August 15, 2008;
Mix.1/2-dependent control of FGF availability during gastrulation is essential for pronephros development in Xenopus.
Although FGFs are known to affect mesoderm
patterning, their influence on intermediate mesoderm
specification during gastrulation is ignored. Here, we show that pronephros
precursors are exposed to FGF, but a strict control of FGF signals is necessary to allow pronephros
development. We provide evidence that this control is mediated by the paired-like homeobox genes Mix.1 and Mix.2. Morpholino-based Mix.1/2 knockdown, or repression of Mix.1 target genes with an enRMix.1 construct, causes an expansion of FGF4
expression in the lateral
marginal zone at gastrula
stage, together with an inhibition of pronephros
development at neurula
stages. Expression of the nephrogenic mesoderm
markers Xlim-1 and XPax-8 can be rescued in Mix.1/2 morphants by intrablastocoelic injections of the FGFR inhibitor SU5402 at mid-gastrula
stage, showing that inhibition of pronephros
development results from an increase of FGF signalling. We further show that Mix.1 overexpression results in the down-regulation of FGF3
, 4, 8 and XmyoD
, in addition to Xbra
. However, cells overexpressing Mix.1 can normally populate somites
, indicating that Mix.1 does not affect their fate cell autonomously. These data support the idea that Mix.1/2 regulates levels and/or duration of FGF signals to which pronephros
precursors are exposed during gastrulation.
[+] show captions
Fig. 1. Pronephros development is sensitive to FGF overexpression, although pronephros precursors express Xbra. (A) Analysis of GFP distribution in transgenic Xbra-GFP embryos. Transverse sections of embryos at the late gastrula stage (NF st. 12.5), mid-neurula stage (NF st. 16) and tailbud stage (NF st. 25). GFP remains detectable long after the transgene has been switched off, i.e., by the end of gastrulation, allowing the identification of Xbra-expressing cells progeny. GFP is detected in the dorsal mesodermal structures, but is absent ventrally. There is no sharp limit between these two domains. Rather, cells expressing low levels of GFP are intercalated with negative cells in the lateral mesoderm. At the late gastrula stage and the mid-neurula stage, GFP is expressed in paraxial (pm) and laterodorsal mesoderm. At the tailbud stage, GFP is strongly detected in somites (s) and pronephros anlage (pa), but is absent from ventral blood islands (vbi). (n) notochord, (nt) neural tube. (B) In situ hybridization analysis of mesodermal markers in response to exogenous FGF4. Analyses of XmyoD, XPax-8 and Xlim-1 expression at the neurula stage, and SCL at the tailbud stage. FGF4 expression was targeted on the right side of the embryos (XmyoD, XPax-8, Xlim-1) or ventrally (SCL). In some cases, β-galactosidase was co-expressed with FGF4 and revealed with Red Gal. XmyoD expression is slightly expanded on the injected side. SCL expression is inhibited in response to exogenous FGF4. XPax-8 and Xlim-1 expression is totally abolished by FGF4 overexpression. (C) RT-PCR analysis of pronephros markers in combined marginal zone explants. Expression of XWT1 and XSMP30 is induced in VMZ explants combined with DMZ explants in response to signals from Spemann's organizer. FGF4 overexpression in VMZ causes a strong inhibition of XWT1 and XSMP30 expression. (D) Expression of a constitutive FGFR (t-R1) impedes mesodermal cells to adopt a pronephric fate. Small aggregates of control cells stained with RLDx and cells expressing t-R1 and GFP were implanted in the dorsolateral marginal zone of unlabelled recipient embryos, as described in Materials and methods. Distribution of fluorescent cells was analyzed at early tadpole stage. Both kinds of cells are able to differentiate into well elongated somitic cells. However, only control cells populate the pronephros.
Fig. 2. Expression of Mix.1 in the marginal zone. (A) Whole mount in situ analysis of Mix.1 expression during gastrulation. Embryos are viewed from the vegetal pole. Dorsal side is up. Mix.1 is expressed in the lateral marginal zone until the mid-gastrula stage, and remains detectable in the ventral marginal zone until completion of gastrulation. (B) Comparison of Mix.1 and Xbra expression in the lateral marginal zone. Examples showing the two halves of a bisected embryo respectively hybridized with Mix.1 or Xbra probes. Bisection was made according to the dotted lines shown in panel A. At the beginning of gastrulation, Mix.1 and Xbra expression domains largely overlap. Overlapping area is smaller at the mid-gastrula stage.
Fig. 4. Effect of Mix.1 and Mix.2 loss of functions upon expression of nephrogenic mesoderm markers. (A) In situ hybridization analyses of XPax-8 or Xlim-1 expression in response to morpholino injection, or repression of Mix.1 target genes with enRMix.1. Morpholinos or enRMix.1 expression was targeted to the right side of the embryo. β-galactosidase was co-expressed and revealed with Red Gal. Mo1, Mo3 and enRMix.1 all cause a severe reduction of XPax-8 and Xlim-1 expression, showing that Mix.1 is required for the establishment of nephrogenic mesoderm. Neural expression of XPax8 in the otic placode (arrowhead) is also affected. In contrast, Mo2 has only a very limited effect on the expression of these markers, showing that Mix.2 alone is dispensable. (B) In situ hybridization analyses of mesodermal markers expression in response to Mo3 injection. Mo3 was targeted to the right side of the embryo (XmyoD, XPax-8, XOsr2) or ventrally (SCL). Although XPax-8 expression is inhibited, XmyoD expression is only slightly expanded on the injected side. XOsr2 expression is also inhibited in lateral mesoderm. In a minority of cases, SCL expression is reduced.
Fig. 5. Effect of Mix.1 and Mix.2 loss of functions upon later development of the pronephros. In situ hybridization analyses of XPax-2 (A–D) and XWT1 (E–H) in Mo1 or Mo3 injected embryos. Morpholinos were targeted on the right side as described in Fig. 4. On control sides, XPax-2 is expressed in nephrostome, tubules and Wolffian duct. Its expression is totally abolished in response to Mo1 or Mo3. Neural expression of XPax-2 in the hindbrain, optic and otic vesicles is also affected on the injected side. XWT1 is expressed in the glomus part of the pronephros on control sides, while it is totally down-regulated on the injected side both with Mo1 and Mo3. (I) Histological analysis at early tadpole stage (NF st. 38) after injection of Mo3 on the right side. Pronephros tubules differentiated on the control side (pt), are absent on the injected side. They are replaced by a mass of uncharacterized tissue (arrow). n: notochord; nt: neural tube; s: somites.
Fig. 6. Inhibition of pronephros development does not result from an effect of Mo3 on the Spemann's organizer. (A) Mo3 does not affect the ability of DMZ explants to induce pronephros markers in VMZ explants. DMZ explants were dissected at the early gastrula stage from embryos previously injected at the 4-cell stage with Cmo and RLDx, or Mo3 and RLDx, in the two dorsal blastomeres. Explanted DMZ were combined with VMZ explants taken from uninjected embryos and combinates were cultured until tailbud stage (NF st. 33/34). In situ hybridization analysis of XSMP30 and XWT1 expression in combinates (left panels) and corresponding RLDx fluorescence (right panels). Induction of XSMP30 and XWT1 expression in the unlabelled VMZ is not affected by Mo3. Undissected DMZ donor embryos injected with Mo3 and cultured until the tailbud stage show head anomalies when compared with Cmo-injected embryos. (B) Mo3 targeted outside of the organizer still inhibits pronephric expression of XPax-8 and XPax-2. Mo3 and RLDx were injected at the 8-cell stage into one ventro-vegetal blastomere. Embryos were cultured until neurula or tailbud stage for the analysis of XPax-8 or XPax-2 expression, respectively. In situ hybridization analysis of XPax-8 and XPax-2 expression (left and up panels, respectively) and corresponding RLDx fluorescence (right and down panels, respectively). Comparison of control and injected sides shows that Mo3 inhibits XPax-8 expression in the mesoderm at the neurula stage. Expression of XPax-2 in the lateral mesoderm is strongly reduced at tailbud stage showing that pronephric development is impaired. However, XPax-8 expression in the otic placode (arrowhead), as well as neural expression of XPax-2 in the hindbrain, optic and otic vesicles are identical on injected and control sides.
Fig. 7. Effect of Mix.1 and Mix.2 loss of functions upon FGF4, FGF8 and Xbra expression during gastrulation. In situ hybridization analyses on mid-gastrula stage embryos (NF st. 10 further cultured 3 h at 25 °C) bisected transversally as described in Fig. 2, except for the expression of Xbra in enRMix.1-injected embryo showing a whole embryo at late gastrula stage (NF st. 10 further cultured 4 h at 25 °C). Cmo, Mo1, Mo3 or enRMix.1 were targeted to the right side of the embryos as in Fig. 4. Red Gal staining results from β-galactosidase co-expressed with morpholinos or enRMix.1. White arrowheads indicate injected sides. In response to Mo1, Mo3 and enRMix.1, FGF4 is strongly up-regulated on the injected side. Its expression is barely detected on the control side and requires longer color development times to be visible, as exemplified for the Cmo example shown. Both Mo1 and Mo3 cause an expansion of FGF8 expression in vegetal cells localized deeper in the vegetal mass or closer to the blastocoel floor. A similar expansion of Xbra expression is visible, although more limited than that of FGF4 and FGF8. Expansion of Xbra expression is more evident at late gastrula stage.
Fig. 8. Inhibition of FGF signalling rescues expression of nephrogenic mesoderm markers in Mo1 and Mo3-injected embryos. (A) Design of the experiment. Morpholinos were injected in the two right blastomeres at the 4-cell stage. At the mid-gastrula stage, a solution of the FGF receptor inhibitor SU5402 was injected into the blastocoel. Embryos were further cultured until the neurula stage, and processed for in situ hybridization analyses of XPax-8 and Xlim-1 expression. (B) Examples of rescue. Left column corresponds to morphant embryos that received intrablastocoelic injection of the solvent alone. These embryos display a strong inhibition of XPax-8 and Xlim-1 expression by Mo3. Right column shows rescue of XPax-8 and Xlim-1 by intrablastocoelic injection of the SU5402 solution. (C) Percentage of embryos with normal levels of XPax-8 or Xlim-1 expression on both sides relative to the total number of analyzed embryos (n) scored in three independent experiments.
Fig. 9. Perturbation of FGF signals during gastrulation by Mix.1 overexpression. High dose of Mix.1 mRNA (100 pg) was injected into the two right blastomeres at the 4-cell stage, and embryos cultured until the mid-gastrula stage for in situ hybridization analysis of FGF-encoding genes (FGF3, 4, 8) or FGF-target genes (Xbra and XmyoD). Expression of all analyzed genes is inhibited indicating that Mix.1 overexpression globally interfered with FGF signals in the marginal zone. At a lower dose that does not block gastrulation (25 pg), Xbra is still down-regulated. Embryos of the same series of injection fixed at later stages express XPax-8 at the neurula stage and XSMP30 at the tailbud stage, showing that pronephros development has not been affected.
osr2 (odd-skipped related 2) gene expression in Xenopus tropicalis embryo via in situ hybridization, NF stage 17, dorsal view, anterior up.