| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
BRIEF REPORT |
Centro de Investigação de Patobiologia Molecular (B.M.C., P.F.B., V.L.) and Serviço de Endocrinologia (L.G.S., V.L.), Instituto Português de Oncologia de Lisboa Francisco Gentil, 1099-023 Lisboa, Portugal; Faculdade de Ciências Médicas da Universidade Nova de Lisboa (L.G.S., V.L.), 1169-056 Lisboa, Portugal; and Instituto de Medicina Molecular (B.M.C.), Faculdade de Medicina de Lisboa, 1600-190 Lisboa, Portugal
Address all correspondence and requests for reprints to: Branca M. Cavaco, Ph.D., Centro de Investigação de Patobiologia Molecular (CIPM), Instituto Português de Oncologia de Lisboa Francisco Gentil, 1099-023 Lisbon, Portugal. E-mail: bcavaco{at}ipolisboa.min-saude.pt.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: Our objective was to map the gene predisposing to familial thyroid epithelial neoplasia in a large Portuguese family.
Methods and Results: The clinical screening of a Portuguese family identified 11 members affected with benign thyroid lesions and five affected with thyroid carcinomas. Linkage analysis excluded the involvement of the fPTC/PRN, NMTC1, MNG1, and TCO loci. To map the gene predisposing to thyroid epithelial neoplasia in this family, a genome-wide linkage analysis was conducted, using DNA samples from 17 family members and high-density single-nucleotide polymorphism arrays. A genome-wide significant evidence of linkage, to a single region on chromosome 8p23.1-p22 was obtained, with a maximum parametric haplotype-based LOD score of 4.41 (
= 0.00). Linkage analysis with microsatellite markers confirmed linkage to 8q23.1-p22, and recombination events delimited the minimal region to a 7.46-Mb span. Seventeen suggestive candidate genes located in the minimal region were excluded as susceptibility genes by mutational analysis. Allelic losses in the 8p23.1-p22 region were absent in seven thyroid tumors from family members, suggesting that the inactivation of a putative tumor suppressor gene may have occurred through other mechanisms.
Conclusions: Our results present evidence for the existence of a novel familial thyroid epithelial neoplasia susceptibility locus on chromosome 8p23.1-p22, providing the basis for the identification of a gene for this disease.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In families with more than three members affected with NMTC, the risk of a familial syndrome exceeds 96% (3). Familial NMTC (FNMTC; MIM 188550), accounts for approximately 5% of all thyroid tumors (4, 5, 6). High incidences of multinodular goiter (MNG) and follicular thyroid adenoma (FTA) are common features in FNMTC families (7, 8). Review of the different families and genetic studies suggest an autosomal dominant inheritance with a reduced penetrance (4, 8).
The involvement of the PTEN, TSHR, TRKA/NTRK1, RET, MET, APC, BRAF, MEK1, MEK2, and K-, N-, and H-RAS genes in the susceptibility to FNMTC has been investigated (9, 10, 11), but no linkage or germline mutations were found. However, in a recent study, we identified somatic BRAF and RAS mutations in thyroid tumors from eight FNMTC families, suggesting that these oncogenes are likely to be involved in the progression of familial thyroid tumors (12).
Four chromosomal regions stand out as candidates for genetic susceptibility to FNMTC: the thyroid carcinoma with oxyphilia locus (TCO; MIM 603386) localized to chromosome 19p13.2 (9), the MNG susceptibility locus (MNG1; MIM 138800) mapped to chromosome 14q32 (13), the papillary thyroid carcinoma (PTC) and papillary renal neoplasia locus identified on chromosome 1p13.2-1q22 (fPTC/PRN; MIM 605642) (10), and a fourth susceptibility locus named non-medullary thyroid carcinoma 1 mapped to chromosome 2q21 (NMTC1; MIM 606240) (14).
In the present study, we describe a large Portuguese family affected with thyroid epithelial neoplasia. Genome-wide linkage analysis of this family, using high-density single-nucleotide polymorphism (SNP) arrays, identified a novel familial thyroid epithelial neoplasia predisposition locus on chromosome 8p23.1-p22.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The criteria for eligibility of the families were that three (or more) first-degree family members had to be affected with NMTC (3). Patients with nodular goiter (NG) or FTA were also considered as affected (7, 8).
We identified a family (family 1, previously described in Ref. 12) presenting an autosomal dominant inheritance of familial thyroid epithelial neoplasia (Fig. 1
and supplemental Table 1, published as supplemental data on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org). We also evaluated six previously reported Portuguese families with FNMTC (families 2–7 in Ref. 12).
|
GeneChip Human Mapping 10K linkage analysis
Linkage to fPTC/PRN, NMTC1, MNG1, and TCO susceptibility loci had been previously excluded in family 1 (12). A genome-wide linkage analysis was conducted using DNA from 17 members of family 1 (Fig. 1
) and the GeneChip Human Mapping 10K Array Xba 131, which contains about 10,000 SNP markers (Affymetrix, Santa Clara, CA). Haplotype-based, single-point, and multipoint parametric linkage analyses of SNP genotypes, assuming a dominant model of inheritance and disease allele frequency 0.001, were undertaken using, respectively, the programs VARIA (Agilent Technologies, Palo Alto, CA), SuperLink version 1.6, and GeneHunter version 2.1r5 (with easyLINKAGEPlus version 5.08) (16). Genome-wide linkage analysis was conducted using a complete-penetrance model and two more conservative, tumor-goiter and tumor-only, models. Details of GeneChip analysis and linkage models are supplied as supplemental material.
Microsatellite linkage analysis
Linkage to chromosome 8p23.1-p22, was further appraised in family 1, using 13 microsatellite markers (Fig. 1 and supplemental Table 2) and assessed in families 2–7. The tumor-goiter model was selected for parametric single-point and multipoint linkage analyses of microsatellite genotypes, using the SuperLink version 1.6 and GeneHunter version 2.1r5 programs, respectively. Multipoint nonparametric linkage analysis was performed using GeneHunter version 2.1r5. Details are supplied as supplemental material.
Gene expression profile in normal thyroid
Analysis of gene expression in a human thyroid total RNA pool, from human thyroids of 65 male and female Caucasians (BD Biosciences, Palo Alto, CA), was carried out, using the Affymetrix Human Genome U133 Plus 2.0 Array. Probe sets were assigned with expression calls A (absent, gene not expressed), P (present, gene expressed), or M (gene marginally expressed). Details are supplied as supplemental material.
Analysis of candidate genes on chromosome 8p23.1-p22
A bioinformatic search for candidate genes, located at the minimal region on chromosome 8p23.1-p22 was conducted, using the Ensembl Genome Browser (http://www.ensembl.org) and NCBI (http://www.ncbi.nlm.nih.gov) databases. A search for germline mutations in 17 candidate genes was undertaken (Table 1) by screening constitutional DNA from three affected members of family 1 (individuals II.15, III.4, and III.5). Details are supplied as supplemental material.
|
Four classic PTC, one PTC of the follicular variant and two FTA, from five patients of family 1 were analyzed for LOH, with the same microsatellite markers used in the linkage analysis on chromosome 8p23.1-p22 (except for D8S1135). Allelic loss was calculated using a previously described method (17).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A large family with familial thyroid epithelial neoplasia was identified (family 1; Fig. 1). The majority of the patients (n = 11; 69%) had benign thyroid lesions (NG or FTA), and the remaining (n = 5; 31%) presented PTC (supplemental Table 1). None of the five patients with PTC presented recurrence or distant metastases or died from thyroid disease. The mean age of diagnosis was 40 yr (range 13–72 yr), in keeping with previously published data (8).
GeneChip Human Mapping 10K linkage analysis in family 1
A genome-wide linkage analysis of family 1 was conducted, using high-density SNP arrays. Using the complete-penetrance model, haplotype-based linkage analysis yielded a maximal significant parametric LOD score of 4.41 ( = 0.00), at the 14-Mb physical position from chromosome 8p22 (supplemental Fig. 1, published as supplemental data on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org). Single-point and multipoint parametric analyses confirmed the involvement of this region, achieving maximal LOD scores of 2.41 and 4.02 ( = 0.00), respectively. Patients II.15 and II.20 presented telomeric (at rs1445493) and centromeric (at rs822318) recombination events, respectively, delimiting the candidate region to a 7.46-Mb region on chromosome 8p23.1-p22 (Fig. 1). Details of analysis using other linkage models are supplied as supplemental material.
Microsatellite linkage analysis
Linkage to chromosome 8p23.1-p22, was further appraised in family 1, using microsatellite markers. Maximal parametric single-point and multipoint LOD scores of 3.15 ( = 0.00) and 3.62 were obtained at the D8S503 locus, and a nonparametric LOD score of 5.15 (P = 0.002) was obtained at the close adjacent marker D8S550, indicating a significant evidence of linkage to chromosome 8p23.1-p22 (supplemental Table 2). One haplotype, from D8S503 to D8S1135, cosegregated with the affected phenotype in 14 of 16 patients (Fig. 1). Patient III.11, who was a recombinant for all the markers, was likely to be a phenocopy. Patient III.13 shared a small region from D8S503 to D8S550 with the 14 affected members. NG was diagnosed in this patient at the age of 13 yr (supplemental Table 1), but her father was asymptomatic. Thus, the possibility that she represented a phenocopy could not be excluded. Individuals III.14 and III.15, who also shared the affected haplotype, are likely to be asymptomatic carriers, because they were 24 and 21 yr old at the last clinical screening, and the mean age of diagnosis in this family is 40 yr. Data from both SNP genome-wide linkage scan and microsatellite fine-mapping delineated the susceptibility region to a 7.46-Mb span, flanked telomerically by rs1445493 and centromerically by rs822318 on chromosome 8p23.1-p22 (chromosome 8: 9.019–16.482 Mb; Ensembl release 30 January 2008; Fig. 1).
In families 2–6, the LOD scores obtained in the microsatellite linkage analyses of chromosome 8p23.1-p22, were in the majority of cases less than –2 (data available upon request). In family 7, one haplotype cosegregated with the disease, but the analysis did not reach a suggestive evidence of linkage.
Chromosome 8p23.1-p22 gene expression profile in normal thyroid
To identify genes on chromosome 8p23.1-p22 that are expressed in normal thyroid, we analyzed the expression profile from a RNA pool of human thyroids. There were 96 probe sets, assigned with P/M/A calls, which corresponded to 22 well established genes and 11 open reading frames (ORFs), predicted or hypothetical genes, mapping to the 8p23.1-p22 linked region (data shown only for candidate genes; Table 1).
Candidate gene selection and mutation screening
Thirty-two genes, 31 ORFs and predicted/hypothetical genes, and eight pseudogenes, mapped to the 7.46-Mb region of linkage on chromosome 8p23.1-p22 (NCBI release 30 January 2008). We screened three patients with thyroid cancer from family 1 for sequence changes in the coding regions and splice sites of 17 genes (Table 1). A number of previously documented synonymous SNPs were detected, but no potentially pathogenic changes were identified.
LOH analysis
A LOH analysis was pursued in seven tumors from affected members of family 1 to investigate the inactivation of a putative tumor suppressor gene at 8p23.1-p22; however, no allelic losses were detected (results not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We have identified a large Portuguese family, with 11 members affected with benign lesions, such as NG or FTA, and five patients affected with thyroid cancer, which did not behave aggressively in any case. To map the gene predisposing to thyroid epithelial neoplasia in this family, we performed a dense genome-wide linkage scan, using high-density SNP arrays. A genome-wide significant evidence of linkage to chromosome 8p23.1-p22 was obtained, with a maximum parametric haplotype-based LOD score of 4.41 ( = 0.00). Using microsatellite markers, scattered across the 8p23.1-p22 region, a maximal significant multipoint LOD score of 3.62 and a nonparametric LOD score of 5.15 (P = 0.002) were achieved, further confirming linkage to this locus. An affected haplotype segregated with thyroid disease in 14 patients, and recombination events delimited the new susceptibility region to a 7.46-Mb span, flanked telomerically by rs1445493 and centromerically by rs822318 on chromosome 8p23.1-p22. Although a smaller region of haplotype sharing between 15 of the 16 patients (D8S503 to D8S550) was also identified, candidate gene selection was extended to the whole 7.46-Mb region.
Thirteen genes that represented plausible candidates on the basis of either their expression in thyroid and/or possible involvement in cancer were selected for mutational analysis. Four ORFs localized in the small region encompassing D8S503 to D8S550 were also screened for mutations (details on candidate gene selection are supplied as supplemental material). Overall, no pathogenic sequence variations were detected in the coding regions of the 17 candidate genes, making it unlikely that defects in these genes underlie thyroid neoplasia in family 1. However, we cannot exclude the possibility of mutations in regulatory elements of these genes or large germline deletions.
We also searched for LOH in seven tumors from affected members of family 1, focusing on chromosome 8p23.1-p22. However, no LOH was detected in these tumors. In a recent work (12), we observed that allelic losses were not common among familial PTC. It is possible that the somatic inactivation of the putative susceptibility gene, located at 8p23.1-p22, has occurred through other mechanisms, such as microdeletions, point mutation, or promoter hypermethylation. Another explanation would be the involvement of a haploinsufficient tumor-suppressor gene. Additionally, we cannot exclude the possibility of the segregation of a proto-oncogene activating mutation in this family.
We found no evidence for the involvement of the novel candidate region (8p23.1-p22) or the four previously identified candidate loci in the other six families analyzed, suggesting further genetic heterogeneity in FNMTC.
Interestingly, a recent work, identified an approximately 1.4-Mb consensus region for prostate cancer at 8p23.1 (18), overlapping the small region shared by the 15 affected members from family 1.
We used, for the first time, SNP arrays in the genome-wide approach to unveil the genetic contributions to familial thyroid epithelial neoplasia and have identified a new candidate region on chromosome 8p23.1-p22, for which we propose the designation of FTEN. There are still 22 known genes and ORFs in the region of linkage, which were not addressed in the present study, that merit consideration for future analysis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: All authors have nothing to disclose.
First Published Online September 2, 2008
Abbreviations: FNMTC, Familial nonmedullary thyroid carcinoma; FTA, follicular thyroid adenoma; LOH, loss of heterozygosity; MNG, multinodular goiter; NG, nodular goiter; NMTC, nonmedullary thyroid carcinoma; PTC, papillary thyroid carcinoma; ORF, open reading frame; SNP, single-nucleotide polymorphism.
Received February 27, 2008.
Accepted August 25, 2008.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
