Exome sequences versus sequential gene testing in the UK highly specialised Service for Limb Girdle Muscular Dystrophy

Abstract

Background: Limb girdle muscular dystrophies are a group of rare and genetically heterogeneous diseases that share proximal weakness as a common feature; however they are often lacking very specific phenotypic features to allow an accurate differential diagnosis based on the clinical signs only, limiting the diagnostic rate using phenotype driven genetic testing. Next generation sequencing provides an opportunity to obtain molecular diagnoses for undiagnosed patients, as well as identifying novel genetic causes of muscle diseases. We performed whole exome sequencing (WES) on 104 affected individuals from 75 families in who standard gene by gene testing had not yielded a diagnosis. For comparison we also evaluated the diagnostic rate using sequential gene by gene testing for 91 affected individuals from 84 families over a 2 year period.

Results: Patients selected for WES had undergone more extensive prior testing than those undergoing standard genetic testing and on average had had 8 genes screened already. In this extensively investigated cohort WES identified the genetic diagnosis in 28 families (28/75, 37%), including the identification of the novel gene ZAK and two unpublished genes. WES of a single affected individual with sporadic disease yielded a diagnosis in 13/38 (34%) of cases. In comparison, conventional gene by gene testing provided a genetic diagnosis in 28/84 (33%) families. Titinopathies and collagen VI related dystrophy were the most frequent diagnoses made by WES. Reasons why mutations in known genes were not identified previously included atypical phenotypes, reassignment of pathogenicity of variants, and in one individual mosaicism for a COL6A1 mutation which was undetected by prior direct sequencing.

Conclusion: WES was able to overcome many limitations of standard testing and achieved a higher rate of diagnosis than standard testing even in this cohort of extensively investigated patients. Earlier application of WES is therefore likely to yield an even higher diagnostic rate. We obtained a high diagnosis rate in simplex cases and therefore such individuals should be included in exome or genome sequencing projects. Disease due to somatic mosaicism may be increasingly recognised due to the increased sensitivity of next generation sequencing techniques to detect low level mosaicism.

Keywords: Collagen VI related dystrophy; Exome; Limb girdle muscular dystrophy; Mosaicism; Myopathy; Titinopathy.

Fig. 1

Clinical characteristics and prior testing in the cohorts selected for WES and those undergoing standard testing. a - The number of individual genes screened by any method per family in WES cohort or standard testing cohorts. b - Comparison of inheritance pattern, serum CK and age of disease onset in cohort selected for WES or undergoing standard testing. Numbers on columns indicate the percentage of patients or, for inheritance pattern, families with a characteristic

Fig. 2

Diagnoses made by standard testing (a) and by WES (b). COL6-RD – Collagen VI related dystrophy, LAMA2 MD –LAMA2 related muscular dystrophy, TAM – Tubular Aggregate Myopathy, IBMPFTD – Inclusion Body Myopathy with Pagets disease and Frontotemporal Dementia

Fig. 3

Sequencing and muscle MRI in Patient WES9. a – variants in collagen VI genes identified by Sanger sequencing and WES in this patient. b – Pedigree. c – Visualisation of exome sequencing reads in patient DNA from blood with arrow indicating chr21:47,410,198 G > T (hg19) corresponding to COL6A1 c.957G > T; d – Sequencing chromatograms in patient DNA and parental DNA with COL6A1 c.957G > T variant indicated by arrows, height of sequencing peak in patient with DNA extracted from both blood and fibroblasts is lower than expected suggestive of mosaicism. Fibroblasts were obtained from skin biopsy taken from the right arm, which clinically is less severely affected than the left. e-i – T1 weighted axial MRI images of pelvis and lower limbs. In the pelvis (E) there is diffuse involvement of the gluteus maximus and medium with the left side more severely affected. In the thighs (F&G) there is sparing of gracilis and vastus medialis on the right, and diffuse involvement of all muscles on the left. There is central sparing of the vastus lateralis in the thighs and a ‘central shadow’ of increased signal intensity in the rectus femoris on the left (F&G) and less marked on the right, both of which are typical of COL6-RD [56]. In the calves there is peripheral involvement of the gastrocnemius and soleus on the left, with less pronounced peripheral involvement of the distal gastrocnemius on the right