BATF2 Promotes HSC Myeloid Differentiation Via Amplification of the Pro-Inflammatory Response during Chronic Infection

Chronic infections affect more than 2 billion people worldwide and can impair the function and differentiation of hematopoietic stem cells (HSCs), which produce all circulating blood cells. However, the molecular mechanisms by which infection suppresses bone marrow function are not fully elucidated. In prior work, we used a mouse model of Mycobacterium avium infection to show that chronic infection leads to pancytopenia, and impaired self-renewal and increased differentiation of HSCs via an interferon gamma (IFNg)-dependent process. We found that HSCs are significantly depleted in this model, and we identified basic leucine zipper ATF-like transcription factor 2 (Batf2) as a key factor responsible for HSC depletion during chronic infection. Here, we hypothesize that BATF2 drives the depletion of HSCs during chronic infection by promoting myeloid differentiation.

To test this hypothesis, we created a Batf2 knock-out animal model and performed flow cytometry analysis on chronically infected wild-type (WT) and Batf2 KO mice. We found that, compared to the WT controls, infected Batf2 KO mice induced fewer myeloid progenitor cells, including MPP3 and GMP subsets in the bone marrow and inflammatory macrophages in the blood. Upon secondary transplant, HSCs from infected Batf2 KO mice were better conserved and maintained better transplant capacity than WT control in chronic infection, indicating loss of Batf2 improved HSC self-renewal.

To assess the role of Batf2 in the cellular response to the infection, we performed a cytokine bead array in infected WT and Batf2 KO mice. Compared to WT, we found that infected Batf2 KO mice displayed lower levels of pro-inflammatory cytokines in the blood, including IFNg, TNFa, and CCL5. Moreover, Batf2 KO mice recruited fewer immune cells to and had fewer granulomas in the spleen in chronic infection compared to WT mice.

To identify the binding partners and gene targets of BATF2 in HSCs, we performed immunoprecipitation and mass spectrometry in IFNg-treated 32D cells, a bone marrow cell line. We found that BATF2 formed a complex with JUN in response to IFNg treatment in 32D cells. Moreover, RNA sequencing in HSCs isolated from infected WT and Batf2 KO mice revealed that genes involved in IFNg response were significantly enriched in WT but not Batf2 KO mice. Specifically, we found amplification of pro-inflammatory signaling pathways including CCL5 in infected WT mice compared to Batf2 KO. CUT&RUN sequencing demonstrated that BATF2 interacts with regulatory elements in IFN response genes. These results indicate that BATF2 interacts with JUN to amplify expression of inflammatory signaling and myeloid differentiation pathways during infection.

To test the validity of this finding, we treated WT and Batf2 KO mice with recombinant CCL5. CCL5 was sufficient to rescue IFNg-induced myeloid differentiation and recruit more immune cells to the spleen in Batf2 KO mice. Conversely, blocking CCL5 receptors using the CCR5 inhibitor maraviroc reduced myeloid differentiation in WT mice, phenocopying Batf2 KO mice. These results suggest that BATF2 induces CCL5 to promote myeloid differentiation and recruit immune cells to sites of infection.

In summary, we demonstrate that BATF2 binds with JUN in HSCs and induces CCL5 to promote myeloid differentiation and stimulate pro-inflammatory responses during chronic infection, leading to HSC differentiation and depletion. We envision that our work will enable the development of therapeutic approaches targeting BATF2 signaling pathways to prevent bone marrow suppression due to chronic infection.

No relevant conflicts of interest to declare.