Dominant retinitis pigmentosa phenotype associated with a new mutation in the splicing factor PRPF31

Autosomal dominant retinitis pigmentosa results from mutations in 14 known proteins, and at least two further loci have been highlighted by genetic linkage in families (reviewed by the RetNet website; http://www.sph.uth.tmc.edu/Retnet/). The known genes include those encoding components of the phototransduction cascade, retinal transcription factors and retinal structural proteins.¹ The list also includes four ubiquitously expressed splicing factors: pre‐mRNA processing factor 8 (PRPF8),² PRPF31,³ PRPF3⁴ and PAP‐1, also known as RP9.⁵,⁶

Splicing is a complex process that involves the precise excision of introns from pre‐mRNA by a macromolecular structure called the spliceosome. Three of the splicing factors implicated in autosomal dominant retinitis pigmentosa (ADRP) are components of the U4/U6‐U5 tri‐snRNP particle, an essential component of the spliceosome.⁷,⁸ Mutations in one of these, PRPF31, have been reported to cause between 5 and 20% of ADRP.⁹,¹⁰ In this report, a new mutation in the PRPF31 gene is described, together with the clinical phenotype.

Cases

The proband was a 33‐year‐old female with a corrected visual acuity of 58 and 51 ETDRS letters in the right and left eye, respectively (approximate Snellen equivalents of 6/18 and 6/36). She had a myopic refraction with a spherical equivalence of −2 dioptres in each eye. Nyctalopia had been present since the middle of the second decade, and she had noticed a decrease in her central vision since the beginning of the third decade. At the most recent examination, she had early posterior subscapsular cataract, bone spicule formation in all four quadrants (fig 1a,b) attenuated arterioles and pale optic discs in each eye. The maculae appeared normal on clinical examination. On Goldman perimetry, the mean visual field to the V4e target measured 6.5° from fixation. Zeiss OCT 3 examination demonstrated a foveal thickness of 170 and 144 microns, respectively, in the right and left eyes, with absence of the third highly reflective band.¹¹

Figure 1 Electropherogram of the mutated (upper) and normal (lower) sequence of PRPF31 in. As the mutation is heterozygous, the upper image shows both the mutated and normal sequences superimposed. The arrow on the mutated sequence denotes the beginning of the deleted sequence, while the arrow on the normal sequence marks the boundary between exon 6 and intron 6. Sequence was generated from PCR‐amplified DNA on a Pharmacia MegaBACE automated DNA sequencer.

Her younger sister had a similar clinical phenotype and age of onset. The 61‐year‐old mother was asymptomatic, with unaided visual acuities of 80 and 81 ETDRS letters (Snellen equivalent of 6/7.5). Fundus examination revealed mild bone spicule attenuation in the peripheral retina (fig 1c,d). Visual field to the V4e target on Goldman perimetry was slightly reduced from normal with a mean of 57.8° from fixation. Foveal thickness on Zeiss OCT 3 examination was 223 and 249 microns in the right and left eyes, respectively. The father of the proband was also asymptomatic with visual acuities of 85 ETDRS letters (Snellen 6/6) in both eyes and normal ocular examinations. No clinical information was available from any other living relative along the maternal line.

Mutation screening

DNA from the proband was included in a large cohort of retinal dystrophy DNAs, which were screened for mutations in a limited set of exons or parts of exons of known retinal degeneration genes. The exons screened were selected from the available literature because they were known mutation hotspots or locations of common founder mutations. Screening was carried out by radioactively labelled single‐strand conformation polymorphism/heteroduplex analysis (SSCP/HA).¹²

One of the sequences screened was PRPF31 exon 6. Screening of this sequence in the proband revealed a large mobility shift suggestive of a deletion. Sequencing revealed a novel 16 bp deletion present in the three female members of the family but absent from the father and from 120 control Caucasian genomic DNAs (240 chromosomes). This sequence change is denoted c.522–527del&IVS6+1to+10del¹³ (fig 2). It deletes codons 175 and 176, the last two in exon 6, encoding glutamine and glycine residues. However, it also deletes the first 10 bp of intron 6, including the exon 6/intron 6 boundary and splice donor site, the mutation abolishing the exon 6 splice donor site. This may give rise to an mRNA transcript which includes intron 6, adding seven novel amino‐acids then terminating the encoded protein, or could lead to the skipping of exon 6.

Figure 2 A and B, retinal photographs taken from the proband showing bone spicule attenuation. C and D, retinal photographs from the mother of the proband. The mild bone spicule attenuation was evident only on slit lamp biomicroscopy using a wide angle lens.

Discussion

This novel mutation in the PRPF31 gene causes a severe phenotype in symptomatic cases, with the onset of nyctalopia in the second decade and loss of acuity from the third. Both the age of onset and the phenotype observed are similar to that described by Sato et al¹⁴ in Japanese families. In addition, this report is the first to demonstrate variable penetrance of the phenotype in an asymptomatic carrier of the mutation. A high level of non‐penetrance has been described previously, both in families with confirmed PRPF31 mutations and in those linked to the RP11 locus before mutations in PRPF31 were identified.¹⁴,¹⁵,¹⁶,¹⁷,¹⁸,¹⁹ Evans et al¹⁵ used the term bimodal expressivity to describe this phenomenon. Sato and colleagues also identified asymptomatic carriers of the mutations in the PRPF31 gene by genetic analysis. One of these was an elderly relative of three generations of symptomatic RP sufferers, though he himself had no ocular abnormalities except for mild cataracts. In our report, the mother of the proband had definite retinal findings and a mildly reduced visual field on Goldman perimetry, though she was totally asymptomatic. This may perhaps imply that the range of phenotypes seen in PRPF31‐RP could be better described as a spectrum of severity, rather than true bimodal expressivity.

The mutation described above is likely to result in a grossly abnormal transcript which may be subject to nonsense mediated decay.²⁰ This brings to 18 the number of published PRPF31 mutations in the literature, comprising six deletions (ranging from one base pair to the whole gene), five splice‐site mutations, two insertion/deletion events, one duplication, one insertion and only three missense mutations.³,¹⁴,¹⁸,¹⁹,²¹,²²,²³,²⁴ The lack of missense changes has led others to speculate that mutations in PRPF31 cause RP due to haploinsufficiency and consequent insufficiency of splicing activity.¹⁸ Wilkie et al²⁵ concluded that reduced mutant protein solubility in two of the known missense mutations, A194E and A216P, also led to splicing insufficiency.

This hypothesis is further supported by the finding that high‐expressing alleles of PRPF31 from the normal parent compensate for a potentially RP‐causing mutation on the opposing chromosome.²⁶ This phenomenon accounts for the variation in severity described above and predicts that the normal second allele of PRPF31 in the mother from the family described herein is a high‐expressing variant which masks the RP symptoms. However, the alleles inherited by her daughters from their normal father are less well expressed, and so these individuals have a much more severe form of RP. To date, the mechanism controlling this level of expression remains unknown. A bimodal phenotype might be explained by a single diallelic polymorphism in a sequence involved in transcription regulation, whereas a spectrum of severity, as observed herein, might imply a more complex interplay between several such polymorphisms. Understanding the basis of this variation in severity, together with the finding of haploinsufficiency as a cause of disease, could have important implications for the testing of potential new treatments for this relatively common retinal degeneration.

Acknowledgements

We thank Yorkshire Eye Research (grant numbers 009 and 006) and the Leeds Teaching Hospitals Charitable Foundation for funding this research, and Mike Stockton in the medical illustration department, St. James's University Hospital, and of course the patient and her family for enabling us to carry out this lesearch and share the information with the wider medical community.