| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Medicine (S.M., A.M.D., X.-H.L., R.E.W., S.R.) and Pediatrics (S.R.) and Committees on Genetics (S.R.) and Molecular Medicine (S.R., R.E.W.), The University of Chicago, Chicago, Illinois 60637; and Division of Endocrinology (S.Y., T.D.), Dokuz Eylül University, TR-35240 Izmir, Turkey
Address all correspondence and requests for reprints to: Samuel Refetoff, University of Chicago, MC 3090, 5841 South Maryland Avenue, Chicago, Illinois 60637. E-mail: refetoff{at}uchicago.edu.
| Abstract |
|---|
|
|
|---|
Objective: The objective of the study was to investigate the genotype of a family with RTH and correlate it to the phenotype.
Design: The DNA was isolated from different tissues, and the sequence of the TRß gene was determined. Clinical studies involved the administration of incremental doses of T3.
Setting: The study was conducted at a referral pediatric endocrinology clinic in Turkey and an academic medical center in the United States.
Main Outcome and Measures: Measurement included markers of thyroid hormone action and sequencing of TRß revealing a R338W mutation.
Patients and Family: We studied two siblings with short stature, panic disorder, psychosis, and high free iodothyronine concentrations with nonsuppressed TSH and their father with similar thyroid function tests without growth or psychiatric abnormalities.
Results: Direct sequencing of gDNA obtained from the fathers leukocytes, buccal mucosa cells, and prostate tissue showed less amplification of the mutant allele (R338W) than the normal allele as confirmed by PCR/restriction fragment length polymorphism analysis. No sequence abnormalities were detected in gDNA from fibroblasts. Similar results were found in mRNA from the leukocytes and fibroblasts. The sensitivity of various tissues to thyroid hormone was not uniform. The progeny had equal amounts of mutant and wild-type gDNA in leukocytes and skin.
Conclusions: The father has a mosaicism for the R338W mutation as it was present in some cell lineages, including his germline, because it was transferred to his children but not in fibroblasts. This indicates that the mutation occurred de novo in early embryonic life. Here is the first report of mosaicism in RTH. The possibility of mosaicism should be considered in subjects with RTH without apparent mutations in the TRß gene.
| Introduction |
|---|
|
|
|---|
Linkage between RTH and the thyroid hormone receptor (TR)-ß gene was shown in 1988 (7). Since then, more than 300 families with the RTH phenotype have been found to harbor mutations in that gene (2). RTH is usually transmitted in an autosomal dominant fashion, but de novo cases are common, and recessive inheritance is rare (8). In our laboratory de novo mutations occurred in 22.5% of the 120 families in which TRß gene mutations were excluded in the parents of an affected individual (2); 44.4% of the de novo mutations occurred in mutagenic CpG dinucleotide hot spots. The mutant TRß molecules have either a reduced affinity for T3 (6, 9) or impaired interaction with one of the cofactors involved in the mediation of thyroid hormone action (10, 11).
However, RTH can occur in the absence of mutations in the TR
or TRß gene and is referred to as non-TR RTH. Since the first demonstration of non-TR RTH (12), 29 subjects belonging to 23 different families have been identified (2, 13, 14, 15). The phenotype is indistinguishable from that in subjects harboring TRß gene mutations. Distinct features are an increased female to male ratio of 2.5:1 and the high prevalence of sporadic cases (2, 13). In only three families had linkage analysis excluded the involvement of the TRß gene (12, 16).
Although not previously reported, the variable presence of a TRß mutation in certain tissues may present as non-TR RTH, depending on the tissue analyzed for the TRß mutation in a subject with RTH. We report for the first time the occurrence of mosaicism in RTH that might be misdiagnosed as non-TR RTH. The affected individual had variable tissue expression of the de novo TRß gene mutation, R338W, which was transmitted to his children.
| Subjects and Methods |
|---|
|
|
|---|
The proposita (II.1), a 36-yr-old Turkish woman, presented with a multinodular goiter, short stature (144 cm), and a depressive disorder. She had serum free T3 of 5.83 pg/ml (1.574.71), FT4 of 4.23 ng/dl (0.81.9), and TSH of 1.21 µIU/ml (0.45), increased thyroidal radioiodide uptake of 52.5% at 24 h, but positive thyroid peroxidase (TPO) and thyroglobulin (TG) antibodies, and normal pituitary magnetic resonance imaging. Her 32-yr-old brother (II.2) was agitated and had evidence of psychosis. He had similar abnormal thyroid function tests (TFTs) and antibodies. Blood samples were obtained from all family members. We found that the father (I.1) also had abnormal TFTs (Fig. 1
), whereas the mother (I.2) had only positive TPO and TG antibodies. The father had normal growth and no psychological problems. Informed consents were obtained for this study approved by the Institutional Review Board of the University of Chicago.
|
Total thyroxine (TT4), total T3, and TSH were measured by chemiluminescence using Elecsys 2010 technology (Roche Molecular Biochemicals GmbH and Hitachi, Ltd., both located in Indianapolis, IN). Total reverse T3 (TrT3) and TG were measured by RIAs. FT4 index (FT4I) was estimated from the resin T4 uptake ratio and the TT4 concentrations. TPO and TG antibodies were measured by an agglutination method.
Clinical studies
To assess the sensitivity of various tissues to thyroid hormone, the father was given incremental doses of L-T3 (50, 100, and 200 µg/d, each for 3 consecutive days). L-T3 was administered orally as a split dose, every 12 h, for a total of six doses per incremental dose (3). Just before beginning the study and 12 h after the last of each incremental dose, blood was drawn for the determination of the hormonal measurements described above as well as for the measurement of serum cholesterol, creatine kinase (CK), ferritin, and SHBG. In addition, body weight and resting pulse rate were recorded. Fifteen healthy volunteers served as controls.
Sequencing of the TRß gene
Genomic DNA (gDNA) was extracted from peripheral blood lymphocytes (PBLs) of the proposita and members of her family. A second blood sample was drawn from the father. In addition, primary cultures of fibroblasts were generated from skin punch biopsies obtained from the proposita and her father. Additional tissues examined were buccal mucosa cells and prostate tissue obtained for diagnostic purposes unrelated to this study. Total RNA was extracted from skin of proposita, her fathers PBLs, and cultured skin fibroblasts (17).
All coding exons of the TRß gene were sequenced from gDNA extracted from PBLs and using primers and conditions described previously (9, 16). The oligonucleotide primers used for amplification and sequencing of the two exons containing nucleotide substitutions were as follows. For exon 4 of the TRß1 gene, primers were 5'-GCCTTCGAAAACTCTGCATCTCA-3' (sense) and 5'-CTTGTTGAAACACAT GATAATGGGCTA-3' (antisense) at 55 C annealing temperature, and those for exon 9 were 5'-CTGTATGTTGTTCCTGACTGGCAT-3' (sense) and 5'-GTGATTGGAATTAGCGCTAG ACAAGCA-3' (antisense) at 59 C annealing temperature. PCR products were confirmed by direct sequencing using ABI PRISM BigDye terminator cycle sequencing reading reaction kits (Applied Biosystem, Foster City, CA).
cDNA was prepared by reverse transcription of mRNA with the Superscript III first-strand synthesis system for RT-PCR (Life Technologies/Invitrogen, Carlsbad, CA) and random hexamers to sequence exons 910 of the TRß gene. The primers for amplification and sequencing were 5'-ACCAGATCATCCTCCTCAAAGG-3' (sense) and 5'-TGAATCCAGTCAGTCTAATCCTCG-3' (antisense). Annealing was at 58 C for 1 min and extension at 72 C for 1 min for each of the 35 cycles.
Genotyping
The mutation was confirmed in gDNA and cDNA by restriction fragment length polymorphism (RFLP) using the restriction endonuclease NlaIII (New England Biolabs, Beverly, MA), which cleaves only the mutant sequences at the site of the nucleotide substitution. The primers, 5'-GTGCGCTATGACCCAGAAAG-3' (sense) and 5'-AGGAGGGCTA CTTCAGTGTC-3' (antisense), amplify a 164-bp segment of exon 9 at 56 C annealing temperature. cDNA was amplified by a two-step PCR. The product of a first PCR amplification using primers located in exons 9 and 10 was used as template for a second PCR to amplify the 164-bp fragment of exon 9 as described above.
The sizes of fragments after restriction with NlaIII were 61, 55, and 48 bp for the mutant allele and 116 and 48 bp for the wild-type (WT) allele. Fragments were well separated on a 16% nondenaturing polyacrylamide gel at 100 V for 1 h.
To determine the relative amounts of mutant and WT alleles, ethidium bromide-strained gels were scanned, and the density of the respective bands was measured using Image J, version 1.34 S (Rasband, W. S., Image J, U.S. National Institutes of Health, Bethesda, MD, http://rsb.info.nih.gov/ij/ (18). Because the mutant allele generated two closely migrating bands, the combined densities were divided by two and corrected for molecular size (x 0.83). The total amount of combined mutant and WT alleles is represented by the density of the 48-kb band (see Fig. 5
). Results are expressed as percentage of the mutant allele relative to the WT allele in heterozygous subjects.
|
All family members were assigned the TRß1 allele haplotypes using a PCR-based silent substitution of nucleotide 461 (TCG
TCA, S54S), which was found in the affected father. The products of PCR amplification using primers for exon 4, as described above, were determined by sequencing.
| Results |
|---|
|
|
|---|
We measured serum TSH at baseline and after administration of incremental doses of L-T3 to the father and 15 normal individuals. Serum TSH level of the father at baseline was 1.3 µU/ml. L-T3 doses of 50, 100, and 200 µg/d suppressed TSH to 46.2, 15.4, and 7.7%, respectively, whereas in control subjects TSH was suppressed to means ± SD of 9.6 ± 4.4, 3.6 ± 1.9, and 2.5 ± 1.7% of baseline (Fig. 2
). Thus, response of the pituitary thyrotrophs to L-T3 in the father was reduced, compared with the normal subjects.
|
|
We identified in the proposita (II.1) and her brother (II.2) a heterozygous mutation at codon 338 (nucleotide 1297 CGG
TGG), located in exon 9 of the TRß gene, which results in the replacement of the normal arginine with tryptophan (R338W). Sequencing of gDNA extracted from PBL of her affected father (I.1) showed a reduced amplification of the mutant allele, compared with the normal allele (Fig. 4
). We obtained a second blood sample and additional tissues from the affected father. These included buccal mucosa cells, prostate tissue, and skin fibroblasts. gDNA extracted from the second blood sample, buccal mucosa, and prostate tissue showed the same results, namely low amounts of the mutant allele relative to the WT allele. However, no sequence abnormalities in exon 9 of TRß gene were detected in gDNA isolated from skin fibroblasts. Findings were similar in cDNA derived from the respective tissues. As shown in Fig. 4
, both TRß gene alleles of the propositas mother (I.2) were WT.
|
TGG in codon 338 creates an additional restriction site (61 and 55 bp) for NlaIII. gDNA extracted from PBLs and cDNA isolated from skin biopsy of the proposita showed the expected additional fragment, whereas these were absent in her mothers gDNA. Similarly gDNA and cDNA isolated from cultured skin fibroblasts of the affected father had only the WT allele. In contrast, gDNA extracted from blood sampled on different occasions, from prostate tissue as well as cDNA isolated from second blood sample of the father showed a weak, though detectable, fragment generated by NlaIII. The average amount of the mutant allele in gDNA of the father extracted from PBL was 8.6%, and that from prostate was 3.9%, compared with 50% of WT in PBLs of the proposita.
To determine the transmission of the mutation to the proposita and her affected brother, family members were haplotyped for a silent substitution of nucleotide 461 (TCG
TCA, S54S) present in the father. As shown in Fig. 6
, both the proposita and her brother inherited the mutant TRß1 allele from their father. Although samples from the parents of the affected father were not available because they were deceased, we could deduce that the mutation occurred de novo in the father during early embryonic life because it was present in some but not all lineages of somatic cells and in the germline (Table 1
).
|
|
| Discussion |
|---|
|
|
|---|
These results demonstrate that the father has a mosaicism for the mutant TRß R338W because it was present in some lineages of somatic cells but not in skin fibroblasts. The transmission of the mutation to his children proves that it is also present in his germline, indicating that the mutation occurred de novo in his early embryonic life. The mutation R338W (CGG
TGG) has been identified in 25 different families (5, 9, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 and unpublished data from our laboratory). This DNA site is likely to be prone to de novo mutation (eight of 25 families) because of its location in a mutagenic CpG dinucleotide hot spot (29). The presence of a single nucleotide difference (G/A 461) in the father, which was also found in his children, demonstrates that it was the paternal allele that underwent the R338W mutation.
The clinical presentation of RTH and the severity of hormone resistance are highly variable. The possibility of variable levels of expression of the mutant allele relative to the normal counterpart has been explored, but results have been inconsistent (30, 31). Somatic mutations in the TRß gene have been reported only in TSH-secreting pituitary tumors presenting as TSH-induced thyrotoxicosis (32, 33). As expected, our clinical studies showed that mosaicism resulted in a variable degree of RTH in different tissues. The thyrotrophs were resistant to thyroid hormone as evidenced by the reduced suppressibility of TSH by L-T3. Failure of ferritin to respond to L-T3 suggests hormonal resistance in red blood cells, spleen, and other tissues that regulate serum ferritin levels. In contrast, the normal responses of SHBG, cholesterol, and CK to L-T3 suggest that the hepatocytes express predominantly the WT-TRß. These results also provide an explanation for the difference in the fathers clinical manifestation, compared to his two children.
Potential mechanisms of non-TR-mediated RTH involve abnormalities at all steps in the mediation of thyroid hormone action (34). Defects in the thyroid hormone transporter MCT8 have been recently identified, but they do not produce a phenotype of RTH (35, 36). Another thyroid hormone transporter, LST-1, was not linked to the RTH phenotype in three families with non-TR RTH studied (Macchia, P. E., R. E. Weiss, and S. Refetoff, unpublished data). Aberrant alternative splicing was the defect causing pituitary RTH in one somatic TRß gene mutation (33). Mutations in the promoter of the TRß gene could reduce the level of gene expression; however, such a defect will produce RTH only in the homozygous state because subjects heterozygous for TRß gene deletion do not manifest RTH (8). Finally, based on mouse models of RTH (37, 38), search for mutations of coregulators in subjects with non-TR RTH have yielded negative results (39). Mosaicism has been reported in a small number of autosomal dominant disorders including pseudoachondroplasia (40), tuberous sclerosis (41), retinoblastoma (42), neurofibromatosis type 2 (43), familial isolated hyperparathyroidism (44), familial hypocalcemia (45), Albright hereditary osteodystrophy (46), and monogenic diabetes mellitus (47). The mosaicism rate for autosomal disorders is 619% (48). Thus, it is reasonable to speculate that TRß gene mutations could have been missed in some patients reported as having non-TR RTH. Often the degree of mosaicism is different among tissues, and the mutation might not be present in the tissue examined. Linkage analysis is also prone to miss the possibility of mosaicism because this method is based on allele inheritance and germline transmission (43, 44). Thus, this is of particular concern in disorders that are frequently sporadic and have increased rates of de novo mutations (49).
Our findings in this family are pertinent to RTH studies because they point to an issue not previously considered in patients with RTH in whom a mutation in TRß is not detected. Therefore, the examination of several different tissues should be considered in cases of putative non-TR RTH.
| Footnotes |
|---|
S.R. is Academic Associate for Quest Diagnostics (San Juan Capistrano, CA).
First Published Online June 27, 2006
Abbreviations: CK, Creatine kinase; FT4, free T4; FT4I, FT4 index; gDNA, genomic DNA; PBL, peripheral blood lymphocyte; RFLP, restriction fragment length polymorphism; RTH, resistance to thyroid hormone; TFT, thyroid function test; TG, thyroglobulin; TPO, thyroid peroxidase; TR, thyroid hormone receptor; TrT3, total reverse T3; TT4, total thyroxine; WT, wild type.
Received April 4, 2006.
Accepted June 16, 2006.
| References |
|---|
|
|
|---|
or ß genes may be due to a defective co-factor. J Clin Endocrinol Metab 81:41964203[Abstract]
-deficient mouse. J Clin Invest 106:7379[Medline]
genes. J Clin Endocrinol Metab 85:36093617This article has been cited by other articles:
![]() |
N K Agrawal, R Goyal, A Rastogi, D Naik, and S K Singh Thyroid hormone resistance Postgrad. Med. J., September 1, 2008; 84(995): 473 - 477. [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 |