A novel homozygous missense substitution p.Thr313Ile in the PDE6B gene underlies autosomal recessive retinitis pigmentosa in a consanguineous Pakistani family

This study was approved by the bioethical committee of Hazara University, Mansehra, Pakistan (Approval No. F.NO:185/HU/Zool/2021/182), and was conducted in accordance with the standards of the Declaration of Helsinki and according to the ARVO statement on the use of human subjects in medical research. Written informed consent was obtained from the parents of the patients. Clinical and demographic information was recorded on a pre-designed questionnaire through face-to-face interviews with the patients and the pedigree of the family was drawn electronically using the pedigree chart designer tool (CeGaT, Tubingen, Germany). Patients were clinically examined by an ophthalmologist at the Layton Rehmat Ullah Benevolent Trust (LRBT) Hospital, Mansehra, Pakistan. Ophthalmic examination included measurement of best-corrected visual acuity (BCVA) using a Snellen chart, slit-lamp biomicroscopy, and ophthalmoscopy, fundus examination after pupillary dilation, fundus photography, spectral domain optical coherence tomography (SD-OCT).

Following collection of saliva samples from patients as well as available healthy members of the family in a saliva self-collection kit (DNA Genotek, Ottawa, Canada), genomic DNA was extracted from saliva samples using a standardized protocol as mentioned in the prepIT-L2P manual (DNA Genotek, Ottawa, ON, Canada). Next quantitative and qualitative assessment of the proband’s DNA was made, and ~ 2.0 μg of genomic DNA was sent to Novogene Co. Ltd (Hong Kong, China) for whole-exome sequencing (WES) analysis. Protocols and platforms used for sequencing have been previously published [20, 21]. Exome data were analyzed using the in-house computational pipeline. Details of filtering steps and variant prioritization strategy were described previously [21]. All short-listed variants were classified based on their molecular profile (nonsense, frameshift, missense, and splice sites). The nature of pathogenicity of all missense variants was checked using numerous online tools such as, but not limited to, Sorting Intolerant from Tolerant (SIFT), Polyphen-2, Mutation Taster, Provean and CADD, etc. Genome-wide autozygosity mapping was performed using AutoMap [22]. PCR-based amplification of the target DNA sequence carrying the potentially pathogenic variant(s) was done using standard PCR conditions/cycles, and by employing a pair of target-specific primers (Forward primer: 5’-TACCAAGGGCAGCACTCAA-3’ and Reverse primer: 5’-CACAGTGCTGGAGTACGGG-3’). PCR products were Sanger sequenced bi-directionally using a commercial facility. Lastly, causality of the potentially pathogenic variant(s) was confirmed by performing a strict genotype–phenotype co-segregation study within the available members of the family.

To examine the evolutionary conservation of the altered (p.Thr313) and adjacent residues, PDE6B protein sequences of different vertebrate species including Human (H. Sapiens, NP_001138764), Chimpanzee (P. troglodytes XP-024211748.1), Cattle (B. taurus, NP-776843.1), Rat (R. norvegicus, NP-001099494.1), Rhesus monkey (M. mulatta, XP-028704029.1), Chicken (G. Gallus, NP-001305369.1), Cat (F. cactus, XP-019685437.1), Horse (E. caballus, XP-023494350.1), Dog (C. lupus, NP-001002934.2), Zebrafish (D. rerio XP-685002.1), Woodpecker (D. Pubescens, XP-009907072.1), and Mouse (M. Musculus, NP-032832.2) were downloaded in fasta format from UniProtKB/Swiss-prot database (http://​www.​uniprot.​org/​). Fasta sequences were aligned using clustal omega multiple sequence alignment package (https://​www.​ebi.​ac.​uk/​Tools/​msa/​clustalo/​).

Swiss Model was used to create homology models for wild (PDE6B and PDE6A) and mutant (p.Thr313Ile_PDE6B) proteins and energy was minimized by using UCSF Chimera 1.1 [23] with the default setting, and the models were evaluated using verify 3D [24], ERRAT [24, 25] and ProCheck [26]. The Ligand cGMP was retrieved from PubChem (135398570) [27]. The ligand was converted to 3D and was energy minimized using Chem 3D version 19.0.0.22 (PerkinElmer, Waltham, MA, USA). Active site residues were predicted using a 3D ligand Site [28]. MOE [29] was used for docking of wild-type and mutant PDE6B protein with cGMP ligand and PDE6A. Interaction between wild and mutated PDE6B with cGMP and PDE6A were analyzed by PyMol [The PyMOL Molecular Graphics System, Version 2.4 Schrödinger, LLC], while wild-type and mutated PDE6B were docked with PDE6A using CLUSPRO 2.0 [30].

The pedigree of the family is shown in Fig. 1. The proband (II.6) is currently a 32-years old female living in Abbottabad district in northern Pakistan. She is the child of unaffected parents (I.1 and I.2) who married consanguineously (parents are first-cousins). The proband has five siblings, all alive, including three brothers and two sisters. Of all the siblings, only one brother (II.1) of the proband has RP while the remaining siblings were all healthy. The proband was suffering from night blindness at the age of 6 years and was diagnosed with RP by an ophthalmologist. She was aware of dark adaptation problems since the age of approximately 5 years. During her last examination in 2020, at age 31 years, the patient’s best-corrected visual acuity was greatly damaged, as both eyes were sensitive only to light perception. As shown in Fig. 2A, the fundus photograph of the proband showed arteriolar attenuation, waxy disc pallor, and mid-periphery bony spicules. There were also associated maculopathy with altered foveal reflex. Red free image with a superimposed false color-coded macular thickness map showed reduced macular thickness, indicating foveal atrophy. There were, however, no evidence of posterior vitreous detachment or abnormal vitreous attachment. The inner retinal layers (Retinal Nerve-fibre layer and Ganglion cell layer) appear grossly intact. There seems no intra retinal hypo/hyper reflective areas in the middle layers of retina namely Inner Plexiform layer, Inner nuclear layer, Outer Plexiform Layer and Outer Nuclear layer. The underline Retinal pigment epithelium shows degenerative changes. The corresponding map of age-matched normal is showing diffuse thinning in the foveal region. There is reduced average thickness, reduced central thickness and reduced total volume of 190 um, 158um and 5.37mm3 respectively (Fig. 2B).

Whole-exome sequencing in proband identified a homozygous missense change (NM_001145292.2:c.938C > T; p.Thr313Ile) in the PDE6B gene (NM_001145292.2). Interestingly, genome-wide homozygosity mapping of WES data localized the PDE6B gene inside an autozygous interval spanning ~ 2.36 Mb; however, total genomic autozygosity in the proband was found to be 293.14 Mb (Fig. 3). Collectively, these data corroborate pedigree-based consanguinity in the studied family.

Upon familial co-segregation study using DNA of the available members of the family, c.938C > T variation showed strict genotype–phenotype correlation. As shown in chromatograms in Fig. 4a, the mutation (c.938C > T) was present in a homozygous state in both patients (II.1, II.6) while proband’s father (I.1) and a healthy brother (II.3) were both heterozygote carrier for the same sequence alteration. This unique sequence variation was not detected in our internal control cohort which contains whole-exome and genome sequencing data from more than 500 unrelated individuals. However, c.938C > T alteration was detected heterozygously only once in the genome aggregation database (gnomAD; http://​gnomad.​broadinstitute.​org/​), which contains sequencing data from over 141,000 unrelated individuals. In-silico analysis of the variant using existing online tools classified the variant as deleterious. Similarly, multiple sequence alignment of the altered residue revealed a high degree of evolutionary conservation across species (Fig. 4b).

Docking scores for wild and mutated protein with cGMP were -16.9951 and -11.0597, which shows destabilization upon mutation (Fig. 5A). Docking results showed that the wild PDE6B interacts with cGMP through His279, Asp319, and Asp439 residues, while the mutant type PDE6B interacts with cGMP through His282, His318, Asp319, Asp326, Leu328 (Fig. 5A). Furthermore, wild-type and mutant PDE6B interacted with its heterodimer PDE6A via distinct hydrogen bonds and residues (Fig. 5B). Native and aberrant interactions of wild-type and mutated PDE6B with PDE6A are shown in supplementary data (Table S1).