Diabetes insipidus, diabetes mellitus, optic atrophy and deafness (DIDMOAD) caused by mutations in a novel gene (wolframin) coding for a predicted ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
1998 Oxford University Press Human Molecular Genetics, 1998, Vol. 7, No. 13 2021–2028
ARTICLE
Diabetes insipidus, diabetes mellitus, optic atrophy
and deafness (DIDMOAD) caused by mutations in a
novel gene (wolframin) coding for a predicted
transmembrane protein
Tim M. Strom1, Konstanze Hörtnagel1, Sabine Hofmann1,3, Florian Gekeler1,5,
Curt Scharfe1, Wolfgang Rabl4, Klaus D. Gerbitz2,3 and Thomas Meitinger1,*
Downloaded from https://academic.oup.com/hmg/article/7/13/2021/634861 by guest on 14 October 2021
1AbteilungMedizinische Genetik, Klinikum Innenstadt, Ludwig-Maximilians-Universität, Goethestraße 29, 80336
München, Germany, 2Institute für Klinische Chemie, 3Diabetesforschung and 4Kinderklinik, Akademisches
Lehrkrankenhaus München-Schwabing, 80804 München, Germany and 5Neurologische Klinik, Klinikum
Grosshadern, Ludwig-Maximilians-Universität, 81377 München, Germany
Received September 7, 1998; Revised and Accepted September 29, 1998 DDBJ/EMBL/GenBank accession nos Y18064, AJ011971
Wolfram syndrome is an autosomal recessive disorder characterized by juvenile diabetes mellitus, diabetes
insipidus, optic atrophy and a number of neurological symptoms including deafness, ataxia and peripheral
neuropathy. Mitochondrial DNA deletions have been described in a few patients and a locus has been mapped
to 4p16 by linkage analysis. Susceptibility to psychiatric illness is reported to be high in affected individuals and
increased in heterozygous carriers in Wolfram syndrome families. We screened four candidate genes in a refined
critical linkage interval covered by an unfinished genomic sequence of 600 kb. One of these genes, subsequently
named wolframin, codes for a predicted transmembrane protein which was expressed in various tissues,
including brain and pancreas, and carried loss-of-function mutations in both alleles in Wolfram syndrome
patients.
INTRODUCTION (4,5). The similarity in phenotype between patients with Wolfram
syndrome and those with certain types of respiratory chain
The first description of Wolfram syndrome (MIM 222300) is diseases led to the investigation of mitochondrial DNA (mtDNA)
attributed to the physician D.J. Wolfram (1), who reported four mutations in Wolfram syndrome patients. Deletions in mtDNA
cases in 1938. The acronym DIDMOAD summarizes the most and morphological mitochondrial abnormalities were reported in
frequent findings: diabetes insipidus, diabetes mellitus, optic some sporadic and familial cases (6–8). However, the mitochon-
atrophy and deafness. The minimal criteria for diagnosis are drial tRNALeu (3243) mutation (2) and deletions in the mitochon-
diabetes mellitus and optic atrophy. Diabetes insipidus, sensori- drial genome have been excluded in >20 patients (9–11).
neuronal deafness, urinary tract atony, ataxia, peripheral neuro- Linkage analysis in families with autosomal recessive Wolfram
pathy, mental retardation and psychiatric illness are additional syndrome has shown significant lod scores for a locus on
symptoms seen in the majority of patients. Diabetes mellitus is chromosome 4p16 between D4S432 and D4S431 (12,13). In two
usually the first symptom, with the age of onset in the second half families in whom linkage has been demonstrated to this locus,
of the first decade (range 1–26 years). The mean onset of optic multiple mtDNA deletions were found in both homozygous and
atrophy is in the first half of the second decade (range 6 weeks to heterozygous family members, suggesting a nuclear disease gene
30 years). The prevalence has been estimated to be 1/770 000 in which interacts deleteriously with the mitochondrial genome (6).
the UK and 1/100 000 in the North American population (2,3). Although the phenotype in Wolfram syndrome is highly variable,
The syndrome is best described as a neurodegenerative disorder no distinct clinical subgroups have been described so far.
involving the central nervous system, peripheral nerves and Psychiatric manifestations are particularly diverse and an in-
neuroendocrine tissue. Affected siblings with unaffected parents, creased predisposition for psychiatric disorders has been pro-
often consanguineous, suggests a recessive mode of inheritance posed for heterozygous carriers (14).
*To whom correspondence should be addressed. Tel: +49 89 5160 4466; Fax: +49 89 5160 4780; Email: thomas@pedgen.med.uni-muenchen.de2022 Human Molecular Genetics, 1998, Vol. 7, No. 13
Table 1. Clinical manifestations in patients with Wolfram syndrome
Family Patient Gender Agea Diabetes Progressive Abnormal Diabetes Renal tract Neurological Other complications Consanguinity
(years) mellitus optic atrophy audiogram insipidus abnormalities abnormalities
1 5519 f 22 6y + + 6y + Ataxia, nystagmus Retarded sexual maturation, depression –
13883 f 11 4y – – – + – – –
2 13775 f 20 9y 12 y 17 y 15 y – – – –
13776 m 17 10 y 14 y 13 y 15 y + – Retarded sexual maturation –
3 13766 f 26 13 y 7y 13 y – – – – +
4 13070 f 22 4y 9y 19 y – + Abnormal EEG Psychiatric illness –
5 13885 f 35 8y + + – + – Cataract –
6 13062 f 25 7y 9y 22 y 7y 24 y Ataxia, nystagmus
7 13076 m 26 11 y 17 y 17 y 17 y – – Retarded sexual maturation, mental retardation –
8 13073 f 35 6y 5y + – 15 y Ataxia Cataract, psychiatric illness, ragged red fibers
Downloaded from https://academic.oup.com/hmg/article/7/13/2021/634861 by guest on 14 October 2021
9 13781 m 19 4y 10 y 6y 14 y 10 y Abnormal EEG Retarded sexual maturation +
10 13782 m 16 14 y 15 y 15 y – 15 y – – –
11 13783 f 12 11 y 6y 6y – – Ataxia –
12 12131 m + + +
12132 m + + +
aAge at time of the study.
We used unfinished genomic sequences from a chromosome 4 a b
sequencing project to identify candidate genes for mutation
screening in patients with Wolfram syndrome. Segregation
analysis in autosomal recessive families led to a refined critical
interval of 600 kb which was shown to contain at least four genes.
One of them, a novel gene coding for a predicted 100 kDa
transmembrane protein, harbored mutations in eight patients with
Wolfram syndrome.
RESULTS
Analysis of candidate genes
We investigated DNA samples from three familial and nine
sporadic patients with a clinical diagnosis of Wolfram syndrome.
Minimum ascertainment criteria were juvenile diabetes mellitus
and progressive optic atrophy (Table 1). mtDNA mutations and
deletions were excluded in all affected individuals. In seven index Figure 1. Refined interval for the Wolfram syndrome locus. (a) The critical
cases (families 1–7, Table 1), an extensive screen for mtDNA interval in 4p is marked with a bar. (b) Chromosome 4p16 haplotypes for family
02 exclude D4S412 and D4S2354 from the critical region assuming that the
mutations has been reported, including single-stranded con- disease locus in this family was linked to the reported interval between D4S432
formation polymorphism (SSCP) analysis of all mitochondrial and D4S431. Mutation analysis in the wolframin gene confirmed that
ND and tRNA genes (11). In five further cases, a deletion screen individuals 13779 and 13780 are heterozygous carriers.
by Southern blotting of mtDNA was performed. In addition, the
three primary LHON mutations (3460, 11778 and 14484) and
three MELAS mutations (3243, 3271 and 3251) were not markers D4S2354 and D4S431, and made unfinished sequences
detected on screening. available electronically to all investigators (http://www-
We performed segregation analysis with 4p markers in three shgc.stanford.edu/Seq ). We analyzed the DNA sequences from
families in which at least two siblings were affected. Assuming six BACs falling into the interval between D4S2354 and D4S431
that the disease locus in these families was linked to the reported (Fig. 2). All six BAC clones harbored gaps which, given the total
interval between D4S432 and D4S431, we screened for single length of the clones, should constitute2023
Human
NucleicMolecular Genetics,
Acids Research, 1998,
1994, Vol. Vol. 7, No.
22, No. 1 13 2023
Downloaded from https://academic.oup.com/hmg/article/7/13/2021/634861 by guest on 14 October 2021
Figure 2. Physical map of the Wolfram syndrome critical region. In the lower part, the position of DNA markers and BACs are drawn to scale as estimated by the
unfinished sequence data (Stanford Human Genome Center). The positions of genes in this region are indicated by boxes. The upper part of the figure shows the
genomic structure of the wolframin gene (WM1). The gene covers a region of ∼33 kb. The genomic sequence contained a gap between exons 1 and 2 and between
exons 7 and 8. Only the first half of exon 1 and 800 bp upstream of the cDNA were covered by genomic sequence. WM1, WM3 and WM4, candidate genes for
mutations in Wolfram syndrome patients; DRP-1, dihydropyriminidase related protein-1; BRg, γ isoform of the B regulatory subunit of protein phosphatase 2A.
revealed at least four known or predicted genes in this region (Fig. Overlapping RT–PCR and 5′-RACE was performed with adult
2): (i) the gene for dihydropyriminidase-related protein-1 brain tissue as template to establish the cDNA sequence of WM1.
(DRP-1, GenBank accession no. D78012); (ii) the human The cDNA consists of 3628 bp encompassing seven small exons
homolog of the rat mRNA for the γ isoform of the B regulatory and a large last exon of 2609 bp. Exons 3–5, 7 and 8 were
subunit of protein phosphatase 2A (BRg, GenBank accession no. predicted by GENSCAN (15), GRAIL (16) and GENEFINDER,
D38261); (iii) WM3, a novel gene partially covered by ESTs exon 6 by GENSCAN and GRAIL and exon 2 by GENSCAN
detected in brain and retinal tissue and with an overall amino acid only. Sequence analysis of the cDNA predicted an ORF of 2670
identity of 58% (similarity 74%) over 397 amino acids to the bp. The sequences around the putative ATG translation start site
KIAA0555 protein (GenBank accession no. 3043634) of un- (cgggccggcaggatgg) contained the –3 purine and +4 guanine
known function; (iv) WM1, also partially covered by ESTs, residues of the Kozak consensus sequence. No in-frame stop
which showed no significant homology to published DNA or codon was present upstream of the predicted start site. A possible
protein sequences. A potential fifth gene, WM4, was defined by promotor was predicted (TSSW program) at the beginning of the
composite cDNA with a TATA box 34 bp upstream of the putative
gene prediction programs only. For this gene, at least 6 exons
transcription start site. The 3′-UTR consisted of 797 bp with a
spread over a genomic region of 40 kb were predicted by at least
polyadenylation signal (AATAAA) 771 bp downstream of the
two programs used. No significant homology could be retrieved stop codon.
in a database search with sequences encompassing WM4.
Primers were designed for a total of 36 exons belonging to the
above genes, with the exception of WM4, and a mutation screen wolframin mutations
was performed by SSCP in all 12 Wolfram syndrome index To assess the possible role of this candidate gene in Wolfram
patients. Aberrant migration patterns suggesting mutations in the syndrome, exons 2–7 were amplified from affected individuals
DRP-1, BRg and WM3 genes were few and sequencing revealed with primer pairs based on exon–intron boundaries (Table 2) and
intronic nucleotide substitutions, a silent mutation in exon 8 of screened for mutations using SSCP. Exon 8 was divided into eight
BRg and an arginine/histidine and a proline/threonine poly- overlapping parts of ∼300 bp for SSCP screening. Sequencing of
morphism in exon 6 of DRP-1. In contrast, screening of WM1 exons with aberrant migration patterns (Fig. 3) revealed muta-
resulted in a significant number of gel shifts, some of which tions in the three familial cases and in five of the nine sporadic
turned out to be due to homozygous deletions in Wolfram cases (Table 3). The mutations included stop, frameshift and
syndrome patients. WM1 was therefore named wolframin and amino acid deletions and insertions, suggesting wolframin loss-
characterized further. of-function mutations as the cause of Wolfram syndrome.2024 Human Molecular Genetics, 1998, Vol. 7, No. 13
Table 2(a) Intron–exon organization of wolframin
Exon Exon length (bp) Donor splice site Intron length (bp) Acceptor splice site
1 157 >6700 ttgacttttcttccaggcag/GATGGACTCC
2 233 GACGGCACCG/gtaagggagcaggctgggaa 9405 ttgtttcttctgtgttaaag/GGCCTACAAA
3 83 ACAGACTGAG/gtgaggactgcggtgccggc 1811 gactggtgtctggcttgcag/GTGGGGAAGC
4 145 GACAGAAGAG/gtgggtctgtgtgaggctta 2065 gaccacatcctatccctcag/GCATCACGTC
5 171 AACGAGCACG/gtgcgaggattcaccctggg 549 atccaccctgtcccctgcag/ATGGAGGGGC
6 81 GGCAGCGAGT/gtgagtgcagcccctgcccc 3043 tgttttctctcatgcttcag/CCAAGAACTA
7 149 GCGTCTGAAG/gtgagtgaccaagaccccgg >6000 acgtaccatctttcccccag/GTGGTCAAGT
8 2609 GGAACCTGCA
The exonic and intronic sequences are indicated in upper case and lower case letters, respectively.
Downloaded from https://academic.oup.com/hmg/article/7/13/2021/634861 by guest on 14 October 2021
Table 2. (b) Primers for PCR amplification and sequencing of wolframin exons
Exon Sense primer Antisense primer PCR product size (bp)
2 CTGGATGTGCCTGACCTTG CCTGAACATCCCCAGCCTG 311
3 GAAGACCCTCATGCCTTGTC ATCTCAGGCACCGACACTTC 272
4 CGGAGAATCTGGAGGCTGAC CAACCCTCCAGAGGCTGTTC 234
5 ACAAGGCCTTTGACCACATC GTGCCCAGGGTGAATCCTC 225
6 CTGTTAATCCACCCTGTCCC GAGTCGCACAGGAAGGAGAG 186
7 CCTCCACCTGAACCCACTCA ACCGGGGTCTTGGTCACTC 301
8-1 TTCCCACGTACCATCTTTCC CACATCCAGGTTGGGCTC 334
8-2 AGAACTTCCGCACCCTCAC TCAGGTAGGGCCAATTCAAG 330
8-3 CTATCGCTGCTGCCCTCC GGGCAAAGAGGAAGAGGAAG 307
8-4 GTGAGCTCTCCGTGGTCATC CCCTCTGAGCGGTACACATAG 346
8-5 ATCCTGGTGTGGCTCACG GTAGAGGCAGCGCATCCAG 300
8-6 GCGTGACTGACATCGACAAC GCTGAACTCGATGAGGCTG 356
8-7 CAGCAGCGAGTTCAAGAGC CCTCATGGCAACATGCAC 300
Table 3. Mutations in nine families with Wolfram syndrome
Family Patient Mutationa Type Exon
1 5519 1380del9b 9 bp deletion 8
2 13775 460+1G→Ab 5′ splice signal 4
4 13070 599delTb Frameshift 5
5 13885 Q366X, 1096C→T Stop 8
6 13062 Q226X, 676C→T Stop 6
Q819X, 2455C→T Stop 8
7 13076 599delT Frameshift 5
Y669C Missense 8
8 13073 Q366X, 1096C→T Stop 8
Q520X, 1558C→T Stop 8
9 13781 1523delATb Frameshift 8
12 12131 2164ins24b 24 bp insertion 8
aNucleotide positions are given counting from the first base of the start codon.
bHomozygous mutations.2025
Human
NucleicMolecular Genetics,
Acids Research, 1998,
1994, Vol. Vol. 7, No.
22, No. 1 13 2025
a b protein sequences. Protein structure prediction programs (17,18)
identified a hydrophobic central domain (amino acids 330–650)
comprising nine helical transmembrane segments. Both pro-
grams preferentially predicted the N-terminal hydrophilic part to
lie extracytoplasmically and the C-terminal part to lie intracyto-
plasmically. PSORT (19) computed a probability of 65% for a
localization in the plasma membrane, 26% for endoplasmatic
reticulum and 4% for a mitochondrial or nuclear localization.
These features suggested wolframin encodes a transmembrane
protein.
Hybridization of a PCR-derived exon 8 fragment (wm1E7-2-5)
against a Southern blot containing EcoRI- and PstI-digested
genomic DNA revealed a single band per digest and no
Figure 3. Northern blot analysis of wolframin. A multiple tissue blot (Clontech) cross-hybridization signals. The bands corresponded to the
was hybridized with probe wm1E7-2-5 (exposure 4 h) and rehybridized with expected EcoRI fragment of 16.4 kb and the expected PstI
β-actin cDNA as control (exposure 30 min). An ∼3.6 kb transcript was observed fragment of 4.4 kb. Probe wm1E7-2-5 was also used to hybridize
Downloaded from https://academic.oup.com/hmg/article/7/13/2021/634861 by guest on 14 October 2021
under stringent washing conditions in heart, brain, placenta, lung, liver, skeletal
muscle, kidney and pancreas. a Zoo blot and specific signals were observed in mammals (data
not shown). No cross-hybridization was observed with genomic
DNA from Drosophila and yeast. Hybridization of the probe to
a multi-tissue northern blot revealed a signal of ∼3.6 kb in all
tissues (Fig. 4). The signal was strong in heart, intermediate in
brain, placenta, lung and pancreas and weak in liver, skeletal
muscle and kidney. The 3.6 kb band corresponded to the length
of the assembled wolframin cDNA.
Mouse homolog of wolframin
Primers derived from mouse EST sequences were used to amplify
the mouse homolog of wolframin by RT–PCR. 5′-Sequences
corresponding to human exons 3 and 4 and 3′-sequences
corresponding to exon 8 were covered by the mouse IMAGE
clones 526533 and 1152973, respectively. IMAGE clone 526533
was presumably mis-spliced and, in addition to exons 3 and 4,
contained sequences corresponding to the 3′-UTR of wolframin.
A sequence was amplified which corresponds to the human
cDNA sequence starting at nucleotide position 316. 5′-RACE
was used to amplify an additional 206 bp resulting in a composite
sequence which covers the entire ORF of mouse wolframin.
Human and murine sequences are highly homologous, with 83%
identity at the nucleotide level and 87% at the amino acid level
(Fig. 5).
DISCUSSION
Figure 4. Mutation analysis in family 02. (a) Co-segregation of Wolfram
syndrome with SSCP band shifts. Affected and unaffected individuals are We have identified a gene mutated in patients with Wolfram
presented by standard symbols. (b) Identification of a homozygous G→A
transition at nucleotide position 460+1 at the donor splice site of intron 4.
syndrome by a positional cloning approach. Loss-of-function
mutations on both alleles were demonstrated in five out of 12
families studied and we conclude that lack of the putative gene
Sequencing of aberrant bands also revealed several intronic and product (wolframin) is causative for the disease. The mutation
eight exonic polymorphisms. Six of the exonic polymorphisms status of a heterozygous missense mutation (Y669C), and of two
are silent substitutions (PM684 C/G, PM1185 C/T, PM1500 C/T, homozygous in-frame mutations (1380del9 and 2164ins24)
PM2433 A/G, PM2469 C/T and PM2565 A/G). Two exonic remains unclear pending functional studies. These sequence
polymorphisms are characterized by either valine or isoleucine at alterations were not found in 200 control chromosomes.
amino acid position 333 (PM997 A/G) and by either histidine or In one of the families (family 05) only a heterozygous stop
arginine at amino acid position 611 (PM1832 A/G). mutation was found. No mutations in either of the two alleles were
detected in three families (families 03, 10 and 11). One of these
Characterization of wolframin (family 03) was reported to be consanguineous. Mutations in exon
1, which was not included in the mutation screen, intronic
The 890 amino acid wolframin protein corresponded to a mutations including deletions or mutations in the regulatory
predicted molecular weight of 100 kDa. BLAST and FASTA flanking regions of the gene could be pathogenic in these families.
searches showed no significant homology to published DNA or Non-allelic heterogeneity provides an alternative explanation (12).2026 Human Molecular Genetics, 1998, Vol. 7, No. 13
Downloaded from https://academic.oup.com/hmg/article/7/13/2021/634861 by guest on 14 October 2021
Figure 5. Alignment of the human wolframin protein sequence and its mouse homolog. Predicted transmembrane regions (17) are indicated by bars under the sequence.
The numbers above the sequence correspond to the family numbers in Table 3 and indicate the positions of the mutations.
Wolfram syndrome is a rare autosomal recessive disorder and, event in a family with only two affected members. Mutation
accordingly, a high proportion of homozygous mutations (five analysis in this family (family 02) has shown that there was
out of eight) were observed, although consanguinity was indeed a homozygous splice site mutation present in the
recorded in only one of the five families. The mutations were wolframin gene. The five genes predicted by EST clusters and
distributed along the entire gene and there was no evidence for a exon prediction programs occupy ∼225 kb of the 600 kb refined
founder mutation in the seven families. No obvious genotype– interval. ESTs derived from the gene coding for DRP-1 (Gen-
phenotype relationship emerged. For example, diabetes insipidus Bank accession no. D78012) and WM3 (SHGC-25149) had
was absent in the index patient of family 04 with a proximal already been localized to this region. The gene encoding BRg
frameshift mutation, but present in the index patient of family 06 (GenBank accession no. D38261) is located tail-to-tail only 17 kb
with the most distal stop mutation. proximal of wolframin. It has been shown to be highly expressed
This cloning project was greatly aided by the unfinished in brain and in spinal cord (20). The 5′-end of wolframin was not
genomic sequences made available from the Stanford Human fully represented on genomic sequences. It is known from
Genome Center. With no chromosomal aberration present in our large-scale sequencing projects that promoter regions are under-
panel of patient samples, gene identification relied on exon represented due to GC-rich regions at upstream regulatory
prediction and the identification of ESTs in the critical linkage sequences. According to the human/mouse synteny map, the
interval. For refining this interval, we relied on a recombination highly homologous mouse wolframin gene should be located on2027
Human
NucleicMolecular Genetics,
Acids Research, 1998,
1994, Vol. Vol. 7, No.
22, No. 1 13 2027
mouse chromosome 5. Flanking markers include the mouse genes (15), GRAIL (16) and GENEFINDER (C. Wilson, L. Hillier and
Msx1 and Drd5. No mouse mutants with degenerative disease P. Green, personal communication) were used as exon prediction
map to this region. programs. Genome-wide repeats were identified using the
mtDNA deletions have been excluded in peripheral leukocytes REPEATMASKER program. GENOTATOR was used to display
of the 12 index patients investigated in the present study. In the results of the exon predictions programs, BLAST and repeats
particular, no mtDNA mutation was found in skeletal muscle searches (23). Sequence alignments were performed using the
from a patient with a homozygous wolframin mutation (family Global Alignment Program (GAP) and CLUSTALW.
08). This argues against a mechanism in which wolframin
mutations trigger a pathogenic mechanism involving both the
nuclear and the mitochondrial genome. Nonetheless, the clinical RT–PCR and RACE
signs of the disease favor a disease mechanism which involves a RT–PCR was performed using 1–2 ng of human or mouse adult
mitochondrial protein. This reasoning gains support from the brain cDNA (Clontech) as templates. RACE was done using the
finding of mitochondrial pathology in one of the patients (family Marathon cDNA Amplification Kit (Clontech). Primers designed
08), in whom we found two compound heterozygous wolframin from the predicted cDNA sequence were used for amplification
mutations. Wolframin is expressed in all tissues tested. Notably, of 0.2–2 kb products and are available on request. The probe
the band observed in the skeletal muscle lane was of low intensity
Downloaded from https://academic.oup.com/hmg/article/7/13/2021/634861 by guest on 14 October 2021
wm1E7-2-5 was obtained with the following primer pair:
compared with the high intensity band seen in heart muscle. It is wm1E7-2F, 5′-AGAACTTCCGCACCCTCAC-3′; wm1E7-5R,
known that expression levels of genes coding for mitochondrial 5′-GTAGAGGCAGCGCATCCAG-3′.
proteins do not solely reflect the number of mitochondria in a cell
or tissue. Secondary structure prediction revealed the signature of
a typical membrane protein: a central hydrophobic domain Northern blot analysis
flanked by two hydrophilic domains. The number of helical
domains crossing the membrane is predicted to be nine, The multiple tissue northern blot contained 2 µg poly(A)+ RNA
positioning the N- and C-termini of wolframin on different sides each from heart, brain, placenta, lung, liver, skeletal muscle,
of the membrane. The low score given for a mitochondrial kidney and pancreas (Clontech). Hybridization with probe
sublocalization is explained by the absence of a typical import wm1E7-2-5 (see RT–PCR above) was performed in Church
signal. With no significant homologies found in the databases, buffer at 65C; washing was with 0.01× SSC at 60C.
this defines a novel membrane protein, which may eventually
provide a new means for studying neurodegenerative processes.
Mutation analysis
Determination of its subcellular localization will reveal the first
clues to its functional role. mtDNA deletions were screened by Southern blotting and
The pleiotropic effects caused by wolframin mutations include mtDNA point mutations at positions 3460, 11778 and 14484
a spectrum of psychiatric disorders. In three patients with (LHON mutations) and 3243, 3271 and 3251 (MELAS muta-
mutations described in this study, psychiatric illness has been tions) were analyzed by RFLP methods as previously described
recorded. A locus for a dominant bipolar disorder has been (11). wolframin exons were amplified with intronic primers
mapped to 4p (21). These linkage studies were repeated in other (Table 2). Amplified fragments were analyzed by SSCP using
patient samples with bipolar disorder (22). The interval defined Hydrolink (AT Biochemical) gels at 20C with and without
by these studies includes markers which flank the wolframin glycerol. Staining was performed with VistraGreen and detection
gene. Combined with the observation that heterozygous carriers performed with a FluorImager (Molecular Dynamics). Variant
in recessive Wolfram syndrome families are predisposed to bands were reamplified and used for direct sequencing with both
psychiatric illness, haploinsufficiency of wolframin could predis- the sense and antisense primer using a Taq DyeDeoxy Terminator
pose to psychiatric illness (14). The gene structure presented here Cycle sequencing kit (ABI). Sequences were determined with an
provides the tools to test this hypothesis. Applied Biosystems 377 automated sequencer (accession nos
Y18064, AJ 011971).
MATERIALS AND METHODS
ACKNOWLEDGEMENTS
Patients
We are grateful to the patients and their families for their
participation in this study and the Stanford Human Genome
A summary of clinical details available from 12 families with
Center for their generation and open dissemination of human
Wolfram syndrome is given in Table 1. Histochemical analysis of
genomic sequence data. We thank W. Burger, M. Dreyer,
a muscle biopsy in patient 13073 (family 08) was compatible with
K.J. Eßer, H.J. Hartmann, W. Hecker, H. Muhle, R. Mühlenberg,
the diagnosis of ragged red fibres. The families were of German
K. Schlecht, W. Vorhoff, U. Wendel and D. Wenzel for sending
(families 3–7, 10 and 11), Turkish (families 1, 2 and 8),
DNA samples and D. Pongratz for histological analysis. We also
Yugoslavian (family 9) and Portuguese (family 12) descent.
thank K.B. Jedele for help in manuscript preparation and
J. Murken for his support. MRI scans were kindly provided by K.
Seelos, Department of Neuroradiology, Klinikum Grosshadern,
Sequence analysis LMU, München. This work was supported by the German
Federal Ministry for Education, Research and Technology
Homology searches against the non-redundant and EST data- (BMBF 01KW9605).
bases were performed using BLAST and FASTA. GENSCAN2028 Human Molecular Genetics, 1998, Vol. 7, No. 13
REFERENCES 11. Hofmann, S., Bezold, R., Jaksch, M., Obermaier-Kusser, B., Mertens, S.,
Kaufhold, P., Rabl, W., Hecker, W. and Gerbitz, K.D. (1997) Wolfram
(DIDMOAD) syndrome and Leber hereditary optic neuropathy (LHON) are
1. Wolfram, D.J. and Wagener, H.P. (1938) Diabetes mellitus and simple optic associated with distinct mitochondrial DNA haplotypes. Genomics, 39, 8–18.
atrophy among siblings: report of four cases. Mayo Clin. Proc., 1, 715–718. 12. Collier, D.A., Barrett, T.G., Curtis, D., Macleod, A., Arranz, M.J., Maassen,
2. Barrett, T.G., Bundey, S.E. and Macleod, A.F. (1995) Neurodegeneration and J.A. and Bundey, S. (1996) Linkage of Wolfram syndrome to chromosome
diabetes: UK nationwide study of Wolfram (DIDMOAD) syndrome. Lancet, 4p16.1 and evidence for heterogeneity. Am. J. Hum. Genet., 59, 855–863.
346, 1458–1463. 13. Polymeropoulos, M.H., Swift, R.G. and Swift, M. (1994) Linkage of the gene
3. Kinsley, B.T., Swift, M., Dumont, R.H. and Swift, R.G. (1995) Morbidity and for Wolfram syndrome to markers on the short arm of chromosome 4. Nature
Genet., 8, 95–97.
mortality in the Wolfram syndrome. Diabetes Care, 18, 1566–1570.
14. Swift, R.G., Polymeropoulos, M.H., Torres, R. and Swift, M. (1998)
4. Fraser, F.C. and Gunn, T. (1977) Diabetes mellitus, diabetes insipidus and
Predisposition of Wolfram syndrome heterozygotes to psychiatric illness.
optic atrophy. An autosomal recessive syndrome? J. Med. Genet., 14,
Mol. Psychiat., 3, 86–91.
190–193.
15. Burge, C. and Karlin, S. (1997) Prediction of complete gene structures in
5. Gunn, T., Bortolussi, R., Little, J.M., Andermann, F., Fraser, F.C. and
human genomic DNA. J. Mol. Biol., 268, 78–94.
Belmonte, M.M. (1976) Juvenile diabetes mellitus, optic atrophy, sensory
16. Uberbacher, E.C. and Mural, R.J. (1992) Locating protein-coding regions in
nerve deafness and diabetes insipidus—a syndrome. J. Pediat., 89, 565–570. human DNA-sequences by a multiple sensor neural network approach. Proc.
6. Barrientos, A., Volpini, V., Casademont, J., Genis, D., Manzanares, J.M., Natl Acad. Sci. USA, 88, 11261–11265.
Ferrer, I., Corral, J., Cardellach, F., Urbano-Marquez, A., Estivill, X. and 17. Rost, B., Casadio, R., Fariselli, P. and Sander, C. (1995) Transmembrane
Downloaded from https://academic.oup.com/hmg/article/7/13/2021/634861 by guest on 14 October 2021
Nunes, V. (1996) A nuclear defect in the 4p16 region predisposes to multiple helices predicted at 95% accuracy. Protein Sci., 4, 521–533.
mitochondrial DNA deletions in families with Wolfram syndrome. J. Clin. 18. Hofmann, K. and Stoffel, W. (1993) TMbase—a database of membrane
Invest., 97, 1570–1576. spanning protein segments. Biol. Chem. Hoppe-Seyler, 347, 166.
7. Rötig, A., Cormier, V., Chatelain, P., Francois, R., Saudubray, J.M., Rustin, P. 19. Nakai, K. and Kanehisa, M. (1992) A knowledge base for predicting protein
and Munnich, A. (1993) Deletion of mitochondrial DNA in a case of localization sites in eukaryotic cells. Genomics, 14, 897–911.
early-onset diabetes mellitus, optic atrophy and deafness (Wolfram syn- 20. Akiyama, N., Shima, H., Hatano, Y., Osawa, Y., Sugimura, T. and Nagao, M.
drome, MIM 222300). J. Clin. Invest., 91, 1095–1098. (1995) cDNA cloning of BR gamma, a novel brain-specific isoform of the B
8. Bundey, S., Poulton, K., Whitwell, H., Curtis, E., Brown, I.A. and Fielder, regulatory subunit of type-2A protein phosphatase. Eur. J. Biochem., 230,
A.R. (1992) Mitochondrial abnormalities in the DIDMOAD syndrome. J. 766–772.
Inherit. Metab. Dis., 15, 315–319. 21. Blackwood, D.H., He, L., Morris, S.W., McLean, A., Whitton, C., Thomson,
9. Seyrantepe, V., Topaloglu, H., Simsek, E., Ozguc, M. and Yordam, N. (1996) M., Walker, M.T., Woodburn, K., Sharp, C.M., Wright, A.F., Shibasaki, Y., St
Mitochondrial DNA studies in Wolfram (DIDMOAD) syndrome. Lancet, Clair, D.M., Porteous, D.J. and Muir, W.J. (1996) A locus for bipolar affective
347, 695–696. disorder on chromosome 4p. Nature Genet., 12, 427–430.
10. Jackson, M.J., Bindoff, L.A., Weber, K., Wilson, J.N., Ince, P., Alberti, K.G. 22. Kennedy, J.L. and Macciaroli, M. (1998) Chromosome 4 workshop.
and Turnbull, D.M. (1994) Biochemical and molecular studies of mitochon- Psychiat. Genet., 8, 67–71.
drial function in diabetes insipidus, diabetes mellitus, optic atrophy and 23. Harris, N.L. (1997) Genotator: a workbench for sequence annotation.
deafness. Diabetes Care, 17, 728–733. Genome Res., 7, 754–762.You can also read
