Depletion of Mature B Cells By Anti CD20-Directed Antibody Induces Trabecular Bone Loss in Young Mice: An Alert for Clinicians?

Targeted B cell therapy by anti-CD20 antibodies has revolutionized the treatment of B cell malignancies, immune cytopenias as well as connective tissue diseases. A growing body of evidence links B cell homeostasis and bone metabolism, including reports demonstrating severe bone loss in transgenic mice lacking IgM⁺ B cells. Here, we addressed the hypothesis that pharmacological depletion of mature B cells induces bone loss in an age-dependent manner.

Ten-week and one-year old C57BL/6J female mice were treated with 150µg and 250µg, respectively, of mouse anti-mouse CD20 (hereafter αCD20), administered in two equal doses, two weeks apart. Four weeks after the first injection, distal femurs and L3 vertebrae (old mice only) were subjected to bone morphometric analysis using µCT. The efficiency of B cell depletion as well as immunophenotyping of the bone marrow (BM) and spleen were assessed by flow cytometry. We used ELISA to measure serum markers of bone formation (N-terminal propeptide of procollagen 1, P1NP)and bone resorption (tartrate-resistant acid phosphatase 5b, TRAcP 5b). The expression of receptor activator of nuclear factor κ-B ligand (RANKL), osteoprotegerin (OPG) and interleukin 7 (IL-7) in both BM containing bone fragment (BM-BF) and spleen was analyzed by means of RT-qPCR. To investigate the osteoclastogenic potential of various BM B cell subsets, total BM cells from wild-type untreated mice were flow-sorted and subjected to an osteoclastogenesis assay in RANKL- and macrophage colony stimulating factor (M-CSF)-containing medium.

In young mice with >90% splenic B cell depletion (5 out of 7), αCD20 treatment (compared to diluent controls) induced a significant trabecular bone loss, reflected by 25% decrease in bone mineral density (30.1±2.5 vs 40±3.7 mgHA/cm³), 44% decrease in connectivity density (10.3±1.5 vs 18.9±2.8 mm^-3), 13% reduction in trabecular number (2.8±0.13 vs 3.24±0.13 mm^-1), and corresponding 14% increase in trabecular separation (p<0.05 for all parameters, n=5-7). However, no effect of αCD20 on bone mass was noted in the old mice (neither in femurs nor in L3 vertebrae, n=6 in each group).

αCD20 treatment induced a compensatory increase in BM pro-B cells in young animals (2.8±0.5% vs 1.8±0.1%, p<0.05) which was even more pronounced in old ones. However, only the former exhibited a 2.5-fold increase in M-CSF receptor (CD115) positivity of this population (p<0.05). In that respect, our osteoclastogenic assays showed that among various BM B cell populations, only CD115⁺ pro-B cells could give rise to functional osteoclasts (18±6.5% OC area vs none). Moreover, serum P1NP levels were significantly lower in the young but not in the old αCD20-treated animals as compared to their corresponding diluent controls (48.2±8.7 ng/mL vs 74.5±4.6 and 27.9±8.2 ng/mL vs 30.3±10.8, respectively), indicating decreased bone formation in the former. αCD20 treatment did not affect the levels of TRAcP 5b, a surrogate marker of the total osteoclast number. Following B cell depletion, the expression of IL-7, known to induce bone loss, was upregulated in both BM-BF and spleen by the magnitude of 37% and 50%, respectively (p<0.05). Yet, the expression of RANKL and OPG in BM-BF and in spleen, were similar in control and in the αCD20 treatment groups.

The current study demonstrates for the first time the adverse skeletal outcomes of pharmacological depletion of mature B cells, conceivably fueled by B cell-derived osteoclastogenesis in the presence of decreased bone formation. Since many patients treated with clinically available anti-CD20 antibodies already have other risk factors for bone loss, e.g. female gender and glucocorticoid therapy, these data also have immediate clinical implications.