A recently published study from the laboratory has shown that acute myeloid leukemia patient samples have undergone different degrees of methylation throughout the CpG island of the gene. We observe three types, those that have a high degree of hypermethylation, those that have essentially no methylation and those that have an intermediate degree of methylation. In all cases the gene was not expressed. When there is no hypermethylation in patient samples at the p15INK4b locus CpG island we presume that there are other mechanisms that inhibit expression of the gene and for some acute myeloid cases we have determined an alternate pathway. In acute myeloid leukemia with inv(16) we have found that the INK4b silencing is caused by another mechanism. In these leukemias the inv(16)-encoded core binding factor beta-smooth muscle myosin heavy chain (CBFbeta-SMMHC) protein targets the promoter and displaces the transcription factor RUNX1 causing transcriptional repression. For the inv(16) acute myeloid leukemia patients, re-expression from the INK4b locus would not be predicted to occur using hypmethylating drugs. Due to their function in the regulation of gene expression histone modifications represent an additional epigenetic pathway that may be altered in AML. A focus in our present research is to address the histone modifications that regulate p15INK4b in AML cells and how these marks are altered in cells that have p15INK4b DNA methylation. For this analysis high density CHIP-on-chip are being utilized that cover the INK4b-ARF-INK4a region of chromosome 9 at a 20 bp resolution. An important observation that has resulted from this study is that in AML patient blasts with p15INK4b DNA methylation, H3K4me3 occupancy is reduced around the p15INK4b promoter when compared to AML blasts without DNA methylation. Therefore, it is possible that acquisition of DNA methylation at an earlier stage of cell development or leukemia development could alter the chromatin structure around the CpG island, preventing the H3K4 histone methyltransferase from accessing the histones in this region. This would suggest also that DNA methylation serves to solidify repression at the promoter by preventing the gene from acquiring activation-associated histone marks. Another important observation in our study is that H3K27me3 enrichment spans p15INK4b and extends into p14ARF and p16INK4a in AML patient blasts irrespective of the DNA methylation status of p15INK4b. Although a previous report implicated increases in the repressive H3K9me3 mark in AML cells with p15INK4b DNA methylation, we found low levels of this mark at p15INK4b in all AML cells, indicating H3K27me3 is the primary repressive mark at this gene. In AML clinical samples without p15INK4b DNA methylation, H3K27me3 enrichment overlaps the area of activation-associated H3K4me3, creating a bivalent histone modification pattern at the p15INK4b promoter typically found at developmentally regulated genes in stem and progenitor cells. Genes in this conformation are typically maintained in a transcriptionally poised state in which expression is low. Accordingly, we find expression of the p15INK4b transcript is low even in samples in which p15INK4b is unmethylated. Since the region of H3K27me3 enrichment is also bound by the EZH2, our results imply that polycomb-mediated repression is important in the regulation of p15INK4b and as well as p14ARF and p16INK4a in AML. It is of interest that the histone modification pattern we observed in AML cell lines did not fully recapitulate the patterns we described in AML clinical samples. Cell lines free of p15INK4b DNA methylation (U-937 and HL-60) resolve the bivalent histone modification state at p15INK4b, displaying high levels of H3K4me3 at the p15INK4b promoter in the absence of H3K27me3 enrichment throughout the INK4b-ARF-INK4a region. Although these cell lines may represent a unique subpopulation or differentiation state of myeloid leukemia cells, changes acquired during the establishment of these cell lines may explain differences between unmethylated cell lines and clinical samples in our studies. Our laboratory has shown that when p15INK4b expression is reactivated with decitibine (an inhibitor of DNA methylation) and Trichostatin A (a deacetylation inhibitor) in AML with p15INK4b DNA methylation H3K4me3 is increased at the promoter region while maintaining H3K27me3 throughout the gene. Therefore, epigenetic reactivation restores the promoter to a bivalent state observed in unmethylated AML clinical samples. These data indicate AML cells with p15INK4b DNA methylation have an altered histone methylation pattern compared to unmethylated samples and that these changes can be reversed by epigenetic therapies. Removal of DNA methylation at p15INK4b was shown to only marginally increase the transcriptional activity of the gene possibly because H3K27me3 is still present at high levels at p15INKb. Since it has been shown that the drug 3-deazaneplanocin A (DZNep) can indirectly deplete H3K27me3 levels and reduce EZH2 protein levels in AML cells. (Fiskus et al., Blood, 2009), we hypothesized that the combined treatment with 5-aza-dC and DZNep can modulate histone methylation at p15INK4b, allowing an increase in H3K4me3 and a decrease in H3K27me3. Resolving the bivalent histone methylation (H3K4me3/ H3K27me3) towards an activation state may improve epigenetic reactivation of the p15INK4b tumor suppressor gene in AML. Our preliminary experiments show that reactivation of p15INK4b by 5-aza-dC is inhibited by DZNep. Furthermore, DZNep treatment had no effect on p15INK4b DNA methylation or gene expression. Further experiments are being performed to determine an explanation for these unexpected result.
| Paul, Thomas A; Bies, Juraj; Small, Donald et al. (2010) Signatures of polycomb repression and reduced H3K4 trimethylation are associated with p15INK4b DNA methylation in AML. Blood 115:3098-108 |
