Resolving the classification rates and molecular architecture of early-onset chronic kidney disease with NephVar

Abstract

The genetic architecture of early-onset chronic kidney disease (CKD) is caused by more than 200 monogenic genes, where their common diagnostic classes include congenital anomalies of the kidney and urinary tract, steroid-resistant nephrotic syndrome, nephronophthisis-related ciliopathies, chronic glomerulonephritis, and urinary stone disease. While advancements in whole-exome and whole-genome sequencing have enabled identification of disease-causing variants, their rates of classification have remained unknown. Likewise, the molecular effects of these pathogenic variants remain unresolved, which is essential for improving personalized treatment approaches. In this study, we collected clinical and biophysical data from 117,373 genetic variants across 129 monogenic genes causing early-onset CKD. This data established the NephVar registry, which aims to be a molecular dictionary for nephrologists to classify variants and resolve their unique molecular effects. Through NephVar, we estimated 1-15% of alleles are reclassified and the time to reclassification per variant is 2-12 years in early-onset CKD. Furthermore, NephVar identified the molecular effects of all variant types, emphasizing missense variants. Our analyses indicate that intrinsically disordered regions of proteins are protective against disease-causing missense alleles across most diagnostic classes, but often occur through a buried loss-of-function (LoF) mechanism. Additionally, we show that the mode of inheritance for these monogenic genes influences clustering patterns of genetic variants, where autosomal dominant (AD) genes are more clustered than those of autosomal recessive (AR) genes. This data accurately predicted the non-LoF effects in INF2, PAX2, GATA3, ACTN4, and LMX1B causing inherited nephrotic syndromes. We demonstrate that variant effect prediction is effective for downgrading variants of unknown significance and classifying AR genes, but challenging for pathogenic alleles in AD genes. Lastly, we propose standards and guidelines for determining non-LoF effects, including gain-of-function and dominant negative, in inherited nephrotic syndrome. Overall, the NephVar renal registry has important implications for defining the molecular architecture and estimating the progress of molecular diagnostics for early-onset CKD.