We have been investigating the molecular mechanism of the induction and patterning of the mesoderm and neuroectoderm in the early frog embryo. Recent evidence has shown that peptide growth factors are essential during the formation of the mesoderm in Xenopus. We are now interested in determining how these secreted molecules act in a concerted way, in space and time, to produce a fully organized embryo. It is known that the mesoderm is induced and crudely patterned at the blastula stages, but is not determined or organized until the gastrula. Therefore we have recently been focusing on the gastrula stages in order to decipher the events that pattern the mesoderm. The frog embryo is very well suited for this analysis. Tissues can be dissected, isolated or transplanted with high precision and ease in the frog embryo. In addition signalling molecules and their receptors can be missexpressed or inhibited using conventional molecular approaches. Also the organized and predictable morphogenetic movements that occur during the gastrula and neurula stages can be followed using time-lapse video microscopy, allowing one to study the signalling events that pattern the embryo as they occur. Finally, by using an approach that we have developed to generate transgenic frog embryos (see section 2), one can control the temporal and spatial expression of developmental genes in the embryo much better than ever before.

Using a combinations of these tools, we have been studying the role of growth factor signalling during early embryonic development in the frog. For example, by injecting RNA's encoding wild-type and mutant growth factors or wild-type and mutant growth factors receptors (i.e. dominant negative), we can produce cells that either secrete growth factors or are unable to respond to growth factors (Amaya, et al. 1991 Cell 66: 257-270; Amaya, et al. 1993 Development 118:477-487; Nutt et al. 2001, Genes Dev. 15:1152-1166) (see Mesoderm Formation). This type of experiments will allow us to determine where and how signals are being transmitted during the patterning of the mesoderm at the gastrula stages.

These experiments, though, cannot address when FGF signalling is needed during these events. Therefore we have also generated transgenic frog embryos that express the dominant negative FGF receptor behind temporal and spatial specific promoters (Kroll and Amaya 1996, Development 122:3173-3183). This approach allows one to generate embryos with intact FGF signalling during the blastula stages, when mesoderm is induced, but with inhibited FGF signalling at the gastrula stages or later. Using this approach we have shown that FGF signalling is required for mesoderm to be maintained. Thus embryos with intact FGF signalling during mesoderm induction but compromised FGF signalling during the gastrula stages do not form muscle or notochord.

The elucidation of the molecular basis of pattern formation and differentiation in frog embryos could be greatly facilitated if wild type and mutant forms of developmental genes could be expressed in a temporal and/or tissue specific manner. Traditionally this experimental approach can be achieved by driving the gene of interest behind temporal and/or tissue specific promoters. In frog embryos, though, this approach has only been marginally successful for two reasons. 1. When DNA is injected into developing embryos, the DNA does not integrate into the frog chromosomes and therefore, the embryo expresses the genes in a highly mosaic pattern. 2. Temporal and/or tissue specific promoters are not expressed with adequate fidelity when the DNA is not integrated into the genome of the frog. For these reasons, we have developed a procedure for the efficient integration of DNA into the chromosomes of early frog embryos (Kroll and Amaya 1996, Development 122:3173-3183; Amaya and Kroll 1999, Meth. Mol. Bio. 97:393-414). Briefly, the approach involves isolating sperm nuclei and decondensing their chromosomes in egg extracts. Then DNA is integrated into the isolated sperm nuclei using restriction enzyme mediated integration (REMI). The sperm nuclei are then transplanted into eggs, leading to embryos that inherit the integrated DNA in all of its cells. Using this approach one can now generate hundreds of transgenic embryos in an afternoon. Since the integration occurs before fertilization, one does not need to go to the next generation to make non-mosaic, fully transgenic embryos (see Muscle Actin Transgenic Embryo, N-Tubulin Transgenic Tadpoles, g-Crystallin Frog). This approach allows us to investigate the role of growth factor signalling during organognesis (Breckenridge et al. 2001, Dev. Biol. 232:191-203; Hartley et al. 2001, Dev. Biol. 238:168-184). For this studies we have recently begun to use the binary Gal4-UAS misexpression system (Hartley et al. 2002, PNAS 99:1377-1382). We have also used transgenesis to study the transcriptional regulation of XMyf-5, a gene expressed soon after mesoderm induction (Polli and Amaya, 2002, Development 129:2917-2927). In addition this technology can be used as an insertional mutagen in combination with a gene trap approach (Bronchain et al. 1999, Current Biology 9:1195-1198). As an extension of this work, we are currently investigating the efficacy of doing a large scale insertional mutagenesis screen in Xenopus tropicalis (see below).

Amphibian embryos have classically been one of the best systems for elucidating the mechanisms of early development. While their advantages are numerous, one large disadvantage is that it has not been exploited at the genetic level. The reasons for this have been: 1. the generation time is too long; 2. the husbandry is cumbersome for maintaining a sufficiently large colony; 3. and in the case of X. laevis, genetic studies would be confounded since it is pseudotetraploid. Therefore we have been interested in incorporating X. tropicalis, a diploid species, as a model system for vertebrate embryological studies (Amaya et al. 1998, TIGS 14:253-255; Nutt et al. 2001, genesis 30:97-100; Khokha et al. 2002, Dev. Dynamics 225:499-510; D'Souza et al. 2003, Dev. Dynamics 226:118-127). It appears to provide several advantages as a model system. For the past several years we have doing preliminary experiments on this species. We have found that it is easy to maintain and breed in the laboratory in high numbers and its generation time is between 4 to 6 months. Furthermore it develops similarly to X. laevis, albeit at a faster rate, and its development appears to be more consistent from batch to batch than X. laevis. The embryos are smaller than X. laevis embryos but large enough for manipulation. As described above, X. tropicalis would be the species of choice for insertional mutagenesis and/or enhancer trap screens. For additional information visit the X. tropicalis website.