The homeobox transcription factor MEIS1 is expressed in hematopoietic stem- and progenitor cells (HSCs & HPCs, respectively) such as human CD34+ cells, whereas its expression in lineage-committed blood cells is restricted to megakaryocytes (MKs). We observed that MEIS1 not only drives megakaryopoiesis but is also indispensable for the differentiation of HPCs towards the erythroid lineage. In cord blood CD34+ cells lentiviral-driven MEIS1 overexpression resulted in a 3-fold increase in BFU-E at the expense of CFU-GM colonies and a 2-fold increase in MK colonies was recorded. Vice versa, silencing MEIS1 led to a reduction in the number of MK-colonies and a near absence of BFU-E, a phenotype strongly reminiscent of the ones observed in knock-out mice and zebrafish. To pinpoint at which stage of hematopoietic commitment MEIS1 expression regulates lineage fate, we sorted CD34+ cells further into the HSC and HSPC subsets. MEIS1 overexpression in HSC and common myeloid progenitors (CMP) induced a 3-fold increase in BFU-E at the expense of CFU-GM consistent with the data obtained in CD34+ cells. Remarkably, MEIS1 overexpression also resulted in erythroid colony formation in the granulocyte-monocyte precursor cells (GMP), a subset that naturally is committed to myeloid differentiation. The results show that MEIS1 drives HPCs and HSCs towards megakaryocyte-erythroid precursor cell (MEP) fating.
To unravel the underlying mechanism of this fating we performed chromatin immunoprecipitation with MEIS1 antibodies in the megakaryoblastic cell line CHRF 288–11 and primary MKs combined with massive parallel sequencing (ChIP-seq). 13,842 MEIS1 binding events were detected in CHRF and 18012 events in MK, respectively. The transcription factors GATA1 and FOG1 (ZFPM1) are critical for the commitment of HSCs and HPCs towards the erythroid lineage. Lately, it has also been noted that FOG1 limits stem cells towards MEP fating and that GATA1 transcription is induced by FOG1 (Mancini et al., EMBO, 2012). No MEIS1 binding sites were observed in the GATA1 promoter, but potential binding events were observed in the FOG2 promoter at position 625 and 567. These ChIP-seq results were replicated by RT-qPCR, which confirmed MEIS1 binding to the promoter of FOG1 but not of GATA1 or PU.1. Interestingly overexpression of MEIS1 resulted in a 2-fold increase in FOG1 and GATA1 transcripts but the level of PU.1, a transcription factor essential for the differentiation towards the granulocytic/monocytic lineage, remained unaltered. We are currently performing ChIP-Seq on CD34+ cells to define the differences in MEIS1 occupancy between MKs and HSCs.
In conclusion, we show that the transcription factor MEIS1 induces a MEP fate by binding to the FOG1 promoter thus positively regulating FOG1 transcription. As MEIS1 does not bind to the GATA1 promoter, we further hypothesize that in human hematopoiesis increased GATA1 expression is mediated by FOG1 as described earlier in the murine model.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.