Novel Rhodopsin Mutations Gly114Val and Gln184Pro in Dominant Retinitis Pigmentosa

Abstract

purpose. To identify mutations in the rhodopsin gene in North American patients with autosomal dominant retinitis pigmentosa (ADRP) and to measure the proportion of cases with rhodopsin mutations.

methods. Single-strand conformation polymorphism (SSCP) analysis and direct genomic sequencing were used to evaluate the coding region and intron splice sites of the rhodopsin gene for mutations in 91 unrelated patients.

results. Nineteen patients heterozygously carried a missense change in the rhodopsin gene (six with Pro23His, two with Pro347Leu, and one each with Thr17Met, Phe45Leu, Gly51Arg, Gly89Asp, Gly114Val, Arg135Trp, Pro171Leu, Gln184Pro, Phe220Leu, Ser297Arg, and Pro347Thr). All these missense changes were previously reported as causes for ADRP except for Gly114Val, Gln184Pro, and Phe220Leu, which were evaluated further by examining the relatives of index patients. The Gly114Val and Gln184Pro alleles cosegregated with ADRP as expected if they were pathogenic. Phe220Leu did not, indicating that it is not a cause of ADRP.

conclusions. Summation of the results of cases in this study with those of 272 unrelated cases of ADRP previously evaluated by our group shows that 90 of 363 (25%) of cases were caused by rhodopsin mutations.

Eleven loci causing autosomal dominant retinitis pigmentosa (ADRP) have been detected by linkage studies, and five of these have been identified: rhodopsin (RHO), RDS, CRX, NRL, and RP1 (www.sph.uth.tmc.edu/RetNet/disease.htm). Although the proportions of cases caused by each gene are not known precisely, the rhodopsin gene is thought to account for the largest number of cases. More than 80 mutations of the rhodopsin gene (RHO, Mendelian Inheritance in Man 180380) have been reported (www.uwcm.ac.uk/uwcm/mg/search/120347.html). Most rhodopsin mutations cause ADRP, whereas a few cause recessive RP ^{1 2} or a dominant form of stationary night blindness. ^{3 4 5}  

We evaluated 91 new index cases with ADRP for mutations in the rhodopsin gene to identify novel mutations and to better estimate the prevalence of rhodopsin mutations in North America. 

Methods

We determined the genetic type of patients through family history and ophthalmic examination including electroretinography. Patients with dominant disease typically had an affected parent or child. Normal control individuals had no symptoms of or blood relatives with retinal degeneration. After obtaining informed consent, we collected 10 to 50 ml of venous blood from each participant and purified leukocyte DNA. We examined and obtained blood samples from relatives of some index patients. The research adhered to the tenets of the Declaration of Helsinki. 

The coding region and the intron splice sites of the rhodopsin gene were screened by single-strand conformation polymorphism (SSCP) analysis in all index cases and normal controls. ⁶ DNA samples with variant SSCP fragments were evaluated by directly sequencing the relevant amplicons. ⁶  

Results

Among 91 unrelated patients with ADRP, we found 19 who were heterozygous for a missense change in the rhodopsin gene (Table 1) . Pro23His was the most frequent mutation (six patients), and Pro347Leu was the next most frequent (two patients). The remaining changes were found in one patient each. Three of the mutations were novel: Gly114Val, Gln184Pro, and Phe220Leu (Fig. 1) , and each of these was absent among control individuals (94 control subjects were evaluated for Gly114Val and 87 for Gln184Pro and Phe220Leu). The missense changes Gly114Val and Gln184Pro cosegregated with ADRP (Fig. 2) . The Phe220Leu allele did not; two of the index patient’s affected relatives did not carry this change (Fig. 2) . 

Discussion

The two novel pathogenic mutations identified in this study, Gly114Val and Gln184Pro, affect amino acid residues that are fairly well conserved among vertebrate opsins. Of 108 reported vertebrate opsin sequences (www.gpcr.org/7tm), ⁷ 98 have a Gly at position 114, whereas 10 have an Ala residue. Gly and Ala residues have small side groups (R⋕H for Gly and R⋕CH₃ for Ala), whereas Val has a larger side group: R⋕CH(CH₃)₂. Gly114 normally occupies the third transmembrane domain. It is the target of a previously described missense mutation that causes ADRP, Gly114Asp. ⁸ Gln184 is near Cys187, which participates in a disulfide bond linkage between the second and third intradiscal loops. Gln is at this position in 86 of 108 sequenced vertebrate rhodopsin homologues, Lys in 19 of them, and His in 1. The fairly high conservation of Gly114 and Gln184, together with the cosegregation of the Gly114Val and Gln184Pro alleles with ADRP in the respective families, is evidence supporting the pathogenicity of these mutations. 

In contrast, the missense variant Phe220Leu affects a residue in the fifth transmembrane domain that is not well conserved. Among 108 vertebrate opsins, ⁷ only 60 have Phe at this position. Two of the opsin homologues have a Leu at position 220, whereas 16 have Ile, 9 have Cys, 7 have Val, 4 have Thr, 3 have Gly, and 2 have Met. Based on these observations and the fact that the Phe220Leu allele did not cosegregate with ADRP, we interpret the Phe220Leu allele as a nonpathogenic variant. To our knowledge, Phe220Leu is only the second known nonpathogenic missense change in the human rhodopsin gene, Val104Ile being the first. ⁹  

A separate mutation affecting residue 220, Phe220Cys, has been previously reported to segregate with ADRP and was interpreted as pathogenic. ¹⁰ However, the report did not specify the size of the family segregating this variant, and it is possible that it is a nonpathogenic variant cosegregating with ADRP by chance. 

When the cases from a previous study by our group ⁸ are combined with the new cases described here, the rhodopsin gene accounts for approximately 25% (90/363) of families with ADRP in North America. The true prevalence of rhodopsin mutations among patients with ADRP may be a bit higher than this because the screening method (SSCP) may miss approximately 10% of point mutations. ^{11 12 13} There were 32 distinct pathogenic mutations, with 22 of them occurring in one family each. The most frequent mutation, Pro23His, accounted for 31 cases, or approximately one third of the rhodopsin-induced cases. Pro23His is found only in North America ¹⁴ and likely arose in a single ancestor. ¹⁵ Excluding Pro23His, the proportion of ADRP families with rhodopsin mutations is 59 of 332 (18%). This is close to the proportion (14/88; 16%) found in the largest European survey reported. ¹⁰ Pro347Leu is the next most frequent mutation (13 patients), accounting for approximately 3.6% of ADRP-affected families. Pro347Leu has been found in Europe, ^{16 17} Asia, ¹⁸ and Africa ¹⁹ and has arisen independently more than once. ⁶  

 Supported by National Institutes of Health Grants EY08683, EY11655, and EY00169 and by the Foundation Fighting Blindness.

 Submitted for publication December 10, 1999; revised March 23, 2000; accepted April 19, 2000.

 Commercial relationships policy: N.

 

Table 1.

 

Reported Cases of Rhodopsin Mutations in ADRP

Mutation	Sequence	Current Study (n = 91)	Previously Reported (n = 272)	Total Reported (n = 363)
Thr17Met	ACG to ATG	1	1	2
Pro23Leu	CCC to CTC		1	1
Pro23His	CCC to CAC	6	25	31
Phe45Leu	TTT to CTT	1	1	2
Gly51Arg	GGC to CGC	1	2	3
Gly51Val	GGC to GTC		2	2
Thr58Arg	ACG to AGG		2	2
Gly89Asp	GGT to GAT	1	3	4
Cys110Tyr	TGC to TAC		1	1
Gly114Asp	GGC to GAC		1	1
Gly114Val	GGC to GTC	1		1
Leu125Arg	CTG to CGG		1	1
Arg135Trp	CGG to TGG	1	2	3
Ala164Glu	GCG to GAG		1	1
Cys167Arg	TGC to CGC		1	1
Pro171Ser	CCA to TCA		1	1
Pro171Leu	CCA to CTA	1	3	4
Glu181Lys	GAG to AAG		3	3
Gln184Pro	CAG to CCG	1		1
Ser186Pro	TCG to CCG		1	1
Gly188Arg	GGA to AGA		1	1
Asp190Asn	GAC to AAC		1	1
Asp190Gly	GAC to GGC		1	1
Phe220Leu*	TTT to TTG	1*		1
Cys264del	TGC del		1	1
Lys296Glu	AAG to GAG		1	1
Ser297Arg	AGC to AGA	1		1
Val345Met	GTG to ATG		1	1
Val345Leu	GTG to CTG		1	1
Pro347Gln	CCG to CAG		1	1
Pro347Ser	CCG to TCG		1	1
Pro347Thr	CCG to ACG	1		1
Pro347Leu	CCG to CTG	2	11	13
Total cases with a pathogenic mutation (%)		18 (20)	72 (26)	90 (25)

*  Nonpathogenic variant (see text).

Figure 1.

 

Sequence of the index patients (identified with pedigree number and position in pedigree shown in Figure 2 ) with the missense changes Gly114Val (A), Gln184Pro (B), and Phe220Leu (C). For each patient, the relevant region of the rhodopsin gene sequence is shown adjacent to the sequence from an individual with the wild-type sequence. The sequences of Gly114Val and Phe220Leu are in the sense direction going from bottom to top, whereas the sequence of Gln184Pro is in the antisense direction.

Figure 2.

 

Pedigrees of families segregating the rhodopsin alleles (A) Gly114Val (family 0240), (B) Gln184Pro (family 0946), and (C) Phe220Leu (family E627). Filled symbols indicate affected individuals. Rhodopsin genotypes are beneath the symbols for those family members who donated blood for this study. M, rhodopsin variant specific for each family; +, wild-type rhodopsin coding sequence; arrow, index case in each family.