RP9 revisited; RP9 p.(H137L) remains a likely cause of dominant splicing factor-Retinitis Pigmentosa

Abstract

Variants in six pre-mRNA processing factors cause autosomal dominant Retinitis Pigmentosa (adRP). The RP9 gene encodes a seventh splicing factor, and in 2002, we published RP9 variants c.410A>T; p.(H137L) and c.509A>G; p.(D170G) as likely causes of adRP in a large multigenerational RP9-linked family and a single case, respectively. It has since been suggested these variants might be artefacts due to simultaneous amplification of the RP9P pseudogene, and no further pathogenic variants have been reported. We therefore rescreened two members of the RP9-linked family by genome sequencing. Examination of the 2 Mb locus defined by crossovers in the original family revealed no other plausible causative variants. Alignment of both short and long-read sequences confirmed that p.(H137L) is in the RP9 gene, not the pseudogene. Screening for p.(H137L) in 1961 RP/Rod-cone dystrophy (RCD) cases from the Leeds patient cohort and UK 100,000 Genomes Project (100kGP) database revealed four further carriers. Including the original family, this variant was therefore present in 5/1962 RP/RCD probands, and is absent from gnomAD, constituting statistically significant enrichment in RP cases. Long-read sequencing of p.(H137L) in available carriers showed this is a UK founder allele. The RP9 p.(D170G) allele was also confirmed as gene, not pseudogene, derived, but is present in 22 individuals in the 100kGP cohort, none with RP, as well as >200 individuals in gnomAD and Biobank, suggesting it is non-pathogenic. In conclusion, RP9 p.(H137L) is strongly associated with RP and remains the only plausible variant accounting for the condition in a large multi-generation adRP family.

Fig. 1. Clinical images from a member of the original RP9-linkedfamily.

Ultrawide field pseudocolour fundus images, autofluorescence and OCT imaging in an affected individual in the mid-twenties, showing typical RP with retained retinal structure at the central macula, delineated by hyper-autofluorescent rings of Robson and a retained ellipsoid line (OCT). Outer-retinal degeneration has occurred anteriorly. RP was diagnosed in second decade, and the patient is night blind and unable to drive due to field loss, but has retained central vision. This is similar to RP seen in other forms of splicing factor RP.

Fig. 2. Comparative alignments between RP9 and RP9P for a region centred upon the c.410A>T; p.(H137L) variant.

A DNA sequence comparative alignment between RP9 and RP9P for a region centred upon the c.410A>T; p.(H137L) variant, which is highlighted in cyan. Non-matching nucleotides are shaded. B IGV views of short read genome sequence from individuals RP9-289 and 296 at the RP9 and RP9P regions, showing the T>A (forward strand, with the gene encoded on the reverse strand) change is present in RP9 and not in RP9P.

Fig. 3. Long-read sequencing amplification products visualised using IGV.

Haplotypes for four samples were created following the selection of reads that correspond to the T or A nucleotide at genomic position chr7:33096550. Variants defined by haplotype chr7: chr7:33096550 A are in linkage disequilibrium with the mutant allele. The upper two panels show the location of reads for each haplotype within a ~500 kb interval of the RP9 locus, with a representation of the genes in the region shown between the two haplotype views; these data confirm that all reads map to RP9 and not the pseudogene (RP9P). The lower panel magnifies the sequenced region; the upper of these show the conserved haplotype associated with the variant chr7:g.33096550A nucleotide, which is shared by all four cases analysed. The lower panel shows the varying haplotypes on the other, non-disease-associated allele. The genomic architecture of RP9 is displayed between the two haplotype panels.

Fig. 4. AlphaFold (upper panel) and AlphaMissense (lower panel) analysis of the RP9 protein.

The sites of the c.410A>T; p.(H137L) and c.509A>G; p.(D170G) variants are denoted by arrows.