ASSESSING GENOMIC SAFETY OF ANTIVIRAL NUCLEOSIDE ANALOGS IN HEMATOPOIETIC STEM CELLS

ABSTRACT

Background: Nucleoside analog (NAs) drugs are used for the treatment of a variety of diseases, such as cancers and viral infections. After phosphorylation of viral and host kinases, NA drugs compete with the corresponding naturally occurring nucleotide during DNA replication of the infected cell. After incorporation, they can lead to mutations, or chain termination. However, because of their mechanism of action, NA are also potentially mutagenic to the genome of physiologically normal cells. Indeed, we have shown that treatment with the antiviral NA ganciclovir (GCV) after stem cell transplantation induces an increased mutation burden in the hematopoietic stem and progenitor cells (HSPCs) of pediatric leukemia patients. Using mutational signature analysis, we provided evidence that GCV-induced mutagenesis contributes to development of relapses and second malignancies in pediatric patients by inducing driver mutations. Over 30 NA drugs have been approved for clinical use and millions of people receive antiviral treatment worldwide to treat viral infections, including COVID-19. However, the mutagenicity in normal cells and potential carcinogenicity is unclear. Aims: Here, we aimed to systematically assess the mutational consequences of antiviral NAs in human HSPCs and identify underlying mechanisms. Methods: By combining in vitro treatment of umbilical cord blood-derived HSPCs with whole-genome sequencing (WGS) analyses, we provide a compendium of mutational consequences of antiviral NAs in a relevant human tissue (i.e., toxicity to the hematopoietic system is often dose-limiting). We treated HSPCs with IC40-60 concentration of the assessed compound followed by clonal expansions to obtain sufficient DNA for WGS. Using established bioinformatic pipelines, we catalogued the somatic mutations and mutational signatures in these cells. Results: At time of writing, 5 out of 7 tested antiviral NAs induce an enhanced mutation burden in exposed HSPCs. For some of this antiviral NAs we were able to identify unique unreported mutational signatures. Of note, the thymidine analog brivudine showed the highest increase in single base substitutions, which were characterized by a T>C signature, depleted for flanking cytosines. Furthermore, like GCV, we also observed a signature characterized by C>ApA substitution after treatment with the penciclovir, a molecule nearly identical to GCV. Currently we are working on machine learning approach to identify relevant mutation characteristics and modes of action as well as to screen cancer genome databases for mutational signature occurrence. Summary/Conclusion: Many compounds of the NA class currently prescribed for the treatment of viral infections are mutagenic to healthy cells. This calls for more thorough screening of these drugs, incorporation of information on mutagenicity to healthy cells in drug safety guidelines and patient surveillance over time.