Heterozygous truncating variants in SUFU cause congenital ocular motor apraxia

Abstract

Purpose: This study aimed to delineate the genetic basis of congenital ocular motor apraxia (COMA) in patients not otherwise classifiable.

Methods: We compiled clinical and neuroimaging data of individuals from six unrelated families with distinct clinical features of COMA who do not share common diagnostic characteristics of Joubert syndrome or other known genetic conditions associated with COMA. We used exome sequencing to identify pathogenic variants and functional studies in patient-derived fibroblasts.

Results: In 15 individuals, we detected familial as well as de novo heterozygous truncating causative variants in the Suppressor of Fused (SUFU) gene, a negative regulator of the Hedgehog (HH) signaling pathway. Functional studies showed no differences in cilia occurrence, morphology, or localization of ciliary proteins, such as smoothened. However, analysis of expression of HH signaling target genes detected a significant increase in the general signaling activity in COMA patient-derived fibroblasts compared with control cells. We observed higher basal HH signaling activity resulting in increased basal expression levels of GLI1, GLI2, GLI3, and Patched1. Neuroimaging revealed subtle cerebellar changes, but no full-blown molar tooth sign.

Conclusion: Taken together, our data imply that the clinical phenotype associated with heterozygous truncating germline variants in SUFU is a forme fruste of Joubert syndrome.

Keywords: COMA; Joubert syndrome; SUFU; congenital ocular motor apraxia; sonic hedgehog.

Fig. 1. Characteristic magnetic resonance image (MRI) features in three subjects with heterozygous SUFU variants.

Four representative T1-weighted MRIs (arranged in horizontal rows) are shown from three individuals with SUFU variants and one adult healthy control (m–p). Panels (a–d) are from subject I.1, family 3, at age 1.5 years; (e–h) from subject II.1, family 2, at age 7.5 years; and (i–l) from subject II.1, family 5, at age 40 years. The first vertical row (a,e,i,m) shows axial views at the level of the upper vermis, indicating folial dysplasia (arrow). The second vertical row (b,f,j,n) illustrates axial views at the level of the superior cerebellar peduncles (arrows), these are more prominent (longer, thicker) compared with normal. The third vertical row (c,g,k,o) shows parasagittal sections demonstrating that the superior cerebellar peduncles (arrows) are thicker and have a more horizontal course compared with normal (o). The fourth vertical row (d,h,l,p) illustrates coronal images revealing irregular folia and vermis splitting (arrows) in the individuals with SUFU variants.

Fig. 2. Pedigrees and genetic characterization of six families with congenital ocular motor apraxia carrying heterozygous loss-of-function variants in SUFU.

(a) Pedigrees of families 1–6 showing segregation of rare deleterious SUFU variants. Unfilled shapes denote healthy, filled shapes affected individuals. (b) Chromatograms of the identified SUFU variants in family 1 (F1: c.83C>A; p.Ser28*), family 2 (F2: c.1099G>T; p.Glu367*), family 3 (F3: c.479delA; p.His160Leufs*20), family 4 (F4: c.1220_1221insT; p.Phe408Valfs*13), family 5 (F5: c.309_310delAG; p.Arg103Serfs*3), and family 6 (F6: c.1333dupG; p.Glu445Glyfs*22) compared with wild-type (WT) sequences of the respective positions. Localization of frameshift or nonsense variants is indicated in red.

Fig. 3. Overview of the identified heterozygous loss-of-function variants in SUFU on genomic and protein level.

Schematic diagram of the SUFU gene (top panel) and protein (bottom panel) showing the localization of six truncating variants identified within this study in affected individuals from six independent families (black). Localization of SUFU variants associated in previous studies with Joubert syndrome are indicated in green and blue; variants associated with Gorlin–Goltz syndrome are marked in red.^,,,.

Fig. 4. Cilia formation and expression of Hedgehog signaling signature genes in congenital ocular motor apraxia (COMA) patient–derived dermal fibroblasts.

(a) Representative pictures of double immunofluorescent staining for visualization of SMO (upper row in monochrome, third and lower row in red) and acetylated tubulin (acetyl.tubulin) (second row in monochrome, third and lower row in green) in control fibroblasts and fibroblasts derived from affected individual II.2 (family 3, first column), individual II.1 (family 2, middle column), and individual II.1 (family 4, right column). Nuclei were visualized by DAPI staining (lower row in blue). Scale bars: 1 µm. (b) Quantitative real-time polymerase chain reaction (PCR)–based expression analyses of the Hedgehog signaling signature genes GLI1, GLI2, GLI3, and PTCH1 normalized to 18S ribosomal RNA (rRNA) (left column) or HPRT (right column) expression levels, respectively, of controls (N = 5) and COMA patient–derived fibroblasts (COMA) (N = 4). Shown results represent data of two different cellular passages per fibroblast culture each analyzed in biological triplicates (gray circles) that were measured in technical triplicates. Total mean values +/- SEM of all analyzed samples are indicated in black. Significant differences were tested by nonparametric Mann–Whitney tests. *p < 0.05; **p < 0.01; ***p < 0.001. No significantly different expression levels were observed in the five independent control fibroblast cultures, different cellular passages, or biological triplicates. For comparison of individual gene expression levels of the four independent COMA patient–derived fibroblast cultures as well as HIP expression levels see Supplemental Data (Fig. S1).