Targeted VNTR long read sequencing resolves a diagnostic bottleneck in ADTKD and detects de novo ADTKD-MUC1.

Abstract

The molecular diagnosis of autosomal dominant tubulointerstitial kidney disease due to MUC1 variants (ADTKD-MUC1) using high-throughput (short-read) sequencing methods remains challenging due to the presence of a coding long variable-number tandem repeat (VNTR) region wherein most known pathogenic variants are located. Here, we used targeted amplicon long-read sequencing to study the MUC1 VNTR in a retrospective cohort study of 78 individuals. Using a bioinformatic pipeline including newly developed specialized software, VNTRtools, we reconstruct patient-specific complete VNTR haplotypes, generate a synthetic VNTR reference, perform long-read realignment to this reference and finally perform variant calling. VNTRtools proved efficient, requiring seconds or minutes per sample to accurately identify all pathogenic MUC1 frameshift variants in positive controls, including atypical variants. Ten new diagnoses of ADTKD-MUC1 were made. Furthermore, we report a confirmed de novo case of ADTKD-MUC1 in a 32-year-old patient. We were able to structurally resolve and phase the inter-individually highly variable VNTRs in most probands, enabling the high-confidence detection of pathogenic frameshift variants in 24 individuals. We also detected 18 previously unreported VNTR repeat unit types, demonstrating the highly polymorphic nature of MUC1's VNTR. We propose a combined approach in which short-read VNTR analysis using the published alignment-free bioinformatic tools is used as a first line test, followed by targeted long-read sequencing with VNTRtools analysis for confirmatory testing and in-depth VNTR characterization. This combined approach will lead to a higher diagnostic confidence in ADTKD-MUC1 - especially in sporadic cases. Complete VNTR haplotype information will likely enable a better genetic understanding of this currently underdiagnosed disorder and may become relevant for future therapeutic approaches like targeted silencing of the pathogenic MUC1 allele.