| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Endocrinology and Metabolism (Y.J., W.T.), Department of Medicine, First Affiliated Hospital to China Medical University, Shenyang 110001, PR China; Chinese National Human Genome Center at Shanghai (Y.J., S.B., X.X., J.Z., S.X., L.J., W.H.), Shanghai 201203, PR China; Department of Public Health (S.B.), China Medical University, Shenyang 110001, PR China; Rui-Jin Hospital affiliated to Shanghai Second Medical University (J.C., W.H.), Shanghai 200025, PR China; and Department of Epidemiology (Y.Y.S.), Bloomberg School of Public Health, Johns Hopkins University, Baltimore, Maryland 21205
Address all correspondence and requests for reprints to: Dr. Weiping Teng, China Medical University, 92 North 2nd Road, Heping District, Shenyang 110001, PR China. E-mail: tengweiping{at}hotmail.com; or Dr. Wei Huang, Chinese National Human Genome Center at Shanghai, 250 Bi Bo Road, Shanghai 201203, PR China. E-mail: huangwei{at}chgc.sh.cn.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
= 0.38). Nonparametric multipoint analysis also showed significant excess allele sharing, with a P value as low as 0.001, at the same locus. Our findings provide evidence for a susceptibility locus for GD on chromosome 5q31 and support the existence of genetic heterogeneity in GD. |
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The etiology of GD is generally thought to fit a multifactorial pattern in which clinical disease phenotypes develop with the interplay of genetic susceptibility and environmental and endogenous factors (1, 2). Evidence for the role of genetic factors is indicated by an increased relative risk of more than 15 in siblings (3), compared with the general population (
s), and an increased concordance rate in monozygotic, compared with dizygotic, twins (4). The lack of a clear pattern of inheritance suggested that multiple genes are involved in influencing the autoimmune events in GD (5). During the last decade, many efforts have been put into the characterization of the genetic background of GD. Several candidate genes have been postulated and tested for a possible contribution to genetic susceptibility to GD by means of both association and linkage analysis. These have included the genes for human leukocyte antigen (HLA) (6, 7, 8, 9), T-cell receptor (10, 11), TSH receptor (12, 13), IL-1 receptor antagonist (14, 15), and cytotoxic T lymphocyte antigen 4 (CTLA-4) (16, 17). Interestingly, HLA (6, 7, 8, 9, 18, 19, 20, 21) and CTLA-4 (16, 17, 22, 23) have been consistently associated with GD in different ethnic groups. However, linkages of GD with HLA and CTLA-4 need to be confirmed in additional studies (24, 25, 26, 27, 28). Other genes responsible for GD remain unknown.
Genome-wide linkage studies provide a powerful tool for disease gene identification. The availability of maps of markers that cover the entire human and mouse (and rat) genomes has enabled the chromosomal localization of almost 40 genes that contribute to autoimmunity in human disease and animal models (3). In one study, 14 multiplex GD pedigrees each with multiple affected individuals, 20 pedigrees that were mixed with both GD and Hashimoto’s thyroiditis (HT) first-degree relatives, and 22 multiplex pedigrees with HT were investigated. This cohort of pedigrees was derived from the North American, Italian, Israeli, and British populations. Three GD susceptibility loci were identified and located on chromosome 14, 20, and X (29). Another study was performed by Sakai et al. (28) using the nonparametric sibling-pair method in 123 Japanese sibling pairs with autoimmune thyroid disease (AITD), including 67 GD sibling pairs, 25 HT sibling pairs, and 31 GD-HT sibling pairs, which demonstrated evidence for linkage of AITD to 5q31-q33. A maximum multipoint LOD score of 3.14 was obtained at D5S436.
In the current study, we performed a genome scan in a homogeneous cohort of 54 Chinese Han multiplex pedigrees with GD (322 individuals) to identify chromosomal regions that contribute to the genetic risk of GD. We hereby report the possible identification of a major susceptibility locus for GD on chromosome 5q31 in Chinese.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Fifty-four Chinese Han pedigrees (322 individuals) were analyzed in the study, all from Liaoning Province (located in the Northeast China). Among these pedigrees, 32 (59%) had two affected first-degree relatives, 15 (28%) had three, 5 (9%) had four, and 2 (4%) had five. Pedigrees were ascertained through a GD proband attending the Outpatient Department of Endocrine Diseases, who were subsequently confirmed having at least one first-degree relative with GD. The diagnosis of GD was based on documented clinical and biochemical evidence of hyperthyroidism, diffused goiter, and the presence of at least one of the following: positive TSH receptor antibody tests, diffusely increased 131I (iodine-131) uptake in the thyroid gland, or presence of exophthalmos. For this study, all other pedigree members, whether thyroid autoantibody positive or negative, were classified as unaffected. For diagnosis accuracy, all individuals classified as affected were interviewed and examined by experienced clinicians. After informed consent, blood samples were collected from all participants for DNA preparation as well as biochemical measurements. This project was approved by the Ethical Committee of the Chinese National Human Genome Center at Shanghai for the involvement of human subjects.
Genotyping
Genomic DNAs were isolated from peripheral blood using standard procedures (30). A linkage mapping set (version 2.0, Perkin-Elmer PE Applied Biosystems) was used for genotyping. PCR was performed in 5 µl reaction volumes containing 20 ng genomic DNA; 0.05 µmol of each primer; 10 mmol/liter Tris-HCl (pH 8.3); 50 mmol/liter KCl; 3.0 mmol/liter MgCl2; 200 µmol/liter of each deoxy (d)-ATP, GTP, TTP, and CTP; and 0.25 U AmpliTaq Gold DNA polymerase (Perkin-Elmer PE Applied Biosystems). Reaction mixtures were heated to 94 C for 10 min and then cycled 44 times as follows: 30 sec at 94 C, 60 sec at 56 C, and 60 sec at 72 C, except that in the first 14 cycles, the annealing temperatures decreased from 63 C to 56 C by 0.5 C per cycle. A final extension of 10 min at 72 C was included. After PCR, 1 µl PCR products was mixed with 0.25 µl internal size standard, 1 µl double-distilled H2O, and 2.75 µl loading solution and then denatured and separated on a MegaBASE 1000 DNA sequencer (Amersham Pharmacia Biotech, Inc., Piscataway, NJ). Allele calling was performed by using Genetic Profiler 1.1 software (Amersham Pharmacia Biotech, Inc.). Each genotype was reviewed manually by two members of the research team to confirm the accuracy of allele calling. Mendelian inheritance was confirmed using pedigree information.
Linkage analysis
Linkage analysis was performed by using both model-free (nonparametric) and model-based (parametric) approaches.
Parametric analysis
In the genetic analysis of common complex diseases, the LOD score methods have been shown to be more powerful than nonparametric methods (31, 32). We thus used the former as our major approach to test for linkage. Two-point and multipoint LOD scores were calculated by using the GENEHUNTER program (33) assuming both dominant and recessive models. For the dominant model, three levels of penetrance were tested (30%, 60%, and 90%); and for the recessive model, four levels of penetrance were evaluated (30%, 60%, 90%, and 100%). A population prevalence of 1% for GD was assumed (34, 35), with a nonsusceptibility genotype penetrance of 0.009; the gene frequency was adjusted according to the model (dominant or recessive) and the penetrance used, assuming Hardy-Weinberg equilibrium. Marker allele frequencies were calculated from all available pedigree data. All LOD scores were calculated under the assumption of heterogeneity using an admixture test (36, 37) as implemented by GENEHUNTER (33).
Nonparametric analysis
Nonparametric linkage (NPL) scores were calculated for the complete marker set by using the multipoint algorithm in GENEHUNTER, version 2.0. As recommended by Kruglyak et al. (33), pedigrees were analyzed using the ALL function of GENEHUNTER, which examines all affected individuals simultaneously and assigns a higher score when more of them share the same allele by descent.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Clinical characteristics of the 54 pedigrees studied are given in Table 1
. Of the 139 affected individuals, 109 (78.4%) were females, with an affected female/male ratio of 3.6:1. This was consistent with that reported in the literature (38). Of the 183 unaffected subjects, 78 (42.6%) were positive for thyroid antibodies, which are similar to the incidence reported by Burek et al. (39). The female/male ratio in the thyroid antibody-positive, unaffected individuals was 1.5:1, which was in accordance with the finding by Tomer et al. (40).
|
A total of 277 microsatellite markers were genotyped, with an average intermarker spacing of 13.4 cM and an average heterozygosity of 71%. We considered any region with a LOD greater than 1.0 or a P value less than 0.05 as potentially interesting. Parametric linkage analyses identified eight markers yielding LOD scores greater than 1.0. Of these, only the marker (D5S436) on chromosome 5q31 achieved a maximum LOD score above 2.0. The other markers were localized on chromosome 1p33, 1q42, 11p14, 12q15, and 15q21 (Table 2). Nonparametric linkage analyses also revealed eight loci showing nominal evidence for linkage (P < 0.05); the majority are the same as those found in parametric analyses, except the one (D10S185) on chromosome 10q23 and the one (D18S1102) on 18q12 (Table 3).
|
|
) and nonparametric linkage analyses (Table 3) revealed the strongest evidence for linkage at marker D5S436 on chromosome 5q31. To further assess the significance of these suggestive findings, we typed four additional markers around D5S436 in this chromosome region, and we obtained evidence in favor of linkage by both parametric and nonparametric analyses. Two-point parametric linkage analysis
A maximum two-point LOD score of 4.31 ( = 0.52) was obtained for marker D5S2090 under the recessive model with complete penetrance. The LOD scores for other markers in the local region of D5S2090 were positive in a geographically logical sequence (Fig. 1). The two-point LOD scores obtained for D5S2090 under all inheritance models tested further demonstrate its major effect in our pedigrees; indeed, three of the recessive models (with a penetrance of 60%, 90%, or 100%) gave LOD scores greater than 3.0.
|
For the multipoint analysis, we used a genetic map for chromosome 5, with sex-average distances between markers (in cM units) as follows: D5S1981 - 10.1 - D5S406 - 7.9 - D5S630 - 20.9 - D5S419 - 12.1 - D5S426 - 6.5 - D5S418 - 6.9 - D5S407 - 9.7 - D5S647 - 9.4 - D5S424 - 8.3 - D5S641 - 26.5 - D5S2027 - 10.7 - D5S471 - 9 - D5S2115 - 6.2 - D5S2017 - 2.4 - D5S436 - 2.7 - D5S2090 - 0.2 - D5S434 - 2.7 - D5S2014 - 4.0 - D5S410 - 7.1 - D5S422 - 10.9 - D5S400 - 21 - D5S408. This order and the cM units were obtained from the Genethon maps (41). Multipoint linkage analysis localized the GD susceptibility locus to within an approximate interval of 2 cM between markers D5S436 and D5S434. The multipoint LOD scores throughout this interval were more than 3.0, with a maximum multipoint LOD score of 4.12 ( = 0.38; Fig. 2). Examination of individual pedigrees showed that 30% of the pedigrees had positive (>+0.1) LOD scores for D5S2090, within the range of 0.1–2.0. No apparent clinical differences in patients were observed in the "linked" (>+0.1) and "unlinked" pedigrees (<-0.1).
|
Multipoint nonparametric linkage analyses were conducted by using the GENEHUNTER program. The peak NPL score occurred at marker D5S2090, the same marker that gave the highest parametric LOD score. The peak NPL score was 2.66 (P = 0.001), which supports the evidence for linkage at this locus (Fig. 3).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
For a complex trait like GD, the multigenic nature and possible heterogeneity among different ethnic, geographic, and racial populations, in both genetic etiology of the disease and marker allele frequencies, make replication of linkages a great challenge. However, using the same marker set but different ascertainment strategies to recruit pedigrees (at least two affected first-degree relatives with GD for our study and at least two affected sibs with AITD for the study of Sakai et al.) and different statistical tests, our genome-wide scan in Chinese replicated the findings of Sakai et al. (28) in Japanese, which localized a susceptibility locus for GD to chromosome 5q31. Especially, results from both groups revealed the maximum LOD score at D5S436 (28). The sample size of this study was not very large, but we did replicate the linkage of GD to 5q31, suggesting that this region may harbor a gene of strong effect. Because we fine-mapped this region further and Sakai et al. (28) did not, we were able to localize the susceptibility locus for GD in this region more accurately, to within an approximate interval of 2 cM between markers D5S436 and D5S434.
Both Chinese and Japanese are Eastern-Asian populations, and it has been shown that Eastern-Asian populations are derived from the same ancient population (42). The fact that linkage of GD with chromosome 5q31 has not been reported in whites thus far suggests that the susceptibility locus for GD in this region may be Eastern-Asian population specific.
In the present study, we obtained a maximum two-point LOD score of 4.31 and a maximum multipoint LOD score of 4.12; both represented significant evidence for linkage (43, 44). The results from parametric tests of linkage are far more supportive of linkage than the nonparametric (NPL) results. This may be due at least in part to the inherently lower power of nonparametric approaches (31, 32).
Genetic heterogeneity, which means mutations in any one of several genes can result in the same phenotype, is a major confounding factor in the mapping of complex disease genes. The existence of genetic heterogeneity hampers genetic mapping because a chromosomal region may cosegregate with a disease in some pedigrees but not in others (45). To test genetic heterogeneity in our data set, the admixture test (36, 37), as implemented in the GENEHUNTER program (33), was employed. The results were significant. The maximum two-point and multipoint LOD scores obtained for marker D5S2090 under the assumption of heterogeneity were 4.31 and 4.12, respectively, significantly greater than the corresponding values without heterogeneity, -2.23 and -7.09, respectively. In addition, we also performed the heterogeneity test using HOMOG (46). We tested the hypothesis that there is linkage and heterogeneity (H2) against the hypothesis that there is neither linkage nor heterogeneity. And the estimated proportion of families linked to the locus of interest was 0.55. The 
2 value was 19.753 and the likelihood ratio that presents the exact odds for this test was 1.947 x 104. As pointed out by Terwilliger and Ott (37), "for considering the test of H2 vs. H0 to be significant, the likelihood ratio must exceed 2000." Thus, we concluded that our data support the existence of genetic heterogeneity in GD. In addition, we want to note that, by default, GENEHUNTER computes two-point HLOD at a recombination fraction of zero, and HOMOG uses all recombination fractions and computes the LOD scores at these recombination fractions. In the heterogeneity test we performed, 
and 
were both estimated ( = 0.55; 
= 0.001, HLOD = 4.294). This may explain why the estimated and HLOD given by two computer programs are slightly different.
Three of the potential GD susceptibility regions identified in this screen of GD pedigrees, 12q15, 1p33, and 18q12, have also been implicated in other human autoimmune diseases. Interval 12q15 includes potential susceptibility loci for Crohn’s disease (47) and multiple sclerosis (MS: 49). The 1p33 and 18q12 intervals showed evidence for linkage with MS (48, 49). The overlapping of the susceptibility loci for different autoimmune disorders supports the notion that there may be common genetic factors that predispose to autoimmunity (50). Vyse and Todd (3) have shown significant overlapping of susceptibility loci for autoimmune diseases in the mouse.
In conclusion, our first attempt of the whole genome screen in Chinese GD pedigrees demonstrated evidence for linkage of GD to chromosome 5q31 in a region that has also been linked to GD in Japanese, but not in whites, suggesting that this region may contain a major locus for GD specific for Eastern-Asian populations. The detection of multiple loci with lower significance levels and the evidence for the existence of genetic heterogeneity support the presumed complexity of the genetic etiology of GD. The characterization of susceptibility genes with effects large enough to be detectable in whole genome screens should provide new insights into the pathogenesis of GD and offer important opportunities for improving diagnostic methods, treatment, and perhaps even strategies to prevent this genetically complicated autoimmune disorder.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
W.T. and W.H. contributed equally to this work.
Abbreviations: AITD, Autoimmune thyroid disease; CTLA-4, cytotoxic T lymphocyte antigen 4; GD, Graves’ disease; HLA, human leukocyte antigen; HT, Hashimoto’s thyroiditis; LOD, limit of detection; NPL, nonparametric linkage.
Received December 12, 2001.
Accepted January 10, 2003.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
and interleukin-1 receptor antagonist genes and Graves’ disease in a North American Caucasian population. J Clin Endocrinol Metab 81:4476–4478[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  C. J Owen, J. A Eden, C. E Jennings, V. Wilson, T. D Cheetham, and S. H S Pearce Genetic association studies of the FOXP3 gene in Graves' disease and autoimmune Addison's disease in the United Kingdom population. J. Mol. Endocrinol., August 1, 2006; 37(1): 97 - 104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Taylor, S. C. Gough, P. J. Hunt, T. H. Brix, K. Chatterjee, J. M. Connell, J. A. Franklyn, L. Hegedus, B. G. Robinson, W. M. Wiersinga, et al. A Genome-Wide Screen in 1119 Relative Pairs with Autoimmune Thyroid Disease J. Clin. Endocrinol. Metab., February 1, 2006; 91(2): 646 - 653. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Hiromatsu, T. Fukutani, M. Ichimura, T. Mukai, H. Kaku, H. Nakayama, I. Miyake, S. Shoji, Y. Koda, and T. Bednarczuk Interleukin-13 Gene Polymorphisms Confer the Susceptibility of Japanese Populations to Graves' Disease J. Clin. Endocrinol. Metab., January 1, 2005; 90(1): 296 - 301. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Bayer, S. Neumann, B. Meyer, F. Ruschendorf, A. Reske, T. Brix, L. Hegedus, P. Langer, P. Nurnberg, and R. Paschke Genome-Wide Linkage Analysis Reveals Evidence for Four New Susceptibility Loci for Familial Euthyroid Goiter J. Clin. Endocrinol. Metab., August 1, 2004; 89(8): 4044 - 4052. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
