New Insight into the Genotype-Phenotype Correlation of PRPH2-Related Diseases Based on a Large Chinese Cohort and Literature Review

Abstract

Variants in PRPH2 are a common cause of inherited retinal dystrophies with high genetic and phenotypic heterogeneity. In this study, variants in PRPH2 were selected from in-house exome sequencing data, and all reported PRPH2 variants were evaluated with the assistance of online prediction tools and the comparative validation of large datasets. All variants were classified based on the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) guidelines. Individuals with pathogenic or likely pathogenic variants of PRPH2 were confirmed by Sanger sequencing. Clinical characteristics were summarized. Ten pathogenic or likely pathogenic variants of PRPH2 were identified in 14 families. In our cohort, the most frequent variant was p.G305Afs*19, accounting for 33.3% (5/15) of alleles, in contrast to the literature, where p.R172G (11.6%, 119/1028) was the most common variant. Nine in-house families (63.8%) were diagnosed with retinitis pigmentosa (RP), distinct from the phenotypic spectrum in the literature, which shows that RP accounts for 27.9% (283/1013) and macular degeneration is more common (45.2%, 458/1013). Patients carrying missense variants predicted as damaging by all seven prediction tools and absent in the gnomAD database were more likely to develop RP compared to those carrying missense variants predicted as damaging with fewer tools or with more than one allele number in the gnomAD database (p = 0.001). The population-specific genetic and phenotypic spectra of PRPH2 were explored, and novel insight into the genotype-phenotype correlation of PRPH2 was proposed. These findings demonstrated the importance of assessing PRPH2 variants in distinct populations and the value of providing practical suggestions for the genetic interpretation of PRPH2 variants.

Keywords: PRPH2; genotype-phenotype correlation; macular degeneration; retinitis pigmentosa.

Figure 1

The genealogy of 14 families with pathogenic or likely pathogenic variants of PRPH2. Squares represent male individuals, and circles indicate female individuals. Affected patients are indicated by black shading, and the proband of each family is indicated by an arrow. Family numbers are listed on the top of genealogies, while variants are listed under the genealogies.

Figure 2

Fundus performance of in-house patients with PRPH2-associated retinopathy. (A–D) Pale optic discs, narrowing of retinal vessels, widespread tapetoretinal degeneration, macular atrophy in different degrees, and bone-spicule pigment deposits in the mid-peripheral area were observed among patients diagnosed with RP. Mottled hyper-fluorescein spots were observed in the whole posterior pole except for the macula area. (E) A small round atrophic lesion in the macula that presented as a hypoautofluorescent lesion in the autofluorescence image was observed in a patient with adult vitelliform macular dystrophy. (F) A patient presented with granular pattern retinal pigment epithelium mottling diffusely throughout the macular area that showed widespread mottled hyper-fluorescein in the late stage of fluorescein fundus angiography.

Figure 3

The genotypic and phenotypic spectrum of variants in PRPH2 in our cohort and reported in the literature. (A) Distribution and frequency of all detected PRPH2 variants in our cohort, the gnomAD database, pathogenic or likely pathogenic PRPH2 variants in our cohort, and reported PRPH2 variants in the literature. The blue, red, and gray lines represent missense, truncation, and in-frame variants, respectively. ID2 indicates the intradiscal D2 loop of the peripherin protein. (B) The three pie charts show the phenotypic spectrum of families with PRPH2 variants identified in our cohort, Asian families, and all reported non-Asian families with PRPH2-associated retinopathy in the literature, respectively.

Figure 4

The exploration of the correlation between genotype and phenotype of PRPH2-associated diseases in this study. (A) In our cohort, among the four patients carrying missense variants that were predicted by at least six out of the seven in silico tools and absent in the gnomAD database (Class 1), three were diagnosed with retinitis pigmentosa (A1). Among the remaining four patients carrying missense variants predicted as damaging by five tools and absent in the gnomAD database or predicted as damaging by six tools and containing one allele number in the gnomAD database, three were identified with macular degeneration (A2). Patients carrying missense variants predicted to be benign by more than three tools and with more than one allele number in the gnomAD database presented phenotypes unrelated to PRPH2 that were not co-segregated among family members and were not aggregated among individuals with inherited retinal dystrophy compared to controls. (B) In total, 49.1% of the 254 families carrying Class 1 missense variants were diagnosed with retinitis pigmentosa (B1), while 69.6% of the 102 families with Class 2 missense variants were diagnosed with macular degeneration (B2), which was statistically significant (p = 0.001). Among all of the 149 reported variants, 90.5% of the Class 1 and Class 2 missense variants were located in the ID2 loop domain of proteins, and only 54.5% of the remaining variants were found to aggregate in the domain (B3). *** represented p ≤ 0.001.