Members of the Cdx family of homeobox containing genes have been identified in a wide range of animal groups. The prototype of the family is the Drosophila caudal gene, which is expressed both maternally and zygotically and is required for the development of posterior structures of the fruit-fly (MacDonald and Struhl, 1986). The vertebrate Cdx genes have prominent domains of expression in the early mesoderm and posterior neural tube (Gamer and Wright, 1993; Subramanian et al., 1995; Marom et al., 1997; Pillemer et al., 1998a). Cdx genes are also expressed in posterior regions of the vertebrate gut and directly regulate the expression of several differentiation products of the mature gut epithelium (Troelsen et al., 1997; Drummond et al., 1998; Park et al., 2000). In addition, Cdx2 is able to direct the differentiation of an undifferentiated intestinal cell line (Suh and Traber, 1996).
The human and mouse genomes contain three caudal-related genes (Cdx1, Cdx2, and Cdx4). The amphibian Xenopus laevis also contains three caudal-related genes, which are designated Xcad1, Xcad2, and Xcad3. Functional studies, in both mouse and Xenopus laevis, indicate that the role of Cdx genes in patterning the anteroposterior axis is conserved in the vertebrate lineage. Overexpression or inhibition of Cad homeodomain transcription factors in Xenopus laevis embryos has demonstrated that these proteins are involved in regulating the expression of Hox genes (Pownall et al., 1996; Epstein et al., 1997; Isaacs et al., 1998). A role for Cdx proteins in Hox gene regulation is also supported by work in other species. Mice homozygous for a null allele of the Cdx1 gene exhibit posterior shifts in the boundaries of Hox gene expression (Subramanian et al., 1995). Conversely, mice carrying transgenes that drive ectopic anterior expression of Cdx genes, show anterior shifts in Hox gene expression (Charité et al., 1998). The activity of the caudal homologue in the primitive chordate Halocynthia roretzi is required for normal Hox gene expression, indicating an ancient origin for the regulatory pathway involving Cdx and Hox genes (Katsuyama et al., 1999). These findings underline the key role that these transcription factors have in the anteroposterior patterning of animals.
Until recently, the amphibian model of choice for developmental studies has been the frog Xenopus laevis, which has many advantages for functional studies. However, Xenopus laevis is not amenable to the powerful genetic analyses that are possible in other animal models such as Drosophila, zebrafish, and mouse. This is due to the long generation time of the animal and the allotetraploidy of its genome. This latter issue means that each gene will have two pseudoallelic variants, which typically differ by 5–10% in DNA sequence identity.
However, recent work using the related species Xenopus tropicalis, which has a much shorter generation time and is diploid, holds out the prospect of a model system that has the advantages of Xenopus laevis for functional studies, but which is also amenable to classical genetics. It has been reported that some Xenopus laevis derived probes can be used to investigate the expression of orthologous genes in Xenopus tropicalis by in situ hybridization (Amaya et al., 1998). However, other techniques commonly used for the quantitative analysis of gene expression such RNAase protection, reverse transcriptase-polymerase chain reaction (RT-PCR), and techniques for inhibiting gene function, such as injection of antisense oligonucleotides will require exact knowledge of target sequences. With this in mind, we report the cloning and sequence of the Cdx family members Xtcad1, Xtcad2, and Xtcad3 from Xenopus tropicalis. We also present a detailed in situ hybridization analysis of the expression patterns of Xtcad1, Xtcad2, and Xtcad3.
RESULTS AND DISCUSSION
Isolation of Xenopus tropicalis Cdx cDNA
Probes derived from the coding sequence of the Xenopus laevis Cad cDNAs were used to probe a Xenopus tropicalis gastrula stage plasmid cDNA library at high stringency. This method identified several clones for each tropicalis Cad cDNA, which we have designated Xtcad1, Xtcad2, and Xtcad3. Single clones of each Xtcad cDNA were completely sequenced. Figure 1A–C shows alignments of the 5′ UTR and coding sequences for each tropicalis and laevis cDNA. Within this region of the cDNAs the identity of Xtcad1 to Xcad1 is 88%, the identity of Xtcad2 to Xcad2 is 94%, and the identity of Xtcad3 to Xcad3 is 93%. Identity of the derived amino acid sequences of Xtcad1, Xtcad2, and Xtcad3 to their laevis orthologues is 90%, 95%, and 94%, respectively.
Outside their homeodomains, mouse and frog caudal-related proteins have relatively low amino acid sequence identity. However, Marom et al. (1997) have identified a set of core conserved motifs, which they have used to classify vertebrate Cdx proteins. By using these criteria, the nearest mouse homologue of Xtcad1 is Cdx2 (43% amino acid identity) and the nearest chicken homologue is CdxC (43% amino acid identity). The nearest mouse and chicken homologues of Xtcad2 are, respectively, Cdx1 (54% amino acid identity) and CdxA (70% amino acid identity). The nearest mouse and chick homologues of Xtcad3 are, respectively, Cdx4 (42% amino acid identity) and CdxB (51% amino acid identity). The alignments of the tropicalis, mouse, and chicken caudal-related proteins are shown in Figure 2A–C.
Expression Patterns of Xtcad1, Xtcad2, and Xtcad3
To determine the expression of the tropicalis Cdx genes during normal development, whole-mount in situ hybridizations were carried out on embryos from several stages between blastula stage 9 and larval stage 42. Figures 3, 4, and 5 show the expression patterns Xtcad1, Xtcad2, and Xtcad3, respectively. During the mid- to late blastula stage 9, there is no expression of the Xtcad genes detectable by in situ hybridization (Figs. 3A, 4A, 5A). The expression of Xtcad1 and Xtcad3 is first detected in a ring around the equator of the embryo within the nascent mesoderm at the early gastrula stage 10+. Figures 3B,C, and 5B,C show dorsal and ventral half embryo views of this expression. The expression of Xtcad2 at this early stage is barely detectable and is somewhat diffuse (Fig. 4B,C). At gastrula stage 10.5 to stage 11, Xtcad expression is in a circumblastoporal ring, although the ring of expression is somewhat narrower and fainter in the dorsal blastopore region (Figs. 3D, 4D, 5D). Down-regulation of expression in this region of the embryo, which corresponds to Spemann's organizer is greatest with Xtcad2 and, in many embryos, expression is absent from the dorsal midline. This dorsal down-regulation of expression has previously been noted for the laevis Cdx genes (Isaacs et al., 1999; Pillemer et al., 1998a,b).
High levels of expression from all three genes persists around the blastopore through gastrula stages. Figures 3E, 4E, and 5E show expression around the closing blastopore at late gastrula stage 12. After the closure of the blastopore at gastrula stage 13, expression is confined to the posterior of the embryo. The expression of Xtcad1 and Xtcad3 extends approximately 30% along the anteroposterior axis from the blastopore (Figs. 3F, 5F). The expression of Xtcad2 is slightly more restricted and extends approximately 25% the length of the anteroposterior axis (Fig. 4F). By late neurula stage 20 embryos, the expression of the Xtcad genes becomes restricted to the neural tube and future tail forming region of the embryo (Figs. 3G, 4G, 5G). Within the neural tube, the Xtcad genes have a striking nested expression pattern, with Xtcad3 having the most anterior expression and Xtcad2 having the most posterior boundary of expression. Such a nested set of expression domains has been previously noted for chick and laevis Cdx genes (Marom et al., 1997; Pillemer et al., 1998a).
To show the distribution of Xtcad transcripts within the different tissue layers in late neurula embryos, we prepared cut face embryos, bisected along the anteroposterior axis, in a slightly parasagittal position. Figure 3H shows that Xtcad1 expression is found in the dorsal neuroectoderm and the ventral ectoderm. There is also some expression in the dorsal mesoderm and in the mesoderm of the circumblastoporal collar. No expression is detected within the yolky ventral endodermal mass. Figure 4H shows that Xtcad2 expression is confined to the posterior dorsal neuroectoderm and the posterior ventral ectoderm. No expression is detected in deeper layers of the mesoderm and endoderm. Figure 5H shows Xtcad3 expression at high levels in the dorsal and posterior ventral ectoderm. Within the mesoderm, Xtcad3 expression is present all around the closed blastopore and, as with Xtcad1 and Xtcad2, no expression is detected within the endoderm.
At tail bud stage 30, Xtcad1 expression is detected in the tail bud and the most posterior neural tube, although in this specimen some staining is seen in the posterior endodermal mass (Fig. 3I). Xtcad2 expression is only detected in the posterior tip of the neural tube and the posterior wall of the neuroenteric canal (Fig. 4I). Xtcad3 expression is found in the tail bud and the posterior neural tube (Fig. 5I).
As the tail bud extends in stage 33 embryos, Xtcad1 expression is maintained in the posterior neural tube, the whole of the tail bud and the posterior neuroenteric canal that lies ventral to the notochord (Fig. 3J). By this stage, Xtcad2 expression in the tail-forming region is not detectable, except for very faint expression in the posterior wall of the neuroenteric canal. The expression pattern of Xtcad3 at this stage is quite similar to Xtcad1, although it has a more anterior limit of expression within the neural tube and expression within the ventral neuroenteric canal is not detectable.
Expression within the developing gastrointestinal tract is a common feature of caudal-related genes. Due to technical difficulties associated with the impermeability of amphibian embryos, it difficult to detect gene expression within the developing gut. To circumvent this problem, guts from stage 40 and stage 42 embryos were dissected out and subjected to whole-mount in situ hybridization. Figure 3K,L shows that Xtcad1 is expressed in the whole of the gut endoderm except for the anterior region fated to form the stomach (Chalmers and Slack, 2000). Xtcad2 is also expressed at high levels in the developing gut (Fig. 4K,L). Interestingly, Xtcad3 expression is not detectable at these stages (Fig. 5K,L) or in later larval stage 47 guts (data not shown). Similar expression patterns have been reported within the gut endoderm for the laevis caudal-related genes (Chalmers et al., 2000).
Data in the present study underline the complex and dynamic nature of the expression from the vertebrate Cdx gene family. Our detailed analysis of the sequence and expression of the Xenopus tropicalis Cdx genes also serves to emphasize several issues to investigators who are considering working with this exciting new animal model system. First, the expression patterns of related genes are highly conserved between Xenopus laevis and Xenopus tropicalis. Second, the sequence identity between orthologous genes in laevis and tropicalis is very high, which accounts for the ability of some in situ hybridization probes to cross-react between the two species. It also means that cross species hybridization approaches to cDNA and gene cloning should be relatively simple. However, significant sequence differences do exist between orthologues in the two species, which means that knowledge of the laevis sequence is likely to be insufficient for the effective design of primers for quantitative PCR applications and antisense inhibition protocols for use with tropicalis.
Embryos were generated by in vitro fertilization by using a sperm suspension and eggs derived from females induced to lay by injection of 100 units of human chorionic gonadotrophin. Embryo culture was at 23°C in 10% Normal Amphibian Medium (NAM) (Slack et al., 1973). Embryos were staged by comparison to the normal table of development of Xenopus laevis (Nieuw- koop and Faber, 1967).
Library Screening and Sequence Analysis
A total of 106 colonies of a Xenopus tropicalis gastrula stage 10–12 cDNA library in the pCS107 vector (kind gift of Dr. Aaron Zorn) were transferred to Hybond-N (Amersham) hybridization membrane and screened with [32P]dCTP random oligonucleotide labelled probes derived from the protein coding regions of Xenopus laevis Xcad1, Xcad2, and Xcad3 cDNAs. Hybridization was carried out overnight at 42°C in the presence of 5× SSC and 50% formamide. Stringency washes were 5× SSC at room temperature for 30 min, 1× SSC at 65°C for 60 min, and 0.1× SSC at 65°C for 90 min. Filters were exposed to Kodak X-omat film overnight. Multiple positive colonies identified on the primary screen were subsequently isolated as pure clones through secondary and tertiary screens. Positive clones were subjected to automated sequencing, and the largest clones corresponding to Xtcad1 (3218 bp), Xtcad2 (2643 bp), and Xtcad3 (2377 bp) were fully sequenced on both strands. The sequences for Xtcad1, Xtcad2, and Xtcad3 have been deposited in the GenBank database with accession numbers AF417107, AF417198, and AF417199.
In Situ Hybridization
Xtcad1 was linearised with NdeI, Xtcad2 template was linearised with EcoRI and Xtcad3 was linearised with PvuII. These templates were used to generate DIG-labelled probes by using a 10× DIG RNA labelling mix (Roche) and T7 RNA polymerase. Embryos were cultured to appropriate stages and then fixed in MEMFA (0.1 M MOPS, 2 mM EDTA, 1 mM MgSO4, 3.7% formaldehyde) for 1 hr at room temperature and stored in 100% ethanol at −20°C until further processing. Embryos were rehydrated through a graded series of ethanol and then rinsed in PBS with 0.1% Tween. Proteinase K treatment was done for between 10 and 15 min at room temperature with 10 μg/ml of proteinase K. Hybridisation was carried out overnight at 60°C in 50% formamide, 5× SSC, 1 mg/ml total yeast RNA, 100 μg/ml heparin, 1× Denhardt's, 0.1% Tween, 0.1% CHAPS, 10 mM EDTA. Extensive washes in 2× SSC and 0.2× SSC at 60°C were followed by washes at room temperature with maleic acid buffer (MAB; 0.1 M maleic acid, 0.15 M NaCl, 0.1% Tween, pH 7.8) and blocking in 2% Roche Blocking Reagent and 20% heat treated lamb serum for 2 hr at room temperature. Embryos were then incubated with anti-DIG antibody at a dilution of 1:2,000 in blocking solution at 4°C overnight. The antibody was detected after extensive washes at room temperature in MAB by a colour reaction by using BM purple precipitating alkaline phosphatase substrate (Roche). Endogenous pigment was partially removed from embryos by bleaching for several hours in phosphate buffered saline containing 5% (v/v) H2O2 after colour development. Digital photography of specimens was carried out by using a Spot Junior CCD camera (Diagnostic Instruments). Image manipulation was carried out by using Adobe Photoshop.
The authors thank Aaron Zorn for the gift of the Xenopus tropicalis cDNA library used in this study. We also thank Bruce Blumberg, Abraham Fainsod, and David Kimelman for Xenopus laevis cDNA probes. This work was supported by the Wellcome Trust (H.V.I. and J.S.R.-H.) and the Biotechnology and Biological Sciences Research Council (I.D.K.).