Hum Genet (2001) 109 : 498–502 DOI 10.1007/s004390100600
O R I G I N A L I N V E S T I G AT I O N
Dawn L. Thiselton · Christiane Alexander · Alex Morris · Simon Brooks · Thomas Rosenberg · Hans Eiberg · Birgit Kjer · Poul Kjer · Shomi S. Bhattacharya · Marcela Votruba
A frameshift mutation in exon 28 of the OPA1 gene explains the high prevalence of dominant optic atrophy in the Danish population: evidence for a founder effect Received: 25 May 2001 / Accepted: 3 August 2001 / Published online: 3 October 2001 © Springer-Verlag 2001
Abstract Dominant optic atrophy (DOA) is a hereditary optic neuropathy characterised by decreased visual acuity, colour vision deficits, centro-coecal scotoma and optic nerve pallor. The gene OPA1, encoding a dynamin-related GTPase, has recently been identified within the genetic linkage interval for the major locus for DOA on chromosome 3q28 and shown to harbour genetic aberrations segregating with disease in DOA families. The prevalence of the disorder in Denmark is reported to be the highest of any geographical location, suggestive of a founder effect. In order to establish the genetic basis of disease in a sample of 33 apparently unrelated Danish families, we screened DNA from affected members for OPA1 gene mutations by heteroduplex analysis and direct sequencing. A novel identical mutation in exon 28 (2826delT) was associated with DOA in 14 pedigrees and led to a frameshift and abnormal OPA1 protein –COOH terminus. Haplotype analysis of a region of ~1 Mb flanking the OPA1 gene using eight polymorphic markers revealed a common haplotype shared by all 14 patients; this haplotype was markedly
D. L. Thiselton and C. Alexander contributed equally to this study D.L. Thiselton (✉) · C. Alexander · A. Morris · S. Brooks · S.S. Bhattacharya · M. Votruba Department of Molecular Genetics, Institute of Ophthalmology, University College London, EC1 V 9EL, UK e-mail:
[email protected], Tel.: +44-207-6086839, Fax: +44-207-6086863 T. Rosenberg · P. Kjer National Eye Clinic for the Visually Impaired, Hellerup, Denmark H. Eiberg University Institute of Medical Biochemistry and Genetics, Copenhagen, Denmark B. Kjer Eye Department, Frederiksborg County Hospital, Helsevej 2, DK-3400, Hillerod, Denmark M. Votruba Moorfields Eye Hospital, London, UK
over-represented compared with ethnically matched controls. Statistical analysis confirmed significant linkage disequilibrium with DOA over ~600 kb encompassing the disease mutation. We have therefore demonstrated that the relatively high frequency of DOA in Denmark is attributable to a founder mutation responsible for ~42% of the examined families and suggest that presymptomatic screening for the (2826delT) mutation may facilitate diagnosis and genetic counselling in a significant proportion of DOA patients of Danish ancestry.
Introduction Kjer type optic atrophy (dominant optic atrophy; DOA) is an eye disease inherited in an autosomal dominant manner with incomplete penetrance (Kjer et al. 1996; Kivlin et al. 1983; Toomes et al. 2001) and a highly variable expression ranging from subclinical cases to those legally registered blind. The disorder is characterised by an insidious onset of visual impairment in early childhood with moderate to severe loss of visual acuity, temporal optic disc pallor, colour vision deficits and centrocoecal scotoma of variable density (Hoyt 1980; Votruba et al. 1998). Histopathological studies (Kjer et al. 1983; Johnston et al. 1979) have pointed towards a primary degeneration of the retinal ganglion cell layer, accompanied by ascending atrophy of the optic nerve. The predominant genetic locus for this disorder (OPA1: MIM 165500) resides in a 1.4-cM interval on 3q28–29 (Jonasdottir et al. 1997). We have recently published the cloning of the disease gene, OPA1 (Alexander et al. 2000; Delettre et al. 2000), which encodes a dynamin-related GTPase that is targetted to mitochondria. Mutations segregating with dominant optic atrophy have been described throughout the coding portion of OPA1, with the majority resulting in truncated polypeptides and located in the highly conserved GTPase domain or towards the 3’ coding portion of the gene (Alexander et al. 2000; Delettre et al. 2000; Pesch et al. 2001; Toomes et al. 2001). The population frequency of dominant optic atrophy is reported to be 1:50,000 (Lyle 1990), rendering
499
it the most common of the hereditary optic neuropathies. However, the frequency of this disease in Denmark is notably much higher, approaching 1:10,000 (Kjer et al. 1996) and possibly reflecting a demographic trait (founder effect) or a particularly high diagnostic sensitivity. To investigate the hypothesis that the high prevalence in this population is attributable to the spread of a limited number of mutations, a cohort of 33 index patients from apparently unrelated 3q28-linked Danish DOA families were analysed for OPA1 mutations and associated haplotypes.
Materials and methods Patients A total of 33 Danish DOA families were analysed. The diagnosis of DOA was based on insidious visual loss during infancy, colour vision defects including a tritan defect, temporal pallor of the optic disc and familial occurrence. Visual fields and visual evoked potentials were measured routinely. Blood samples were obtained from patients with informed consent and genomic DNA was prepared from these samples by standard methods (Sambrook et al. 1989). OPA1 mutation screening All 28 coding exons and splice sites of the OPA1 gene were amplified with primers described previously (Pesch et al. 2001) and screened for DNA sequence alterations by heteroduplex analysis (Keen et al. 1991). Polymerase chain reaction (PCR) products were allowed to re-anneal and then analysed by gel electrophoresis on MDE Flowgen gels run at 180 V for 14–18 h on a Hoeffer 600S apparatus. PCR products showing heteroduplexes were purified (QIAquick PCR purification kit-250, Qiagen) and bidirectionally sequenced with fluorescent dideoxynucleotides (PE Biosystems, Foster City, USA) prior to separation on an ABI automated sequencer (Applied Biosystems, Foster City, USA).
Results Upon OPA1 mutation screening, we discovered a heteroduplex derived from the exon 28 amplification product to be present in 14 of the Danish DOA patients. Examination of the corresponding DNA sequence in all cases revealed an identical 1-bp deletion (2826delT) leading to a frameshift (Fig. 1). Manual scrutiny of the sequence disruption revealed that the 1-bp deletion actually results from the deletion of 11 bp (2826–2836) being replaced with a novel insertion of 10 bp (Fig. 1). The consequent frameshift leads to a delayed stop, downstream of the original stop codon, and results in the replacement of the last 18 amino acids of the protein with 22 novel amino acids (NH2-DAPKKNLMLSLKLFIRRNKLSE-COOH). Segregation of the mutation with disease was demonstrated by heteroduplex analysis in additional affected individuals of three index pedigrees that have been wellcharacterised previously and in which disease is linked to 3q28: IOA-168, IOA-184 and IOA-161 (Eiberg et al. 1994; Kjer et al. 1996; Jonasdottir et al. 1997). The results of this analysis are shown in Fig. 2 and, together with the finding that this mutation was not observed in 50 ethnically matched controls (data not shown), support the prediction that this OPA1 gene mutation is likely to be responsible for the disease phenotype. Its discovery in
Haplotype and statistical analysis Genotypes were constructed by using eight polymorphic markers closely linked to the OPA1 gene, viz. microsatellites D3S3669, D3S1523, D3S3642, D3S3590, D3S2305 and D3S3562 and SNPs OPA1 IVS8 +32 (c→t), IVS18+51 (t→g)). The SNPs at IVS8+32 (c/t) and IVS18+51 (t/g) were discovered during mutation screening of other DOA pedigrees (data not shown). To test the null hypothesis of no association of marker alleles with disease mutation, the frequency distribution of alleles between patient and control chromosomes was analysed using the linkage disequilibrium programs CLUMP (v1.5; Sham and Curtis 1995) and DISLAMB (v2.1; Terwilliger 1995), accessed via the Human Genome Mapping Resource Project Centre at http://www.hgmp.mrc.ac.uk/. DISLAMB uses a likelihood-ratio test (LRT) to measure the allelic association of a single marker to the disease locus, together with the parameter λ, which is the proportion of increase of the ancestral allele in disease chromosomes, relative to its population frequency. In this model, the frequency of the associated allele equals p + λ(1-p), where p is the allele frequency in the general population. For this analysis, the associated allele is not predetermined, but rather λ is estimated over all alleles. The disease gene frequency was taken to be 0.0001.
Fig. 1 Electropherogram section showing the exon 28 OPA1 (2826delT) mutation found in 14/33 Danish patients (A) and a normal control (B). The sequence shown is the reverse complement, containing a nucleotide deletion (underlined) and insertion (in bold) of CTTGAATTTCTCTAACTTT to CTTTGGAGCATCCACTTT, which results in (2826delT) on the forward strand and a frameshift (indicated by an arrow). Nucleotide designation commencing with 1 at position 56 (translation start) of GenBank entry AB011139
500
Fig. 2 Segregation of the exon 28 (2826delT) mutation in two Danish index pedigrees by heteroduplex analysis. Solid symbols Individuals affected with DOA, open symbols unaffected individuals. Lane assignments correspond to individuals in the pedigrees shown directly above the gel image. Heteroduplexes are seen in affected individuals from different branches of each family
14/33 Danish DOA families led us to suspect that this genetic aberration represents a founder allele that would explain the high prevalence of the disease in Denmark (Kjer et al. 1996).
To investigate the possibility of a common lineage in the Danish DOA patients harbouring the exon 28 mutation, we conducted haplotype analysis on DNA from all 14 individuals, with genotypes also being determined for 42–50 ethnically matched control individuals, using eight polymorphic markers within or flanking the OPA1 gene locus and spanning ~1 Mb. In 11 of the Danish families, a single sample only was available, thus it was not possible to determine phase. Nevertheless, all exon 28 mutation carriers had genotypes compatible with their sharing the same haplotype for six out of the eight markers studied (Table 1). As expected, linkage disequilibrium was observed, with all 14 DOA patients carrying a single predominant allele showing elevated frequency over controls, for markers spanning the D3S1523–D3S2305 interval (Table 1). These “ancestral” alleles presumably represent the chromosome upon which the founding mutation occurred and form a haplotype (shown bold in Table 1), which is markedly over-represented in the patient group compared to controls (χ2=50, P
