MicroRNA-155 influences B-cell function through PU.1 in rheumatoid arthritis

Abstract

MicroRNA-155 (miR-155) is an important regulator of B cells in mice. B cells have a critical role in the pathogenesis of rheumatoid arthritis (RA). Here we show that miR-155 is highly expressed in peripheral blood B cells from RA patients compared with healthy individuals, particularly in the IgD^-CD27^- memory B-cell population in ACPA⁺ RA. MiR-155 is highly expressed in RA B cells from patients with synovial tissue containing ectopic germinal centres compared with diffuse synovial tissue. MiR-155 expression is associated reciprocally with lower expression of PU.1 at B-cell level in the synovial compartment. Stimulation of healthy donor B cells with CD40L, anti-IgM, IL-21, CpG, IFN-α, IL-6 or BAFF induces miR-155 and decreases PU.1 expression. Finally, inhibition of endogenous miR-155 in B cells of RA patients restores PU.1 and reduces production of antibodies. Our data suggest that miR-155 is an important regulator of B-cell activation in RA.

Figure 1. CD20, CD3, CD68 and CD21 in synovial tissue from early and long-standing RA.

(a) Representative pictures of CD68 (red)/CD21 (DAB, brown) and CD20 (DAB, brown)/CD3 (red) in synovial tissue from early, long-standing RA and OA patients (original magnification, × 20); (b-c) CD20 staining score for lining and sublining areas of synovial tissue from early (n=27) and long-standing RA (n=33) patients; * CD20 staining in early RA versus OA (n=14) patients; ** CD20 staining in long-standing RA versus OA patients; (d–e) CD3 staining score for lining and sublining areas in synovial tissue from early (n=27) and long-standing RA (n=33) patients; * CD3 staining in early RA versus OA patients; ** CD3 staining in long-standing RA versus OA (n=14) patients; (f-g) CD68 staining score for lining and sublining areas in synovial tissue from early (n=27) and long-standing RA (n=33) patients; * CD68 staining in early RA versus OA (n=14) patients; ** CD68 staining in long-standing RA versus OA patients. Data are mean±s.d., *P≤0.05, **P≤0.01; Mann-Whitney U-test; ns=not significant.

Figure 2. miR-155 expression in CD19⁺ cells from RA, PsA and HC.

(a) miR-155 expression in peripheral blood (PB) CD19⁺ cells isolated from RA (n=24) and HC (n=9); *P<0.0001 RA versus HC; Mann-Whitney U-test; (b) FACS analysis of B-cell subpopulations in PB of RA (n=24) and HC (n=9); *P=0.02 and **P=0.05 RA versus HC; Mann-Whitney U-test; (c) miR-155 expression in CD19⁺ cells isolated from PB of RA patients based on disease duration and ACPA positivity; * P=0.032 ACPA^- ERA versus HC; **P=0.0004 ACPA⁺ ERA versus HC and ***P=0.002 ACPA⁺ ERA versus ACPA^- ERA; ^P=0.04 ACPA^- LSRA versus HC; ^^P=0.0002 ACPA⁺ LSRA versus HC and ^^^P=0.002 ACPA⁺ LSRA versus ACPA^- LSRA; § P=0.04 ACPA⁺ ERA versus ACPA⁺ LSRA; Data are shown as mean±s.d., 25–75 percentiles; Mann-Whitney U-test; (d) miR-155 expression in paired samples (n=10) of CD19⁺ cells isolated from PB and synovial fluid (SF) of RA patients; *P=0.04 SF CD19⁺ cells versus PB CD19⁺ cells; Wilcoxon test; (e) miR-155 expression in sorted B-cell subpopulations from PB of RA patients (n=6) and HC (n=6) in three independent experiments. *P=0.02; Mann-Whitney U-test; (f) miR-155 expression in PB-derived CD19⁺ cells of RA and PsA patients; * P=0.032 ACPA^- ERA versus HC; **P=0.0004 ACPA⁺ ERA versus HC and ***P=0.002 ACPA⁺ ERA versus ACPA^- ERA; ^P=PsA versus HC; ^^P=0.25 ACPA^- ERA versus PsA; §P=0.01 ACPA⁺ ERA versus PsA; data are shown as mean±s.d., 25–75 percentiles; Mann-Whitney U-test.

Figure 3. miR-155 expression in synovial RA B cells and in synovial tissue from RA and OA.

(a–c) Double immunofluorescence staining of miR-155 (green) and CD20 (red) in RA synovial tissue with follicular histological pattern. (a) Representative picture for grade 3 lymphocyte aggregate (original magnification × 40; orange=double positive cells) with upper insert showing control staining (scramble probe/isotype) and a bottom insert showing × 100 magnification of miR-155⁺CD20⁺ cell; n=5. (b) Representative pictures of the distribution of miR-155 (green channel) and (c) CD20 (red channel) are shown. (d) Representative picture of miR-155 in situ hybridization on synovial tissue from RA with follicular (n=5) and diffuse (n=5) pattern (5-bromo-4-chloro-3′-indolyphosphate, purple); (e) qPCR assessing miR-155 expression in synovial tissue from RA with follicular (n=10) and diffuse (n=10) pattern and OA patients (n=5). *P=0.03 diffuse RA versus OA; **P=0.001 diffuse versus follicular RA; §P<0.0001 follicular RA versus OA patients. Data are shown as mean±s.d., 25–75 percentiles; Mann-Whitney U-test. (f) miR-155 expression in CD19⁺ cells isolated from PB and SF of RA patients with different synovial pattern; *P=0.01 PB CD19⁺ cells from follicular versus diffuse RA; **P=0.001 PB CD19⁺ cells from follicular RA versus HC; ^P=0.01 SF CD19⁺ cells from follicular versus diffuse RA; ^^P=0.001 SF CD19⁺ cells from follicular versus HC; ^^^P=0.02 SF CD19⁺ cells from follicular RA versus PB CD19⁺ cells from follicular RA; data are shown as mean±s.d., 25–75 percentiles; Mann-Whitney U-test. DAPI, 4′,6-diamidino-2-phenylindole.

Figure 4. miR-155 in CD19⁺ cells is induced by inflammatory and maturation factors.

miR-155 expression in PB CD19⁺ cells isolated from HC (n=5) after in vitro stimulation with CD40L (2 μg/ml), IL-21 (50 ng/ml), anti-IgM (10 μg/ml), CpG (1 μg/ml), Interferon alpha (400 IU/ml), IL-6 (30 ng/ml) and BAFF (20 ng/ml) for 24–72 h; *P<0.05. Data are shown as mean±s.d.; Mann-Whitney U-test. The data were generated in three independent experiments.

Figure 5. PU.1 expression in CD19⁺ cells of RA patients.

(a) Expression of PU.1 in CD19⁺ cells isolated from PB and SF of RA patients compared to HC, **P=0.002 SF-derived CD19⁺ cells from ACPA⁺ RA versus HC; *P=0.001 SF-derived CD19⁺ cells from ACPA^- RA versus HC; and §P=0.001 SF-derived CD19⁺ cells from ACPA⁺ RA versus PB-derived CD19⁺ cells from ACPA⁺ RA; data are shown as mean±s.d., 25–75 percentiles; Mann-Whitney U-test. (b) Expression of PU.1 in CD19⁺ cells in paired RA PB and SF samples; *P=0.002; Wilcoxon test; (c) Inverse correlation between miR-155 and PU.1 expression in SF derived CD19⁺ cells in RA patients [r= −0.65; P=0.04; 95%CI (−0.38 to −0.88)]; Spearman Rank Correlation coefficient; (d) PU.1 expression in PB CD19⁺ cells isolated from HC (n=5) after in vitro stimulation with CD40L (2 μg/ml), IL-21 (50 ng/ml), anti-IgM (10 μg/ml), CpG (1 μg/ml), Interferon alpha (400 IU/ml), IL-6 (30 ng/ml) and BAFF (20 ng/ml) for 24–72 h; *P<0.05 CD40L, anti-IgM, IL-21, CpG, IFN-alpha, BAFF or IL-6 versus Baseline RPMI; data are shown as mean±s.d.; Mann-Whitney U-test.

Figure 6. PU.1 expression in synovial tissue and synovial CD20⁺ cells of RA patients.

(a–d) PU.1 Immunohistochemical staining (DAB) in synovial tissue of RA patients. (a) Representative PU.1 staining (Brown-DAB) in synovial tissue with diffuse pattern; (b) Representative PU.1 staining (Brown-DAB) in mononuclear cells in synovial lining of synovial tissue with diffuse pattern; (c,e) CD20 (DAB) and CD3 (RED) staining (original magnification × 20) and (d,f) PU.1 (DAB) staining in lymphocyte synovial follicules (original magnification × 20); (g) Representative immunofluorescence staining for PU.1 expression (TRITC, red) in synovial CD20⁺ cells (FITC, green) in RA patients (original magnification × 20); (h,i) PU.1 expression (TRITC) in CD20⁺ cells (FITC) within lymphocyte synovial aggregates (original magnification × 40); (j) PU.1 expression (TRITC) in the synovial lining layer (original magnification × 40). orange = double positive cells.

Figure 7. Inhibition of miR-155 in RA PB B cells reduces antibody production.

(a) Inhibition of miR-155 in RA PB B cells derepresses PU.1. PB CD19⁺ cells from RA patients (n=6) were transfected with miR-155 inhibitor or control inhibitor (CI) using the Neon Transfection System. Expression of PU.1, IRF4, PRDM1, PAX5 and housekeeping 18S were determined by qPCR 48 h later. Data are presented as mean±s.e.m. of biological replicates. *P<0.05 anti-miR-155 versus CI; Wilcoxon test; (b) miR-155 inhibition reduces IgG production by RA B cells. RA PB CD19⁺ cells (n=5) were transfected with miR-155 inhibitor or control mimic and stimulated with plate bound human CD40 ligand (2 μg/ml) in the presence of BAFF (100 ng/ml), IL-21 (50 ng/ml) and anti-IgM F(ab)2-fragments (5 μg/ml) or left in the presence of BAFF and IL-6 (spontaneous production of antibodies). Each condition was performed in n=10 (spontaneous) and n=14 replicates (stimulated). * P<0.05 stimulated miR-155 inhibitor transfected cells versus stimulated control transfected cells;≠P<0.05 stimulated versus spontaneous production of antibodies; Wilcoxon test. The data was generated in three independent experiments.

Figure 8. Rationale for the use of antagomiR-155 in the treatment of RA synovitis.

(a) miR-155 control of SHIP-1 and PU.1 is crucial for the regulation of inflammatory response in macrophages and for antibody production and terminal differentiation of B cells; (b) In RA synovial compartment, macrophages and B cells are characterized by higher expression of miR-155 due to endogenous TLR ligands and other soluble factors, for example, CD40L, IL21, CpG, IFN-alpha, IL-6 and BAFF. This leads to the downregulation of SHIP-1 and PU.1 expression and thus to the amplification of the inflammatory cascade by macrophages and increase in antibody production and B-cell maturation; (c) Thus, the administration of antagomiR-155 could restore the normal expression of SHIP-1 and PU.1 acting both on macrophages and B cells and facilitate the resolution of autoimmunity and inflammation in RA. SHIP-1, inositol polyphosphate-5-phosphatase; PU.1, Spi-1 proto-oncogene; TLR, Toll like receptor BAFF, B-cell activating factor; CD40L, CD40 ligand, BCR, B-cell receptor.