Abstract
Lynch syndrome (LS) is the most prevalent hereditary cancer syndrome as it affects an estimated 1 of 279 people. LS is caused by germline mutations in mismatch repair (MMR) genes MLH1, MSH2, MSH6, PMS2, and EPCAM. Deletion mutations of EPCAM have been found to affect MSH2 expression. These heterozygous mutations do not cause the cancer phenotype, but increase the somatic mutation rate, which increases cancer risk. PMS2 is the most frequently mutated MMR gene, yet the number of LS cases does not reflect this mutation rate. Historically, there have been sequencing challenges because of numerous pseudogenes, which have impacted the identification and classification of PMS2 mutations. These mutations must be examined to understand genetic variability and characterize pathogenicity risk rates. Single nucleotide variants that cause amino acid modifications, are considered missense variants. Some missense variants result in protein loss-of-function; however, 87.9% of PMS2 missense variants are classified on ClinVar as variants of uncertain significance (VOUSs). LS VOUSs are clinically inactionable without additional evidence supporting their pathogenicity. Typically, LS risk is mitigated through frequent cancer screenings and, in some cases, prophylactic surgery. After a PMS2 missense VOUS, c.86G>C, was identified through screening families suspected of hereditary breast and ovarian cancer (HBOC) within the Lynch Memorial Biobank, we developed a mid-throughput in vitro assay to functionally assess MMR activity of PMS2 VOUSs. First, PMS2 knockout cell lines were generated using CRISPR-Cas9 technology in the human haploid HAP1 cell line. Individual PMS2 variants were produced through site-directed mutagenesis and delivered into the PMS2-deficient cell line via lentiviral transduction. PMS2 variant expression was induced by doxycycline prior to MMR activity assessment through 6-thioguanine (6-TG) treatment. To measure the MMR activity of variants, we analyzed the cell viability of PMS2 variant lines post 6-TG treatment. High confidence variants were selected and run through the model. As expected, well-established benign variants, c.59G>A and c.1408C>T, were sensitive to 6-TG treatment, indicating MMR proficiency. A well-established pathogenic variant, c.2444C>T, also displayed 6-TG sensitivity; however, after model corrections, was 6-TG resistant. We pose a comprehensive approach for analyzing PMS2-specific MMR activity of missense variants ready for large scale validation.