Systematic Classification of WT1 Zinc Finger Missense Mutations Using Multiplex Assays of Variant Effect (MAVEs)

Background: Patients with Mendelian nephrotic syndrome (NS) face a high risk of progressing to end-stage kidney disease. Genetic screening is increasingly accessible and can guide clinical decisions. However, it remains challenging to accurately class variant's pathogenicity. MAVEs are a new category of functional assay that can be applied to many variants at one time in a pooled fashion. Here we apply MAVEs to systematically and prospectively test variants in the zinc finger domains of WT1, one of the most common genes underyling Mendelian NS. Methods: We derived clonal WT1 knockout (KO) cell lines from KBM7, a myeloid leukemia cell line that normally expresses WT1. We then synthesized a saturation mutagenesis library for WT1 cDNA, encompasing every possible missense mutation across three zinc-finger motifs (exons 7-9, n=1028 mutations). This library was stably integrated into KBM7 WT1 KO cells using a lentiviral indcuble expression construct, such that each cell expresses a different WT1 variant. Because WT1 loss of function is growth-promoting in KBM7 cells, we performed serial outgrowth of the WT1 library population for four passages to enrich for loss-of-function mutatinos. We then deeply sequenced the WT1 library from cells at passage 4 v.s. baseline, and calculated an enrichment score of each mutation. Results: We observed perfect separation between nonsense and synonymous mutations in functional scores. Missense variants were nearly all deleterious to some extent (89%), with nearly a third of these being equivalently deleterious to a nonsense variant, indicating a strong constraint against coding variation in this key DNA-binding domain. Nevertheless, some mutations have relatively mild effects, especially R435Q, which is also present in apparently unaffected individuals in the general population. Mutations at essential residues that coordinate Zn2+ ions were particularly deleterious, comparable to truncating mutations. Conclusion: The completion of a variant-to-effect map for this region validates the feasibility of extending this approach to the full WT1 protein. The maps will serve as a freely available ‘lookup table’ of functional evidence to support the clinical interpretation of newly encountered variants. This approach can improve the accuracy of genetic diagnosis for NS and remove one barrier to future genotype-guided therapies. Funding: Other NIH Support - 5K12HD028820-28:Clinical Scientist Institutional Career Development Program Award