Fig. 1: A highly optimized SGE protocol to assay VHL variants.
From: Saturation genome editing maps the functional spectrum of pathogenic VHL alleles
a, CRISPR knockout screening data from the Cancer Dependency Map (DepMap)26 reveal VHL loss widely leads to reduced growth in cell lines lacking VHL mutations (n = 29 kidney-derived and n = 1,048 other lines; boxplot: center line, median; box limits, upper and lower quartiles; whiskers, 1.5× interquartile range; all points shown). b, CRISPR-induced editing of VHL was performed in HAP1 cells (day 0), and outcomes were quantified by NGS. Distributions of InDel scores, calculated as the log2 ratio of day 13 frequency to day 6 frequency, show frameshifting, and in-frame InDels are strongly depleted in parental HAP1 (left) compared to HAP1–HIF1A–knockout (–KO) (right) (median InDel score −3.20 versus −0.20, respectively; Wilcoxon rank-sum two-sided P = 5.6 × 10−18). c, The strategy to perform SGE across the complete coding sequence of VHL is shown, with ClinVar variant counts for all ‘pathogenic’ and ’likely pathogenic’ variants (red) and VUS (orange) displayed from gnomAD39. SGE regions were designed to tile exons 1–3, as well as a region of intron 1. A total of 480 SNVs in ClinVar are in SGE regions, of which 269 are VUS. (Introns not to scale.) d, For each SGE region, a library of oligos containing all possible SNVs was synthesized and cloned into a vector with homology arms to facilitate genomic integration via CRISPR-induced HDR. Variants present in cells were quantified over time via amplicon sequencing, and function scores were calculated to reflect variants’ effects on fitness. e,f, Function scores for synonymous, nonsense and canonical splice site SNVs are shown for a single SGE region (exon 2) assayed in normal media (e) or media supplemented with 2.5 µM DAB (f).
