Next generation sequencing (NGS) in the UK steroid resistant nephrotic syndrome (SRNS) study reveals complex genetic heterogeneity

Abstract

This national cohort was developed within the rare renal disease registry (RaDaR) with detailed clinical phenotyping. Sequencing was undertaken for (470) genes in the first 24 patients with SRNS (either acquired or congenital) as a pilot project. Nimblegen Roche multiplex capture array (12×135KB) was used to capture exons from genomic DNA and sequencing was undertaken on an Illumina GAII in-house. Data was processed using CLC Genomics workbench software for mapping against human genome (hg19/GRch37) and detecting SNPs (single nucleotide polymorphisms) and DIPs (deletion insertion polymorphisms). Any variations were checked against dbSNP (132) and the Human Genome Mutation Database (professional version). All amino acids changes were further analysed in a prediction tool MutPred (www.mutpred.mutdb.org) for probability of pathogenicity. 97% of reads mapped to the reference sequence and 76% of reads were on target genes. On average, 3400 SNPs were found per patient, 185 being non-synonymous with approximately 10 novel per patient. An average total of 2 DIPs were found per patient. SNP and DIP analysis revealed several known disease causing variations in Nephrin, Podocin, PLCe1 but also novel mutations in these and other genes. In one patient an isolated COQ2 nephropathy (single previous case report) was revealed without other organ involvement. In families, addition of single heterozygote mutations on top of homozygous disease causing mutations altered phenotype causing early onset of disease. All variations considered clinically significant were verified with Sanger sequencing with no false positives identified. All novel variations were evaluated within MutPred and several were identified as being probably pathogenic, warranting further investigation. Although knowledge of genes involved in this condition has evolved rapidly in the last ten years, prevalence in the UK cohort has never been tested before. This study suggests a detection-of-mutation rate in sporadic disease of approximately 25% and in familial of 40% and also reveals that tri-allelic hits are not uncommon and may alter phenotype. Genotype - phenotype analysis of the extended cohort will allow for development of prognostic variants which could ultimately be used in the early clinical setting.