Background/Purpose:  Humanized nonobese diabetic/Shi-scid/γc^null (NOG) mice, which are reconstituted with human hematoimmune system by transplantation with human CD34^＋ hematopoietic stem cells (termed as humanization), have been shown to reproduce key aspects of human Epstain-Barr virus (EBV) infection. We previously demonstrated that EBV-infected humanized NOG (hu-NOG) mice developed erosive arthritis similar to that in Rheumatoid arthritis (RA). Furthermore we showed that human osteoclasts were present in the bone erosion sites and that human osteoclasts actually differentiated in vitro from osteoclast progenitors settled in the bone marrow of EBV-infected hu-NOG mice. However, the mechanisms for the development of human osteoclasts in the EBV-infected hu-NOG mice including cells supplying human receptor activator nuclear factor-κB ligand (RANKL), an essential molecule for osteoclast differentiation, have not been elucidated yet. In this study, we investigated the roles of EBV in RANKL expression responsible for osteoclastogenesis in EBV-infected hu-NOG mice.

Methods: Humanization of NOG mice and inoculation of hu-NOG mice with 5 to 20 × 10² TD₅₀ (50% transforming dose) of EBV (B95-8) were performed as described in our previous study. After 8 to 10 weeks of the EBV inoculation, when erosive arthritis can be observed at a high rate, peripheral blood cells (PBCs) collected from the mice were stained with monoclonal antibodies (Abs) specific to human RANKL, CD19 and CD20 by multicolor immunofluorescence technique and analyzed by flow cytometry. To examine the effects of EBV on RANKL expression, not only EBV-free cord blood mononuclear cells (CBMCs) obtained after informed consent but also Ramos cells, an EBV-negative B-cell lymphoma cell line, were incubated with culture medium containing EBV (5 to 20 × 10² TD₅₀). Then, both cell types were stained with those Abs as mentioned above and analyzed by flow cytometry. Quantification of EBV DNA was performed by a real-time PCR assay.

Results:  RANKL ⁺ cells were found in PBCs of the EBV-infected hu-NOG mice, most of which cells showed positivity for the B-cell markers, CD19 and CD20. In contrast, almost no RANKL ⁺ cells were detected in that of un-infected hu-NOG mice. Inoculation of CBMCs with the EBV in vitro induced RANKL on CD19⁺CD20⁺ B cells. The positivity for RANKL on B cells increased dramatically, reaching nearly 90% after 3 to 4 weeks of EBV inoculation. Real-time PCR assay showed a high level of EBV DNA (5 × 10⁴ copies/μg of DNA) in these cells, indicating that the cells were actually infected with EBV. The experiments with Ramos cells showed that inoculation with the same dose of EBV as that with CBMCs markedly induced RANKL on the Ramos cells around 3 weeks after the EBV inoculation, whereas the cells without EBV inoculation remained negative for RANKL. EBV DNA was detected (1 × 10⁴ copies/μg of DNA) in the EBV-inoculated Ramos cells.

Conclusion:  Our present study demonstrates that EBV infection induces RANKL expression on B cells and that B cells, the primary target of EBV, are the predominant RANKL⁺ population in the PBCs of EBV-infected hu-NOG mice, suggesting that EBV-induced RANKL⁺ B cells play roles in osteoclastogenesis causing bone erosion in these mice.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/epstein-barr-virus-induced-expression-of-receptor-activator-nuclear-factor-%ce%bab-ligand-on-b-cells-is-possibly-responsible-for-erosive-arthritis-in-epstein-barr-virus-infected-humanized-nonobese-dia/