CD3 complex is known to induce significant increases in CD27 and CD30 membrane expression.^13,19 Here, we demonstrated that on peripheral blood T cells the CD27^low phenotype is not stable and can be converted into an intermediate phenotype via CD27^high after T-cell activation. This response may contribute to down-regulation of cellular responses by reducing the number of receptors on the cell surface, indicating that the CD27^intermediate phenotype is a reflection of differentiation rather than of activation. In addition, costimulation with SAg plus LPS caused down-regulation of the CD27^high phenotype, but did not impair T-cell proliferation or survival (Table 1). Down-regulation of CD27 on activated T cells by LPS-mediated signals is intriguing since LPS signals are generally associated with positive, rather than negative, regulation of monokines or monocyte- and B-cell-accessory molecules.²³ Several monokines secreted by activated monocytes are known to stimulate T-cell function (reviewed in ref. 24). Thus, we investigated whether monokines could be involved in the regulation of CD27 surface expression on SAg lymphoblasts. However, we found no regulatory role for IL-1, IL-6, IL-12, or TNF-α in the regulation of CD27 expression. No significant effects were observed following addition of IL-2, IL-4, or IFN-γ derived from TCR/CD3 triggering. Further investigations of the roles of cytokines are necessary for the report that cytokines provide the signal for CD27 transcription.²⁵ When the physical interaction between T cells and monocytes was prevented by mechanical dispersion after 48 hr of stimulation, LPS could no longer induce inhibition of the surface expression of CD27 on T cells (Fig. 4), implying that a direct cell-to-cell contact between T-cell and monocyte is required for this inhibition. In addition, even PBMC preincubated with LPS followed by washing out showed inhibition (data not shown). Thus, monocytes during priming by LPS seem to deliver a costimulatory signal via accessory molecules for T-cell regulation.

It is also possible that the increased expression of accessory molecules such as B7 induced by LPS is sufficient to regulate T-cell function. Indeed, there is a clear correlation between CD80 (B7-1) expression on monocytes after LPS stimulation and their capacity to regulate T-cell functions.⁹ We compared the expression of CD80 on monocytes from healthy donors with and without LPS stimulation, and surprisingly found no significant induction of CD80 expression on monocytes by LPS (Fig. 5a). It has recently been shown that healthy donors can be separated into two groups, with regard to the induction of CD80 expression by LPS; responders and non-responders.⁹ The results presented here were compatible with non-responders. The expression of CD80 was only induced through monocyte activation by SAg stimulation, but was inhibited by addition of LPS. The inhibition of CD80 expression 24 hr after addition of LPS resulted in inefficient delivery of positive signals via CD28/CD80 interaction. On the other hand, CD86 was constitutively present on resting monocytes, and no significant changes were induced in its expression by SPM-2 or LPS stimulation. When CD152 (CTLA-4) appeared on activated T cells, the expression of CD80 recovered on LPS-treated monocytes. CD152 on activated T cells plays a role in negative regulation of T-cell activation; that is, CD152 may scavenge B7 ligands, rendering them unable to bind CD28 and thus reducing T-cell responses.²⁶ Indeed, CD152 expression is induced after protein kinase C activation via phorbol myristate acetate stimulation, which causes marked down-regulation of CD27.^19,27 Our studies with a mAb to CD152 demonstrated surface expression on activated T cells after 48 hr of SAg-activation, although CD28 expression was not markedly altered by stimulation. CD152 was found to be preferentially expressed on the CD30⁺ subset (Fig. 6). CD30 signalling can down-regulate CD28 expression within 48 hr comparable to up-regulation of CD152 expression on T cells.^28,29 In addition, CD30 expression on activated T cells is higher in Th2 than Th1 clones.²⁰ Recently, it has been reported that the level of expression of CD152 is much higher in Th2 than Th1 clones, suggesting that Th2 cells use common molecules for signal transduction.³⁰ Therefore, higher levels of expression of CD152 in CD30⁺ T cells might have resulted in these cells being much more sensitive to CD152-dependent down-regulation of CD27^high expression.

We investigated the roles of these costimulatory signals during LPS-induced down-regulation of CD27 expression using soluble blocking peptides. As shown in Fig. 7, CD27^high expression on CD30⁺ T cells induced by SAg was inhibited by CD28 or CD80 peptide, while down-regulation of CD27^high expression by LPS was restored by CD80 peptide. Unexpectedly, the addition of CD28 peptide significantly affected down-regulation by LPS. The most likely explanation for this result is that CD28 either provides effective adhesion for T-cell–monocyte interactions, thereby also permitting CD152 interactions, or CD28 signals are in some way required to promote CD152 expression. These experiments indicated that superantigenic stimulation of T cells is dependent on B7 costimulatory interactions and that costimulatory signals through CD28 and/or CD152 are required for down-regulation of CD27 expression by LPS. These observations are at least partially consistent with those of a previous study of the T-cell response against the bacterial superantigen SEB in mice lacking CD28 in which CD28 costimulation was necessary for T-cell activation after stimulation with SEB.³¹ The results of the present study suggested that the LPS released from the Gram-negative flora of the host’s gastrointestinal tract may induce down-regulation of receptors expressed on SAg-activated T cells, which results in inability to perceive overactivation signals via the CD27–CD27 ligand (CD70) interaction. These results suggested the existence of additional suppressive factors masking the ligand specificity of the active state induced by SAg. Thus, it is likely that under particular conditions LPS prevents SAg-driven aberrant responses of T-cell-dependent immune systems.

Abbreviations

CAP

cytoplasmic membrane-associated protein

LPS

lipopolysaccharide

PBMC

peripheral blood mononuclear cells

SAg

superantigen

SEB

staphylococcal enterotoxin B

SPM

Streptococcus pyogenes mitogen

SPM-2

Streptococcus pyogenes mitogen-2

TCR

T-cell receptor

Th

T helper