AZA and DAC treated samples were not associated with CGIs, and no array based methylation data was obtained for these genes. However, DNA demethylation was detected in the non-CGI promoters of the top three up-regulated genes by bisulfite sequencing. These data indicate that prolonged low dose treatments are capable of demethylating CpG sites at non-CGI promoters and that this may have an effect on gene expression. The impact of demethylating agents on AML cell lines has recently been evaluated in several studies using bisulfitemodified target DNA arrays. Here we have extended previous observations by investigating the effect of prolonged low-dosage treatment with AZA and DAC in a model, which is likely to be more similar to the clinical situation than previous short-term and/or high-dose treatments. Furthermore, we have investigated the effects in the SKM-1 cell line, which was derived from overt leukaemia following MDS and hence may provide a better model for investigating the relationship between demethylating treatments and MDS. We have used McrBC fragmentation in combination with standard CpG island arrays to robustly distinguish differential CGI methylation profiles in cells proliferating normally. Most of the CGIs are located at either TSS or within gene bodies. Gene-body CGIs are significantly more highly methylated than TSS CGIs. However, this epigenetic mark was preferentially lost at TSS CGIs after prolonged treatment with AZA or DAC. Demethylating agents are thought to act as nucleoside Niraparib side effects analogues that incorporate into DNA, causing specific inactivation of DNMT1. This effect is non-specific and cannot per se explain the selectivity of demethylation observed. In contrast, the de novo methyltransferase DNMT3B are targeted to specific loci and it is possible that their activity contributes to the specificity of the demethylation observed. However, we found a decrease in both DNMT1 and DNMT3B protein levels as a result of AZA or DAC treatment and hence it is unlikely that DNMT3B plays a strong role in the maintenance of DNA methylation at demethylation resistant loci. DNMT1 recognizes hemi-methylated DNA and causes the methylation of the non-methylated strand. A reduction in the level of active DNMT1 should thus lead to the presence of more hemi-methylated DNA VE-822 resulting in a passive demethylation during cell proliferation. It is not known whether DNMT1 is differentially targeted to different genomic regions in SKM-1 cells, but it is feasible that since DNMT1 recognizes hemi-methylated DNA, that it may be preferentially associated with regions of DNA containing high levels of methylated CpGs. In fact, genome-wide mapping data of DNMT family proteins suggests that DNMT1 is depleted in TSS and enriched in the gene bodies. On the other hand, active DNA demethylation mediated by the TET family of methylcytosine deoxygenases may also play a part in selectivity of demethylation. The Tet1 protein binds preferentially to TSSs and less intensively throughout gene bodies. Therefore, a reduction in overall activity of DNMTs may have a stronger demethylation effect at regions that are normally less methylated, such as promoter regions. A similar study on the effect of nanomolar-scale demethylating agents on both AML and breast cancer cell lines has recently been reported. The authors of this study concluded that low-dose DAC affected a sub-population of clonogenic cells, rather than directly inducing cytotoxicity, to produce an antitumor ��memory�� response. These effects were accompanied by sustained promoter demethylation and gene re-expression in key cellular regulatory pathways. In agreement with their data, genes involved in the immune response of the ‘Triggering Receptor Expressed on Myeloid Cells’ -1 signalling pathway were activated after low-dose DNMT inhibitor treatment. Moreover, we demonstrate that the enhanced expression of a subset of these molecules follows DNA demethylation during the course of treatment. Activation of the TREM-1 signalling pathway is a feature of mature differentiated myelomonocytic cells. TYROBP constitutively associates with TREM-1 to mediate the induction of intracellular signals that lead to inflammatory cytokine TNF-a and chemokine IL-8 production. Further investigation into the epigenetic regulation of the TREM-1 pathway may extend our knowledge of the molecular basis of hematopoiesis and myeloid cell differentiation.