Resistance to Azacitidine Is Determined at Cellular Level By Lower Expression of Nucleoside Metabolizing Enzymes

Introduction: Azacitidine is at present the standard of treatment of myelodysplastic syndromes (MDS). We demonstrated that MDS patients responsive to azacitidine have significantly higher expression of the azacitidine-activating enzyme uridine-cytidine kinase-1 (UCK1) in bone marrow mononuclear cells (Valencia et al. Leukemia 2014). Correlation of the expression of nucleoside transporter, activating and disactivating enzymes with clinical and in vitro response to hypomethylating drugs azacitidine and decitabine has been suggested by several authors. Yet, the crucial role of these enzymes has to be ascertained, as well as their possible different importance in determining resistance.

Objectives: To confirm that the cellular expression of nucleoside metabolizing enzymes plays a major role in cellular resistance and significantly impacts on clinical response to azacitidine.

Methods: Two cell lines, SKM1 sensitive (SKM1-S) and SKM1 resistant (SKM1-R) to azacitidine (Cluzeau et al. Oncogene 2014) were analyzed for expression of UCK1, UCK2, hENT1, hCNT3, RRM1 and RRM2 by quantitative PCR, using an ABI GeneAmp® 5700. The specific oligonucleotides and TaqMan® probes were acquired among the Assay-on-Demand® Gene Expression Products (Applied Biosystems). Corresponding proteins were quantitated by western blotting in both cell lines. The expression of UCK1 and UCK2 was blunted by siRNAs in order to determine their role in in vitro sensitivity to azacitidine.

For UCK2 silencing in SKM1-S, specific siRNAs were used (OriGene Technologies, Inc., Rockville, MD, USA). SKM1-S cells were cultured at a density of 1x10⁶ cells/ml in 9 ml of RPMI 1940 medium. After 72 h of transfection, cells were treated for further 48h with azacitidine at the concentrations of 0,1 and 1 μM. After assessment of effective gene silencing, apoptosis and DNA methylation were evaluated by Annexin V test, MSP and pyrosequencing in cells exposed to azacitidine.

In addition, the expression of nucleoside metabolizing enzymes was evaluated prospectively in IPSS high risk MDS patients treated with azacitidine 75mg/m²/7 days every 28 days.

Results: SKM1-R cells did not express UCK1, UCK2, hENT1, hCNT3, RRM1 and RRM2. Corresponding proteins were also not expressed. In order to clarify the specific role of UCK2 protein in the mechanisms of resistance to azacitidine, we knocked-down UCK2. A reduction of apoptosis was observed in UCK2-silenced SKM-1 S after azacitidine 0.1 μM treatment: 31%±0,85% Annexin V-positive cells versus 21%±0,35% in non-silenced control SKM1-S cultures (P= 0.054). Hypomethylation induced by in vitro azacitidine treatment was also hampered by absence of expression of UCK2.

Quite surprisingly, in MDS cases, gene expression of UCK1, UCK2, hENT1, hCNT3, RRM1 and RRM2 in primary cellsdid not predict different clinical response to azacitidine.

Conclusions: We demonstrated that UCK1, UCK2, hENT1, hCNT3, RRM1 and RRM2 and the corresponding proteins are absent in an in vitro generated azacitidine-resistant cell line SKM1-R. These abnormalities could be major determinants of the induced resistance to the nucleoside analogue. Supporting the importance of these enzymes in resistance to nucleosides, and the major role played by UCK2, its silencing induced by synthetic siRNAs significantly decreased azacitidine effects. Prospective evaluation of the predictive role of cellular expression of genes involved in azacitidine metabolism is ongoing in a larger cohort of MDS patients.

Establishing the mechanisms of resistance to azacitidine could lead us to optimize treatment of MDS patients whose scarce susceptibility could be predictable.