Biphasic intestinal development in amphibians: embryogenesis and remodeling during metamorphosis.

Intestinal development in anurans is a biphasic process. The embryogenesis of intestine resembles that in higher vertebrates. The subsequent remodeling process during metamorphosis to produce an adult organ is controlled by TH. Recent progress in studying TH action and its application to amphibian metamorphosis has provided considerable insights into the remodeling process. One possible model for the TH-induced gene regulation cascade of intestinal remodeling is presented in Fig. 9. It is assumed that TRs function as heterodimers with RXRs. In the absence of TH, the TR-RXR heterodimers can bind to TH response elements present in the target genes and repress the basal transcription of these genes (Fondell et al., 1993; Damm et al., 1989; Sap et al., 1989; Baniahmad et al., 1992; Ranjan et al., 1994). The binding of TH leads to conformational changes in the receptor complexes that in turn activate gene transcription. The products of these early response genes then participate in the activation of the remaining gene regulation cascade. Exactly how this occurs remains unknown. Interestingly, the early response genes include not only transcription factors but also other proteins such as metalloproteinases. The transcription factors could activate or repress downstream TH response genes directly. Other proteins are likely to assert their effect indirectly. For example, they could modify the ECM or cell surface. In addition, they could regulate and/or participate in signal transduction by growth factors. The cooperation between these complex intra- and extracellular processes eventually results in the degeneration of the larval organ and formation of the adult tissue. This simplified scheme immediately raises many questions. Although the mRNAs for TRs and RXRs are present in the intestine and the other tissues during metamorphosis (Yaoita and Brown, 1990; Kawahara et al., 1991; Y.-B. Shi, unpublished observations), it is unknown whether the mRNA levels reflect the protein levels. It also remains to be tested whether TR-RXR is indeed the functional complex in vivo and whether the heterodimer is responsible for the activation of the early response genes isolated to date. The majority of the early response genes are ubiquitous. Of the few intestine-specific genes, none of them have yet been identified by sequence analysis. It is of great interest to understand how the same genes expressed in tissues undergoing drastically different changes can exert their biological effects. It is likely that together with existing proteins in the intestine, these early genes regulate tissue-specific downstream genes, which in turn determine the tissue-specific transformation. An important issue is to establish the identity of these downstream genes.