Exome sequencing of a pedigree with Caudal Regression Syndrome (CRS)

Abstract

Caudal Regression Syndrome (CRS) is a rare, congenital disorder characterized by varying degrees of agenesis of the caudal spine, and is usually associated with general anomalies of other systems. Although maternal diabetes is a well-known risk factor, the exact cause of CRS is still unknown. CRS mainly occurs sporadically, except for Currarino syndrome (CS) or CRS sub-phenotype IV, which follows an autosomal-dominant pattern of inheritance. We present a three generation Caucasian family of 11 individuals of whom 6 are affected with CRS subtype III or V based on the sacral/coccygeal agenesis pattern. A presacral mass was present in 5 of the 6 affected individuals. The medical history of family was re-assessed and the phenotype/individual ID was confirmed. G-banding karyotype of the affected members was normal. To identify the genetic defect underlying the disorder, DNA from the 11 family members were submitted to whole genome single nucleotide polymorphism genotyping (Omni2.5) and whole exome sequencing. Further, using 31,520 pruned tag SNPs, parametric linkage analysis using MERLIN revealed 4 linkage peaks with LOD score ≈1.8 under the autosomal dominant inheritance model. The linkage peak regions lied in chr1, chr11 and chr12 and spanned 19.57Mb (ranges from 1.3Mb to 8.2 Mb; ≈ 33.18cM). Whole exome sequencing for the identification of causal variants was performed using Illumina HiSeq platform with TruSeq Exome enrichment kit, and SOLID5500 platform with Agilent SureSelect Exome enrichment kit. The sequencing data retrieved ≈65K recalibrated single nucleotide variants/insertions/deletions for each individual on average and the output showed a concordance rate of 96.2% with Omni2.5 (reads mapping by bwa/bfast+bwa/mrsFast; post-processing and variants calling by GATK pipeline/mpileup/Varscan/freebayes). Besides, we performed structural variants (SVs) calling from Omini2.5 density file using PennCNV, iPattern, QuantiSNP, or CNVPartition as well as from the exome sequencing data using BreakDancer-Max, VariationHunter, Pemer, SpitRead or ExomeDepth. Filtering for causal variants was performed according to 1) novel small variants within the linkage peak regions which were shared only by patients; 2) large structural variants (>0.1Mb, based on the tag SNP resolution for linkage analysis) each of which were examined for the inheritance pattern observed. Several plausible genes with functional mutations have been identified.