ARTICLE
Study of three intragenic polymorphisms in the
Machado-Joseph disease gene (MJD1) in
relation to genetic instability of the (CAG)
n
tract
Patr´ıcia Maciel
1,5
, Claudia Gaspar
1,5
, Laura Guimar ˜aes
5
, Jun Goto
3
,
Iscia Lopes-Cendes
1
, Sean Hayes
1
, Karin A rvidsson
1
, Aureliano Dias
1
,
Jorge Sequeiros
2,5
, Alda Sousa
4,5
and Guy A Rouleau
1
1
Centre for Research in N euroscience, McG ill University and the M ontreal G eneral Hospital Research Institute,
Montreal, Q u ´ebec, Canada
2
L ab. G en ´etica M ´edica, ICBA S
3
Departm ent of N eurology, Tokyo University M edical School, Tokyo, Japan
4
Gen´etica A plicada, ICBA S, Universidade do Porto, Portugal
5
UnIG E N e, IB M C Universidade do Porto, Portugal
Intergenerational instability is one of the most important features of the disease-associated
trinucleotide expansions, leading to variation in size of the repeat among and within families,
which manifests as variable age at onset and severity, and is probably the basis for the
occurrence of anticipation. Several factors are known to affect the degree of instability, namely
the type of repeated sequence, its initial size, the presence or absence of interruptions in the
repetitive tract and the gender of the transmitting parent. A recent study demonstrated the
effect of an intragenic polymorphism (C
987
GG/G
987
GG) in the Machado-Joseph disease
causative gene, immediately downstream of the CAG repeat, on the intergenerational
instability of the expanded repeat. Surprisingly, there was an effect not only of the specific
allele in cis to the disease chromosome, but also of the allele on the normal chromosome,
suggesting the existence of an interaction between the normal and expanded alleles that affects
the fidelity of replication of the (CAG)
n
tract. This effect could be a direct effect of the
polymorphism studied or, alternatively, this polymorphism could be in disequilibrium with
some other flanking sequence which affects the instability of the repetitive (CAG)
n
tract. In
order to confirm the previous results in a different population and to distinguish between a
direct and indirect effect of the CGG/GGG polymorphism, we typed 70 parent–progeny pairs
for which the variation in the (CAG)
n
length in the MJD 1 gene was known, for three
intragenic polymorphisms: C
987
GG/G
987
GG and two additional, newly described ones,
TA A
1118
/TA C
1118
and A
669
TG/G
669
TG. We also typed a control population of 125 individuals
for the A
669
TG/G
669
TG, C
987
GG/G
987
GG and TAA
1118
/TA C
1118
polymorphisms, in an attempt
to identify any association between haplotype and (CAG)
n
length in normal chromosomes,
Correspondence: Patr´ıcia Maciel, U nIG E Ne, IBMC - Uni-
versidade do Porto, R. Campo Alegre, 823, 4050 Porto,
Portugal. Tel: 351 2 6074941; Fax: 351 2 6099157 or 5506519;
E-mail: patricia.maciel@writeme.com
Received 24 February 1998; revised 8 July 1998; accepted 18
August 1998
European Journal of Human Genetics (1999) 7, 147–156
© 1999 Stockton Press All rights reserved 1018–4813/99 $12.00
t
http://www.stockton-press.co.uk/ejhg
suggestive of an instability-predisposing effect of the repeat-flanking sequences, which could
have led to the origin of the MJD mutation in the human population. We confirmed the effect
of the C
987
GG/G
987
GG polymorphism on intergenerational instability when present in trans.
Our results suggest that this effect is restricted to a small region of the gene, immediately
downstream of the CAG repeat, which includes this particular nucleotide substitution and the
stop codon of the MJD 1 cDNA, and is not a more widespread chromosomal effect. The lack
of a significant association of any specific intragenic haplotype with larger CAG repeats in
normal chromosomes, together with the absence of an effect of the intragenic haplotype in cis
on the intergenerational instability of the expanded (CAG)
n
in MJD families does not indicate
the existence of an instability-predisposing haplotype.
Keywords: ataxia; neurodegenerative disorder; polyglutamine; trinucleotide repeats; dynamic
mutations; haplotypes; origin of mutation
Introduction
Machado-Joseph disease (MJD) is an autosomal domi-
nant multisystem degeneration of adult onset
1
with
variable clinical presentation. Its manifestations include
cerebellar ataxia and progressive external ophthalmo-
plegia, associated in a variable degree with pyramidal
signs, dystonia, rigidity, amyotrophies, and peripheral
neuropathy.
1–4
The mutation associated with this dis-
order is the expansion of a (CAG )
n
tract in the coding
region of the M JD 1 gene,
5
which in normal individuals
contains from 12 to 41 repeat units and in expanded
chromosomes from 61 to 86 repeat units.
5–7
Although the mechanisms of trinucleotide repeat
instability are still not understood, a variety of factors
have been implicated as modulating intergenerational
repeat instability in the diseases associated with
dynamic mutations, namely the sex of transmitter,
6–18
the expanded repeat size in the transmitting par-
ent,
8–12,18–20
the presence or absence of an interruption
of the repeat,
16,21–25
and even the sex of the
descendent.
26
We and others have previously shown that in MJD
families the sex of the transmitting parent has a
significant effect on intergenerational instability,
6,7,17
with male meiosis associated with larger variations,
both contractions and expansions, of repeat size.
However, there was no association between size of
CAG repeat length in the transmitting parent and
change in CAG repeat length during transmission,
6
nor
with the presence of the CA A interruption at the 5' end
of the CAG repeat, since both control and expansion-
carrying individuals have this interruption.
17,27
A single-base substitution polymorphism (C
987
GG/
G
987
GG) at the 3' end of the CAG repeat in the MJD1
gene
5
was recently demonstrated to be associated with
differential intergenerational instability of (CAG )
n
length.
17
Surprisingly, this was not observed uniquely
for the allele in the disease chromosome (in cis of the
expansion); the combination with the allele in the
normal chromosome (in trans of the expansion) was
shown to effect the degree of intergenerational instabil-
ity.
28
The biological relevance of these findings remain
unclear: whilst it is possible that this association results
from a direct effect of the base substitution at the site
examined, it may alternatively reflect the effect of a
sequence or gene closely linked to this site.
In order to confirm the previous observations in a
different population and to distinguish between a direct
and indirect effect of the C
987
GG/G
987
GG polymor-
phism, we analysed the influence of three intragenic
single-base substitution polymorphisms, C
987
GG/
G
987
GG and two additional, newly described ones,
TA A
1118
/TAC
1118
and A
669
TG /G
669
TG,
29
on interge-
nerational instability of the expanded CAG tract in the
MJD1 gene in our set of families. In addition, we
studied a control population for the three intragenic
polymorphisms, in an attempt to identify an association
between haplotype and (CAG )
n
length in normal
chromosomes that might give an insight into the origin
of the MJD mutation in the human population.
Methods
Subjects
Our study is based on 117 individuals with a previously
detected (CAG)
n
expansion in the M JD 1 gene, including
affected individuals and asymptomatic carriers, constituting
70 parent–progeny pairs, belonging to 21 different families.
Given that the great majority of families in our MJD
population was traceable to a Portuguese-Azorean origin, we
used a control population of Azorean origin for the study of
the association between haplotypes and (CAG)
n
length,
Intragenic polymorphisms and instability in MJD
t
P Maciel et al
148
including 124 normal individuals with no known relationship
to an MJD family.
Blood samples were obtained after informed consent from
all individuals, and genomic DNA was extracted from
lymphocytes as described elsewhere.
30
Size of the CAG Repeat and Typing of Intragenic
Polymorphisms
Amplification of CAG repeat-containing fragment of the
MJD1 gene was performed by PCR using previously
described conditions,
5
and the size of the PCR products was
determined by denaturing polyacrylamide gel electrophoresis,
in parallel with an M13 sequence ladder.
The intragenic polymorphism C
987
GG/G
987
GG, situated at
the 3' end of the CAG repeat, was detected by allele-specific
PCR using primers A SP1 (ACTCTG TCCTG A-
TAGGTCCCC) or ASP2 (ACTCTG TCCTGA-
TAGGTCCCG) in combination with MJD 52.
5
The stop
codon polymorphism TA A
1118
/TAC
1118 29
was detected by
allele-specific PCR using primers ASP3 (G CAAA A ATCA -
CATGGAGCTCG ) or ASP4 (G CAAA A ATCACATG-
GAGCTCG ) in combination with MJD52. In both cases, the
PCR reaction was performed using the same conditions used
for amplification of the CAG repeat, except for the annealing
step, which was performed at 61°C for 30 s. The PCR products
were analysed on a denaturing 6% polyacrylamide gel and
visualised by autoradiography.
The A
669
TG /G
669
TG polymorphism
29
was detected by
SSCP analysis using a 0.5X MD E gel (FMC BioProducts)
with 5% glycerol for electrophoresis; the PCR amplification
was performed using primers MJDIVSR (TACTAG AG CT-
TATTTG CCAG) and MJD 734R (CAG AG CCCTCTG -
CAAATCCT), in a solution containing 1X PCR buffer
(Perkin Elmer), 0.2 mM dNTP-A, 0.03 mM dATP, 8 ng/µl each
primer, 0.2 µCi
35
S dATP, 0.1 µg/µl BSA , 0.12 U AmpliTaq
(Perkin Elmer). The amplification conditions were as follows:
initial denaturation for 5 min at 94°C, followed by 1 min at
94°C, 1 min at 56°C and 1 min at 72°C for 30 cycles and a final
elongation period of 5 min at 72°C. When necessary, the
polymorphisms were typed for both parents and progeny, in
order to determine phase.
Statistical Methods
In order to analyse the associations between intragenic
polymorphisms and the size of the CAG repeats in normal
chromosomes (of control individuals), we compared the
frequencies of the alleles in three classes of normal chromo-
somes, grouped according to size of (CAG )
n
(up to 21, 22 to
26, and more than 27 repeat units), using Fisher’s exact test
2 3 k and Pearson χ
2
test.
The intergenerational changes in CAG repeat length were
analysed by comparison of absolute change in size of the
expanded allele during transmission. Fisher’s exact test
(2 3 2) was used to determine if intergenerational change in
repeat number between parent and offspring, expressed as
two classes of variation (difference of more than two repeat
units between parent and progeny – large variation, and
difference of up to two repeat units – small variation), was
associated with sex of transmitting parent.
The distribution of the variable intergenerational changes
in CAG repeat length as defined is highly skewed, therefore
non-parametric methods are appropriate for hypothesis
testing; Mann-Whitney U test or Kruskal-Wallis A NOVA
were used to assess associations between alleles or haplotypes
and instability.
Linkage disequilibrium analysis was performed using Fish-
er’s exact test.
All analyses were performed using procedures of SPSS,
release 6.1.3, ©SPSS Inc. (1989–1995) and E xact 2 3 k- exact
test for a 2 3 k table, version 2.04, ©JH A bramson & PM
Gahlinger (1993–1996).
Results
No Apparent Association between the
Alleles of Intragenic Polymorphisms in cis of
the (CAG)
n
Expansion and Amplitude of
Intergenerational Variation of Expanded
(CAG)
n
Size
We detected no significant difference in the distribution
of absolute variation of (CAG )
n
length during transmis-
sion between individuals carrying either the C or G
alleles of the C
987
GG/G
987
GG polymorphism in cis with
the expanded allele (Mann-Whitney U test, P = 0.150).
The comparison of absolute variation in repeat size
between individuals carrying either the A or the C
variants of the TAA
1118
/TAC
1118
polymorphism in cis
with the expanded CAG also revealed absence of a
significant difference in the median of the absolute
variation in repeat size between the two groups (Mann-
Whitney U test, P = 0.450). The same was observed for
the A
669
TG /G
669
TG polymorphism (Mann-Whitney U
test, P = 0.700).
No association was detected between the presence of
any particular (complete or partial) intragenic haplo-
type in cis (defined by the three intragenic polymor-
phisms mentioned above) with increased instability
(Kruskal-Wallis test, P = 0.378).
A borderline significant effect of the C
987
GG/
G
987
GG polymorphism in cis with the expanded allele
was detected only when the stronger effect of the allele
in trans was taken into account (see below).
Influence of the Genotype of Intragenic
Polymorphisms on the Intergenerational
Instability of the Expanded (CAG)
n
When the transmission events were divided into groups
according to genotype of the transmitting parent for the
C
987
GG/G
987
GG polymorphism, taking both the alleles
in cis and in trans into account (Table 1), there was a
significant difference in the median of the absolute
variation in repeat length during transmission between
the groups, with the parental (CAG)exp-CGG/(CAG )
normal-G G G genotype associated with increased insta-
bility (Kruskal-Wallis test, P = 0.028).
Intragenic polymorphisms and instability in MJD
P Maciel et al
t
149
When the genotype of the stop codon polymorphism
was considered, the (CAG )exp-TAA/(CAG)normal-
TAC polymorphism was associated with larger inter-
generational variation, but the difference was not
statistically significant (Kruskal-Wallis test,
P = 0.080).
The parental genotype for the A
669
TG /G
669
TG poly-
morphisms did not have a significant effect on instabil-
ity (Kruskal-Wallis test, P = 0.632).
Effect of the Alleles of Intragenic
Polymorphisms Present in the Normal
Chromosome on the Amplitude of
Intergenerational Variation of Expanded
(CAG)
n
Size in the MJD Chromosome
A highly significant association was observed (Table 2)
between the presence of the G variant of the C
987
GG/
G
987
GG polymorphism on the normal chromosome of
the transmitting parent and an increased amplitude of
intergenerational variation in expanded repeat size
(Mann-Whitney U test, P = 0.019).
A similar analysis for both the TAA
1118
/TAC
1118
and
the A
669
TG /G
669
TG polymorphisms revealed an asso-
ciation of the C variant of the stop codon polymor-
phism present in the normal chromosome of the
transmitter with an increased intergenerational insta-
bility of the expanded repeat (Mann-Whitney U test,
P = 0.024), but no such association was detected for the
A
669
TG /G
669
TG polymorphism (Mann-Whitney U test,
P = 0.291).
Given these results, we also analysed the effect of the
polymorphisms present in cis of the expanded allele on
its instability controlling for the effect of the polymor-
phism present in trans. Considering only the individuals
that had a GGG variant in trans (which is associated
with increased intergenerational instability), when we
compared the degree of instability between G G G and
CG G carriers in cis (n = 20 and n = 33, respectively),
the difference in instability was of borderline sig-
nificance, in spite of a small sample size (Mann-
Whitney U test, P = 0.055), indicating an increased
instability in carriers of the C variant in cis. For the stop
codon polymorphism, however, considering the pairs
where the progenitor had the C variant in the normal
chromosome, the difference between C (n = 23) and G
(n = 16) carriers was not significant (Mann-Whitney U
test, P = 0.19).
We did not detect a significant association of any
particular intragenic haplotype (defined by the three
intragenic polymorphisms mentioned above) in the
normal chromosome of the transmitter with increased
intergenerational instability of the expanded repeat
(Kruskal-Wallis, P = 0.321). However, when partial
haplotypes, including only two of the intragenic poly-
morphisms, were analysed, the haplotype defined by
the C
987
GG/G
987
GG and TAA
1118
/TAC
1118
polymor-
phisms seemed to have an effect on instability, with the
GC haplotype in trans associated with larger variations
of the expanded repeat size (Kruskal-Wallis,
P = 0.042). The partial haplotypes defined by the
A
669
TG /G
669
TG and C
987
GG/G
987
GG polymorphisms,
and by the A
669
TG /G
669
TG and TA A
1118
/TAC
1118
poly-
morphisms were not associated with differential insta-
bility (Kruskal-Wallis, P = 0.270 and P = 0.197,
respectively).
Combined Analysis of the Effects of
Intragenic Polymorphisms and Gender of
the Transmitting Parent
Knowing that gender of transmitting parent is an
important factor determining the degree of inter-
generational instability of the expanded (CAG)
n
in
MJD, we analysed the influence of intragenic polymor-
phisms on instability separately for maternal and
paternal transmissions.
As is shown in Table 1, the significance of the effect of
genotype of intragenic polymorphisms on intergenera-
tional instability was increased when we considered
maternal transmissions only: for paternal transmissions,
no significant effect of the genotype was observed.
Similarly, when the analysis of the effect of the
polymorphisms in trans of the expanded allele was
Table 1 Association of genotypes (alleles in cis+trans of the
expanded repeat) of the three intragenic polymorphisms in
the M JD1 gene with increased intergenerational instability of
the repeat (Kruskal-Wallis)
polym orphism all male fem ale
transm issions transm issions transm issions
A
669
TG /G
669
TG p=0.632 p=0.914 p=0.304
C
987
GG/G
987
GG p=0.028 (C+G) p=0.623 p=0.014 (C+G)
TA A
1118
/TAC
1118
p=0.080 p=0.685 p=0.027 (C+C)
Table 2 A ssociation of alleles of the three intragenic
polymorphisms in the MJD 1 gene in trans of the expanded
(CAG )
n
with increased intergenerational instability of the
repeat (Mann-Whitney test)
polym orphism all male fem ale
transm issons transmissons transm issons
A
669
TG /G
669
TG p=0.291 p=0.942 p=0.092
C
987
GG/G
987
GG p=0.019 (G) p=0.301 p=0.040 (G)
TA A
1118
/TAC
1118
p=0.024 (C) p=0.650 p=0.003 (C)
Intragenic polymorphisms and instability in MJD
t
P Maciel et al
150
performed separately for paternal and maternal trans-
missions, we observed a significant effect only in
maternal transmissions (Table 2).
The haplotype analysis confirmed these observations:
the G C haplotype defined by the C
987
GG/G
987
GG and
TA A
1118
/TAC
1118
polymorphisms in trans was asso-
ciated with increased instability in maternal transmis-
sions only (Kruskal-Wallis, P = 0.029).
The direct comparison between haplotypes G A and
GC defined by the C
987
GG/G
987
GG and TAA
1118
/
TA C
1118
polymorphisms in maternal transmissions
revealed a borderline significant difference in the
degree of instability among carriers of these haplotypes
(P = 0.047, Mann-Whitney U test), which was not
detected when the two genders were analysed together
(P = 0.656, Mann-Whitney U test). The comparison of
the CC and GC haplotypes revealed no significant
difference in the degree of instability (P = 0.555,
Mann-Whitney U test).
Larger Sizes of the (CAG)
n
in MJD1 are not
Associated with Specific Alleles of the
Intragenic Polymorphisms
The distribution of CAG repeat sizes in the normal
chromosomes of our Portuguese-Azorean control pop-
ulation is grossly trimodal, with peaks at 15, 24 and 28
repeats. To test for association between intragenic
polymorphisms and (CAG )
n
length in normal chromo-
somes, three groups of normal alleles (defined by
(CAG)
n
lengths from 15 to 21, 22 to 26 and 27 to 36)
were compared. The frequencies of the C and G alleles
of the C
987
GG/G
987
GG polymorphism were signifi-
cantly different between groups (Fisher’s exact test,
P = 0.0021). However, the detailed analysis of the
distribution of these alleles revealed that, although
there is an increased relative frequency of the C allele
among the chromosomes with large CAG repeats when
compared with small and medium CAG repeats, this
allele is not exclusively associated with the larger
repeat lengths or even the predominant allele in this
group of chromosomes (Figure 1a).
The same analysis was performed for the TAA
1118
/
TA C
1118
polymorphism in 231 normal chromosomes,
and we observed a difference in distribution
(P < 0.0001). But although there was an enrichment of
allele A among the large CAG repeat-containing
normal chromosomes when compared with the small or
medium size repeats, no predominance or exclusive
association of this allele with larger repeats was
observed (Figure 1b). Similar results were obtained for
the A
669
TG /G
669
TG polymorphism (n = 154,
P = 0.0051 – Figure 1c). A summary of these results is
presented in Table 3.
Haplotypes could be determined for 133 individuals,
and their distribution according to (CAG )
n
sizes
revealed that four different haplotypes are present in
Figure 1a Distribution of the alleles of the C
987
GG/G
987
GG
polym orphism according to (CAG )
n
size, in a control popula-
tion (n = 232) of Portuguese-A z orean origin and in the M JD
fam ilies of Portuguese descent. 1b D istribution of the alleles
of the TA A
1118
/TA C
1118
polym orphism according to (CAG )
n
size, in a control population (n = 231) of Portuguese-A zorean
origin and in the M JD families of Portuguese descent.
1c D istribution of the alleles of the A
669
TG/G
669
TG polym or-
phism according to (CAG )
n
size, in a control population
(n = 154) of Portuguese-A zorean origin and in the M JD
fam ilies of Portuguese descent
Intragenic polymorphisms and instability in MJD
P Maciel et al
t
151