Abstract
Introduction: A hallmark of multiple myeloma (MM) is the low levels of uninvolved immunoglobulin (Ig) levels. B-cell maturation antigen (BCMA) is a receptor expressed in mature non-malignant and malignant B lymphocytes, including plasma cells. Its ligands are BAFF (B-cell activating factor) and APRIL (a proliferation inducing ligand). We previously demonstrated that BCMA is present in the serum of MM patients (pts) and that its levels predict survival (Sanchez et al. Br J Haematol 2012). We hypothesized that circulating BCMA binds it ligands, preventing normal plasma cell development in MM patients which may explain their reduction in uninvolved Ig levels.
Methods: BCMA-Fc and control Ig were obtained and reconstituted in PBS (R&D Systems). Retro-orbital bleeds were performed on SCID mice implanted with the human xenografts LAGλ-1, LAGk-1A or LAGk-2 following BCMA treatment. Human BCMA and mouse BAFF, IgM and IgA levels were measured with ELISA (R&D Systems & Bethyl Laboratories). Human IgA and IgG levels were determined in MM patients using nephelometry (Immage 800, Beckman Coulter). Hevylite® Assays (Binding Site) were used to quantify the levels of heavy-light chain isoform pairs in MM patients.
Results: Recombinant mouse BAFF (rmBAFF) was mixed with recombinant human BCMA (rhBCMA) and incubated on plates coated with anti-mouse BAFF antibody (Ab). An anti-BCMA detection Ab was added, and BCMA-BAFF complexes were easily detected at concentrations of human BCMA and BAFF present in MM pts. We determined if mBAFF from SCID mice would also bind rhBCMA. SCID mice were bled; their plasma isolated and incubated with rhBCMA, and showed that rhBCMA formed complexes with mBAFF.
To determine what effect human BCMA had on Ig levels in immune competent mice, rhBCMA-Fc or control Ig-Fc (100 mg) was injected into C57 Bl and Balb/c mice, and plasma IgA and IgM levels were measured. Following rhBCMA-Fc injection, a marked decrease in both antibody classes was observed in both mouse strains. Decreases in IgA levels were observed following BCMA treatment when compared to baseline plasma IgA on days 4 and 6 (P = 0.0031 and P = 0.0064, respectively), and the control groups (P = 0.0087 and P = 0.0221). Samples were also analyzed for mouse IgM levels with similar marked reductions when compared to the untreated (P = 0.0001) and Ig-Fc (P = 0.0088) groups. To determine if rhBCMA-mBAFF complexes formed in vivo, an ELISA was performed. Plates were pre-coated with a monoclonal mouse anti-BAFF capture Ab. Plasma samples were incubated and an anti-human-BCMA detection Ab was added. rhBCMA-mBAFF complexes were detected at high levels in plasma samples from mice dosed 4 and 6 days previously with BCMA-Fc whereas no complexes were found in samples in control Ig-Fc or untreated mice. We evaluated mBAFF levels in SCID mice with and without human MM. We have previously shown high levels of human BCMA in the plasma of MM tumor-bearing mice (Sanchez et al. Br J Haematol 2012). On days 14, 21 and 28 post-tumor implantation, mice implanted with LAGλ-1 had markedly lower plasma levels of mBAFF (P = 0.0143, P = 0.0002 and P = 0.0008, respectively) when compared to age and sex-matched mice not bearing any tumors. We obtained similar results using two additional human MM xenograft models (LAGk-1A and LAGk-2). Next, we determined whether serum BCMA levels inversely correlated with uninvolved Ig levels in MM pts. For pts with IgA (n = 134) or IgG (n = 313) MM, higher BCMA levels (> 100 ng/ml) correlated with below normal levels of uninvolved IgG in IgA MM and uninvolved IgA in IgG MM, whereas lower BCMA levels (< 100 ng/ml) correlated with normal uninvolved levels (P < 0.0001). Using the Hevylite Assay, similar results were observed for the levels of BCMA compared to uninvolved IgG isoforms in both pts with involved IgG lambda (n = 62, P = 0.0006) and IgG kappa (n = 117, P < 0.0001) MM.
Conclusions: Our laboratory previously reported that BCMA is present in the serum of MM patients, correlates with response to treatment and predicts survival. We now demonstrate the formation of circulating BCMA-BAFF complexes in MM, and administration of recombinant BCMA to normal mice results in marked reductions in their antibody levels. We also show that BCMA levels inversely correlate with uninvolved Ig levels in MM pts. Thus, the lack of normal antibody production in MM pts results in part from circulating BCMA binding its ligands, preventing production of normal antibody-producing cells.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.