The western clawed frog Xenopus tropicalis is an important model for vertebrate development that combines experimental advantages of the African clawed frog Xenopus laevis with more tractable genetics. Here we present a draft genome sequence assembly of X. tropicalis. This genome encodes over 20,000 protein-coding genes, including orthologs of at least 1,700 human disease genes. Over a million expressed sequence tags validated the annotation. More than one-third of the genome consists of transposable elements, with unusually prevalent DNA transposons. Like other tetrapods, the genome contains gene deserts enriched for conserved non-coding elements. The genome exhibits remarkable shared synteny with human and chicken over major parts of large chromosomes, broken by lineage-specific chromosome fusions and fissions, mainly in the mammalian lineage.

Gastrulation movements are critical for establishing the three principal germ layers and the basic architecture of vertebrate embryos. Although the individual molecules and pathways involved are not clearly understood, non-canonical Wnt signals are known to participate in developmental processes, including planar cell polarity and directed cell rearrangements. Here we demonstrate that the dual-specificity transcriptional repressor Kaiso, first identified in association with p120-catenin, is required for Xenopus gastrulation movements. In addition, depletion of xKaiso results in increased expression of the non-canonical xWnt11, which contributes to the xKaiso knockdown phenotype as it is significantly rescued by dominant-negative Wnt11. We further demonstrate that xWnt11 is a direct gene target of xKaiso and that p120-catenin association relieves xKaiso repression in vivo. Our results indicate that p120-catenin and Kaiso are essential components of a new developmental gene regulatory pathway that controls vertebrate morphogenesis.

There is great interest in the patterning and morphogenesis of the vertebrate nervous system, but the morphogenetic movements involved in early neural development and their underlying cellular mechanisms are poorly understood. This paper describes the cellular basis of the early neural morphogenesis of Xenopus laeuis. The results have important implications for neural induction. Mapping the fate map of the midneurula (Eagleson and Harris: J. Neurobiol. 21:427440, 1990) back to the early gastrula with time-lapse video recording demonstrates that the prospective hindbrain and spinal cord are initiaUy very wide and very short, and thus at the beginning of gastrulation all their precursor cells lie within a few cell diameters of the inducing mesoderm. In the midgastrula, the prospective hindbrain and spinal cord undergo very strong convergence and extension movements in two phases: In the first phase they primarily undergo thinning in the radial direction and lengthening (extension) in the animal-vegetal direction, and the second phase is characterized primarily by mediolateral narrowing (convergence) and anterior-posterior lengthening (extension). These movements also occur in sandwich explants of the gastrula, thus demonstrating the local automomy of the forces producing them. Tracing cell movements with fluorescein dextran-labeled cells in embryos or explants shows that the initial thinning and extension occurs by radial intercalation of deep cells to form fewer layers of greater area, all of which is expressed as increased length. The subsequent convergence and extension occurs by mediolatera1 intercalation of deep cells to form a longer, narrower array. These results establish that a similar if not identical sequence of radial and mediolateral cell intercalations underlie convergence and extension of the neural and the mesoderm tissues (Wilson and Keller: Development, 112:289-300, 1991). Moreover, these results establish that radial and mediolateral intercalation are the principal neural cell behaviors induced by the planar signals emanating from the dorsal involuting marginal zone (the Spemann organizer) in the early gastrula (Keller et al: Develop. Dynamics, 193: 218-234,1992 cord, thus producing the massive convergence and extension movements that narrow and elongate these regions of the nervous system in the late gastrula. A more general significance of these results is that neural induction is best analyzed and understood in terms of the dynamics of the morphogenetic processes involved.

Pilot forward genetic screens in Xenopus tropicalis have isolated over 60 recessive mutations. Here we present a simple method for mapping mutations to chromosomes using gynogenesis and centromeric markers. When coupled with available genomic resources, gross mapping facilitates evaluation of candidate genes as well as higher resolution linkage studies. Using gynogenesis, we have mapped the genetic locations of the 10 X. tropicalis centromeres, and performed fluorescence in situ hybridization to validate these locations cytologically. We demonstrate the use of this very small set of centromeric markers to map mutations efficiently to specific chromosomes.

We present a genetic map for Xenopus tropicalis, consisting of 2886 Simple Sequence Length Polymorphism (SSLP) markers. Using a bioinformatics-based strategy, we identified unique SSLPs within the X. tropicalis genome. Scaffolds from X. tropicalis genome assembly 2.0 (JGI) were scanned for Simple Sequence Repeats (SSRs); unique SSRs were then tested for amplification and polymorphisms using DNA from inbred Nigerian and Ivory Coast individuals. Thus identified, the SSLPs were genotyped against a mapping cross panel of DNA samples from 190 F2 individuals. Nearly 4000 SSLPs were genotyped, yielding a 2886-marker genetic map consisting of 10 major linkage groups between 73 and 132 cM in length, and 4 smaller linkage groups between 7 and 40 cM. The total effective size of the map is 1658 cM, and the average intermarker distance for each linkage group ranged from 0.27 to 0.75 cM. Fluorescence In Situ Hybridization (FISH) was carried out using probes for genes located on mapped scaffolds to assign linkage groups to chromosomes. Comparisons of this map with the X. tropicalis genome Assembly 4.1 (JGI) indicate that the map provides representation of a minimum of 66% of the X. tropicalis genome, incorporating 758 of the approximately 1300 scaffolds over 100,000 bp. The genetic map and SSLP marker database constitute an essential resource for genetic and genomic analyses in X. tropicalis.