Human Molecular Genetics, 2008, Vol. 17, No. 24
doi:10.1093/hmg/ddn304
Advance Access published on September 20, 2008
4012–4021
WFS1 mutations are frequent monogenic causes
of juvenile-onset diabetes mellitus in Lebanon
Pierre A. Zalloua1,2,3, Sami T. Azar2,3, Marc Delépine4, Nadine J. Makhoul1, Hervé Blanc5,
May Sanyoura1, Anne Lavergne5, Karmen Stankov6, Arnaud Lemainque4,
Patrick Baz2,7 and Cécile Julier5,6,
1
Lebanese American University, School of Medicine, Chouran, Beirut 1102 2801, Lebanon, 2Chronic Care Center,
Hazmieh, Lebanon, 3Division of Endocrinology, American University of Beirut, Bliss St, Beirut, Lebanon, 4Centre
National de Génotypage, Institut de Génomique, Commissariat à l’Énergie Atomique, 2 rue Gaston Crémieux, CP
5721, 91057 Evry Cedex, France, 5Genetics of Infectious and Autoimmune Diseases, Institut Pasteur, 75015 Paris,
France, 6Genetics of Diabetes, Inserm U730, Centre National de Génotypage, 2 rue Gaston Crémieux, CP 5721,
91057 Evry Cedex, France and 7Department of Ophthalmology, Hotel Dieu Hospital, Beirut, Lebanon
Received August 11, 2008; Revised September 8, 2008; Accepted September 16, 2008
Most cases of juvenile-onset diabetes (JOD) are diagnosed as type 1 diabetes (T1D), for which genetic studies
conducted in outbred Caucasian populations support the concept of multifactorial inheritance. However, this
view may be partly challenged in particular population settings. In view of the suggestive evidence for a high
prevalence of Wolfram syndrome (WFS) in Lebanon, the phenotypic variability associated with WFS1
mutations, and the high consanguinity rate in Lebanon, we aimed to evaluate the contribution of WFS1
mutations as monogenic determinants to JOD in Lebanon. We performed a family-based genetic study,
with linkage analysis followed by systematic mutation screening of WFS1 exons in all JOD probands. The
study population consisted of an unbiased recruitment of all juvenile-onset insulin-dependent diabetic
patients from a specialized diabetes pediatric clinic in Beirut, Lebanon. Homozygous or compound heterozygous WFS1 mutations were found in 22 of the 399 JOD probands (5.5%), resulting in WFS (17 probands) or in
non-syndromic non-autoimmune diabetes mellitus (DM, five probands). These accounted for 12.1% (21/174)
of probands in consanguineous families, compared with 0.4% (1/225) in non-consanguineous families. Of
the 38 patients identified with homozygous or compound heterozygous WFS1 mutations, 11 (29%) had
non-syndromic DM, all of whom carried a particular WFS1 mutation, WFS1 LIB, encoding a protein with an
extended C-terminal domain. This mutation resulted in a delayed onset or absence of extrapancreatic
features. These results underscore the major impact of population-specific factors, such as populationspecific mutations and founder effects, and family structure in the genetic determinism of JOD.
INTRODUCTION
Type 1 diabetes (T1D) is among the most common chronic
childhood diseases in Caucasian populations. Its etiology is
thought to be principally autoimmune, resulting in the destruction of insulin-producing pancreatic b cells, leading to complete dependence on insulin replacement therapy generally
at a young age. Epidemiologic and genetic studies support
the hypothesis of multifactorial contributions, with genetic
and environmental risk factors (1). Genes located within the
major histocompatibility complex locus are major risk
factors of T1D, and genome-wide linkage and association
scans in outbred Caucasian populations consistently support
the model of multiple additional genes with minor contribution to T1D risk (2,3). Hence, monogenic factors are
unlikely to play a major role in T1D in these populations.
However, the extent of monogenic contributions may vary
considerably between populations, depending on ethnic or
To whom correspondence should be addressed. Genetics of Diabetes, Inserm U730, Centre National de Génotypage, 2 rue Gaston Crémieux, CP 5721,
91057 Evry Cedex, France. Tel: þ33 160878330; Fax: þ33 160878383; Email: [email protected]
# The Author 2008. Published by Oxford University Press. All rights reserved.
For Permissions, please email: [email protected]
Human Molecular Genetics, 2008, Vol. 17, No. 24
population-specific factors, including founder effects and
the presence of consanguinity. They may also vary between
families, depending on their structure (presence or not of multiple affected siblings, and inter-parental consanguinity).
Several rare recessive syndromes in which insulindependent juvenile diabetes is associated with other clinical
manifestations have been described, including Wolfram
syndrome (WFS), Wolcott – Rallison syndrome and thiamineresponsive megaloblastic anemia (4). Although these syndromes are very rare in outbred Caucasian populations, their
incidence may be increased in some populations or contexts,
due to founder effects and/or the high frequency of consanguineous marriages. WFS (DIDMOAB, OMIM 222300) is
characterized by juvenile-onset diabetes (JOD) mellitus,
optic atrophy (OA), diabetes insipidus (DI) and sensorineural
deafness, and is caused by recessive mutations in the Wolframin gene (WFS1) (5,6). It is generally considered to be very
rare, with an estimated prevalence of 1/770 000 in UK (7).
However, based on previous reports, the prevalence of this
syndrome may be much higher in some countries, including
Lebanon (8). In addition to WFS, WFS1 mutations and
genetic variants have been implicated in low-frequency nonsyndromic hearing loss (dominant mutations) and psychiatric
diseases (9,10), and common WFS1 variants have recently
been shown to be associated with type 2 diabetes (T2D)
(11,12). In view of the large phenotypic variability associated
with WFS1 mutations with regard to diabetes and extrapancreatic manifestations, the possibility of some genetic
overlap between typical WFS and non-syndromic diabetes
could also be anticipated. In this study, we aimed to evaluate
the contribution of WFS1 mutations and variants in syndromic
and non-syndromic JOD in Lebanese families.
RESULTS
Strong contribution of WFS1 locus to JOD in Lebanon in
WFS and non-syndromic DM detected by linkage analysis
We studied 408 nuclear families from Lebanon, which had
been ascertained through a patient with insulin-dependent
JOD, with a total of 455 JOD patients, including nonsyndromic and syndromic cases. Twenty seven patients from
17 of these families had clinical features characteristic of
WFS (Tables 1 and 2). The high frequency of WFS patients
among Lebanese JOD probands (17/408, or 4.2%), based on
our survey, confirms the high prevalence of this syndrome in
Lebanon that had been suggested in a previous study (8).
Linkage analysis performed under a highly penetrant rare
recessive model in all multiplex, consanguineous or extended
families showed highly significant evidence of linkage
near the WFS1 locus on chromosome 4p (Fig. 1), with LOD
(logarithm of odds) scores with heterogeneity of 8.66
(estimated proportion of linked families: a ¼ 13.8%). After
excluding WFS families and the 11 neonatal or other syndromic diabetes families, there was evidence of residual linkage
at WFS1 locus (LOD ¼ 2.47, a ¼ 7.1%) (Fig. 1), suggesting
either that some WFS patients may have been underdiagnosed,
or that some WFS1 mutations may result in an incomplete
phenotype restricted to diabetes, or that frequent variants in
4013
the WFS1 gene or another gene near WFS1 may contribute
to diabetes susceptibility in Lebanon.
WFS1 mutations in WFS and non-syndromic DM patients
By sequencing WFS1 exons in WFS and non-syndromic DM
probands and their parents from multiplex, consanguineous
or extended families, and in probands from simplex nonconsanguineous families, we identified 173 WFS1 DNA
variants, 46 of which were predicted to affect the primary
sequence of the protein, 38 of which were novel (Supplementary Material, Table S2).
In the WFS patients, we identified 10 variants compatible
with the status of WFS mutations, nine of which were novel
(Table 1). All the WFS patients carried a homozygous
mutation (or a combination of two mutations in complete
linkage disequilibrium) that was predicted to have deleterious
consequences on the protein function, as it resulted in a truncated protein, deleted one amino acid or was predicted to be
damaging based on PolyPhen analysis. A double mutation,
WFS1 LIB, associating the 707VI non-synonymous change
and the F884fs951X frameshift on the same haplotype, was
homozygous in 7/17 (41.2%) of the WFS probands, and
10/27 of all WFS patients. F884fs951X frameshift affects
the most C-terminal endoplasmic reticulum (ER) luminal
domain, replacing the last seven amino acids by 67 other
amino acids, while 707VI is predicted to have no major
functional consequence on protein function.
In non-syndromic DM patients, two variants were compatible
with the status of mutations, based on our criteria (Methods):
WFS1 LIB and F646fs708X (Supplementary Material,
Table S2). WFS1 LIB was homozygous in four non-syndromic
DM probands (10 non-syndromic DM patients) and compound
heterozygous with F646fs708X in one non-syndromic DM
proband (Table 2). Re-analysis of linkage data after excluding
the corresponding families cancelled the residual linkage
peak near WFS1, indicating that WFS1 mutations in WFS and
non-syndromic DM patients accounted for the totality of the
evidence of linkage detected at WFS1 locus (Fig. 1).
WFS1 LIB mutation is associated with delayed
or absent extrapancreatic manifestations
Of the 38 JOD patients from our study population diagnosed
with homozygous or compound heterozygous WFS1
mutations, 11 (27%) had non-syndromic DM; 11/21 (52.4%)
of the patients homozygous or compound heterozygous for
WFS1 LIB mutation had only DM, without any evidence of
OA, DI or deafness, compared with none (0/17) of the patients
homozygous for any other mutation (Table 2). Using logistic
regression analysis, we showed that the absence of extrapancreatic manifestations was dependant on the age of the patients
at last examination (P ¼ 0.036) and on the nature of the
mutation (P ¼ 0.0001) (Table 3A). In a Kaplan –Meier analysis, we showed that the appearance of OA was significantly
delayed in patients homozygous for the WFS1 LIB mutation
compared with those homozygous for other WFS1 mutations
(Fig. 2, P ¼ 0.0003). Patients homozygous for the WFS1 LIB
mutation exhibited reduced expression of other extrapancreatic features, including hearing impairment, DI, cataract and
4014
Exon
Nucleotide
changea
Nature of
change
Reference
sequence
Mutation
Amino acid
change
Protein
domainb
PolyPhen
predictionc
Amino acid
conservation
Reference
Number of probands with
homozygous mutations (total
number of patients) d
Exon4
Exon8
Exon8
Exon8
Exon8
490
1552
2694
1752
2106
SNP
SNP
SNP
SNP
Del(8 bp)
G
C
C
T
TGCTGTTC
A
G
T
G
–
107GE
461TSe
842LFe
528YD
F646fs708X
Damaging
Benign
Damaging
Damaging
NA
Yes
no
Yes
Yes
NA
This study
2 (2)
This study o 2 (5)
This study
This study
1 (1)
This study
2 (4)g
Exon8
Exon8
Exon8
2238
2289
2819
SNP
SNP
Del(C)
T
G
C
G
A
–
690CG
707VIf
F884fs951Xf
Damaging
Benign
NA
Yes
Yes
NA
Exon8
Exon8
2637–2639
2716–2717
Del(ATC)
Ins(22 bp)
ATC
–
–
CTGCATGGCCC
AGCTCTCGCCC
823I/del
T857fs946X
Out
Out
In
TM
Frameshift/
truncation
In
In
In, frameshift/
extension
In
In, frameshift/
extension
NA
NA
NA
NA
This study
1 (1)
)
This study
7 (10)g
Hansen
(2005)
This study
1 (3)
This study
1 (1)
WFS, Wolfram syndrome; DM, diabetes mellitus; NA, not available. For references to known mutations, see Supplementary Material, Table S2.
a
Position on NM_006005.2.
Out: cytoplasmic domain, In: ER domain, TM: transmembrane domain.
c
PolyPhen prediction was used to assess the consequence of missense mutations (see Materials and Methods). We indicated as ‘damaging’ mutations scored as ‘possibly’ or ‘probably’ damaging by this
program.
d
All the mutations identified in WFS patients were homozygous.
e
Mutations 461TS and 842LF were in complete LD and systematically associated with the same haplotype (461TS þ 842LF).
f
Mutations 707VI and F884fs951X were in complete LD and systematically associated with the same haplotype (707VI þ F884fs951X, referred to as WFS1 LIB in the text).
g
The double mutation 707VI þ F884fs951X was also found to be homozygous in 10 non-syndromic DM patients and compound heterozygous (together with F646fs708X) in one non-syndromic patient (see
text and Table 2).
b
Human Molecular Genetics, 2008, Vol. 17, No. 24
Table 1. WFS1 mutations identified in Lebanese WFS patients
Human Molecular Genetics, 2008, Vol. 17, No. 24
4015
Table 2. Clinical characteristics and nature of WFS1 mutations in Lebanses WFS and non-syndromic DM patients carrying WFS1 mutations
Family
Patient
Clinical
diagnosis
Mutationa
Age at
onset of
diabetes
Age at last
examination
OA
Age at
onset
of OA
Deafness
Age at
onset of
deafness
DI
Age at
onset
of DI
Cataract
Epilepsy
LIB1
LIB1
LIB1
LIB1
LIB2
LIB2
LIB3
LIB3
LIB4
LIB5
LIB5
LIB6
LIB6
LIB6
LIB7
LIB7
LIB8
LIB9
LIB9
LIB10
LIB11
LIB12
LIB13
LIB14
LIB14
LIB15
LIB16
LIB17
LIB18
LIB18
LIB18
LIB19
LIB19
LIB19
LIB20
LIB20
LIB21
LIB22
1
2
3
4
1
2
1
2
1
1
2
1
2
3
1
2
1
1
2
1
1
1
1
1
2
1
1
1
1
2
3
1
2
3
1
2
1
1
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFS
WFSb
DM
WFS
WFS
WFSb
DM
DM
DM
DM
DM
DM
DM
DM
DM
DM
461TS þ 842LF
461TS þ 842LF
461TS þ 842LF
461TS þ 842LF
707VI þ F884fs951X
707VI þ F884fs951X
F646fs708X
F646fs708X
107GE
707VI þ F884fs951X
707VI þ F884fs951X
823I/del
823I/del
823I/del
F646fs708X
F646fs708X
461TS þ 842LF
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X
107GE
690CG
528YD
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X
T857fs946X
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X
707VI þ F884fs951X/
F646fs708X
4
6
7
6
7
10
5
3
4
7
4
2
1
5
7
5
4
7
8
6
11
5
9
3
3
7
5
6
7
5
5
3
7
5
8
4
7
6
29
31
33
24
32
24
19
15
23
29
21
16
13
11
16
9
16
20
19
22
21
19
33
12
8
15
12
13
14
12
9
23
20
16
16
7
23
12
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
No
No
No
No
No
No
No
No
No
No
8
21
11
9
23
17
12
6
5
19
13
7
2
7
11
5
12
15
9
15
12
10
24
11
–
8
6
8
–
–
–
–
–
–
–
–
–
–
No
Yes
Yes
Yes
No
Yes
Yes
No
Yes
No
No
No
No
No
No
No
No
No
No
No
Yes
No
Yes
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
–
24
20
–
–
17
8
–
1
–
–
–
–
–
–
–
–
–
–
–
13
–
24
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
No
Yes
Yes
No
No
No
Yes
No
Yes
Yes
No
No
No
No
Yes
No
Yes
No
No
Yes
Yes
No
Yes
No
No
Yes
No
No
No
No
No
No
No
No
No
No
No
No
–
24
18
–
–
–
12
Yes
Yes
No
No
Yes
No
No
No
Yes
No
No
No
Yes
No
No
No
No
No
No
No
No
No
Yes
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Yes
No
No
No
No
Yes
Yes
No
No
No
No
No
No
No
No
No
Yes
No
No
No
No
Yes
No
No
No
No
No
No
No
No
No
No
Yes
No
2
24
–
–
–
–
13
–
12
–
–
21
14
–
25
–
–
7
–
–
–
–
–
–
–
–
–
–
–
–
WFS, Wolfram syndrome; DM, diabetes mellitus; OA, optic atrophy; DI, diabetes insipidus. For references to known mutations, see Supplementary
Material, Table S2.
a
With the exception of LIB22-1 patient, all the mutations were found in the homozygous status.
The initial diagnosis of these patients was DM, and was revised to WFS as OA developed during the course of this study.
b
neuropsychiatric complications, with the possible exception of
epilepsy, showing the opposite trend (Table 3A); in contrast,
ages at onset of diabetes were similar in both groups
(WFS1 LIB: 5.95, others: 5.24, P ¼ 0.32) (Table 3B). With
the exception of epilepsy reported in one patient, nonsyndromic DM patients with WFS1 mutations had no extrapancreatic features, at age up to 23 years. Glutamic acid decarboxylase (GAD) autoantibody level was negative in all
patients with WFS1 mutations tested, seven WFS patients
and two non-syndromic DM patients, supporting that the
disease process is non-autoimmune in all cases.
Among WFS1 LIB homozygous individuals, the risk of
having WFS (compared with non-syndromic DM) was
family dependent (P ¼ 0.002, after correcting for the age at
last examination). This observation could not be accounted
for by other WFS1 variants, since the entire WFS1 sequence
was identical in all the WFS1 LIB haplotypes. This suggests
the contribution of additional factors to the clinical outcome
of WFS1 LIB patients, which may be variants located in
flanking regions of the WFS1 gene. Alternatively, this could
be explained by other genetic or environmental factors that
may be shared within families; the role of a differential diagnosis by family is unlikely, since all patients were examined
by the same ophthalmologist.
Impact of family structure on the risk of
WFS1-monogenic contribution
Based on our sequence screening survey, 5.5% (22/399) of all
Lebanese JOD probands (including all WFS and non-WFS
syndromic cases) were homozygous or compound heterozygous for WFS1 mutations. We tested the impact of family
4016
Human Molecular Genetics, 2008, Vol. 17, No. 24
Figure 1. Linkage analysis of JOD in consanguineous, multiplex or extended Lebanese families: chromosome 4p linkage (WFS1 locus). Parametric linkage
analysis in all consanguineous, multiplex or extended families (solid line), after excluding WFS families and 11 neonatal or other syndromic diabetes families
(interrupted line), and after excluding these and other families with homozygous or compound heterozygous WFS1 mutations identified in patients (dotted line).
structure, namely parental consanguinity and multiplicity of
diabetic siblings, on the frequency of monogenic determinism
due to WFS1 mutations. As expected, the frequency of WFS1
mutations as a causal determinant of JOD (i.e. homozygous or
compound heterozygous status for WFS1 mutations) was
increased in probands with consanguineous parents, reaching
12.1% (21/174) overall, compared with 0.4% (1/225) in nonconsanguineous simplex families. This risk reached 46.2%
(12/26) in multiplex consanguineous families, compared
with 6.1% (9/148) in simplex consanguineous families.
The WFS1 LIB mutation was found to be prevalent in
Lebanese JOD. Although all other WFS1 mutations in our
study were absent in non-transmitted parental alleles to WFS
and non-syndromic DM probands (N ¼ 395 alleles),
WFS1 LIB mutation was present in one non-transmitted parental
allele (allele frequency: 0.0025). This mutation was absent in
1000 Lebanese population Controls studied. Its presence in
non-transmitted Control alleles is likely to be a consequence
of familial endogamy.
There were no cases of T1D in parents of WFS or nonsyndromic DM patients with WFS1 mutations (i.e. parents
who are obligate WFS1 mutation carriers, all mutations
combined). Comparable risk of T2D was observed among
parents of patients with or without WFS1 mutations, after
adjusting for age and sex (6.4 versus 6.3%, respectively,
data not shown). Hence heterozygous status for these WFS1
mutations does not significantly affect the risk of T1D or
T2D. Conversely, there was no distorted transmission of
alleles at the T2D-associated common SNPs rs10010131 and
rs734312 (611RH) to JOD patients in our study (Supplementary Material, Table S2).
DISCUSSION
WFS1 encodes wolframin, a transmembrane glycoprotein
localized to the ER, with high expression in pancreatic b
cells, neurons and some other tissues (13), where it regulates
Ca2þ homeostasis (14). To date, WFS1 mutations have been
implicated in WFS, low-frequency non-syndromic hearing
loss and psychiatric diseases, and common variants have
recently been shown to be associated with T2D. We now
show evidence that some WFS1 mutations may result in nonsyndromic JOD, and that WFS1 mutations are responsible for a
large proportion of JOD in some population subgroups, reaching 12.1% of probands in Lebanese consanguineous families.
This situation results from the combination of populationspecific founder effects responsible for the high prevalence
of WFS1 LIB mutation in Lebanese patients, the non-syndromic
DM phenotype frequently associated with this mutation, and
the high rate of consanguinity. Intra-population stratification
further increases this effect, as all WFS1 LIB patients belonged
to the same socio-cultural community, although they were
geographically dispersed over Lebanon (data not shown).
Similar clustering of homozygous mutations has been previously reported in Lebanon for b-thalassemia (b-globin
gene) and familial Mediterranean fever (MEFV gene)
Human Molecular Genetics, 2008, Vol. 17, No. 24
4017
Table 3. Effect of the nature of the mutation on clinical manifestations of WFS
(A) Extrapancreatic clinical manifestations in patients homozygous for WFS1 mutations
Trait
Optic atrophy
Hearing impairment
Diabetes insipidus
Cataract
Epilepsy
Neuropsychiatric complicationsb
P-valuea
Number positive for the trait (%)
WFS1 LIB (N ¼ 20)
Other (N ¼ 17)
10 (50%)
1 (5%)
3 (15%)
1 (5%)
5 (25%)
3 (15%)
17 (100%)
7 (41.2%)
8 (47.1%)
5 (29.4%)
1 (5.9%)
9 (52.9%)
0.0001
0.008
0.06
0.08
0.03
0.01
(B) Age at onset of diabetes
Phenotype (number)
Homozygous for WFS1 LIB mutation;
numbers, mean (95% CI)
Homozygous for ‘other’ mutations;
numbers, mean (95% CI)
P-valuec
WFS (N ¼ 27)
DM (N ¼ 10)
All (N ¼ 37)
N ¼ 10, 6.50 (5.03–7.97)
N ¼ 10, 5.40 (4.13–6.67)
N ¼ 20, 5.95 (4.97–6.93)
N ¼ 17, 5.24 (4.11–6.36)
N¼0
N ¼ 17, 5.24 (4.18–6.29)
0.17
0.32
Notes: All individuals were homozygous for WFS1 mutations, either WFS1 LIB, or all others grouped (other). We excluded one compound heterozygous
WFS1 LIB/other from this analysis.
CI, confidence interval.
a
Logistic regression test (likelihood ratio test), taking into account the age at last examination.
b
Neuropsychiatric complications were: memory loss, disorientation to time, place or person, delusions or hallucinations, or confusion.
c
ANOVA test comparing patients homozygous for WFS1 LIB with those homozygous for other mutations.
(15,16), providing strong evidence for the combined impact of
consanguinity and founder effect, resulting from the religious
and socio-cultural ethnic endogamy in the Lebanese population. Hence, the impact of WFS1 mutations on JOD is
strongly dependent on population-specific factors, ranging
from almost none in outbred Caucasians patients (7), consistent with the absence of genome-wide significant linkage and
SNP association observed at WFS1 locus in outbred Caucasian
populations (16,17), to very high in Lebanese consanguineous
families, and especially those with multiple affected siblings.
WFS1 mutations may also have a strong impact in other populations or patients’ subgroups with high frequency of autoantibody negative T1D. Interestingly, association of common
population-specific WFS1 variants with T1D has been reported
in Japan (18), although the role of monogenic determinants
has not been demonstrated in this population.
Although we were able to show that WFS1 LIB mutation is
associated with significant delay in the appearance of OA,
compared with other WFS1 mutations, a longer follow-up of
these patients, or a specific study of adult patient populations,
would be needed in order to determine if a subset of the nonsyndromic WFS1 LIB patients may be exempted from extrapancreatic manifestations during their lifetime. Remarkably, some
of these patients were still with non-syndromic DM, with no
sign of OA and full visual acuity, up to 23 years of age.
A mild WFS phenotype has been previously described in
two siblings homozygous for F884fs951X, included in the
WFS1 LIB mutation (707VI mutation was newly identified in
our study and was not described in these patients), who had
DM and OA, but no deafness, DI or neurologic abnormalities
(19). A compound heterozygous F884fs951X mutation has
also been reported in another WFS patient who had no deafness or DI (20). WFS patients carrying a homozygous 4 bp
deletion with similar protein consequences (F883fs950X)
also had no DI or deafness (21,22). All these patients had
OA and had been diagnosed as having WFS. Two hypotheses
may explain the selective mild WFS/non-syndromic DM features of WFS1 LIB mutation: (i) b cells may be more sensitive
than other cells to the pathway affected in WFS; b cells are
highly secreting cells, which makes them particularly sensitive
to ER stress (23), a proposed mechanism by which WFS1
mutations may lead to WFS (24,25). This may result in an
obligate diabetes phenotype, but no or delayed extrapancreatic
features in case of ‘mild’ mutations; (ii) the C-terminal region
of wolframin affected in the WFS1 LIB mutation may be critical
for b-cell-specific function. Recently, this region has been
shown to interact with Naþ/Kþ ATPase b1 subunit, which
is critical for b cell function and may involve mechanisms
that are partly independent of ER stress pathways (26).
Evidence for genotype/phenotype correlation on the clinical
variability of WFS was shown in a recent meta-analysis, with a
dose–effect association of the number of inactivating mutations
on disease severity (20). However, the effect of specific mutations
on this variability could not be detected in the previous study,
owing to the large variety of mutations. In addition, the
WFS1 LIB mutation, which would have been considered as an
inactivating mutation in this study (frameshift) is actually associated with a milder disease phenotype. Additional molecular
studies in patients and Controls cell lines will be needed to determine the underlying mutation-dependant mechanisms resulting
in variable clinical expression.
The full WFS or non-syndromic DM expression of WFS1 LIB
showed familial clustering, suggesting the contribution of
additional modifier factors, most likely genetic; Wfs1-deficient
mice appear to have a ‘milder’ disease than human WFS (27),
as extrapancreatic features have not been described to date,
4018
Human Molecular Genetics, 2008, Vol. 17, No. 24
Figure 2. Effect of WFS1 LIB mutation on the appearance of OA. All individuals included in this analysis were homozygous for a mutation in the WFS1
gene, either WFS1 LIB or another one (grouped under ‘other’). We did not
include in this analysis the unique patient who was a compound heterozygous
for WFS1 mutations, heterozygous for WFS1 LIB, who had no sign of OA at 12
years of age. Arrows indicate two patients homozygous for ‘other’ mutations
who had a notably late onset of OA (age 21 years: mutation 461TS þ 842LF,
and age 24 years: mutation 528YD).
further supporting this hypothesis. Some studies have suggested
a role of mitochondrial mutations in WFS (28). In particular the
possible role of Leber hereditary optic neuropathy mutations in
WFS has been proposed (29). We did not identify significant
differences between the entire mitochondrial sequence of full
WFS and non-syndromic DM WFS1 LIB patients that may
explain this effect (data not shown); however, the small
number of independent homozygous WFS1 LIB patients provided
limited power to this analysis.
Our findings extend the role of WFS1 mutations as a
frequent monogenic determinant of JOD in some population subgroups and suggest the existence of significant genetic
heterogeneity between patients with JOD, strongly dependent
on populations and family structure. To our knowledge,
our observation of non-syndromic DM caused by a specific recessive WFS1 mutation constitutes the first description of recessive
monogenic inheritance in diabetes which is not syndromic or neonatal. Based on these results, we anticipate that in some populations, or in particular subgroups of patients, monogenic
causes of diabetes may largely exceed the estimation of 1–2%,
with important consequences for diagnosis and patients management (30–32). The increasing genetic knowledge on these monogenic forms of diabetes suggests the need for some revision of the
current diabetes classification, to integrate genetic characteristics,
when applicable. Generally, our results highlight the need to
consider population origin and family structure, in addition to ethnicity, in genetic studies of diabetes and other multifactorial diseases. We anticipate that monogenic determinants, which remain
undetected in the general multifactorial background in outbred
populations, may contribute significantly to multifactorial diseases in specific situations.
MATERIALS AND METHODS
Patients and families
We recruited 408 nuclear families through a proband with
JOD, with a total of 455 patients and 1520 of their family
members, who were followed at the Chronic Care Center
(Beirut, Lebanon), a pediatric clinic. The Chronic Care
Center is the country’s sole professional center for the
treatment of JOD, where 90% of patients from the various
communities and social structures from Lebanon come for
treatment. All patients and their families visiting the center
between January 2001 and March 2003 were asked to participate in the current study, by answering a questionnaire and
providing a blood sample for DNA studies (establishment of
cell lines was not included in the study protocol). All patients
and family members, or their guardians in case of children,
who accepted to participate, gave their written informed
consent after the study was explained to them. The study protocol was approved by the Research and Ethics Committee of
the Chronic Care Center.
All patients included in the study were with ketoacidosis or
ketosis with severe symptoms of acute onset at presentation
and continuous dependence on insulin within 6 months of
diagnosis, fulfilling the criteria for T1D according to the
World Health Organization criteria (33). Patients from four
families were suspected of Maturity Onset Diabetes of the
Young on the basis of incomplete insulin-dependency and
dominant transmission, and these families were excluded
from this study. A secondary inclusion criterion for this
study was that families had to have both parents, or one
parent and at least one sibling, available for DNA study. All
but 10 families had both parents available. In addition to the
medical examination and routine biochemical analyses, a standardized medical history questionnaire and demographic
information were obtained on all patients. Age at onset of diabetes was before 18 and 26 years for probands and all patients
respectively (mean age at onset of all patients: 9.7 years). Epidemiologic characteristics of an earlier subset of this collection have been reported (34,35). Patients presenting with
other clinical manifestations in addition to diabetes mellitus
(DM) that may reveal known or unknown syndromic forms
of diabetes, were not excluded, and these additional clinical
manifestations were recorded. Hence this recruitment provides
an unbiased representation of all patients diagnosed with JOD
in Lebanon, and may include increased clinical heterogeneity
compared with ‘typical’ T1D.
DNA was extracted from blood collected on EDTA, using
standard procedures. The presence of anti-GAD autoantibodies was tested at disease onset or retrospectively in a
subset of 97 consecutive diabetic patients, using the
GAD-AB radioimmunoassay kit (RSR, Limited, Cardiff,
UK) following recommended procedures. Positive values
(.1.0 U/ml) were observed in 68% of patients tested within
5 years after disease onset. Patients were medically evaluated
at 1- or 2-month-interval visits, with clinical examinations by
the Center’s endocrinologist, dietician and clinical psychologist. Fundoscopic examination and other routine ophthalmic tests were performed by a specialized ophthalmologist
twice a year, and every 2 months for WFS patients. Diagnosis
of WFS was based on the presence of OA confirmed by the
presence of white papilla with regular and well-demarcated
borders, in addition to DM. All the diabetic patients who
were discovered to have a mutation in the wolframin gene
(WFS1) during the course of this study were re-examined
and evaluated by the ophthalmologist and the endocrinologist
Human Molecular Genetics, 2008, Vol. 17, No. 24
for any clinical symptoms of WFS every 2 months following
genetic diagnosis, and throughout the duration of the study.
Information on consanguinity between probands’ parents was
obtained through a questionnaire, and statistically re-evaluated
based on genome-wide microsatellite genotyping done on all
multiplex, consanguineous or extended families, and on 40
additional probands from non-consanguineous families and
their parents. Information on consanguinity obtained by
questionnaire and inferred from microsatellite data were
highly concordant. In particular, no consanguinity was detected
in the 40 non-consanguineous probands in this analysis. Frequencies of first and all degree consanguineous marriages
between probands’ parents were 25.7 and 43.4% respectively
(general population estimates: 14 and 25%, respectively, 36).
Families were classified as consanguineous, multiplex
(multiple affected siblings), extended (multiple related
nuclear families), or simplex (single affected child) nonconsanguineous. Of a total of 408 nuclear families, 220
belonged to consanguineous, multiplex or extended families,
and 192 were consanguineous or multiplex. Several non-typical
T1D patients were identified based on clinical diagnosis or/and
subsequent genetic studies and re-examination of patients. These
included 17 families with WFS, and 11 other families where
patients had neonatal diabetes or were affected by additional
extrapancreatic clinical manifestations, suggesting recessive
monogenic inheritance. A role of WFS1 gene has been excluded
in these 11 families, based on linkage analysis, WFS1 mutation
screening and/or direct gene identification of the causative
defect. These 11 families were included in our study in order
to achieve an exhaustive prevalence survey of genetic causes
of JOD in Lebanon.
One thousand unrelated consenting healthy adults (.40
years) representative of the socio-cultural diversity of the
Lebanese population were used as population Controls.
These individuals are part of an ongoing larger population
study in Lebanon (37).
Microsatellite genotyping in WFS1 gene region
We combined microsatellite genotyping data from a microsatellite genome scan performed in all consanguineous,
multiplex or extended families, excluding those initially
diagnosed as having WFS, and additional genotyping of
three microsatellite markers located near WFS1 gene,
D4S431, D4S2366 and D4S2935, was performed in all these
families, including those with WFS. The microsatellite
genome scan used .400 microsatellite markers (Linkage
Mapping Set 2, Applied Biosystems, with modifications).
Mutation and polymorphisms screening in WFS1 gene
Screening of WFS1 mutations and polymorphisms was performed by DNA sequencing in all WFS probands (N ¼ 17)
and their parents (N ¼ 32), in all non-syndromic diabetic probands except nine due to technical failure (N ¼ 371) and all
available parents from multiplex, consanguineous or extended
families (N ¼ 363). We amplified all WFS1 exons, including
intron/exon boundaries, from genomic DNA by PCR using
primers shown in Supplementary Material, Table S1A.
Sequence interpretation was performed using the Genalys soft-
4019
ware (38). After identification of a prevalent WFS1 Lebanese
mutation (WFS1 LIB), we genotyped it in duplicate in all study
individuals, using a TaqMan assay (Applied Biosystems), with
primers described in Supplementary Material, Table S1B.
There was 100% concordance between sequencing and genotyping results. We also screened 1000 unrelated Lebanese
Controls for this mutation using the same TaqMan assay.
Defining the mutation status of variants, in WFS
and non-syndromic DM patients
The mutation status of variants was assessed based on WFS1
sequence data generated on patients (WFS: N ¼ 17 and nonsyndromic DM: N ¼ 371) and their parents (N ¼ 32 and 363
respectively). WFS1 variants were considered to be compatible
with WFS mutations if they satisfied simultaneously the following four criteria: (i) being homozygous or compound heterozygous with other mutations in WFS patients; (ii) never being
homozygous or compound heterozygous with other mutations
in non-diabetic individuals (parents); (iii) being absent (or
present at a very low allele frequency, in the case of more
prevalent mutations) in parental alleles that were not transmitted
to diabetic children (Control alleles); (iv) predicted to affect the
coding sequence of the protein (coding variant or alteration of
splice site). We used the same strategy to assess WFS1 mutation
status for non-syndromic DM (test population: non-syndromic
DM probands). In addition to the Control alleles from the
study population (non-transmitted parental alleles), a large
Control population (N ¼ 1000) was used to assess the frequency
of a more prevalent WFS1 mutation in the Lebanese population.
Prediction of the functional consequences
of WFS1 missense variants
Missense variants that were compatible with WFS or nonsyndromic DM mutations, based on the criteria mentioned
earlier, were evaluated for their predicted impact on protein
function using the PolyPhen computational program (39).
We also assessed the species-conservation of amino acids
by sequence alignment. Scores of ‘possibly damaging’ or
‘probably damaging’ given by this program for a conserved
amino acid were estimated to be likely deleterious to protein
function.
Linkage and association studies, and statistical analyses
We revised the family structures and corrected genotyping
errors resulting in Mendelian inconsistencies, using the PEDCHECK program (40). Family structures and consanguinity
estimates were determined using the IBSCHECK program
(S. Heath, unpublished). Linkage analyses were performed
under a highly penetrant recessive model (penetrance: 0.95,
disease allele frequency: 0.001, no phenocopy), allowing for
heterogeneity, using MERLIN (41). Complex or extended
pedigrees were simplified when necessary to enable computation with this program, while maintaining the necessary consanguinity information. Hardy – Weinberg equilibrium testing,
and family-based association and transmission disequilibrium
test studies were performed using the PLINK package (42).
Additional statistical analyses, including logistic regression
4020
Human Molecular Genetics, 2008, Vol. 17, No. 24
analyses and Kaplan – Meier survival analysis were done using
STATVIEW or JMP7 statistical packages.
11.
SUPPLEMENTARY MATERIAL
Supplementary Material is available at HMG Online.
12.
FUNDING
13.
This study was supported by institutional support from Inserm
and the Pasteur Institute of Paris, and a joined grant from the
Juvenile Diabetes Research Foundation, the Fondation pour la
Recherche Médicale and Inserm to C.J. We thank the Hospices
Civils de Lyon for their support.
14.
15.
ACKNOWLEDGEMENTS
We thank the patients and their families for agreeing to participate in this study. We thank Valérie Senée and Anne Boland
for their help with DNA samples management, and Fumihiko
Matsuda and Victor Renault for their help in database integration. We thank the Chronic Care Centre, Beirut, Lebanon,
for their collaboration and support. We are very grateful to
Douglas Cavener, Bernhard Boehm and Marc Nicolino for
discussions and critical reading of the manuscript.
16.
17.
18.
Conflict of Interest statement. None declared.
19.
REFERENCES
1. Rich, S.S. (1990) Mapping genes in diabetes. Genetic epidemiological
perspective. Diabetes, 39, 1315– 1319.
2. Concannon, P., Erlich, H.A., Julier, C., Morahan, G., Nerup, J., Pociot, F.,
Todd, J.A. and Rich, S.S. (2005) Type 1 diabetes: evidence for
susceptibility loci from four genome-wide linkage scans in 1,435
multiplex families. Diabetes, 54, 2995–3001.
3. Todd, J.A., Walker, N.M., Cooper, J.D., Smyth, D.J., Downes, K.,
Plagnol, V., Bailey, R., Nejentsev, S., Field, S.F., Payne, F. et al. (2007)
Robust associations of four new chromosome regions from genome-wide
analyses of type 1 diabetes. Nat. Genet., 39, 857–864.
4. Barrett, T.G. (2007) Differential diagnosis of type 1 diabetes: which
genetic syndromes need to be considered? Pediatr. Diabetes, 8, 15–23.
5. Inoue, H., Tanizawa, Y., Wasson, J., Behn, P., Kalidas, K.,
Bernal-Mizrachi, E., Mueckler, M., Marshall, H., Donis-Keller, H.,
Crock, P. et al. (1998) A gene encoding a transmembrane protein is
mutated in patients with diabetes mellitus and optic atrophy (Wolfram
syndrome). Nat. Genet., 20, 143–148.
6. Strom, T.M., Hortnagel, K., Hofmann, S., Gekeler, F., Scharfe, C., Rabl, W.,
Gerbitz, K.D. and Meitinger, T. (1998) Diabetes insipidus, diabetes mellitus,
optic atrophy and deafness (DIDMOAD) caused by mutations in a novel
gene (wolframin) coding for a predicted transmembrane protein. Hum. Mol.
Genet., 7, 2021–2028.
7. Barrett, T.G., Bundey, S.E. and Macleod, A.F. (1995) Neurodegeneration
and diabetes: UK nationwide study of Wolfram (DIDMOAD) syndrome.
Lancet, 346, 1458–1463.
8. Medlej, R., Wasson, J., Baz, P., Azar, S., Salti, I., Loiselet, J., Permutt, A.
and Halaby, G. (2004) Diabetes mellitus and optic atrophy: a study of
Wolfram syndrome in the Lebanese population. J. Clin. Endocrinol.
Metab., 89, 1656– 1661.
9. Khanim, F., Kirk, J., Latif, F. and Barrett, T.G. (2001) WFS1/Wolframin
mutations, Wolfram syndrome, and associated diseases. Hum. Mutat., 17,
357– 367.
10. Cryns, K., Sivakumaran, T.A., Van den Ouweland, J.M., Pennings, R.J.,
Cremers, C.W., Flothmann, K., Young, T.L., Smith, R.J., Lesperance,
M.M. and Van Camp, G. (2003) Mutational spectrum of the gene in
20.
21.
22.
23.
24.
25.
26.
27.
Wolfram syndrome, nonsyndromic hearing impairment, diabetes mellitus,
and psychiatric disease. Hum. Mutat., 22, 275–287.
Sandhu, M.S., Weedon, M.N., Fawcett, K.A., Wasson, J., Debenham,
S.L., Daly, A., Lango, H., Frayling, T.M., Neumann, R.J., Sherva, R. et al.
(2007) Common variants in confer risk of type 2 diabetes. Nat. Genet., 39,
951–953.
Franks, P.W., Rolandsson, O., Debenham, S.L., Fawcett, K.A., Payne, F.,
Dina, C., Froguel, P., Mohlke, K.L., Willer, C., Olsson, T. et al. (2008)
Replication of the association between variants in WFS1 and risk of type 2
diabetes in European populations. Diabetologia, 51, 458–463.
Hofmann, S., Philbrook, C., Gerbitz, K.D. and Bauer, M.F. (2003)
Wolfram syndrome: structural and functional analyses of mutant and
wild-type wolframin, the WFS1 gene product. Hum. Mol. Genet., 12,
2003– 2012.
Osman, A.A., Saito, M., Makepeace, C., Permutt, M.A., Schlesinger, P.
and Mueckler, M. (2003) Wolframin expression induces novel ion channel
activity in endoplasmic reticulum membranes and increases intracellular
calcium. J. Biol. Chem., 278, 52755– 52762.
Makhoul, N.J., Wells, R.S., Kaspar, H., Shbaklo, H., Taher, A., Chakar, N.
and Zalloua, P.A. (2005) Genetic heterogeneity of Beta thalassemia in
Lebanon reflects historic and recent population migration. Ann. Hum.
Genet., 69, 55–66.
Jalkh, N., Genin, E., Chouery, E., Delague, V., Medlej-Hashim, M.,
Idrac, C.A., Megarbane, A. and Serre, J.L. (2008) Familial Mediterranean
fever in Lebanon: founder effects for different MEFV mutations. Ann.
Hum. Genet., 72, 41– 47.
Qu, H.Q., Grant, S.F., Bradfield, J.P., Kim, C., Frackelton, E.,
Hakonarson, H. and Polychronakos, C. (2008) Association analysis of the
type 2 diabetes loci in type 1 diabetes. Diabetes, 57, 1983– 1986.
Awata, T., Inoue, K., Kurihara, S., Ohkubo, T., Inoue, I., Abe, T., Takino,
H., Kanazawa, Y. and Katayama, S. (2000) Missense variations of the
gene responsible for Wolfram syndrome (WFS1/wolframin) in Japanese:
possible contribution of the Arg456His mutation to type 1 diabetes as a
nonautoimmune genetic basis. Biochem. Biophys. Res. Commun., 268,
612–616.
Hansen, L., Eiberg, H., Barrett, T., Bek, T., Kjaersgaard, P., Tranebjaerg,
L. and Rosenberg, T. (2005) Mutation analysis of the WFS1 gene in seven
Danish Wolfram syndrome families; four new mutations identified.
Eur. J. Hum. Genet., 13, 1275– 1284.
Cano, A., Rouzier, C., Monnot, S., Chabrol, B., Conrath, J., Lecomte, P.,
Delobel, B., Boileau, P., Valero, R., Procaccio, V. et al. (2007)
Identification of novel mutations in WFS1 and genotype-phenotype
correlation in Wolfram syndrome. Am. J. Med. Genet. A, 143A,
1605– 1612.
Hardy, C., Khanim, F., Torres, R., Scott-Brown, M., Seller, A., Poulton,
J., Collier, D., Kirk, J., Polymeropoulos, M., Latif, F. et al. (1999) Clinical
and molecular genetic analysis of 19 Wolfram syndrome kindreds
demonstrating a wide spectrum of mutations in WFS1. Am. J. Hum.
Genet., 65, 1279– 1290.
Sam, W., Qin, H., Crawford, B., Yue, D. and Yu, S. (2001) Homozygosity
for a 4-bp deletion in a patient with Wolfram syndrome suggesting
possible phenotype and genotype correlation. Clin. Genet., 59, 136–138.
Shi, Y., Taylor, S.I., Tan, S.L. and Sonenberg, N. (2003) When translation
meets metabolism: multiple links to diabetes. Endocr. Rev., 24, 91–101.
Yamada, T., Ishihara, H., Tamura, A., Takahashi, R., Yamaguchi, S.,
Takei, D., Tokita, A., Satake, C., Tashiro, F., Katagiri, H. et al. (2006)
WFS1-deficiency increases endoplasmic reticulum stress, impairs cell
cycle progression and triggers the apoptotic pathway specifically in
pancreatic beta-cells. Hum. Mol. Genet., 15, 1600– 1609.
Fonseca, S.G., Fukuma, M., Lipson, K.L., Nguyen, L.X., Allen, J.R., Oka,
Y. and Urano, F. (2005) WFS1 is a novel component of the unfolded
protein response and maintains homeostasis of the endoplasmic reticulum
in pancreatic beta-cells. J. Biol. Chem., 280, 39609–39615.
Zatyka, M., Ricketts, C., da Silva Xavier, G., Minton, J., Fenton, S.,
Hofmann-Thiel, S., Rutter, G.A. and Barrett, T.G. (2008)
Sodium-potassium ATPase 1 subunit is a molecular partner of Wolframin,
an endoplasmic reticulum protein involved in ER stress. Hum. Mol.
Genet., 17, 190–200.
Ishihara, H., Takeda, S., Tamura, A., Takahashi, R., Yamaguchi, S.,
Takei, D., Yamada, T., Inoue, H., Soga, H., Katagiri, H. et al. (2004)
Disruption of the WFS1 gene in mice causes progressive beta-cell loss and
impaired stimulus-secretion coupling in insulin secretion. Hum. Mol.
Genet., 13, 1159– 1170.
Human Molecular Genetics, 2008, Vol. 17, No. 24
28. Bu, X. and Rotter, J.I. (1993) Wolfram syndrome: a
mitochondrial-mediated disorder? Lancet, 342, 598–600.
29. Hofmann, S., Bezold, R., Jaksch, M., Obermaier-Kusser, B., Mertens, S.,
Kaufhold, P., Rabl, W., Hecker, W. and Gerbitz, K.D. (1997) Wolfram
(DIDMOAD) syndrome and Leber hereditary optic neuropathy (LHON)
are associated with distinct mitochondrial DNA haplotypes. Genomics,
39, 8–18.
30. Barrett, T.G. (2007) Differential diagnosis of type 1 diabetes: which genetic
syndromes need to be considered? Pediatr. Diabetes, 8 (Suppl. 6), 15–23.
31. Murphy, R., Ellard, S. and Hattersley, A.T. (2008) Clinical implications
of a molecular genetic classification of monogenic beta-cell diabetes.
Nat. Clin. Pract. Endocrinol. Metab., 4, 200–213.
32. Slingerland, A.S. (2006) Monogenic diabetes in children and young
adults: challenges for researcher, clinician and patient. Rev. Endocr.
Metab. Disord., 7, 171– 185.
33. Alberti, K.G. and Zimmet, P.Z. (1998) Definition, diagnosis and
classification of diabetes mellitus and its complications. Part 1: diagnosis
and classification of diabetes mellitus provisional report of a WHO
consultation. Diabet. Med., 15, 539–553.
34. Zalloua, P.A., Shbaklo, H., Halaby, G., Terwedow, H., Xu, X. and
Azar, S.T. (2002) Type-2 diabetes family history delays the onset of
type-1 diabetes. J. Clin. Endocrinol. Metab., 87, 3192–3196.
35. Zalloua, P.A., Terwedow, H., Shbaklo, H., Halaby, G., Xu, X. and
Azar, S.T. (2003) Host and environmental factors defining the
36.
37.
38.
39.
40.
41.
42.
4021
epidemiology of type 1 diabetes mellitus in a group of Lebanese children
and young adults. J. Pediatr. Endocrinol. Metab., 16, 759– 769.
Khlat, M. (1988) Consanguineous marriage and reproduction in Beirut,
Lebanon. Am. J. Hum. Genet., 43, 188– 196.
Abchee, A., Puzantian, H., Azar, S.T., Shbaklo, H., Nasrallah, A.,
Sawaya, F.J., Alam, S. and Zalloua, P.A. (2006) Predictors of coronary artery
disease in the Lebanese population. Thromb. Res., 117, 631–637.
Takahashi, M., Matsuda, F., Margetic, N. and Lathrop, M. (2003)
Automated identification of single nucleotide polymorphisms from
sequencing data. J. Bioinform. Comput. Biol., 1, 253– 265.
Sunyaev, S., Ramensky, V., Koch, I., Lathe, W. III, Kondrashov, A.S. and
Bork, P. (2001) Prediction of deleterious human alleles. Hum. Mol.
Genet., 10, 591–597.
O’Connell, J.R. and Weeks, D.E. (1998) PedCheck: a program for
identification of genotype incompatibilities in linkage analysis.
Am. J. Hum. Genet., 63, 259– 266.
Abecasis, G.R., Cherny, S.S., Cookson, W.O. and Cardon, L.R. (2002)
Merlin – rapid analysis of dense genetic maps using sparse gene flow
trees. Nat. Genet., 30, 97– 101.
Purcell, S., Neale, B., Todd-Brown, K., Thomas, L., Ferreira, M.A.,
Bender, D., Maller, J., Sklar, P., de Bakker, P.I., Daly, M.J. et al. (2007)
PLINK: a tool set for whole-genome association and population-based
linkage analyses. Am. J. Hum. Genet., 81, 559– 575.