 1998 Oxford University Press
Human Molecular Genetics, 1998, Vol. 7, No. 2
177–186
Molecular genetic analysis of autosomal dominant
cerebellar ataxia with retinal degeneration (ADCA type II)
caused by CAG triplet repeat expansion
Jurgen Del-Favero, Luc Krols, Andrej Michalik, Jessie Theuns, Ann Löfgren,
Dirk Goossens, Anita Wehnert, Dirk Van den Bossche, Karel Van Zand,
Hubert Backhovens, Nicole van Regenmorter1, Jean-Jacques Martin2 and
Christine Van Broeckhoven*
Laboratory of Neurogenetics, Flanders Interuniversity Institute for Biotechnology, Born-Bunge Foundation, University of
Antwerp, Department of Biochemistry, Universiteitsplein 1, B-2610 Antwerpen, Belgium, 1Centre de Génétique,
Free University of Brussels, Campus Erasme, 808 Route de Lennik, B-1070 Brussels, Belgium and 2Laboratory of
Neuropathology, Born-Bunge Foundation, University of Antwerp, Department of Medicine, Universiteitsplein 1, B-2610
Antwerpen, Belgium
Received September 30, 1997; Revised and Accepted November 12, 1997
Autosomal dominant cerebellar ataxia with retinal
degeneration (ADCAII) was previously mapped by
linkage analysis studies to chromosome 3p12–p21.1
(SCA7). Positional cloning efforts have recently identified a novel gene, SCA7, containing a translated CAG
repeat, expanded in SCA7 patients. We cloned the
SCA7 gene from a yeast artificial chromosome (YAC)
clone contig spanning the SCA7 candidate region.
Using a combination of genomic sequencing and
cosmid-based exon trapping, two expressed sequence
tags were identified. Sequencing of the corresponding
cDNA clones and RT-PCR analysis identified the
full-length SCA7 cDNA. Together, our sequence data
defined the intron/exon boundaries of the first two
coding exons of the SCA7 gene, with the first exon
containing the expanded CAG repeat. Further, sequence
comparison with the published SCA7 cDNA identified
one additional putative exon in the 5′-UTR region of the
SCA7 gene. The SCA7 gene was mapped on the YAC
contig in the 2.5 cM interval between D3S1600 and
D3S1287. In one extended Belgian SCA7 pedigree the
expanded alleles ranged from 38 to at least 55 repeats
with allele lengths being inversely correlated with
onset age of ADCAII symptoms. The SCA7 repeats
increased in length in successive generations. Normal
alleles had from four to 18 repeats, with 10 repeats
being the most common allele.
INTRODUCTION
The autosomal dominant cerebellar ataxias (ADCA) are a heterogeneous group of neurodegenerative disorders characterized by
progressive degeneration of the cerebellum, brain stem and spinal
cord (1). Clinically, ADCA has been divided into three groups:
ADCA types I–III (1). In ADCA type I (ADCAI) progressive
cerebellar ataxia is associated with ophthalmoplegia, optic
atrophy, extrapyramidal signs and dementia. ADCA type II
(ADCAII) always presents with pigmentary macular dystrophy
but is variably associated with ophthalmoplegia, dementia or
extrapyramidal signs (2,3). ADCA type III (ADCAIII) is often
referred to as the ‘pure’ cerebellar syndrome.
ADCAI is genetically heterogeneous, with five genetic loci,
designated spinocerebellar ataxia (SCA) 1, 2, 3, 4 and 6, being
assigned to five different chromosomes. Four ADCAI genes
(SCA1, 2, 3 and 6) have been cloned and shown to contain an
expanded CAG repeat in their coding sequence producing an
elongated polyglutamine tract in the corresponding protein (4–7).
In each case the expanded repeats are variable in size. Also, the
expanded repeats are unstable and usually increase in size when
transmitted to successive generations, explaining the anticipation
phenomenon observed in ADCAI families. All ADCAI disorders
share common features, i.e. a negative correlation between
number of repeats and onset age and the appearance of symptoms
above a certain repeat length specific for each disease gene.
However, their pathophysiology differs depending on the
location of the repeat in the gene. In contrast to ADCAI, both
ADCAII and ADCAIII are most likely homogeneous disorders.
The pure form ADCAIII was assigned to chromosome 11p–q11
(SCA5) but the corresponding gene has not yet been cloned (8).
Three independent studies mapped ADCAII to chromosome
3p12–p21.1 (SCA7) (3,9,10). More recent linkage studies in
ADCAII families refined the SCA region significantly (11,12).
Informative recombinants defined a candidate region of 9.3 cM
between D3S1312 and D3S3635 (11). We reduced the SCA7
region to the 12 cM interval between D3S1300 and D3S1285
based on informative recombinants in an extended Belgian
pedigree (12). Recently the SCA7 gene was cloned and shown to
*To whom correspondence should be addressed. Tel: +32 3 820 26 01; Fax: +32 3 820 25 41; Email: [email protected]
172
Human Molecular Genetics, 1998, Vol. 7, No. 2
Figure 1. Genetic and physical mapping of the SCA7 region. The SCA7 candidate regions based on informative recombinants in family CA-1 are indicated: A, Krols
et al. (12); B, this study. The distances between STR markers in cM correspond to the sex-average genetic map (16). The CEPH megaYACs in the YAC contig spanning
the SCA7 candidate region between D3S1300 and D3S1285 are represented by horizontal lines. The positive PCR results obtained on the YACs with STR, STS and
EST markers are represented by the following symbols: circles, STR markers; triangles, ESTs; squares, STSs. Negative PCR results suggesting internal deletions in
the YACs are shown with brackets. The physical location of the SCA7 gene is indicated.
contain an expanded CAG repeat in SCA7-linked ADCAII
families (13). The SCA7 protein was designated ataxin-7, a
protein of 892 amino acids with unknown functions.
RESULTS
DNA diagnosis in family CA-1
Subsequent to the genetic fine mapping of SCA7 to the 12 cM
region between D3S1300 and D3S1285 (12; Fig. 1), DNA
diagnosis of ADCAII based on haplotype analysis was included
in the genetic counseling program of members of the Belgian
SCA7 family CA-1 (14). In total 16 at-risk individuals had DNA
diagnosis of SCA7 and their DNA was analyzed with 12 simple
tandem repeat (STR) markers located between D3S1300 and
D3S1285, including four STR markers that we had not used
before: D3S3631, D3S3571, D3S1228 and D3S3644 (Fig. 2). All
individuals had a complete clinical work-up, which was done
blind from the DNA data. Three individuals had been diagnosed
previously as mildly affected (VI-27) or probably affected (V-23
and VI-29) (12,14). Their clinical work-up confirmed that they
are very mild (V-23) to mildly affected (VI-27 and VI-29), which
Figure 2. Partial pedigree of Belgian SCA7 family CA-1. The detailed clinical and pathological characteristics of family CA-1 have been described in Martin et al.
(14). Filled circles and squares are very mild to severely affected patients segregating the SCA7 repeat that had an extensive clinical follow-up examination (12,14;
this study). Shaded symbols are individuals that are obligate carriers of the SCA7 mutation but had no clinical examination. The individuals in the pedigree are numbered
as in Martin et al. (14). The reconstructed haplotype for the STR markers between D3S1300 and D3S1285 are given below the haplotype, with the disease haplotype
boxed. The expanded SCA7 CAG repeat is given in number of CAG repeats. Informative recombinants are indicated with arrowheads. Onset ages in years are shown
below the disease haplotype except for the asymptomatic mutation carriers in generation VI, where age at clinical examination is indicated. Onset ages in brackets
have been arbitrarily fixed at age of clinical examination. An asterix indicates patients that had autopsy confirmation. For reasons of confidentiality, haplotypes are
given only for patients and at-risk individuals that had a complete clinical examination and genetic counseling.
173
Human
Genetics,
1998,
7, No.
NucleicMolecular
Acids Research,
1994,
Vol. Vol.
22, No.
1 2
173
174
Human Molecular Genetics, 1998, Vol. 7, No. 2
is in agreement with them carrying the risk haplotype (Fig. 2). Of
the eight individuals that showed no symptoms, five did not
segregate the risk haplotype (V-17, V-25, V-27, V-30 and VI-34)
while the other three did (VI-31, VI-35 and VI-38). The latter
three are young individuals aged 34, 28 and 24 years respectively
at examination. Clinical examination of five other at-risk
individuals showed that they are very mildly (V-26 and VI-33) to
mildly (V-19, V-22 and V-24) affected and all five carry the at-risk
haplotype.
Analysis of the haplotypes showed that the recombinant
detected by D3S1285 in patient VI-27 (12) is also seen with
D3S3644, reducing the candidate region further (Fig. 2). Also, a
recombination has occurred between D3S1600 and D3S1287 in
the asymptomatic individual V-30, mapping the SCA7 gene
centromeric of D3S1600 (Fig. 2). Individual V-30 was 63 years
old at examination; she had no complaints, a normal neurological
examination and normal laboratory tests.
Physical mapping of the SCA7 region
DNA from 22 yeast artificial chromosomes (YACs) selected for the
region contained between D3S1300 and D3S1285 was tested by
PCR amplification for the presence of the STR markers used in the
genetic analysis of family CA-1 as well as for sequence tagged sites
(STSs) and expressed sequence tags (ESTs) previously located in
this region (15). A total of 23 markers were used and the YACs were
ordered into a clone contig based on their marker content (Fig. 1).
The order of the markers is in agreement with previously published
YAC maps (11,15). However, our YAC contig map contains an
additional four markers D3S2984, D3S3424, D3S4010 and
D3S1228. With the exception of one marker, D3S3631, the order of
the STR markers is also in agreement with the genetic map of this
region (16). The location of D3S3631 between D3S1300 and
D3S1312 has also been reported by David et al. (11).
From the YAC contig we selected two overlapping YACs, 882D9
and 830A12, for gene cloning experiments, since together they
spanned the candidate region defined by the informative
recombinants with D3S1312 and D3S3635 (11; Fig. 1). The size of
the two selected YACs was determined by pulsed field gel
electrophoresis. The size of YAC 830A12 was estimated at 1.5 Mb,
which is much larger than the reported 660 kb (17). Also, in 882D9
two YACs were detected of 1.6 Mb and 840 kb respectively, which
are most likely the result of co-cloning. Later we showed by
hybridization that the SCA7 gene is contained within the 1.6 Mb
YAC clone (data not shown).
Isolation of CAG-positive cosmids
Yeast DNA was prepared from both YACs 882D9 and 830A12,
partially digested with Sau3AI and subcloned in the BamHI site of
the exon trapping cosmid vector sCOGH6 (18). Cosmid clones
containing human DNA inserts were identified by colony
hybridization using the Alu consensus sequence (19). Alu-positive
cosmid clones were subsequently hybridized with a radiolabeled
(CTG)10 oligonucleotide. This resulted in selection of 41 Alu/CAGcontaining cosmids clones, of which 33 were derived from YAC
882D9 and eight from YAC 830A12. Cosmid DNA was prepared,
digested with Sau3AI, separated on agarose gel, blotted and
hybridized with the radiolabeled (CTG)10 oligonucleotide. Eight
cosmids were selected for further study since they contained
different sized CAG-positive Sau3AI fragments. The selected
cosmids were Sau3AI digested, plasmid subcloned and CAGpositive plasmids were selected by colony hybridization with the
radiolabeled (CTG)10 oligonucleotide. Sequencing showed that
in four plasmids the CAG-containing sequence was identical to
a yeast sequence. The four remaining plasmids contained human
DNA sequences. The sequence of the 600 bp Sau3AI insert
fragment of one of these plasmids (cos14-1) showed homology
with the (CAG)5CTGCAG-containing mRNA coding for the
human FAN protein (GenBank accession no. X96586). All three
remaining plasmids (cos5-10, cos18-3 and cos21-4) contained a
1.8 kb Sau3AI fragment; however, one plasmid contained an
additional Sau3AI fragment of 1.4 kb, probably due to incomplete
digestion of the cosmid DNA during plasmid subcloning
(cos18-3). Sequencing showed that both ends of the 1.8 kb
fragment are identical in cos5-10 and cos21-4 while only one end
is identical in cos18-3. The complete sequence of the 1.8 kb
Sau3AI fragment of cos21-4 was determined by transposonbased sequencing. Together with an internal 350 bp sequence of
the partial 1.4 kb Sau3AI fragment in cos18-3, obtained by
transposon-based sequencing, a genomic sequence of 2.18 kb was
obtained containing an uninterrupted CAG repeat with 10 CAG
units surrounded by a highly CG-rich sequence (Fig. 3A)
(GenBank accession no. AF032102). Hybridization of the
gridded 41 Alu/CAG-positive cosmids with the cos14-1 and
cos21-4 plasmid inserts identified additional positive cosmids. In
total 14 cosmid clones were positive for cos21-4 (all subclones
from YAC 882D9), while cos14-1 recognized seven positive
clones (five subclones from YAC 882D9 and two from 830A12).
Identification of the SCA7 CAG repeat
To test whether one of the two CAG repeats identified was expanded
in ADCAII patients we designed flanking primer pairs and PCR
amplified genomic DNA of patients from family CA-1. PCR
amplification of the cos14-1 CAG repeat produced only the
expected PCR fragment (data not shown). PCR amplification of the
cos21-4 CAG repeat from genomic DNA using primers H1 and H2
followed by agarose gel electrophoresis and hybridization with the
radiolabeled (CTG)10 oligonucleotide showed not only the expected
PCR fragment of 279 bp but also a longer PCR fragment that
differed in size between patients (Fig. 4A). These data suggested that
the CAG repeat in cos21-4 is expanded in SCA7 patients. To test this
hypothesis we developed a fluorescence-based PCR test allowing
Figure 3. The SCA7 gene. (A) Genomic sequence of the 2.18 kb fragment containing the SCA7 CAG repeat. Exonic sequences are indicated with capital letters, the
CAG repeat and ATG start codon in exon 1 of the SCA7 gene are boxed, the sequences of primers H1, H2 and 1220R are indicated with arrows and the NotI (N) and
Sau3AI (S) restriction sites are underlined. The two regions that show sequence similarities to known ESTs are underlined: region 1 (positions 47–542) is homologous
to ESTs H41756, H40290 and H40285; region 2 (positions 1829–1898) to EST AA099235. (B) Schematic representation of the SCA7 cDNA. cDNA clones 258172
and 510718 correspond to ESTs N30892 and AA099235 respectively. The RT-PCR product was obtained by RT-PCR amplification of lymphoblast RNA with primers
H2 and 1220R. The horizontal lines represent the sequences of the cDNA clones and the RT-PCR fragment containing the SCA7 CAG repeat drawn to scale. The SCA7
cDNA consensus sequence is 3465 bp long. The additional 98 bp sequence in the 5′-UTR, absent from the published SCA7 cDNA (GenBank accession no. AJ000517).
175
Human
Genetics,
1998,
7, No.
NucleicMolecular
Acids Research,
1994,
Vol. Vol.
22, No.
1 2
175
176
Human Molecular Genetics, 1998, Vol. 7, No. 2
Based on these results, we concluded that the cos21-4 CAG repeat
was most likely the SCA7 repeat that when expanded causes
ADCAII. To confirm this we PCR amplified with primer set H1 and
H2 the cos21-4 CAG repeat from cDNA, synthesized with reverse
transcriptase (RT) from RNA extracted from cultured lymphoblasts,
of two SCA7 patients of family CA-1. The PCR fragments were
separated on agarose gel and the longer PCR fragment was plasmid
subcloned and sequenced. The sequences revealed that the increase
in length of the PCR fragment was the sole result of an increase in
number of CAG repeats to 43 in patient VI.-27 and 39 in patient
VI.-29 (Fig. 2), confirming that the cos21-4 (CAG)10 is the SCA7
CAG repeat. The latter experiment also indicated that the SCA7
CAG repeat could be located in the coding region or in the
untranslated region of the SCA7 gene, since the CAG repeat
fragment was amplified from cDNA.
Isolation of SCA7 cDNA
Figure 4. Amplification of the SCA7 CAG repeat. (A) Genomic DNA of
members of family CA-1 was amplified with primers H1 and H2 (Fig. 3) and
PCR fragments separated by 2% agarose gel electrophoresis, blotted and
hybridized with radiolabeled oligonucleotide (CTG)10. Numbering of the
individuals in the pedigree is as in Figure 2. (B) Genomic DNA of CA-1 patient
VI-35 (Fig. 2) was PCR amplified with primers H1 and H2, with primer H2
labeled with the fluorescent dye JOE (Applied Biosystems). PCR fragments
were separated on a 6% polyacrylamide sequencing gel and analyzed with
GENESCAN 672 software on an ABI 373A automated sequencer (Applied
Biosystems). The lengths of the normal and expanded SCA7 alleles were
estimated at 279 and 378 bp respectively.
analysis of CAG repeat fragments on an ABI automated sequencer
equipped with GENESCAN software for size determination of the
PCR fragments (Fig. 4B). First, we determined the polymorphic
character of the CAG repeat in 50 control individuals. The size
differences between polymorphic fragments were interpreted in
terms of number of triplet repeats and were shown to vary from 4
to 18 CAG repeats, with 10 CAG repeats in the most common allele
(80%). Next, we analyzed segregation of the CAG repeat in the
complete CA-1 family (Fig. 2). In patients and carriers of the atrisk haplotype the number of CAG repeats varied from 38 to 55. In
one patient, VI-85, with juvenile onset of ADCAII the exact size
could not be determined, but was minimally 55 repeats. Furthermore, we observed a high degree of mosaicism in all expanded
alleles from the analyzed patients (Fig. 4B). The location of the
CAG repeat in the SCA7 candidate region was determined by PCR
amplification of YACs from the YAC contig (Fig. 1). The CAG
repeat mapped between markers D3S1600 and D3S1287, a location
that is in agreement with the genetic mapping data obtained in family
CA-1, mapping SCA7 centromeric of D3S1600.
The genomic sequence of the 2.18 kb fragment containing the
SCA7 CAG repeat was subjected to a BLASTN analysis to
identify homologous clones present in databases. Strong
homology was obtained between the SCA7 CAG repeat and
surrounding sequences (positions 1348–1581) in the 2.18 kb
fragment (Fig. 3) with the NotI jumping clone J32Z180D,
previously mapped to chromosome 3 (GenBank accession no.
X95831). Also, two other regions within the 2.18 kb fragment
showed significant homologies with ESTs: region 1 upstream of
the CAG repeat with ESTs H41756, H40290 and H40285; region
2 downstream of the CAG repeat with EST AA099235 (Fig. 3A).
Analysis of the 2.18 kb genomic sequence using the GRAIL
program (20) predicted two putative exons of 324 and 69 bp
respectively containing respectively the SCA7 CAG repeat and
EST AA099235. Subsequent RT-PCR amplification of lymphoblast cDNA using primers H2 and 1220R (Fig. 3) produced a
fragment of 404 bp. The cDNA clone 510718, corresponding to
EST AA099235, was obtained from the UK HGMP resource
centre. The cDNA had been isolated from a colon cDNA library
(Stratagene z178a12r1). Restriction digestion identified an insert
of 2.5 kb that was sequenced by transposon-based sequencing
(Fig. 3B). The total length of the 510718 cDNA was 2461 bp,
containing 137 bp of 3′-UTR but no CAG repeat (GenBank
accession no. AF032103).
Exon trapping of SCA7 cosmids
The 14 cosmids containing the SCA7 CAG repeat cloned in the
exon trapping cosmid vector sCOGH6 (18) were used in an exon
trapping experiment. Trapped products were obtained for seven
cosmids and sequencing analysis revealed that two of them
contained an identical 98 bp fragment which upon BLASTN
analysis identified one EST, N30892, corresponding to a
placental cDNA clone 258172. The other five trapped products
were unique and ranged from 28 to 476 bp but did not identify
homologies with known expressed sequences. The cDNA clone
258172 was obtained from the UK HGMP resource center and
the 750 bp insert was sequenced (GenBank accession no.
AF032104). Sequence comparison identified an 89 bp overlap
between clone 258172 and the SCA7 CAG exon (Fig. 3). Since
the placenta cDNA library was cloned after NotI digestion, clone
258172 ended at the NotI site located 10 bp upstream of the SCA7
CAG repeat (Fig. 3), the same NotI site that is present in the NotI
jumping clone J32Z180D. The trapped exon is located in the
177
Human
Genetics,
1998,
7, No.
NucleicMolecular
Acids Research,
1994,
Vol. Vol.
22, No.
1 2
3′-region of clone 258172, immediately upstream of the SCA7
CAG exon (Fig. 3B).
The sequences of cDNA clones 258172 and 510718 and of the
RT-PCR products containing the SCA7 CAG repeat (Fig. 3B) were
used to put together a consensus sequence for the SCA7 cDNA of
3465 bp, comprising an open reading frame (ORF) of 2727 bp, a
5′-UTR of 652 bp and a 3′-UTR of 137 bp. The SCA7 cDNA
sequence was verified by RT-PCR sequencing of lymphoblast
cDNA using four overlapping primer sets (GenBank accession no.
AF032105).
DISCUSSION
Several lines of evidence had indicated that ADCAII is caused by
a CAG triplet repeat expansion (3,14,21–23). This was confirmed
by recent cloning of the SCA7 gene, containing an expanded CAG
repeat in the 5′-coding region (13). We independently cloned the
SCA7 gene using a combination of genomic sequencing, cosmid
based exon trapping, cDNA sequencing and RT-PCR analysis.
We subcloned two overlapping YACs from the SCA7 region in
a cosmid-based exon trap vector, sCOGH6. Human CAG-positive
cosmids were selected and cosmids containing the SCA7 CAG
repeat were used in exon trapping experiments. Also, the genomic
sequence of a 2.18 kb CAG-containing fragment was determined by
transposon-based sequencing. BLASTN analysis of the genomic
and trapped sequences identified two ESTs, one colon cDNA clone
510718 and one placental cDNA clone 258172. Sequencing of the
two cDNA clones together with RT-PCR sequencing of the SCA7
CAG repeat defined a consensus SCA7 cDNA of 3465 bp, with the
CAG repeat located in an ORF of 2727 bp. The exon containing the
SCA7 repeat is the first coding exon of the SCA7 gene and the SCA7
CAG repeat is translated into a polyglutamine stretch ranging from
38 to at least 55 amino acids in ADCAII patients of the Belgian
SCA7 family CA-1 (12,14).
Sequence comparison with the published lymphoblast SCA7
cDNA sequence (13) showed that the coding sequence in our
SCA7 cDNA was identical with the exception of 15 nt at the
3′-end of the ORF, which result in five different amino acids
(VGNGL instead of PKARP). Also, the published 3′-UTR
sequence is 734 bp long and completely different from the 137 bp
sequence we obtained from colon cDNA clone 510718. However,
sequencing of a 3′-RACE product obtained from lymphoblast
RNA and of a fetal liver/spleen EST, H78392, identified by
BLASTN analysis of our SCA7 cDNA consensus sequence,
provided the same 137 bp 3′-UTR. Further, the 5′-UTR sequence
contained within placental cDNA clone 258172 is 98 bp longer
than in the published lymphoblast SCA7 cDNA (13). RT-PCR
sequencing of lymphoblast cDNA confirmed the published
5′-UTR, showing the absence of this 98 bp exon and suggesting
alternative splicing of 5′-UTR exons. From the position of the
trapped 5′-UTR exon and the putative alternative spliced 5′-UTR
exon we predict that the 5′-UTR of the SCA7 gene contains at
least three exons. However, alternative splicing in the 5′-UTR of
SCA7 will have to be demonstrated by analyzing 5′-UTR
sequences by RT-PCR analysis of RNA obtained from different
tissue or cell sources or by sequencing cDNA clones isolated from
tissue-specific cDNA libraries.
The rationale behind the use of the sCOGH6 cosmid vector in
our cloning experiments was that we could use the CAG-positive
cosmids immediately in exon trapping experiments. Trapped
exons would help us in the identification of the SCA7 gene since
177
they can be used in screening cDNA libraries or EST databases.
Although we were successful in trapping the 5′-UTR exon
immediately upstream of the SCA7 CAG exon from two
CAG-positive cosmids, our experiences suggest that there are a
number of limitations to the use of the sCOGH6 trapping vector.
First, although all selected cosmids were shown to contain the
1.8 kb fragment and thus had to contain the first exon of the SCA7
gene with the CAG repeat, this exon was never trapped. Also, the
SCA7 CAG exon was not trapped together with the adjacent
5′-UTR exon from the two cosmids that were successful in the
trapping experiment. However, the sequences of the intron/exon
boundaries of the SCA7 CAG exon confirm the consensus splice
site sequences. This observation is of particular importance if a
similar strategy is followed in cloning CAG repeat-associated
genes in other diseases, such as bipolar affective disorder (24) or
spastic paraplegia (25). Also, the availability of the sCOGH6
cosmid sublibrary of YAC 882D9 was expected to be helpful in
determining the exon/intron boundaries of the SCA7 gene by exon
trapping of SCA7 cDNA-positive cosmids. We screened filters of
the cosmid library with cDNA 510718 containing the ORF of the
SCA7 cDNA and identified 10 positive cosmid clones, all
different from our initial set of CAG-positive cosmids. However,
analysis of these SCA7 cosmids showed that several of them had
partially or completely lost their insert DNA, indicating that
recombinant sCOGH6 clones are unstable. Instability had not
been observed in previous experiments using an earlier version of
the sCOGH cosmid vector, sCOGH2 (18,26). sCOGH6 is a
derivative of sCOGH2 and was obtained by removal of the SVNeo
gene and mutagenesis of the EcoRI site, upstream of the
metallothionein-1 promoter. What exactly is causing the
instability of recombinant clones in sCOGH6 is unclear.
The SCA7 gene was genetically localized in the 5 cM interval
between D3S1600 and D3S3635 (13). We were able to refine the
location of the SCA7 gene to the 2.5 cM interval between
D3S1600 and D3S1287 (Fig. 1). This location is compatible with
our previous mapping data (12), as well as with the observation
in this study of an informative recombinant in one unaffected
individual with D3S1600, mapping SCA7 centromeric of this
marker (Fig. 2). When the SCA7 repeat was identified the
recombinant individual was shown not to carry an expanded CAG
repeat, mapping the recombinant event between D3S1600 and the
SCA7 gene.
We developed a fluorescence-based PCR test for the SCA7
repeat allowing its analysis on an automated sequencer, in our
case an ABI 373A system (Applied Biosystems). In the normal
population the SCA7 CAG repeat ranged from 4 to 18 repeats,
with 10 repeats as the most common allele (80%) (n = 50). In
family CA-1 the SCA7 CAG repeat ranged from 38 to 55 repeats,
with onset ages ranging from 20 to 65 years and inversely
correlated with repeat length (Fig. 2). In one patient (VI-85) with
an onset age of 3 years the SCA7 repeat was 55 repeats or longer,
but its exact length could not be determined because of the low
quality of the DNA sample. The range of SCA7 CAG repeat
lengths and of onset ages is similar to that described by David et
al. (13), who also found the majority of their patients having
repeat lengths between 35 and 55, with only a few juvenile onset
cases having longer repeats. Also, we observed somatic mosaicism for each expanded allele analyzed from leukocyte DNA. In
seven affected parent/child transmissions in family CA-1 the
mean increase in repeat length was 4.25 repeats (range 0–15
repeats). In four affected parent/child transmissions onset age
178
Human Molecular Genetics, 1998, Vol. 7, No. 2
decreased from a mean onset age of 60 years (range 56–65 years)
in the parents to 31.2 years (range 21–36 years) in the affected
children (VI-27, VI-29, VI-33 and V-21). The large decrease in
onset age between two successive generations is hard to reconcile
with the rather small increase in repeat length. However, in three
of the four affected parent/child transmissions the onset age had
been arbitrarily set at age at examination (VI-27, VI-29 and
VI-30). These individuals received a complete clinical work-up
within the framework of genetic counseling and were diagnosed
with very mild to mild symptoms of cerebellar ataxia and retinal
lesions in the absence of opthalmoplegia, decreased vibration
sense and pyramidal signs. This active clinical evaluation at a
young age may have biased onset age estimation in these children.
On the other hand, all four affected parent/child transmissions
were of paternal origin and it was previously shown that
anticipation in SCA7 families is more pronounced in paternal
transmission of ADCAII (3,14,27). Further, the large repeat in the
juvenile case (VI-85) is also of paternal origin; however, we have
not yet been able to analyse her father’s DNA. Interestingly, the
increment in repeat length (+15) is larger in this branch of the
pedigree, suggesting that here other genetic factors contribute to
gonadal mosaicism or that certain repeat lengths are more prone
to expansion than others.
The SCA7 gene is a novel gene coding for a protein, ataxin-7, of
unknown function that is ubiquitously expressed as a 7.5 kb
transcript (13). Also, we detected a 7.5 kb transcript in various
tissues of the brain upon Northern blot hybridization with colon
cDNA clone 510718 containing the SCA7 cDNA ORF (data not
shown). It was suggested, based on its amino acid composition and
its nuclear localization (21,22), that ataxin-7 might be a potential
transcription factor (13). However, more work, including functional
studies in cellular and animal models, are needed to unravel the
normal and pathological function of this new protein. Meanwhile,
the availability of an easy test for SCA7 repeat expansion will be
very helpful in differential diagnosis of ADCA patients and in
delineating the ADCAII phenotype.
primer was chemically labeled with the fluorescent dye FAM or
JOE (Applied Biosystems) and alleles were separated on a 6%
polyacrylamide sequencing gel. The polymorphic alleles were
analyzed on a model ABI 373A sequencer using GENESCAN
672 software (Applied Biosystems, Foster City, CA).
YAC contig mapping
A total of 22 CEPH YACs were selected for the SCA7 candidate
region, contained between D3S1300 and D3S1285, from three
different databases (Baylor College of Medicine Genome center,
http://kiwi.imgen.bcm.tmc.edu: 8088/; Whitehead Institute for Biomedical Research, http://www.genome.wi.mit.edu/; the Généthon
server, http://www.genethon.fr/genethon_en.html/) based on their
marker content. YACs were obtained from CEPH (Paris, France)
and the YAC Screening Center Leiden (Leiden, The Netherlands).
The YAC clones were plated on selective agar plates and a single
colony from each plate was grown in selective liquid culture
medium for YAC DNA extraction using standard procedures.
YAC DNA was PCR amplified with the 12 STR markers used
for the genetic analysis, D3S1239 and the trinucleotide repeat
marker D3S4010. Further screening was performed with five
STSs (D3S2984, D3S3424, D3S4003, D3S3395, D3S3068 and
GATA-61E08) and with three ESTs (D3S3934, D3S3119 and
D3S3913). Published primers were used in a standard PCR
reaction. PCR amplifications were done in the presence of
positive (human genomic DNA) and negative (water) controls
and each marker was tested at least twice for each YAC/marker
combination. YAC sizes were determined by pulsed field gel
electrophoresis from embedded total yeast DNA. Pulsed field gel
electrophoresis analysis was carried out using the CHEF Mapper
XA apparatus (BioRad). Conditions for optimal separations were as
determined by the embedded algorithm. YACs were visualized by
ethidium bromide staining and their size was estimated using the
chromosomes of Saccharomyces cerevisae YP148 as a size marker.
YACs were also visualized after Southern blot hybridization with a
radiolabeled Alu consensus probe or the SCA7 CAG repeat fragment
of cos21-4.
MATERIALS AND METHODS
Genetic and clinical studies
Family CA-1 is an extended SCA7 family of which the clinical
and pathological characteristics have been described in detail
(12,14). Following our more recent refinement of the SCA7
locus, DNA diagnosis was offered by the genetic counselor to
affected and at-risk family members of family CA-1. Risk
haplotypes were reconstructed using 12 STR markers located
between D3S1300 and D3S1285. A counseling program similar
to that used for Huntington’s disease was followed. In addition,
each individual who requested a DNA diagnosis received an
extensive neurological examination followed by laboratory
work-up, including brain-evoked potentials, electroretinography,
visually evoked potentials, ophthalmological examination and
magnetic resonance imaging (MRI). The clinical examination
was done blind from the DNA analysis.
After informed consent, genomic DNA was isolated from
whole blood and the STR markers were analyzed using published
primers in PCR amplification. The PCR reaction was carried out
in a total volume of 25 µl containing ∼100 ng genomic DNA,
25 pmol each primer, 1.25 mM dNTPs, 20 mM Tris–HCl, pH 8.3,
50 mM KCl, 1.5 mM MgCl2 and 1 U Taq DNA polymerase. One
Subcloning in sCOGH6
High molecular weight yeast DNA embedded in 1% low melting
point agarose was digested with the restriction enzyme Sau3AI
using optimal conditions for partial digestion (28) after overnight
incubation of the enzyme/DNA mixture on ice, generating
fragments in the size range 20–100 kb. Yeast plugs were
subsequently treated with agarase (Epicentre Technologies)
according to the manufacturer’s protocol. The partially digested
yeast DNA was dephosphorylated with 1.5 U HK phosphatase
(Epicentre Technologies)/µg DNA for 2 h followed by heat
inactivation of the phosphatase for 20 min at 65C. Next the DNA
was ligated in a 1:10 molar ratio to sCOGH6 vector, digested with
XbaI, dephosphorylated and digested with BamHI to generate the
vector arms and the cloning site (17). After ligation the cosmid
DNA was packaged using Gigapack gold II packaging extract
(Stratagene) and recombinant cosmids were transfected into
XL1-BlueMR (Stratagene).
Identification of CAG-positive clones
For each YAC ∼10 000 cosmids were plated, representing a
15–20 times coverage of the yeast genome. Cosmid clones
179
Human
Genetics,
1998,
7, No.
NucleicMolecular
Acids Research,
1994,
Vol. Vol.
22, No.
1 2
containing human inserts were identified by colony hybridization
with a radiolabeled Alu consensus probe and arrayed in 96-well
plates. Gridded replica filters were prepared on Hybond-N+
membranes (Amersham) and were hybridized with [γ-32P]ATP
end-labeled (CTG)10 oligonucleotide. Cosmid DNA was isolated
using the alkaline extraction method. Cosmid DNA was digested
with NotI to determine insert size by agarose gel electrophoresis.
The pGem3zf(–) vector (Promega) was digested with BamHI to
completion and subsequently dephosphorylated with HK
phosphatase (Epicentre Technologies) and purified. Cosmid
DNA was digested with 2 U Sau3A I/µg DNA for 4 h and purified
by phenol/chloroform extraction and ethanol precipitation.
Sau3AI-digested cosmid DNA was ligated overnight to the
BamHI-digested vector with 1 U ligase (Boehringer). The ligated
DNA was dialyzed against 1× TE on VS 0.025 µm filters
(catalogue no. VSWP 02500, Millipore) for at least 2 h. Aliquots
of 2 µl dialyzed DNA were used to electroporate electrocompetent XL1-Blue cells, prepared according to Sheng et al.
(29). Transformants were plated on ampicilin agar plates, prepared
for colony hybridization and hybridized with [γ-32P]ATP endlabeled (CAG)10 oligonucleotide. Hybridizations were done overnight in 7% SDS, 0.5 M Na2HPO4/NaH2PO4, pH 7.2, 1 mM EDTA.
Hybridization temperatures were 58C for the γ-32P-labeled
(CAG)10 oligonucleotide and 65C for the [α-32P]-labeled
probes. Filters were washed in 5× SSC (1× SSC is 0.15 M sodium
chloride, 0.015 M sodium citrate) at 58C for 15 min, followed
by a 15 min wash at 58C in 2× SSC and exposed to Kodak X-ray
film for 4 h at room temperature for the oligonucleotide probes.
[α-32P]-Probed filters were washed in 2× SSC, 1% SDS for
15 min, 1× SSC, 1% SDS for 15 min and 0.5× SSC, 1% SDS
for 30 min, followed by overnight exposure to Kodak X-ray film.
Plasmid DNA was prepared from an overnight 3 ml liquid bacterial
culture with the High Pure Plasmid Isolation Kit according to the
manufacturer’s protocol (Boehringer Mannheim). The DNA was
digested with Sau3AI to determine insert size by agarose gel
electrophoresis.
Sequencing
Purified plasmid DNA was sequenced by PCR cycle sequencing
using the fluorescent T7 dye termination system (Applied
Biosystems) and the pUC/M13 forward and reverse primers. For
transposon-based sequencing plasmid inserts were recloned in
the pOCUS-2 vector supplied with the Locus-Pocus subcloning
system (Novagen). Transformation, mating and analysis of the
clones were performed according to the manufacturer’s protocol.
RNA was isolated from ∼107 EBV-transformed lymphoblast
cells as described (30) and treated with RNase-free DNase
(Promega). 3′-RACE products were obtained by nested PCR on
first strand cDNA synthesized with the 3′-RACE system (Gibco
BRL). A first PCR was done with the primers SCA7.3
(5′-TCCATCAAGAGGATGAGTGT-3′) and the 3′-RACE
adapter primer supplied with the kit. A nested PCR was done with
primer SCA7.2 (5′-CATGAACAATGTCCACATGA-3′) and
UAP (Gibco BRL). The PCR product was gel purified and PCR
cycle sequenced with primer SCA7.2. For the RT-PCR reactions
first strand cDNA was synthesized by random priming using the
Superscript II system (Gibco BRL) and PCR amplified using the
primer set H1 and H2 or H2 and 1220R (Fig. 3). The SCA7 CAG
fragments were size selected on 2% agarose gel, subcloned in
179
pGem3zf(–) and sequenced. Verification of the lymphoblast cDNA
sequence was done by RT-PCR using four overlapping primer sets.
First strand cDNA was amplified with a standard PCR of 30 cycles
using primer pairs: 562F and SCA7.H (5′-TGTGACTGTCCCTATCTTAAGG-3′); SCA7.8 and ctg3R (5′-GCATCCCATTCAGCAAAGTA-3′ and 5′-CGTAATAGCCTCTTCCTATCTG3′); SCA7.6 and SCA7.E (5′-TAAACCTAAACCTCACACCC-3′
and 5′-GCCAGATAGCTGACTCCACA-3′); SCA7.5 and
SCA7.A (5′-TCTCCACACGTATTCCTCAC-3′ and 5′-CCAAACACATTTCTGAGCTG-3′). The sequences of the amplification
products were determined with cDNA primers 562F, H2, H1,
SCA7.2, SCA7.8, ctg3R, SCA7.6, SCA7.5, H1IC
(5′-CCTCTGCCCAGTCCTGAA-3′), SCA7.I (5′-ACAGAAACCAAATATTGGCA-3′), SCA7.7 (5′-TATTAGCCGAGCACAAAAAC-3′), SCA7.G (5′-CTACAAAAGGCTTTGATTCG-3′),
SCA7.E
(5′-GCCAGATAGCTGACTCCACA-3′),
SCA7.4
(5′-AATCCAGCAAATCTTTGAGG-3′), SCA7.3 (5′-TCCATCAAGAGGATGAGTGT-3′), SCA7.2 (5′-CATGAACAATGTCCACATGA-3′), 562E (5′-ATTGCAGCAGATGAAGATCC-3′) and
562A (5′-CTCGTTCGTCTTAATTCTGTTC-3′). Sequences were
analyzed on an automated DNA sequencer model ABI 373A
(Applied Biosystems) with the fluorescent T7 dye termination
system (Applied Biosystem). The obtained cDNA sequences were
aligned with DNAStar to generate the cDNA consensus sequence.
Exon trapping
Exon trapping was essentially as published (18), with minor
modifications. Cosmid DNA was purified by standard phenol/
chlorofrom extractions. After electroporation (750 V, 900 µF) the
V79 cells were transferred immediately to a 100 mm tissue
culture dish containing 10 ml prewarmed DMEM + 10% FCS, to
minimize cell death. To avoid DNA contamination of the exon
trap products total RNA was first pretreated with RNase-free
DNase (Promega). First strand cDNA synthesis was performed
with primer hGHex4/5 (5′-CTTCCAGCCTCCCCATCAGC-3′),
spanning two exon/intron boundaries. Aliquots of 5 µl first strand
cDNA were used as template in a standard PCR reaction with
primer hGHex1/2 (5′-ATGGCTACAGGCTCCCGGAC-3′).
PCR amplification of the SCA7 repeat
Aliquots of 100 ng genomic DNA was mixed with 10 pmol
primers H1 and H2 (Fig. 3) in a total volume of 25 µl, containing
20 mM Tris–HCl, pH 8.4, 50 mM KCl, 0.25 mM dNTPs, 0.05%
W1 (Gibco BRL), 1 mM MgCl2, 10% DMSO and 1.5 U Taq
DNA polymerase (Gibco BRL). Samples were denatured at 94C
for 2 min followed by 30 cycles of denaturation (94C, 1 min),
annealing (54C, 1 min) and extension (72C, 2 min) and a final
extension at 72C for 10 min. PCR product were analyzed on a
2% agarose gel, blotted onto Hybond-N+, hybridized with a
radiolabeled (CTG)10 oligonucleotide, washed and exposed to
Kodak X-ray film for 4 h at room temperature. In the case of
fluorescent detection of the SCA7 CAG repeat the H2 primer was
chemically labeled with the fluorescence dye JOE and PCR
fragments were separated on a 6% polyacrylamide sequencing
gel. The SCA7 repeat length was determined on a model ABI
373A sequencer using GENESCAN 672 software (Applied
Biosystems) with the GENESCAN 2500ROX marker as internal
size standard.
180
Human Molecular Genetics, 1998, Vol. 7, No. 2
ACKNOWLEDGEMENTS
The authors are grateful to the patients and family members of
family CA-1 for their cooperation in the research project, to
Nicole Datson and Johan den Dunnen for providing the sCOGH6
vector and to Marleen Van den Broeck for help with the
sequencing. The work described in this paper was in part funded
by the Fund for Scientific ResearchFlanders (Belgium), a
research project of the University of Antwerp, Belgium and an
EU grant BIOMED2 CT97-2466. L.K. and J.T. are grant holders
of the Institute for Science and Technology.
REFERENCES
1. Harding,A.E. and Deufel,T. (1993) Clinical features and classification of
inherited ataxias. Adv. Neurol. 61, 1–14.
2. Weiner,L., Konigsmark,B., Stoll,J. and Magladery,J. (1967) Hereditary
olivopontocerebellar atrophy with retinal degeneration. Report of a family
through 6 generations. Arch. Neurol., 16, 376.
3. Benomar,A., Krols,L., Stevanin,G., Cancel,G., LeGuern,E., David,G.,
Ouhabi,H., Martin,J.J., Durr,A. and Zaim,A. (1995) The gene for autosomal
dominant cerebellar ataxia with pigmentary macular dystrophy maps to
chromosome 3p12–p21.1. Nature Genet., 10, 84–88.
4. Orr,H., Chung,M., Banfi,S., Kwiatkowski,T., Servadio,A., Beaudet,A.,
McCall,A., Duvick,L., Ranum,L. and Zoghbi,H. (1993) Expansion of an
unstable trinucleotide CAG repeat in spinocerebellar ataxia type 1. Nature
Genet., 4, 221–226.
5. Pulst,S., Nechiporuk,A., Nechiporuk,T., Gispert,S., Chen,X., LopesCendes,I., Pearlmans,S., Starkman,S., Orozco-Diaz,G., Lunkes,A., DeJong,P., Rouleau,G.A., Auburger,G., Korenberg,J.R., Figueroa,C. and
Sahba,S. (1996) Moderate expansion of a normally biallelic trinucleotide
repeat in spinocerebellar ataxia type 2. Nature Genet., 14, 269–276.
6. Kawaguchi,Y., Okamoto,T., Taniwaki,M., Aizawa,M., Inoue,M., Katayama,S.,
Kawakami,H., Nakamura,S., Nishimura,M., Akiguchi,I., Kimura,J.,
Naruyima,S. and Kakizuka,A. (1994) CAG expansions in a novel gene for
Machado–Joseph disease at chromosome 14q32.1. Nature Genet., 8, 221–228.
7. Zhuchenko,O., Bailey,J., Bonnen,P., Ashizawa,T., Stockton,D.W., Amos,C.,
Dobyns,W.B., Subramony,S.H., Zoghbi,H. and Lee,C.C. (1997) Autosomal
dominant cerebellar ataxia (SCA6) associated with small polyglutamine
expansions in the alpha 1A-voltage-dependent calcium channel. Nature
Genet., 15, 62–69.
8. Ranum,L., Schut,L., Lundgren,J., Orr,H. and Livingston,D. (1994)
Spinacerebellar ataxia type 5 in a family descended from the grandparents of
President Lincoln maps to chromosome 11. Nature Genet., 8, 280–284.
9. Gouw,L., Kaplan,C., Haines,J., Digre,K., Rutledge,S., Matilla,A., Leppert,M.,
Zoghbi,H. and Ptacek,L. (1995) Retinal degeneration characterizes a
spinocerebellar ataxia mapping to chromosome 3p. Nature Genet., 10, 89–93.
10. Holmberg,M., Johansson,J., Forsgren,L., Heijbel,J., Sandgren,O. and
Holmgren,G. (1995) Localization of autosomal dominant cerebellar ataxia
associated with retinal degeneration and anticipation to chromosome
3p12–p21.1. Hum. Mol. Genet., 4, 1441–1445.
11. David,G., Giunti,P., Abbas,N., Coullin,P., Stevanin,G., Horta,W., Gemmill,R.,
Weissenbach,J., Wood,N., Cunha,S., Drabkin,H., Harding,A.E., Agid,Y. and
Brice,A. (1996) The gene for autosomal dominant cerebellar ataxia type II is
located in a 5-cM region in 3p12–p13: genetic and physical mapping of the
SCA7 locus. Am. J. Hum. Genet., 59, 1328–1336.
12. Krols,L., Martin,J.J., David,G., Van Regemorter,N., Benomar,A.,
Lofgren,A., Stevanin,G., Durr,A., Brice,A. and Van Broeckhoven,C. (1997)
Refinement of the locus for autosomal dominant cerebellar ataxia type II to
chromosome 3p21.1–14.1. Hum. Genet., 99, 225–232.
13. David,G., Abbas,N., Stevanin,G., Durr,A., Yvert,G., Cancel,G., Weber,C.,
Imbert,G., Saudou,F., Antoniou,E., Drabkin,H., Gemmill,R., Giunti,P.,
Benomar,A., Wood,N., Ruberg,M., Agid,Y., Mandel,J.L. and Brice,A. (1997)
Cloning of the SCA7 gene reveals a highly unstable CAG repeat expansion.
Nature Genet., 17, 65–70.
14. Martin,J., Van Regemorter,N., Krols,L., Brucher,J., de Barsy,T., Szliwowski,H.,
Evrard,P., Ceuterick,C., Tassignon,M., Smet-Dieleman,H., Heyez-Delatte,F.,
Willems,P. and Van Broeckhoven,C. (1994) On an autosomal dominant form of
retinal-cerebellar degeneration: an autopsy study of five patients in one family.
Acta Neuropathol., 88, 277–286.
15. Hudson,T.J., Stein,L.D., Gerety,S.S., Ma,J., Castle,A.B., Silva,J., Slonim,D.K.,
Baptista,R., Kruglyak,L. and Xu,S.H. (1995) An STS-based map of the human
genome. Science, 270, 1945–1954.
16. Dib,C., Faure,S., Fizames,C., Samson,D., Drouot,N., Vignal,A., Millasseau,P.,
Marc,S., Hazan,J., Seboun,E., Lathrop,M., Gyapay,G., Morissette,J. and
Weissenbach,J. (1996) A comprehensive genetic map of the human genome
based on 5,264 microsatellites. Nature, 380, 152–154.
17. Cohen,D., Chumakov,I. and Weissenbach,J. (1993) A first generation
physical map of the human genome. Nature, 336, 698–701.
18. Datson,N.A., van de vosse,E., Dauwerse,H.G., Bout,M., van Ommen,G.B.
and den Dunnen,J.T. (1996) Scanning for genes in large genomic regions:
cosmid-based exon trapping of multiple exons in a single product. Nucleic
Acids Res., 24, 1105–1111.
19. Batzer,M.A., Alegria-Hartman,M. and Deininger,P.L. (1994) A consensus
Alu repeat probe for physical mapping. Genet. Anal. Tech. Appl., 11, 34–38.
20. Uberbacher,E.C. and Mural,R.J. (1991) Locating protein-coding regions in
human DNA sequences by a multiple sensor–neural network approach. Proc.
Natl Acad. Sci. USA, 88, 11261–11265.
21. Trottier,Y., Lutz,Y., Stevanin,G., Imbert,G., Devys,D., Cancel,G., Saudou,F.,
Weber,C., David,G. and Tora,L. (1995) Polyglutamine expansion as a
pathological epitope in Huntington’s disease and four dominant cerebellar
ataxias. Nature, 378, 403–406.
22. Stevanin,G., Trottier,Y., Cancel,G., Durr,A., David,G., Didierjean,O., Burk,K.,
Imbert,G., Saudou,F., Abada-Bendib,M., Gourfinkel-An,I., Benomar,A.,
Abbas,N., Klockgether,T., Grid,D., Agid,Y., Mandel,J.L. and Brice,A. (1996)
Screening for proteins with polyglutamine expansions in autosomal dominant
cerebellar ataxias. Hum. Mol. Genet., 5, 1887–1892.
23. Lindblad,K., Savontaus,M.L., Stevanin,G., Holmberg,M., Digre,K., Zander,C.,
Ehrsson,H., David,G., Benomar,A., Nikoskelainen,E., Trottier,Y., Holmgren,G.,
Ptacek,L.J., Anttinen,A., Brice,A. and Schalling,M. (1996) An expanded CAG
repeat sequence in spinocerebellar ataxia type 7. Genome Res., 6, 965–971.
24. Lindblad,K., Nylander,P., De bruyn,A., Sourey,D., Zander,C., Engström,C.,
Holmgren,G., Hudson,T., Chotai,J., Mendlewicz,J., Van Broeckhoven,C.,
Schalling,M. and Adolfsson, R. (1995) Detection of expanded CAG repeats
in bipolar affective disorder using the repeat expansion detection (RED)
method. Neurobiology, 2, 55–62.
25. Nielsen,J., Koefoed,P., Abell,K., Hasholt,L., Eiberg,H., Fenger,K., Niebuhr,E.
and Sorensen,S. (1997) CAG repeat expansion in autosomal dominant pure
spastic paraplegia linked to chromosome 2p21–p24. Hum. Mol. Genet., 6,
1811–1816.
26. Cruts,M., Backhovens,H., Wang,S., Van Gassen,G., Theuns,J., De Jonghe,C.,
Wehnert,A., De Voecht,J., De Winter,G., Cras,P., Bruyland,M., Datson,N.,
Weissenbach,J., den Dunnen,J.T., Martin,J.J., Hendriks,L. and Van
Broeckhoven,C. (1995) Molecular genetic analysis of familial early-onset
Alzheimer’s disease linked to chromosome 14q24.3. Hum. Mol. Genet., 4,
2363–2371.
27. Benomar,A., Le Guern,E., Durr,A., Ouhabi,H., Stevanin,G., Yahyaoui,M.,
Chkili,T., Agid,Y. and Brice,A. (1994) Autosomal-dominant cerebellar ataxia
with retinal degeneration (ADCA type II) is genetically different from ADCA
type I. Annls Neurol., 35, 439–444.
28. Del-Favero,J. and Jacobs,M. (1996) Method for rapid restriction analysis of
YAC clones. BioTechniques, 21, 404–405.
29. Sheng,Y., Mancino,V. and Birren,B. (1995) Transformation of Escherichia coli
with large DNA molecules by electroporation. Nucleic Acids Res., 23,
1990–1996.
30. Chomczynski,P. and Sacchi,N. (1987) Single-step method of RNA isolation by
acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem.,
162, 156–159.