Autosomal Recessive Rod-Cone Dystrophy Associated With Compound Heterozygous Variants in ARL3 Gene

Abstract

Purpose: ARL3 (ADP-ribosylation factor-like 3) variants cause autosomal dominant retinitis pigmentosa (RP) or autosomal recessive Joubert syndrome. We found a family with rod-cone dystrophy (RCD) and verified it was associated with compound heterozygous variants in ARL3 gene. Methods: Ophthalmic examinations including optical coherence tomography and electroretinogram (ERG) were performed. Targeted next generation sequencing (NGS) was performed for the proband using a custom designed panel. Sanger sequencing and co-segregation were conducted in the family members. Changes of protein structure mediated by the variants were studied in vitro. ARL3 protein stability and its interaction with RP2 protein were assessed by cycloheximide chase assay and co-immunoprecipitation (Co-IP) assay. Results: Visual acuity of the 18-year-old male proband was 0.25 in the right and 0.20 in the left eye, while his non-consanguineous parents and sister was normal. The proband showed signs of RCD, including nyctalopia, peripheral field loss, bone-spicule deposits in the retina, and reduced ERG responses. The father, aged 50 years old, showed visual acuity of 1.0 in both eyes. Unlike the proband, he presented late onset and mild cone-rod dystrophy (CRD), including macular atrophy, central scotomata, moderate reduction in photopic ERG responses. None of all the family members had hearing abnormality, mental dysplasia or gait instability. We identified two novel compound heterozygous variants (c.91A>G, p.T31A; c.353G>T, p.C118F) in ARL3 in the proband, while his father only had variant c.91A>G. Bioinformatics analysis indicated amino acid positions of the two variants are highly conserved among species. The in silico tools predicted the variants to be harmful. Protein structure analysis showed the two variants had potential to alter the protein structure. Based on the ACMG guidelines, the two variants were likely pathogenic. In addition, the ARL3 mutations destabilized ARL3 protein, and the mutation c.353G>T disrupted the interaction between ARL3 and RP2 in HEK293T cells. Conclusions: We showed novel compound heterozygous variants in ARL3 were associated with early onset of autosomal recessive RCD, while c.91A>G along may be associated with a late onset of dominant CRD. The two variants in ARL3 could be causative by destabilizing ARL3 protein and impairing its interaction with RP2 protein.

Keywords: ARL3; RP2; compound heterozygous variants; cone-rod dystrophy; rod-cone dystrophy.

Figure 1

Pedigree and Clinical features of the proband with RCD. (A) Pedigree of the family with autosomal recessive RCD. (B) Visual field showed tunnel vision. (C) Fundus photographs showed peripheral bone spicule pigmentation. (D) SS-OCT showed atrophy of the retinal outer layers of the binocular macular area. (E) ERG showed severe reduced rod responses and to less extent cone responses.

Figure 2

Clinical features of the father with CRD. (A) Visual field showed central scotoma. (B) Fundus photographs showed circular degeneration around macular fovea. (C) SS-OCT revealed the binocular macular area outer retinal was thinned by atrophy. (D) ERG showed a moderate decrease in the binocular cone and rod systems.

Figure 3

Two novel ARL3 variants (c.91A>G, p.T31A; c.353G>T, p.C118F) were identified in the RCD family. (A) Sanger sequencing in all family numbers. (B) Multiple alignments of Thr31 and Cys118 of ARL3 protein from different species. Both variants occurred on the conserved residues of the ARL3 protein.

Figure 4

Tertiary structure prediction of mutant proteins. (A) The schematic structures of the original and the mutant amino acids. The backbone was colored in red; the side chain was colored in black. (B) The 3D-structure of the wild-type and mutant proteins. The main structure of the ARL3 protein was colored in yellow. The side chains of the wild-type and the mutant residues were colored in green and red, respectively. (C) Superimposition of the structures of ARL3 (yellow) in complex with its interactors: RP2 (blue; PDB: 3BH6) (Veltel et al., 2008), ARL13B (gray; PDB: 5DI3) (Gotthardt et al., 2015), and UNC119A (orange; PDB: 4GOJ) (Ismail et al., 2012). On the right side is a zoomed-in view of some structures, including GNP (pink), Mg²⁺ (green), and the side chains of Thr31 and Cys118 (red).

Figure 5

The ARL3 variants T31A and C118F impaired the stabilities of the encoded proteins in HEK293T cells. HEK293T cells were transfected with empty vector, wild-type, ARL3 T31A and C118F vectors for 24 h and then treated with 100 μg/ml cycloheximide (CHX) for 1, 3, and 6 h, respectively. ARL3 protein levels were detected using the flag antibody. The protein levels of β-actin were used as endogenous control (n = 1).

Figure 6

ARL3 variants T31A and C118F disrupted the combination between ARL3 and RP2. Plasmids of HA-RP2-WT were co-transfected with wild-type or mutant Flag-ARL3 plasmids into HEK293T cells. ARL3 was immunoprecipitated with anti-Flag antibody. Western blotting was performed to detect the specific proteins indicated on the left side of each panel (Error bars indicate means ± SD; n = 3, ***p < 0.001).

Figure 7

All the reported ARL3 pathogenic mutations were labeled in the schematic diagram of ARL3 gene and protein. The ARL3 mutations c.91A>G (p.T31A) and c.353G>T (p.C118F) were indicated in red.