Drs. In der Rieden and Lloret Vilaspasa contributed equally to this work.
Special Issue Research Article
Xwnt8 directly initiates expression of labial Hox genes
Article first published online: 21 JUL 2009
Copyright © 2009 Wiley-Liss, Inc.
Special Issue: Special Issue on WNT Signaling in Development and Disease
Volume 239, Issue 1, pages 126–139, January 2010
How to Cite
In der Rieden, P. M. J., Vilaspasa, F. L. and Durston, A. J. (2010), Xwnt8 directly initiates expression of labial Hox genes. Dev. Dyn., 239: 126–139. doi: 10.1002/dvdy.22020
- Issue published online: 15 DEC 2009
- Article first published online: 21 JUL 2009
- Manuscript Accepted: 22 APR 2009
- EU Network. Grant Number: LSHM-CT-2003-504468
- Hox genes;
- WNT signal
Hox transcription factors play an essential role in patterning the anteroposterior axis during embryogenesis and exhibit a complex array of spatial and temporal patterns of expression. Their earliest onset of expression in vertebrates is during gastrulation in a temporally collinear sequence in the presomitic/ventrolateral mesoderm, and it is not clear which upstream signal transduction events initiate this expression. Using Xenopus, we present evidence that Xwnt8 is necessary for initiation of this collinear sequence by activating Hox-1 expression in three Hox clusters: hoxd, hoxa, and hoxb. All three labial genes appear to be direct targets of canonical Wnt signaling through Tcf/Lef. In addition, Xwnt8 loss- and gain-of-function leads to indirect regulation of other Hox genes: Hoxb4, Hoxd4, Hoxa7, Hoxc6, and Hoxc8. These findings shed new light on the early role of Wnt8 as well as of a proposed WNT gradient in patterning the Xenopus central nervous system (Kiecker and Niehrs  Development 128:4189–4201). Developmental Dynamics 239:126–139, 2010. © 2009 Wiley-Liss, Inc.
Hox proteins are involved in the specification of positional identities along the anteroposterior (AP) axis and other embryonic axes in a wide range of animal species, including vertebrates (Bürglin et al.,1991; McGinnis and Krumlauf,1992; Bürglin and Ruvkun,1993; Lawrence and Morata,1994; Manak and Scott,1994). Hox genes in tetrapod vertebrates are organized in four clusters located on different chromosomes. The clusters are thought to have arisen by tandem duplication of a single gene, followed, in vertebrates, by duplication of the cluster itself (Schughart et al.,1989; Ruddle et al.,1994; Bailey et al.,1997; Greer et al.,2000). As a consequence, Hox genes occupying the same relative position along the 5′ to 3′ chromosomal coordinate, named paralogous genes, share more similarity in sequence and expression pattern than do adjacent Hox genes on the same chromosome (Greer et al.,2000). A phenomenon of particular interest is that Hox genes located at the 3′ end of a cluster are expressed earlier and more anteriorly than the subsequently more 5′ located genes (Gaunt et al.,1988; Duboule,1994; Gaunt and Strachan,1996). How spatiotemporal colinearity in the expression of Hox genes is regulated is intriguing but, to date, not well understood.
Recently, the early Hox expression patterns have been analyzed in Xenopus gastrula stage embryos (Wacker et al.,2004). This revealed temporally collinear initiation of expression of a sequence of Hox genes within a horseshoe-shaped domain in ventrolateral marginal zone mesoderm at different stages during gastrulation, followed by sequential dorsalization of each Hox expression zone into a stable AP zone in axial mesoderm and the neural plate. Comparable patterns of early Hox gene expression are seen in gastrulae or early mesoderm of other vertebrates, and have been best characterized in chicken (Gaunt and Strachan,1996; Iimura and Pourquié,2007). To gain further insight into the regulation of spatiotemporal Hox expression in marginal zone mesoderm, we wished to identify factors involved the initiation of mesodermal Hox expression.
In Xenopus, Xwnt8 expression is first detected in late blastula stage embryos. Expression is found in all cells of the marginal zone with the exception of the cells centered on the dorsal midline. This pattern of expression in the ventrolateral marginal zone persists during gastrulation (Christian and Moon,1993). Ectopic expression of Xwnt8 posteriorizes neuroectoderm (Fredieu et al.,1997; Erter et al.,2001; Kiecker and Niehrs,2001), a feature also known for Hox genes (Charité et al.,1994; Hooiveld et al.,1999; Maconochie et al.,1997; McNulty et al.,2005). Conversely, gain-of-function for Xdkk1, a secreted Wnt antagonist (Glinka et al.,1998), down-regulates expression of Hoxd1 in neuroectoderm of Xenopus embryos (Kiecker and Niehrs,2001). In mouse and chick embryos, the expression patterns of the Xwnt8 orthologs have been considered to be indicative for a possible function in the regulation of expression of labial-type Hox genes. In chick embryos, the expression of Cwnt8C immediately precedes the localization of Hoxb1 expression to rhombomere 4 (Hume and Dodd,1993). In mouse embryos, expression of Mwnt8 is found in the presumptive rhombomere 4 region (Bouillet et al.,1996). In Caenorhabditis elegans, Wnt/WG signaling elements are involved in the regulation of ceh-13, the labial ortholog of the worm (Streit et al.,2002). Furthermore, expression of Cwnt8C in the chick coincides with that of Hoxb1 at the onset of gastrulation in the nascent primitive streak, the site of ingression of mesodermal and endodermal cells, and remains to be expressed across this structure during gastrulation (Chapman et al.,2004). These properties make Xwnt8 a good candidate to fulfill the role of initiator of Hox expression in marginal zone mesoderm of Xenopus embryos.
Xwnt8 is a member of the Wnt family of secreted glycoproteins, which act as ligands, activating receptor-mediated signal transduction pathways (reviewed in Moon et al.,2002, and references therein). After binding of Xwnt8 to suitable receptors, intracellular signals are transduced by the canonical Wnt pathway (Darken and Wilson,2001), which acts through a rise in cytosolic and subsequent nuclear levels of β-catenin, influencing the function of Tcf/Lef transcription factors. Misexpression of synthetic Xwnt8 mRNA on the presumptive ventral side, before the activation of the zygotic genome, leads to formation of a secondary axis (Sokol et al.,1991), while later activation of Xwnt8 expression leads to posteriorization of the primary axis (Christian and Moon,1993). In Xenopus embryos, it has been shown that β-catenin–induced axis formation is mediated by means of the transcription factor XTcf3 (Molenaar et al.,1996). The early and late effects of ectopic Xwnt8 on axis formation can be mimicked by timed activation of an activated form of XTcf3 (Darken and Wilson,2001).
So, is Xwnt8 signaling involved in the initiation of mesodermal Hox expression? To answer this question, we used the following strategies. First, we made a detailed description of the time dependent early expression pattern of Xwnt8 and examined whether the expression of Xwnt8 coincides with the expression of Hoxd1, Hoxb4, and Hoxc6 during gastrulation. These three genes were chosen because they are expressed in well-defined spatial domains in the early neuroectoderm, corresponding to mid-hindbrain (the identities of rhombomeres 4 and 5) (Hoxd1, Kolm and Sive,1995b), posterior hindbrain (Hoxb4, Harvey and Melton,1988), and anterior spinal cord (Hoxc6, Oliver et al.,1988; De Robertis et al.,1989), in addition, the spatiotemporally collinear expression of these genes in ventrolateral mesoderm has been described (Wacker et al.,2004). We report a significant overlap in expression between Xwnt8 and the Hox genes examined in ventrolateral mesoderm during gastrula stages.
Next, we analyzed the effects of Xwnt8 loss-of-function, using a morpholino-based strategy, on development and on the expression of several early Hox genes during gastrulation. Xwnt8 loss-of-function leads to an anteriorization of the developing embryo, accompanied by a reduction in expression of Hoxd1, Hoxd4, Hoxb1, Hoxb4, Hoxb5, Hoxa1, Hoxa7, and Hoxc8. Expression of Hoxc6 is up-regulated.
Next, we performed Xwnt8 gain-of-function experiments. This results in posteriorized embryos and an up-regulation of most of the Hox genes that were regulated by Wnt8MO. To investigate whether the observed effects on Hox expression by Wnt8 gain-of-function are direct, we undertook an approach using fusion of an activated form of XTcf3 to the ligand-binding domain of the glucocorticoid receptor, which allows hormonal regulation of nuclear translocation. Activation of a dominant positive form of XTcf3 shortly before gastrulation, leads in a direct manner to up-regulation of Hoxd1, Hoxb1, and Hoxa1 expression. All other Wnt-sensitive genes examined were regulated indirectly. In this report, we present evidence that Xwnt8 function is necessary and sufficient to directly induce the expression of Hoxd1, Hoxb1, and Hoxa1 in gastrula embryos. Furthermore, we show that Hoxd1 expression in gastrula mesoderm can be initiated, in a direct manner, by Tcf/Lef signaling.
We believe that these findings suggest that Wnt8 is a specific initiator of expression in the Hoxd, Hoxa, and Hoxb clusters. If this is true, it is the first identification of an early initiator of Hox cluster expression.
Xwnt8 and Anterior Hox Genes Have Partially Overlapping Expression Domains During Gastrulation
If Xwnt8 is involved in the initiation of Hox gene expression, it needs to be coexpressed with Hox genes. Because Wnt family members are secreted factors, their functional domains could extend beyond the borders of their mRNA expression domains, but nonetheless, overlapping expression of Xwnt8 and Hox genes could reveal functional relations. We compared the detailed expression patterns of Xwnt8 and three early Hox genes in gastrula and early neurula stage embryos.
Early during gastrulation Xwnt8 is expressed in a horseshoe-like pattern in the mesoderm, showing a gap of expression corresponding to the organizer mesoderm (Fig. 1A). During the progression of gastrulation, Xwnt8 expression expands in the animal direction (Fig. 1B,C). Expression of Xwnt8 is lost at the ventralmost side of the embryo around stage 12 (Fig. 1C). Expression is maintained in dorsolateral mesodermal domains close to the blastopore, and in involuted mesoderm (Fig. 1C). During early neurulation, three domains of Xwnt8 expression can be observed on either side of the midline: a domain in the paraxial mesoderm, a domain in the presumptive hindbrain neuroectoderm, where Xwnt8 expression anterior boundary coincides with the anterior expression domain in paraxial mesoderm, and a posterior domain in dorsolateral mesoderm (Fig. 1E).
Expression of Hoxd1 starts in a horseshoe-like pattern in the marginal zone mesoderm at stage 10.25 (Fig. 2B) and two dorsolateral domains become prominent as gastrulation progresses (Fig. 2B). At stage 11.5, the ectoderm overlying the dorsolateral mesodermal expression domains starts to express Hoxd1 (Fig. 2B). Early during neurulation (St.12.5), expression of Hoxd1 can be found anteriorly in ectoderm, and in lateral mesoderm extending backward to the almost closed blastopore (Fig. 2B). The expression patterns of Hoxd1 and Xwnt8 in gastrula stages show a clear overlap (compare Fig. 2A with 2B). During early gastrulation, the overlap can be found in marginal zone mesoderm. At stage 13, expression of both genes is found in neuroectoderm. In paraxial and ventrolateral mesoderm, expression of Xwnt8 is within the domain of Hoxd1 expression (compare Fig. 2A with 2B).
Initiation of Hoxb4 expression during gastrulation takes place later than Hoxd1 initial expression (stage 10.5), but in a similar nested domain in marginal zone mesoderm (Fig. 2C). At stage 12, ectoderm overlying the dorsolateral mesodermal expression domains starts to express Hoxb4 (Fig. 2C). At stage 13, this ectodermal expression is located more posteriorly than the ectodermal expression of Hoxd1 (compare Fig. 2B with 2C). Expression of Hoxb4 overlaps with that of Xwnt8 in marginal zone mesoderm (compare Fig. 2A with 2C). At stage 12, this overlap is restricted to the dorsolateral domain of Hoxb4 expression; at stage 13 Xwnt8 and Hoxb4 are coexpressed in paraxial mesoderm, while no overlap can be observed in neuroectoderm. Expression of Hoxc6 is initiated in a similar pattern to that of Hoxd1 and Hoxb4, starting at stage 11.5 in marginal zone mesoderm (Fig. 2D). At stage 13, expression of Hoxc6 can be observed in ectoderm overlying the dorsolateral mesodermal expression domain (Fig. 2D), with its anterior expression boundary located posterior to the most anterior expression of Hoxb4 (compare Fig. 2C with 2D). The expression patterns of Hoxc6 and Xwnt8 overlap in marginal zone mesoderm but not in neuroectoderm. The expression overlap in posterior dorsolateral mesoderm persists during later gastrula stages (compare Fig. 2D with 2A).
These results are consistent with the possibility that Xwnt8 could serve a role as an initiator of Hox gene expression during gastrulation.
Xwnt8 Loss-of-Function Leads to Anteriorization of Embryos and Loss of Hoxa1, Hoxb1, and Hoxd1 Expression as Well as of Other Hox Genes
To investigate whether Xwnt8 is of importance for the early expression of Hoxd1, Hoxb4, and Hoxc6, as well as for other Hox genes, we used a loss-of-function method using an Xwnt8 morpholino antisense oligonucleotide (MOXwnt8). Several loss-of-function strategies have been used to study the function of Xwnt8: dnWnt8 (Hoppler et al.,1996), Xdkk-1 (Glinka et al.,1998), and Sizzled (Salic et al.,1997). The advantage of a morpholino-based approach is the reported high specificity (reviewed in Heasman,2002, and references therein). By binding of the morpholino to sequences overlapping, or lying adjacent to, the start site of translation, the targeted mRNA is not translated (reviewed in Heasman,2002). This approach results in a potentially more specific Xwnt8 loss-of-function method as compared to overexpressing antimorphic forms of Xwnt8 or Wnt antagonists.
MOXwnt8 was injected into the animal hemisphere of embryos at the one-cell stage or (in both cells) at the two-cell stage, resulting in spreading of the MOXwnt8 all over the embryo; subsequently, the embryos were allowed to develop until control embryos reached stage 24 (Fig. 3A) or stage 35 (Fig. 3B). Knocking down Xwnt8 function by injection of MOXwnt8 leads to anteriorization of the embryo in a concentration dependent manner (Fig. 3). In MOXwnt8 injected embryos, the axis was reduced, and an enlargement of the cement gland was observed (compare Fig. 3A with 3C and 3D, and Fig. 3B with 3E). This phenotype has also been reported for the other Xwnt8 (or Wnt) loss-of-function methods mentioned above: dnWnt8 (Hoppler et al.,1996), Xdkk-1 (Glinka et al.,1998), Sizzled (Salic et al.,1997). In zebrafish embryos, injection of morpholinos directed against both the Zwnt8 ORFs found (Erter et al.,2001; Lekven et al.,2001) leads to comparable effects to those we observed for Xenopus using the MOXwnt8. A control morpholino (MOcontr), unrelated in sequence to MOXwnt8, was injected in the same amounts as the MOXwnt8. Abnormalities were not observed in the development of the embryos injected with MOcontr (data not shown). The specificity of the MOXwnt8 was further shown by rescue of the Xwnt8 loss-of-function phenotype with CS2-Xwnt8 morpholino insensitive (MOI) plasmid DNA (see the Experimental Procedures section for details). 64 ng of MOXwnt8 and 20 pg of MOI CS2-Xwnt8 were injected either singly or in combination into the animal hemisphere of embryos at the one-cell stage. The embryos receiving both the MOXwnt8 and the MOI CS2-Xwnt8 show a clear reduction in size of the cement gland as compared to the single injection of the MOXwnt8 (Fig. 3F).
After confirming that the MOXwnt8 is a valid Xwnt8 loss-of-function reagent, we investigated its effects on the expression patterns of 6 Hox genes: Hoxd1, Hoxa1, Hoxb1, Hoxb4, Hoxd4, and Hoxc6. Mesodermal expression of Hoxd1 was strongly down-regulated, and the distance between the two dorsolateral domains of expression in marginal zone mesoderm was increased by Xwnt8 loss-of-function (Fig. 4A). Ectodermal expression of Hoxd1 was also down-regulated in injected embryos (Fig. 4A). Hoxa1 was regulated weakly in the early gastrula mesoderm but apparently not in the early neurula (stage 13) neuroectoderm (Fig. 4B). Hoxb1 is also down-regulated in the neuroectoderm at early neurula (Fig. 4C). Expression of Hoxb4 in mesoderm and ectoderm was also modestly altered by Xwnt8 loss-of-function (Fig. 4D). Hoxd4 was down-regulated too (Fig. 4 E). Expression of Hoxc6 was ectopically up-regulated in dorsal mesoderm of stage 10.5 embryos and in mesoderm and dorsal ectoderm of embryos at st. 12 (Fig. 5F and see Fig. 5G for views on dissected embryos). In situ hybridizations were performed on embryos injected with 64 ng of MOcontr. For all markers studied, injection of the control morpholino results in unaltered expression (data not shown). To confirm these results and at the same time extend our analysis, we used reverse transcriptase-polymerase reaction (RT-PCR) to examine the effect of Xwnt8 loss-of-function on expression of several Hox genes (Fig. 6). This confirmed the effects on Hoxa1, Hoxb1, Hoxd1, and late Hoxc6 (whereas st. 11 expression levels seemed rather down-regulated), but showed a stronger down-regulation of Hoxb4 (at st. 13) and an earlier down-regulation of Hoxb1 (stage 11) at a stage where the Hoxb1 in situ hybridization yielded too low a signal to estimate possible effects of Xwnt8 loss-of-function. In addition, early Hoxa7 (at st. 11) and neurula Hoxc8 and more slightly Hoxb5 (st.13) were found to be down-regulated, Hoxd4 could not be detected in our PCRs.
We further checked the effects of the Xwnt8 loss-of-function on the expression of the organizer mesoderm and anterior neural plate marker Otx2. We found that its domain was expanded at stage 10.5 after in situ hybridization analysis.
We interpreted these results as a molecular confirmation of the morphological phenotype observed after injection of Xwnt8MO; namely, an anteriorization of the mid-axial region of the embryo.
Ectopic Expression of Xwnt8 After the Mid-blastula Transition Leads to an Up-regulation of Expression of Hox Genes
To study the effects of Xwnt8 gain-of-function on gastrulation and neurulation, we designed a construct driving expression of Xwnt8 after the mid-blastula transition (MBT). This avoids the early, dorsalizing, activity found following Xwnt8 synthetic mRNA injections (Smith and Harland,1991; Sokol et al.,1991). To this end, we generated a plasmid containing the full-length coding region of Xwnt8 in the CS2+ vector (Rupp et al.,1994) and named the construct CS2-Xwnt8. The CS2+ vector harbors a sCMV promoter leading to efficient expression and subsequent translation of the derived mRNA in Xenopus embryos after the MBT (Turner and Weintraub,1994; Kühl et al.,1996). A mutated MO insensitive version of this was used for rescue experiments. Embryos at the one-cell stage were injected into the animal hemisphere with 100 pg of CS2-Xwnt8 plasmid; this results in a clear posteriorization (Fig. 3F). Next, we assayed for the early expression of the same Hox genes examined above, Otx2, the mesodermal marker Xbra and the posterior marker Xcad3, in CS2-Xwnt8 injected embryos. Strong up-regulation of the expression of Hoxd1 could be observed (Fig. 5A). Not only is the expression domain larger as compared to control embryos, but the expression also appears earlier and can be observed in the organizer field (Fig. 5A). Later during gastrulation, ectopic Hoxd1 expression continues to be present in ectoderm and mesoderm in the midline of the embryo, and is expanded in anterior direction (Fig. 5A). Hoxa1 was up-regulated in the gastrula and closed up its horseshoe pattern over the organizer field to form a ring, after Xwnt8 overexpression; at stage 13 however, Hoxa1 expression looked no longer up-regulated (Fig. 5B). CS2-Xwnt8 injections cause strong ectopic up-regulation of Hoxb1 at both stage 11 and stage 13 (Fig. 5C); up-regulation appeared randomly distributed in any region of the embryo and was observed in both ectoderm and mesoderm in different individuals (Fig. 5C presents an example). Expression of Hoxb4 was up-regulated in a similar way to Hoxa1 in the early gastrula (stage 11) and, like this later gene, it did not seem up-regulated any longer at stage 13 (Fig. 5D) by Xwnt8 gain-of-function (Fig. 5B). Expression of Hoxd4 (Fig. 5E) was up-regulated during gastrula as well as neurula stages. Wnt8 gain-of-function. caused ectopical expression of Hoxc6 much earlier than its normal expression and was found in dorsal (organizer) mesoderm, tissue that normally does not express Hox genes. Later in gastrulation, an expansion of the endogenous horseshoe-shaped domain is observed (data not shown). In early neurula stages, an anterior expansion of the expression of Hoxc6 is observed in neuroectoderm (Fig. 5F). The effect of Xwnt8 injection on Otx2 expression was opposed to that seen on most Hox genes in that it was down-regulated (Fig. 5G). Expression of the mesodermal marker Xbra was unaltered by the Xwnt8 gain-of-function (Fig. 5H), suggesting that formation of mesoderm was not affected by exogenous Xwnt8. Expression of the posterior marker Xcad3 was up-regulated in mesoderm (data not shown) and ectoderm of injected embryos (Fig. 5I), confirming the posteriorizing nature on neuroectoderm of CS2-Xwnt8 injection. The different effects that the misexpression of Xwnt8 has on the expression of Hoxd1 Hoxa1, Hoxb4, Hoxd4, and Hoxc6 demonstrate the complex and dynamic regulation of the expression of Hox genes in marginal zone mesoderm and suggest a role for Xwnt8 in this regulation. These results were extended by using RT-PCR to monitor Hox expression (Fig. 6). This could confirm the effects of Xwnt8 ectopic expression on Hoxd1, Hoxa1, Hoxb1, and Hoxb4 (only detected strong enough at st. 13) and Hoxc6 It also showed that Hoxa7 is up-regulated (at both st. 11 and 13). Hoxb5 was not visibly altered and Hoxc8 seemed rather down-regulated.
Labial Type Hox Genes Are Direct Targets of Canonical Wnt Signaling
It is has been shown that Xwnt8 uses the canonical Wnt pathway before and after the onset of gastrulation, stabilizing cytosolic β-catenin and activating gene expression through Tcf/Lef transcription factors (Darken and Wilson,2001). To investigate whether the induction of anterior Hox genes by Xwnt signaling is direct, we made use of an activated, hormone inducible form of XTcf3, TVGR (Darken and Wilson,2001). Embryos were injected into the animal hemisphere at the one-cell stage with 100 pg of TVGR. CHX was added at stage 9.5, followed 0.5 hr later by addition of DEX to the appropriate samples, well before the onset of gastrulation and the initiation of Hox gene expression. At stage 11, the embryos were harvested for RT-PCR. The results are shown in Figure 6. Because CHX was added before the onset of gastrulation, induction of Hoxd1, Hoxa1, Hoxb1, Hoxb4, and Hoxc6 expression in control embryos is reduced or absent (Fig. 7). Expression of Hoxd1, Hoxa1 and Hoxb1 is directly activated by the TVGR, while none of the other Hox genes examined were directly induced. The expression of all other Hox genes studied is, however, up-regulated when DEX was added in absence of CHX, demonstrating indirect induction of expression of Hoxb4, Hoxc6, and Hoxa7 (data not shown).
We checked and extended these results using in situ hybridization. This gave the same result (Fig. 8). In addition, it was noticeable that, induced expression after activation of TVGR was confined to the usual NOM mesodermal domain, while hoxa1 and hoxd1 expression following activated TVGR with CHX is not. It extends into the organizer mesoderm and into the ectodermal animal cap. We suspect that CHX blocks the production or action of repressors that normally define the limits of the Hox expression domains.
Ectopic Xwnt8 Directly Initiates Expression of Hoxa1, Hoxb1, and Hoxd1
We report that ectopic Xwnt8 expression can initiate Hoxd1, Hoxa1 and Hoxb1 expression in mesoderm and ectoderm of Xenopus gastrula stage embryos. Xwnt8 is thus able to ectopically induce expression of the three earliest expressed Hox genes in Xenopus and can do this earlier than initiation of endogenous expression and in tissues normally not expressing these Hox genes as well as in endogenously expressing tissues. Kiecker and Niehrs (2001) have reported that the injection of pCSKA-Xwnt8 (CSKA-X8, Christian and Moon,1993) into Xenopus embryos does not alter the expression of Hoxd1, an apparent contradiction to the results shown in this report, because we do observe an increased expression of Hoxd1 by ectopic Xwnt8. In our hands, the pCSKA-Xwnt8 construct was also not able to initiate the expression of Hoxd1 in mesoderm or ectoderm. This could be due to the specific untranslated region (UTR) sequences contained in the pCSKA-Xwnt8 plasmid, as UTR sequences are known to affect the stability of mRNA and to regulate the translation of the messenger (reviewed in Derrigo et al.,2000, and references therein). Our results show that Xwnt8 is capable of initiating the expression of Hoxd1 and Hoxa1. The Wnt induction of these two genes and of Hoxb1 is direct. Activation in the presence of cycloheximide by means of a hormone inducible VP16 activated form of the transcription factor Xtcf3-TVGR (Darken and Wilson,2001) induced expression of Hoxd1 and Hoxa1 and Hoxb1. Tcf3 mediates action of the canonical Wnt pathway and is known to induce Wnt8 targets (Darken and Wilson,2001). Hoxa1, Hoxd1 and Hoxb1 were the only three of 7 Hox genes examined to be regulated directly. Other Hox genes: Hoxa7, Hoxc6, and Hoxd4 were regulated indirectly by the Wnt pathway. Hoxb4 was not obviously Wnt regulated. See also below.
Endogenous Xwnt8 Signaling Is Necessary for Endogenous Expression of Hoxa1, Hoxd1, and Hoxb1, as Well as Other Hox Genes, in Dorsolateral Mesoderm and Neuroectoderm
The necessity of Xwnt8 function for Hoxa1, Hoxd1, and Hoxb1 expression in marginal zone mesoderm is shown by Xwnt8 loss-of-function experiments, where a strong reduction in expression of these genes and other genes can be observed. A mechanism whereby different inputs are capable of starting Hox expression from different Hox paralog groups or from different Hox clusters could be of importance in the regulation of Hox gene expression and thus for patterning the anteroposterior axis. The effect of Xwnt8 loss-of-function on expression of Hoxc6 expression is striking. Hoxc6 is up-regulated in dorsal mesoderm and ectoderm, tissues normally not expressing Hoxc6, significantly earlier than endogenous expression. This result is confirmed by multiple experiments using both in situ hybridization and RT-PCR, but it is hard to understand. A partial explanation could lie in the fact that Wnts are known to cause repression as well as activation. Perhaps we are concerned here with a balance between activation and repression on the same gene such that, while normal levels do not induce it, very high or very low ones do. The necessity of Wnt signaling for the expression of labial-type Hox genes is also supported by findings in C. elegans (Streit et al.,2002). It was found that expression of ceh-13, the labial ortholog of C. elegans, depends on Wnt signaling. Strikingly, regulatory elements of ceh-13 can act as WG response elements in transgenic Drosophila embryos. This evidence points, together with our results, to a conserved and ancient mechanism, wherein labial-type Hox gene expression is dependent on Wnt signaling. However, to our knowledge, it has never previously been reported that induction of labial-type Hox genes by Wnt signaling can be accomplished when protein synthesis is inhibited and that this induction is, therefore, direct.
Early Role of Xwnt8
Our findings above indicate a specific early role for Xwnt8 during gastrulation. It directly activates expression of the three earliest expressed labial Hox genes (Hoxd1, Hoxa1, and Hoxb1), and consecutively it possibly initiates expression of the Hoxa, Hoxd, and Hoxb clusters. In agreement, Xwnt8 activates expression of other Hox genes, but their activation is indirect. The specificity of these effects is emphasized by the fact that not all Hox genes examined are strongly regulated by Xwnt8 in the early embryo, even though all of these genes are expressed. This is presumably the first report of a direct initiation factor necessary for early expression of Hox genes in the early embryo. Presumably, Wnt8 also induces other factors needed for progression. One factor that could be involved is Raldh2 (enzyme responsible for the synthesis of retinoic acid), which is strongly induced by the Wnt pathway (unpublished observations). Retinoid signaling is, in turn, known to regulate neuroectodermal expression of 3′ anterior Hox genes (Lumsden and Krumlauf,1996; Gosave and Durston,1997; Durston et al.,1998; Bel-Vialar et al.,2002). It is also known that knocking out all three labial Hox genes knocks out other Hox genes in the early neural plate. Altogether, these factors could act in parallel to coordinate an appropriate Hox gene expression profile in the early embryo, which is crucial for the overall patterning of the anterior-to-posterior axis.
Is Xwnt8 Involved in Generating a Gradient?
The different effects of Xwnt8 function on the expression of different Hox paralog groups and Hox clusters may contribute to the generation of an early Hox pattern. This pattern is initiated in mesoderm, followed by the appearance of the Hox sequence in neuroectoderm. A posterior to anterior positive gradient of β-cat/Wnt signaling in neuroectoderm has previously been postulated to underlie the embryo's neuroectodermal AP pattern (Kiecker and Niehrs,2001). In fact, there are now many observations that make the action of one or more Wnt gradients likely in the AP patterning of the developing central nervous system (CNS). These range from gradients acting within the anterior brain (Lagutin et al.,2003) through patterning of the fore-mid-hindbrain region (Nordström et al.,2002) and distinguishing between head and trunk to patterning of the posterior CNS (Nordström et al.,2006). Some of the findings feature interactions between Wnt and other pathways, with Wnt being a source of graded information and other pathways sometimes being permissive.
The following point of contact with our own observations is very interesting. Nordström et al. (2006) documented the responses of three Hox genes to Wnt3A protein. These are all Hoxb genes, representing different positions along the posterior neuraxis. These are, however, responses of gastrula neural plate explants to subsequent long periods (up to 44 hr) of exposure to WNT protein, leading to a much later assay of Hox expression. The Hox gene response also requires fibroblast growth factor, and, in the case of Hoxb4, retinoic acid. This study differs from our own in several respects. It is in chicken, the response is to Wnt3A, and it examines exclusively a neural plate cell population (and not the earliest mesodermal Hox expression population). It is likely that the last parameter is crucial.
In our own study, the very early responses of Hox genes seem not to fit a gradient function of Wnt8. We are concerned with gastrula stages, where AP patterning genes (Hox genes and Otx2) are already expressed, but where the AP axis is not yet obviously set up. At this stage, the information for the APHox sequence seems to be contained in a temporally collinear sequence of Hox expression in the gastrula's nonorganizer mesoderm. If a gradient of anteroposterior patterning information were to spread from the future posterior tissues to future more anterior tissues, we expect the Hox genes to be functional downstream of this gradient and, as a consequence, to respond to changes in it. The observed effects on Hox expression by Xwnt8 loss- and gain-of-function make the existence of such an early gradient unlikely or at least, argue against regulation of all of these Hox genes by thresholds on an early source of WNT signal. In Xwnt8 loss-of-function, posterior Hox genes would be expected to be down-regulated, considering that the source of the gradient is thought to be in the posterior part of the embryo. This is in conflict with the observed induced expression of Hoxc6 in dorsal mesoderm and ectoderm, and with an enhanced level of Hoxc6 expression in ventrolateral mesoderm of embryos injected with MOXwnt8.
Differently affected expression of Hoxa1, Hoxd1, and Hoxb1 as seen in early Xwnt8 loss- and gain-of-function is also not consistent with a simple model wherein a gradient of Wnt signaling along the anteroposterior axis is used to provide positional information within the trunk (Hox expressing) part of the axis. According to such a model, a gene expressed more anteriorly could be expressed in more posterior tissues in response to loss-of-function for the gradient. The results of MOXwnt8 experiments contradict this idea; they show a strong down-regulation of Hoxd1 expression in embryos with reduced Xwnt8 signaling, and never a posterior expansion of the Hoxd1 expression zone. Up-regulation of Hoxc6 expression observed in Xwnt8 gain-of-function is much weaker and in a significantly smaller domain than up-regulated Hoxd1 expression, while leaving expression of Hoxb4 little affected. These results also contradict a model whereby an anteroposterior gradient of Wnt signaling is used to pattern the early primary axis. We propose a model wherein Xwnt8 is involved in initiating a pattern of Hox expression in the ventrolateral marginal zone mesoderm of the embryo by initiating the expression of Hoxd1, Hoxa1 and Hoxb1, in a direct manner. After the initial activation, the Hox cascade continues, by progressive temporally collinear opening and expression of the Hox clusters. Hox genes from other paralogue groups are induced indirectly because they depend on Wnt initiation of Hox cluster opening but also require other factors than just Wnt signaling, thereby creating steps in the Hox code. This leads to our conclusion that Xwnt8, and perhaps other Wnts, play an important part in setting up the early Hox codes. In fact this code of collinearly expressed Hox genes can be considered as a map of positional information along the anteroposterior axis. We note that there is evidence for interactions among Hox genes in the early embryo. Most notably, simultaneous MO loss of function of all three Hox1 (labial Hox) genes anteriorizes the most anterior Hox expressing part (hindbrain region) of the early neuraxis, deleting expression of all Hox 1, 2, and 3 paralogues and deleting the anterior parts of the expression domains of Hox 4, 5, and 6 paralogues. The anterior part of the neural plate that normally makes the hindbrain is converted to the identity of rhombomere 1 (expressing Gbx2).
DNA Constructs and Morpholino (MO) Designs
MOXwnt8, supplied by Gene Tools, LLC, has the sequence: 5′-tttgcatgatgaaggctgctatccg. The MOcontr has the sequence 5′-cctcttacctcagttacaatttata. Embryos were injected into the animal pole at the one-cell stage with 32 or 64 ng of MOXwnt8, dissolved in water, in a volume of 4 or 8 nl, respectively, or with the MOcontr using the same conditions.
The CS2-Xwnt8 construct was made by cloning the full-length coding region of Xwnt8 obtained by PCR using the CSKA-X8 plasmid (Christian and Moon,1993) as template and the following primers: f: 5′-gaggaattccggatagcagccttcatcatgcaaaacacc, r: 5′-ctactcgagtctccggtggcctctgttcttcc, containing and EcoRI or an XhoI restriction site, respectively, in the CS2+ vector (Rupp et al.,1994) using the restriction sites in the primers. A total of 50 pg, in a volume of 8 nl, of this plasmid was injected, dissolved in water, into the animal pole of embryos at the one-cell stage. For rescue experiments, we used a plasmid (MO insensitive CS2Xwnt8) (MOI CS2Xxwt8) that does not contain an intact Xxwt8 MO binding site. There are a total of 7 mismatches between the MO and its (now non complementary) binding site. Also, while the concentrations of MO used were 32 and 64 ng/ml, those of the construct were 50 pg. Considering the different molecular weights of these two reagents, this delivers approximately a 20,000-fold difference in molarity. Stoichiometric titration of the MO by this construct or the mRNA transcribed from it is inconceivable.
Synthetic capped mRNA was made using the Ambion mMessage mMachine Kit, 100 pg of TVGR mRNA was injected into the animal pole of one-cell stage embryos.
Xenopus Embryos, Culture, Treatments, and Microinjections
Pigmented Xenopus laevis embryos were obtained by in vitro fertilization, and after de-jelling in a 2% cysteine solution (pH 8.0), cultured in 0.1× Marc's Modified Ringers (MMR) (Sive et al.,2000), containing 50 μg/ml gentamycin at 14–21°C. Embryos were injected in 1× MMR + 4% Ficoll and afterward transferred to 1× MMR + 1% Ficoll, and cultured in this medium for 1 to 7 hr, after which they were transferred and to 0.1× MMR in which they were cultured until harvesting. Staging of the embryos was performed according to Nieuwkoop and Faber (Nieuwkoop and Faber,1967).
Treatments with cycloheximide (CHX) and dexamethasone (DEX) were performed using concentrations of 10 μM. CHX treatment was started 30 min before DEX addition. Embryos were harvested 4 hr after addition of CHX (Kolm and Sive,1995a).
Numbers: For each treatment, we used a minimum of 100 embryos treated with each of several concentrations in each of a minimum of four experiments. In a first experiment, dose dependence of the morphological phenotype was established. In each of at least three subsequent experiments, we injected batches of at least 100 embryos with each of at least two chosen doses and processed them for in situ hybridization or RT-PCR. The morphological and in situ phenotypes shown were obtained with very high penetrance (>90%).
Whole-Mount In Situ Hybridization and Antisense Probes
Whole-mount in situ hybridizations were performed (Harland,1991), with minor modifications. The antisense RNA probes were generated by runoff in vitro transcription using DIG RNA labelling mix (Roche), and T7 or Sp6 RNA polymerase (Promega). The probes were generated using the following templates: Hoxd1 (Sive and Cheng,1991); Hoxa1: a 1312-bp Hoxa-1 fragment; Hoxb1: a 666-bp Hoxb-1 fragment; Hoxb4: a 708-bp fragment containing the complete Hoxb-4 ORF cloned in pGEMTE; Hoxd4: EST XL094L20; Hoxc6: a 998-bp Hoxc-6 fragment in pGEM1 containing a part of the homeodomain and extending into the 3′ UTR; Xcad3 (Pownall et al.,1996); Xbra: pSP73Xbra (Smith et al.,1991); Otx-2: a 220-bp OTX-2 fragment (Pannese et al., 1995).
Total RNA was extracted using Tri-Pure reagent (Roche). First-strand cDNA was subsequently synthesized using Superscript KSII polymerase (Gibco-BRL), primed with an Oligo dT15 according to the manufacturer's instruction. RT-PCR assays were performed in the exponential phase of amplification as described (Busse and Séquin,1993) using Tfl polymerase (Promega) in buffer containing 20 mM TrisAc, pH 9.0, 75 mM KAc, 10 mM NH4SO4, 1.7 mM MgSO4, and 0.05% Tween-20. The primers used are: Hoxd1: f: 5′-agggaactttgcccaactctcc r: 5′-gtgcagtacatgggtgtctggc; Hoxa-1: f: 5′-atgtggacctgtccctagcagc r: 5′-tgctttgcagctcaatgagacc; Hoxb-1: f: 5′-tttggttgtcttgggaggatttct r: 5′-ataatggggatggaaggtttgttg; Hoxb4 (Hooiveld et al.,1999); Hoxb-5: f: 5′cgtcagtctcggaggagg r: 5′-aatgtgagcggctcatacag; Hoxc-6 f: 5′-cagagccagacgtggactattcatccagg r: 5′-caaggtaactgtcacagtatggagatgatggc; Hoxc-8 f: 5′-cacatgttacaacgccgaggccacc r: gagtgtgagttccttgctctccttagtctcctcttcctc; ODC f: 5′-gtcaatgatgggtgtatggatc r: 5′-tccattccgctctcctgagcac.
We thank C. McNulty, G. Mainguy and S. Wacker for critically reading the manuscript. P.I.d.R. was supported by a NWO-grant and EU “HPRN” and “Quality of life” programs. A.J.D. thanks EU for the network grant: “Cells into Organs” for support.
- 1997. Phylogenetic reconstruction of vertebrate Hox cluster duplications. Mol. Biol Evol 14: 843–853. , , , .
- 2002. Initiating Hox gene expression: in the early chick neural tube differential sensitivity to FGF and RA signaling subdivides the HoxB genes in two distinct groups. Development 129: 5103–5115. , , .
- 1996, A new mouse member of the Wnt gene family, mWnt-8, is expressed during early embryogenesis and is ectopically induced by retinoic acid. Mech Dev 58: 141–152. , , , , , .
- 1993. The Caenorhabditis elegans homeobox gene cluster. Curr Opin Genet Dev 3: 615–620. , .
- 1991. Nematode homeobox cluster. Nature 350: 703. , , , , , , , , .
- 1993. Molecular analysis of the Wnt-1 proto-oncogene in Ambystoma mexicanum (axolotl) embryos. Differentiation 53: 7–15. , .
- 2004. Expression analysis of chick Wnt and frizzled genes and selected inhibitors in early chick patterning. Dev Dyn 229: 668–676. , , , , .
- 1994. Ectopic expression of Hoxb-8 causes duplication of the ZPA in the forelimb and homeotic transformation of axial structures. Cell 78: 589–601. , , , .
- 1993. Interactions between Xwnt8 and Spemann organiser signalling pathways generate dorsoventral pattern in the embryonic mesoderm of Xenopus. Genes Dev 7: 13–28. , .
- 2001. Axis induction by Wnt signalling: target promoter responsiveness regulates competence. Dev Biol 234: 42–54. , .
- 1989. Determination of axial polarity in the vertebrate embryo: homeodomain proteins and homeogenetic induction. Cell 57: 189–191. , , .
- 2000. RNA-protein interactions in the control of stability and localization of messenger RNA (review). Int J Mol Med 5: 111–123. , , , .
- 1994. Temporal colinearity and the phylotypic progression: a basis for the stability of a vertebrate Bauplan and the evolution of the morphologies through heterochrony. Dev Suppl 1994: 135–142. .
- 1998. Retinoids and related signals in early development of the vertebrate central nervous system. Curr Top Dev Biol 40: 111–175. , , , .
- 2001. Wnt8 is required in lateral mesodermal precursors for neural posteriorization in vivo. Development 128: 3571–3583. , , , , .
- 1997. Xwnt-8 and lithium can act upon either dorsal mesodermal or neurectodermal cells to cause a loss of forebrain in Xenopus embryos. Dev Biol 186: 100–114. , , , , .
- 1996. Temporal colinearity in expression of anterior Hox genes in developing chick embryos. Dev Dyn 207: 270–280. , .
- 1988. Spatially restricted domains of homeo-gene transcripts in mouse embryos: relation to a segmented body plan. Dev Suppl 104: 169–179. , , .
- 1998, Dickkopf-1 is a member of a new family of secreted proteins and functions in head induction. Nature 391: 357–362. , , , , , .
- 1997. Neural induction and patterning in embryos deficient in FGF signaling. Int J Dev Biol 41: 57–65. , .
- 2000. Maintenance of functional equivalence during paralogous Hox gene evolution. Nature 403: 661–665. , , , .
- 1991. In situ hybridisations: an improved whole mount method for Xenopus embryos. Methods Cell Biol 36: 685–695. .
- 1988. Microinjection of synthetic Xhox-1A homeobox mRNA disrupts somite formation in developing Xenopus embryos. Cell 1988 53: 687–697. . .
- 2002. Morpholino oligos: making sense of antisense? Dev Biol 243: 209–214. .
- 1999. Novel interactions between vertebrate Hox genes. Int J Dev Biol 43: 665–674. , , , , , , , .
- 1996. Expression of a dominant-negative Wnt blocks induction of MyoD in Xenopus embryos. Genes Dev 10: 2805–2817. , , .
- 1993. Cwnt-8C: a novel Wnt gene with a potential role in primitive streak formation and hindbrain organization. Development 119: 1147–1160. , .
- 2007. Hox genes in time and space during vertebrate body formation. Dev Growth Differ 49: 265–275. , .
- 2001. A morphogen gradient of Wnt/β-catenin signalling regulates anteroposterior neural patterning in Xenopus. Development 128: 4189–4201. , .
- 1995a. Efficient hormone-inducible protein function in Xenopus laevis. Dev Biol 171: 267–272. , .
- 1995b. Regulation of the Xenopus labial homeodomain genes, HoxA1 and HoxD1: activation by retinoids and peptide growth factors. Dev Dyn 167: 34–49. , .
- 1996. Dominant negative expression of a cytoplasmically deleted mutant of XB/U-cadherin disturbs mesoderm migration during gastrulation in Xenopus laevis. Mech Dev 54: 71–82. , , , .
- 2003. Six3 repression of Wnt signaling in the anterior neuroectoderm is essential for vertebrate forebrain development. Genes Dev 17: 368–379. , , , , , , , , , .
- 1994. Homeobox genes: their function in Drosophila segmentation and pattern formation. Cell 78: 181–189. , .
- 2001. Zebrafish wnt8 encodes two proteins on a bicistronic transcript and is required for mesoderm and neurectoderm patterning. Dev Cell 1: 103–114. , , , .
- 1996. Patterning the vertebrate neuraxis. Science 274: 1109–1115. , .
- 1997. Cross-regulation in the mouse HoxB complex: the expression of Hoxb2 in rhombomere 4 is regulated by Hoxb1. Genes Dev 11: 1885–1895. , , , , , , , .
- 1994. A class act: conservation of homeodomain protein functions. Dev Suppl 1994: 61–77. , .
- 1992. Homeobox genes and axial patterning. Cell 68: 283–302. , .
- 2005. Knockdown of the complete Hox paralogous group 1 leads to dramatic hindbrain and neural crest defects. Development 132: 2861–2871. , , , , .
- 1996. XTcf-3 transcription factor mediates β-catenin-induced axis formation in Xenopus embryos. Cell 86: 391–399. , , , , , , , , .
- 2002. The promise and perils of Wnt signaling through beta-catenin. Science 296: 1644–1646. , , , .
- 1967. Normal table of Xenopus laevis (Daudin). Amsterdam: North Holland. , .
- 2006., An early role for WNT signaling in specifying neural patterns of Cdx and Hox gene expression and motor neuron subtype identity. PLoS Biol 4: e252. , , , .
- 2002. An early role for WNT signaling in specifying neural patterns of Cdx and Hox gene expression and motor neuron subtype identity. Nat Neurosci 5: 525–532. , , .
- 1988. Differential antero-posterior expression of two proteins encoded by a homeobox gene in Xenopus and mouse embryos. EMBO J 7: 3199–3209. , , , .
- 1996. eFGF, Xcad3 and Hox genes form a molecular pathway that establishes the anteroposterior axis in Xenopus. Development 122: 3881–3892. , , , .
- 1994. Gene loss and gain in the evolution of the vertebrates. Dev Suppl 1994: 155–161. , , , .
- 1994. Xenopus embryos regulate the nuclear localisation of XMyoD. Genes Dev 8: 1311–1323. , , .
- 1997. Sizzled: a secreted antagonist expressed in the ventral marginal zone of Xenopus embryos. Development 124: 4739–4748. , , , .
- 1989. Duplication of large genomic regions during the evolution of vertebrate homeobox genes. Proc Natl Acad Sci U S A 86: 7067–7071. , , .
- 1991. Retinoic acid perturbs the expression of Xhox.lab genes and alters mesodermal determination in Xenopus laevis. Genes Dev 5: 1321–1332. , .
- 1991. Injected Xwnt-8 RNA acts early in Xenopus embryos to promote formation of a vegetal dorsalizing center. Cell 67: 753–765. , .
- 1991. Expression of a Xenopus homolog of Brachyury (T) is an immediate-early response to mesoderm induction. Cell 67: 79–87. , , , , .
- 1991. Injected Wnt RNA induces a complete body axis in Xenopus embryos. Cell 87: 741–752. , , , .
- 2002. Conserved Regulation of the Caenorhabditis elegans labial/Hox1 Gene ceh-13. Dev Biol 242: 96–108. , , , , , , , , .
- 1994. Expression of achaete-scute homolog 3 in Xenopus embryos converts ectodermal cells to a neural fate. Genes Dev 8: 1434–1447. , .
- 2004. Timed interactions between a Hox cascade and the Spemann organiser generate positional information during vertebrate gastrulation. Dev Biol 268: 207–219. , , , , .