Structure, diversity, and evolution of the 45-bp VNTR in intron 5 of the USH1C gene.

Abstract

Usher syndrome type IC is a rare, autosomal recessive sensorineural disorder caused by mutations in the USH1C gene, which encodes a PDZ-domain protein named harmonin. The Acadian-specific 216G-->A mutation in exon 3 and a variant 9-repeat VNTR allele (designated VNTR(t,t)) in intron 5 are in complete linkage disequilibrium. (The usual form of the allele is referred to as VNTR(t).) To gain insight into the structure, diversity, and evolution of the VNTR, we analyzed individuals from seven different populations, as well as nonhuman primates and rodents. The 2-, 3-, and 6-repeat VNTR alleles were the most common. Four novel alleles containing 1, 5, 7, and 10 repeats were detected with frequencies of 0.002, 0.029, 0.005, and 0.001, respectively. The USH1C VNTR region is highly conserved among primates, but not between primates and rodents. Five unrelated individuals had a 3-repeat VNTR(t,t) allele. Haplotype analysis indicates that the 9-repeat VNTR(t,t) and the 3-repeat VNTR(t,t) alleles arose independently. However, the 9-repeat VNTR(t,t) and 6-repeat VNTR(t) alleles shared the same haplotype, suggesting an expansion from 6(t) to 9(t,t).

Full text

Page 1

Structure, diversity, and evolution of the 45-bp VNTR in
intron 5 of the USH1C gene
$
Sevtap Savas,aBen Frischhertz,bMark A. Batzer,c
Prescott L. Deininger,band Bronya J.B. Keatsa,*
aDepartment of Genetics, Louisiana State University Health Sciences Center, 533 Bolivar Street, New Orleans, LA 70112, USA
bTulane Cancer Center and Department of Environmental Health Sciences, New Orleans, LA 70112, USA
cDepartment of Biological Sciences, Biological Computation and Visualization Center, Louisiana State University, Baton Rouge, LA 70803, USA
Received 23 June 2003; accepted 4 September 2003
Abstract
Usher syndrome type IC is a rare, autosomal recessive sensorineural disorder caused by mutations in the USH1C gene, which encodes a
PDZ-domain protein named harmonin. The Acadian-specific 216G!A mutation in exon 3 and a variant 9-repeat VNTR allele (designated
VNTR(t,t)) in intron 5 are in complete linkage disequilibrium. (The usual form of the allele is referred to as VNTR(t).) To gain insight into
the structure, diversity, and evolution of the VNTR, we analyzed individuals from seven different populations, as well as nonhuman primates
and rodents. The 2-, 3-, and 6-repeat VNTR alleles were the most common. Four novel alleles containing 1, 5, 7, and 10 repeats were
detected with frequencies of 0.002, 0.029, 0.005, and 0.001, respectively. The USH1C VNTR region is highly conserved among primates, but
not between primates and rodents. Five unrelated individuals had a 3-repeat VNTR(t,t) allele. Haplotype analysis indicates that the 9-repeat
VNTR(t,t) and the 3-repeat VNTR(t,t) alleles arose independently. However, the 9-repeat VNTR(t,t) and 6-repeat VNTR(t) alleles shared the
same haplotype, suggesting an expansion from 6(t) to 9(t,t).
D 2003 Elsevier Inc. All rights reserved.
Keywords: USH1C; Acadian; VNTR; Diversity; Minisatellite expansion; Haplotypes; Primates; Rodents
Usher syndrome type I is a rare, autosomal recessive
disorder characterized by congenital severe to profound
hearing loss, early onset of retinitis pigmentosa, and absence
of vestibular function. So far, seven different loci for Usher
syndrome type I have been mapped and five of these genes
have been identified (Hereditary Hearing Loss home page:
http://www.uia.ac.be/dnalab/hhh/). The USH1C gene is lo-
cated on chromosome 11p15.1 and is predicted to encode
several protein isoforms containing PDZ domains, suggest-
ing a role in signal transduction processes [1–3]. Families
with Usher syndrome type IC have been reported in several
countries, including Pakistan, Lebanon, and some in Europe
[2–5]. However, Usher syndrome type IC is most frequent
in the Acadian population of southwestern Louisiana, with
the founder effect explaining the relatively high frequency
of Usher syndrome type IC in this population [6,7].
Molecular analysis of Acadian Usher syndrome type IC
patients has demonstrated association with the 216G!A
mutation in exon 3 [2,8]. In addition, a purine-rich 45-bp
variable number of tandem repeat (VNTR) polymorphism
that accounts for almost the entire sequence of intron 5 of
USH1C was found by Verpy et al. [3]. They reported alleles
with 2, 3, 4, 6, and 8 repeats in a French control group with
frequencies of 0.410, 0.190, 0.015, 0.375, and 0.010, respec-
tively. The VNTR structure is characterized by the last repeat
unit containing a ‘‘T’’ nucleotide in the eighth nucleotide
position instead of the ‘‘G’’ nucleotide observed at the same
position in the previous units (VNTR(t)). However, in
Acadian Usher syndrome type IC patients the VNTR is
characterized by the presence of a 9-repeat allele with the
last two repeat units containing a ‘‘T’’ at the eighth nucleo-
tide position (9VNTR(t,t)) (Fig. 1). This 9VNTR(t,t) allele is
in complete linkage disequilibrium with the 216G!A mu-
tation in the Acadian population [8]. A 9VNTR allele was
0888-7543/$ - see front matter D 2003 Elsevier Inc. All rights reserved.
doi:10.1016/j.ygeno.2003.09.006
$Sequence data from this article have been deposited with the
GenBank Data Library under Accession Nos. AF467846–AF467854.
* Corresponding author. Fax: +1-504-568-8500.
E-mail address: bkeats@lsuhsc.edu (B.J.B. Keats).
www.elsevier.com/locate/ygeno
Genomics 83 (2004) 439–444

Page 2

also detected in an individual of Hispanic origin, but this
allele contained only one of the T-containing repeat units at
its downstream end [8].
The 216G!A mutation leads to an in-frame deletion of
13 amino acids, and the resulting USH1C transcript was
found to be unstable [2]. Verpy et al. [3] proposed that the
expanded 9VNTR(t,t) allele in Acadian Usher syndrome
type IC patients could be responsible for the disease,
possibly through interference with transcription/posttran-
scription processes. Alternatively, it may be that homozy-
gosity for both 216G!A and 9VNTR(t,t) is necessary for
disease expression or, conversely, that the association of the
9VNTR(t,t) allele with the 216G!A mutation is simply a
result of the founder effect.
To characterize the structure, diversity, and evolution of
the VNTR fully, and to gain insight into the formation of
the 9-repeat VNTR(t,t) allele in the Acadian Usher syn-
drome type IC population, we have studied DNA samples
from seven human populations (Asian, Sub-Saharan Afri-
can, African-American, Hispanic, Egyptian, Northern Eu-
ropean, and Acadian), five nonhuman primates, and four
rodents.
Results
The VNTR-length allele frequencies and heterozygosity
values in the seven populations are shown in Table 1. The 2-,
3-, and 6-repeat alleles were the most frequent, accounting
for approximately 95% of the chromosomes, with the 2-
repeat allele being the most frequent (46.3%). The 3- and the
6-repeat alleles had frequencies of 24.6 and 24.0%, respec-
tively. Comparing different populations, almost all the
alleles had 2 or 3 repeats in Sub-Saharan Africans (98.3%)
and African-Americans (93.5%), while the frequency was
lower in Egyptians (79.4%), Asians (77.5%), Hispanics
(67.3%), Acadians (60.7%), and Europeans (60.5%). The
6-repeat allele was not found in Sub-Saharan Africans and
was rare in African-Americans (2.6%), but was relatively
common in Acadians (36.6%), Europeans (33.9%), His-
panics (22.5%), Egyptians (19.0%), and Asians (15.5%).
The 4- and 8-repeat alleles were rare, accounting for collec-
tively less than 2% of the alleles. Estimated heterozygosity
values varied from 44.3% in Sub-Saharan Africans to 71.1%
in the Hispanic population.
Four new allele sizes (1, 5, 7, and 10 repeats) were
detected with frequencies of 0.002, 0.029, 0.005, and 0.001,
respectively. The 1-repeat allele was found in one Hispanic
and one Sub-Saharan African (Bantu speaker). The 7-repeat
allele was found only in the Asian population. A single
Hispanic sample was heterozygous for a 10-repeat allele, the
largest size for this VNTR detected so far. Alleles with 7 or
more repeats were rare and were not found in the African-
American, Egyptian, or Sub-Saharan African populations.
Transmission of the VNTR was analyzed in 90 parent–
offspring pairs of European and Acadian control families.
This data set included mainly 2-, 3-, and 6-repeat VNTR
alleles, and no evidence of meiotic instability in repeat size
was detected. We also followed the transmission of the
Fig. 1. Nomenclature and structure of the USH1C 45-bp VNTR. The
shaded rectangle represents a repeat unit with a ‘‘T’’ at the eighth position
and the unshaded rectangle represents a repeat unit with a ‘‘G’’ at the eighth
position.
Table 1
Repeat size distribution of VNTR alleles in seven populations
Repeat
size
European
(248)
Acadian
(186)
Hispanic
(98)
Asian
(142)
Egyptian
(58)
African-American
(78)
Sub-Saharan African
(60)
Total
(870)
1
2
3
4
5
6
7
8
9
10
0.010
0.398
0.275
0.010
0.062
0.225
0.017
0.683
0.300
0.002
0.463
0.246
0.010
0.029
0.240
0.005
0.002
0.002
0.001
0.379
0.226
0.016
0.036
0.339
0.435
0.172
0.005
0.017
0.366
0.585
0.190
0.007
0.028
0.155
0.028
0.007
0.466
0.328
0.016
0.500
0.435
0.013
0.026
0.0260.190
0.004
0.005 0.010
0.010
The sample size (number of alleles) is given in parentheses following the population name. The estimated heterozygosity values for each population are
Europeans (68.9%), Acadians (64.7%), Hispanics (71.1%), Asians (59.6%), Egyptians (63.9%), African-Americans (55.9%), and Sub-Saharan Africans
(44.3%) and for the entire population (66.6%).
S. Savas et al. / Genomics 83 (2004) 439–444
440

Page 3

9VNTR(t,t) allele in 18 Acadian Usher syndrome type IC
families (108 parent–offspring pairs) and found that it was
stably inherited.
To determine if other alleles had the VNTR(t,t) structure,
both Allele Refractory Mutation System (ARMS) (described
in [8]) and sequence analyses were performed. Two samples
(one Acadian and one Hispanic) were heterozygous for 9-
repeat alleles, and as reported previously [8], the Acadian
individual carried the 9VNTR(t,t) allele and also the
216G!A mutation, while the Hispanic individual did not
have the 216G!A mutation, and ARMS analysis produced
only a 117-bp fragment, indicating that neither allele had the
VNTR(t,t) structure. Confirmation of the VNTR(t) structure
of this Hispanic 9-repeat allele as well as seven 5-repeat
alleles (three Acadian, two European, two Asian), two 7-
repeat alleles (two Asian), and the 10-repeat allele (Hispan-
ic) was obtained by ARMS analysis of reamplified gel-
purified USH1C VNTR PCR fragments. Sequence analysis
of both the Bantu speaker and the Hispanic samples with 1-
repeat alleles demonstrated the presence of a ‘‘G’’ in the
eighth nucleotide position (1VNTR(g)). Additionally, on the
other chromosome the Bantu speaker sample had a 2-repeat
allele with ‘‘G’’ nucleotides at the eighth position in both
repeating units (2VNTR(g,g)).
ARMS analysis of 152 samples with 2-, 3-, and 6-repeat
alleles demonstrated the presence of VNTR(t,t) alleles in
four unrelated individuals. One of these samples was
Acadian, and the other three were European in origin. All
of these samples were heterozygous for a 3-repeat VNTR
allele, and genotyping of family members followed by direct
sequencing showed that the VNTR(t,t) structure was asso-
ciated with the 3-repeat VNTR allele (3VNTR(t,t) allele) in
all cases. A fifth 3VNTR(t,t) allele was detected in an
Acadian Usher syndrome type IC parent, who was hetero-
zygous for the 9VNTR(t,t) and the 3VNTR(t,t) alleles, as
well as the 216G!A mutation.
To determine if the Acadian 9VNTR(t,t) and the
3VNTR(t,t) alleles derived from each other by repeat expan-
sion/contraction, we screened these five 3VNTR(t,t) indi-
viduals for the 216G!A mutation by restriction enzyme
analysis [2] and performed haplotype analysis using 10
intragenic polymorphic sites [5] by direct sequencing (Table
2). DNA samples were available from family members of
four of these five individuals and they were analyzed to
determine the haplotype associated with the 3VNTR(t,t)
allele.
None of the samples had the 216G!A mutation, except
the parent of the Usher syndrome type IC patient. In two
European and two Acadian families, the exact haplotype
associated with the 3VNTR(t,t) allele was determined
(Table 2). Two different haplotypes (haplotypes A and B)
were found in these families. Haplotype A was identified in
one European and two Acadian families. In the remaining
European family, haplotype B, which differed from haplo-
type A only at the indel site, was found. In contrast, the
haplotype associated with the 9VNTR(t,t) allele, haplotype
C [5], differs at seven of the nine substitutional poly-
morphisms and is, as such, very different from haplotypes
A and B.
We also constructed haplotypes for 43 Acadian chromo-
somes (20 2-repeat, 7 3-repeat, and 16 6-repeat VNTR
alleles) by taking advantage of the availability of DNA
samples from family members. All of these chromosomes
had the VNTR(t) structure, and the associated haplotypes
are listed in Table 3. Eleven different haplotypes were
found. Five were associated with the 2VNTR(t) allele, with
one (haplotype B in Table 2) accounting for more than 50%
of these chromosomes. The 16 Acadian 6VNTR(t) alleles all
had the same haplotype as the 9VNTR(t,t) allele (haplotype
C in Table 2). In contrast, a different haplotype was found
for each of the seven Acadian 3VNTR(t) alleles, suggesting
that this allele is much older than the 6VNTR(t) allele.
Haplotype A was not observed in any of these chromo-
somes. As well as being found for all 6VNTR(t) and
9VNTR(t,t) alleles, haplotype C was also associated with
one of the 2VNTR(t) alleles and one of the 3VNTR(t)
alleles.
Table 2
Haplotypes obtained for the 3-repeat VNTR(t,t) and the 9-repeat VNTR(t,t)
upon analysis of 10 intragenic USH1C polymorphic sites
Polymorphism Intragenic
location
Haplotype
A
Haplotype
B
Haplotype
Ca
IVS1-61G!Ab
IVS1-47G!T
IVS1-45C!G
IVS1-41insCC
USH1C VNTR
IVS13+21C!G
IVS13?12A!Gb
IVS13?42G!A
IVS13?45A!G
IVS13?108A!G
1188A!G
The haplotype spans approximately 6 kb.
aRef. [5].
bVariants identified during this study.
Intron 1
Intron 1
Intron 1
Intron 1
Intron 5
Intron 13
Intron 13
Intron 13
Intron 13
Intron 13
Exon 14
G
G
G
?
3VNTR(t,t)
C
G
G
A
A
A
G
G
G
+
3VNTR(t,t)
C
G
G
A
A
A
A
G
C
?
9VNTR(t,t)
G
G
A
G
G
G
Table 3
Haplotypes obtained for 43 Acadian control chromosomes upon analysis of
10 intragenic USH1C polymorphic sites
Haplotype2VNTR(t)3VNTR(t) 6VNTR(t)Total
1
2
3
4
5
6
7
8
9
10
11
GGG+VCGGAAA (B)
GGG+VCGAGGG
GGG+VCGGAAG
AGC?VCGAGGG
AGC?VCAAGGG
ATG?VGGAGGG
ATG?VCGAGGG
ATG?VCAAGGG
GGG+VGGAGGG
AGC?VGGAGGG (C)
AGG?VGGAGGG
Total
11
4
3
—
—
—
—
—
—
1
1
20
—
—
—
1
1
1
1
1
1
1
—
7
—
—
—
—
—
—
—
—
—
16
—
16
11
4
3
1
1
1
1
1
1
18
1
43
V indicates the location of the VNTR within the haplotype.
S. Savas et al. / Genomics 83 (2004) 439–444
441

Page 4

To determine whether the VNTR sequence was also
present in nonhuman primates and rodents, we analyzed
five nonhuman primates (owl monkey, green monkey,
gorilla, pygmy chimp, and rhesus macaque) and four
rodents (rat, mouse, gerbil, and guinea pig) (Fig. 2). The
pygmy chimp and gorilla were found to be homozygous for
the 2VNTR(g,g) allele. The gorilla VNTR sequence was
identical to that of human, and the pygmy chimp’s sequence
differed from the human and gorilla sequences at a single
base pair. The green monkey and rhesus macaque contained
a single copy of the VNTR sequence, in which 6 nucleotides
differed from the human sequence, including an ‘‘A’’ at the
eighth position. Among the primates, the owl monkey had
the most divergent sequence; it was 81 bp long with a ‘‘G’’
at the eighth nucleotide position. Closer examination
revealed an additional 36-bp segment in the middle of the
VNTR sequence. It is not clear whether this is the ancestral
sequence that was deleted in other species or resulted from
some form of amplification in the owl monkey. Excluding
the owl monkey, the primates’ VNTR regions showed more
than 85% identity.
The DNA fragment amplified in rodents differed consid-
erably from that in nonhuman primates. Fig. 2 shows that
the guinea pig and gerbil sequences were identical, while the
mouse and rat sequences differed at nine positions. In all
organisms analyzed, there was a strong similarity in se-
quence characteristics in terms of the striking richness of G
residues.
Discussion
Minisatellite sequences (10–100 bp) are common in
eukaryotic genomes. Many studies have focused on those
that show mitotic and/or meiotic instability due to DNA
replication/recombination mechanisms, in which polarity
(the gain/loss of the repeating units at one end of the repeat
array) is commonly observed, particularly for meiotic insta-
bility [reviewed in 10,11]. It is not known why the USH1C
VNTR most downstream repeat unit contains a ‘‘T’’ nucle-
otide at the eighth position while the previous units contain
a ‘‘G’’ at the same position. However, considering the fact
that USH1C VNTR alleles with more than 2 repeat units
have almost exclusively a single repeat unit at the 3Vend with
the ‘‘T’’ nucleotide [3], it seems likely that the expansion of
the USH1C VNTR usually occurs within the more homo-
geneous ‘‘G’’-containing segments of the array and possibly
at the 5Vend. However, the Acadian USH1C 9VNTR(t,t) and
perhaps the 3VNTR(t,t) alleles may represent rare, reverse
polarization events.
Genotyping of samples originating from seven different
ethnic groups showed that the alleles varied in size from 1 to
10 repeats, with the most common allele sizes being 2, 3,
and 6 repeats. The 2-repeat VNTR allele is the most
frequent, suggesting that it is the oldest allele, although
the variation in haplotypes on the 3VNTR(t) allele suggests
that it is also much older than the 6-repeat allele. Of interest
is the rarity of 1-, 4-, 5-, and >7-repeat alleles in the samples
analyzed. Alleles with 7 or more repeats were not found in
the African-American, Egyptian, or Sub-Saharan African
populations. The 1-repeat allele was observed in only two
(one Sub-Saharan African and one Hispanic) samples. Both
of these alleles were found to have the 1VNTR(g) structure.
A 2VNTR(g,g) allele was detected in the Sub-Saharan
African sample with 1VNTR(g), and the gorilla and pygmy
chimp also had 2VNTR(g,g) alleles. Thus, the ancestral
VNTR sequence probably had a ‘‘G’’ at the eighth nucle-
otide position. In addition, because the pygmy chimpanzee
and the gorilla both have the 2VNTR(g,g), this is likely to
be the ancestral duplication state prior to the great ape
divergence. However, because we looked at only one
individual from each primate species, we cannot make
Fig. 2. Comparisons of the human VNTR sequence with those in nonhuman primates, the mouse with rat, and the guinea pig with gerbil. In the human
sequence, the ‘‘G’’ nucleotide is shown at the eighth position. The observed number of repeats (if greater than 1) is given in parentheses. An additional 36-bp
sequence present in the middle of the VNTR in the owl monkey is bracketed (* represents bases identical to the reference species).
S. Savas et al. / Genomics 83 (2004) 439–444
442

Page 5

conclusions regarding the possibility of other VNTR alleles.
The VNTR was not found in rodents, consistent with
reduced selective pressure in noncoding regions during
evolution.
Considering the fact that most human alleles have the
VNTR(t) structure, the G!T change at the eighth position
of the last repeating unit of this VNTR probably occurred
prior to migration out of Africa. The VNTR(t) structure also
indicates that expansion of this minisatellite is largely within
the region containing repeats with the ‘‘G’’ nucleotide at the
eighth position. However, the presence of the 9VNTR(t,t)
structure in Acadian Usher syndrome type IC patients [3]
and the 3VNTR(t,t) suggests occasional expansion of the
‘‘T’’-containing repeat unit.
Haplotype analyses demonstrated very different haplo-
types for the Acadian USH1C 9VNTR(t,t) allele and the
3VNTR(t,t) alleles, suggesting that they arose independent-
ly. Also, detection of an Acadian USH1C carrier with both
the 9VNTR(t,t) and the 3VNTR(t,t) alleles indicated that
homozygosity for VNTR(t,t) is not associated with the
Usher syndrome type IC phenotype. Absence of the (t,t)
structure in other VNTR alleles and the low frequency of the
3VNTR(t,t) allele (2.6% of the total and 6% of 3VNTR
alleles) suggest that the molecular events leading to the (t,t)
structure are uncommon.
Haplotypes A and B, which are associated with
3VNTR(t,t) alleles, differ only at the indel polymorphic site
in intron 1 and may be derived from the same event.
Identification of unrelated individuals for whom the
3VNTR(t,t) allele is associated with haplotype A suggests
identity by descent. Surprisingly, though, haplotype A was
not found for any of the VNTR(t) chromosomes analyzed.
Haplotype B, on the other hand, was present on 11 of 20
2VNTR(t) chromosomes. It therefore seems likely that this
3VNTR(t,t) allele is derived from a 2VNTR(t) allele by a
single expansion step at the 3Vend.
Identification of the same haplotype for all of the
9VNTR(t,t) and the Acadian 6VNTR(t) chromosomes is
consistent with the 9VNTR(t,t) allele having arisen from a
6VNTR(t) allele. However, the mechanism is not obvious;
at least two events seem necessary. In addition, the presence
of the 216G!A mutation on all 9VNTR(t,t) chromosomes,
but not on any of the 6VNTR(t) chromosomes, suggests that
multiple mutation events took place within the USH1C gene
on a chromosome containing a 6VNTR(t) allele.
Analyses of mini- and microsatellite marker loci as well
as sequence polymorphisms from both the nuclear and the
mitochondrial genomes have, in general, shown greater
variation in African populations than in non-African
populations [12–15], suggesting that the African popula-
tion is the oldest and has accumulated genetic diversity
over a longer time period. However, we found that the
VNTR in intron 5 of USH1C had a lower heterozygosity
in the Sub-Saharan African population (44.3%) than in
others (African-American, 55.9%; Asian, 59.6%; Egyptian,
63.9%; Acadian, 64.7%; European, 68.9%; Hispanic,
71.1%). In particular, the almost complete lack of larger
alleles in the African population would be consistent with
a founder effect that gave rise to larger alleles in the non-
African populations. Either because of the timing of that
large founder allele (presumably a 6-repeat allele) or
because of increased frequency of that allele due to
random drift in smaller non-African populations, the larger
alleles may be limited to the non-African population. This
model would predict that the shorter alleles are quite
stable, but that the larger alleles are less stable and are
all derived from the same founder, such as has been found
for the unstable triplet repeat sizes in the Friedreich ataxia
GAA expansion [16].
Materials and methods
DNA samples
Genomic DNA samples from 435 unrelated individuals
were analyzed: 124 Europeans (16 German, 18 British, 19
French, 73 American), 93 Louisiana Acadians, 71 Asians
(10 Chinese and 61 Thai), 39 African-Americans, 49
Hispanics, 29 North Africans (Egyptians), and 30 Sub-
Saharan Africans (4 Nigerians, 14 Zulu, 12 Bantu speakers).
We also genotyped samples from members of 17 unaffected
families (10 European, 7 Acadian) and 18 Acadian Usher
syndrome type IC families to determine if the VNTR repeat
size remains stable when the allele is transmitted from
parent to offspring. In addition, genomic DNA samples
from each of five nonhuman primates (gorilla (Gorilla
gorilla), pygmy chimp (Pan paniscus), green monkey
(Cercopithecus aethiops), owl monkey (Aotus trivirgatus),
and rhesus macaque (Macaca mulatta)), and four rodents
(mouse (Mus musculus), rat (Rattus norvegicus), gerbil
(Meriones unguiculatus), and guinea pig (Cavia porcellus))
were analyzed.
Polymerase chain reaction (PCR) and sequence analysis
The VNTR repeat size was determined following the
methods of Verpy et al. [3]. PCR-amplified fragments
containing repeat sizes that had not been previously reported
were reamplified from new DNA dilutions to confirm the
results. To determine if an allele had the VNTR(t) or
VNTR(t,t) structure, ARMS analysis, using the allele-spe-
cific primer given in Savas et al. [8], was applied to DNA
samples from 152 unrelated individuals belonging to four
different ethnic groups (32 Acadian, 20 Sub-Saharan Afri-
cans, 48 European, 52 Asian). The sequence containing the
VNTR region was amplified and the gel-eluted fragment
was either manually sequenced (1-repeat allele) or used as a
template in ARMS analysis. A 117-bp ARMS PCR product
indicates that the structure is VNTR(t), whereas a 162-bp
PCR product together with 117-bp product indicates the
VNTR(t,t) structure.
S. Savas et al. / Genomics 83 (2004) 439–444
443