Molecular Mechanism of Frog Metamorphosis

Yun-Bo Shi, PhD, Head, Section on Molecular Morphogenesis
Liezhen Fu, PhD, Staff Scientist
Biswajit Das, PhD, Research Fellow
Maria Rosaria Fiorentino, PhD, Visiting Fellow
Kenta Fujimoto, PhD, Visiting Fellow
Rachel Heimeier, PhD, Visiting Fellow
Smita Mathew, PhD, Visiting Fellow
Hiroki Matsuda, PhD, Visiting Fellow
Xuedong Wang, PhD, Visiting Fellow
Alexis Oetting, BS, Predoctoral Fellow

We investigate the molecular mechanisms of amphibian metamorphosis. The control of this developmental process by thyroid hormone (TH) offers a unique paradigm in which to study gene function in post-embryonic organ development. During metamorphosis, distinct organs undergo vastly different changes. Some, such as the tail, undergo complete resorption while others, such as the limb, develop de novo. In the frog, most larval organs persist through metamorphosis but are dramatically remodeled. For example, tadpole intestine in Xenopus laevis is a simple tubular structure consisting primarily of a single layer of larval epithelial cells. During metamorphosis, it is transformed, through specific cell death and selective cell proliferation and differentiation, into an organ with a multiply folded adult epithelium surrounded by elaborate connective tissue and muscles. The wealth of knowledge from past research and the ability to manipulate amphibian metamorphosis both in vivo, by using transgenesis or hormone treatment of whole animals, and in vitro, by using organ cultures, offer excellent opportunities to study in vivo the developmental function of TH receptors (TRs) and their underlying mechanisms and to identify and characterize the function of genes critical for post-embryonic organ development in vertebrates.

Mechanism and function of TR during development

Buchholz,¹ Das, Heimeier, Matsuda, Oetting, Sato,² Wang, Washington,³ Shi

We have proposed a dual function model for TR during frog development, that is, the heterodimers between TRs and RXRs (9-cis retinoic acid receptors) bind to target genes constitutively in vivo. In premetamorphic tadpoles, the heterodimers repress gene expression in the absence of TH to prevent metamorphosis, thus ensuring a proper tadpole growth period. When TH is present either from endogenous synthesis during development or is added exogenously to the rearing water of premetamorphic tadpoles, TR/RXR heterodimers activate TH-inducible genes to initiate metamorphosis. By using a transgenic approach, we previously showed that TR is both necessary and sufficient for the metamorphic effects of TH, whereas the non-genomic action of TH, while it exists, plays a minor role, if any, during this post-embryonic process. Thus, metamorphosis provides the first example of TR mediating directly and sufficiently the developmental effects of TH in individual organs by regulating target gene expression in these organs.

More recently, we designed dnN-CoR, a dominant negative form of the TR-binding co-repressor N-CoR (nuclear receptor co-repressor); dnN-CoR contains only the receptor-interacting domain of N-CoR. We overexpressed dnN-CoR under the control of a heat shock–inducible promoter in tadpoles through transgenesis and observed significant derepression of TH-response genes in transgenic animals. More important, transgenic tadpoles developed faster than wild-type siblings, with an acceleration of as much as 7 days in the 30-day experiment. By the end of the experiment, the animals reached early stages of metamorphosis. The data suggest that unliganded TR recruits co-repressors to control metamorphic timing, as predicted by our model.

In addition, the complexity of metamorphic changes in different organs argues for the presence of distinct gene regulation programs regulated by TR. Knowledge of systematic gene regulation will help identify not only molecular markers but also important cellular pathways or critical genes for future mechanistic studies. Thus, we have begun to use the recently developed Xenopus laevis cDNA array to analyze genome-wide gene expression changes associated with TH-induced intestinal remodeling. Our initial analysis of animals treated with TH for varying numbers of days has provided a molecular description of the gene-regulation pathways associated with various metamorphic processes in the intestine. However, the analysis cannot determine whether TRs directly or indirectly regulate the TH-response genes thus identified. Given that directly regulated genes act earlier in the metamorphic process and are more likely to be important regulators of metamorphosis, we are currently employing cDNA array analysis to identify such genes by using RNA from animals treated with TH in the presence or absence of protein synthesis inhibitors. By determining tissue-specific direct target genes of TR, we should be able to gain insight into how TH controls tissue-specific changes during metamorphosis.

While the above studies using Xenopus laevis as a model have led to several important in vivo findings about the function and mechanisms of TR action during development, the lack of genomic sequence information and the frog’s tetraploid genome and lengthy developmental cycle make it difficult for further analyses of TR functions in this species. However, the closely related species Xenopus tropicalis offers many advantages. Toward exploiting X. tropicalis for genome-wide and genetic studies of TR function, we analyzed the expression profiles of TRs and RXRs during X. tropicalis development. We showed that their expression correlates with organ transformations and that TR/RXR heterodimers are capable of repressing and activating gene expression in vivo in the absence and presence of TH, respectively. Furthermore, TRs are bound to endogenous target genes in X. tropicalis tadpoles. Our results support a role for TRs in mediating the metamorphic effects of TH in X. tropicalis. More important, the similarities in the expression and function between X. tropicalis and X. laevis TRs and RXRs, as demonstrated by our study, pave the way for taking advantage of existing morphological, molecular, and cellular knowledge of X. laevis development and the greater genetic and sequence suitability of X. tropicalis to dissect the molecular pathways governing tissue- and organ-specific transformations during vertebrate post-embryonic development.

Buchholz DR, Heimeier RA, Das B, Washington T, Shi Y-B. Pairing morphologywith gene expression in thyroid hormone-induced intestinal remodeling andidentification of a core set of TH -induced genes across tadpole tissues. DevBiol 2007;303:576-590.
Buchholz DR, Paul BD, Fu L, Tomita A, Shi Y-B. Molecular and developmental analyses of thyroid hormone receptor function in Xenopus laevis, the African clawed frog. Gen Comp Endocrinol 2006;145:1-19.
Sato Y, Buchholz DR, Paul BD, Shi Y-B. A role of unliganded thyroid hormone receptor in premetamorphic development in Xenopus laevis. Mech Dev 2007;124:476-488.
Stewart D, Tomita A, Shi Y-B, Wong J. Chromatin immunoprecipitation for studying transcriptional regulation in Xenopus oocytes and tadpoles. Methods Mol Biol 2006;322:165-81.
Wang X, Matsuda H, Shi Y-B. Developmental regulation and function of thyroid hormone receptors and 9-cis retinoic acid receptors during Xenopus tropicalis metamorphosis. Endocrinology 2008;149:5610-5618.

Roles of co-factors in gene regulation by TR

Choi,⁴ Fu, Heimeier, Hsia,⁵ Matsuda, Paul,⁶ Shi

TR regulates gene transcription by recruiting co-factors to target genes. In the presence of TH, TR can bind to co-activators while unliganded TR binds to co-repressors. Many biochemical and molecular studies have been conducted on such co-factors. Much less is known, however, about whether and how they participate in gene regulation by TR in various biological processes in vivo. We investigate how TR uses different co-factors in the context of development in various organs.

Among co-repressors, we have studied the role of N-CoR and SMRT (silencing mediator of retinoid and thyroid receptors) in gene repression by TR in premetamorphic tadpoles. We previously showed that both are expressed and, more important, bound to TH-response genes in such tadpoles. Furthermore, the repressors are released from the target genes upon TH treatment of premetamorphic tadpoles or during natural metamorphosis. Our studies with the dnN-CoR, as summarized above, have demonstrated in vivo a role of co-repressor complexes in gene repression by unliganded TR during premetamorphic tadpole development. On the activator side, we used the chromatin immunoprecipitation (ChIP) assay to show that Xenopus co-activators SRC3 (steroid receptor co-activator 3) and p300 are both recruited to TH-responsive promoters. More important, by using transgenesis to overexpress dominant negative forms of SRC3 and p300, we showed that co-activators, particularly the SRC3-p300 family of co-activator complexes, are required for gene regulation by liganded TR and for TH-dependent metamorphosis. These findings represent the first known example of specific co-activator complexes playing critical roles in the developmental function of a nuclear receptor in vertebrates.

The SRC/p300 complexes also contain the methyltransferases CARM1 and PRMT1, which have both been implicated in TR function in mammalian cell culture studies. Thus, to investigate further the role and mechanisms of the SRC/p300 complexes in development, we cloned and characterized Xenopus CARM1 and PRMT1. We obtained two alternative splicing forms of Carm1, CARM1a and CARM1b. Both isoforms are expressed throughout metamorphosis, suggesting that they play a role during the process. Surprisingly, transcriptional analysis in the Xenopus oocyte revealed that overexpression of CARM1b had little effect on TR-mediated transcription, although CARM1a expectedly enhanced gene activation by liganded TR. ChIP assays showed that liganded TR recruited both endogenous CARM1a and overexpressed CARM1a and CARM1b to a target promoter. However, the binding of liganded TR to the target promoter was lower when CARM1b was overexpressed and was accompanied by a slight reduction in histone methylation at the promoter. The results suggest that CARM1 plays a role in TR-mediated transcriptional regulation during frog development and that alternative splicing regulates the enzyme’s function. Likewise, our studies in the frog oocyte model system have shown that liganded TR recruits PRMT1 to enhance transcriptional activation by TH. We are currently investigating the in vivo function of these methyltransferases during metamorphosis.

In addition to histone modifications involving complexes such as SRC-p300, transcriptional regulation by TR involves chromatin remodeling. In particular, the BRG1-containing, ATP-dependent chromatin remodeling complexes have been implicated in gene regulation by nuclear receptors. To investigate their potential involvement in metamorphosis, we showed that the expression of BRG1, a chromatin-remodeling enzyme, is upregulated at the climax of Xenopus laevis metamorphosis while BAF57, a BRG1-binding protein in BRG1-containing chromatin remodeling complexes, is constitutively expressed during development. Consistent with this finding, TH treatment of premetamorphic tadpoles led to upregulation of the expression of BRG1 but not of BAF57. Studies using a reconstituted TH-dependent Xenopus oocyte transcription system, in which we can study TR function in the context of chromatin, revealed that BRG1 enhances the transcriptional activation by ligand-bound TRs in a dose-dependent manner while a remodeling-defective BRG1 mutant inhibited the activation, suggesting that the process relies on chromatin remodeling. Further studies showed that BAF57 interacted with BRG1 in oocytes and cooperatively enhanced gene activation by TR with BRG1 in vivo. Chromatin immunoprecipitation revealed that BAF57 was recruited to the TR-regulated promoter in the presence of TR and TH. Together, the findings suggest a role for BRG1-/BAF57-containing chromatin remodeling complexes in TR-regulated gene expression during metamorphosis.

Heimeier RA, Hsia VS-C, Shi Y-B. Participation of BAF57 and BRG1-containing chromatin remodeling complexes in thyroid hormone-dependent gene activation during vertebrate development. Mol Endocrinol 2008;22:1065-1077.
Matsuda H, Paul BD, Choi CY, Shi Y-B. Contrasting effects of two alternative splicing forms of coactivatorassociated arginine methyltransferase 1 on thyroid hormone receptor-mediated transcription in Xenopus laevis. Mol Endocrinol 2007;21:1082-1094.
Paul BD, Buchholz DR, Fu L, Shi Y-B. SRC-p300 coactivator complex is required for thyroid hormone induced amphibian metamorphosis. J Biol Chem 2007;282:7472-7481.

Regulation and function of the matrix metalloproteinases during TH-induced tissue remodeling

Fiorentino, Fu, Hasebe,⁷ Mathew, Shi; in collaboration with Ishizuya-Oka

We previously identified several TH-response genes encoding matrix metalloproteinases (MMPs) during intestinal metamorphosis. MMPs are Zn²⁺-dependent proteases capable of cleaving various proteins of the ECM (extracellular matrix). The ECM not only provides essential physical support within an organism but also influences cell fate during developmental and pathological processes. Upregulation of MMP genes has been observed in many developmental processes such as limb morphogenesis and in diverse pathological conditions such as arthritis, wound healing, and cancer metastasis. However, it has been difficult to investigate the roles of MMPs in mammals due to the lack of good models and the relatively subtle phenotypes in limited tissues in knockout mice. The activation of MMP genes by TH together with the observed ECM remodeling during metamorphosis suggests that MMPs participate in this process through ECM remodeling. Frog metamorphosis affects essentially all organs in a tadpole within a short developmental period and thus offers an excellent opportunity to investigate the in vivo functions of MMPs and associated mechanisms during development.

Expression and function studies in organ cultures have shown that the MMP stromelysin-3 (ST3) is directly regulated by TR at the transcription level and is required for ECM remodeling and TH-induced cell death in the animal intestine. Furthermore, in premetamorphic tadpoles, transgenic overexpression of ST3, but not of a catalytically inactive mutant, is sufficient to cause ECM remodeling and apoptosis in the intestinal epithelium, demonstrating that ST3 is necessary and sufficient for larval epithelial cell death. Toward understanding the mechanism by which ST3 affects tissue remodeling, we have also identified the 37 kDa laminin receptor (LR), a cell surface receptor for the ECM protein laminin, as an in vivo substrate of ST3. LR binds to ST3 and can be cleaved by ST3 at two sites in the extracellular domain between the transmembrane domain and laminin binding sequence, separating the cells from laminin. Interestingly, ST3 cleavage sites in LR are conserved in human LR, and high levels of LR are known to be present in tumor cells, which are often surrounded by fibroblasts expressing ST3. Thus, LR may be a conserved substrate of ST3, and its cleavage by ST3 may alter cell-ECM interactions, thereby mediating the effects of ST3 on cell fate and behavior during development and pathogenesis. To investigate the functional importance of LR cleavage by ST3 during development, we carried out a series of mutational analyses on the two cleavage sites in LR. Our findings revealed that, in addition to primary sequence at the cleavage site (positions P3 through P3′, with the cleavage occurring between P1 and P1′), flanking sequences/conformation influenced the cleavage of LR by ST3. Moreover, alanine substitution studies led to the surprising finding that the surrounding sequence and/or conformation dictated the site of cleavage in LR by ST3. The results have important implications for our understanding of substrate recognition and cleavage by ST3 and underscore the importance of studying ST3 cleavage in the context of full-length substrates. More important, we plan to use the LR cleavage mutants in transgenic studies to investigate the role of LR cleavage by ST3 during metamorphosis.

In addition to ST3, several other MMPs are regulated by TH during metamorphosis, consistent with the complex nature of the ECM, which would require several enzymes for its remodeling and degradation during metamorphosis. Our earlier studies showed that the MMPs GelA and MT1-MMP (a membrane-type MMP) are coordinately regulated during metamorphosis. Using Xenopus embryogenesis as a model, we showed that Xenopus GelA and MT 1-MMP are co-expressed during embryogenesis and associated with each other in vivo. Furthermore, we showed through overexpression of both MMPs in developing embryos that GelA and MT1-MMP play cooperative roles during development, at least in part through the activation of pro-GelA by MT1-MMP. We are now analyzing the role of these MMPs during metamorphosis by using transgenic overexpression in a manner similar to our ST3 overexpression experiments.

Fu L, Hasebe T, Ishizuya-Oka A, Shi Y-B. Roles of matrix metalloproteinases and ECM remodeling during thyroid hormone-dependent intestinal metamorphosis in Xenopus laevis. Organogenesis 2007;3:14-19.
Fu L, Tomita A, Wang H, Buchholz DR, Shi Y-B. Transcriptional regulation of the Xenopus laevis stromelysin-3 gene by thyroid hormone is mediated by a DNA element in the first intron. J Biol Chem 2006;281:16870-16878.
Hasebe T, Hartman R, Fu L, Amano T, Shi Y-B. Evidence for a cooperative role of gelatinase A and membrane type-1 matrix metalloproteinase during Xenopus laevis development. Mech Dev 2007;124:11-22.
Ishizuya-Oka A, Shi Y-B. Regulation of adult intestinal epithelial stem cell development by thyroid hormone during Xenopus laevis metamorphosis. Dev Dyn 2007;236:3358-3368.
Shi Y-B, Fu L, Hasebe T, Ishizuya-Oka A. Regulation of ECM remodeling and cell fate determination by matrix metalloproteinase stromelysin-3 during thyroid hormone-dependent postembryonic development. Pharmacol Ther 2007;116:391-400.

Thyroid hormone regulation of adult intestinal epithelial stem cell development during metamorphosis

Hasebe,⁷ Shi; in collaboration with Ishizuya-Oka

As the larval epithelial cells undergo apoptosis during intestinal metamorphosis, the adult epithelium develops. Currently, the origin of the adult epithelial stem cells is unknown, although some evidence supports the notion that they derive from dedifferentiated larval epithelial cells. We previously isolated sonic hedgehog (Shh), a cell-cell signaling molecule, as a direct TR target gene during intestinal remodeling. Furthermore, Shh is highly expressed in proliferating adult epithelial cells during metamorphosis. These results suggest that Shh is involved in the larval-to-adult remodeling of the Xenopus laevis intestine and is likely to play a role in the development and/or proliferation of adult epithelial stem cells. Indeed, organ culture studies have allowed us to show that exogenous Shh enhances epithelial cell proliferation. The Shh signal is known to be regulated at post-translational levels in other animal species. In particular, the pan-hedgehog inhibitor hedgehog-interacting protein (Hip) has been implicated in the regulation of Shh signaling. Using real-time RT-PCR and in situ hybridization, we showed that Hip expression is transiently upregulated during both natural and TH-induced metamorphosis and that Hip mRNA is localized in the connective tissue adjacent to the adult epithelial primordia expressing Shh. Interestingly, the expression of bone morphogenetic protein-4 (BMP-4), an Shh target gene, is remarkably low where Hip is strongly expressed, suggesting that Hip regulates Shh signaling. Finally, we demonstrated that Hip binds to the N-terminal fragment of processed Shh in vivo, indicating that Hip may suppress Shh signaling by sequestering Shh and thus might play a role in regulating the spatial and temporal signaling of Shh during intestinal stem cell development and proliferation.

Hasebe T, Kajita M, Shi Y-B, Ishizuya-Oka A. Thyroid hormone-up-regulated hedgehog interacting protein is involved in larval-to-adult intestinal remodeling by regulating sonic hedgehog signaling pathway in Xenopus laevis. Dev Dyn 2008;237:3006-3015.
Ishizuya-Oka A, Shi Y-B. Thyroid-hormone regulation of stem cell development during intestinal remodeling. Mol Cell Endocrinol 2008;288:71-78.

¹Daniel Buchholz, PhD, former Postdoctoral Fellow
²Yukiyasu Sato, MD, PhD, former Visiting Fellow
³Theresa Washington, PhD, former Postdoctoral Fellow
⁴Choel Yong Choi, PhD, former Visiting Fellow
⁵ShaoChung Hsia, PhD, former Visiting Fellow
⁶Bindu Diana Paul, PhD, former Visiting Fellow
⁷Takashi Hasebe, PhD, former Visiting Fellow

Collaborator

Atsuko Ishizuya-Oka, PhD, Nippon Medical School, Tokyo, Japan

For further information, contact shi@helix.nih.gov.