Next-generation sequencing identifies unexpected genotype-phenotype correlations in patients with retinitis pigmentosa

Abstract

Retinitis pigmentosa (RP) is an inherited degenerative disease causing severe retinal dystrophy and visual impairment mainly with onset in infancy or adolescence. Targeted next-generation sequencing (NGS) has become an efficient tool to encounter the enormous genetic heterogeneity of diverse retinal dystrophies, including RP. To identify disease-causing mutations in unselected, consecutive RP patients, we conducted Sanger sequencing of genes commonly involved in the suspected genetic RP subtype, followed by targeted large-panel NGS if no mutation was identified, or NGS as primary analysis. A high (70%) detection rate of disease-causing mutations was achieved in a large cohort of 116 unrelated patients. About half (48%) of the solved RP cases were explained by mutations in four genes: RPGR, EYS, PRPF31 and USH2A. Overall, 110 different mutations distributed across 30 different genes were detected, and 46 of these mutations were novel. A molecular diagnosis was achieved in the majority (82-100%) of patients if the family history was suggestive for a particular mode of inheritance, but only in 60% in cases of sporadic RP. The diagnostic potential of extensive molecular analysis in a routine setting is also illustrated by the identification of unexpected genotype-phenotype correlations for RP patients with mutations in CRX, CEP290, RPGRIP1, MFSD8. Furthermore, we identified numerous mutations in autosomal dominant (PRPF31, PRPH2, CRX) and X-linked (RPGR) RP genes in patients with sporadic RP. Variants in RP2 and RPGR were also found in female RP patients with apparently sporadic or dominant disease. In summary, this study demonstrates that massively parallel sequencing of all known retinal dystrophy genes is a valuable diagnostic approach for RP patients.

Fig 1. Distribution of mutations in consecutive, unrelated RP patients.

Genes mutated in patients who were reversely phenotyped (based on the genetic findings) are indicated by asterisk. The number of patients with mutations in the different genes is displayed on the right. Black, autosomal recessive. Blue, autosomal dominant. Green, X-linked.

Fig 2. Right eye of a patient with RP due to an CRX mutation.

(A) widefield false-color image, fundus AF with (B) 488 nm and (C) 787 nm excitation light, and (D) spectral-domain optical coherence tomography. Only one eye is shown due to high symmetry between eyes.

Fig 3. Left fundus of a 70-year-old women with RP caused by compound-heterozygous CEP290 mutations.

(A) widefield fundus imaging, (B, C) near infrared reflectance imaging, (D) spectral-domain optical coherence tomography. Only one eye is shown due to high symmetry between eyes.

Fig 4. Right eye of a patient with RP and a homozygous RPGRIP1 mutation.

Fundus color image (A), fundus AF with (B) 488 nm and (C) 787 nm excitation light, (D, E) spectral-domain optical coherence tomography. Only one eye is shown due to high symmetry between eyes.

Fig 5. Left eye of a patient with RP and a homozygous MFSD8 mutation.

(A) widefield false-color image, fundus AF with (B) 488 nm and (C) 787 nm excitation light, (D) spectral-domain optical coherence tomography. Only one eye is shown because there was high similarity of both eyes.

Fig 6. Phenotype of an RP patient with an RP2 mutation.

A-D right eye, E-H left eye Right eye: (A, E) widefield fundus imaging and (B, F) widefield fundus autofluorescence, (C, D, G, H) spectral-domain optical coherence tomography.

Fig 7. Phenotype of an RP patient with an RPGR mutation.

A-E right eye, F-J left eye: Fundus color imaging (A, D), fundus autofluorescence with 488 nm excitation light (B, E) and (C, F) spectral-domain optical coherence tomography.

Fig 8. Phenotype of an RP patient with an RPGR mutation.

(A) widefield fundus AF, (B) widefield false-color image and (C, D) spectral-domain optical coherence tomography. Only one eye is shown due to high symmetry between eyes.