Targeted resequencing identifies genes with recurrent variation in cerebral palsy

Abstract

A growing body of evidence points to a considerable and heterogeneous genetic aetiology of cerebral palsy (CP). To identify recurrently variant CP genes, we designed a custom gene panel of 112 candidate genes. We tested 366 clinically unselected singleton cases with CP, including 271 cases not previously examined using next-generation sequencing technologies. Overall, 5.2% of the naïve cases (14/271) harboured a genetic variant of clinical significance in a known disease gene, with a further 4.8% of individuals (13/271) having a variant in a candidate gene classified as intolerant to variation. In the aggregate cohort of individuals from this study and our previous genomic investigations, six recurrently hit genes contributed at least 4% of disease burden to CP: COL4A1, TUBA1A, AGAP1, L1CAM, MAOB and KIF1A. Significance of Rare VAriants (SORVA) burden analysis identified four genes with a genome-wide significant burden of variants, AGAP1, ERLIN1, ZDHHC9 and PROC, of which we functionally assessed AGAP1 using a zebrafish model. Our investigations reinforce that CP is a heterogeneous neurodevelopmental disorder with known as well as novel genetic determinants.

Keywords: Genetic testing; Molecular medicine; Neurodevelopmental disorders; Paediatric neurological disorders.

Fig. 1

Scatterplots of CADD Phred vs PolyPhen2 scores for variants in genes with an overrepresentation of pathogenic variants in cerebral palsy cases compared to 1000 genomes controls. Controls—filled dots, CP cases—crossed dots

Fig. 2

AGAP1 morphant zebrafish show gross developmental defects, neurological deficits and reduced motility. a Representative images of developmental phenotypes observed at 24 hpf and 72 hpf in AGAP1 morphant zebrafish larvae. At 24 hpf, we frequently observed gross developmental delay, including reduced pigment, with some larvae exhibiting necrosis in the head. At 72 hpf, AGAP1 morphant larvae frequently showed milder developmental delay and a curved tail. b, c Phenotype frequencies observed in morphant larvae at 24 and 48 hpf, respectively. AGAP1 morpholino was injected in a concentration range to assess dosage dependence of the phenotypes in morphant larvae. d–g AGAP1 morphant larvae show decreased activity and reduced escape response to stimuli at 96 hpf, which is partially rescued by co-injection of human AGAP1 mRNA. CTRL^MO + CTRL^mRNA; n = 48, CTRL^MO + AGAP1^MO; n = 42, AGAP1^MO + CTRL^mRNA; n = 40, AGAP1^MO + AGAP1^mRNA; n = 24. d AGAP1 morphant larvae (dark grey) show reduced activity compared to controls during 1 min in the dark (p = 0.0002). Larvae co-injected with human AGAP1 mRNA and AGAP1 morpholino (light grey, rescue) show significantly greater activity than AGAP1 morphants (p < 0.0001). There is no significant difference in activity between AGAP1 morphants co-injected with human AGAP1 mRNA (light grey, rescue) and either control. Complete statistics can be found in Supplementary Table 9. e AGAP1 morphant larvae display both a reduced escape response and an altered trajectory in the 250 ms following tap stimulation compared to larvae co-injected with Control morpholino and either Control mRNA or human AGAP1 mRNA. Larvae co-injected with AGAP1 morpholino and human AGAP1 mRNA show an increased reaction compared to both controls. Complete statistics can be found in Supplementary Table 10. ***p < 0.001; ns, not significant. f Activity of larvae measured over a time course in darkness and then following the light in the apparatus being switched on. Error bars are s.e.m. g Response to light measured over 250 ms following the light stimulus. All groups except AGAP1 morphant larvae display a characteristic spike in activity immediately following the light stimulus. AGAP1 morphant larvae (black bars) show no significant increase in activity in response to light (p = 0.426). Error bars are s.e.m. *p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant