Cytotoxic T lymphocyte-associated antigen-4-Ig (CTLA-4-Ig) suppresses Staphylococcus aureus-induced CD80, CD86, and pro-inflammatory cytokine expression in human B cells

Abstract

Background: Cytotoxic T lymphocyte-associated antigen-4-Ig (CTLA-4-Ig) competes with CD28 for binding CD80/CD86 on antigen-presenting cells (APCs) to limit T cell activation. B cells are believed to be important APCs in the pathogenesis of autoimmune diseases and express CD80/CD86 after activation; however, relatively little is known about the effect of CTLA-4-Ig on B cells. This study tested the impact of CTLA-4-Ig on human B cell responses.

Methods: Human blood B cells were purified from healthy donors and activated in the presence of CTLA-4-Ig or the L6-Ig control protein in vitro. RT-q-PCR and immunofluorescence staining were performed to detect activation marker expression. ELISA was conducted to measure cytokine secretion. The CD80/CD86 levels on the surface of the memory B cells in the blood of 18 patients with rheumatoid arthritis (RA) were detected using immunofluorescence staining.

Results: CTLA-4-Ig suppressed the expression of Staphylococcus aureus (SAC)-induced CD80, CD86, TNFA, and IL6 in human B cells at the transcriptional level. Furthermore, CTLA-4-Ig concomitantly decreased SAC-induced CD80/CD86 surface expression on and TNF-α and IL-6 secretion from B cells. On the other hand, T cell-dependent (TD) stimulation-induced B cell activation, proliferation, plasma cell differentiation, and antibody secretion were not affected by CTLA-4-Ig. As expected, TD stimulation-induced surface CD80 was hindered by CTLA-4-Ig. Notably, a blockade of CD80/CD86 on the surface of the memory B cells was observed in the patients with RA after abatacept (CTLA-4-Ig) treatment. In a portion of the RA patients, restoration of CD80/CD86 staining on the surface of the memory B was detected starting in the 3rd month of abatacept treatment. Interestingly, the surface levels of CD80/CD86 on the patients' memory B cells positively correlated with disease activity.

Conclusions: We found that CTLA-4-Ig directly suppressed SAC-induced B cell activation in vitro. Obstruction of CD80 and CD86 on the surface of the memory B cells was detected in the RA patients after abatacept treatment. Blocking CD80/CD86 on B cells by CTLA-4-Ig may hinder T cell activation and associated with the disease activity of RA in vivo. Our findings indicate that CTLA-4-Ig may regulate humoral responses by modulating B cell activation and interfering T cell-B cell interaction.

Keywords: Abatacept; CD80; CD86; CTLA-4; IL-6; Rheumatoid arthritis; TNF-α.

Fig. 1

CTLA-4-Ig binds human memory and activated B cells. a CTLA-4-Ig binds peripheral blood memory B cells. Human peripheral blood mononuclear cells (PBMCs) were stained with 10 μg/ml biotinylated-CTLA-4-Ig or control-protein (Ctrl-Ig), followed by streptavidin-PE together with anti-CD20 and anti-CD27 antibodies and analyzed using flow cytometry. Left histograms, one representative binding of CTLA-4-Ig gated on CD20⁺CD27⁻ naïve B cells (bottom) or on CD20⁺CD27⁺ memory B cells (top) is shown (n = 3). Gray peaks, Ctrl-Ig; blue line, CTLA-4-Ig. Middle and right histograms, anti-CD80 or anti-CD86 antibody was used to examine CD80 and CD86 expression on naïve and memory B cells. Gray peak, isotype control; blue line, anti-CD80 or anti-CD86 antibody. b CD19⁺ B cells were purified from the blood of healthy donors and the purity was determined by anti-CD20 antibody staining. On representative result is shown. c CTLA-4-Ig binds activated B cells. Purified CD19⁺ B cells from blood were stimulated with anti-IgM and anti-CD40 antibodies for 3 days and the levels of CTLA-4-Ig binding (left), CD80 (middle), and CD86 (right) were examined as described in a. Gray peaks, Ctrl-Ig (10 μg/ml) or isotype control; blue line, CTLA-4-Ig (10 μg/ml) or anti-CD80/86 antibody. Data are representative of three independent experiments

Fig. 2

CTLA-4-Ig suppresses Staphylococcus aureus-induced CD80 and CD86 levels on B cells in vitro. a Purified CD19⁺ B cells were stimulated with SAC (Staphylococcus aureus Cowan strain, SAC) or anti-IgM (5 μg/ml) plus anti-CD40 (1 μg/ml) antibodies in the presence of CTLA-4-Ig (100 μg/ml) or Ctrl-Ig (100 μg/ml) for 2 days. The expression of CD80, CD86, and CD69 was measured by RT-q-PCR, and GAPDH level was used to normalize the readout. After deducing the value of the medium control, the value of the control-Ig-treated B cells obtained from each donor was considered to be 100% for calculation of the relative level. Each dot represents one donor (n = 7~8). b The effect of CTLA-4-Ig on the level of CD80, CD86, or CD69 on the surface of the B cells was analyzed using flow cytometry (n = 4). The percentage of positive cells in each experiment is shown. c CD19⁺ B cells were stimulated as described in a. After acidic elution to remove bound CTLA-4-Ig, the CD80 and CD86 on the surface of the treated cells were examined using immunofluorescence staining. The percentage of positive cells in each experiment is shown (n = 4)

Fig. 3

CTLA-4-Ig suppresses SAC-induced TNF-α and IL-6 production in B cell. a Purified B cells were stimulated with SAC or anti-IgM plus anti-CD40 antibodies in the presence of CTLA-4-Ig (100 μg/ml) or Ctrl-Ig (100 μg/ml) for 8 h (TNFA) or 48 h (IL6 and LTA), and gene expression was measured as described in Fig. 2b. b Concentration of TNF-α or IL-6 in the supernatant of the SAC-stimulated B cells was analyzed by ELISA. c CTLA-4-Ig suppressed proinflammatory cytokine production by naïve and memory B cells. Sorted naïve and memory B cells were stimulated with SAC in the presence of CTLA-4-Ig or Ctrl-Ig for 8 h (TNFα) or 48 h (IL-6), and the supernatant was collected for ELISA (n = 6). The readout for the Ctrl-Ig treatment results was considered 100% for the calculation of the relative level in each donor

Fig. 4

CTLA-4-Ig has no effect on SAC- and TD-induced proliferation of human B cells. a CTLA-4-Ig did not affect T-dependent or SAC-induced B cell proliferation. Fifty thousand purified B cells per well were seeded in a 96-well plate and stimulated with SAC or anti-IgM (5 μg/ml) plus anti-CD40 (1 μg/ml) antibodies in the presence of CTLA-4-Ig (100 μg/ml) or Ctrl-Ig (100 μg/ml) for 3 days, and then 1 μCi [³H]-thymidine was added to each well and incubated for an additional day before being harvested (n = 7). The readout of the medium only group was used to calculate the incorporation (fold) of [³H]-thymidine for each donor sample. The bar graph shows the cumulative incorporation (fold) of the samples from 7 donors. b CTLA-4-Ig had no effect on naïve or memory B cell proliferation. Sorted naïve and memory B cells were stimulated with SAC in the presence of CTLA-4-Ig or Ctrl-Ig, and [³H]-thymidine uptake was measured. The bar graph shows the cumulative incorporation in the samples of 3 donors

Fig. 5

CTLA-4-Ig has no effect on T cell-dependent stimulation- or SAC-induced plasma cell differentiation or antibody secretion. Purified B cells were stimulated with anti-IgM (5 μg/ml), anti-CD40 (1 μg/ml) antibody, and IL-21 (100 ng/ml), or with SAC in the presence of Ctrl-Ig or CTLA-4-Ig for 4 days, and the cells (a and c) were stained with anti-CD27 and anti-CD38 antibodies to mark plasma cells (CD38^highCD27^high; n = 3). One representative result is shown. b ELISPOT assay results after TD stimulation. Two thousand B cells (activated for 4 days) were seeded on a 96 well plate precoated with anti-hIgG antibody in triplicate, and the cells were incubated 16 h at 37 °C in complete RPMI medium. After removing the cells, anti-hIgG antibody conjugated with biotin was added to mark antibody secreting cells (ASC). The mean spot number was quantified by an AID ELISPOT reader and analyzed by AID ELISPOT software. Upper, one representative result. Lower, cumulative ELISPOT assay results of samples from 3 donors. d IgM secretion after SAC stimulation for 3 days. IgM in the supernatant of the activated B cells was measured using ELISA

Fig. 6

Levels of CD80/CD86 on the surface of the memory B cells correlate with disease severity in the RA patients. a Surface level of CD80 or CD86 on peripheral blood memory B cells of RA patients during abatacept treatment was monitored for 4 months (n = 18). PBMCs from the RA patients were stained with anti-CD20 and anti-CD27 antibodies in the presence of anti-CD80 or anti-CD86 antibodies. The analysis of CD80 or CD86 level was gated on memory (CD20⁺CD27⁺) cells. The surface level (ΔMFI) was calculated as the mean fluorescence intensity level of CD80/CD86 minus the isotype control. b Comparison of the responses of the RA patients with or without decreased CD80 or CD86 levels on the surface of the memory B cells. EULAR, European League Against Rheumatism collaborative initiative classification criteria. The patients were classified as good responders, moderate responders, and non-responders according to the DAS28-ESR levels and their improvement from the baseline 6 months after abatacept treatment