Abstract
Mutations in the thyroid hormone receptor β (TRβ) gene result in resistance to thyroid hormone. However, it is unknown whether mutations in the TRα gene could lead to a similar disease. To address this question, we prepared mutant mice by targeting mutant thyroid hormone receptor kindred PV (PV) mutation to the TRα gene locus by means of homologous recombination (TRα1PV mice). The PV mutation was derived from a patient with severe resistance to thyroid hormone that has a frameshift of the C-terminal 14 aa of TRβ1. We knocked in the same PV mutation to the corresponding TRα gene locus to compare the phenotypes of TRα1PV/+ mice with those of TRβPV/+ mice. TRα1PV/+ mice were viable, indicating that the mutation of the TRα gene is not embryonic lethal. In drastic contrast to the TRβPV/+ mice, which do not exhibit a growth abnormality, TRα1PV/+ mice were dwarfs. These dwarfs exhibited increased mortality and reduced fertility. In contrast to TRβPV/+ mice, which have a hyperactive thyroid, TRα1PV/+ mice exhibited mild thyroid failure. The in vivo pattern of abnormal regulation of T3 target genes in TRα1PV/+ mice was unique from those of TRβPV/+ mice. The distinct phenotypes exhibited by TRα1PV/+ and TRβPV/+ mice indicate that the in vivo functions of TR mutants are isoform-dependent. The TRα1PV/+ mice may be used as a tool to uncover human diseases associated with mutations in the TRα gene and, furthermore, to understand the molecular mechanisms by which TR isoforms exert their biological activities.
The thyroid hormone, T3, has profound effects on growth, development, and homeostasis. These biological activities are mediated mainly by thyroid hormone receptors (TRs) that are ligand-dependent transcription factors (1). Three ligand-binding TR isoforms have been identified, TRα1, TRβ1, and TRβ2, which are derived from the TRα and TRβ genes, by alternative splicing of the primary transcripts. Each TR isoform has a unique developmental and tissue-specific expression (1, 2). Studies using a gene-inactivation approach indicate that these TR isoforms have distinct and common functions in vivo (3). The action of TR depends not only on the types of DNA elements on the T3 target genes but also on a host of corepressor and coactivator proteins (1, 2).
Resistance to thyroid hormone (RTH) is a syndrome characterized by reduction in the sensitivity of tissues to the action of thyroid hormones. Mutations in the TRβ gene result in TRβ mutants, which mediate the clinical phenotype by interfering with transcription of T3-regulated genes by means of a dominant negative effect. This disease is manifested by elevated levels of circulating thyroid hormones associated with normal or high levels of serum thyroid-stimulating hormone (TSH) (4). The other clinical features include short stature, decreased weight, tachycardia, cardiac disease and hearing loss, attention-deficit hyperactivity disorder, decreased IQ, and dyslexia (4). Based on the extensive sequence homology in the functional domains of α and β TR and their similar in vitro functional characteristics, it is intriguing that no TRα mutations have ever been found in RTH patients.
It has been postulated that mutations of the TRα gene could be embryonic lethal, inconsequential, or not associated with abnormalities of RTH. To test these possibilities and to understand further the functions of TRα mutants in vivo, we prepared mutant mice by targeting the PV mutation to the TRα gene locus by means of homologous recombination (TRα1PV mice). PV has a mutation in exon 10 of the TRβ gene, a C insertion at codon 448, which produces a frameshift of the carboxyl-terminal 14 aa of TRβ1 (5). We knocked in the same PV mutation to the corresponding TRα gene locus and compared the phenotypes of mice with a similar mutation at the TRβ locus (TRβPV mice; ref. 6). TRα1PV/+ mice show significant mortality, and TRα1PV/PV mice were rarely obtained and died shortly after birth. TRα1PV/+ mice were dwarfs and exhibited increased mortality, reduced fertility, and mild thyroid failure. These different phenotypes in the pituitary–thyroid axis of TRα1PV/+ and TRβ1PV/+ mice are consistent with the fact that no TRα mutations could be identified in RTH patients. The abnormal regulation patterns of T3 target genes differed in the tissues of TRα1PV/+ and TRβ PV/+ mice. The distinct phenotypes exhibited by TRα1PV/+ mice indicate that the in vivo signaling pathways of TR mutants are isoform-dependent.
Materials and Methods
Preparation of TRα1PV Mutant Mice.
The targeting vector, pmTRα-1, was modified as described by Wikstrom et al. (7). A 522-bp fragment containing exons 9 and 10 of the TRα gene was released by digesting the plasmid with StuI and SalI. In its place, a 209-bp fragment containing the mutated TRα sequence (TRα1PV) was ligated to give an intermediate plasmid (pmTRα1PV-Intermediate). This 209-bp fragment was obtained by PCR by using the mouse TRα1 cDNA as a template and the following two primers: 5′ primer, 5′-AGTCAGGAGGCCTACCTGCTGGCGTTTGAGCACTAC; 3′ primer, 5′-TGAGTCGTCGACAGATCTTCAGTCTAATCCTCGAACGGATCCAAGAACAAAGGGGGGAAGAGTTCTGTGGGGGCACTCGACTTTCATGTGGAG.
The 3′ primer contained a C insertion at the mouse TRα1 cDNA nucleotide position 1173 followed by a human PV sequence, and a BamHI site was placed downstream of the stop codon (see Fig. 2A). The resulting DNA fragment contained a StuI and SalI at the 5′ and 3′ ends, respectively. The sequence of this insert was confirmed by DNA sequencing.
Figure 2.
Targeting of the PV mutation onto the TRα gene locus by homologous recombination. (A) Schematic representation of the TRα1PV targeting vector; the 10-kb targeting vector contains the PV mutation in exon 9. The locations of the NeoR and the HSV-TK genes are indicated. (B) Restriction map of the wild-type TRα gene locus; the expected sizes of the fragments digested by BamHI, which were detected by internal and external probes, respectively, are indicated. (C) Restriction map of the targeted TRα1PV gene locus. The expected sizes of the fragments digested by BamHI, which were detected by the internal and external probes, are indicated. SV40, simian virus 40.
To enhance the selection efficiency, a 1.854-kb herpes simplex virus thymidine kinase (HSV-TK) cDNA was placed 1.2 kb upstream of the pmTRα1PV-Intermediate. The HSV-TK cDNA was restricted from pPyfEnh TK (8) by XhoI and HindIII and blunted by using Klenow fragment. pmTRα1PV-Intermediate was restricted with EcoRV, and both ends were blunted by treating with Klenow. The blunt-ended HSV-TK cDNA was ligated to the blunt-ended pmTRα1PV-Intermediate to give the final targeting vector that we designated as pTRα1PV (see Fig. 2A). The inserted HSV-TK cDNA was confirmed by sequencing.
The targeting vector was linearized by XbaI digestion and transfected (25 μg) into the TC-1 embryonic stem cells. The selection of the recombinant clones was performed as described (6). From six transfection experiments, six positive recombinant clones were identified by Southern blotting analysis. The recombinant clones were microinjected into C57BL/6J blastocysts to produce chimeras that were crossed with NIH Black Swiss mice to establish germ-line transmission. Mice harboring the targeted PV mutation were designated as TRα1PV mice (see Fig. 2C).
F-2 mice were genotyped by PCR by using two sets of primers. For the identification of the wild-type allele, the sequence of the 5′ primer (W5) is 5′-TCTTGTCCTCGGGCCTCATGCC, and the 3′ primer (W3) is 5′-CTCTGGCCGCCTGAGGCTTTAG. For the identification of the mutant allele, the sequence of the 5′ primer (M5) is 5′-CTGTGCGTGGACAAGATCGAGA, and the 3′ primer (M3) is 5′-CTGACCGCTTCCTCGTGCTTTACG. PCRs were carried out for 35 cycles (94°C, 30 s; 63°C, 30 s; 74°C, 30 s for the determination of the wild-type allele) (94°C, 30 s; 63°C, 30 s; 74°C, 30 s for the mutant allele) in a buffer containing MgCl2 by using TaKaRa EX-Taq polymerase (Intergen, Purchase, NY).
Northern Blot Analyses.
Total RNA (5–10 μg) was used for Northern blot analysis. After electrophoresis, RNA was transferred onto membranes (Hybond N+, Amersham Pharmacia), which were hybridized with appropriate probes. cDNA probes for the mouse wild-type TRα1, glycoprotein α common subunit (α-SU), TSHβ, growth hormone (GH), malic enzyme (ME), type 1 deiodinase (D1), myelin basic protein (MBP), and Pcp2 genes were labeled with [α-32P]dCTP by using a random primer hexamer protocol. For quantification, the intensities of the mRNA bands were normalized against the intensities of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA. Thus, the blots were stripped and rehybridized with 32P-labeled cDNA for GAPDH. The quantification of the bands was performed by using a Molecular Dynamics PhosphorImager.
Transient Transfection Assay.
Transient transfection experiments were carried out by using CV1 cells as described by Zhu et al. (9). Briefly, cells were transfected with plasmids containing the reporter pdoublePal-TK-Luc (0.4 μg; a generous gift of J. L. Jameson, Northwestern University School of Medicine, Chicago) and the expression plasmid for TRβ1 (pCLC51; 0.2 μg), or for TRα1 (pCLC61; 0.2 μg) in the absence or presence of the expression plasmid for TRα1PV mutant (pCDNA3.1 TRα1PV; 0.2 or 1 μg) or for TRβ1PV (pCLC51PV; 0.2 or 1 μg). Five hours after transfection, cells were incubated in T3-deficient medium (Td medium). Twenty hours after transfection, T3 (100 nM) was added and incubated for an additional 24 h. Cells were lysed, and the luciferase activity was determined. The values were normalized against the protein concentrations that were determined by the BCA protein assay kit (Pierce).
Hormone Assays.
The serum levels of total T4 (TT4) and T3 (TT3) were determined by using a Gamma Coat T4 and T3 assay RIA kit, respectively, and free T4 (fT4) was determined by using Clinical Assays Gamma Coat Free T4 (direct one-step) (DiaSolin, Stillwater, MN) according to the manufacturer's instructions. TSH levels in serum were measured as described (10).
Data Analysis.
All data are expressed as mean ± SE. Statistical analyses used the Student's t test, and P < 0.05 was considered significant. Group analyses were also conducted by using single-factor ANOVA.
Results
The TRα1PV Interferes with the Transactivation Activity of the Wild-Type TRα1 and TRβ1 in Vitro.
Because naturally occurring TRα1 mutants have never been identified in RTH patients, we first evaluated whether an identical mutation in the corresponding TRα gene resulted in functional impairment of the wild-type TRs. An expression vector containing both the cytomegalovirus and T7 promoters was constructed by a C insertion at the nucleotide position 1180 of the mouse TRα1 cDNA followed by the human TRβ1PV mutant sequence (5). This C insertion led to a frameshift mutation to give the TRα1PV protein sequence. TRα1PV protein had a total of 409 aa, 1 aa shorter than the w-TRα1. We prepared TRα1PV protein by in vitro transcription/translation and evaluated its T3 and transactivation activities. Similar to that found for TRβ1PV (11), TRα1PV had totally lost the T3 binding activity (data not shown). Fig. 1 shows that, consistent with the loss of T3 binding activity and similar to TRβ1PV (bar 6 vs. 2), TRα1PV had also lost its T3-dependent transactivation activity (bar 4 vs. 2). Comparison of bars 9 and 10 with bar 8 shows that the T3-dependent transactivation activity mediated by w-TRβ1 was repressed 85% and 92% by the cotransfection of the TRα1PV expression plasmid at the TRα1PV/w-TRβ1 ratios of 1 and 5, respectively. A similar extent of repression on the w-TRα1 transactivation activity by TRα1PV (TRα1PV/w-TRα1 plasmid ratios of 1 and 5 for bars 15 and 16, respectively) was also detected (80% and 92% repression shown in bars 15 and 16 vs. bar 14, respectively). These results indicate that TRα1PV was a potent dominant negative mutant receptor that interferes with the transcriptional activity of the w-TRβ1 and w-TRα1. Interesting, we found that TRβ1PV exerted less dominant negative action than TRα1PV on w-TRβ1 transactivation activity (65% and 80% repression at TRβ1PV/w-TRβ1 plasmid ratios of 1 and 5 for bars 11 and 12, respectively) but exerted a slightly more potent dominant negative action than TRα1PV on w-TRα1 transcriptional activity (92% and 97% repression at TRβ1PV/w-TRα1 plasmid ratios of 1 and 5 for bars 17 and 18, respectively).
Figure 1.
TRα1PV and TRβ1PV repress the transactivation activity of w-TRα1 and w-TRβ1. CV1 cells (1.25 × 105 cells/well) were cotransfected with pdoublePal-TK-Luc (0.4 μg) and the expression plasmid for w-TRα1 (pCLC61; 0.2 μg) or w-TRβ1 (pCLC51; 0.2 μg) in the absence or presence of TRα1PV (pCDNA3.1 TRα1PV; 0.2 or 1 μg) or TRβ1PV (pCLC51PV; 0.2 or 1 μg). The luciferase activity was determined as described in Materials and Methods. The data are expressed as mean ± SD (n = 3).
Generation of Mice with Targeted TRα1 Mutant PV.
Fig. 2A shows the targeting vector that contained the PV mutation in exon 9 (see Materials and Methods). Other features of the 10-kb targeting vector included a BamHI site, which was placed immediately after the stop codon for screening purposes (Fig. 2A). Positive embryonic stem cells were identified by using Southern blot analysis by both internal and external probes (Fig. 2A). BamHI digests of the genomic DNA revealed 20- and 12.5-kb fragments when hybridized with the internal probe encompassing the PV mutation site in exon 9 (Fig. 2C; data not shown), whereas the wild-type clones showed the expected 20-kb band (Fig. 2B). When the same BamHI digests were hybridized with the external probe (Fig. 2C), a 9-kb fragment was detected together with the 20-kb fragment derived from the wild-type allele (Fig. 2C; data not shown). These results indicate that the PV mutation was correctly targeted onto the TRα gene locus.
The mice carrying the TRα1PV gene were identified by PCR analysis by using genomic DNA prepared from mouse tails (Fig. 3). Using the primers pairs specifically for detecting the wild-type (W5 and W3; Fig. 3A) and mutant genes (M5 and M3; Fig. 3B), the PCR products with sizes of 586 bp (lane 4 of Fig. 3C) and 313 bp (lanes 1 and 3 of Fig. 3C) were detected for the mutant mice and wild-type mice, respectively. The genotypes of the mutant mice were further confirmed by Southern blot analysis (data not shown).
Figure 3.
Genotyping of TRα1PV/+ mice by PCR. Using the primer pairs of M5 and M3 shown in B, a 586-bp fragment was obtained from TRα1PV/+ mice (C, lane 4). Using the primer pairs W5 and W3 as indicated in A, a 313-bp fragment was obtained from the genomic DNA of the same mice (C, lane 3). When the 586-bp fragment was absent, the mice were wild type (C, lane 2).
Fertility and Survival.
TRα1PV/+ mice were fertile; however, the fertility of TRα1PV/+ mice was reduced as shown by decreases in the frequencies of pregnancy and the litter size (Table 1). Compared with the wild-type mice, the success rate in mating was reduced to 62.5% for both female and male TRα1PV/+ mice (n = 8). The litter size was reduced from 10–12 to 3–5 per litter. When both female and male TRα1PV/+ mice were used in the mating, the frequency of pregnancy was further reduced to 21.4%, and the average litter size was 3.0 (n = 14). These results indicate that both the male and female TRα1PV/+ mice were less fertile than wild-type mice. As a result of the reduction in the fertility in both male and female TRα1PV/+ mice, we had no success in obtaining homozygous offspring despite numerous repeated attempts for 18 months. There was only one homozygous neonate that died shortly after birth with unknown cause.
Table 1.
Reduced fertility in TRα1PV/+ mice
Mice | Pregnancy/mating, % | Litter size, number of pups |
---|---|---|
Wild-type × wild-type | >99 (n = 10) | 10–12 |
F-TRα1PV/+ × M-wild-type | 62.5 (n = 8) | 3.2 |
M-TRα1PV/+ × F-wild-type | 62.5 (n = 8) | 5.1 |
M-TRα1PV/+ × F-TRα1PV/+ | 21.4 (n = 14) | 3.0 |
In addition, even though both male and female TRα1PV/+ mice were viable, the mortality rate was high, in that 30% of mice died (24 of 79). This finding is in contrast to the absence of death for the wild-type siblings (0 of 99) and tow deaths for the TRβPV/+ mice (2 of 49) during the same observation period (Fig. 4). Death occurred frequently before reaching adulthood, in that 83% (20 of 24) and 96% (23 of 24) of deaths occurred by 6 weeks and 11 weeks of age, respectively. No obvious cause of death could be found. These results suggest that mutation of the TRα gene seemed to compromise the ability of TRα1PV/+ mice to survive, especially during early life. Because it was not possible to obtain homozygous mice, the characterization of the phenotypes was focused on the TRα1PV/+ mice.
Figure 4.
Comparison of survival rates of TRα1PV and TRβPV/+ mice. The survival rates of TRα1PV/+ (n = 79), TRβPV/+ (n = 49), and wild-type sibling (n = 99) mice were determined within 300 days.
The TRα1PV Gene Is Expressed in All Expected T3 Target Tissues Examined.
Two sets of primers flanking the mutated exon 9 were used to assess the expressed RNA in different tissues. Using the primer pairs of 5N and 3N (Fig. 5A) or 5N and 3PV (Fig. 5B), cDNA fragments with sizes of 444 bp or 304 bp were expected for the expression of wild-type and mutant alleles, respectively. Representative results from the expression of mutant PV in the heart and liver are shown in Fig. 5C. For the TRα1PV/+ mice, both 444 and 304 bp were detected (lanes 3 and 4 for the heart and 7 and 8 for the liver; Fig. 5C). Using this analysis, we found that the TRα1PV allele was expressed in the cerebrum, cerebellum, pituitary, muscle, white and brown adipocytes, lung, spleen, and kidney (data not shown), indicating that the PV mutant allele was detected in the expected T3 target tissues.
Figure 5.
Expression of TRα1PV mRNA in the tissues of TRα1PV mice by reverse transcription–PCR (RT-PCR) analysis. Total RNA was isolated from the heart (C, lanes 1–4) and liver (C, lanes 5–8). RT-PCR was carried out by using the primer pairs of 5N and 3N as shown in A and 5N and 3PV shown in B for the w-TRα1 cDNA and TRα1PV cDNA, respectively. The 444-bp and 304-bp fragments represent the expression of the wild-type and mutant alleles, respectively, as shown in C. The genotypes are marked.
The expression of the TRα1PV allele was further confirmed by Northern blot analysis by using cDNA encoding TRα1. Wild-type TRα1 mRNA was expressed with a size of 5.0 kb (12) and the TRα1PV mRNA as a size of 1.8 kb of TRα1PV/+ mice. In TRα1PV/+ mice, the expression of TRα1 mRNA was weaker than that in wild-type mice as expected (data not shown).
TRα1PV/+ Mice Are Dwarfs.
TRα1PV/+ mice exhibited severe growth impairment evident shortly after birth. TRα1PV/+ mice were dwarfs, and by 4 weeks of age, the mean weight of TRα1PV/+ male pups was 10.07 ± 2.9 g (n = 10), which was ≈40% less than that of their wild-type siblings 16.39 ± 1.39 g (n = 23; P < 0.0001; Fig. 6A). The lengths of male TRα1PV/+ were also shorter. By 4 weeks of age, the mean length of TRα1PV/+ male pups was 6.6 ± 0.3 cm (n = 10), which was 17% shorter than that of their wild-type siblings at 8.0 ± 0.3 cm (n = 23; P < 0.0001; Fig. 6A). The weight and length differences persisted into adulthood.
Figure 6.
Impairment in growth of male (A) and female (B) TRα1PV/+ mice. The weights and lengths of wild-type and TRαPV/+ mice were measured over the first 10 postnatal weeks. Significant differences in weights and lengths were detected in both male and female TRαPV/+ mice (P < 0.0001).
Similar growth impairment was also detected in female TRα1PV/+ mice. At the age of 4 weeks, the mean weight of female TRα1PV/+ pups was 10.08 ± 1.12 g (n = 7), which was 30% less than that of their wild-type siblings (14.78 ± 1.2 g; n = 23), and the mean length was 6.6 ± 0.2 cm (n = 7), which was 15% shorter than that of their wild-type siblings (7.7 ± 0.2 cm; n = 23; P < 0.0001; Fig. 6B).
Mild Thyroid Failure in TRα1PV/+ Mice.
To evaluate whether the mutation of the TRα gene leads to functional impairment in the pituitary–thyroid axis, we determined the serum levels of TT4, fT4, TT3, and TSH. As shown in Fig. 7A, the mean TT4 concentration was 4.77 ± 1.07 μg/dl (n = 21) in TRα1PV/+ mice, which was not significantly different from that in the wild-type mice (4.48 ± 1.13 μg/dl; n = 27). This was further confirmed by fT4 (Fig. 7B) in that the mean concentration was 49.79 ± 7.72 pmol/liter (n = 13) and 44.68 ± 8.65 pmol/liter (n = 19) for the TRα1PV/+ and wild-type mice, respectively (P = 0.1, no significant differences).
Figure 7.
Concentrations of total T4 (A), free T4 (B), and total T3 (C) in wild-type and TRα1PV/+ mice. No significant differences in TT4 (A) and fT4 (B) between wild-type and TRα1PV/+ mice were detected. TT3 was 15% higher in TRα1PV/+ mice than in wild-type mice (P < 0.005 by Student's t test).
However, as shown in Fig. 7C, a significant 15% increase was found in the mean TT3 concentration for TRα1PV/+ mice (1.67 ± 0.23 ng/ml; n = 21) as compared with their wild-type siblings [1.43 ± 0.31 ng/ml; n = 26; P < 0.005 by Student's t test: F(1,47) = 8.502, P = 0.005 by ANOVA]. Furthermore, histological assessment indicates that neither abnormal follicular morphology nor lymphoid infiltration was detected in the thyroid glands of TRα1PV/+ mice (data not shown). These findings are in great contrast to TRβPV/+ mice, which have a 2- to 4-fold higher levels of circulating thyroid hormone and enlarged thyroid follicles (6).
Intriguingly, TRα1PV/+ mice had a significantly higher TSH level (Fig. 8). The mean TSH concentration in TRα1PV/+ mice was 47.58 ± 16.16 ng/ml (n = 21), which was 1.7-fold higher than that in their wild-type siblings [27.45 ± 13.48 ng/ml; n = 27; P < 0.0001 by Student's t test; F(1,48) = 21.005, P = 0.00004 by ANOVA]. The lack of lymphoid infiltration in the thyroid glands of TRα1PV/+ mice makes autoimmune thyroid disease less likely to account for the results of thyroid function tests. The elevated TSH, together with lower T4/T3 ratios (2.85 for TRα1PV/+ mice vs. 3.1 for their wild-type siblings), suggested mild thyroid failure in TRα1PV/+ mice. The causes of the mild thyroid failure are not clear at present.
Figure 8.
Elevated serum TSH levels in male (A) and female (B) TRα1PV/+ mice. The mean TSH concentrations in male (A) and female (B) TRα1PV/+ mice were significantly higher than those of wild-type mice (P < 0.02 and P < 0.0003, respectively).
Abnormal Expression Patterns of T3-Target Genes in Tissues of TRα1PV/+ Mice.
As a first step to understanding the molecular action of TRα1PV in vivo, we evaluated the expression of T3 target genes in the tissues of TRα1PV/+ mice. Fig. 9A shows the expression of TSH and GH mRNA in the pituitary gland. TSH consists of two polypeptides, the TSH-specific β subunit and the common α-SU, whose expression is suppressed by T3. There was no difference in the expression levels of TSHβ between wild-type and TRα1PV/+ mice (Fig. 9Ab). It is not clear why there is a lack of concordance between the circulating TSH level and the TSHβ mRNA expression. One possibility could be that the circulating TSH in TRα1PV/+ mice is more stable than that in the wild-type mice. However, a 2.3-fold increase in the expression of α-SU was detected in TRα1PV/+ mice (Fig. 9Aa), indicating a selective activation of the α-SU gene in the pituitary by thyroid hormone. In TRβPV/+ mice, which have 2- to 4-fold elevated thyroid hormone levels, a 1.7-fold activation of α-SU gene was observed (ref. 6 and Table 2). However, no effect on the expression of the GH gene was observed (Fig. 9Ac), which was similar to that seen in TRβPV/+ mice (ref. 6 and Table 2).
Figure 9.
Abnormal expression patterns of T3-target genes in the tissues of TRα1PV/+ and TRβPV/+ mice. Total RNA was isolated from the pituitary (A), liver (B), and cerebellum (C) of TRα1PV/+ mice or the cerebellum of TRβPV/+ mice (D). Northern blot analyses were carried out as described in Materials and Methods. The levels of the expression of the T3-target genes were normalized by using GAPDH mRNA. Quantification was performed by using Molecular Dynamics PhosphorImager. The mean fold of changes is indicated (n = 3–4 for each group of 12-week-old male mice).
Table 2.
Comparison of the expression of T3-target genes in the tissues of TRα1PV/+ and TRβPV/+ mice
Target gene | mRNA Fold of mRNA in wild-type mice
|
|
---|---|---|
TR1αPV/+ | TRβPV/+ | |
α-SU* | 2.3 | 1.7 |
TSHβ* | 1.0 | 1.0 |
GH† | 1.0 | 1.0 |
ME† | 2.2 | 0.6 |
D1† | 9.2 | 0.8 |
MBP† | 1.4 | 1.0 |
Pcp2† | 1.2 | 1.0 |
T3 negatively regulated genes.
T3 positively regulated genes.
A different profile of abnormal regulation of T3 target genes was detected in the liver. ME and D1 are T3-positively regulated genes and were activated 2.2-fold (Fig. 9Ba) and 9.2-fold (Fig. 9Bc) in TRαPV/+ mice, respectively, suggesting a hypersensitivity of these two genes to thyroid hormone. Previously, an increased sensitivity of target genes to thyroid hormone in mice with complete deficiency of TRα (TRα1−/− TRα2−/−) was also reported (13). This finding is in contrast to the abnormal expression patterns of these two genes observed in TRβPV/+ mice in which a repression was detected (Table 2; ref. 6). The increased expression of D1 in the liver is consistent with the increased circulating levels of TT3 in TRα1PV/+ mice.
More differences in the abnormal expression patterns of T3 target genes were also found in the cerebellum. Fig. 9 C and D compares the expression of MBP and Pcp2 in the cerebellum of TRα1PV/+ and TRβPV/+ mice, respectively. No changes in the expression of MBP and Pcp2 were found in TRβPV/+ mice, whereas an activation of 1.4-fold and 1.2-fold was detected for the expression of MBP and Pcp2 genes, respectively, in TRα1PV/+ mice (Fig. 9 C and D and Table 2).
Discussion
The generation of the TRα1PV knock-in mice has provided evidence that the mutation of the TRα gene is not inconsequential. The mild thyroid failure in the TRα1PV/+ mice, which is clearly distinct from the hyperactive thyroid exhibited by the TRβPV and TRβΔ337T) knock-in mice (6, 14), is consistent with the fact that no TRα mutations have ever been identified in RTH patients. RTH patients are diagnosed based on high circulating levels of thyroid hormones associated with inappropriately normal or elevated TSH (4). The findings that TRα1PV/+ mice were viable and the majority of them could survive into adulthood suggested that the mutation of one of the TRα gene alleles is not embryonic lethal. Therefore, it is reasonable to postulate that mutations of the TRα gene could occur in humans, but the phenotypes would be expected to be distinct from those of RTH. The present mouse model could be used as a tool to uncover human diseases associated with the mutations of the TRα gene.
Mice deficient in TRα1 (TRα1 knockouts) exhibit mild hypothyroidism with lower levels of thyroid hormones and TSH, normal reproduction ability, and mild abnormal heart rate and body temperature (15, 16). The mild phenotype exhibited by mice deficient in TRα1 suggests that TRβ could compensate for the overlapping functions (3). Although it is not known whether TRα1PV/+ mice also have the phenotypes of abnormality in heart function and body temperature, the known phenotypes of TRα1PV/+ mice characterized so far are clearly distinct from mice deficient in TRα1. It is remarkable that the mutation of one of the TRα alleles led to more severe phenotypes than the inactivation of both TRα1 alleles.
Mice deficient in both TRα1 and TRα2 (TRα1−/− TRα2−/− mice) exhibited severe hypothyroidism, growth arrest, reduced postnatal survival, and intestine abnormality (17). The reduced postnatal survival could be rescued by injection of T3 in TRα1−/− TRα2−/− mice (17). The known phenotypes of TRα1PV/+ mice clearly are different from those of TRα1−/− TRα2−/− mice in that TT4 and TT3 were nearly normal such that the reduced postnatal survival is most likely not because of thyroid hormone deficiency. Moreover, the known phenotypes of TRα1PV/+ mice are also distinct from TRα1−/− TRβ−/− mice (3, 18), which manifest more severe phenotypes, notably in the differences of the impairment of the pituitary–thyroid axis. The latter exhibit hyperactive thyroid glands with very high thyroid hormone levels together with extraordinarily high TSH (3, 18). These observations suggest that TRα1PV could act by means of a gain of function, rather than simply as a result of interfering with the functions of both TRα1 and TRβ functions. This notion is further supported by the transgenic mice expressing v-erbA, which is a non-T3 binding homolog of TRα1 (19). The phenotypes of the transgenic mice expressing v-erbA overlap in many aspects to TRα1PV/+ mice in that increased mortality, reduced fertility, and reduction in weight were observed (19).
As summarized in Table 3, the phenotypes exhibited by the TRα1PV/+ mice are clearly different from those of TRβPV/+ mice. Therefore, the in vivo signaling pathways of α and β TR mutants apparently are distinct. At present, the molecular mechanisms by which these two TR mutant isoforms exert their distinct phenotypes are not clear. Based on the abnormal expression patterns of T3 target genes in the pituitary, liver, and cerebellum of TRβPV/+ mice (Table 2), TRβPV could interfere with the functions of wild-type TRs by means of a dominant negative effect (6). TSHβ and α-SU are T3-negatively regulated genes. Instead of being repressed by the high thyroid hormone levels in TRβPV/+ mice, no repression and further activation was observed for TSHβ and α-SU, respectively (Table 2). GH, ME, and D1 are T3-positively regulated genes. However, no activation was seen in the expression of the GH gene, and repression was detected in the expression of ME and D1 genes in TRβPV/+ mice (Table 2). However, we could not exclude the possibility that TRβPV could also act by means of a T3-independent pathway.
Table 3.
Distinct phenotypes manifested by TRα1PV/+ and TRβPV/+ mice
Phenotype | TRα1PV/+ mice | TRβPV/+ mice |
---|---|---|
Dysfunction of the pituitary–thyroid axis | Mild thyroid failure | Hyperactive thyroid |
Impaired growth | Dwarfs | No |
Reduced survival | Yes | No |
High mortality | Yes | No |
Reduced fertility | Yes | No |
Expression of ME and DI mRNA | Activation | Repression |
Expression of MBP and Pcp2 mRNA | Activation | No change |
In contrast to TRβPV, the abnormal expression patterns of T3 target genes in vivo shown in Table 2 do not support the notion that TRα1PV acted by means of a dominant negative effect. In contrast to TRβPV/+ mice, the expression of α-SU in the pituitary, ME and D1 in the liver, and MBP and Pcp2 in the cerebellum was activated. This observation is not because of the possibility that TRα1PV cannot interfere with the transcriptional activity of the wild-type TRs. In vitro studies have shown that TRα1PV did not bind T3, lacked in vitro transactivation activity, and acted to repress strongly the transactivation activity of both w-TRβ1 and w-TRα1. Therefore, the dominant negative effect of TRα1PV could be masked in vivo by coregulatory proteins that act in a TR mutant isoform-dependent way (20). Alternatively, it is possible that TRα1PV structurally and functionally is intrinsically different from TRβ1PV; therefore, it could act by means of a T3-independent pathway differently from TRβ1PV. This notion is supported by the recent x-ray crystallographic studies of the ligand binding domains of the wild-type TRα and TRβ isoforms (21). The two wild-type TR subtypes differ in the loop between helices 1 and 3, which could affect both ligand recognition and interaction with coactivators and corepressors (21). In line with these findings, it is reasonable to postulate that the differences in the structures of the TRα1PV and TRβ1PV isoforms could interact and recruit corepressors differently, leading to distinct functional consequences. The important in vivo role of corepressors in the action of mutant TRβ is exemplified by the manifestation of specific and deleterious action in mutant mice harboring TRβ(Δ337T) mutant (14). However, the validation of these possibilities needs further study.
Acknowledgments
We thank J. L. Jameson for the generous gift of pdoublePal-TK-Luc and L. Zhao for her assistance in the transient transfection experiments.
Abbreviations
- T3
3,3′,5-triiodo-l-thyronine
- T4
l-thyroxine
- TR
thyroid hormone receptor
- GH
growth hormone
- RTH
resistance to thyroid hormone
- PV
mutant thyroid hormone receptor kindred PV
- TSH
thyroid-stimulating hormone
- MBP
myelin basic protein
- ME
malic enzyme
- α-SU
α-glycoprotein subunit
- D1
type 1 deiodinase
- GAPDH
glyceraldehyde-3-phosphate dehydrogenase
- HSV-TK
herpes simplex virus thymidine kinase
References
- 1.Cheng S-y. Endocrinol Metab Dis. 2000;1:9–18. [Google Scholar]
- 2.Yen P. Physiol Rev. 2001;81:1097–1142. doi: 10.1152/physrev.2001.81.3.1097. [DOI] [PubMed] [Google Scholar]
- 3.Forrest D, Vennstrom B. Thyroid. 2000;10:41–52. doi: 10.1089/thy.2000.10.41. [DOI] [PubMed] [Google Scholar]
- 4.Weiss R E, Refetoff S. Endocrinol Metab Dis. 2000;1:97–108. doi: 10.1023/a:1010072605757. [DOI] [PubMed] [Google Scholar]
- 5.Parrilla R A, Mixson A J, McPherson J A, McClaskey J H, Weintraub B D. J Clin Invest. 1991;88:2123–2130. doi: 10.1172/JCI115542. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kaneshige M, Kaneshige K, Zhu X, Dace A, Garrett L, Carter T A, Kazlauskaite R, Pankratz D G, Wynshaw-Boris A, Refetoff S, et al. Proc Natl Acad Sci USA. 2000;97:13209–13214. doi: 10.1073/pnas.230285997. . (First Published November 7, 2000; 10.1073/pnas.230285997) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wikstrom L, Johansson C, Salto C, Barlow C, Baas F, Thoren P, Vennstrom B. EMBO J. 1998;15:455–461. doi: 10.1093/emboj/17.2.455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mansour S L, Thomas K R, Capecchi M R. Nature (London) 1988;336:348–352. doi: 10.1038/336348a0. [DOI] [PubMed] [Google Scholar]
- 9.Zhu X-G, Kaneshige M, Bhat M K, Zhu Q, Mariash C N, McPhie P, Cheng S-y. Mol Cell Biol. 2000;20:2604–2618. doi: 10.1128/mcb.20.7.2604-2618.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Pohlenz J, Maqueem A, Cua K, Weiss R E, Van Sande J, Refetoff S. Thyroid. 1999;9:1265–1271. doi: 10.1089/thy.1999.9.1265. [DOI] [PubMed] [Google Scholar]
- 11.Meier C A, Dickstein B M, Ashizawa K, McClaskey J H, Muchmore P, Ransom S C, Menke J B, Hao E-H, Usala S J, Bercu B B, et al. Mol Endocrinol. 1992;6:248–258. doi: 10.1210/mend.6.2.1569968. [DOI] [PubMed] [Google Scholar]
- 12.Lazar M A, Hodin R A, Darling D S, Chin W W. Mol Endocrinol. 1988;2:893–901. doi: 10.1210/mend-2-10-893. [DOI] [PubMed] [Google Scholar]
- 13.Macchia P E, Takeuchi Y, Kawai T, Cua K, Gauthier K, Chassande O, Seo H, Hayashi Y, Samarut J, Murata Y, et al. Proc Natl Acad Sci USA. 2001;98:349–354. doi: 10.1073/pnas.011306998. . (First Published December 19, 2000; 10.1073/pnas.011306998) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hashimoto K, Curty F H, Borges P P, Lee C E, Abel E D, Elmquist J K, Cohen R N, Wondisford F E. Proc Natl Acad Sci USA. 2001;98:3998–4003. doi: 10.1073/pnas.051454698. . (First Published March 6, 2001; 10.1073/pnas.051454698) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Johansson C, Vennström B, Thorén P. Am J Physiol. 1998;275:R640–R646. doi: 10.1152/ajpregu.1998.275.2.R640. [DOI] [PubMed] [Google Scholar]
- 16.Wilkström L, Johansson C, Saltó C, Barlow C, Campos Barros A, Baas F, Forrest D, Thorén P, Vennström B. EMBO J. 1998;17:455–461. doi: 10.1093/emboj/17.2.455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Fraichard A, Chassande O, Plateroti M, Roux J P, Trouillas J, Dehay C, Legrand C, Gauthier K, Kedinger M, Malaval L, et al. EMBO J. 1997;16:4412–4420. doi: 10.1093/emboj/16.14.4412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Göthe S, Wang Z, Ng L, Nilsson J, Campos-Barros A, Ohlsson C, Vennström B, Forrest D. Genes Dev. 1999;13:1329–1341. doi: 10.1101/gad.13.10.1329. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Barlow C, Meister B, Lardelli M, Lendahl U, Vennstrom B. EMBO J. 1994;13:4241–4250. doi: 10.1002/j.1460-2075.1994.tb06744.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lin K H, Wu Y H, Chen S L. Endocrinology. 2001;142:653–662. doi: 10.1210/endo.142.2.7927. [DOI] [PubMed] [Google Scholar]
- 21.Wagner R L, Huber B R, Shiau A K, Kelly A, Cunha Lima S T, Scanlan T S, Apriletti J W, Baxter J D, West B L, Fletterick R J, et al. Mol Endocrinol. 2001;15:398–410. doi: 10.1210/mend.15.3.0608. [DOI] [PubMed] [Google Scholar]