Life Without Thyroid Hormone Receptor

Abstract

Thyroid hormone (T3) is critical not only for organ function and metabolism in the adult but also for animal development. This is particularly true during the neonatal period when T3 levels are high in mammals. Many processes during this postembryonic developmental period resemble those during amphibian metamorphosis. Anuran metamorphosis is perhaps the most dramatic developmental process controlled by T3 and affects essentially all organs/tissues, often in an organ autonomous manner. This offers a unique opportunity to study how T3 regulates vertebrate development. Earlier transgenic studies in the pseudo-tetraploid anuran Xenopus laevis revealed that T3 receptors (TRs) are necessary and sufficient for mediating the effects of T3 during metamorphosis. Recent gene knockout studies with gene-editing technologies in the highly related diploid anuran Xenopus tropicalis showed, surprisingly, that TRs are not required for most metamorphic transformations, although tadpoles lacking TRs are stalled at the climax of metamorphosis and eventually die. Analyses of the changes in different organs suggest that removal of TRs enables premature development of many adult tissues, likely due to de-repression of T3-inducible genes, while preventing the degeneration of tadpole-specific tissues, which is possibly responsible for the eventual lethality. Comparison with findings in TR knockout mice suggests both conservation and divergence in TR functions, with the latter likely due to the greatly reduced need, if any, to remove embryo/prenatal-specific tissues during mammalian postembryonic development.

Thyroid hormone (T3) plays a critical role in metabolism and organ physiology, and thyroid diseases are one of the most prevalent types of metabolic disorders (1-6). Insufficient or excessive T3 in humans results in abnormal metabolic rates and adversely affects the physiological functions of many organs, such as heart and liver (7-11). T3 also contributes to human health by regulating development. Peak levels of T3 are present in the several months around birth in humans, when many organs/tissues mature into their adult forms, during a period referred to as postembryonic development (12, 13). T3 deficiency during human postembryonic development causes defects, such as goiter formation and cretinism (14, 15). The roles of T3 in development and adult organ function are conserved in other vertebrates such as mouse and anurans (1, 2, 12, 16). Perhaps the most dramatic process regulated by T3 is the metamorphosis of anurans, such as the highly related pseudo-tetraploid Xenopus laevis and diploid Xenopus tropicalis, which involves drastic transformation of essentially every organ as the tadpole is metamorphosed into frog (12, 16-18).

T3 binds to nuclear T3 receptors (TRs), which are DNA-binding transcription factors that are located predominantly in the nucleus (2, 19-22). Extensive genetic studies have provided strong support for TRs in mediating the developmental effects of T3 in vertebrates. Here I review some studies on the role of TR during vertebrate development, with a focus on Xenopus laevis and Xenopus tropicalis.

Gene Regulation by TR and a Dual-Function Model for TR During Xenopus Development

Two TR genes, TRα and TRβ genes, exist in all vertebrates, although in some species one or both TR genes are duplicated, and they encode highly conserved proteins that bind T3 with high affinities (20). In mammals, the TRα gene encodes 2 major proteins, TRα1 and TRα2, due to alternative splicing at the 3′-end, with TRα2 unable to bind T3. There are also several other less-studied proteins encoded by the TRα gene via alternative translation start sites/splicing (23). Similarly, the mammalian TRβ gene produces 2 T3-binding TR proteins due to the use of 2 promoters with distinct tissue specificities (24). In addition, transcription from downstream promoters in the mouse TRα and TRβ genes can produce truncated proteins that lack DNA- or both DNA- and hormone-binding capacities (23, 25), although little is known about their roles in vivo. In the anuran Xenopus laevis, there are 2 TRα genes and 2 TRβ genes due to duplication in the pseudo-tetraploid genome (26, 27), while only a single gene for each exists in the highly related diploid species Xenopus tropicalis (28). Each of the duplicated Xenopus laevis TRβ genes also has 2 promoters (26). However, it is worth pointing out that it is still unknown if Xenopus TRα and TRβ genes encode other protein forms due to alternative splicing, downstream promoters, and/or alternative translation start sites.

TRs belong to the superfamily of nuclear hormone receptors that also include 9-cis retinoic acid receptors (RXRs), which form heterodimers with TRs (2, 20, 22, 29, 30). T3 can both activate and repress target genes via TR. Most studies on gene regulation by TR have been on T3-inducible genes. For such genes, TRs bind to T3 response elements (TREs) in target genes mainly as TR/RXR heterodimers and repress or activate transcription in the absence or presence of T3, respectively (2, 22, 29-33). Unliganded TRs recruit corepressor complexes while liganded TR recruits coactivator complexes to target promoters to remodel chromatin and regulate transcription (Fig. 1) (2, 34-48).

Figure 1.

A dual-function model for TR in frog development. Upper: The levels of endogenous T3 and T4 (49), and the relative expression of TRα, TRβ, and RXRα during development in Xenopus laevis. The embryos hatch around stage 35 and the tadpole begins to feed around stage 45 (50). The mRNA levels for TRα (solid line), TRβ (broken line), and RXRα (bold line) (51, 52) are plotted on arbitrary scales. Lower: The expression of T3-inducible genes during three periods of frog development. T3-inducible genes are expressed at basal levels during embryogenesis when the levels of T3 and TRs are low or nondetectable. After onset of tadpole feeding at stage 45 but before stage 54, there is little T3 and the expression of high levels of TRα leads to the binding of unliganded TR-RXR heterodimers to TREs in TR target genes and the recruitment of corepressor complexes such as those containing N-CoR and histone deacetylase (HDAC)-3. This results in gene repression, accompanied with increased levels of gene repression histone marks and reduced levels of gene activation histone marks. After stage 55, the binding of endogenous T3 to TR causes corepressor complex release and the recruitment of coactivator complexes such as those containing histone acetyltransferases SRC and p300 and histone methyltransferase PRMT1, leading to opposite changes in histone modification and the loss of nucleosomes around TRE, which are likely critical for the activation of target genes and metamorphosis.

The TRα and TRβ TR genes have distinct expression profiles during Xenopus development. There are few TR mRNAs during embryogenesis. TRα mRNA in whole animals increases around stage 41, shortly after tadpole hatching (stage 35/36), and reaches high levels by the onset of feeding (stage 45) (Fig. 1) (51, 53). It remains high throughout metamorphosis (Fig. 1). In contrast, TRβ mRNA level is low throughout premetamorphosis but is upregulated during metamorphosis, paralleling the rise in plasma T3 concentration (Fig. 1) (26, 51, 53, 54). In addition, RXR genes, particularly RXRα, are also expressed in premetamorphic and metamorphosing tadpoles (Fig. 1) (52). Furthermore, there is a temporal correlation of TR mRNA levels with organ-specific metamorphosis (19, 28, 52). These expression profiles and the ability of TR to regulate gene expression in a T3-dependent manner have led to a dual-function model (Fig. 1) (19, 55). That is, T3-inducible genes are expressed at basal levels for embryogenesis. By stage 45 when feeding begins (50), embryonic development is complete and these genes are no longer needed. The expression of TR and RXR genes, particularly TRα and RXRα, enables the binding of the target genes by TR/RXR heterodimers, which recruit corepressors to repress the genes since there are little T3 present during premetamorphosis. This in turn prevents premature metamorphosis and thus ensures a proper period of tadpole growth. By stage 54, endogenous T3 level rises, leading to the binding of T3 to TR. The liganded TR then releases the corepressor complexes from the target genes and recruits coactivator complexes to activate the genes to levels even higher than those during embryogenesis. This in turn activates the metamorphic process.

Extensive studies over the years have been carried out to test this model in Xenopus laevis, including transient and transgenic overexpression of wild-type and mutant TRs and cofactors during development. These studies demonstrated that TR is both necessary and sufficient to mediate the effects of T3 during metamorphosis (56-67). They further showed that corepressor complex recruitment by unliganded TR controls the timing of the onset of metamorphosis (68) while coactivator recruitment by liganded TR is important for metamorphosis, particularly the rate of metamorphic progression (69-71).

Mechanistically, chromatin immunoprecipitation (ChIP) assays have revealed that histone deacetylase-containing corepressor complexes are associated with the TRE regions of endogenous TR target genes during premetamorphosis (63, 72, 73), while coactivator complexes containing histone acetyltransferases and methyltransferases are associated with the TRE regions during either T3-induced or natural metamorphosis (69-71, 74-77). More importantly, ChIP assays have demonstrated chromatin remodeling at TR target genes during development. Consistent with the observation that unliganded TR can bind to a TRE in chromatin (31, 52), ChIP assays have shown that during premetamorphosis, when there is little T3, TR is bound to TREs in TR target genes and recruits corepressor complexes, leading to increased levels of transcription repression histone marks, such as trimethylation of histone H3K27, and reduced levels of transcription activation histone marks, such as histone H3 acetylation and H3K79 methylation (Fig. 1). Conversely, during T3-induced or natural metamorphosis, the opposite changes in histone modifications occur in the TRE regions (Fig. 1) (46, 57, 78-80). Furthermore, ChIP assays on total histones in the TRE regions have shown that liganded TR causes a local loss of about 2 to 3 nucleosomes (Fig. 1) (79), in agreement with studies on T3-responsive promoters in the reconstituted Xenopus oocyte transcription system (32, 81). This loss of nucleosomes is likely mediated by the recruitment of chromatin remodeling complexes such as those containing Brg1 and BAF57 (40, 82). Thus, TR regulates target gene transcription at least in part via chromatin remodeling, including the removal of the nucleosome near the TRE and histone modifications, to facilitate transcriptional machinery assembly in the promoter region.

Unliganded TR Controls Metamorphic Timing

Perhaps most important test for the dual-function model relies on functional studies of endogenous TRs during Xenopus development. These have been made possible in recent years by the adaptation of Transcription Activator-Like Effector Nucleases (TALEN) and Clustered, Regularly Interspaced, Short Palindromic Repeats (CRISPR) nuclease-mediated gene-editing to knockout genes in Xenopus laevis and Xenopus tropicalis (83-88). While Xenopus laevis has been used for most earlier studies on anuran metamorphosis, its pseudo-tetraploid genome makes it harder for knockout studies. Thus, several laboratories have taking advantages of the diploid genome of Xenopus tropicalis to investigate the function of endogenous TRs during development (89-102). As expected from the low levels or lack of expression of TR during embryogenesis, knocking out individual TR genes or both TRα and TRβ has no obvious effect on embryogenesis (Table 1). Consistent with the existence of only 2 TR genes in vertebrates, removing both TR genes abolishes tadpole responses, including morphological changes and regulation of direct TR target genes, to exogenous T3 treatment of premetamorphic tadpoles, which should have little endogenous T3 as the thyroid gland is not yet fully developed (100). It is worth pointing out that T3 also has nongenomic effects via binding to cytoplasmic and plasma membrane proteins, at least in cultured cells (103-105). The availability of premetamorphic TR double-knockout tadpoles that lack endogenous T3 offers an opportunity to study such TR-independent effects in vivo, even though T3 treatment of such tadpoles does not induce metamorphic changes. More importantly, knocking out both TR genes accelerates premetamorphic tadpole development, enabling the tadpoles to initiate metamorphosis (reaching the onset of metamorphosis or stage 54) at younger age (100), supporting the dual-function model. Similar premature initiation of metamorphosis occurs in tadpoles with TRα knockout but not TRβ knockout (Table 1) (89-94, 99). These findings suggest that TRα is the main receptor controlling metamorphic timing, consistent with the expression profiles of the 2 TR genes during premetamorphosis (Fig. 1).

Table 1.

Developmental Effects of Knocking Out Individual or Both TR Genes in Xenopus tropicalis

		Embryogenesis stages 1-45	Premetamorphosis 45-54	Prometamorphosis 54-58	Early metamorphic climax 58-62	Late climax 62-66
Rate of development	TRα KO	No effect	Acceleration	Delay	No effect	No effect
	TRβ KO	No effect	No effect	No effect	Minor delay	Delay
	TRα+β KO	No effect	Acceleration	Further delay	Delay, then death	N/Aa
Limb	TRα KO	No effect	Acceleration	Delay
	TRβ KO	No effect	No effect	No effect
	TRα+β KO	No effect	Acceleration	Further delay		N/Aa
Bone	TRα KO
	TRβ KO
	TRα+β KO			Premature ossification	Premature ossification, additional vertebrae	N/Aa
Intestine	TRα KO	No effect	No effect	No effect	Delay	Delay
	TRβ KO	No effect	No effect	No effect	Delay	Delay
	TRα+β KO	No effect	No effect	No effect	Complex effectsb	N/Aa
Dorsal skin	TRα KO
	TRβ KO
	TRα+β KO				Inhibition of larval keratin gene repression	N/Aa
Gill	TRα KO
	TRβ KO
	TRα+β KO				Inhibition of gill resorption	N/Aa
Tail	TRα KO	No effect	No effect	No effect	No effect	No effect
	TRβ KO	No effect	No effect	No effect	Delay	Delay, especially notochord
	TRα+β KO	No effect	No effect	No effect	Inhibition	N/Aa

Note that “No effect” refers to little or no obvious effect reported and that a blank space indicates no reported data.

Abbreviations: KO, knockout; N/A, not applicable; TR, T3 receptor.

^aDevelopment is stalled at stage 61 for up to 2 weeks or so before eventual death, while wild-type tadpoles begin tail resorption around stage 61 and complete the process in a week or so.

^bInhibition of both larval epithelial cell death and adult epithelial stem cell development, but premature formation of adult epithelial folds. Y. Shibata and Y.-B. Shi, unpublished observation.

An unexpected finding from the TR knockout studies is that knocking out TRα alone or both TRα and TRβ leads to faster tadpole growth during premetamorphosis; that is, the resulting tadpoles are larger than the wild-type siblings of the same age, likely due to the de-repression of the growth hormone genes, while no such effect has been observed for animals lacking just TRβ (89-94, 99, 100). Interestingly, despite the faster growth of tadpoles lacking either TRα alone or both TRα and TRβ, these animals are smaller than the wild-type siblings at the onset of metamorphosis (stage 54) due to the fact that wild-type siblings develop much slower during premetamorphosis, thus having extra time to grow before initiating metamorphosis at stage 54 (93, 94, 100). Thus, unliganded TRs, mainly TRα, appear to have 2 functions during premetamorphosis, controlling rates of growth and development through independent pathways.

Liganded TRα and TRβ Have Distinct, Stage-Dependent Roles in Regulating the Rate of Metamorphic Progression

Metamorphosis can be subdivided into prometamorphosis and metamorphic climax. Prometamorphosis covers the period from stage 54, the onset of metamorphosis when plasma T3 and T4 become detectable, to stage 58, when forelimbs erupt from the enclosures (Fig. 1) (50, 106). Metamorphic climax (stages 58 to 66) refers to the period of most dramatic morphological changes when tadpole stops feeding, the levels of T3 and T4 reach maximal levels and tail resorption takes place (as reflected most obviously by tail length shortening during the late climax) (Fig. 1, Table 1) (50, 106). Knocking out TR genes has distinct effects on different periods of metamorphosis. TRα knockout delays prometamorphic development; that is, the resulting tadpoles advancing more slowly between stage 54 and 58 compared with the wild-type siblings, but has little effect at metamorphic climax stages (Table 1) (89, 90, 92-94), likely due to compensation by the rising expression of TRβ at climax stages (Fig. 1). On the other hand, TRβ knockout has little effect on prometamorphosis, likely reflecting low levels of TRβ expression during this period (Fig. 1) but causes developmental delays at climax stages (Table 1) (91, 92, 99). Despite these effects, animals lacking either TR gene can complete metamorphosis and develop to reproductive adults with no obvious morphological abnormalities. On the other hand, when both TR genes are knocked out, more severe developmental delays occur during both prometamorphosis and early metamorphic climax (Table 1). The TR double-knockout animals are developmentally stalled near the end of early climax at stage 61 for about 2 weeks before eventual death (Table 1) (100). These findings indicate that TRα and TRβ have both independent and redundant/compensatory roles in mediating the metamorphic effects of T3 and that at least one copy of either TR gene is required to ensure the completion of metamorphosis and animal survival. While the underlying mechanisms remain unclearly, they likely involve the recruitment of coactivators by liganded TR to the target genes. Consistently, a recent study showed that knockout of the coactivator SRC3 leads to inhibition of T3-induced gene expression and metamorphosis (77).

TR Is Not Required for Metamorphosis of Many Organs/Tissues but Is Essential for the Removal of Larval Organs/Tissues

The ability of tadpoles lacking both TRα and TRβ to develop up to metamorphic climax indicates that many organs/tissues can metamorphose without any TR. On the other hand, tadpoles completely devoid of T3 are developmentally stalled around stage 51/52 (16, 106). Thus, considering the ability of unliganded TR to repress T3-inducible genes, it seems that de-repression of TR target genes due to the knockout of both TR genes is sufficient to allow many organs and tissues to undergo metamorphosis.

On the other hand, TR double-knockout tadpoles are developmentally stalled at stage 61 for 2 weeks or so before death without any noticeable reduction in tail length (Fig. 2, Table 1) (100). Under the same conditions, wild-type tadpoles at stage 61 reach the end of metamorphosis with tail completely resorbed within 1 week (Fig. 2). While the exact cause of the death due to TR double knockout remains to be determined, tail resorption appears to be blocked in the absence of TR.

Figure 2.

TR is essential for the completion of metamorphosis and tadpole survival. TR double-knockout (TR-DKO) tadpoles can develop up to stage 61 at the metamorphic climax. Compared with wild-type (WT) animals at stage 61, TR-DKO tadpoles have normal limbs and body structure but larger gills, suggesting reduced or inhibited gill resorption. For WT animals, tail resorption begins around stage 61 and completes within 1 week or so by stage 66. The TR-DKO animals are developmentally stalled at stage 61 for about 2 weeks and then die.

Comparisons of gene expression and tissue morphology between wild-type and TR double-knockout tadpoles at stage 61 have revealed that many adult organ/tissues appear to develop quite well in the absence of TR (Fig. 2 and Table 1) (100). For example, the forehead and body structure as well as the limbs are similar between TR double-knockout and wild-type tadpoles (Fig. 2). The activation of adult keratin gene in the dorsal skin and bone ossification also take place in stage 61 tadpoles lacking TR (Table 1), although ossification actually occurs precociously in the TR double-knockout tadpoles during metamorphosis (Fig. 3).

Figure 3.

TR double knockout leads to premature and increased hypochord ossification. Whole mount staining, with Alizarin red (for ossification) and Alcian blue (for cartilages), of wild-type (TRα ^(+/+)TRβ ^(+/+)) tadpoles at stage 58 (A), stage 62 (B), and stage 63 (C); and a double-knockout (TRα ^(−/−)TRβ ^(−/−)) tadpole at stage 58 (D). Note that in wild-type tadpoles, the ossification of the hypochord (arrowhead) was not detected at stage 58 (A) but present at stage 62 (B) and increased further at stage 63 (C). The TR double-knockout tadpole at stage 58 already had strong ossification of the hypochord (D), indicating premature ossification due to the knockout. In addition, the TR double-knockout animal at stage 58 also had additional ossified vertebrae at the end of the body, i.e., increased ossification. Double arrowheads: Notochord. Bars: 1 mm. See (100) for more details.

Interestingly, the resorption of the gills is inhibited by TR double knockout (Fig. 2). Both gills and tail are larval specific organs and the resorption of either one during metamorphosis requires TR, raising the possibility that liganded TR plays an essential role in larval tissue resorption. Consistent with this, the repression of larval keratin genes during metamorphosis is also inhibited in stage 61 tadpoles lacking TR (Table 1). In addition, the intestine, like the body skin, undergoes extensive remodeling to the adult form with a multiply folded epithelium that can self-renew via proliferation of adult epithelial stem cells, resembling adult mammalian intestine (107-110). This process transforms a simple tubular structure with a single epithelial fold into a complex structure with multiple epithelial folds in a process that involves complete degeneration of the larval epithelium via apoptosis and concurrent de novo formation of adult epithelial stem cells, followed by their proliferation and differentiation. In the TR double-knockout tadpoles, both larval epithelial cell death and adult stem cell formation and proliferation are blocked (Shibata Y and Shi Y-B, unpublished observation). On the other hand, the formation of adult epithelial folds takes place prematurely in the absence of TRs, accompanied by precocious development of the connective and muscles (Shibata Y and Shi, Y-B, unpublished observation). These observations again suggest that removal of larval tissues requires liganded TR while the formation of adult structure can occur without TR.

Conservation in TR Function

There are only 2 known TR genes, TRα and TRβ, in all vertebrates, with the exception of duplication of TRα and/or TRβ in some species. Like in Xenopus, the expression of TRα is activated before the synthesis of endogenous T3, and earlier and at higher levels than TRβ expression during development in vertebrates such as mouse (111, 112) and fishes (113-116). Thus, there are likely conserved roles of TR during vertebrate development, including a need for gene repression by unliganded TR, particularly TRα, during early development before the maturation of the thyroid gland. Complete deletion of all known T3 receptors encoded by TRα and TRβ has been carried out in Xenopus tropicalis, zebrafish, and mouse. Zebrafish lacking all TR genes (2 duplicated TRα genes and the TRβ gene) can develop into adulthood with defects in cone development and chromophore usage in the retina, although the effects of the knockout on other tissues/organs have not been reported (117).

In both mouse and Xenopus, there are multiple tissue defects suggesting considerable overlap in TR functions. For example, in the mouse, phenotypes of growth and reproduction (118), bone and intestinal maturation (118, 119), auditory function (120), and function of the pituitary-thyroid axis (118, 119) are considerably worse in the TR double-knockout mice than in mice with knockout of either TRα or TRβ. These findings have led to proposals that the total levels of active TR proteins in some cells may be more important than which particular TR isoform is present (112), similar to what is observed during Xenopus tropicalis development (Table 1). Thus, in both mouse and Xenopus tropicalis, it is often more beneficial to have 2 rather than 1 TR gene to provide greater sensitivity and flexibility in setting the total TR levels expressed in different cell types. This would support the concept that T3 signaling is sensitive to quantitative mechanisms for receptor levels.

It is remarkable that in the absence of all known T3 receptors, development proceeds to metamorphic climax in the frog with many adult organs formed, and at least in the mouse, to adulthood (118). In mice, most individuals can survive beyond 1 year, despite being small and having multiple tissue defects. It has been speculated that these completely TR-deficient mice have a milder phenotype than severely hypothyroid mice potentially because in hypothyroid mice, unliganded TR causes more damage than does the absence of TR (118). This is likely due to transcriptional repression of target genes by unliganded TRs (121, 122), leading to abnormal development of adult organs. In this regard, it is interesting to note that unliganded TR also prevents adult organ development in tadpoles as TR double knockout leads to premature development of many adult tissues in Xenopus. A major difference between mouse and Xenopus is that the latter has a biphasic developmental process with tadpoles able to live in the absence of T3 for years (106). The transition from a tadpole to a frog involves not only the development of adult tissues but also the removal of some organs entirely, such as the gills and tail, and others partially, such as the intestine and skin. Such a drastic tissue degeneration process requires TR and is absent during mouse development, which may underlie the different outcome upon TR double knockout, that is, tadpole death at the metamorphic climax versus survival of the mouse through postembryonic development to adulthood.

There are also specific as well as shared roles for TR isoforms in mammals (112), as also seen during Xenopus development. Most studies on TR knockout animals are on adult organs in mouse but on tadpole organs during metamorphosis in frog or on different organs in the 2 species, making it difficult to compare organ-specific function of TRs between mouse and frog. One organ that has been analyzed in TR knockout animals in both mouse and frog is the intestine. The adult Xenopus intestine is formed during metamorphosis while the mouse intestine matures into the adult form after birth, with both processes occurring when plasma T3 level peaks and involving the formation of adult intestinal stem cells, suggesting a conserved role of T3 in intestinal development (Fig. 4). Knockout of TR genes, particularly TRα, leads to intestinal defects in adult mice, including reduced stem cell proliferation, with more severe impairments in the TR double-knockout animals than TRα knockout animals (119, 123). These resemble what observed during Xenopus metamorphosis (Table 1). Furthermore, gene expression analyses suggest that in both mouse and Xenopus, TRα regulates some of the known stem cell pathways, such as the WNT-β catenin pathway, in the intestine (102, 124-126).

Figure 4.

The maturation of the mouse intestine after birth (upper panel) resembles the remodeling of the Xenopus intestine during metamorphosis (lower panel). In both cases, adult stem cells are formed from larval/neonatal epithelial cells when plasma T3 level becomes high. In mouse, the cells in the intervillus region develop into adult stem cells expressing protein arginine methyltransferase 1 (PRMT1) and hedgehog (hh) (green cells with irregular-shaped dark nuclei) and invaginate into the underlying connective tissue to form the crypts. In Xenopus, most larval epithelial cells undergo apoptosis in response to rising levels of T3 during metamorphosis but some larval epithelial cells undergo dedifferentiation to develop into adult stem cells expressing, as in mouse intestine, high levels of PRMT1 and sonic hedgehog (hh) (green cells with irregular-shaped dark nuclei). In addition, adult stem cells in Xenopus also express well-known adult stem cell markers LGR5 and Msi1 as in adult mouse intestinal stem cells. Subsequent differentiation of the descendants of these adult stem cells in both mouse and Xenopus leads to the formation of adult epithelial cells to replace the suckling-type or larval-type epithelial cells, accompanied by the establishment of a self-renewing epithelial system with stem cells located in the bottom of the crypt (mouse)/epithelial fold (Xenopus). Note that in both mouse and Xenopus intestine, T3 also regulates known stem cell pathways, such as the WNT pathway. Modified after (108).

Conclusion

Since the discovery of a substance(s) in the thyroid could accelerate frog metamorphosis over a century ago (18), amphibian metamorphosis has been used as a model to study postembryonic development in vertebrates, particularly the role of T3 (12, 17, 127). While there are other models for studying the function of T3 during metamorphosis/postembryonic development (13, 115, 116, 128, 129), anuran metamorphosis is uniquely suitable for dissecting the mechanisms by which T3 regulates vertebrate development, due to easy manipulation and the lack of maternal influence. The essential role of T3 for metamorphosis and the molecular properties of TR have led to a dual-function model that can be easily tested via molecular and genetic studies in vivo. Indeed, earlier transgenic and molecular analyses in Xenopus laevis have provided strong evidence to support the dual functions of TR in development. The more recent gene knockout studies have not only provided evidence to support such dual functions for endogenous TRs but also revealed a few interesting details and surprises. First, knocking out either TRα alone or both TRα and TRβ accelerate premetamorphic tadpole growth in a process independent of the developmental progression toward metamorphosis. Second, TRs are not required for the development of adult organs/tissues during metamorphosis. Third, TRs are essential for resorption of larval organs/tissues during metamorphosis. Thus, unlike the prevailing hypothesis that gene activation by liganded TR is essential to ensure complete metamorphosis of different tissues/organs, the knockout studies suggest that the repression by unliganded TR in premetamorphic tadpoles mainly function to prevent premature adult organ development, while liganded TR functions to ensure proper larval tissue degeneration, which is critical animal survival and completion of metamorphosis.

There are many similarities between anuran metamorphosis and mammalian postembryonic development, including the presence of peak levels of T3 in the plasma and TR expression prior to the maturation of the thyroid gland (13, 130). On the other hand, mice lacking both TR genes can develop into adulthood, unlike the lethal phenotype for TR double-knockout tadpoles at the metamorphic climax. Given the fact that anuran metamorphosis involves resorption of a large fraction of the tadpole, including the gills and tail, and the fact that TR is required for resorption of larval tissues but not for adult organ development, it is tempting to speculate that the reason for the different survival outcomes of TR double-knockout mice and tadpoles is that coordinated removal of larval tissues/organs, which occurs during anuran metamorphosis but not postembryonic mouse development, is essential for animal survival. In this regard, it will be of interest to analyze the effect of complete TR knockout in other animal species, such as flatfish (115), that involve extensive larval tissue removal during postembryonic development.

Glossary

Abbreviations

ChIP

chromatin immunoprecipitation

RXR

retinoid X receptor

T3

thyroid hormone (triiodothyronine)

TR

T3 receptor

TRE

T3 response elements

Additional Information

Disclosures: The author has nothing to disclose.

NIH Statement: This is an un-copyedited author manuscript copyrighted by The Endocrine Society. This may not be duplicated or reproduced, other than for personal use or within the rule of “Fair Use of Copyrighted Materials” (section 107, Title 17, U.S. Code) without permission of the copyright owner, The Endocrine Society. From the time of acceptance following peer review, the full text of this manuscript is made freely available by The Endocrine Society at http://www.endojournals.org/. The final copyedited article can be found at http://www.endojournals.org/. The Endocrine Society disclaims any responsibility or liability for errors or omissions in this version of the manuscript or in any version derived from it by the National Institutes of Health or other parties. The citation of this article must include the following information: author(s), article title, journal title, year of publication and DOI.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.