Abstract
Individuals with Down syndrome (DS) are genetically predisposed to hematologic malignancies, including acute myeloid leukemia (AML) before the age of 5. AML and myelodysplasia in children with DS are collectively defined as myeloid leukemia associated with DS (ML-DS) by WHO in 2016. ML-DS patients are highly sensitive to cytarabine (Ara-C)-based chemotherapy resulting in significantly higher overall survival rates compared to non-DS children with AML. However, relapsed/refractory (R/R) ML-DS patients have dismal clinical outcomes, worse than their non-DS AML counterparts, highlighting the need for a better understanding of ML-DS Ara-C-resistance mechanisms and the development of effective therapies for this vulnerable group of patients. There are currently no biomarkers to predict chemotherapy responses at diagnosis; therefore, we are seeking to enhance the frontline chemotherapy regimen to improve initial treatment response to delay or prevent relapse of ML-DS.
The therapeutic efficacy of Ara-C depends on its intracellular metabolism. Ara-C is phosphorylated by deoxycytidine kinase (dCK) into its active form, Ara-CTP, which incorporates into DNA, causing DNA damage and apoptosis. We have demonstrated that R/R ML-DS cells exhibit decreased levels of dCK, rendering them unresponsive to Ara-C–based chemotherapy. Additionally, dCK is suppressed by elevated intracellular deoxynucleotide triphosphates (dNTPs) that facilitate DNA repair, further diminishing Ara-C efficacy. Ribonucleotide reductase (RNR) consisting of 2 subunits, RRM1 and RRM2, is the rate-limiting enzyme in the biosynthesis of dNTPs. Sterile Alpha Motif and Histidine-Aspartate Domain–Containing Protein 1 (SAMHD1) which hydrolyzes excess dNTPs balances dNTP levels in the cells. Multiple studies have found that SAMHD1 functions to revert Ara-CTP back to Ara-C prior to DNA incorporation making it a crucial enzyme implicated in Ara-C resistance. Thus, RNR and SAMHD1 could play a role in Ara-C resistance in ML-DS.
Our proteomics data show overexpression of both RRM1 and SAMHD1 in Ara-C–resistant CMY cells compared to sensitive CMK cells. Moreover, metabolomics analysis revealed significantly reduced dNTP levels in CMY cells, suggesting functional hyperactivation of SAMHD1 and enhanced hydrolysis of intracellular dNTPs. Based on these findings, we hypothesized that targeting RNR would lower dNTP pools, reduce SAMHD1 activity and result in an increase of intracellular Ara-CTP and improved therapeutic efficacy of Ara-C. We propose hydroxyurea (HU), an FDA-approved RNR inhibitor used to reduce leukemic burden in AML patients, as a promising agent to improve Ara-C effectiveness in ML-DS.
To begin to test this hypothesis, we generated CMY knockdown models using lentiviral shRNA targeting RRM1 and SAMHD1 (designated CMY shRRM1 and CMY shSAMHD1). Knockdown efficiency was confirmed by western blot. Following Ara-C treatment, Annexin V and propidium Iodine staining and flow cytometry analysis revealed increased apoptosis in both knockdown lines, with a more robust effect observed in shSAMHD1 cells when treated with Ara-C. Metabolomics analysis further showed that SAMHD1 knockdown led to intracellular accumulation of dNTPs and Ara-CTP, consistent with its enzymatic activity.
We then tested the efficacy of HU as a therapeutic strategy to enhance Ara-C activity. Flow cytometry analysis showed increased apoptosis and strong synergy between HU and Ara-C in both CMK and CMY cell lines. Ongoing studies involve in vivo evaluation of the efficacy of HU and Ara-C combination in comparison to Ara-C alone in an initial diagnostic ML-DS patient derived xenograft model.
Together, our findings identify dysregulated nucleotide metabolism, driven by RNR and SAMHD1, as a key mechanism of Ara-C resistance in ML-DS. Genetic targeting of these enzymes restores Ara-C sensitivity in resistant cells. Repurposing HU, a low-toxicity, FDA-approved agent, represents a promising strategy to enhance Ara-C efficacy and delay or prevent relapse in ML-DS. These results support further preclinical and translational studies evaluating nucleotide metabolism as a therapeutic vulnerability in pediatric AML.
