EPO Modulation of Thioredoxin Interacting Protein (TXNIP), and Key Roles for TXNIP during EPO-Dependent Human Erythroid Progenitor Cell Growth and Development

Via contemporary investigatory approaches, important new EPO action mechanisms continue to be discovered.  As recent illustrating examples, lineage tracking studies have demonstrated guiding effects of EPO on the developmental fate of myeloerythroid progenitors [J Exp Med. 211:181-8].  Profiling of EPO response genes in primary erythroid progenitor cells (EPCs) has uncovered a pathway of EPO cytoprotection against leached lysosomal cathepsins in stressed erythroblasts [J Exp Med. 210:225-32].  And in a context of iron metabolism, Erythroferrone has been identified as an EPO-induced inhibitor of hepcidin [Nat Genet. 46:678-84].

In a unique approach to defining EPO’s cell and molecular actions, our laboratory has launched post-translational modification based LC-MS/MS proteomic analyses in human erythroid precursor cell systems. In interrogations of novel EPO/EPOR targets modified within a category of “p-TPP” motifs, TXNIP proved to be rapidly phosphorylated >10 fold due to EPO at novel C-terminal p-T349 and p-S358 sites.  In parallel, multi-fold increases in TXNIP levels also were observed within 20 minutes of EPO exposure. To gain functional insight into TXNIP’s roles during EPO-dependent erythropoiesis, LOF studies were performed via lentiviral shRNA mediated inhibition of TXNIP.  As analyzed first in UT7epo cells, TXNIP knockdown (>90% efficiency) attenuated cell proliferation ~2.5-fold with EPO- dose dependency. This was reflected further in selective attenuation of cell cycle progression at S-phase (with only limited effects of TXNIP knockdown on cell survival observed). LOF experiments were also performed in human bone marrow derived CD34+ primary hematopoietic progenitor cells. Following initial plating, hematopoietic progenitors transduced with TXNIP shRNA exhibited a limited transient lag in growth (as compared directly with shRNA-NT transduced controls), but by day 4 of culture, and thereafter, expanded at essentially normal levels.  Subsequently, erythroid progenitors with inhibited TXNIP expression prematurely committed to a program of late erythroblast differentiation. Specifically, TXNIP knockdown resulted in elevated frequencies of GPA-high erythroblasts (33.4+/-1.3% vs controls at 6.9+/-1.1%, p = 0.0001) and decreased KIT-high expression (6.4+/-3.0% vs controls at 62.5+/-7.1%, p = 0.0002). In addition, visibly obvious increases in erythroblast hemoglobinization due to TXNIP knockdown also were observed. 

The present investigations thus employ a unique PTM LC-MS/MS approach to identify TXNIP as a new EPO/EPOR target modified at novel C-terminal sites upon EPO expression, and in association with an indicated stabilization of TXNIP.  LOF studies further reveal pro-erythropoietic roles for TXNIP as a novel mediator of EPO-dependent erythropoiesis with important effects exerted on erythroid precursor cell proliferation, together with a key requirement for TXNIP in governing a transition to a program of late erythroblast differentiation.  Significance is further underlined by the nature of TXNIP as an emerging target within beta cells for small molecule inhibition in type-2 diabetics, a group highly representative among anemic chronic renal disease patients [J Diabetes Res Apr 7 2015:801348].  In particular, based on our present findings, such inhibition of TXNIP would be predicted to further compromise erythropoiesis in this CKD patient population.