Abstract
Abstract 159
The anemias of chronic disease (ACD) are a common complication of malignancy, inflammation and kidney disorders. In ACD, there is dysregulation of iron homeostasis, decreased proliferation of erythroid progenitors, diminished production of erythropoietin (EPO), and shortened lifespan of RBC. Multiple pathophysiologic mechanisms have been implicated in the development of ACD, including elevated production of hepcidin and inflammatory cytokines, IFNγ, TRAIL, Interleukins-1β, 6, 10, 15, & TNFα. These cytokines are thought to directly inhibit erythroid differentiation through unknown mechanisms. The current study addressed the hypothesis that inhibition of erythropoiesis in ACD may arise through synergistic effects of iron deprivation and specific inflammatory cytokines. To identify relevant cytokines, candidate factors were applied to primary human erythroid progenitors in iron replete and restricted cultures. Peripheral blood human CD34+ progenitors from healthy donors underwent standard prestimulation for 72 hours, followed by culture in unilineage erythroid medium (4.5 U/ml EPO + 25ng/ml SCF) for 4-5 days under iron replete (100% transferrin saturation) or iron restricted (15% transferrin saturation) conditions. Candidate cytokines were screened for effects on viability, proliferation, and differentiation using cell counting and flow cytometric analysis of the erythroid cell surface marker GPA and the megakaryocytic antigen CD41a. Contrary to previous reports, the majority of cytokines (TRAIL & Interleukins-1β, 6, 10, 15) showed no effects on erythroid proliferation or differentiation under iron replete or restricted conditions. By contrast, both IFNγ and TNFα displayed potent inhibitory effects under iron restricted conditions but only weak effects in iron replete cultures. Typically, iron restriction alone reduced the proportion of GPA+ cells by 50%, whereas IFNγ or TNFα combined with iron restriction caused a 90% reduction. While both cytokines cooperated with iron restriction in blocking upregulation of GPA and promoting cell death, each cytokine also had distinctive effects on morphology and differentiation. IFNγ enhanced megakaryocytic development, while TNFα retained cells as immature, CD34+ progenitors. The synergistic inhibition of erythroid differentiation with iron restriction and TNFα was confirmed in vivo using a murine model of dietary iron deprivation coupled with continuous infusion of low-dose TNFα. Regarding the mechanism for this synergy, we have previously shown that erythroid iron deprivation leads to inactivation of the aconitase enzymes, which normally convert citrate to isocitrate, and that provision of exogenous isocitrate abrogates the erythroid inhibition associated with iron deprivation. Accordingly, participation of this pathway was assessed in the more potent erythroid inhibition associated with IFNγ or TNFα plus iron deprivation. Strikingly, isocitrate administration not only abrogated effects due to iron deprivation but also those due to the inflammatory cytokines, leading to complete rescue of erythroid differentiation. To address the underlying basis for erythroid cross-talk of iron and cytokine signaling, we screened pathways implicated in iron metabolism and inflammation. Two relevant pathways identified were Jun kinase (JNK) and calmodulin-associated kinase II (CAMKII), important in TNFα and IFNγ signaling, respectively. In particular, TNFα and iron deprivation synergized in the activation of JNK, and IFNγ and iron deprivation synergized in activating CAMKII. In both cases, isocitrate partially restored the activation to basal levels. As an important negative control, iron deprivation did not affect IFNγ activation of STAT1 phosphorylation, indicating that its effects were not due to upregulation of receptor expression or function. Altogether, these data suggest that among the various cytokines implicated in ACD, only IFNγ and TNFα synergize with iron deprivation in the inhibition of erythropoiesis. These actions occur through cross-talk between intracellular signaling pathways, specifically pathways involving aconitase and cytokine-activated kinases. The connection of aconitase/metabolism with inflammation is novel and has implications for clinical treatment of ACD, as well as for new understanding of erythroid and inflammatory signaling.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.