Production of RANKL by Memory B Cells: A Link Between B Cells and Bone Erosion in Rheumatoid Arthritis

Abstract

Objective: Rheumatoid arthritis (RA) is a systemic autoimmune disease that often leads to joint damage. The mechanisms of bone damage in RA are complex, involving activation of bone-resorbing osteoclasts (OCs) by synoviocytes and Th17 cells. This study was undertaken to investigate whether B cells play a direct role in osteoclastogenesis through the production of RANKL, the essential cytokine for OC development.

Methods: RANKL production by total B cells or sorted B cell subpopulations in the peripheral blood and synovial tissue from healthy donors or anti-cyclic citrullinated peptide-positive patients with RA was examined by flow cytometry, real-time polymerase chain reaction, enzyme-linked immunosorbent assay, and immunohistochemical analysis. To define direct effects on osteoclastogenesis, B cells were cocultured with CD14+ monocytes, and OCs were enumerated by tartrate-resistant acid phosphatase staining.

Results: Healthy donor peripheral blood B cells were capable of expressing RANKL upon stimulation, with switched memory B cells (CD27+IgD-) having the highest propensity for RANKL production. Notably, switched memory B cells in the peripheral blood from RA patients expressed significantly more RANKL compared to healthy controls. In RA synovial fluid and tissue, memory B cells were enriched and spontaneously expressed RANKL, with some of these cells visualized adjacent to RANK+ OC precursors. Critically, B cells supported OC differentiation in vitro in a RANKL-dependent manner, and the number of OCs was higher in cultures with RA B cells than in those derived from healthy controls.

Conclusion: These findings reveal the critical importance of B cells in bone homeostasis and their likely contribution to joint destruction in RA.

Figure 1

B cells produce RANKL upon stimulation. Total peripheral blood mononuclear cells were isolated from the whole blood of a healthy donor and cultured in the presence of anti-CD40 and phorbol 12-myristate 13-acetate (PMA) for 48 hours. Medium alone was used as a control. A, Cells were collected and surface expression of RANKL was analyzed by flow cytometry (right), with live B cells gated based on CD19 expression (left). B, Left, Representative histograms show the percentages of RANKL+ B cells under stimulated conditions or in cultures with fluorescent-minus-one (FMO) or medium alone as controls. Right, Dot plots show the percentages of RANKL+ B cells after stimulation or with medium alone. Symbols represent individual samples; horizontal lines with bars show the mean ± SEM. C, B cells were purified with magnetic beads conjugated with anti-CD19 antibody and cultured with the indicated stimuli or left unstimulated (unstim) as control. RANKL mRNA expression was detected by real-time polymerase chain reaction (PCR) after 48 hours of culture. D, Levels of RANKL and osteoprotegerin (OPG) mRNA were measured by quantitative PCR after 48 hours of culture under stimulated or unstimulated conditions. Bars show the mean ± SEM fold expression relative to unstimulated controls, in 8–13 samples per group in C and 4 samples per group in D. E, Levels of soluble RANKL (sRANKL) in culture supernatants were measured by enzyme-linked immunosorbent assay after 7 days of culture under stimulated conditions or with medium alone as control. Symbols represent individual samples; horizontal lines with bars show the mean ± SEM. * = P < 0.05 by unpaired t-test; ** = P < 0.01 and **** = P < 0.0001 by Wilcoxon’s signed rank test. IL-6 = interleukin-6; TNFα= tumor necrosis factor; NS = not significant.

Figure 2

Memory B cells from patients with rheumatoid arthritis (RA) produce the highest levels of RANKL compared to other B cell sub-populations. B cells isolated from RA patients were separated into switched memory (SM), unswitched memory (USM), double-negative (DN), and naive B cell subsets, based on expression of CD27 and IgD. Each subset was stimulated to produce RANKL. A, Percentages of RANKL+ cells were compared between the different B cell subsets. B, Percentages of RANKL+ switched memory B cells were compared between RA patients and healthy controls (HC). In A and B, symbols represent individual samples; horizontal lines with bars show the mean ± SEM. C, Percentages of RANKL+ B cells were compared between CD95+ and CD95 − subsets. Bars show the mean ± SEM of 9 samples per group. * = P < 0.05 by unpaired t-test; ** = P < 0.01 by Wilcoxon’s signed rank test; *** = P < 0.001 by one-way analysis of variance with Tukey’s multiple comparison test.

Figure 3

Synovial B cells from patients with rheumatoid arthritis (RA) spontaneously express RANKL. Freshly isolated mononuclear cells from RA synovial fluid were stained for B cell markers and RANKL (ex vivo) to examine spontaneous RANKL expression by B cells. Cells were also stimulated with anti-CD40 and phorbol 12-myristate 13-acetate (PMA) for 48 hours, and then analyzed for RANKL expression by flow cytometry. A, Left, Representative histograms showing enrichment of B cells with the switched memory (SM) (CD27+IgD−) phenotype in RA synovial fluid, as compared to that in RA peripheral blood B lymphocytes (PBL; CD19+). Right, Percentages of RA synovial fluid B cells in the switched memory (SM), unswitched memory (USM), double-negative (DN), and naive B cell phenotypes. B, Representative dot plots showing the percentages of RANKL+ B cells in RA synovial fluid compared to fluorescent-minus-one (FMO) controls, ex vivo and under stimulated conditions. C, Top left, Spontaneous expression of RANKL in RA synovial fluid B cells compared to RA synovial T cells. Top right, CD95 expression in RA switched memory B cells compared to CD19+ peripheral blood B lymphocytes. Bottom, Expression of RANKL after stimulation of RA synovial B cells compared to RA peripheral blood B lymphocytes and healthy control (HC) B cells. Symbols represent individual samples; horizontal lines with bars show the mean ± SEM. * = P < 0.05; *** = P < 0.001, by analysis of variance with Tukey’s multiple comparison test.

Figure 4

B cells support osteoclastogenesis in a RANKL-dependent manner, and B cells from patients with rheumatoid arthritis (RA) are more proficient in this function. B cells were stimulated to produce RANKL as described in Patients and Methods, and then fixed in cultures. Isolated healthy control (HC) monocytes (CD14+) were added to each culture along with 10 ng/ml recombinant RANKL (recRANKL) and 25 ng/ml macrophage colony-stimulating factor; monocyte cultures without B cells were included as a negative control. In total, 3–4 wells were set up for each individual. Anti-RANKL was added at a concentration of 1 μg/ml. A, Tartrate-resistant acid phosphatase (TRAP)–positive osteoclasts (OCs) (arrows) were identified after coculture with unstimulated (U) or stimulated (S) B cells in the absence or presence of a neutralizing antibody against RANKL. Representative photomicrographs are shown. B, TRAP+ OCs with ≥3 nuclei were enumerated in each culture well in the presence or absence of stimulation and with or without anti-RANKL. C, Representative photomicrographs show OC formation in cocultures of B cells from a healthy control donor and a patient with RA. D, TRAP+ OCs were enumerated in each culture well in cocultures with B cells from 4 RA patients and 4 healthy control donors. Symbols represent individual wells; horizontal lines with bars show the mean ± SEM. * = P < 0.05; ** = P < 0.01; *** = P < 0.001; **** = P < 0.0001, by one-way analysis of variance with Tukey’s multiple comparison test.

Figure 5

CD27+ memory B cells in synovial tissue from patients with rheumatoid arthritis (RA) express RANKL in close proximity to RANK+ cells. RA synovial tissue sections were immunostained with antibodies to CD20 (B cell marker), CD27 (memory B cell marker), CD3 (T cell marker) (not shown), RANKL, and RANK. A and B, CD27+ memory B cells were abundant in the RA synovium (A) and expressed RANKL in close proximity to abundant RANK+ cells (B). Representative results are shown. Boxed areas in left panels are shown at higher magnification in bottom right panels. Top right panels are views of RA synovial tissue from another region. Original magnification × 20. C, Left, A large fraction of the RANKL+ cells in RA synovial tissue were CD20+ B cells, comparable to the percentage of RANKL+ cells that were T cells. Right, A significantly higher fraction of CD20+ B cells expressed RANKL compared to CD3+ T cells (P < 0.0001). D, RANKL+ B cells were visualized near RANK+ osteoclast precursors (arrows in #1 insets), with neighboring multinucleated osteoclasts (arrows in #2 insets). Boxed areas are shown at higher magnification in the insets on the right. Original magnification × 20. E, Left, The percentage of B cells with adjacent RANK+ cells in RA synovial tissue is shown. Right, A large fraction of RANK+ osteoclast precursors had adjacent B cells. B cells expressing RANKL, RANK, CD20, and CD3 were enumerated in 7–8 random 200 × fields. In C and E, symbols represent individual samples; horizontal lines with bars show the mean ± SEM.