| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Departments of Medicine (S.P., R.E.W., N.C., S.R.) and Pediatrics (S.R.), University of Chicago, Chicago, Illinois 60637; the Division of Clinical and Molecular Genetics, Henry Ford Hospital (C.E.J.), Detroit, Michigan 48202; Caylor-Nickel Clinic (D.D.), Bluffton, Indiana 46714; and Howard Hughes Medical Institute and the Department of Pediatrics, Division of Medical Genetics, University of Iowa (J.C.B. and V.C.S.), Iowa City, Iowa 52242
Address all correspondence and requests for reprints to: Dr. Samuel Refetoff, University of Chicago, MC3090, 5841 South Maryland Avenue, Chicago, Illinois 60637. E-mail: refetoff{at}medicine.bsd.uchicago.edu
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We identified a high incidence of severe hypothyroidism due to a complete iodide organification defect in the youngest generation of five nuclear families belonging to an inbred Amish kindred. Genealogical records permitted us to trace their origin to an ancestral couple 7–8 generations back and to identify an autosomal recessive pattern of inheritance. Initial studies of homozygosity by descent using two polymorphic markers within the TPO gene showed no linkage to the phenotype. In fact, 4 of 15 affected siblings from 2 of the nuclear families were heterozygous, resulting in homozygosity values of 73% and 53% in affected and unaffected family members, respectively. A genome-wide homozygosity screen using DNA pools from affected and unaffected family members localized the defect to a locus close to the TPO gene. Linkage analysis using 4 additional polymorphic markers within the TPO gene reduced the number of homozygous unaffected siblings to zero without altering the percent homozygosity initially found in the affected. Sequencing of the TPO gene revealed 2 missense mutations, E799K and R648Q. TPO 779K was found in both alleles of the 11 affected homozygotes, both mutations were present in each of the 3 affected compound heterozygotes, and there were no TPO mutations in 1 subject with hypothyroidism of different etiology. These results demonstrate the power of the DNA pooling strategy in the localization of a defective gene and the pitfalls of linkage analysis when 2 relatively rare mutations coexist in an inbred population.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Iodide is a key element in the synthesis of the principal thyroid hormones, T4 and T3. The active transport of iodide into the thyroid follicular cell is dependent on the sodium iodide symporter (4, 5). Further processing involves the covalent binding of iodine to tyrosine residues (iodide organification) within the TG molecule and their subsequent coupling to form thyroid hormone. The enzyme responsible for tyrosine iodination and coupling is thyroid peroxidase (TPO), a glycoprotein located at the apical membrane of the thyroid follicular cell. TPO activity requires the presence of hydrogen peroxide (H2O2) generated by a still uncharacterized system present in normal thyroid tissue. Thus, in addition to TPO abnormalities (6, 7, 8), defective thyroidal iodide organification could result from defects in the generation of H2O2 (9, 10) and defective expression of the TG gene (11, 12). Furthermore, the association of congenital deafness and iodide organification defect (Pendred’s syndrome) has been recently found to segregate with mutations in a gene closely related to sulfate transporters (13).
Patients with organification defects have a variable degree of primary hypothyroidism and thyroid gland enlargement depending on the severity of the defect. In the untreated patient, a complete defect produces severe hypothyroidism with resulting physical and mental retardation and a large goiter. The proportion of thyroidal iodide discharged by ClO4- is a reliable index of the severity of the defect. A discharge of 20–45% indicates a partial and mild defect, but values greater than 60% and 90% are typical of severe and complete defects (7, 14). With the exception of subjects that have quantitative defects of TG synthesis, all cases of organification defect have elevated serum TG levels when their TSH concentrations are increased.
We identified a high incidence of severe hypothyroidism due to a complete organification defect in the last generation of 5 nuclear families belonging to an inbred Amish kindred. The families were descendants of a single ancestral couple 7–8 generations back. Initial studies showed that tight linkage of the phenotype to the TPO gene was unlikely, as 4 of 15 affected siblings were heterozygous. As, theoretically, the phenotype could arise from other defects than TPO that are involved in iodide organification, we initiated a genome-wide homozygosity mapping. Using DNA pools of affected and unaffected family members, the defect was mapped to a locus close to the TPO gene. Sequencing of the TPO gene identified 2 missense mutations, E799K and R648Q. Eleven affected subjects were homozygous for 1 of the mutations, 3 were compound heterozygotes for both mutations, and 1 had hypothyroidism of a different etiology.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The study group was composed of 5 consanguineous nuclear
families belonging to an Old Order Amish kindred. Their common
ancestors came from Switzerland, after a short stay in France, to
settle in Adams County, IN in 1853 (15, 16). DNA samples were obtained
from 55 subjects, and serum samples were available from 50 subjects.
Fifteen individuals, all belonging to the youngest generation of each
of the 5 nuclear families, had neonatal hypothyroidism. Of the
remaining 40, 10 were the parents of the affected generation of each of
the 5 nuclear families, 18 were unaffected siblings, 11 were
grandparents, and 1 was a great grandparent (Fig. 1
). Subjects are identified in the text
and figures by the family number, generation number, and chronological
order of birth. DNA samples were also obtained from an unrelated Dutch
family (family IV) (7) harboring 1 of the 2 TPO gene mutations found in
the Amish kindred.
|
Tests of thyroid function
Serum total T4 (normal range, 64–154 nmol/L), T3 (normal range, 1.38–2.8 nmol/L; upper limit of normal, 3.25 nmol/L in children <5 yr), and rT3 (normal range, 0.22–0.46 nmol/L) concentrations were measured by double antibody RIAs (Diagnostic Products Corp., Los Angeles, CA; and Biodata S.p.A., Guidonia Montecelio, Italy), and TSH (normal range, 0.4–4.0 mU/L) was determined by a third generation chemiluminescence assay (Corning/Nichols Institute Diagnostics, San Juan Capistrano, CA). The serum free T4 index (normal range, 77–135) was calculated as the product of the serum total T4 and T4 resin uptake values (17). The RIA for serum TG was an in-house assay, as previously reported (18). TPO and TG autoantibodies were measured by agglutination (Fujirebio, Inc., Tokyo, Japan).
For the ClO4- discharge test, individuals were given orally 100–120 µCi (adults) or 68 µCi (children <12 yr) Na123I, and epithyroid counts were obtained every 15 or 30 min over a period of 4 h. NaClO4- was administered (1 g to adults or 0.5 g to children) orally 2 h after the ingestion of radioiodide. The decay of 123I was corrected by counting a standard placed in a neck phantom at each point interval, and counts obtained over the thigh were used for the subtraction of blood background. The percent thyroidal radioiodide uptake (RAIU) was calculated at each point interval, and the effect of ClO4- was expressed as the percentage of the RAIU at 2 h, the time point at which NaClO4- was administered.
Haplotyping and genotyping
Venous blood samples were collected in ethylenediamine tetraacetate. Genomic DNA was isolated from leukocytes as previously described (19). Buccal brushing collected in 50 mmol/L NaOH was used as source of DNA in one infant (4VIII4) from whom blood could not be obtained.
The following tetranucleotide repeat markers were used to haplotype a region on chromosome 2 suspected to contain the mutant gene on the basis of results obtained from genome-wide screening: D2S2976, D2S1329, D2S2952, D2S423, D2S1400, and D2S2961 purchased from Research Genetics, Inc. (Huntsville, AL).
Ten nanograms of DNA were amplified in a 10-µL PCR reaction volume in
the presence of 0.2 µmol/L each of deoxy (d)-ATP dCTP, dGTP, and
dTTP; 0.08 U Taq polymerase; 0.4 pmol
[
-32P]dATP end-labeled forward primer, 3.6 pmol
unlabeled sense primer, and 4 pmol reverse primer. PCR was carried out
in a Perkin-Elmer 9600 thermocycler (Foster City, CA).
Initial denaturation for 3 min at 95 C was followed by 35 cycles (each
for 1 min at 95 C, 1 min at 57 C, and 1 min at 72 C) and was terminated
by a 10-min extension at 72 C. After the addition of 2 vol 95%
formamide, the reaction products were electrophoresed in 6% denatured
polyacrylamide gel, transferred to Whatman paper (Clifton,
NJ), dried, and exposed to x-ray film for 2–16 h.
Seven intragenic polymorphic markers were used to haplotype the TPO
alleles. All consisted of single nucleotide differences identified by
specific restriction endonucleases as shown in Table 1. The relevant areas of the gene were
amplified by PCR using 50 ng template DNA in a 50-µL reaction volume
that included 0.1 U Taq polymerase; 50 pmol of each primer;
and 0.2 mmol/L each of dATP, dCTP, dGTP, and dTTP. PCR was carried out
under the conditions described above with the annealing temperatures
reported in Table 1. The products of amplification were digested with
the appropriate endonucleases, and DNA fragments were separated by 10%
or 14% PAGE and visualized under UV light after staining with ethidium
bromide.
|
). Genome-wide screening
We used DNA pooling to perform genome-wide homozygosity mapping
(20). In this approach, pools of equal amounts of DNA from each
affected member of the same nuclear family are used as templates for
amplification with primers for genetic markers. Markers not linked to
the genetic disorder are expected to have multiple alleles in the
pooled DNA sample, whereas linked markers are expected to show a shift
in allele frequency toward a single allele. Accordingly, four pools of
genomic DNA were prepared. Two were separate pools from affected
individuals belonging to families 2 (2VII1,
2VII2, 2VII4, 2VII6) and 3
(3VII2, 3VII5, 3VII6,
3VII8, 3VII10) (see Fig. 1). The corresponding
controls of pooled DNAs were derived from unaffected siblings
(3VII3, 3VII4, 3VII7,
3VII9, 3VII11, 3VII12) of family 3 and
two unaffected siblings (2VII3 and 2VII5) and two
grandparents (2V3 and 2V4) of family 2. Before
pooling, individual samples were diluted to 100 ng DNA/µL and
separately submitted to PCR with several primer pairs to ensure that
they could be equally well amplified. The primers (screening set
version 6 from Research Genetics, Inc.) used to amplify
short tandem repeat polymorphic markers (STRPs) were developed by the
Cooperative Human Linkage Center (21).
PCRs were performed with 40 ng pooled DNA in a volume of 8.3 µL containing 1.25 µL buffer [100 mmol/L Tris-Cl (pH 8.8), 500 mmol/L KCl, 15 mmol/L MgCl2, and 0.01% (wt/vol) gelatin]; 200 µmol/L each of dATP, dCTP, dGTP, and dTTP; 2.5 pmol/L of each primer; and 0.25 U Taq polymerase. Tetra- and trinucleotide repeat markers were amplified in sets of 2 or 3 per PCR reaction under the following conditions: 35 cycles each of 30 s at 95 C, 30 s at 55 C, and 30 s at 72 C. The products of amplification were analyzed on 6% polyacrylamide gels in denaturing conditions (7.7 mol/L urea). The gels were silver stained according to the method of Bassam et al. (22), and results were scored independently by 2 investigators.
Sequencing the TPO gene
All 17 exons of the TPO gene, including the exon/intron junctions and the promoter region, were sequenced. The oligonucleotide primers used for amplification by PCR were those reported by Bikker et al. (7), modified by eliminating the GC clamp sequence. The products of amplification were sequenced by the d-Rhodamine Terminator Cycle Sequencing Ready Kit (Perkin-Elmer) using the same primers as those used for amplification and 25 cycles each consisting of 10 s at 96 C, 5 s at 50 C, and 4 min at 60 C. The reaction products were subjected to electrophoresis in an ABI Prism 337 automatic DNA sequencer (Perkin Elmer). The TPO gene of the following individuals was sequenced: 2VII2, 4VIII2, and one normal individual, unrelated to the Amish kindred, as a control. The sequence and nucleotide position numbering are according to Kimura et al. (23) and adopted by Bikker et al. (7)
Data analysis
Linkage analyses were conducted using the FASTLINK version (24, 25) of the LINKAGE package (26) and GENEHUNTER (27). Linkage analyses in pedigrees with as many loops as are present in this Amish genealogy are challenging, and exact calculation of likelihood in the pedigree including all known loops is not possible with current algorithms and computer hardware, even for a single marker locus. Thus, only part of the genealogical information can be used. We calculated the evidence for linkage in several ways: 1) each of the five nuclear families was analyzed as an independent, but inbred (first cousin mating) pedigree; 2) the five families were analyzed as a single pedigree, but with all loops broken; and 3) the families were analyzed as five independent families with no inbreeding loops. By contrasting the results of these analyses we were able to determine the effect of the inbred nature of the pedigree on the evidence for linkage even though we could not calculate exact lod scores for the actual pedigree with all inbreeding loops. Analyses were conducted assuming a fully penetrant, rare, autosomal recessive model for transmission of disease with no phenocopies.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Phenotypes were determined on the basis of clinical history of congenital hypothyroidism and findings on physical examination, such as the presence of obvious mental retardation and goiter. All affected subjects were receiving thyroid hormone supplementation at the time of our initial evaluation. In all subjects the diagnosis of hypothyroidism was confirmed by the detection of a high concentration of TSH in serum.
There were a total of 16 affected subjects (5 female and 11 male) and
19 unaffected siblings (12 female and 7 male). However, DNA samples
could not be obtained from 1 affected (5VII4) and 1
unaffected (3VII1) subject of the youngest generation (Fig. 1).
Early developmental delay was documented in five subjects, and three other subjects had gross mental retardation. Because in earlier years neonatal screening was not routinely practiced in this Amish community, the diagnosis of congenital hypothyroidism was not identified during early infancy. Thyroid hormone treatment was delayed in these first-born subjects of three families (1VII4, 3VII2, and 5VII4) until 5–10 months of age when there was already neurological damage. Most cases came to medical attention because of poor eating, large tongue, and constipation. Medical records reported typical features of cretinism with developmental delay, sleepiness, umbilical hernia, and goiter. At the time of our study, seven subjects had goiters (2–5 times the normal size), and one subject with severe mental retardation [who refused to give blood for testing (5VII4)] required thyroid surgery at the age of 14 yr because of a large symptomatic multinodular goiter. Minimal thyroid gland enlargement (1.5–2 times the normal size) was detected in only two unaffected siblings (5VII5 and 5VII9).
Hypothyroidism was confirmed in all 14 affected subjects from whom
serum was available for testing. In 10 of these 14 subjects, the serum
TSH concentration was clearly above the upper limit of normal,
suggesting insufficient thyroid hormone replacement therapy. In 3 of
the 4 subjects with normal or suppressed TSH values, discontinuation of
treatment for 3–6 weeks yielded high serum TSH levels (Fig. 2, see arrows). In the fourth
subject (4VIII2), the serum TSH level was above 100 µU/mL
before the initiation of treatment. The diagnosis of hypothyroidism in
the 2 affected individuals for whom serum samples were not available
(4VIII4 and 5VII4) was supported by the results
of previous serum T4 and TSH determinations. Of note is the
elevation of serum TG levels concomitant with that of TSH (Fig. 2),
suggesting that a quantitative defect in this molecule causing
hypothyroidism was unlikely.
|
(families 1, 2, and 3),
administration of NaClO4 produced a rapid discharge of the
RAIU in the five affected subjects, but not in their parents. The
discharge reached at least 79% within 1 h of the oral
administration of ClO4-. The mean ±
SD maximal discharge was 88.3 ± 6.1% (range,
79–92%) in the affected subjects and 7.4 ± 4.7% in the
unaffected parents.
|
Six subjects (one man and five women) had autoimmune thyroid disease based on the detection of circulating TPO autoantibodies. Three were grandparents (2V2, 5V2, and 4VI3), two were parents (3VI2 and 5VI2), and one was an unaffected sibling (3VII2). Two of them also had TG autoantibodies; one of these (3VI2) had subclinical thyrotoxicosis (TSH, 0.08 mU/L), and the other (5V2) had mild hypothyroidism (TSH, 11.0 mU/L; free T4 index, 4.1).
Initial search for linkage of the phenotype to the TPO gene
Two highly polymorphic intragenic markers were used for
haplotyping to search for the suspected linkage of the iodide
organification defect phenotype to the TPO gene. These markers are
located in exons 1 and 11 of the TPO gene (Table 1). Given the high
inbreeding and the recessive mode of inheritance of congenital
hypothyroidism, we searched for homozygosity by descent within this
candidate gene. Three alleles were identified, but the frequency of the
most common allele was not significantly different between the affected
(87%) and unaffected individuals (71%; Table 2). More importantly,
four affected individuals were heterozygous for this putative involved
allele, and the haplotypes of a pair of affected and unaffected
heterozygous siblings of family 3 were identical (Fig. 4A). To eliminate a possible error in
sample labeling, blood was drawn again from these individuals and
several additional members of their families, but results confirmed the
initial findings.
|
|
As, theoretically, the phenotype could arise from defects in any
step involved in iodide organification, which in addition to TPO
activity requires the generation of H2O2, we
initiated a genome-wide homozygosity mapping study. The method of
comparing DNA pools of affected and unaffected members of the same
nuclear family was used to reduce the effect of population isolation
and inbreeding in biasing the analysis. Of the 314 STRPs, at an average
distance of 10 centimorgans (cM), used for screening, 4 failed to
amplify DNA from pools of both families, and 9 failed to amplify DNA
from pools of 1 of the 2 families. In addition, 3 markers gave results
that were uninformative. Thus, of the 298 STRPs that gave informative
results with DNA from both families, 4 showed divergent patterns
between DNA pools of affected and unaffected family members (Fig. 5). This divergence, exhibiting a trend
toward a single allele in affected individuals, was found in loci on
chromosomes 2, 6, 9, and 14. Screening individual DNAs with these
markers and closely adjacent STRPs, excluded chromosomes 6, 9, and 14
as linked loci and revealed a possible linked locus on the short arm of
chromosome 2 (Fig. 6), covering the
region where the TPO gene has been mapped.
|
|
Although the initial screening using two intragenic polymorphic
markers suggested that linkage of the phenotype to the TPO gene was
unlikely, based on the encouraging results of the genome-wide screen we
undertook the analysis of additional polymorphic TPO markers to
increase the content of information of the previous linkage study.
Typing all individuals with four additional intragenic polymorphic
markers increased the number of identified alleles from three to seven
(Fig. 4B). The identification of allele (A1) was particularly useful
because its frequency was more common in affected individuals (83%)
than in unaffected siblings (31%). Seventy-three percent of affected
individuals were homozygous for this allele, whereas only 3% of all
unaffected subjects and none of the unaffected siblings were homozygous
for any TPO allele (Table 2). However,
the additional markers only reconfirmed the heterozygosity of four of
the affected individuals.
Table 3 summarizes the results of linkage
analyses using all six markers from the TPO region. Although there was
substantial evidence for linkage to this general region in analyses of
the single pedigree with no loops, the estimate of the recombination
frequency precluded tight linkage to the TPO locus. Moreover, once
inbreeding was taken into account, there was substantial evidence
against tight linkage to TPO in families 3 and 4, whereas families 1,
2, and 5 provided strong evidence for linkage and were consistent with
a localization at TPO. Allowing for a small, but nonzero, probability
of phenocopies (nonzero penetrance for heterozygous and nonsusceptible
homozygotes) increases the evidence for linkage in family 3
substantially, but has little effect on the negative lod scores in
family 4, where the pattern of haplotypes is consistent with either
looser linkage to the TPO locus or the presence of multiple mutations
within the genealogy.
|
, family 3), indicating a different etiology
of his thyroid hormone deficiency. In contrast, subject
4VIII1 with confirmed hypothyroidism who was also
heterozygous for the TPO gene, as were his two affected siblings, had
an almost complete ClO4--mediated discharge of
his RAIU (Fig. 3, family 4). Thus, affected members of this nuclear
family have an iodide organification defect as do affected members of
the other three nuclear families that are homozygous for TPO gene
alleles. Linkage analyses in which individual 3VII8 was assumed to be unknown with respect to affection status (as the phenotype in this individual was different from that in all other affected individuals) improved the lod score in family 3 substantially (to 2.74), leaving only the heterozygosity of the affected individuals from family 4 to exclude a single mutation at TPO as accounting for the observations in this pedigree.
Identification of mutations in the TPO gene
The exonic portions and the intron-exon junctions of the TPO gene of subjects 2VII2 and 4VII2 and an unrelated normal control individual were sequenced. Two mutations were identified. One is a missense mutation in exon 14, a G to A transition at nucleotide 2485, that replaces the normal Glu (GAG) at codon 799 with a Lys (AAG). The second, also a missense mutation due to a G to A transition, is located at nucleotide 2033 in exon 11, replacing the normal Arg (CGG) at codon 648 with a Gln (CAG). Both mutations occur at CpG dinucleotide hot spots. In addition, sequencing of the TPO gene revealed two new polymorphisms: C/G at nucleotide 102 (codon 4, CTC and CTG, both Leu) and T/C at nucleotide 2630 [codon 867, GTG (Val) and GCG (Ala)].
Confirmation of and genotyping for the two mutations identified in exons 14 and 11 of the TPO gene
Fragments were amplified with a mismatched primer (Table 1) that
creates a TaqI restriction site in the presence of the
wild-type (WT) nucleotide G2485 in exon 14, but not when it is replaced
by the mutant nucleotide A2485. The mutant nucleotide A2033 in exon 11
creates a new DdeI restriction site that is absent in the
presence of the WT G2033. The polyacrylamide gel in Fig. 7 shows the identification of the two
mutations by digestion with the two restriction endonucleases of the
corresponding DNA fragments amplified from individuals that harbor
these mutations and those that do not.
|
. The
mutation in nucleotide 2485, resulting in the missense mutation E799K,
was present in at least 1 allele of all affected individuals except
3VII8; the latter proved to have hypothyroidism not due to
an iodide organification defect (Fig. 2, family 3). Furthermore, with
the exception of affected subjects belonging to family 4, all others
were homozygous for this mutation. Affected individuals of family 4 had
in addition a different mutation located at nucleotide 2033 in exon 11
of the TPO gene, resulting in the missense mutation R648Q. They were
therefore compound heterozygous, having inherited different mutant
alleles from their parents. The mutation in exon 11 could be traced to
the maternal grandmother (4VI2) and great grandfather
(4V5; Fig. 8). None of the 43 families that expressed the 10
known TPO gene mutations had the substituted nucleotide 2033 in exon 11
(7, 8, 28, 29), indicating that R648Q is not a polymorphism.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The sex distribution among affected subjects was 2:1 in favor of males (10 male and 5 female); the overall sex distribution for the entire generation was 1:1 (18 male and 17 female). Although this is unexpected in an autosomal recessive disease, it is in agreement with the reported predominance of the iodide organification defect in males (14). A delay in the institution of thyroid hormone replacement resulted in severely delayed physical development and irreversible mental retardation. This and the nearly 90% discharge of thyroidal iodide induced by ClO4- attest to the severity of the organification defect.
Genealogical records permitted us to trace the origin of the 5 families to a common ancestral couple 7–8 generations back. Given the autosomal recessive pattern of inheritance, we expect homozygosity at the level of the disease locus in affected individuals due to identity by descent from a common progenitor (30). Thus, the initial finding, from 2 of the nuclear families, that 4 of 15 affected subjects had different TPO alleles was interpreted as indicative of noninvolvement of the TPO gene in the phenotype.
Having confirmed that this result was not the consequence of errors of sample labeling, it was logical to turn to other candidate genes. A deficiency in iodination substrate was excluded by the high serum TG levels and the presence of ample colloid in thyroid tissue removed surgically from one affected subject. A defect in the system that generates H2O2 or in the control of TPO gene expression were therefore, considered. Although data indicate that NADH or NADPH oxidase is involved in the production of H2O2 (31, 32), the enzyme active in thyroid tissue is still unknown (33). Currently little is known about how perturbations in several of the ubiquitous transcription factors regulating TPO expression (34) would affect thyroid function. For these reasons we turned next to genetic mapping through genome-wide screening.
In this approach of homozygosity mapping we applied the strategy of DNA pooling, which has been very successful in identifying recessive disease loci in general and in genetically isolated populations in particular (20, 35, 36). The genome screen narrowed the search to four loci, each on a different chromosome. To our surprise, screening individual DNAs with these markers and some others at closer proximity, excluded three loci and pointed to a possible linked locus on the short arm of chromosome 2 that covered the region where the TPO gene has been mapped (37).
With this information at hand we went back to the haplotyping of the TPO gene, adding 4 more polymorphic markers, expanding the number of alleles identified from 3 to 7. The result was a reduction in the number of homozygotes among unaffected individuals to 1 of 34 without altering the number of homozygotes in the affected group (11 of 15). There were several interpretations of these data. One possibility was that the affected heterozygotes had a different form of hypothyroidism, a hypothesis that could be tested by performing ClO4- discharge tests. Another possibility was that despite high inbreeding, there were different mutations in the TPO gene, as observed in some nonconsanguineous families (7) .
Considering the difficulty and time required to perform the in vivo ClO4- discharge test, this approach as well as the sequencing of the TPO gene were undertaken at the same time. The entire coding region and flanking intronic regions were sequenced from DNA belonging to a heterozygous affected subject of family 4, an affected homozygote, and an unrelated normal individual. Identification of the two missense mutations in the TPO gene almost coincided with the completion of the ClO4- discharge tests 3 and 6 weeks after the discontinuation of L-T4 treatment. One of five hypothyroid individuals in family 3 who was heterozygous (3VII8) failed to discharge radioiodide from his thyroid gland in response to ClO4-. He also did not harbor either of the two mutations identified in the TPO gene of the other members of this kindred. The homozygotes, encompassing all affected subjects of families 1, 2, 3, and 5, harbored the E799K mutation on both alleles. The presence of an iodide organification defect in an affected individual of family 4 was confirmed by the positive discharge test. All affected members of this family were compound heterozygous for two mutations, E799K and R640Q, each inherited from one of the parents.
It is of interest that both mutations are G to A transitions in
mutational CpG dinucleotide hot spots. CpGs are commonly the site of
single base substitutions causing human genetic diseases and DNA
polymorphisms (38, 39, 40). The mechanism is methylation of a purine
followed by deamination, resulting in C
T and GA substitutions
that escape the DNA repair mechanism.
The mutation producing E799K has been previously reported in a Dutch
family with a complete iodide organification defect (7). Drs. Bikker
and de Vijlder provided DNA samples from members of this family (Dutch
family IV) for haplotyping to determine the possibility of common
ancestry. This was not unlikely, considering that some of the
progenitors of the Amish kindred that fled the persecutions in Alsace
in the early 18th century had moved to several European countries,
including the Netherlands. The results of the screen, using polymorphic
markers within the TPO gene, definitely excluded a common origin of
this TPO gene mutation (Fig. 9). The new
mutation, R648Q, was traced to the maternal great grandfather
(4V5; Fig. 1). His origin from a separate Indiana Amish
community could not be traced to the founders, although one of his
sisters (4V2) has intermarried within the same family. These findings
and the fact that both mutations occur in CpG hot spots led us to
hypothesize that the mutations possibly arose recently. This is
supported by the failure to observe familial cretinism in the preceding
generations despite inbreeding.
|
This study underlines the power of the DNA pooling strategy in the localization of defective genes with likely recessive inheritance. Inclusion of one of five pooled DNA samples that did not belong to the proper phenotype due to the failure to recognize the different etiology of the hypothyroidism did not affect the outcome of the analysis. This study also points out the pitfalls of linkage analysis when two relatively rare mutations coexist in an inbred population. Zlotogora et al. (41) observed several different mutations in genes for two different diseases in a small geographic area and suggested that a high mutation rate and selective advantage of the carriers could be responsible. In the Amish kindred reported herein, the explanation for a similar observation is a mutation brought into the inbreeding community as well as possibly a high mutation rate. When dealing with mutations caused by transitions at CpG hot spots, the possibility that they are of recent origin should not be overlooked, especially in the interpretation of genetic data in inbred populations.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received October 29, 1998.
Revised November 24, 1998.
Accepted December 4, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. Deladoey, N. Pfarr, J.-M. Vuissoz, J. Parma, G. Vassart, S. Biesterfeld, J. Pohlenz, and G. Van Vliet Pseudodominant Inheritance of Goitrous Congenital Hypothyroidism Caused by TPO Mutations: Molecular and in Silico Studies J. Clin. Endocrinol. Metab., February 1, 2008; 93(2): 627 - 633. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Avbelj, H. Tahirovic, M. Debeljak, M. Kusekova, A. Toromanovic, C. Krzisnik, and T. Battelino High prevalence of thyroid peroxidase gene mutations in patients with thyroid dyshormonogenesis Eur. J. Endocrinol., May 1, 2007; 156(5): 511 - 519. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Pfarr, G. Borck, A. Turk, U. Napiontek, A. Keilmann, W. Muller-Forell, P. Kopp, and J. Pohlenz Goitrous Congenital Hypothyroidism and Hearing Impairment Associated with Mutations in the TPO and SLC26A4/PDS Genes J. Clin. Endocrinol. Metab., July 1, 2006; 91(7): 2678 - 2681. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Krohn, D. Fuhrer, Y. Bayer, M. Eszlinger, V. Brauer, S. Neumann, and R. Paschke Molecular Pathogenesis of Euthyroid and Toxic Multinodular Goiter Endocr. Rev., June 1, 2005; 26(4): 504 - 524. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S M Park and V K K Chatterjee Genetics of congenital hypothyroidism J. Med. Genet., May 1, 2005; 42(5): 379 - 389. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Caron, C. M. Moya, D. Malet, V. J. Gutnisky, B. Chabardes, C. M. Rivolta, and H. M. Targovnik Compound Heterozygous Mutations in the Thyroglobulin Gene (1143delC and 6725G->A [R2223H]) Resulting in Fetal Goitrous Hypothyroidism J. Clin. Endocrinol. Metab., August 1, 2003; 88(8): 3546 - 3553. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Fugazzola, N. Cerutti, D. Mannavola, G. Vannucchi, C. Fallini, L. Persani, and P. Beck-Peccoz Monoallelic Expression of Mutant Thyroid Peroxidase Allele Causing Total Iodide Organification Defect J. Clin. Endocrinol. Metab., July 1, 2003; 88(7): 3264 - 3271. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Visual Diagnosis: Two Infants Who Have Coarse Facial Features and Growth and Developmental Delay Pediatr. Rev., January 1, 2003; 24(1): 16 - 22. [Full Text] [PDF]
|
|
|
|
|
|
D.-M. Niu, B. Hwang, Y.-K. Chu, C.-J. Liao, P.-L. Wang, and C.-Y. Lin High Prevalence of a Novel Mutation (2268 insT) of the Thyroid Peroxidase Gene in Taiwanese Patients with Total Iodide Organification Defect, and Evidence for a Founder Effect J. Clin. Endocrinol. Metab., September 1, 2002; 87(9): 4208 - 4212. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Kopp Perspective: Genetic Defects in the Etiology of Congenital Hypothyroidism Endocrinology, June 1, 2002; 143(6): 2019 - 2024. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Bakker, H. Bikker, T. Vulsma, J. S. E. de Randamie, B. M. Wiedijk, and J. J. M. de Vijlder Two Decades of Screening for Congenital Hypothyroidism in the Netherlands: TPO Gene Mutations in Total Iodide Organification Defects (an Update) J. Clin. Endocrinol. Metab., October 1, 2000; 85(10): 3708 - 3712. [Abstract] [Full Text]
|
|
|
|
|
|
S. Neumann, H. Willgerodt, F. Ackermann, A. Reske, M. Jung, A. Reis, and R. Paschke Linkage of Familial Euthyroid Goiter to the Multinodular Goiter-1 Locus and Exclusion of the Candidate Genes Thyroglobulin, Thyroperoxidase, and Na+/I- Symporter J. Clin. Endocrinol. Metab., October 1, 1999; 84(10): 3750 - 3756. |
X174 digested with
HaeIII.
