The combination of the selective BCL-2 inhibitor venetoclax (VEN) with the hypomethylating agent (HMA) decitabine (DEC) is highly effective in treating older ineligible for intensive chemotherapy AML patients. However, approximately 15-34% of newly diagnosed (ND) and 60% of relapsed/refractory (R/R) patients fail to achieve complete remission (CR) (DiNardo et al. New Engl J Med. 2020, DiNardo et al. Lancet Haematol. 2020). Factors that reduce efficacy remain to be elucidated. To dissect underlying mechanisms of HMA-VEN resistance, we performed RNA-seq utilizing primary samples from patients treated on a prospective clinical trial of 10-day DEC with VEN (NCT03404193). Previously, we reported that alterations of transcriptome signatures of energy-metabolism and cell migration/adhesion pathways were upregulated in primary resistant ND group, and immune-system-associated genes were downregulated in R/R group (Yamatani et al. ASH 2021). This study investigated transcriptomic and DNA methylation signatures associated with DEC10-VEN response or resistance in ND and R/R AML. Because DEC was reported to strongly induce interferon-stimulated genes (ISGs) through gene methylation modifications, we focused on 116 ISGs. An integrated analysis of DNA methylation status around the transcription start site (TSS) was performed using the Infinium methylated EPIC array (Illumina). Forty-four pre-treatment samples from 32 patients with AML from DEC-VEN clinical trial, ND patient group (CR/CRi 14; non-responder, NR 11) and R/R patient group (CR/CRi 6; NR 13), were analyzed. First, we performed the comparative analysis of NR vs CRs. In the ND group, NR cases showed significantly higher gene expression and lower TSS methylation of an urea cycle enzyme, ASS1 (Argininosuccinate Synthase 1) with RNA-seq Log2 fold-change (FC) 3.6 and methylation array Δβ (differential methylation β-value) -0.11, p < 0.05, and of an actin-based molecular motor, MYO1C (Myosin 1C), Log2FC 2.5, Δβ = -0.18, p< 0.05, than in CR cases. In the R/R group, gene expression levels of seven ISGs, including CCL3 (C-C motif chemokine 3), were significantly lower in NR than in CRs, with no difference in methylation status. Next, changes in gene expression levels and TSS methylation status after DEC-VEN treatment were investigated in non-responders in ND group (pre 19, post 12) and R/R group (pre 16, post 4). After DEC-VEN treatment, PPARG (Peroxisome Proliferator-Activated Receptor γ), a transcription factor that activates fatty acid metabolism and negatively regulates subsets of IFN-γ target genes, was significantly upregulated along with its TSS demethylation both in ND (Log2FC 4.9, Δβ = -0.14, p < 0.05), and R/R groups (Log2FC 3.7, Δβ = -0.06, p < 0.05). CDC42EP4, an effector protein of CDC42 involved in the organization of the actin cytoskeleton, was upregulated after the treatment in the R/R group (Log2FC 2.7, Δβ = -0.14, p < 0.05). Based on these findings, we validated transcriptional and methylation changes of ASS1, PPARG, MYO1C and CDC42EP4 by RT-qPCR and bisulfite pyrosequencing using HL60 and MV4-11 AML cell lines in which DEC (IC50, HL60, 249 nM; MV4-11, 2090 nM) and VEN (IC50, HL60, 18 nM; MV4-11, 64 nM) produced additive or synergistic combination effect. Cells were treated by DEC and/or VEN under BM-derived mesenchymal stem cells (MSCs) co-culture conditions to mimic the BM microenvironment. DEC upregulated gene expression levels with TSS demethylation of PPARG, MYO1C and CDC42EP4 both in HL60 and MV4-11, and of ASS1 only in HL-60. VEN treatment alone or in combination with DEC did not induce significant changes in the expression of these tested genes. Of note, PPARG expression was further upregulated under MSCs co-culture condition without TSS methylation status change in AML cells. DEC treatment induced no significant changes of PPARG in MSC. These findings suggest that not only TSS demethylation but also interaction with BM stromal cells may be involved in PPARG upregulation associated with energy production in AML cells. Taken together, DEC-VEN treatment stimulates two key factors, energy metabolism and interaction with the BM microenvironment, which may reduce therapeutic efficacy by increasing gene expression of PPARG,MYO1C,ASS1, and CDC42EP4 through TSS demethylation and/or interaction with BM stromal cells. Experimentally, the consequences of overexpression / knockdown of these genes will be further validated.