A Shared Signaling Network That Maintains Erythroid Homeostasis By Activating Stress Erythropoiesis Regulates Inflammation: Implications for the Anemia of Chronic Inflammation

Anemia is a common secondary pathology resulting from inflammatory diseases including cancer or infection. Its exact prevalence is difficult to determine, yet its contributions to the morbidity and mortality of patients and its negative impact on quality of life are clear. Despite the diverse set of factors that can lead to inflammatory anemia, its core pathology of hyperinflammation, iron dysregulation, and lack of red cell production suggests the possibility of a common etiology. Inflammation induces pro-inflammatory cytokines including TNFα, IL-1β and IFNγ that drive myelopoiesis at the expense of steady state bone marrow erythropoiesis. In addition, other cytokines increase the expression of hepcidin, haptoglobin and hemopexin by the liver, leading to the sequestration of iron. While limiting iron can be beneficial in the context of infection, the consequence of this restriction is a further reduction in red cell production in the bone marrow. To compensate for the loss of bone marrow erythropoiesis, inflammation induces stress erythropoiesis in the spleen or liver.

Stress erythropoiesis is regulated by different signals which include BMP4 and GDF15 and utilizes stress erythroid progenitors that are distinct from steady state erythroid progenitors. Our work shows that in contrast to steady state erythropoiesis, pro-inflammatory cytokines like TNFα promote the proliferation of stress erythroid progenitors, while anti-inflammatory signals such as PGJ2 and IL-10 promote their differentiation. These studies demonstrate that the expansion and differentiation stages of stress erythropoiesis are coordinated with, and influenced by, signals that initiate and resolve inflammation. In addition, we show that this regulation is reciprocal. Signals that regulate the differentiation of stress erythroid progenitors (GDF15 and BMP4) promote the resolution of inflammation. Mice infected with the model gut pathogen Citrobacter rodentium, exhibit stress erythropoiesis in the spleen, while steady state erythropoiesis in the bone marrow is suppressed until pathogen clearance. We observed that hepcidin expression in the liver increases initially, but then decreases as the expression of erythroferrone by stress erythroid progenitors increased in the spleen, but not the bone marrow. Using mice mutant for GDF15 (GDF15-/-) and for BMP4 signaling (flexed-tail f/f), which exhibit defective stress erythropoiesis, we observed that the expression of hepcidin was dysregulated suggesting that stress erythroid progenitors are responsible for iron regulation at this time. In addition, infection of mutant mice led to increased lethality. During peak infection, we observed morphological differences in the colons of these mice indicative of increased inflammation and systemic infection. These changes were associated with increased expression of pro-inflammatory genes, as well as decreased numbers of FoxP3+ regulatory T-cells (Tregs). Using naïve CD4+ T-cells isolated from uninfected control, f/f or GDF15-/- mice, we observed significantly altered gene expression from mutant T-cells following Treg induction in vitro. However, the addition of BMP4 and GDF15 into these cultures rescued Treg development of mutant naïve T-cells and enhanced Treg development of naïve control T cells. Analysis of the BMP4 and GDF15 signaling pathways in both stress erythroid progenitor differentiation and in Treg development revealed that in both systems these signals converge on the transcription factor HIF1α.

Taken together these data support a new model showing that the loss of steady state erythropoiesis due to pro-inflammatory signals is balanced by the activation of stress erythropoiesis by those same factors. Similarly, the differentiation of stress erythroid progenitors appears to regulate iron, and is itself regulated by the same signals that drive the development of Tregs and the expression of anti-inflammatory cytokines during immune resolution. This work supports a novel model where initiation and resolution of inflammatory immune responses are co-regulated with stress erythropoiesis, which allows for a robust immune response while maintaining erythroid homeostasis. Furthermore, this model predicts that alterations to this shared signaling network will underlie the development of chronic inflammatory anemia.