PU.1 is a hematopoietic transcriptional regulator that is necessary for the development of both myeloid and B cells. To identify new PU.1 target genes in neutrophil development PU.1 was introduced into mouse 503 PU.1-null cells using lentiviral gene transfer and microarray analyses of two independent 503 PU.1-rescued and parental 503 cells were compared. The BCL2A1 gene was found to be more than 50-fold induced in 503 PU.1- restored as compared to the parental 503-null cells. BCL2A1 (also known as BFL-1/A1) is an anti-apoptotic member of the BCL2 family. BCL2A1 was initially identified as a tissue-specific BCL2-related factor that is induced by different reagents such as granulocyte macrophage colony-stimulating factor (GM-CSF) or all-trans retinoic acid (ATRA) during myeloid differentiation. Upregulation of BCL2A1 in granulocytes may promote a time-dependent survival. To follow up on our microarray findings we evaluated loss of PU.1 function in human NB4 acute promyelocytic leukemia (APL) cells using lentivector delivered, short hairpin (sh) RNAs targeting PU.1. Knockdown efficacy upon ATRA-treatment in the two shPU.1 expressing NB4 cell lines was 67 and 30%, respectively. Silencing of PU.1 markedly reduced BCL2A1 mRNA induction upon ATRA-treatment from 167-fold in control cells to 47- and 112-fold in the two PU.1 knockdown NB4 cell lines, respectively (Figure A). Co-transfection of PU.1 with a human BCL2A1 promoter reporter resulted in a 7-fold activation, suggesting PU.1 can directly regulate BCL2A1. Co-transfection with NF-kappaB, used as positive control, induced the BCL2A1 promoter 14.5-fold. Moreover, in vivo binding of the transcription factor PU.1 to 2/8 putative PU.1 binding sites in the BCL2A1 promoter was shown by chromatin immunoprecipitation in HL60 promyelocytic cells further supporting a role for PU.1 regulation of BCL2A1. Evaluation of BCL2A1 and PU.1 mRNA expression in CD34+ hematopoietic progenitors, granulocytes, and primary acute myeloid leukemia (AML) cells was assessed using real-time quantitative RT-PCR. BCL2A1 and PU.1 mRNA levels were significantly lower in primary AML patient samples (n=80; p<0.0001) and in CD34+ progenitor cells (n=4; p=0.0095) than in granulocytes (n=6; Figure B). In line with this observation, we found that upon ATRA therapy BCL2A1 levels were increased in 5/5 APL patients and PU.1 mRNA levels in 4/5 APL cases, respectively. Altogether, these results clearly indicate that PU.1 and BCL2A1 are co-regulated during granulocyte differentiation. Lastly, we confirmed earlier data showing that ATRA-pretreatment of NB4 cells and thus induction of PU.1 and BCL2A1, rendered these cells less sensitive to arsenic trioxide (As2O3)- induced cell death. Conversely, NB4 PU.1 knockdown cells were markedly more sensitive to As2O3 -induced cell death upon ATRA-pretreatment than the parental NB4 control cells. The increase in sensitivity to As2O3 correlated with the lower BCL2A1 levels found in the PU.1 knockdown cells. In summary, we identified the anti-apoptotic BCL2A1 gene as direct, transcriptional target of PU.1 in myeloid leukemic cells. We hypothesize that PU.1-dependent induction of BCL2A1 is necessary for the survival of normal, terminally differentiated myeloid cells. Furthermore, aberrant expression of PU.1 in erythroleukemia may result in elevated BCL2A1 levels that support increased survival of erythroblasts in this particular type of leukemia.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author