SEARCH

SEARCH BY CITATION

Keywords:

  • wnt11b;
  • wnt11;
  • X. laevis;
  • Pronephros;
  • development;
  • Holtfreter sandwich cultures

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

In this study, we aimed to establish if known wnt signaling molecules could be responsible for inducing early pronephros specification, using a novel and effective in vitro bioassay in Xenopus embryos. Anterior somites have the unique biological activity to signal to unspecified intermediate mesoderm to induce pronephros formation in Xenopus embryos. We have used a molecular candidate gene approach to analyze both canonical and noncanonical wnt expression in isolated anterior and posterior somites and dissected presumptive pronephros, pronephric anlagen, and pronephros from stage 12.5–35 embryos. We have identified potential candidate wnt genes expressed in the right time and place to specify pronephric development. These candidates were then directly tested in an in vitro pronephrogenesis assay based on Holtfreter sandwich cultures. Results revealed that noncanonical wnt11b and wnt11 can induce pronephros formation in vitro. Loss-of-function experiments confirmed that these genes are necessary for normal pronephros development. Developmental Dynamics 239:148–159, 2010. © 2009 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

In larval amphibians, the functional excretory organ, the pronephros, is a paired organ consisting of a single nonintegrated nephron, derived from the intermediate mesoderm, that lies between the anterior paraxial mesoderm (presumptive somites) and the lateral plate. During the early gastrula stage unspecified intermediate mesoderm requires inductive signals during early neurulation to specify the pronephric anlagen, which then undergoes morphogenesis to give rise to a functional renal unit, the nephron (Vize et al.,1997; Brändli,1999). Specification of all three components of the pronephros, the glomus, proximal tubule, and intermediate/distal tubule, occurs between stages 12.5 and 14 (Brennan et al.,1998,1999). The induction of the pronephric anlagen from intermediate mesoderm starts with a signal coming from the anterior somites (Seufert et al.,1999; Mauch et al.,2000; Mitchell et al.,2007). However, the molecular identification of this inductive signal has not been yet elucidated. We hypothesized that a member of the family of wnt signaling molecules might be a candidate pronephric inducing molecule.

Expression analysis of Wnt family genes in mouse embryos have identified expression patterns of six Wnt molecules in mammalian kidney development Wnt2b, Wnt4, Wnt6, Wnt7b, Wnt9b, and Wnt11. A further eight family members can be identified by sequencing of kidney specific cDNAs (reviewed in Merkel et al.,2007). In the metanephros, Wnt9b is expressed throughout the Wolffian duct and ureteric bud during development, and in the newborn and adult kidney (Qian et al.,2003), suggesting that Wnt9b is required for multiple aspects of the urogenital system development (Carroll et al.,2005). Wnt9b−/− mice die within 24 hr of birth from kidney agenesis. The ureteric bud invades the metanephric mesenchyme but fails to induce Wnt4, which is normally secreted by the induced metanephric mesenchyme where its expression is required in the pretubular aggregates. Wnt4−/− mice also lack kidneys due to a failure in mesenchymal/epithelial transition and fail to survive more than 24 hours. Wnt6 is also expressed throughout the ureteric bud/proximal collecting bud system with the exception of the ureteric bud tips (Itäranta et al.,2002). Wnt11 is secreted by the ureteric bud tips where it participates in promoting ureteric branching in parallel with the Ret/Gdnf pathway (Kispert et al.,1996; Majumdar et al.,2003; Chi et al.,2004). Although Wnt11 is expressed in the tips of the growing ureter where tubules are induced, Wnt11 itself has not been shown to be able to induce renal tubule formation in a cellular assay, although Wnt9b, Wnt6, and Wnt4 all do (Kispert et al.,1998).

In the Xenopus, several wnt molecules have been shown to be expressed in the developing pronephros. In X. laevis, wnt4 expression is restricted to the anterodorsal portion of the proximal tubule anlagen corresponding to the dorsal tips of the growing tubule (Carroll and Vize,1999) and normal expression levels are essential for tubulogenesis (Saulnier et al.,2002). We have shown by both reverse transcriptase-polymerase chain reaction (RT-PCR) and in situ analysis that wnt9b is expressed during X. laevis pronephros development (Tételin and Jones, unpublished observations), although in X. tropicalis embryos, wnt9b was shown to be expressed in other developing organs excluding the pronephros (Garriock et al.,2007), an expression pattern which contrasts with that observed in mouse (Qian et al.,2003; Carroll et al.,2005). wnt9a (formerly wnt14) is detected throughout the length of the newly differentiated pronephric intermediate/distal tubule in both Xenopus species (Garriock et al.,2007; Tételin and Jones, unpublished observations). Canonical wnt6 is observed in organs undergoing EMT and MET, and is detected in the Xenopus pronephric anlagen and differentiated pronephric tubules during embryogenesis (Lavery et al.,2008) and also in the Xenopus adult kidney (Wolda and Moon,1992).

In this study, we have analyzed the expression of both canonical and noncanonical wnt signaling molecules in dissected anterior versus posterior somites at stage 17 and during pronephros formation from stage 12.5 until the stage of pronephros functionality, stage 35. We have developed a novel assay to study pronephric specification in vitro, utilizing Holtfreter sandwich cultures, in which unspecified intermediate mesoderm and anterior somitic tissue are co-cultured. We have tested directly the ability of the candidate pronephric inducer wnt molecules to specify intermediate mesoderm to form pronephric structures and identify wnt11 and wnt11b as having this biological activity. We confirm this activity by two independent loss-of-function experiments utilizing either a dominant-negative or morpholino approach.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

wnt Signaling Molecules Are Expressed in Both Somitic and Pronephric Tissues

We have chosen to analyze the expression of wnt signaling molecules present at stage 17 in X. laevis anterior and posterior somites to identify potential candidate pronephric inducers. Stage 17 is the earliest stage at which the somites can be accurately dissected, as indicated by Myf5 expression (data not shown) and has been shown to have activity in a Holtfreter sandwich assay for pronephrogenesis (Seufert et al.,1999; Mei et al.,2001). We also have examined the temporal expression pattern of wnt signaling molecules throughout key stages of pronephros development by RT-PCR analysis.

The expression profiles of the wnt molecules identified within the somites and the pronephric anlagen can be divided into three expression pattern groups (Fig. 1). Group A, which includes wnt11b, wnt4, wnt9b, and wnt11, are either equally expressed in both anterior and posterior somites or more strongly expressed in the anterior somites and are also expressed in the developing pronephros. Noncanonical wnt11b (Garriock et al.,2007, previously described as wnt11) is weakly expressed in the pronephros from the earliest stage tested, stage 12.5, and gradually increases in expression until stage 35. The canonical/noncanonical wnt4 is expressed weakly in the pronephros at stage 15 and 20, and is more strongly and constantly expressed in the pronephros from stage 25 to 35. Canonical wnt9b is weakly detected in the pronephros at stage 15, with significant expression from stage 20 until stage 35. The noncanonical wnt11 (Garriock et al.,2007; previously described as wnt11-R) is very weakly expressed in the pronephros at stage 25 and more strongly from stages 30 to 35. Group B includes wnt8 and wnt5a, which are expressed more strongly in the posterior somites than in the anterior somites. Canonical wnt8 is expressed during the early stages of pronephros development only, from stages 12.5 to 20. Noncanonical wnt5a is strongly expressed throughout all stages of pronephros development. Group C includes wnt9a, wnt6, and wnt7b, which are not expressed in either somite compartments at stage 17. wnt9a is expressed in the pronephros anlagen from stage 15 until stage 35. Canonical wnt6 is weakly expressed throughout pronephros development from stage 15 to stage 35. Finally, canonical wnt7b is expressed within the pronephros only during later stages of development, from stages 25 to 35.

thumbnail image

Figure 1. wnt signaling molecules are expressed in both somitic and pronephric tissues. For each X. laevis embryo (anterior to the top, dorsal side to the right), the red shape indicates the somites or the pronephric anlagen dissected at various stages of development (somites were dissected at stage 17 and pronephric anlagen from stage 12.5 to stage 35). At stage 17, anterior somites correspond to somites 3–5 and posterior somites are unsegmented. Total RNA was extracted from each sample and cDNA templates were used in reverse transcriptase-polymerase chain reaction (RT-PCR) analysis. eef1a1 or odc were used to equalize cDNA input into the RT-PCR samples. The black triangle indicates the canonical wnt molecules; the black square indicates the noncanonical wnt molecules; and the question mark indicates that it is unclear whether the wnt molecule acts by means of the canonical or noncanonical pathway. The wnt molecules tested within the somites and the pronephric anlagen are divided in three groups on the basis of their expression pattern. Group A (wnt11b, wnt4, wnt9b, and wnt11) are expressed within the anterior somites and throughout pronephros development with the exception of wnt11 which is only expressed late within the developing pronephros. Group B (wnt8 and wnt5a) is more expressed in the posterior than in the anterior somites and wnt8 is expressed only during the early stage of the pronephros development (up to stage 20) while wnt5a is expressed throughout pronephros formation. Finally, group C (wnt9a, wnt6, and wnt7b) is not expressed in the somites at stage 17, wnt9a and wnt6 are expressed in the developing pronephros from stage 15 onward, while wnt7b shows expression within the pronephros during later stages of development, from stage 25 to 35, the last stage tested.

Download figure to PowerPoint

In conclusion, the noncanonical wnt molecules wnt11, wnt11b, and wnt5a and the canonical/noncanonical wnt4 are expressed in the somites in Xenopus embryos at stage 17, whereas the canonical wnt, wnt9b, wnt8, wnt7b, wnt9a, and wnt6 are weakly expressed or absent in such tissues. All the wnt molecules tested show characteristic temporal expression patterns within the developing pronephros, which are independent of the canonical or noncanonical classes to which they belong. Since it has been shown that the anterior somites are the tissue responsible for kidney induction (Seufert et al.,1999), the wnt molecules with a potential inducing signal for kidney formation would be those expressed in the anterior somites such as wnt11, wnt9b, wnt11b, and wnt4. We have directly tested their potential to induce pronephros using a Holtfreter pronephrogenesis assay.

Holtfreter Sandwich Cultures Can Be Used as a Direct Assay of Pronephrogenesis

To investigate the role of candidate wnt signaling molecules in Xenopus pronephrogenesis, an in vitro pronephrogenesis assay using Holtfreter sandwich cultures was developed. Here, we describe the development and the validation of this pronephros induction assay. Ectodermal animal caps were dissected from either control uninjected embryos or embryos overexpressing the wnt signaling molecules of interest, at stage 9 (Fig. 2A,a). Unspecified intermediate mesoderm was dissected from normal embryos at stage 10.5–11, in a wedge of 60° to 70° from the vertical midline (Fig. 2A,b). Finally, anterior somitic tissue was removed from normal embryos at stage 17 (Fig. 2A,c). The animal caps were used to wrap unspecified intermediate mesoderm and somitic mesoderm (control caps), or unspecified intermediate mesoderm alone (wnt overexpressing caps; Fig. 2B). Cultures were allowed to develop until stage 40–41 equivalent according to the stage of control sibling embryos to those used for the intermediate mesoderm dissection. At this stage, the pronephric tubules have fully specialized, and can be double immunostained for the presence of pronephric structures. Figure 3 shows control whole-mount immunostained stage 40–41 embryos and sandwich cultures. In control embryos, 3G8 stains the luminal surface of the highly convoluted anterior tubules mass (Fig. 3e; Vize et al.,1995), distal tubules were visualized using 4A6 immunostaining, which decorates the entire surface of the cells, the tube lumen is not visible (Fig. 3j; Vize et al.,1995). Immunostaining thus provides an accurate means of distinguishing pronephric proximal tubules from pronephric intermediate and distal tubules.

thumbnail image

Figure 2. Dissection of explants and assembly of Holtfreter sandwich cultures. A: Dissections of explants. Animal caps are dissected from embryos at stage 9 (a). The intermediate mesoderm is removed from the overlying epidermis from normal embryos at stage 10.5–11, in a wedge of 60° from the vertical midline (b). Finally, the anterior somites are dissected from embryos at stage 17 (c). B: A photograph of a sandwich assembly with intermediate mesoderm (IM) and somites (S) placed together in a first animal cap. Intermediate mesoderm looks much whiter in color than somitic mesoderm. A second animal cap was used to cover and seal the Holtfreter sandwich culture.

Download figure to PowerPoint

thumbnail image

Figure 3. Holtfreter sandwich cultures can be used as a direct assay of pronephrogenesis. The sandwich cultures combinations were assayed for the presence of pronephric proximal tubules using the specific antibody 3G8 and for pronephric intermediate/distal tubules using the specific antibody 4A6. Two animal caps cultured together do not give any tubule structure (a,f). Two animal caps containing intermediate mesoderm alone do not give rise to pronephric structures (b,g). Two ectodermal wraps containing somites alone do not result in tubule formation (c,h). However, intermediate mesoderm and somites wrapped together into two animal caps can develop pronephric proximal tubules (d) and pronephric intermediate/distal tubules (i). Whole embryos immunostained at stage 40–41 show proximal tubule morphology (e) and intermediate/distal morphology (j).

Download figure to PowerPoint

Two control animal caps, sandwiched together and cultured until stage 40–41, do not form either 3G8 or 4A6 positive pronephric tissue (Fig. 3a,f; Table 1) and, therefore, provide an inert wrap for the culture of embryonic explants. Unspecified intermediate mesoderm (stage 10.5–11), the tissue from which the pronephros is derived, cultured alone inside two control animal caps formed pronephric tubules in a small percentage of cases (3G8, 6%; 4A6, 5%; N = 108, Table 1). Similarly, anterior somites cultured alone inside two animal caps showed, a small percentage of sandwich cultures formed pronephric tubules (3G8, 3%; 4A6, 3%; N = 38; Table 1). However, when intermediate mesoderm and anterior somites were cultured together inside two animal caps, a significant increase in the formation of both proximal and intermediate/distal tubule structures occurred (3G8, 25%; N = 92; 4A6, 13%; N = 85; Table1; Fig. 3d,i, respectively), thus confirming that somite is able to induce pronephric tissue from unspecified intermediate mesoderm. We hypothesized that wnt signaling molecules secreted from the anterior somites could be the key molecules in the induction of pronephros and tested this directly in our in vitro pronephrogenesis culture assay.

Table 1. Non-canonical wnt11b and wnt11 Can Induce Pronephric Tubule Formation in Holtfreter Sandwich Cultures
  1. Key: AC, animal caps; IM, intermediate mesoderm; S, somites.

inline image

Holtfreter Sandwich Cultures Expressing Some wnt Molecules Form Pronephric Tissues

Animal caps overexpressing wnt do not express nervous system, kidney, or muscle markers.

To test whether wnt overexpressing Holtfreter sandwich cultures can give an inductive signal to unspecified intermediate mesoderm to differentiate into pronephros, it was essential to verify that wnt signaling molecules do not have the ability to induce formation of pronephric or other tissues when cultured on their own. X. laevis embryos were injected with 2 ng of either wnt4, wnt5a, wnt6, wnt7b, wnt8, wnt9a, wnt9b, wnt11b, or wnt11 mRNA into the animal pole at the one-cell stage. After overnight incubation, animal caps were dissected and incubated until stage 21 or 30–32 equivalent relative to undissected embryos. RT-PCR analyses for the presence of epidermis, kidney, muscle, and nervous system tissues were carried out. The animal caps overexpressing wnt molecules contained epidermal tissue confirmed by the presence of epidermis markers such as cytokeratin and bmp4, but did not form pronephric, somitic, or nervous tissues confirmed by the absence of expression of the pronephric markers lhx1, pax2, pax8, and wt1, muscle markers, myf5, myod1, and actc1, nervous system markers ncam and shh (Supp. Fig. S1, which is available online). In all cases, wnt overexpressing whole embryos cultured to stage 30–32 showed abnormal morphology when compared with control embryos, indicating that the mRNA were biologically active (data not shown).

wnt overexpressing Holtfreter sandwich cultures possess distinct morphologies.

wnt Holtfreter sandwich cultures were set up as previously described, with stage 10.5–11 intermediate mesoderm dissected from control uninjected embryos cultured alone inside two animal caps dissected from embryos overexpressing 2 ng of wnt4, wnt5a, wnt6, wnt7b, wnt8, wnt11b, or wnt11 mRNA and cultured until stage 40–41 equivalent (Fig. 4). It proved impossible to get survival in sandwich cultures from embryos injected with wnt9b mRNA to these late stages, therefore, no sandwich data are available for this molecule. Intermediate mesoderm cultured alone inside two control animal caps showed the rounded, expanded shape characteristic of mesoderm containing structures (Fig. 4a). The positive control Holtfreter sandwich cultures also looked round and clearly showed mesodermal activity (Fig. 4b). Commonly, the positive control Holtfreter sandwich cultures and the noncanonical wnt5a, wnt11b, wnt11, (and wnt4, data not shown) injected sandwich cultures contained a few pigmented cells probably derived from neural crest (Tucker,2004), but exhibited the same rounded morphology (Fig. 4c, d and e, respectively). This was in stark contrast to the canonical wnt Holtfreter sandwich cultures. wnt6 Holtfreter sandwich cultures developed eye-like structures (Fig. 4f arrow). wnt7b Holtfreter sandwich cultures displayed large numbers of pigmented cells (possibly melanocytes), as indicated by the arrow (Fig. 4g). Finally, wnt8 Holtfreter sandwich cultures formed extended axial structures that resembled tails (Fig. 4h, arrow).

thumbnail image

Figure 4. Holtfreter sandwich cultures overexpressing wnt mRNAs have distinctive morphological features. Holtfreter sandwich cultures were allowed to develop until stage 40–41 equivalent according to control embryos of the same developmental stage as those used for the intermediate mesoderm dissection. Intermediate mesoderm alone cultured in normal animal caps give sandwich cultures with a rounded shape, which clearly contain mesodermal structures (a). The positive control, containing intermediate mesoderm and somites cultured together in normal animal caps, also shows mesodermal activity (b). Noncanonical wnt5a, wnt11b and wnt11 Holtfreter sandwich cultures show comparable morphology to the positive sandwich cultures (c, d, and e, respectively). However, canonical wnt6 Holtfreter sandwich cultures developed eye-like structures as shown by the arrow (f). Canonical wnt7b Holtfreter sandwich cultures show the presence of pigmented cells (possibly melanocytes), as indicated by the arrow (g). Canonical wnt8 Holtfreter sandwich cultures formed extended axial structures that resemble tail-like structures as shown by the arrow (h).

Download figure to PowerPoint

To conclude, the positive control Holtfreter sandwich cultures and the noncanonical wnt Holtfreter sandwich cultures showed a highly similar rounded morphology, whereas canonical wnt Holtfreter sandwich cultures developed structure more characteristic of dorsal tissues, including some nervous tissue. These Holtfreter sandwich cultures were then subjected to 3G8/4A6 antibody staining for the presence of pronephric tubules.

Noncanonical wnt11b and wnt11 can both induce pronephric tubules formation in Holtfreter sandwich cultures.

Initially we focused on those wnt molecules that were expressed in the anterior somites by RT-PCR. This analysis had demonstrated that wnt11b and wnt11 are expressed both in anterior somites and during pronephros development in early X. laevis embryos. Therefore, wnt11b and wnt11 constitute potential pronephric inducers, a hypothesis which can be directly investigated using the Holtfreter sandwich culture model. The rationale for this experiment is that, if either of these wnt molecules are the natural inducing molecules, secretion from the overexpressing animal caps will induce the uncommitted intermediate mesoderm to form pronephric tissue. Figure 5A,B shows that two animal caps dissected from embryos overexpressing 2 ng of wnt11b or wnt11 mRNA cultured on their own do not have the ability to form any pronephric structure (Fig. 5a, b, e, and f, respectively; Table 1). However, animal caps expressing wnt11b, cultured with intermediate mesoderm, form pronephric proximal tubules and intermediate/distal tubules at frequencies similar to the positive control pronephrogenic sandwiches (3G8, 23%; N = 137; 4A6, 20%; N = 124; Table1; Fig. 5c,d). Similarly, when two animal caps from embryos overexpressing 2 ng of wnt11 mRNA were cultured with unspecified intermediate mesoderm, 17% of the cultures formed both pronephric proximal tubules and intermediate/distal tubules (3G8, 17%; 4A6, 17%; N = 83; Table 1; Fig. 5g,h). These frequencies are statistically significant (Supp. Tables S1 and S2).

thumbnail image

Figure 5. wnt11b and wnt11 can specify pronephros formation in Holtfreter sandwich cultures from uncommitted intermediate mesoderm. Animal caps overexpressing 2 ng of wnt11b or wnt11 mRNA were assembled with or without stage 10.5–11 intermediate mesoderm. Sandwich cultures were fixed at stage 40–41 equivalent and immunoassayed for the presence of the pronephric tubules using the antibodies 3G8 and 4A6. Two caps overexpressing wnt11b or wnt11 cultured without insert do not have the ability to form tubules (panel A, a, b and panel B, e, f, respectively). However, two caps overexpressing wnt11b or wnt11 and cultured with intermediate mesoderm form pronephric proximal tubules (panel A, c and panel B, g) and pronephric intermediate/distal tubules (panel A, d and panel B, h).

Download figure to PowerPoint

To confirm that pronephros was being induced in these sandwich cultures. RT-PCR analysis was carried out with primers for genes expressed in the developing pronephric proximal intermediate and distal tubules. Positive control sandwiches together with test wnt11b-injected sandwiches were positive for the early tubule markers pax2 and pax8. Of interest, the sandwiches were also positive for the glomus podocyte marker, tcf21 (Xpod1; Simrick et al.,2005), indicating that the third pronephric component, the glomus, was also present (Supp. Fig. S2). These data also indicate a concentration effect of wnt11b on the expression of these markers. Only wnt11b Holtfreter combinations were analyzed in this way.

In conclusion, unspecified intermediate mesoderm that received either a wnt11b or wnt11 signal from the injected animal caps can differentiate into pronephric proximal, intermediate, and distal tubules and probably glomus, suggesting that both wnt11b and wnt11 have the necessary biological activity to play an important role in Xenopus pronephros induction by either a direct or indirect mechanism.

Canonical/noncanonical wnt4 cannot induce pronephric tubules in Holtfreter sandwich cultures.

wnt4 is expressed equally in the anterior and posterior somites in Xenopus embryos at stage 17 and is expressed in the developing pronephros from stage 15 until stage 35 and has also been shown to play a role in pronephros tubulogenesis (Saulnier et al.,2002). Therefore, wnt4 constitutes a very interesting candidate for the Holtfreter pronephrogenesis assay. Our analyses showed that wnt4 Holtfreter sandwich cultures formed differentiated proximal and intermediate/distal tubular structure in only 9% of the cultures (3G8, 9%; 4A6, 9%; N = 23; Table 1), a result that was statistically not significant (χ2 test; Supp. Tables S1 and S2). In conclusion, we suggest that, although small numbers of sandwich cultures were assayed, wnt4 does not act as a pronephros inducer, but carries out other functions in kidney formation.

Because group B and C wnt molecules identified by their distributions in Figure 1 were not expressed in the correct tissue to be the natural inducing molecule, they were unlikely to have natural pronephric inducing activity. This was indeed the case for the noncanonical wnt5a and the canonical wnt7b. In neither case did expression in animal caps stimulate expression of pronephric markers above background levels (data not shown). However, expression of either of the canonical molecules wnt8 and wnt6 could. In fact, wnt6 induced pronephric structures at a higher level than the positive control (3G8, 41%; 4A6, 66%; N = 64; data not shown). In both of these cases, the sandwich cultures showed significant morphological differentiation, suggesting that dorsalization of the inserted intermediate mesoderm had occurred. This conclusion is supported by analysis of these sandwiches by triple immunostain for 3G8, 4A6, and 12/101, which detects somitic muscle. In 25 cultures containing somite, pronephros, or both tissues, 60% showed the simultaneous expression of both pronephric and somitic tissue. This was not the case in wnt11b-injected sandwiches in which pronephros clearly developed in the absence of detectable somitic tissue (data not shown). We suggest, that the high level of kidney induction in these wnt6-injected sandwiches was likely to be due to the inductive action of this somitic tissue.

Inhibition of wnt11b and wnt11 Functions Using Dominant-Negative (dnwnt11b) and Antisense Morpholino Approaches Affects the Normal Development of the Pronephric Kidney

With the aim of investigating the function of wnt11b in Xenopus pronephrogenesis, in vivo, we used the dnwnt11b to down-regulate wnt signalling (Tada and Smith,2000; Smith et al.,2000). A minimum of 1 ng of dnwnt11b can inhibit elongation in animal caps treated with activin, thus confirming that the dnwnt11b was active (data not shown). A total of 1.5 ng of the dnwnt11b mRNA was injected into one V2 ventral blastomere of 8-cell stage X. laevis embryos. Fate mapping of blastomeres at the 8-cell and 32-cell stages identified blastomere V2 of an 8-cell stage embryo to be fated to contribute at high frequency to pronephric structures (Dale and Slack,1987; Moody,1987a,b). The lineage tracer GFP was used to follow and select the correctly targeted embryos. Embryos were left to develop until stage 40–41 before being subjected to double antibody staining for pronephros phenotype. Figure 6A shows a normal Xenopus embryo immunostained in purple, with the specific pronephric proximal tubule antibody, 3G8, and in pink, with the specific pronephric intermediate/distal tubule antibody, 4A6. Figure 6B shows that injection of 1 ng of GFP mRNA alone did not affect the normal pronephros formation (Fig. 6 Ba,b) and, therefore, constitutes the control experiment. However, injection of 1.5 ng of dnwnt11b resulted, on the injected side, in 26% and 15% (N = 54) of such embryos showing an absence of the pronephric proximal and intermediate/distal tubules, respectively, and 35% of the embryos showed drastic reduction of the pronephric tubules structure (Fig. 6B,c), while the pronephros on the uninjected side was not affected (Fig. 6B,d). A total of 38% and 50% of the embryos developed normal proximal and intermediate/distal tubules, respectively, suggesting that inhibition of wnt11b alone is not sufficient to completely inhibit pronephros formation. Because wnt11 and wnt11b are highly related, we hypothesized that wnt11 might be compensating for the defects caused by the dnwnt11b, a hypothesis supported by a double morpholino approach. Results showed that co-injection of 15 ng of each wnt11b (previously characterized by Pandur et al.,2002) and wnt11 (previously characterized by Garriock et al.,2005) morpholinos resulted in higher percentage and more severe abnormal pronephric phenotype than the one observed of each individual morpholino injection (Fig. 6C). Injection of 15 ng of wnt11b or wnt11 individual morpholino resulted in 44% (N = 36; Fig. 6C,c) and 54% (N = 72; Fig. 6C,e) of the embryos developing reduced pronephric tubules, respectively. Co-injection of 15 ng of each wnt11b and wnt11 morpholinos resulted in reduced pronephric structures (3G8, 72%; 4A6, 60%; N = 110; Fig. 6C,g) and in the most extreme cases, embryos did not form any pronephros (3G8, 7%; 4A6, 21%; N = 110; data not shown). We suggest that wnt11b and wnt11 are essential for normal pronephros formation and that it is probable that these two genes can compensate for each other in pronephros formation.

thumbnail image

Figure 6. Inhibition of wnt11b and wnt11 functions results in abnormal pronephros formation in in vivo Xenopus embryos. A: A normal, stage 40, X. laevis embryo immunostained with the specific pronephric proximal tubule antibody, 3G8 (purple) and with the specific pronephric intermediate/distal tubule antibody, 4A6 (pink). β-GAL mRNA acts as a lineage tracer to follow those cells which inherited injected material, thus checking for correct targeting. Here, such cells are stained red in the somites resulting from mRNA injection into one V2 blastomere at the eight-cell stage embryo. B: Shows that inhibition of wnt11b using the dnwnt11b affects the normal development of Xenopus pronephros. Injection of GFP mRNA alone does not disturb normal pronephros development (B, a and b). However, injection of the dnwnt11b results in reduction of the pronephric tubule on the injected side (B, c) and in normal tubule on the uninjected side (B, d). Panel C confirms that inhibition of wnt11b and wnt11 function using morpholino knockdown results in abnormal pronephros formation. Injection of a control standard morpholino did not affect normal pronephros development (C, a and b). Injection of 15 ng of wnt11b morpholino shows reduced pronephric tubules tissues on the injected side (C, c), while the pronephric components remain normal on the uninjected side (C, d). A similar phenotype is observed with injection of 15 ng of wnt11 morpholino (C, e and f). However, when 15 ng of each wnt11b and wnt11 morpholinos was co-injected, reduction of pronephric tubules is greatly accentuated (C, g) when compared with the uninjected side (C, h) and in some cases is totally lost.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

We report the systematic survey of wnt signaling molecules in both the signaling anterior somites of neurula stage embryos and the responding presumptive pronephric tissue, pronephric anlagen, or differentiating pronephros. This approach has identified several candidate molecules, which could be, by either direct or indirect means, responsible for pronephric inducing activity. It has also eliminated some candidates, because they are not expressed in this tissue at the correct time to be able to participate in this biological activity. We have proceeded to test the majority of these molecules (with the exception of wnt9b) directly for this ability to influence pronephrogenesis.

We have shown that Holtfreter sandwich cultures are a reliable in vitro bioassay for pronephrogenic activity initially as a culture chamber for the in vivo interacting tissues and also as an assay for biologically active wnt inducing activity. This assay is powerful because it simplifies the complex endogenous embryonic tissue interactions and serves to define without doubt, anterior somites as the tissue capable of converting unspecified intermediate mesoderm into differentiated pronephric tissues (Seufert et al.,1999). Historically, these sandwich cultures have been used to assay neural inducing activity and mesodermal inducing activity (Holtfreter,1934; Jones and Woodland,1987; Gerhart,1996) but this is the first time they have been used as a vehicle to express signaling molecules within the ectoderm to affect presumptive pronephric tissue sandwiched inside the culture. In vitro signaling assays have been developed by the Kirschner group to assay the activity of signaling molecules translated in oocytes on ectoderm engrafted onto the surface of the oocytes. As in our assay, although technically difficult, these approaches can yield novel and potentially unexpected biological activities in the tested mRNAs (Lustig and Kirschner,1995).

When candidate wnt mRNAs were injected into the animal caps used in the Holtfreter combinations, each wnt resulted in explants of characteristic and different morphology. We suggest that each canonical wnt molecule overexpressed in animal caps confers specific dorsal properties to the pre-existing (intermediate) mesoderm, including formation of somitic tissue, which subsequently and secondarily causes neuralization of the ectoderm and induces differently patterned neural tissue, depending on which wnt molecule was overexpressed in the animal caps (Moon,1993; Heasman et al.,1994; Chen et al.,2000). Thus, wnt6 Holtfreter cultures formed eye-like structures, suggesting that wnt6 might act as an eye inductive factor, although this observation needs to be confirmed by either in situ hybridization using a specific eye marker, such as pax6 (Hirsch and Harris,1997) or analyzed by immunohistology (Massé et al.,2007), and wnt8 cultures displayed axial tail-like structures. It is also possible that the dorsalizing activity of wnt, generated organizer tissue, which in combination with lateral mesoderm, in Holtfreter cultures, generate eye-like structures (Sedohara et al.,2003).

In this study, we have assayed directly the potential pronephric inducing activity of wnt11b, a gene related to Wnt11 but that is absent from mammalian genomes. wnt11b is involved in various biological mechanisms such as dorsal–ventral axis formation (Ku and Melton,1993; Schroeder et al.,1999; Tao et al.,2005; Jessen et al.,2005; Kofron et al.,2007; White and Heasman,2008), dorsalization of pre-existing mesoderm (Ku and Melton,1993), normal convergent extension movements during gastrulation (Ku and Melton,1993; Conlon et al.,1996; Kispert et al.,1995; Smith et al.,2000; Tada and Smith.,2000), normal heart development (Pandur et al.,2002), and for neural crest cell migration (De Calisto et al.,2005). We show here that wnt11b, which is expressed in the anterior somites and throughout pronephros formation, can significantly induce the intermediate mesoderm to differentiate in pronephric glomus, and proximal and intermediate/distal tubules in Holtfreter sandwich cultures. This induction takes place in the absence of the formation of somite tissue, ruling out a secondary mechanism. Our immunohistochemical analyses showed that proximal and intermediate/distal tubules could form independently as unconnected tissue blocks, or with proximal and more distal tubule elements apparently joined. Examples of this were found in all samples including the positive control with somites and intermediate mesoderm.

These studies cannot distinguish whether this is a direct or indirect effect and do not rule out the possibility of involvement of as yet unidentified additional signals. However, we further confirm the central role of this molecule in vivo by the inhibition of its functions by two distinct approaches: a dominant-negative and a morpholino knockdown. Both approaches result in a reduction in size of pronephric components without affecting other axial structures. Moreover, double morpholino experiments inhibiting both the function of wnt11b and the related wnt11 showed that such injected embryos developed more severe abnormal pronephros phenotypes than when either morpholino was injected on its own, suggesting that the two genes share redundant function during organogenesis.

In mouse embryos, Wnt11, which is expressed in the tip of the ureter bud, cannot induce tubulogenesis in vitro (Kispert et al.,1998). Our data suggest that the Xenopuswnt11, although expressed in the anterior somites and in the differentiated pronephric tubules and capable of inducing pronephros formation in Holtfreter sandwich cultures, might also not be the natural source of signal for pronephric induction. Indeed, in situ analysis of wnt11 shows that it is expressed only in the dorsal tips of the somites, whereas wnt11b is expressed throughout the somite (Supp. Fig. S3). From an evolutionary perspective, it is interesting to speculate that wnt11b, is needed in organisms that used the pronephros as an essential embryonic kidney form, such as Xenopus and Zebrafish. Garriock et al., (2007) have shown that the wnt11b copy has probably been lost in mammals, a loss that correlates with the brief rudimentary function of pronephros in this vertebrate Class.

The canonical wnt6 also induced pronephric structures in Holtfreter sandwiches. We previously hypothesized that overexpression of wnt6 mRNA in the epidermal animal caps dorsalized the engrafted intermediate mesoderm, which could result in the formation of somites. This was supported by the high frequency of muscle found in the explants, which itself could induce pronephros and would explain the high frequency of pronephric structure contained in wnt6 overexpressing sandwiches. However, even though wnt6 is detected in the Xenopus pronephric anlagen and differentiated pronephric tubules (Lavery et al.,2008, and this study), it is not expressed in the anterior somites and so is unlikely to be the natural source of signal to induce the intermediate mesoderm into pronephros. In fact, wnt6 has been shown to be required for the epithelialization leading to somite formation and subsequent compartmentalization into the sclerotome and the dermomyotome in chick embryos (Schmidt et al.,2004; Geetha-Loganathan et al.,2006; Schubert et al.,2002; Rodríguez-Niedenführ et al.,2003). Wnt6 has been shown to be capable of inducing mouse kidney tubule development in vitro, although this does not necessarily suggest that it actually plays such a role in vivo (Itäranta et al.,2002). While we expect the presence of somitic tissue to be the explanation of the likely cause of high frequency of pronephros formation in these explants, it is formally possible that additional, as yet uncharacterized factors are induced by wnt6 expression, and are capable of increasing the efficiency of kidney induction in these conjugates.

In conclusion, we have developed a direct assay for pronephrogenesis using the in vitro advantages of the classic Holtfreter sandwich culture. We have used this, together with data on the endogenous levels of wnt RNAs to identify roles for wnt11b and wnt11 activity in specifying the development of pronephros from unspecified intermediate mesoderm. We have confirmed this important role identified in vitro, by knockdown of wnt11b function by two independent methods in vivo. Future experiments will establish the mechanism by which wnt11b achieves this function.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Eggs and Embryos

X. laevis eggs were obtained by in vitro fertilization following standard procedures. Staging of embryos was according to Nieuwkoop and Faber (1994).

Microinjection

dnwnt11b (Smith et al.,2000; Tada and Smith,2000), wnt11, and wnt11b oligo morpholinos were injected together with lineage tracers GFP or β-GAL into one V2 blastomere at the eight-cell stage, which contributes to the pronephric lineage (Moody,1987a,b). wnt11 and wnt11b oligo morpholinos sequences were accessed from published data, wnt11b (5′-CCAGTGACGGGTCGGAGCCATTGGT-3′) (Pandur et al.,2002) and wnt11 (5′-CTTCATCTTCAAAACCCAATAACAA-3′) (Garriock et al.,2005). Injections were performed in 5% Ficoll in full-strength Barth X.

Holtfreter Sandwich Cultures

Lateral plate mesoderm from stage 10.5–11 embryos and presumptive anterior somites from stage 17 embryos were cultured together and separately within normal or 2-ng wnt overexpressing ectodermal wraps in full-strength Barth X for approximately 4 hr. Sandwich cultures were then switched into 0.5% Barth X with 10 μg/ml gentamycin and cultured at room temperature until stage 40 equivalent, according to the stage of control embryos used for lateral plate dissection. Holtfreter sandwich cultures were finally fixed in MEMFA (0.5 M MOPS pH 7.4, 100 mM ethyleneglycoltetraacetic acid, 1 mM MgSO4, 4% formaldehyde) before being subjected to immunohistochemistry for the presence of pronephric proximal, intermediate/distal tubules with antibodies 3G8 and 4A6, respectively. wnt clones were as follows: wnt4 (McGrew et al.,1992), wnt5a (Moon et al.,1993), wnt6 (Lavery et al.,2008), wnt7b (Chang and Hemmati-Brivanlou,1998), wnt8 (Christian et al.,1991), wnt9b (Garriock et al.,2007), wnt11b (Ku and Melton,1993), and wnt11 (Garriock et al.,2005).

Whole-Mount Immunohistochemistry

MEMFA fixed embryos or Holtfreter sandwich cultures were washed in water, dehydrated through serial dilutions of methanol, and then rehydrated in serial dilutions of methanol/phosphate buffered saline (PBS). Pigmented embryos were bleached (0.9% H2O2, 5% formamide, 0.5× standard saline citrate) on a light box. Samples were subsequently washed 3 times and blocked in PBT (2 mg/ml bovine serum albumin, 0.1% Triton X-100 in PBS) for 2 hr at 4°C on a nutator. Antibodies were added to samples and incubated rocking overnight at 4°C, with dilutions as follows in parentheses: primary antibodies 3G8 (1:40) and 4A6 (1:1) (Vize et al.,1995), secondary antibody, alkaline phosphatase conjugated goat anti-mouse (1:500) (Sigma F-2012). When fluorescent staining was required, antibodies 3G8 and 4A6 were used neat and somite monoclonal antibody 12/101 (1:500) (Brockes and Kintner,1984), secondary fluorescein isothiocyanate goat anti-mouse (1:25) (Invitrogen), TRITC goat anti-mouse (1:50) (Invitrogen), and Alexa fluor 405 goat anti-mouse (1:200) (Invitrogen). Samples were washed, after being incubated with antibody, 5 times for 1 hr with PBT at 4°C. Samples were equilibrated in color buffer (0.1 M Tris-Cl pH 9, 25 mM MgCl2, 100 mM NaCl, 0.1% Tween-20) at room temperature for 1 hr and then color substrate was added for 3G8 staining (color buffer with 1% v/v NBT/BCIP, Roche), for 4A6 staining (Fast Red TR/Napthol AS/MX, Sigma) according to the manufacturer's instructions. When sufficient signal had developed, the reaction was stopped with PBS. Samples were re-fixed in MEMFA, dehydrated in methanol, and cleared in Murrays (1:2 benzyl alcohol, benzyl benzoate) before being scored under white or ultraviolet light.

RT-PCR Analyses

Total RNA isolated from X. laevis whole embryos, dissected pronephric anlagen, pronephros, and somites was used to generate first-strand cDNA. RT samples without reverse transcriptase were also prepared to check for contamination of genomic DNA. The quantity of input cDNA was verified by similarity of odc (ornithine decarboxylase) or eef1a1 (eukaryotic translation elongation factor 1 alpha 1) signals. Linearity of amplification was confirmed by PCR with consecutive doubling dilutions of input cDNA. Primers used in this study were as follows: wnt4 (5′-GAGTGGAATGCAAGTGTC-3′ and 5′-TACACTGCCGACCAGTTG-3′), wnt5a (5′-ACTCCAAGCACAGACCTTAG-3′ and 5′-CATGACCTCCTGGATAACTC-3′), wnt6 (5′-CACGACCTGCTGGAAGAAGA-3′ and 5′-ACGTCCAGAGCAGTGCTGTT-3′), wnt7b (5′-CCGCACTTCTCCATACACTG-3′ and 5′-ACTGTCACCGGAGTGC- CTTA-3′), wnt8 (5′-AGATGACGGCATTCCAGA-3′ and 5′-TTCGGAACGGAAGCATGT-3′), wnt9a (5′-GGCGAATAAGTCCAGCAAGG-3′ and 5′-CCGGAGAGTACTTGCTCATC-3′), wnt9b (5′-AGAATGGAGCGATGTACC-3′ and 5′-GGCGAGAGTTGCTTCCAACA-3′), wnt11b (5′-GAAGTCAAGCAAGTCTGCTGG-3′ and 5′-GCAGTAGTCAGGGGAACTAACCAG-3′), wnt11 (5′-AGCTCATGCACCTGCACAAC-3′ and 5′-TCCTGCACCGGCCTTATATC-3′), odc (5′-GGAGCTGCAAGTTGGAGA-3′ and 5′-TCAGTTGCCAGTGTGTGGTC-3′), and eef1a1 (5′-CAGATTGGTGCTGGATATGC-3′ and 5′-CACTGCCTTGATGACTCCTA-3′).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  9. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
DVDY_22012_sm_SupFigS1.tif212KSupp. Fig. S1. Animal caps overexpressing wnt molecules do not induce nervous system, kidney, or muscle markers. Single cell X. laevis embryos were injected with 2 ng of either wnt4, wnt5a, wnt6, wnt7b, wnt8, wnt11b, or wnt11 mRNA. Animal caps were dissected at stage 9 and left to develop to stage equivalent 21 whole embryo (WE). Total RNA was extracted from each sample before being subjected to reverse transcriptase-polymerase chain reaction (RT-PCR) analysis. eef1a1 was used to equalize the cDNA samples. Primers for specific tissues markers were used to test the presence of epidermis, kidney, muscle, and nervous system in animal caps overexpressing wnt molecules. Animal caps overexpressing wnt molecules contain epidermis tissue but do not show the presence of kidney, muscle, or nervous system tissues.
DVDY_22012_sm_SupFigS2.tif166KSupp. Fig. S2. Additional molecular pronephric markers are expressed in Holtfreter sandwich cultures overexpressing wnt11b. Sandwich combinations were set up as indicated on the figure: control animal caps cultured on their own or with intermediate mesoderm or with both intermediate mesoderm and somites, and animal caps overexpressing 2 or 3 ng of wnt11b mRNA cultured on their own or with intermediate mesoderm. Reverse transcriptase-polymerase chain reaction (RT-PCR) analyses confirmed the presence of both pronephric tubules specification markers, pax2 and pax8, in wnt11b Holtfreter sandwich cultures and moreover showed the presence of the pronephric glomus marker tcf21 in such cultures. odc was used to equalize the cDNA samples. Primers used for this study were as follows: pax2 (5′-TCGGAAGAAGAGTGGTCTAC-3′ and 5′-GGTATTCATATTCCGCATTC-3′), pax8 (5′-CCAACAGCAGCATCAGATC-3′ and 5′-CAATGACACCTGGCCGGATA-3′), and tcf21 (5′-TCTCAGTGATGTGGAGGACTT-3′ and 5′-TGACGCAGGTGAGCTATGTAA-3′).
DVDY_22012_sm_SupFigS3.tif758KSupp. Fig. S3. wnt11 and wnt11b show different expression patterns in vivo. Whole Xenopus embryos were harvested for in situ hybridization using untranslated region gene-specific wnt11 and wnt11b in situ probes. At stage 34, wnt11b is clearly expressed throughout the segmented somites (a), whereas wnt11 is observed in the neural tube and at the very dorsal-most tip of the somites (b).
DVDY_22012_sm_SupTabS1.tif382KSupp. Tables S1 and S2. Statistical analyses show that some wnt molecules but not all can significantly induce the formation of proximal tubule (Table 1) and intermediate/distal tubule (Table 2) in Holtfreter sandwich cultures. The chi squared (χ2) test allows a comparison of the observed number of Holtfreter sandwich cultures assayed for tubule formation when wnt molecules were overexpressed in animal caps and cultured with unspecified intermediate mesoderm with a theoretically expected distribution. For each assay, the χ2 value determines whether the difference in tubule formation between wnt sandwich cultures and the background control (AC+ IM) was significant. The χ2 calculated value is compared with tabulated value for one degree of freedom at 95% probability. When the χ2 calculated value exceeds the tabulated value of 3.8, then tubule formation is related to the overexpression of the tested wnt molecule.
DVDY_22012_sm_SupTabS2.tif364KSupp. Tables S1 and S2. Statistical analyses show that some wnt molecules but not all can significantly induce the formation of proximal tubule (Table 1) and intermediate/distal tubule (Table 2) in Holtfreter sandwich cultures. The chi squared (χ2) test allows a comparison of the observed number of Holtfreter sandwich cultures assayed for tubule formation when wnt molecules were overexpressed in animal caps and cultured with unspecified intermediate mesoderm with a theoretically expected distribution. For each assay, the χ2 value determines whether the difference in tubule formation between wnt sandwich cultures and the background control (AC+ IM) was significant. The χ2 calculated value is compared with tabulated value for one degree of freedom at 95% probability. When the χ2 calculated value exceeds the tabulated value of 3.8, then tubule formation is related to the overexpression of the tested wnt molecule.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.