XB-ART-40327Development. November 1, 2009; 136 (21): 3585-95.
Notch activates Wnt-4 signalling to control medio-lateral patterning of the pronephros.
Previous studies have highlighted a role for the Notch signalling pathway during pronephrogenesis in the amphibian Xenopus laevis, and in nephron development in the mammalian metanephros, yet a mechanism for this function remains elusive. Here, we further the understanding of how Notch signalling patterns the early X. laevis pronephros anlagen, a function that might be conserved in mammalian nephron segmentation. Our results indicate that early phase pronephric Notch signalling patterns the medio-lateral axis of the dorso-anterior pronephros anlagen, permitting the glomus and tubules to develop in isolation. We show that this novel function acts through the Notch effector gene hrt1 by upregulating expression of wnt4. Wnt-4 then patterns the proximal pronephric anlagen to establish the specific compartments that span the medio-lateral axis. We also identified pronephric expression of lunatic fringe and radical fringe that is temporally and spatially appropriate for a role in regulating Notch signalling in the dorso-anterior region of the pronephros anlagen. On the basis of these results, along with data from previous publications, we propose a mechanism by which the Notch signalling pathway regulates a Wnt-4 function that patterns the proximal pronephric anlagen.
PubMed ID: 19793883
Article link: Development.
Grant support: G12713 Biotechnology and Biological Sciences Research Council , G1988 Biotechnology and Biological Sciences Research Council
Genes referenced: dll1 gal.2 gata3 gnl3 hey1 lfng lhx1 myh4 myh6 notch1 nphs1 odf3 rfng slc5a2 sst t tbx2 wnt4 atp1b1
Morpholinos referenced: hey1 MO2 lfng MO1 lfng MO2 suclg2 MO1 wnt4 MO1
Article Images: [+] show captions
|Fig. 1. Mis-activating early pronephric Notch signalling caused ectopic medial pronephrogenesis. (A-J) X. laevis embryos were injected at the eight-cell stage into a V2 blastomere to target tissues including the presumptive pronephric region. notch-ICD mRNA or deltaSTU mRNA was co-injected with beta-galactosidase mRNA to act as a lineage tracer [blue (A,B); red (C-J) staining on the injected side]. At stage 41, 3G8 is ectopic (white arrows) and 4A6 is reduced in embryos injected with notch-ICD mRNA (A). deltaSTU reduces 3G8 staining (white arrow) but has no effect on 4A6 staining (B). At stage 36, Na+K+ATPase expression is reduced after overexpression of notch-ICD (C), but only the proximal domain of Na+K+ATPase expression is reduced upon deltaSTU overexpression (D). slc5a2 expression is inhibited after overexpression of notch-ICD (E) and deltaSTU (F). At stage 32, odf3 (white arrows, G) and nephrin (black arrow, I) expression is ectopic upon notch-ICD mRNA injection. odf3 (white circle, H) and nephrin (J) are inhibited by deltaSTU. Asterisk denotes injected side.|
|Fig. 2. wnt4 overexpression induces the same phenotypes as injection of notch-ICD mRNA. (A-E) X. laevis embryos were injected as previously described. wnt4 mRNA was co-injected with beta-galactosidase mRNA to act as a lineage tracer [blue (A); red (B-E) staining on the injected side]. At stage 41, 3G8 is ectopic (white arrows) and 4A6 is reduced on the injected side (A). At stage 36, Na+K+ATPase (B) and slc5a2 (C) expression is reduced upon wnt4 overexpression and at stage 32 odf3 (D) and nephrin (E) expression is ectopic (white arrows). Asterisk denotes injected side.|
|Fig. 3. Notch signalling regulates pronephric wnt4 expression. (A-D) X. laevis embryos were injected as previously described, with mRNA co-injected with beta-galactosidase mRNA to act as a lineage tracer (red staining on the injected side). Embryos were cultured to stage 28 when wnt4 (A,B), delta1 (C) and serrate1 (D) expression was detected by in situ hybridisation. Mis-activation of Notch signalling by injection of notch-ICD mRNA causes ectopic wnt4 expression (white arrow, A), whereas deltaSTU mRNA inhibits wnt4 expression (B). wnt4 overexpression does not have an effect on delta1 or serrate1 expression levels (B). Expression of both these genes is altered in the anteroposterior register; frequently, expression is more distal (white brackets) and rarely ectopic (white arrow, C). Asterisk denotes injected side.|
|Fig. 4. wnt4 acts downstream of the Notch pathway in the pronephros. (A-D) X. laevis embryos were injected as previously described and left to develop to stage 32, when in situ hybridisation detecting expression of nephrin and slc5a2 was performed. (A) Knockdown of wnt4 translation using a MO-inhibited nephrin expression. (B) Co-injection of notch-ICD mRNA and the wnt4 MO also reduces nephrin expression. Co-injection of su(H)DBM mRNA with wnt4 mRNA, causes ectopic glomus formation (white arrow, C) and proximal tubulogenesis (white arrow, D). Asterisk denotes injected side.|
|Fig. 5. hrt1 is induced by notch-ICD mRNA injection and promotes wnt4 expression. (A-E) X. laevis embryos were injected as previously described and left to develop to stage 28, when in situ hybridisation detecting expression of either wnt4 (A-C) or hrt1 (D,E) was performed. Embryos injected with hormone-inducible hrt1-hGR mRNA, but not incubated in dexamethasone to activate the message, have normal pronephric wnt4 expression (A). Embryos injected with hrt1-hGR mRNA but incubated in dexamethasone from stage 18 onwards, display ectopic wnt4 expression (B). MO depletion of HRT-1 inhibited wnt4 expression (C). Overexpression of notch-ICD causes a large increase in pronephric hrt1 expression (D), whereas hrt1 expression is largely unaffected by wnt4 mRNA injection (E). Asterisk denotes injected side.|
|Fig. 6. Early notch-ICD and wnt4 overexpression promotes medial pronephrogenesis at the expense of lateral pronephric cell fates. (A-C) X. laevis embryos injected with notch-ICD mRNA, cultured to stage 32, and in situ hybridised for nephrin expression, were paraplast sectioned (A). notch-ICD overexpression causes the domain of nephrin expression to extend laterally. Paraplast wax sectioning of stage 41 embryos injected with notch-ICD mRNA highlights the severity of this phenotype (B,C). Normal pronephric layout in section can be observed on the uninjected side of B. On the injected side of this embryo, a mass of non-distinct cells can be observed proximally (B), and in the distal region ectopic nephrostomal tubules and podocytes can be detected (C). Regions of interest are outlined (dashed lines). Asterisk denotes injected side. gl, glomus; cc, coelomic cavity; pt, proximal tubules; end, endoderm; nc, notochord; som, somites; nt, neural tube; ns, nephrostomes; dt, distal tubules.|
|Fig. 7. lfng and rfng are expressed in appropriate temporal and spatial regions for a role in regulation of pronephric Notch signalling. (A,B) Whole-mount in situ hybridisation was carried out with a DIG-labelled anti-sense RNA probe for lfng (A) and rfng (B). lfng expression is detected in the dorso-anterior pronephros anlagen between stages 22 and 32. Transverse section of a stage 30 embryo clearly identifies this pronephric expression. At stage 38, pronephric lfng expression is lost. rfng expression in the pronephros is detected slightly later, at stage 26, and persists to stage 32/33. By stage 36, no pronephric expression of rfng could be detected. pn, pronephros.|
|Fig. 8. rfng overexpression caused ectopic proximal pronephrogenesis. (A-I) X. laevis embryos were injected with rfng mRNA as previously described. rfng overexpression causes ectopic Lim-1 (A) and hrt1 expression (B) at stage 28. At stage 41, 3G8 and 4A6 immunostaining is ectopic; frequently the entire proximal pronephros is duplicated in the distal region (white arrow, C). Similarly, Na+K+ATPase (D), slc5a2 (E) and nephrin (F) expression is also ectopic. GATA-3, which is expressed solely in the distal tubule, is completely absent (G). Co-injection of rfng mRNA with the wnt4 MO abrogates these effects; nephrin (H) and slc5a2 (I) expression is reduced in these embryos. Asterisk denotes injected side.|
|Fig. 9. rfng promotes expression of wnt4 and components of the Notch pathway. (A-C) X. laevis embryos were injected as previously described and cultured to stage 28, with beta-galactosidase mRNA co-injected to act as a lineage tracer (red staining on injected side). Ectopic wnt4 (A), delta1 (B) and serrate1 (C) expression was detected after rfng mRNA injection. Asterisk denotes injected side. (D) A model for early stage proximal pronephrogenesis regulated by the Notch signalling pathway. We propose a pool of cells undergoing Notch signalling and located on the lateral side of the dorso-anterior pronephros anlagen at tail bud stages of development, mediate medio-lateral patterning, permitting the glomus and tubules to develop independently. Blue arrows indicate proposed genetic hierarchy; red arrows indicate directional secretion of Wnt4. See text for full description.|
|Fig. S1. Targeting su(H)DBM mRNA to the pronephros causes the same pronephric phenotypes as deltaSTU overexpression. (A-D) X. laevis embryos were injected into a V2 blastomere with su(H)DBM mRNA, which encodes a dominant-negative form of Su(H). Embryos were left to develop to various stages of development, when in situ hybridisation detecting expression of Na+K+ATPase, slc5a2, nephrin and odf3 was performed. su(H)DBM overexpression inhibited expression of the proximal domain of Na+K+ATPase (A) and completely inhibited expression of slc5a2 (B), nephrin (C) and odf3 (D). These results are identical to the effect deltaSTU overexpression has on pronephrogenesis. Asterisk denotes injected side.|
|Fig. S2. Hormone-inducible Notch constructs produce the same pronephric phenotypes as non-inducible Notch constructs.X. laevis embryos were injected into a V2 blastomere with hormone-inducible notch-ICD-hGR mRNA and su(H)DBM-hGR mRNA, left to develop to stage 18 then incubated in culture medium containing dexamethasone and DMSO (induced), or just DMSO alone (non-induced). Embryos were then left to develop to stage 34 when whole-mount in situ hybridisation for nephrin and slc5a2 was performed. (A, B) Embryos injected with notch-ICD-hGR mRNA that was activated at stage 18 showed a reduction in slc5a2 expression (A), but in non-induced embryos, slc5a2 expression was not affected (B). (C, D) Similarly embryos injected with su(H)DBM-hGR mRNA that was activated at stage 18 had reduced slc5a2 expression (C), but when not incubated in dexamethasone, these embryos displayed no pronephric phenotype (D). (E, F) Embryos injected with notch-ICD-hGR mRNA that was subsequently activated at stage 18 had ectopic nephrin expression (E), but when not activated no glomus abnormalities were observed (F). When embryos injected with su(H)DBM-hGR mRNA had their message activated at stage 18, nephrin expression was reduced. (G) Non-activated su(H)DBM-hGR mRNA had no effect on nephrin expression. Inducible Notch constructs produce similar phenotypes to non-inducible constructs and therefore the non-inducible constructs used in this work have specific effects on early pronephrogenesis. Asterisk denotes injected side.|
|Fig. S3. Wnt-4 overexpression promotes glomus formation and proximal tubulogenesis when co-injected with DeltaSTU.X. laevis embryos were injected into a V2 blastomere with a combination of wnt4 mRNA and deltaSTU mRNA, then left to develop to stage 34 where whole mount in situ hybridisation for nephrin and slc5a2 expression was performed. wnt4 overexpression with concomitant inhibition of Notch signalling caused ectopic nephrin (A) highlighting that Wnt-4 acts downstream of Notch signalling in the proximal pronephros. In addition, slc5a2 expression is not significantly reduced by co-injection of wnt4 deltaSTU and frequently the expression domain of slc5a2 is more distal (see dorsal view of pictured embryo). Asterisk denotes injected side.|
|Fig. S5. lfng misexpression inhibits pronephros formation, knockdown of rfng translation has no effect on pronephrogenesis.X. laevis embryos were injected at the eight-cell stage into a ventro-vegetal blastomere to target the presumptive pronephric region, with beta-galactosidase mRNA to act as a lineage tracer (blue staining on injected side). (A-D) Embryos were cultured to stage 41, when they were fixed, stained for the lineage label and correctly targeted embryos were whole-mount antibody stained for epitopes in the nephrostomes and proximal tubules, detected by 3G8 (purple stain, white arrow), and intermediate and distal tubules, detected by 4A6 (red stain, dashed white arrow). Injection of a control MO in addition to beta-galactosidase mRNA had no effect on pronephros development (A and E). Injection of lfng mRNA (B), lfng MO1 (C) and lfng MO2 (D) inhibited pronephrogenesis, as observed by reduced 3G8 and 4A6 staining. Additionally, we have observed the effects of lfng mis-expression on slc5a2, nephrin and Na+K+ATPase expression and found all domains of the pronephros did not form (data not shown). Injection of the rfng MO had no effect on pronephrogenesis, 3G8 and 4A6 staining was normal on the injected side. We also did not detect any effect of Rfng depletion on slc5a2, nephrin and Na+K+ATPase expression, showing all pronephric compartments formed (data not shown). Asterisk denotes injected side.|
|Fig. S6. Xbrachyury (t) expression is perturbed by lfng mis-expression. Since embryos injected with lfng mRNA or MOs had gross developmental phenotypes (such as a curled tail and shortness in length), we hypothesised that lfng mis-expression had earlier effects on development that affected pronephrogenesis indirectly. To determine this, X. laevis embryos were injected into both cells of a two-cell-stage embryo in the equatorial region with beta-galactosidase mRNA to act as a lineage tracer (red staining). Embryos were left to develop to stage 10-11, fixed, stained for the lineage label and whole-mount in situ hybridised for expression of a marker of the early mesoderm, Xbrachyury (Smith et al., 1991). (A) Injection of a control MO had no effect on Xbrachyury expression. (B-D) Overexpression of lfng (B) and depletion of lfng by injection of lfng MO1 (C) or lfng MO2 (D) statistically significantly reduced Xbrachury expression (P>0.05). These results confirm that Lfng mis-expression has early effects on mesodermal development that could be the cause of inhibited pronephrogenesis.|
|Fig. S7. lfng mis-expression perturbs myogenesis and early pronephros development.X. laevis embryos were injected at the eight-cell stage into a ventro-vegetal blastomere as previously described, with β-gal mRNA to act as a lineage tracer (red staining on injected side). Embryos were cultured to stage 24 when whole-mount double in situ hybridisation for expression of the gene encoding myosin heavy chain (MHC, a terminally differentiated muscle marker, white arrow) and lim1 (a marker of the early pronephros anlagen, yellow arrow) was detected. (A) Injection of a control MO had no effect on either lim1 or MHC expression. (B-D) Injection of lfng mRNA and depletion of lfng translation by injection of either lfng MO1 or lfng MO2 completely inhibited lim1 expression and significantly reduced MHC expression on the injected side. As the myotome is required to signal to the intermediate mesoderm to induce pronephrogenesis (Mauch et al., 2000; Mitchell et al., 2007; Seufert et al., 1999), perturbed myogenesis would be a cause of inhibited pronephrogenesis. Taken together with an early inhibition of Xbrachyury expression, we believe that lfng mis-expression causes gross developmental defects that indirectly prevent pronephrogenesis. Consequently, we are unable to confirm a role for Lfng in pronephros development. Asterisk denotes injected side. It is unsurprising lfng mis-expression affects development more severely than rfng mis-expression as it has been shown to have a much higher catalytic activity (Rampal et al., 2005) and it is likely to be the major Fringe protein during development; Rfng has been suggested to have no role in mouse development (Visan et al., 2006; Zhang et al., 2002).|
|Fig. S8. rfng overexpression can rescue the effect of Lfng depletion in the pronephros. We previously showed rfng translation is not inhibited by either lfng MO1 or lfng MO2 (see Fig. S1E in the supplementary material). Given the previously characterised functional homology between vertebrate Fringe proteins, we co-injected rfng mRNA with lfng MO1 and lfng MO2 and attempted to rescue expression of the proximal tubule marker, slc5a2. X. laevis embryos were injected at the eight-cell stage into a ventro-vegetal blastomere to target the presumptive pronephric region, with beta-galactosidase mRNA to act as a lineage tracer (red staining on injected side). (A-G) Embryos were cultured to stage 34 when whole-mount in situ hybridisation for expression of slc5a2 was detected. Injection of a control MO had no effect on proximal tubulogenesis (A). rfng overexpression, as reported in Fig. 6C, caused ectopic expression of slc5a2 in the distal pronephric region (white arrow, B). Injection of either lfng MO1 (C) or lfng MO2 (D) completely inhibits slc5a2 expression; these embryos have gross developmental defects, highlighted by curling and perturbed growth of the tail region (black arrows). Upon co-injection of rfng mRNA with lfng MO1 (E) or lfng MO2 (F), slc5a2 expression is rescued, and in some cases over-rescue is achieved with ectopic slc5a2 expression being observed (white arrows). Overall, gross developmental defects detected after single injections of lfng MO1 or lfng MO2 were also rescued, the embryos look healthy and morphologically similar to controls. A histogram illustrates rescue of Lfng depletion by rfng overexpression in the pronephros, with much fewer embryos having reduced slc5a2 expression after co-injection of lfng MO1 and lfng MO2 with rfng mRNA than after single injections of lfng MO1 or lfng MO2 (G). We propose that Rfng is able to rescue the effects of Lfng depletion. This conclusion suggests functional homology between Rfng and Lfng exists. Asterisk denotes injected side. In addition, we have also rescued the inhibitory effects lfng MO1 and lfng MO2 had on nephrin expression with rfng mRNA and also the effect lfng MO2 had on slc5a2 and nephrin expression with lfng mRNA.|