This laboratory is exploring molecular mechanisms underlying amphibian metamorphosis. The control of this developmental process by thyroid hormone (TH) offers a unique paradigm in which to study gene function in postembryonic organ development. During metamorphosis, various organs undergo vastly different changes. Some, like the tail, undergo complete resorption, while others, such as the limb, are developed de novo. Most of the larval organs persist through metamorphosis but are dramatically remodeled to function in a frog. For example, the tadpole intestine in Xenopus laevis is a simple tubular structure consisting mainly of a single layer of primary epithelial cells. During metamorphosis, a process of specific cell death and selective cell proliferation and differentiation transforms the intestine into a multiply folded adult epithelium with 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 in organ cultures, offer excellent opportunities both to study the developmental function of thyroid hormone receptors (TRs) and the underlying mechanisms in vivo and to identify and functionally characterize genes that are critical for postembryonic organ development in vertebrates.

Function of TR during Development
Buchholz, Hsia, Sachs,a Shi
We have proposed a dual function model of TR action based on our earlier studies in the oocyte, developing embryos, and tadpoles. The model postulates that heterodimers between TR and RXR (9-cis retinoic acid receptor) activate gene expression during metamorphosis when TH is present while, in premetamorphic tadpoles, they repress gene expression to prevent metamorpho-sis, thus ensuring a tadpole growth period. Our studies have since provided strong support and mechanistic insights for such a model. Most recently, we have shown that the corepressors N-CoR and SMRT are recruited to TH-response genes in premetamorphic tadpoles and are released upon treatment of the tadpoles with TH, indicating that unliganded TR recruits these corepressors to repress target genes in tadpoles. In agreement with the observation that these corepressors can form complexes containing histone deacetylases (HDACs), TH treatment leads to a rise in local histone acetylation at the TH response genes, at least in the intestine and tail, arguing that histone acetylation is an important factor in gene regulation by TR.

To investigate the function of TR in vivo, we have adapted a sperm-mediated transgenic method to generate transgenic animals expressing a dominant-negative TR. Phenotypic analyses indicate that over-expression of the dominant-negative TR inhibits TH-induced metamorphosis. More important, we have shown that the dominant-negative TR specifically blocks the expression of TH response genes that we and others have already identified. These results provide molecular evidence for the central role of TR in regulating the response genes that mediate the effects of TH on metamorphosis.

Roles of Cofactors in Gene Regulation by TR
Amano, Jones,b Obata, Paul, Tomita, Wadec, Shi
TR regulates gene transcription by recruiting cofactors to target genes. In view of the several biochemical and molecular studies already completed on such cofactors, we are investigating how TR utilizes different cofactors in the context of development in various organs. We have recently characterized one coactivator, the Xenopus TRIP7. However, our studies using the frog oocyte model system suggest that Xenopus TRIP7 has a relatively small effect on TR function, and thus its role in developmental gene regulation by TR may be limited. Accordingly, we have begun to study p300 and SRC, evidence for whose involvement in TR function has been obtained from in vitro and tissue culture cell studies. We have obtained the cDNA clones for Xenopus p300, SRC1, and SRC2 and have shown that they are expressed during metamorphosis.

With regard to the corepressors mentioned above, we have shown that the corepressors NCoR and SMRT are recruited to TH response genes in developing tadpoles and that histone acetylation is involved in gene regulation by TR in tadpoles. Both NCoR and SMRT are known to interact with other proteins. To investigate how these corepressors participate in the repression by unliganded TR, we have isolated three NCoR-corepressor complexes. However, due to low abundance, we have yet to determine the identities of the components in the complexes. On the other hand, in HeLa cells, the TBL1 (transducingblike 1) protein has been shown to exist in an N-CoR complex that was identified by others. We have cloned the frog homolog and intend to investigate whether TBL1 is present in our NCoR complexes and whether it participates in NCoR-dependent gene repression by TR.

We intend to use chromatin immunoprecipitation assays to determine the recruitment of both corepressors and coactivators by TR to target genes in tadpoles. To investigate directly the function of these factors in vivo, we will precociously over-express wild-type or dominant-negative cofactors and analyze the resulting effects on animal development and target gene expression (and local changes in chromatin structure, when appropriate). Before undertaking transgenic studies, we will characterize the cofactors/mutants for their ability to influence TR function in the frog oocyte system that we established earlier for studying gene function in the context of chromatin. This approach should help in the rational design of our transgenic studies on the molecular mechanisms responsible for the effects of the cofactors.

Involvement of Matrix Metalloproteinases during TH-Induced Tissue Remodeling
Amano, Damjanovski,d Fu, Wei, Shi; in collaboration with Ishizuya-Oka
In an effort to identify genes important for post-embryonic development, we have isolated many TH response genes during metamorphosis. Expression analyses and other studies have led us to focus on the TH-response genes encoding matrix metalloproteinases (MMPs) for func-tional investigations. MMPs are extracellular enzymes capable of digesting various extra-cellular matrix (ECM) components. Our earlier studies led us to propose that the MMP stomelysin-3 (ST3) is directly or indirectly involved in ECM remodeling, which in turn influences cell behavior. By using a function-blocking antibody against the catalytic domain of ST3, we have demonstrated in organ cultures that blocking ST3 function inhibits TH-induced apoptosis of larval intestinal epithelial cells as well as the invasion of the proliferating adult epithelial cells into the connective tissue. These effects are accompanied by an inhibition of the remodeling of the basal lamina or basement membrane, the ECM that separates the connective tissue from the epithelium. The results support the argument that ST3 is directly or indirectly involved in ECM remodeling, which in turn influences cell behavior. To investigate directly the roles of MMPs in developing animals, we are employing the transgenic approach to express wild-type and mutant MMPs in Xenopus embryos and tadpoles. In our initial study, we over-expressed Xenopus MMPs stromelysin-3 (ST3) and collagenases-4 (Col4) under the control of a ubiquitous promoter and observed that embryos with over-expressed ST3 or Col4, but not the control green fluorescent protein (GFP) or a mutant ST3, died in a dose-dependent manner during late embryogenesis. This lethality in early development prevented us from investigating the roles of MMPs during metamorphosis. Thus, we have developed a double-promoter approach described below to over-express MMPs under an inducible promoter. In our preliminary study, we generated transgenic tadpoles by using a double-promoter construct in which ST-3 is under the control of the heat shock–inducible promoter. Heat shock at tadpole stages led to over-expression of stromelysin-3 in all organs, although no visible morphological changes of the tadpoles were observed for four days. Analysis of the intestine showed that over-expression of stromelysin-3 caused premature apoptosis in the tadpole epithelium, consistent with our earlier organ culture studies. We are currently investigating how the transgenic animals undergo natural and TH-induced metamorphosis.

Development of a Double-Promoter Trans-genic Approach
Buchholz, Fu, Shi
In our first set of transgenic experiments, we over-expressed three different MMPs under the control of the ubiquitous CMV promoter. We found that all the animals expressing the wild-type MMPs, but not a catalytically inactive stromelysin-3 or GFP, died at various stages of development, none developing into metamorphosing tadpoles. These initial studies identified two problems with the existing approach. The first is embryonic lethality caused by MMP expression under a ubiquitous promoter, and the second is the difficulty in identifying transgenic animals. Normally, transgenic animals or plants are identified by PCR typing, which is tedious and slow. Alternative approaches include the use of GFP-fusion proteins and cotransgenesis involving two plasmids, one expressing the desired gene and the other a marker gene. The first approach is limited by the possibility that GFP may not be visible if the transgene is not expressed on the surface of the animal or if the fusion protein has reduced fluorescence or is expressed only at certain developmental stages. The second is also of limited use because of the relatively low efficiency of chromosomal integration of both plasmids in a single transgenic animal.

To overcome these problems, we have developed a double-promoter approach. In the construct for transgenesis, we use a lens-specific promoter (g-crystalline promoter) to drive the expression of GFP and a second, heat shock–inducible promoter (or another promoter of interest) to drive the expression of the desired gene. This approach allows us to generate transgenic animals that can be easily identified by checking the eyes under UV and that can be maintained and treated in the same containers as nontransgenic controls. We have validated the approach by using PCR typing of transgenic animals and demonstrating that both promoters function independently and properly as designed.