Abstract
Major progress in understanding the pathobiology of human bone marrow disorders associated with aging has come from identifying recurrent, acquired mutations in the hematopoietic stem and progenitor cell (HSPC) compartment. However, causal roles for some mutations, and mechanistic pathways in cases lacking mutations, remain unclear. Complex changes in the transcriptional repertoire and the epigenome may contribute independently of driver mutations. A key HSPC alteration observed in aging, and exaggerated in marrow disorders, consists of lineage skewing toward myeloid output, usually at the expense of erythropoiesis - the basis of which remains unknown. From mining of validated RNA-seq datasets, we discovered RUNX3 as a factor down-regulated with aging in human and murine HSPCs, correlated with diminished expression of key erythroid genes Gata1, Klf1, Gypa, and Epor. While widely characterized in solid malignancies as a tumor suppressor, RUNX3 in hematopoiesis has been minimally examined. However, overlapping function with Runx1 in hematopoiesis has been described in zebrafish and murine models. Runx3 deficiency in zebrafish blocked transition to definitive hematopoiesis during development, recapitulating Runx1 findings. Murine HSC knockout studies exhibited an age-dependent granulocytic hyperplasia with a myeloproliferative phenotype, and when combined with Runx1 knockout, rapid-onset marrow failure involving Mac1+ granulocyte progenitor expansion, and severely blunted erythropoiesis.
To explore the role of RUNX3 in human hematopoiesis, CD34+ HSPC underwent expression analysis and lentiviral shRNA knockdown (kd). Notably, in unmanipulated progenitors, immunoblot showed RUNX3 to be expressed in undifferentiated CD34+ cells as well as in CD235a+ erythroid cells. Immunofluorescence revealed an initial cytoplasmic predominance followed a nuclear shift upon erythroid induction. In vivo expression in erythroid progenitors was confirmed by immunostaining of human marrow samples. In uni-lineage cultures monitored by flow cytometry, and in colony formation assays, RUNX3 kd of ~60% blocked erythroid output, while sparing granulopoiesis. When cells were maintained in HSPC expansion medium, RUNX3 kd had no effect on growth or viability but suppressed both features on transfer of cells to erythroid medium.
To stage the defect in RUNX3-deficient HSPC, multi-parametric flow cytometry and mass cytometry (CyTOF) interrogated progenitor composition. In these studies, RUNX3 kd blocked entry into the erythroid lineage and retained cells in a GMP-like state, based on diminished CD36 and CD71 expression, and increased CD45RA and CD123 expression, respectively. RNA sequencing of control and RUNX3-deficient progenitors briefly cultured in expansion or erythroid media revealed diminished expression of erythroid master regulators such as GATA1, KLF1, and several globin genes, as well as an increase in the myeloid master regulator GFI1. These findings recapitulate RNA-seq data from aged murine HSPCs.
Because of its persistent expression during erythroid differentiation, RUNX3 also underwent functional analysis in committed progenitors including the pro-erythroblastic HUDEP-2 line and primary sorted human CD36+ cells. RUNX3-deficient HUDEP-2 cells lost their capacity for inducible hemoglobinization, and RUNX3-deficient CD36+ progenitors displayed a similar inability to execute erythroid maturation, based on a failure to upregulate CD235a. These data suggest an additional later role in erythroid differentiation.
As evidence of its human clinical relevance, RUNX3 expression was found to be diminished in HSPCs purified from elderly individuals with Unexplained Anemia (UA), as compared with age-matched non-anemic control subjects. UA HSCs showed significant impairment in erythroid colony formation, with no changes in granulopoieis. The frequency of MEPs was found to be increased in UA marrow, and UA MEPs subjected to colony formation showed blunted CFU-E outgrowth in response to TGFb, signaling of which is known to be dependent on RUNX3 in other cell types. Our findings thus highlight RUNX3 as a human hematopoietic transcription factor downregulated in aging, and critical in the maintenance of balanced lineage output. We further suggest that its deficiency may contribute to aging-associated HSPC perturbations.
Disclosures
No relevant conflicts of interest to declare.
Dysregulation of RUNX3 in Aged Human HSPCs May Contribute to Perturbations in Erythropoiesis and Balanced Lineage Output
