Identifying Novel Pharmacogenetic Markers for Cancer Treatments: Exploiting Cancers Deficient in the Nucleotide Pool Sanitizing Hydrolases NUDT15 AND NUDT18

The NUDIX hydrolases are a super family of enzymes with purported enzymatic activities of nucleotide hydrolases. The activity of the founding member, NUDT1 (also known as MTH1) is established to be a nucleotide hydrolase for d-8-oxoGTP, which is a mutagenic precursor if incorporated into DNA. Therefore, its activity serves to sanitize the nucleotide pool of mutagenic precursors. While known to be highly conserved among all organisms, the enzymatic and biological functions for nearly all the family members remain largely unexplored and unknown. Substrates of family members have been found to include canonical (deoxy)nucleotide triphosphates, oxidized (deoxy)nucleotide triphosphates, and non-nucleoside polyphosphates. Understanding the role of these enzymes would provide valuable information to aid in determining routes of cancer treatments when attempting to alter contents of the nucleotide pool. Of particular interest for this study are the members NUDT15 and NUDT18. NUDT15 has recently been identified as having a role in the metabolism of the active metabolites of 6-thioguanine (6- TG) and the anti-viral acyclovir (ACV), with both compounds showing greater therapeutic effect in NUDT15-deficient cells. NUDT15 also has an interesting link with the cell cycle regulating protein RB1, where a loss in copy number of RB1 also tends to create a loss in NUDT15 copy number as an apparent passenger deletion. Though this mechanism is not understood, both genes being within 500 kb of each other on Chromosome 13, and carcinogenic processes that drive loss of the important tumor suppressor protein RB1 also causes loss of NUDT15. The role of NUDT18, on the other hand, remains largely unexplored with very little known of its enzymatic activities or biological functions. Here, using CRISPR generated NUDT15 and NUDT18 knock out cell lines, we attempted to identify novel substrates for NUDT15 and NUDT18 using previously published in vitro screenings as well as publicly available database mining. We also characterize the Achilles heel of NUDT15 deficiency, namely cellular sensitivity to thiopurine drugs used as anticancer and immunosuppressive agents, and the antiviral agent acyclovir (ACV). We define the mechanism of action for 6-thioguanine (6- TG) and ACV treatments in NUDT15-deficient ovarian and prostate cancer cells. Using long-term treatment selectivity assays, we show with live-cell imaging that NUDT15- deficient ovarian cancer cells are more sensitive over time to 6-TG and ACV compared to NUDT15-proficient cells. Cell cycle analysis of NUDT15-deficient ovarian cancer cells shows a G2/M phase arrest following 6-TG treatment and a dramatic S-phase arrest after ACV treatment, consistent with their hypothesized mechanisms of action after incorporation into DNA. Also, using assays to assess DNA damage, we show that 6-TG and ACV treatments cause increased DNA damage in the NUDT15-deficient cells. Collectively, these results support using 6-TG or ACV as treatments in cancers found to be deficient in NUDT15. Because RB1 loss is a common carcinogenic event in a substantial subset of all cancers including prostate and NUDT15 is an apparent passenger loss in these, cancers, this represents a therapeutic window exploitable with extensively used, inexpensive therapeutics with highly defined dosing regimens and known, manageable side effects. In other words, this represents a true personalized medicine option for a subset of patients for which molecular genetic diagnosis is readily available