Recently, it was reported that reduced level of PU.1 expression developed acute myeloid leukemia (AML) in mice. To clarify the mechanisms of AML, first we established PU.1 knockdown K562 (K562PU.1KD) cells, that express reduced level of PU.1 by stably transfected PU.1 siRNA. Flow cytometric analysis demonstrated that among other lineage markers, expressions of myeloid CD13/CD33 markers are significantly decreased in K562PU.1KD cells compared to control cells. We next tried to identify PU.1 target genes to elucidate the mechanisms of AML. To solve this, microarray and real time PCR analyses were performed. We revealed that several genes including metallothionein (MT) s (MT-1G, MT-1A) were markedly induced in K562PU.1KD cells. MTs are a group of low-molecular weight, cysteine rich intracellular proteins. A number of studies have shown an increased expression of MTs in various human solid tumors. To verify the relevance, we established PU.1 over-expressed K562 (K562PU.1OE) cells and found the significant suppression of these genes. Furthermore, we revealed a negative correlation between PU.1 and MTs mRNA expression in 43 primary AML specimens (MT-1G; R=−0.50, p<0.001, MT-1A; R=−0.57, p<0.0005). These results suggest that MTs are bona fide targets of PU.1. To clarify MTs regulation by PU.1, chromatin immunoprecipitation assay was employed and demonstrated an increased association of acetylated histone H3 and H4 with the MT-1G and MT-1A promoter in K562PU.1KD cells, compared to control cells. As there are 38 CpG sites in approximately −430 to +170 bp from transcription start site in MT-1G and MT-1A promoter, we next performed bisulfite DNA sequencing analysis to examine the methylation status of these CpG sites. We revealed that the proportion of the methylated CpG sites in MT- 1G, as well as MT-1A promoter were significantly decreased in K562PU.1KD cells (MT- 1G: control cells, 29–41% vs K562PU.1KD cells, 16–29%; MT-1A: control cells, 12% vs K562PU.1KD cells, 4%). In contrast, methylations of these sites were markedly increased in K562PU.1OE cells (MT-1G: control cells, 36% vs K562PU.1OE cells, 59%; MT-1A: control cells, 12% vs K562PU.1OE cells, 18%). These suggest that downregulation of PU.1 leads to the impairment of the recruitment of DNA methyltransferase (DNMT) and histone deacetylase, results in the aberrant gene expression in leukemia cells. Therefore, we next hypothesized that there is some specific effect of epigenetic drugs on PU.1 transgenic cells. From WST assay and trypan blue dye exclusion assay, we observed that K562PU.1KD cells were resistant to DNMT inhibitor 5-aza-2′-deoxycytidine (5-Azadc). Viable cells were approximately 20% increased in K562PU.1KD cells in the presence of 5-Azadc, and 20% decreased in K562PU.1OE cells compared to control cells respectively. These suggest that expression of PU.1, MTs and proportion of methylated CpG sites in MTs’ promoter may useful therapeutic marker to predict the effect of 5-Azadc on leukemia cells.

Disclosures: Takahashi:Kitasato University: Research Funding.

Author notes

Corresponding author