L3MBTL1: A Polycomb Protein At the Node of Crosstalk Between the BMP4 and Hippo Signaling Pathways in Erythropoiesis.

Abstract 2296

The Hippo signaling pathway, first discovered in Drosophila, is emerging as an important regulator of stem cell behavior. Upon still-unclear upstream stimuli, the hippo pathway kinase cascade phosphorylates and inhibits the function of YAP, a transcription coactivator, by inducing its cytoplasmic retention. While recent evidences indicate that inhibition of YAP affects cell fate decisions, and proliferation, in many tissues, little is known about the relevance of this pathway in hematopoiesis. However, the interaction of YAP with Smad1, identified in flies and human cells (Alarcon C. et al. Cell 2009), prevents smurf-mediated Smad1 degradation, potentially enhancing BMP signaling.

Our ongoing studies have indentified crosstalk between the BMP4 and the Hippo pathways in hematopoietic cells, and in induced-pluripotent stem (iPS) cells that we differentiated towards the erythroid lineage. This crosstalk involves the chromatin-binding, Polycomb protein L3MBTL1, which clearly regulate the effects of BMP on the erythroid differentiation of hematopoietic stem/progenitor cells and on fetal globin gene expression.

We find that the Lats2 kinase, a core component of the Hippo pathway, physically interacts with L3MBTL1 and that treatment with BMP4 or Erythropoietin decreases the expression of both proteins in various hematopoietic cells, including primary human cord blood-derived CD34+ cells. By altering L3MBTL1 levels in K562 cells, we were able to show that the L3MBTL1-Lats2 interaction enhances Lats-mediated phosphorylation and the cytoplasmic retention of YAP. Furthermore, L3MBTL1-depleted iPS cells have an enhanced smad-mediated transcriptional response; by analyzing the gene expression profile of these cells, we found increased expression of several BMP target genes (such as HHEX and ID genes), suggesting that L3MBTL1 negatively titrates the BMP4 signaling pathway at least in part by affecting YAP phosphorylation and localization. Gene Set Enrichment Analysis confirmed enrichment of many smad-related genes, and yet, these cells presented enhanced smad1/5/8 phosphorylation by WB analysis, indicating that BMP4 signaling is triggered by L3MBTL1 depletion. We also found that hematopoietic differentiation of L3MBTL1-KD iPS cells generates high-fetal globin gene expressing erythroid progeny, suggesting a role for the BMP4 signaling pathway and the targeting of L3MBTL1 in the treatment of hemoglobinopathies.

To further evaluate the effect of BMP4 signaling on hematopoietic cells that lack L3MBTL1, we analyzed the stress erythroid response of L3MBTL1 KO mice: while no difference was observed at baseline in the null mice compared to wt littermates, the L3mbtl1 null mice had a more severe anemia, with increased leukocytosis, and thrombocytosis post-hydrazine (PHZ) or Epo. We found a significant increase in the colony-forming ability of the l3mbtl1 null spleen and bone marrow cells, compared to controls, as well as increased spleen size and an expansion of the spleen erythroid compartment. Thus, l3mbtl1 null hematopoietic stem cells are more sensitive to the PHZ-mediated cytokine storm, which includes BMP4. Interestingly, the L3mbtl1 null BM and spleen cells showed diminished expression of Lats2 and phospho-YAP, consistent with our in vitro findings.

In conclusion, these investigations have shown that L3MBTL1 not only negatively titrates the BMP4 signaling pathway, but also provides a nodal point for crosstalk between the BMP4 and Hippo signaling pathways in erythropoiesis. Thus, these data provide insights into possible novel treatments for genetic red cell disorders (such as β-thalassemia) and for acquired bone marrow failure syndromes such as Epo-resistant anemia.