P38α/JNK Signaling Restrains Erythropoiesis By Suppressing Ezh2-Mediated Epigenetic Silencing of Bim

A remarkable feature of erythropoiesis is the coordination of proliferation, differentiation and apoptosis of erythroid cells to precisely achieve erythropoietic homeostasis to avoid anemia and polycythemia. Anemia is a common disease arising from various causes, including Myelodysplastic syndromes, thalassemia, cancer chemotherapy, chronic kidney disease and hemorrhage. The pro-erythropoietic factor erythropoietin (EPO) is often employed for anemia therapy. However, questions have been raised about the safety of EPO given its potential for tumor promotion in cancer-related anemia. Moreover, many acute and chronic anemias, including hemolysis, sepsis and genetic bone marrow failure diseases such as Diamond-Blackfan anemia are untreatable with EPO. To overcome these hurdles, new molecular mechanisms need to be identified that physiologically restrain erythropoiesis by acting as molecular brakes to prevent over-active erythropoiesis caused by pro-erythropoietic signals. Inhibiting these restraining mechanisms could provide alternative approaches to treat anemia in an EPO-independent fashion. P38 MAPK (Mitogen-activated protein kinase) is an important pathway involved in diverse biological processes. P38 modulates cell proliferation, controls cell survival and decides cell fate during differentiation. P38 pathway functions mainly by phosphorylating and activating important transcription factors in response to different stimuli, including ATF2, CREB, and MEF2. There are four members within the P38 MAPK family, including P38α, P38β, P38γ, and P38δ. These members are encoded by different genes and have different tissue expression patterns. Among them, P38α is ubiquitously expressed. P38α modulates the function of different cell types. There are two distinct developmental defects reported in global P38α knockout mice by two separate groups using different mouse strains. One displayed embryonic death with highly anemic appearance due to reduced EPO production and another showed even earlier embryonic lethality due to placental developmental defects. In a P38α conditional mice model in which Cre recombinase was expressed in the whole mouse embryo but not in the placenta by crossing to MORE-Cre mice, no anemia or EPO defects were observed. However, the intrinsic and cell autonomous role of P38α in adult steady-state or stressful erythropoiesis has not been established. Loss of P38α causes activation of JNK in the liver. P38 inhibitors are in clinical trials and have the potential for the treatment of human disease. Therefore, it is important to understand the down-stream targets and functional outcomes induced by P38α deficiency. Using primary human erythroblasts derived from human CD34⁺ hematopoietic stem and progenitor cells (HSPCs) and P38α conditional knockout mice, we find that P38α acts as a molecular brake during anemia recovery through integrating apoptotic signals and by shortening the lifespan of erythroblasts to prevent potential over-active erythropoiesis caused by pro-erythropoietic signaling. Loss of P38α in erythroblasts activates JNK through augmented Map3k4 via a negative feedback circuit revealed by gene expression profiling. Functionally, JNK serves as a pro-survival signal independent of EPO by compromising Bim expression via stabilizing the epigenetic silencer Ezh2 in erythroblasts. JNK-controlled Cdk1 activity modulates full interaction of Ezh2 to the E3 ligase Smurf2 through multiple threonine phosphorylation sites within Ezh2. Our findings identify a key signaling cascade involving P38α/JNK/Cdk1/smurf2/Ezh2/Bim in fine tuning stress erythropoiesis. We propose that inhibition of P38α may provide an alternative strategy for the treatment of anemia.