Osteoblast-derived ATP maintains bone marrow plasma cells

Plasma cells reside in the bone marrow but the niche factors that maintain them are poorly defined. Reporting in Nature, Ishikawa et al. show that ATP is released in a regulated manner by osteoblasts and signals via P2RX4 on bone marrow plasma cells to protect them from apoptosis.

The authors began their study by assessing haematopoietic and immune cell populations in mice deficient for PANX3, which is a gap junction protein that controls the regulated release of ATP from osteoblasts in the bone marrow. Panx3^−/− mice have reduced bone marrow cellularity owing to an overall reduction in bone length, but otherwise have relatively normal bone marrow architecture. Thymic T cell, splenic T cell, splenic B cell and peritoneal B1 B cell compartments all appeared to be normal in Panx3^−/− mice. However, serum titres of IgM, IgG and IgA were substantially reduced in Panx3^−/− mice, as were antibody-secreting plasma cell populations in the bone marrow. Subset analyses showed that both newly generated and mature long-lived plasma cell populations were reduced in the bone marrow from Panx3^−/− mice. By contrast, antibody-secreting plasma cells in the spleen were not reduced. Therefore, the generation or maintenance of bone marrow plasma cells seems to be impaired in the absence of PANX3.

Credit: S. Bradbrook/Springer Nature Limited

PANX3 is mainly expressed by osteoblasts, so the authors next examined numbers of plasma cell numbers in co-cultures of bone marrow cells and differentiated osteoblasts from wild-type or Panx3^−/− mice. They found a marked decrease in IgG-secreting cells when wild-type bone marrow cells were cultured on Panx3^−/− osteoblast feeder cells. This was not associated with any differences in levels of mRNAs encoding CXCL12, SCF, IL-6 or APRIL, which are factors known to support plasma cell survival. Instead, the authors found reduced concentrations of extracellular ATP (eATP), which suggests that PANX3-mediated release of ATP by osteoblasts might support plasma cells. In accordance, the addition of ATP, but not ADP or AMP, supported the recovery of plasma cell numbers.

To identify the relevant mechanism of ATP signalling, the authors assessed P2RX receptor activity on bone marrow plasma cells. Treatment of mice with global P2RX inhibitors reduced the frequencies of bone marrow plasma cells. As P2RX4 is expressed by both mouse and human bone marrow plasma cells, the authors next generated mice with B-lineage-restricted deletions of P2RX4. Specifically, they generated P2rx4^fl/fl;Mb1-cre mice, in which P2rx4 is mutated in early B cell precursors, and P2rx4^fl/fl;CD20-TAM-Cre mice, in which P2rx4 can be conditionally deleted from mature B cells using tamoxifen. In P2rx4^fl/fl;Mb1-cre mice, steady-state levels of serum immunoglobulin were reduced as well as bone marrow plasma cells, whereas splenic plasma cell numbers were not changed. Similarly, P2rx4^fl/fl;CD20-TAM-Cre mice fed tamoxifen showed a reduction in steady-state serum immunoglobulin titres and bone marrow plasma cells, but had an intact splenic plasma cell compartment.

To assess antigen-specific plasma cell responses, the authors fed P2rx4^fl/fl;CD20-TAM-Cre mice tamoxifen and then immunized them with NP-KLH (a T cell-dependent hapten–carrier). They observed a substantial reduction in NP-specific plasma cells in the bone marrow but not in the spleen, and no defect in NP-specific germinal centre B cells.

Treatment of bone marrow cultures with the P2RX4-specific inhibitor 5-BDBD also suppressed antibody-producing plasma cells, and eATP could not rescue plasma cells numbers in the presence of 5-BDBD. In addition, wild-type mice treated with 5-BDBD showed reduced numbers of bone marrow plasma cells but other B cell and plasma cell populations were intact.

Pathogenic autoantibodies are a feature of the autoimmune disease systemic lupus erythematosus (SLE), and the authors therefore examined whether targeting P2RX4 could have therapeutic potential. In the NZB/W mouse model of SLE, treatment with 5-BDBD after the onset of disease reduced titres of IgG antibodies against double-stranded DNA and overall proteinuria. Similar findings were seen when 5-BDBD was therapeutically administered in a separate mouse model of autoantibody-associated proteinuria.

Finally, the authors explored whether eATP–P2RX4 signalling supports plasma cells by promoting ER homeostasis in the cells. In support of this, a series of experiments indicated that inhibition of P2RX4 in plasma cells causes an acute ER stress response that drives apoptosis via the ATF4–CHOP–BIM pathway.

“bone marrow plasma cells use P2RX4 to sense the regulated release of ATP from osteoblasts”

The authors conclude that bone marrow plasma cells use P2RX4 to sense the regulated release of ATP from osteoblasts and that this supports cell survival by preventing ER stress-driven apoptosis. Importantly, these findings suggest that the eATP–P2RX4 pathway could be targeted to deplete pathogenic plasma cells in SLE and could be explored in other disease settings, including plasma cell cancers. These data also highlight that sensing of eATP can have homeostatic roles, as well as pro-inflammatory ones, in host immune regulation.