Probing the RecQL4 homolog Hrq1: A genetic suppressor screen identifies critical functional features

The DNA repair helicase RECQL4 is a potent oncogene whose dysregulation is a hallmark of multiple cancers and diseases of genomic instability. Overexpression of RECQL4 is observed in human breast, ovarian, and liver cancers, where its expression is a biomarker for poor prognosis. Conversely, germline loss-of-function perturbations in the ATPase domain of the protein are linked to genetic diseases like Rothmund-Thomson syndrome that are characterized by high risk of developing osteosarcoma and lymphoma. This bifurcated role highlights that RECQL4 dosage and activity must be precisely tuned. Additionally, RECQL4 has a highly conserved function in the repair of DNA interstrand crosslink (ICL) damaging agents such as platinum-based chemotherapeutics, making this an important protein to understand for overcoming chemoresistance. However, mechanistic studies of RECQL4 are difficult to study in vivo as well as overexpress and purify because its N terminus contains an essential replication-initiation domain equivalent to Saccharomyces cerevisiae's Sld2. Therefore, we study the S. cerevisiae homologue of RECQL4, Hrq1. We use this model to discover more details about this protein's functions that can then be linked back to the human system. The ATPase-dead mutant of Hrq1 ( hrq1-K318A ) is profoundly hypersensitive to the ICL agent diepoxy butane (DEB), exhibiting a lethality far greater than the complete deletion strain ( hrq1Δ ). This defines a dominant-negative "protein-trapping" or "roadblock" phenotype. To identify how this toxic protein-DNA complex is resolved, we generated spontaneous suppressor colonies by plating hrq1-K318A strains on DEB. Whole-genome sequencing revealed that over 90% of screened colonies in the hrq1-K318A strain had a secondary mutation within the hrq1 gene itself. Crucially, spot dilution assays confirmed that these intragenic suppressors rescue the K318A hypersensitivity back to the level of the hrq1Δ deletion. This provides powerful genetic evidence for the "roadblock" hypothesis, demonstrating the suppressors function by "defusing" the toxic K318A protein, reverting its phenotype to a simple loss-of-function. Further analysis of those results has revealed that many of the mutations are at highly conserved residues amongst RECQL4 and its homologues, suggesting important functional properties at those sites. Finally, we compared these genetic suppressors to a rationally-designed, DNA-binding-deficient mutant ( hrq1-R739A ). While in vitro assays confirmed reduced DNA binding, this mutation in vivo provided only a slight rescue of the K318A toxicity. This finding suggests the dominant-negative "trap" is a complex intermediate stabilized by more than DNA-binding alone-likely including key protein-protein interactions-and reveals critical mechanistic nuances for the design of future RECQL4-targeting therapeutics.