| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Division of Endocrinology and Metabolism, Departments of Medicine and Psychiatry (D.A.G.), Mount Sinai School of Medicine, New York, New York 10029
Address all correspondence and requests for reprints to: Yaron Tomer, M.D., Division of Endocrinology and Metabolism, Box 1055, Mount Sinai Medical Center, One Gustave L. Levy Place, New York, New York 10029. E-mail: ytomer{at}smtplink.mssm.edu
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
= 0.01),
assuming a recessive mode of inheritance and a penetrance of 0.3.
Multipoint analysis yielded a maximum LOD score of 2.5 between markers
D14S81 and D14S1054. There was no evidence for heterogeneity in our
sample. These data again suggest the presence of a major Graves’
disease susceptibility gene (GD-1) on chromosome 14q31. This locus is
close to the recently identified multinodular goiter-1 locus. |
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Although the genetic susceptibility to AITD has been studied extensively, the genes causing AITD have not been identified. Several candidate genes have been examined in the past. These have included the human leukocyte antigen (HLA) genes, Ig heavy chain genes (IgH), T cell receptor genes, the CTLA-4 gene, the thyroid peroxidase gene, and the TSHR gene (reviewed in Refs. 3, 4). Of these, only the HLA (6, 7) and CTLA-4 (8, 9) genes have shown positive associations with AITD. In particular, the HLA associations with AITD have received much attention. GD has been reported to be associated with HLA-DR3 (7) and HLA-DQA110501 (10, 11) in Caucasians, and HT has been associated with HLA-DR5 (12, 13), HLA-DR3 (14), and HLA-DQw7 (DQB110301) (15). Similarly, associations have been reported between AITD (GD and HT) and the 106 allele of a microsatellite inside the CTLA-4 gene, giving a relative risk of 2.0–2.8 (8, 9, 16). However, we have found that both the HLA and the CTLA-4 loci were not linked with either GD or HT (17, 18, 19, 20). Thus, the HLA and CTLA-4 genes confer a modulating effect on the development of AITD, but as neither is linked, they must be of minor significance in the overall genetic susceptibility to AITD. The major necessary genes (21) contributing to the development of AITD remain to be identified.
To map the major necessary genes (21) for the development of AITD we
have recently screened several candidate genes and candidate genetic
regions for linkage with AITD. Our strategy was to test genes involved
in immunoregulatory pathways and genes encoding for the major thyroid
autoantigens. In addition, genes linked to sex hormone functions were
also tested, as the AITDs are 5–10 times more common in females (22).
We have rejected linkage of AITD to several immunoregulatory genes in
addition to the HLA region (23) and the CTLA-4 gene (20, 23), including
the IgH gene and T cell receptor 
and ß genes (20). No linkage was
found between AITD and several other candidate genes tested, including
the thyroglobulin gene, the thyroid peroxidase gene (23), the estrogen
receptor and ß genes, and the aromatase gene (24). However, in a
preliminary screen of the TSHR region on chromosome 14q31 using 19 AITD
families (8 with GD) we found evidence for linkage between a locus in
the region (GD-1) and GD (23). This locus was not linked with HT or
AITD as a whole (i.e. GD+HT) (23). We now report a full
screening of chromosome 14 and fine mapping of the GD-1 locus using
densely spaced microsatellite markers and a larger dataset of 53
families. Our results support our previous preliminary data, and we
have mapped the GD-1 locus to the same region as the recently reported
multinodular goiter-1 (MNG-1) locus (25). The presence of thyroid
follicular hyperplasia in both GD and MNG (1) raises the possibility
that GD-1 and MNG-1 may constitute a thyroid disease gene grouping.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
GD was diagnosed using the following criteria: 1) documented clinical and biochemical hyperthyroidism requiring treatment, 2) a diffuse goiter, and 3) the presence of TSHR antibodies and/or diffuse uptake on 131I thyroid scan. HT was diagnosed by 1) documented clinical and biochemical hypothyroidism requiring thyroid hormone replacement, and 2) the presence of autoantibodies to thyroid peroxidase and thyroglobulin. Fifty-three families (323 individuals) were used in the current study (32 families with GD). All families were multiplex (>1 affected subject) and multigenerational. Families were ascertained through an index case with AITD who confirmed having at least 1 other first degree relative with AITD. Although as many relatives as possible were recruited from each family, the minimum requirement for participation in the study (nuclear family) was a family consisting of 4 first degree relatives (including the index case) from 2 generations. On the average, our families had 6.1 members. Each participant was interviewed, examined, and signed a consent, and all the pertinent clinical and laboratory data were recorded and stored in our database. At the same time blood was collected for DNA purification as well as for thyroid function and antibody testing. DNA was purified using the PureGene kit (Gentra Systems, Minneapolis, MN) according to the manufacturer’s instructions (26).
Microsatellite screening of chromosome 14
For the chromosome 14 screen, microsatellites were selected using the Genethon linkage maps (27) and were analyzed according to the method of Weber (28). Oligonucleotides for amplification of the microsatellites were designed according to published sequences in the Genome Database (http://gdbwww.gdb.org/). Fluorescent-labeled primers were purchased from PE Applied Biosystems (Foster City, CA). PCRs were performed as previously described (20). Two microliters of the pooled PCR products were mixed with 0.5 µL internal size standard and 10 µL deionized formamide, denatured, and separated using an ABI 310 genetic analyzer (PE Applied Biosystems). Allele calling was performed using the Genotyper 2.0 software (PE Applied Biosystems). The results were then automatically exported to our database (Ingres database) for linkage analysis. This eliminated possible errors in copying results by inputting them manually.
Analysis of candidate genes on chromosome 14
Chromosome 14 contains at least six candidate genes for thyroid
autoimmunity: 1) TSHR gene, 2) IgH gene, 3) T cell receptor gene,
4) MNG-1 gene, 5) insulin-dependent diabetes mellitus-11
(IDDM-11) gene, and 6) estrogen receptor ß (ESRß) gene. Three of
these candidate genes (IgH, T cell receptor , and ESRß) have been
previously analyzed and have not been found to be linked with GD, HT,
or AITD (GD+HT) (20, 24). The TSHR gene MNG-1 and IDDM-11 loci were
studied individually. Microsatellite markers were selected inside these
genes/loci, and they were amplified and analyzed as described
above.
Linkage analysis
This was performed using parametric methods of likelihood-maximization [classical logarithm of odds (LOD) score analysis]. Two-point LOD scores for the different markers studied were computed using LIPED software (29), and multipoint LOD scores were calculated using the GeneHunter program (30). Two-point LOD scores were computed for dominant and recessive models and for a range of penetrances (0.2–0.9). For the 16 markers used to screen chromosome 14, linkage analysis was performed for recombination fractions of 0.01 and 0.05 (as the average distance between the markers was <10 cM). For the 3 candidate genes tested by markers inside these genes, linkage was analyzed at a recombination fraction of 0.01. In the largest epidemiological survey of thyroid diseases, the prevalence of AITD was estimated to be 1% (31, 32). We, therefore, assumed that the disease prevalence was 0.01, and adjusted the gene frequency according to the model (dominant or recessive) and penetrance used, assuming Hardy-Weinberg equilibrium. Multipoint LOD scores were computed for the whole chromosome 14 using 19 markers (including the 3 candidate genes markers; see Results). For the multipoint analysis we assumed recessive inheritance and 0.3 penetrance based on the model giving the maximum LOD score in the 2-point LOD score analysis (see Results).
Models
The AITDs include GD and HT. At present it is not clear whether these are two distinct disorders with different etiologies, or whether common etiological factors are involved in the pathogenesis of these disorders. Indeed, both disorders can occur in the same family, and in our dataset 34% of the families included first degree relatives with GD and HT. Therefore, we have analyzed the data using three models: 1) all AITD patients were considered as affected; 2) only GD patients were considered as affected (under this model, HT patients were considered unaffected even if they had relatives with GD); and 3) only HT patients were considered as affected (under this model, GD patients were considered unaffected even if they had relatives with HT). Family members with thyroid autoantibodies alone were classified as unaffected. In addition, we tested the dataset for heterogeneity. Heterogeneity testing was also performed using the GeneHunter program (30), as were extended haplotypes for recombinant mapping. Each extended haplotype generated by GeneHunter was verified manually.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Table 1
shows some of the clinical
characteristics of the 53 families studied: 14 (26%) had GD affected
members only, 21 (40%) had HT affected members only, and 18 (34%)
were mixed with GD and HT affected first degree relatives. Of the 138
affected individuals, 121 (87.7%) were females, and the affected
female/male (F:M) ratio (7.1:1) was comparable to that reported
previously (33). Of the clinically and biochemically unaffected family
members, 35% had thyroid antibodies similar to the incidence reported
in previous studies (34, 35). Interestingly, the F:M ratio in the
thyroid antibody-positive unaffected individuals was 1:1,
i.e. much lower than the F:M ratio in the affected family
members, suggesting different pathogenic mechanisms for the development
of the AITDs and the propensity to secrete thyroid autoantibodies.
|
Simulation studies were performed to assess the power of our 53
families to detect linkage and to assess the maximum attainable LOD
scores for our dataset. We assumed a penetrance of 30% for our
simulations based on the reported 30% concordance rate in monozygotic
twins (36). The simulation software used for the power calculations
(37) generated 53 families that were homologous to the families in our
study. The simulations demonstrated that using a dataset of 53 families
we had the statistical power to reject linkage (i.e. LOD
score <-2) out to 10 cM ( = 0.1) at a penetrance of 0.3. Thus, our
families were sufficient to reject linkage for the tested markers.
Simulations also showed that we had the power to detect linkage at a
penetrance of 0.3 using the 53 families. The maximum LOD scores were
more than 4.0 for data generated with a marker at 0.01 recombination
fraction from the disease gene and more than 2.8 for a marker at 0.05
recombination fraction from the disease gene. The theoretical maximum
attainable LOD score in our dataset assuming the recessive model and a
recombination fraction of 0.01 was 6.7.
Screening chromosome 14
Analysis of microsatellite markers on chromosome 14 showed no
linkage with either HT or AITD as a whole (data not shown). However,
when considering GD as affected, positive LOD scores were obtained on
chromosome 14q31 (Fig. 1A). The maximum
two-point LOD score was 2.1 for marker D14S81, obtained for the
recessive model, at a penetrance of 30% and a recombination fraction
of 0.01 (Fig. 1B). Moreover, the LOD scores of other markers in the
local region of D14S81 were positive in a geographically logical
sequence (Fig. 1A).
|
For the multipoint analysis, we generated a genetic map for
chromosome 14, with sex-averaged distances between markers (in
recombination fraction units) as follows: D14S261 - 0.07 -
D14S283-0.12-D14S80-0.13 -
D14S70 - 0.06 - D14S288 - 0.08 -
D14S276-0.12-D14S63-0.07-D14S258
- 0.11 - D14S74 - 0.05 - TSHR - 0.04 - D14S68 - 0.005 - D14S67
(IDDM-11) - 0.095 - D14S280 - 0.03 - D14S973 - 0.01 - D14S81 - 0.05 -
D14S1054 (MNG-1) - 0.04 - D14S65 - 0.08 - D14S78 - 0.09 - D14S292.
These order and recombination fractions were obtained from the Genethon
maps (27) and were verified on our dataset. Multipoint linkage analysis
localized the GD-1 susceptibility locus on chromosome 14 to within an
approximate interval of 3 cM between markers D14S81 and D14S1054
(MNG-1). The multipoint LOD scores throughout this interval were more
than 2.2, with a maximum multipoint LOD score of 2.5 obtained 2 cM
telomeric to D14S81 (Fig. 2). This
represents significant evidence for linkage (38, 39). Further testing
showed a lack of heterogeneity in the linked region ( = 0.99;
maximum heterogeneity LOD score = 2.5). The GeneHunter
nonparametric LOD score at the linked region was 1.9, further
supporting the evidence for linkage in this region. Examination of
individual families showed that only 3 of 32 GD families had negative
LOD scores for marker D14S81, contributing a LOD score of -0.61 to the
total LOD score. These results also indicated that significant
heterogeneity was very unlikely at this marker locus.
|
Our Graves’ families consisted of some with GD only and some with
both GD and HT phenotypes (mixed families). To examine whether the
genetic influence of GD-1 was different in the group of Graves’
families in which some family members had HT, we tested them
separately. In this analysis the HT individuals in the mixed families
were classified as unaffected. Subgrouping the GD families into those
in which HT did and did not appear demonstrated that the LOD scores for
marker D14S81, which gave the highest two-point LOD score, were
positive in both subsets of families, contributing approximately
equally to the total LOD score (Table 2).
This suggested that the contribution of GD-1 to the susceptibility to
GD was similar in all families with GD that we examined, whether they
included only GD-affected members or GD- and HT-affected members.
Subgrouping of our Hashimoto’s families into HT-only families and
families with both GD and HT (mixed families) did not change the
negative linkage results obtained for HT (in this analysis the GD
patients in the mixed families were classified as unaffected). In both
groups of families the LOD scores were negative for all markers on
chromosome 14 (data not shown).
|
The GD families were also subgrouped by their country of origin.
The LOD scores obtained for marker D14S81 after subgrouping the GD
families demonstrated that both the North American families and the
European families contributed approximately equally to the total LOD
score (Table 2
). This suggested that the contribution of GD-1 to the
susceptibility to GD was similar in the North American and European
families with GD that we examined and provided further evidence for a
lack of heterogeneity in our population.
Recombinant mapping of GD-1
An interval of 22 cM surrounding the GD-1 locus was analyzed by
recombinant mapping in all families. Critical meiotic recombination
events between markers D14S81 and D14S1054 were detected in two
individuals (31-203, 31-304) in one large informative family (Fig. 3). These data provided support for the
localization of the GD-1 locus telomeric to D14S81 and centromeric to
D14S1054, as suggested by the multipoint analysis. Interestingly, the
D14S1054 marker was linked to MNG according to a recent study in one
large family (MNG-1) (25).
|
The three candidate genes on chromosome 14 (TSHR, MNG-1, and
IDDM-11) were analyzed using specific markers located inside these
loci. For all of these candidate genes we were able to reject linkage
to HT and AITD (GD+HT; Table 3). When
considering GD as affected, linkage could not be strictly rejected, but
there was no evidence for linkage for any of the candidate genes except
for MNG-1. The maximum two-point LOD score for the MNG-1 locus, when
considering GD as affected, was 1.65 at a recombination fraction of
0.01. Indeed, the multipoint analysis of chromosome 14 localized GD-1
to within 2 cM centromeric of the MNG-1 locus (Fig. 2). This raised the
possibility that MNG-1 and GD-1 constituted the same locus.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In our previous analysis of 8 GD families, a 2-point LOD score of 2.1 was obtained using a dominant model at 80% penetrance, whereas the recessive model gave a low positive LOD score of 0.4 (23). With our larger dataset, which included 32 GD families, the maximum LOD score of 2.1 was obtained for the recessive model at 30% penetrance, whereas the dominant model gave a maximum LOD score of 0.73. This suggested that the recessive model better approximated the true genetic inheritance of GD-1 (37). Our preliminary analysis probably reflected the effects of a small sample size.
There are several factors that may reduce LOD scores when true linkage exists. These include heterogeneity in the dataset, too small a sample size, and the presence of other contributing genes or epistatic genes. Our analysis of the individual families as well as a computer analysis (using the GeneHunter algorithm for heterogeneity testing) showed that heterogeneity, if it existed in our dataset, was minimal. The power calculations demonstrated that our dataset was large enough to detect linkage with a LOD score of more than 2.8 if only one gene contributed to disease susceptibility. Therefore, other epistatic genes must exist, and the effect of GD-1 alone on the genetic susceptibility to GD is only moderate. Indeed, the reduced penetrance (30%) suggests that several genes of varying effects are involved in the pathogenesis of GD. When more loci for Graves’ disease are identified, this hypothesis could be tested by performing a linkage analysis for several loci together. However, because complex multifactorial diseases appear to be caused by more than one gene of varying effects (i.e. oligogenic diseases), the identification of susceptibility genes for these diseases has been difficult (42). For example, in IDDM at least four loci have been confirmed for linkage with the disease (41, 43), but only two of them (HLA and insulin-variable number tandem repeat genes) have been identified (44, 45).
As in all complex and common diseases, the genetic susceptibility to AITD lacks a simple Mendelian pattern of inheritance and probably involves several genes with varying penetrances and interactions. This makes the identification of susceptibility genes for complex diseases very difficult. One possible solution is to test for linkage using different models. We tested our markers for linkage with AITD using two models, dominant and recessive. In addition, we decided a priori to subgroup the AITDs based on clinical criteria into HT and GD, and test each of these independently and as a whole. Although these multiple analyses enabled us to maximize the LOD score with respect to the model and phenotype of the disease, they may have weakened the evidence for linkage because of multiple testing. It was recently suggested that in cases when dominant and recessive models are used, the maximum LOD score obtained should be reduced by 0.3 (46). In our case this would result in a maximum LOD score of 2.2, still giving strong evidence in favor of linkage (38). Thus, analyzing for two genetic models did not significantly change the strength of our linkage results and enabled us to find the model that best approximated the inheritance of GD-1.
In addition to screening chromosome 14, we tested six specific
candidate genes on this chromosome [three of them, IgH, TCR, and
ESRß, were previously reported (20, 24)]. Candidate genes are genes
of known sequence and location that could be involved in disease
pathogenesis by virtue of their involvement in pathways associated with
the pathological manifestations of the disease. Linkage tests using
candidate genes offer a rapid way to test the hypothesis that these
particular genes are major contributors to the disease. The candidate
gene approach has already been used successfully in maturity-onset
diabetes of the young (47, 48). Potential candidate genes for
autoimmune thyroid diseases include 1) genes participating in immune
regulatory pathways, 2) genes controlling the expression of the thyroid
autoantigens, 3) genes involved in sex hormone function (in view of the
female preponderance of AITD) (22), and 4) IDDM genes (in view of the
association between IDDM and AITD) (49). The six chromosome 14
candidate genes tested in this and our previous studies belong to all
of these categories. None of the candidate genes studied, except MNG-1,
demonstrated evidence of linkage to GD, HT, or AITD as a whole (20, 24). Of note is the evidence against linkage of Graves’ disease with
the TSHR gene. Previous work has shown a weak association between the
TSHR gene and GD (50), but other studies could not confirm these
results (51). Indeed, we and others were unable to demonstrate TSHR
mutations in thyroid tissues from GD patients (52, 53). However, a gene
close to the TSHR locus could theoretically confer susceptibility to GD
by influencing the expression of the TSHR gene. Our results showing no
evidence for linkage between GD and the TSHR gene made such a
possibility highly unlikely and confirmed the results of a previous
linkage study (54).
In conclusion, we confirmed evidence for a new GD susceptibility gene (GD-1) on chromosome 14q31. GD-1 was mapped to within 2 cM of the recently reported MNG-1 locus (25), which has been linked to MNG. In both GD and MNG, the thyroid gland may show follicular hyperplasia, and MNG is often associated with AITD (1). This raises the possibility that GD-1 and MNG-1 may represent a thyroid disease gene grouping. A newly isolated growth factor, SEL1L, was recently mapped to 14q31 (55), and this would be an interesting candidate to examine in both GD and MNG.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Additional members of the International Consortium for the Genetics
of Autoimmune Thyroid Disease include Drs. Meir Berezin (Tel-Hashomer,
Israel), Rhoda Cobin (New York, NY), Luca Chiovato and Aldo Pinchera
(Pisa, Italy), Sandra McLachlan (Los Angeles, CA), Bernard Rees Smith
(Cardiff, UK), and Fred Clark and Eric Young (Newcastle upon Tyne,
UK).
Received June 30, 1998.
Revised August 10, 1998.
Accepted August 18, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J.-Y. Hsiao, M.-C. Hsieh, K.-J. Tien, S.-C. Hsu, S.-J. Shin, and S.-R. Lin Association between a C/T Polymorphism in Exon 33 of the Thyroglobulin Gene Is Associated with Relapse of Graves' Hyperthyroidism after Antithyroid Withdrawal in Taiwanese J. Clin. Endocrinol. Metab., August 1, 2007; 92(8): 3197 - 3201. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Tomer and T. F. Davies Searching for the Autoimmune Thyroid Disease Susceptibility Genes: From Gene Mapping to Gene Function Endocr. Rev., October 1, 2003; 24(5): 694 - 717. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Collins, J. M. Heward, J. Carr-Smith, J. Daykin, J. A. Franklyn, and S. C. L. Gough Association of a Rare Thyroglobulin Gene Microsatellite Variant with Autoimmune Thyroid Disease J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 5039 - 5042. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Sakai, S. Shirasawa, N. Ishikawa, K. Ito, H. Tamai, K. Kuma, T. Akamizu, M. Tanimura, K. Furugaki, K. Yamamoto, et al. Identification of susceptibility loci for autoimmune thyroid disease to 5q31-q33 and Hashimoto's thyroiditis to 8q23-q24 by multipoint affected sib-pair linkage analysis in Japanese Hum. Mol. Genet., June 1, 2001; 10(13): 1379 - 1386. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Ban, M. Taniyama, and Y. Ban Vitamin D Receptor Gene Polymorphism Is Associated with Graves' Disease in the Japanese Population J. Clin. Endocrinol. Metab., December 1, 2000; 85(12): 4639 - 4643. [Abstract] [Full Text]
|
|
|
|
 |
|
 A. P. Weetman Graves' Disease N. Engl. J. Med., October 26, 2000; 343(17): 1236 - 1248. [Full Text] [PDF]
|
|
|
|
|
|
R. Nithiyananthan, J. M. Heward, A. Allahabadia, A. H. Barnett, J. A. Franklyn, and S. C. L. Gough A Heterozygous Deletion of the Autoimmune Regulator (AIRE1) Gene, Autoimmune Thyroid Disease, and Type 1 Diabetes: No Evidence for Association J. Clin. Endocrinol. Metab., March 1, 2000; 85(3): 1320 - 1322. [Abstract] [Full Text]
|
|
|
|
|
|
T. H. Brix, K. O. Kyvik, and L. Hegedüs A Population-Based Study of Chronic Autoimmune Hypothyroidism in Danish Twins J. Clin. Endocrinol. Metab., February 1, 2000; 85(2): 536 - 539. [Abstract] [Full Text]
|
|
|
|
|
|
Mapping the Major Susceptibility Loci for Familial Graves' and Hashimoto's Diseases: Evidence for Genetic Heterogeneity and Gene Interactions J. Clin. Endocrinol. Metab., December 1, 1999; 84(12): 4656 - 4664. [Abstract] [Full Text]
|
|
|
|
|
|
J. M. Heward, A. Allahabadia, M. Armitage, A. Hattersley, P. M. Dodson, K. Macleod, J. Carr-Smith, J. Daykin, A. Daly, M. C. Sheppard, et al. The Development of Graves' Disease and the CTLA-4 Gene on Chromosome 2q33 J. Clin. Endocrinol. Metab., July 1, 1999; 84(7): 2398 - 2401. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
