Azacitidine (AZA) and Decitabine (DAC) Show Differential Effects On DNA Methyltransferase (DNMT) Levels in Patients with High Risk (HR) Myelodysplastic Syndrome (MDS)

Abstract Abstract 4906 Background: The mechanism of action (MOA) of AZA and DAC is considered to be predominantly via DNMT inhibition. However these drugs are also cytotoxic as nucleoside analogs, and AZA inhibits RNA translation and protein synthesis. Both are clinically efficacious in HR MDS but patient responses differ. We hypothesized that the relative contributions of these functions to overall MOA may be at the basis of this clinical divergence. Previous studies by others (Hollenbach et al, PLOS One 5(2) e9001) in AML cell lines demonstrated that both drugs reduced DNMT1 levels at clinically relevant concentrations and induced apoptosis. However, they differed in effects on the cell cycle: AZA was active across cell cycle phases while DAC induced a G2-M shift. Here we apply single cell network profiling (SCNP) to analyze the effect of AZA or DAC on cell cycle and DNMT levels in bone marrow mononuclear cell (BMMC) specific subsets (CD34+, nRBC, lymphocytes, myeloid cells) from primary HR MDS and healthy donors (HEA). Methods: SCNP is a multiparametric flow cytometry-based assay that simultaneously measures, at the single cell level, extracellular surface markers and changes in intracellular proteins in response to extracellular modulators (including drugs). SCNP was used to measure cell cycle (DAPI stain), apoptosis (amine aqua and c-Parp) and levels of DNMT1, DNMT3a and DNMT3b at baseline and after treatment for 24 hours with AZA (2. 5uM) and DAC (0. 625uM) in AML cell lines (U937, TF-1), primary BMMC from HEA > 60 years of age (n=15) and HR MDS (n=10, RAEB, int-2/high) patients. The Mann-Whitney U statistic was used for analysis. Results: Cell cycle and apoptosis. Treatment of U937 cells with AZA increased the subG1 cell fraction (i. e. apoptotic cells) from 0 to 12% and decreased the percentage of cells in S phase from 37% to 24%. Conversely treatment with DAC increased the fraction of cells in S phase (from 37% to 54%), consistent with DAC intercalation into DNA, with no significant effects on the subG1 fraction. In TF-1 cells, only AZA significantly induced apoptosis as measured by both amine aqua (10%) and c-Parp (5%). Similar results were observed in primary HR MDS samples: in 4 of 8 HR MDS samples AZA treatment resulted in apoptosis in CD34+ cells (median= 20. 5% induction) while DAC did not affect CD34+ cell viability. DNMTs levels. In U937 cells AZA decreased all 3 DNMTs levels (DNMT1(41%), DNMT3a(26%), DNMT3b(40%)), more effectively than DAC (4%, 4% and 2% decrease, respectively). In primary samples, nRBC subsets from HR MDS showed higher basal levels of DNMT3a and DNMT3b compared to healthy age matched BMMC (p= 0. 003 and p=0. 029 respectively). In the latter, AZA and DAC both decreased the levels of DNMT1, DNMT3a and DNMT3b in all BMMC cell subsets (lymphocytes, nRBC, myeloid and CD34+). The effects of these drugs in HR MDS contrast with healthy samples, as summarized in Table 1. Specifically: 1) both drugs increased the levels of all three DNMTs in lymphocytes, 2) DNMT1 levels were increased by both drugs in lymphocytes, myeloid and CD34+cell subsets, and 3) effects of the two drugs diverged with respect to effects on DNMT 3a and 3b in myeloid cell subsets (Table 1). Conclusions: In vitro treatment for 24 hrs with clinically relevant concentrations of AZA or DAC decreased baseline DNMTs levels in cell lines and healthy BMMC, consistent with their role as promoters of genomic hypomethylation. However, these drugs have strikingly different effects on primary HR MDS BMMC cell subsets. While MDS is not typically considered a disorder of lymphocytes, DNMTs levels increased in response to either AZA or DAC suggesting an alteration in HR MDS lymphocytes vs. HEA lymphocytes. The combined differential activities of AZA and DAC on cell cycle, intracellular DNMTs levels and apoptosis in specific cell subsets (in particular DNMT3a levels in myeloid and CD34+ HR MDS BMMC and absence of DAC-induced apoptosis) suggest the inhibition of post-transcriptional processing by AZA contributes significantly to cytotoxicity and may explain differences in clinical activities of the 2 drugs. Disclosures: Cohen: Nodality, Inc.: Consultancy. Huang:Nodality, Inc.: Employment, Equity Ownership. Hawtin:Nodality, Inc.: Employment, Equity Ownership. Ware:Nodality, Inc.: Employment, Equity Ownership. Cesano:Nodality, Inc: Employment, Equity Ownership.