Novel Mutations in the TULP1 Gene Causing Autosomal Recessive Retinitis Pigmentosa

Abstract

purpose. To assess the contribution of TULP1 to autosomal recessive retinitis pigmentosa (arRP).

methods. Fifteen exons of the gene were screened by single-strand conformation polymorphism analysis of 7 (of 49) arRP pedigrees showing cosegregation with TULP1 locus markers.

results. In one of the seven families two allelic mutations, IVS4–2delAGA and c.937delC, were found in exons 5 and 10, respectively.

conclusions. Two novel mutations in TULP1 were found to be associated with arRP. That they both compromise the gene product supports their pathogenicity. This gene was present in no more than 2% of a panel of 49 Spanish families affected by arRP.

Retinitis pigmentosa (RP) is a group of inherited eye diseases with a world wide prevalence of 1 in 4000. A hallmark of RP is its extreme heterogeneity, both clinical and genetic. ¹ The autosomal recessive forms of RP (arRP) have been the subject of intense research during the past 10 years. At least six genes and seven loci have been found to be involved so far (RetNet web site http://www.sph.uth.tmc.edu/RetNet/). However, these explain only a small number of the known cases of arRP. 

The RP14 locus was identified at chromosome 6p, linked to the marker D6S291, by linkage analysis on a large Dominican arRP pedigree. ² Recently, mutations in the TULP1 gene have been identified in this pedigree ³ and in other patients with arRP. ^{4 5} This gene belongs to a recently identified family encoding proteins (63%–90% amino acid identity in the C-terminal regions) of unknown function present in plants, invertebrates, and vertebrates. The TULP1 cDNA was originally cloned by North et al., ⁶ and high expression was reported only in the retina. 

Homozygosity analysis and linkage to D6S291 and D6S439 has been assessed in a panel of 49 Spanish families affected by arRP. ⁷ After haplotype construction, TULP1 mutations were searched for in seven families. Two novel mutations were identified in family M-141. The severity of their predicted effect, together with the cosegregation analysis, strongly support their involvement in arRP. 

Materials and Methods

Families and Samples

Seven Spanish arRP pedigrees, comprising one consanguineous and six nonconsanguineous families, were used in this study. Pedigrees for six of them (B-31, B-26, M-56, M-141, M-235, and V-3) are shown in Bayés et al. ⁹ Control samples were obtained from Centre d’Études du Polymorphisme Humaine and volunteers working at the Department of Genetics, University of Barcelona. 

This research followed the tenets of the Declaration of Helsinki. Informed consent was obtained from all subjects after the nature and possible consequences of the study had been explained. 

Ophthalmologic Examination

Patients were studied according to the same clinical protocol. In every case a complete ophthalmologic examination was performed, including visual acuity testing with the best correction, computerized visual field testing (recorded on an Octopus 500 instrument; Interzeag AG, Schlieren, Switzerland), and biomicroscopy and fundus examination after pupillary dilation. Cone, rod, mixed, and photopic flicker (30 Hz) electroretinograms (ERGs) and electro-oculograms were performed and recorded according to the standard testing protocols for clinical electroretinography. 

DNA Analyses

Genomic DNA was prepared from peripheral blood as described. ⁸ The 15 TULP1 exons were amplified in polymerase chain reaction (PCR) by using primers anchored in the corresponding flanking intron sequences (sequences and PCR conditions were kindly provided by Alan Wright). 

Amplification in a thermal cycler (Perkin–Elmer Applied Biosystems, Foster City, CA), was performed in a total volume of 50 μl. Each reaction contained 100 ng genomic DNA, 20 picomoles of each primer, 200μ M dNTPs, and 1.25 U Taq polymerase (Promega, Madison, WI) in a buffer containing 1.0 to 1.5 mM MgCl₂ with or without 10% dimethyl sulfoxide. Reactions were generally subjected to 35 cycles of 94° for 40 seconds, X° for 30 to 40 seconds (where X° is the annealing temperature between 52° and 60°), and, optionally, 72° for 30 seconds. A final extension step was performed at 72° for 5 minutes. 

Single-strand conformation polymorphism (SSCP) analyses were performed as described previously. ⁹ Each PCR-amplified fragment was assayed under three different conditions, combining acrylamide and glycerol concentrations and running temperatures. The DNA sequence was obtained for each PCR sample showing an aberrant SSCP pattern. Sequencing was performed directly on PCR products using a dye terminator cycle sequencing premix kit (Thermosequenase; Amersham Pharmacia Biotech, Uppsala, Sweden). In addition, PCR products of exons 5 and 10 of the index patient were cloned using a ligation kit (Sureclone; Amersham Pharmacia Biotech), and clones were sequenced as described. 

The deletion IVS4–2delAGA was confirmed by MspI digestion after PCR amplification, by using a forward primer with a mismatch (italics): 5′ TTATGAAGAGTTTCTACCTCC 3′, which generated an MspI restriction site in the mutant allele. 

Results

TULP1 Analysis

Haplotype construction in 49 Spanish arRP pedigrees using markers D6S291 and D6S439 allowed us to exclude TULP1 as the cause of the disease in 42 of them. The seven remaining pedigrees were further analyzed for mutations in the TULP1 gene coding sequence. PCR amplification of the 15 exons was performed in one affected member of each family and a control, followed by SSCP. Two aberrant patterns were identified in exons 5 and 10, respectively: both in the patient sample of family M-141 (Fig. 1A ). Sequencing of the PCR products revealed the deletions IVS4–2delAGA and c.937delC (GenBank accession number U82468), as shown in Figures 1B and 1C . To confirm both mutations, PCR products of exons 5 and 10 of the index patient were cloned. Sequences of three independent mutant clones for exon 5 and five for exon 10 unambiguously established both deletions. Furthermore, the deletion IVS4–2delAGA was confirmed by restriction digestion, as explained in the Materials and Methods section. SSCP analysis of the remaining members of this family showed that the exon 5 mutation was maternal, whereas the exon 10 deletion was paternal. Consistent with the haplotype analysis, the healthy sister did not carry either mutation. Neither did any of 50 control individuals. 

Both deletions are predicted to impair protein function, because IVS4–2delAGA eliminates the splicing acceptor site of intron 4 together with the first nucleotide of exon 5, and c.937delC causes a frame shift leading to a premature stop codon in exon 10 (Fig. 1C) . 

No other novel mutations or polymorphisms were found in the TULP1 gene of any of the patients. However, the variant c.394del24 (E120-D127del), reported by Gu et al. ⁵ appeared in heterozygosis in 2 of 50 control individuals. 

Clinical Examination of Patients

The index patient, II.2 on the pedigree (Fig. 1) , reported no previous history of other sensorial or neurologic disorders. He had nystagmus that had appeared shortly after birth. Before the age of 5, he had shown poor dark adaptation, together with visual field defects and color vision alterations. Based on these findings and a typical funduscopy appearance, retinitis pigmentosa was diagnosed at 6 years of age. Since then, there has been progression of the visual field constriction and visual acuity impairment, which was less than 20/200 at the age of 20. Ophthalmologic examination 6 years later showed bilateral horizontal nystagmus. There were absolute scotomae in his visual fields, his best corrected visual acuity was counting fingers in both eyes, and he showed bilateral subcapsular opacity. Fundus findings at age 26 included normal optic disc and macula, slight constriction of arteriolar vessels, and typical bone spicule pigmented lesions in the midperiphery. Rod, mixed, and cone ERGs were completely abolished in both eyes, indicating severe impairment of functional retina. 

Patient II.3 did not report any relevant previous disorder, other than two episodes of seizures at 6 months and 15 years of age. She had poor vision, which was secondary to congenital nystagmus, and she reported having night blindness and progressive peripheral visual field constriction since the age of 2. At that time, funduscopy showed typical RP changes. Symptoms progressed rapidly, and visual fields were restricted to 20° of central visual field diameter by the age of 8. At this time, rod, mixed, and cone ERG responses were completely extinguished. At present, she has myopia, with a refraction of− 4.50° to −2.75° × 178° in the right eye and −3.75° to −2.75° × 5° in the left eye, and her binocular visual acuity is counting fingers (1 m in the right eye and 2.5 m in the left eye). Visual fields are restricted to a very small inferior paracentral area in both eyes. The fundus shows myopic changes: posterior staphyloma and Fuchs’ spots. 

Neither of the two patients is obese nor shows other endocrinologic disorders or hearing impairment. 

Discussion

Two patients with RP are described who are compound heterozygotes for two novel mutations in the TULP1 gene: IVS4–2delAGA and c.937delC. Both mutations are small deletions, whose predicted effect would be a severe impairment of the gene product. IVS4–2delAGA destroys the intron4–exon5 junction. Because the 100% conserved dinucleotide AG (positions −2 and −1) is missing, splicing at this site cannot occur. Possible outcomes include skipping of exon 5 and usage of alternative neighboring cryptic 3′ acceptor sites. Using the formula developed by Shapiro and Senapathy, ¹⁰ we have calculated the score value of the wild-type 3′ acceptor site (WT), as well as that of the mutant site (MS), and two alternative sites (AS) located in intron 4 and exon 5, respectively (not shown). WT scores 90%, and MS, AS1, and AS2 score 80%, 72%, and 70%, respectively. All these values are within the range of functional sites. Usage of the mutant site would result in a deletion of five nucleotides in the mature mRNA, leading to a translation frame shift and a premature stop codon within exon 6. On the other hand, skipping of exon 5 would imply a Glu-to-Gly substitution and an in-frame deletion of 50 amino acids. 

Mutation c.937delC clearly leads to a truncated gene product in exon 10. Even though the structural domains of TULP1 are still unknown, the absence of 44% of the residues at the C-terminal conserved moiety is unlikely to be compatible with function. 

Figure 2 is a summary of the TULP1 mutations reported so far in patients with arRP. Although the amount of data gathered is not very large, there is a bias in the distribution of pathogenic mutations, most of which are in exons 10 to 15, and all the other mutations probably compromise the expression of the C-terminal region. These exons contain the sequences most conserved among members of the tub family and thus it could be assumed that they are critical to TULP1 function. On the other hand, the N-terminal half could be less crucial. In this respect, the finding of an in-frame deletion of 24 nucleotides in exon 5 (c.394del24, E120-D127del) in two control individuals of our series is noteworthy. This same deletion was reported by Gu et al. ⁵ in 2 of 155 patients with arRP (mainly of German origin). Although it appeared in heterozygosis and no other mutation could be found in the other allele of the patients, Gu et al. suggested a disease-causing effect based on the deletion of eight acidic residues in the protein and their inability to locate the mutation in control chromosomes. Although our data are not sufficient to rule out the pathogenicity of c.394del24 (E120-D127del), its presence at a frequency of approximately 2% in control chromosomes suggests that it may be a polymorphism. 

Concerning the phenotype, neither affected individual of family M-141 was obese or had any hearing impairment, but both showed a very severe visual handicap, as shown by the early age of onset, the rapid progression, and the abolished ERGs. Although the age of onset has not been reported in all cases described, the overall clinical features associated with TULP1 mutations rank among the most severe for RP. This indicates that TULP1 is essential for proper function of the retina. Further research should be conducted in model organisms to shed light on the TULP1 function. 

Although TULP1 mutations seem to account for only 2% of arRP cases worldwide, knowledge of its function will undoubtedly improve our understanding of the normal and affected retina. 

 EP and LH contributed equally to this work.

²  Present address: Department of Medical Genetics, Haartman Institute, Helsinki, Finland.

 Research was funded by grants from CIRIT, Spanish Ministry of Education Grants SAF96-0329 and UE98-0014, and European Union Grant BIO4-CT97-2123; automated DNA sequencing was performed at the Serveis Científico–Tècnics at the Universitat de Barcelona; and LH is the recipient of a postdoctoral fellowship from the Swedish Foundation for International Cooperation in Research and Higher Education (STINT) and a travel grant from the Swedish Society of Medicine; and EB and MB are the recipients of predoctoral fellowships from Universitat de Barcelona (EB) and CIRIT (MB; Catalan Government), respectively.

 Submitted for publication January 11, 1999; revised May 6, July 19, and September 13, 1999; accepted October 5, 1999.

 Commercial relationships policy: N.

 

Figure 1.

 

TULP1 mutations in family M-141. (A) SSCP patterns of exons 5 and 10 for all members of the family and a control individual. (B) Changes, at the nucleotide level, associated with mutations IVS4–2delAGA and c.937delC. For the latter, the translation frame shift and the premature stop codon are shown. (C) Exon 5 (reverse strand) and exon 10 sequence chromatograms of a mutant, a wild-type, and an heterozygous sample. The positions of the deletions are marked by vertical arrows in the mutant chromatograms, and their corresponding positions are boxed in the wild-type chromatograms. In the heterozygous sample chromatograms, overlapping sequence patterns are observed starting at the points of the deletions (marked by arrowheads).

Figure 2.

 

TULP1 mutations involved in arRP, drawn on a scheme of the genomic structure. Only mutations explaining both alleles of a patient are included. Lines connect heterozygous mutations found together in a patient. The two novel mutations reported here are in bold type and boxed. Solid exons: C-terminal sequence most conserved among members of the tub family. ^{3 5 6}

The authors thank Alan Wright for information and technical assistance in the analysis of TULP1 and Robin Rycroft for revising the English.