WES in nonprogressive congenital ataxia: Diagnostic yield and identification of novel loci

Abstract

Objective: To explore the usefulness of whole-exome sequencing (WES) in the etiologic study of 14 unrelated patients with congenital ataxia (CA). Methods: All patients presented within the first year of life with hypotonia and developmental delay, followed by overt ataxic signs thereafter; intellectual disability was present in 10/14. Family history was negative but there were two consanguineous couples. All cases lacked any features allowing direct genetic testing, such as characteristic hindbrain abnormalities or metabolic biomarkers. MRI performed at the outset (age range 6–24 months) was normal (n = 6) or showed mild cerebellar atrophy (n = 8), in one case with hypomyelination. WES capture was performed in all 14 trios using Nimblegene SeqCap EZ Human Exome version 3, followed by sequencing on a HiSeq 2500 using 2×100bp read-length. Prioritization of causal variants was performed with the eDiVA software (http://www.ediva.crg.es/), which performs read alignment using bwa-mem, multi-sample variant prediction using GATK and variant prioritization using a random forest model trained on ClinVar and an in-house disease variant knowledge database. Results: We identified 14 variants in 10 trios. Twelve were Sanger-validated; five of these led to genetic diagnosis, comprising heterozygous de novo variants in CACNA1A, ITPR1 and STXBP1 and compound heterozygous variants in PIEZO2, and GJC2. In retrospect, pupillary abnormalities, arthrogryposis, seizures and hypomyelination could have represented useful hints to diagnosis in four cases. We additionally found homozygous variants in the new candidate gene DNAH14, encoding an axonemal dynein heavy chain, and heterozygous variants in SLC39A8, encoding a divalent cation transporter, previously associated with intellectual disability and cerebellar atrophy. Conclusions: In our CA cohort, WES allowed for diagnosis of 5/14 cases and identified a putative novel locus plus a new phenotype related to a known gene. We suggest that WES should be considered early in the diagnostic algorithm of CA with negative or unspecific imaging findings.