SEARCH

SEARCH BY CITATION

Keywords:

  • mouse mutant;
  • Irx genes;
  • neural tube;
  • patterning;
  • floor plate

Abstract

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

Sonic hedgehog (Shh), produced by the notochord and floor plate cells of the neural tube, plays a critical role in organizing dorsal–ventral patterning in the developing neural tube. We have investigated neural tube development in mouse embryos homozygous for the Fused toes (Ft) mutation, a deletion composed of genes of the Iroquois B (IrxB) cluster and of Fts, Ftm, and Fto. In Ft mutants starting from embryonic day 10.5, the floor plate appeared to degenerate and the notochord failed to undergo ventral displacement from the spinal cord. Consistent with the loss of Shh signalling from the floor plate, V3 neuron generation was reduced in Ft/Ft embryos and the domain of motor neuron generation expanded ventrally at the expense of V2 neurons. These data support the idea that Ft genes play an important role in dorsal–ventral patterning of the neural tube acting to define the extent of motor neuron generation; moreover, the data reveal a previously unanticipated function for Ft genes in the maintenance of the floor plate. Developmental Dynamics 233:623–630, 2005. © 2005 Wiley-Liss, Inc.


INTRODUCTION

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

Development of the vertebrate nervous system is one of the most complex processes of embryogenesis. Due to its simple anatomy, the spinal cord has emerged as a model to study vertebrate neurogenesis (Jessel, 2000). In the ventral half of the spinal cord, five distinct neuronal subtypes are generated at characteristic dorsal–ventral positions; these are motor neurons (MN) and V3, V2, V1, and V0 interneurons, respectively (Briscoe and Ericson, 2001). The generation of these cell types is dependent on the secreted signalling molecule Sonic hedgehog (Shh; Chiang et al., 1996). Initially, Shh is produced by axial mesodermal midline cells of the notochord. Shh signaling from the notochord induces development of the floor plate—ventral midline cells of the neural tube—which also expresses Shh together with several other genes, including the winged helix transcription factor FoxA2 (Dodd et al., 1998). The different progenitor domains are defined by the combinatorial expression of subsets of transcription factors. These can be subdivided into two classes based on their mode of regulation by Shh signaling. Class I genes are repressed by Shh while class II genes require Shh signaling for their expression. Moreover, cross-repressive interactions between class I and class II proteins appear to refine and maintain the distinct progenitor domains (Briscoe et al., 2000). The class I group includes Pax7, Dbx1, Irx3, Pax6, and Dbx2, whereas Nkx2.2, Nkx6.1, and Olig2 belong to the class II proteins. A series of gain- and loss-of-function experiments have provided evidence of the cross-repressive interactions between Pax6 and Nkx2.2, and Nkx6.1 and Dbx2, respectively (Briscoe et al., 1999, 2000; Sander et al., 2000; Wijgerde et al., 2002; Vallstedt et al., 2001; Jacob and Briscoe, 2003). The expression of Nkx2.2 defines the ventral limit of MN generation: dorsal expansion of Nkx2.2 results in the production of V3 neurons at the expense of MNs (Ericson et al., 1997), whereas loss of Nkx2.2 results in the ventral expansion of MN generation. Conversely, the dorsal limit of MN generation appears to be determined by Irx3 as ventral misexpression of Irx3 represses Olig2 expression (Novitch et al., 2002), inducing V2 neurons at the expense of motor neurons (Briscoe et al., 2000). However, whether Irx3 or other Irx genes are necessary to determine the dorsal limit of MN generation remains unclear.

In Drosophila the Iroquois complex (IRO-C) consists of three homeobox genes, araucan (ara), caupolican (caup), and mirror (mirr), which are involved in development of the peripheral nervous system and the regionalization of the eye and wing imaginal discs (Gomez-Skarmeta and Modollel, 1996; McNeill et al., 1997; Grillenzoni et al., 1998). Ara and caup are suggested to function as positive regulators of the proneural genes achaete and scute, which encode for basic helix–loop–helix (bHLH) transcription factors (Gomez-Skarmeta and Modollel, 1996). Studies in Xenopus indicate that Iroquois (Irx) genes are expressed in a broad domain in the neural plate before the activation of the proneural gene XASH3 (Xenopus Achaete Scute Homologue; Gomez-Skarmeta et al., 1998). The mouse homologue Mash1 is expressed in the progenitor population of the dorsal neural tube in an overlapping domain with Irx expression.

The Fused toes (Ft) mouse mutation is characterized by a 1.6 Mb deletion on chromosome 8, which affects six genes, including the genes of the Iroquois B cluster, Irx3, Irx5, and Irx6 (Peters et al., 2002) and three other genes Fts, Fto, and Ftm. Embryos homozygous for the Ft mutation are retarded and die between embryonic day (E) 10.5 and E14.5, probably due to a malformation of the heart. Furthermore, they show severe abnormalities of craniofacial and forebrain structures, polydactyly of the forelimbs, and syndactyly of fore- and hindlimbs (van der Hoeven et al., 1994; Grotewold and Rüther, 2002). In addition, embryonic turning and heart looping are randomized (Heymer and Rüther, 1999), which suggests a defect in the midline and as a consequent disturbance in neural tube development (Izraeli et al., 1999; Murcia et al., 2000; Przemeck et al., 2003). Based on the described function of Irx genes in neural tube specification and the observed phenotypes in Ft/Ft embryos, we asked whether neural tube development was affected in Ft mutant mice and here we report this analysis.

RESULTS

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

Floor Plate Maintenance Is Disrupted in Ft/Ft Embryos

The mammalian genome contains six members of the Irx gene family organized in two clusters (IrxA and IrxB; Peters et al., 2000). Expression analyses have shown that all Irx genes except Irx4 are expressed in distinct patterns in the neural tube, and additionally, Irx1/Irx2 and Irx3/Irx5 are expressed in the notochord (Houweling et al., 2001). Furthermore, for Irx3, a member of the IrxB cluster, a function in ventral neural tube specification has been suggested by its ability to repress MN generation (Briscoe et al., 2000). In Ft mice, the entire IrxB cluster (Irx3, 5, and 6) is deleted (Peters et al., 2002). Besides the genes of the IrxB cluster and Fts, which is also expressed in the neural tube, two other genes, Ftm and Fto are deleted in Ft mutant mice as well. We, therefore, analyzed patterning of the neural tube in Ft/Ft embryos.

First we analyzed Shh, which is expressed by the notochord and floor plate and controls the generation of distinct classes of ventral neurons (Ericson et al., 1997; Briscoe et al., 1999, 2000; Pierani et al., 1999). In situ hybridizations for Shh at forelimb levels indicated that, although Shh RNA expression was present in the floor plate and notochord of E10.5 embryos, by E11.5, Ft/Ft embryos lacked Shh expression in the floor plate (Fig. 1A–D). Whole-mount in situ hybridizations at E11.5 embryos confirmed that Shh expression was absent from the majority of the spinal cord; however, expression of Shh was detectable in the most posterior and anterior parts of the neural tube. Although in situ hybridizations indicated that Shh was induced in the floor plate of Ft/Ft embryos, antibody staining suggested that defects in Shh production were already apparent at E10.5: in Ft/Ft embryos, Shh protein was down-regulated or absent in some regions of the spinal cord (Fig. 1F).

thumbnail image

Figure 1. Shh expression in the floor plate disappears between embryonic day (E) 10.5 and E11.5. AL: Shown are in situ hybridizations (A–D) and immunohistochemistry (E–L) on transverse sections at forelimb levels of E10.5 and E11.5 embryos. A,B: At the RNA level, Shh expression is detectable in notochord and floor plate of both E10.5 wild-type and Ft/Ft embryos. D: At E11.5, Ft/Ft embryos lack expression of Shh in the floor plate. The expression in the notochord remains normal, but the notochord fails to displace from the neural tube in Ft/Ft embryos (compare positions of notochord in C and D). F,H: At the protein level, Shh is not detectable in the floor plate of Ft/Ft embryos at both stages. J,L: Fox2A is still present at E10.5 (J), but not longer at E11.5 (L). E,F,I: The position of the notochord is marked by white arrows. E–H,J,L: Analysis of Nkx2.2, which marks V3 progenitor cells, shows that expression is greatly reduced to the ventral midline in Ft/Ft embryos (compare E–H), whereas Irx3 expression is not found in Ft/Ft embryos, as expected (J,L).

Download figure to PowerPoint

To determine whether the defect in Shh expression by floor plate cells of Ft/Ft embryos represented a more general deficit in floor plate development, we examined the expression of the floor plate markers FoxA2, Bmp6, and Lmx1b. FoxA2 was detectable in ventral midline cells of Ft/Ft embryos at E10.5 (Fig. 1J). However, by E11.5, expression of FoxA2, Bmp6, and Lmx1b was absent in mutant embryos (Figs. 1L, 2B,D).

thumbnail image

Figure 2. Dorsal cell fate is unaffected in Ft/Ft embryos. Shown are in situ hybridizations on sections of embryonic day (E) 11.5 wild-type (+/+) and Ft/Ft embryos. AH: The dorsal expression domains of Lmx1b (A,B), Bmp6 (C,D), Lh2B (E,F), and Wnt3a (G,H) are unaltered in Ft/Ft embryos, whereas the ventral expression domains of Lmx1b (A,B) and Bmp6 (C,D) in the floor plate are not detectable in Ft/Ft embryos. IL: Analyses of the expression of Pax2 (I,J) and Pax6 (K,L) show an expansion to the ventral side for both (bars in K,L mark the distance between the floor plate and expression of Pax6).

Download figure to PowerPoint

In contrast to the loss of Shh expression in the floor plate, Shh expression appeared unaffected in the notochord. However, the notochord stayed located close to the neural tube, rather than undergoing the characteristic ventral displacement (Fig. 1C,D). Together, these data indicate that there is a general defect in floor plate maintenance in Ft embryos. Although floor plate induction is initiated, its maintenance or maturation is affected, resulting in the degeneration of the floor plate and ventral midline defects.

Dorsal Neural Tube Patterning Is Unaffected by the Ft Mutation

The effect on floor plate development in Ft/Ft embryos raised the possibility that dorsal–ventral patterning of the neural tube might be disturbed. To test this possibility, we first examined markers for dorsal progenitor domains and differentiated neurons. Pax7, a class 1 gene, is expressed in the dorsal half of the neural tube (Briscoe et al., 2000; Wijgerde et al., 2002) and was unaffected in Ft/Ft embryos (compare Fig. 3A–D). Moreover, no alteration of Lh2b expression in dI1 neurons was detected. Wnt3a and Bmp6 expression in the roof plate was normal, as was expression of Lmx1b in the dorsal half of the neural tube (see Fig. 2). The Pax family member Pax2 is expressed in several types of interneurons in the dorsal and intermediate neural tube, including dI4, dI6, V0, and V1neurons (Burill et al., 1997; Helms and Johnson, 2003). In E11.5 Ft/Ft embryos, Pax2 expression was maintained in the dorsal neural tube, but we found a ventral expansion of its expression domain (Fig. 2I, J). Together, these data provide evidence that dorsal neural tube patterning is unaffected in Ft/Ft embryos.

thumbnail image

Figure 3. Ft/Ft embryos show alterations in ventral cell fate and an expansion of the motor neurons domain. AH: Immunohistochemistry was performed on transverse sections of embryonic day (E) 10.5 and E11.5 wild-type (+/+) and Ft/Ft embryos. B,D: The dorsal border of the Nkx6.1 expression domain becomes more diffuse in Ft/Ft embryos, because some Nkx6.1-expressing cells were found dorsal to the normal limit of Nkx6.1 expression. A–H: Pax7 (A–D) seems to be mostly unaltered, while the motor neuron progenitor domain marked by Olig2 expression (E,F) is found to be expanded in Ft/Ft embryos at E10.5, as is the expression of HB9 at E11.5 (G,H), which marks the differentiated motor neurons. G,H: The V2 domain marked by Chx10 expression is reduced at E11.5. E,F: The ventral Mash1expression domain is found to be slightly reduced (brackets).

Download figure to PowerPoint

Motor Neuron Generation Expands Ventrally in Ft/Ft Mutants

We next turned our attention to the ventral neural tube and examined the V3 progenitor domain defined by expression of Nkx2.2. Because Nkx2.2 is activated by high levels of Shh (Briscoe et al., 1999), the loss of Shh expression in the floor plate would be expected to result in diminishing of Nkx2.2 expression. Consistent with this expectation, in E11.5 Ft/Ft embryos, Nkx2.2 expression was reduced or absent in the ventral neural tube (Fig. 1H). At E10.5, Nkx2.2 expression was still present but occupied a smaller domain of cells that included the ventral midline adjacent to the notochord (Fig. 1F). The ventral boundary of the Pax6 expression domain normally complements the dorsal boundary of the Nkx2.2 expression domain. In Ft/Ft embryos, a slightly ventral expansion of Pax6 expression was observed (see bars, Fig. 2K,L).

Nkx2.2 is required for the generation of V3 neurons and represses the generation of MNs (Briscoe et al., 1999). We therefore examined V3 neuron and MN generation in Ft/Ft embryos. Expression of Olig2, a bHLH gene expressed in MN progenitors of the ventral neural tube, was expanded toward the ventral midline in Ft/Ft embryos (Fig. 3F). Consistent with this finding, analysis of the MN markers HB9 (Fig. 3H) and Isl1/2 (not shown) indicated a ventral expansion of MN generation. Moreover, the production of V3 neurons was greatly reduced, in situ hybridization analysis for the V3 marker Sim1 showed no detectable signal in E11.5 Ft/Ft embryos (data not shown).

Motor Neuron Generation Expands Dorsally in Ft/Ft Embryos

Previous studies have suggested that Irx3 limits the dorsal extent of MN generation. The ventral boundary of the Irx3 expression domain shares a common border with the dorsal expression domain of Olig2. Furthermore, Irx3 and Olig2 cross-repress one another's expression to define the pV2/pMN boundary, and ectopic expression of Irx3 results in a reduction in MN generation (Briscoe et al., 2002). To analyze whether Irx3 or other IrxB genes are necessary to define the dorsal boundary of MN generation, we first examined Nkx6.1. Nkx6.1 is expressed in ventral progenitors of V3, V2, and MN neurons (Qiu et al., 1998; Briscoe et al., 1999) and is important for the generation of MNs (Sander et al., 2000). In Ft/Ft embryos, expression of Nkx6.1 is maintained in the ventral regions of the neural tube. However, the sharply delimited dorsal boundary of Nkx6.1, normally seen in wild-type embryos was lost, and some Nkx6.1-expressing cells were observed in a more dorsal position, intermingled with more other progenitor cells (Fig. 3B,D). Comparison with littermate controls suggested that Olig2 expression expanded into the region that normally generates V2 neurons (Fig. 3F). Moreover, a new sharp dorsal boundary of Olig2 expression appeared to form. To confirm the dorsal expansion of Olig2, we analyzed the expression of the mammalian acheate scute homologue Mash1. Mash1 is broadly expressed throughout dorsal progenitor cells, whereas in ventral progenitor cells, scattered expression was found in the V2 progenitor domain (Gowan et al., 2001). In Ft/Ft embryos, the dorsal part of the Mash1 domain appeared unaltered; however, the ventral domain of expression was reduced (Figs. 3E,F, 4A,B, brackets). Analysis of two other proneural genes, Math1 and Ngn1, showed no obvious changes in expression (data not shown).

thumbnail image

Figure 4. Although V2 neurons are reduced; pV1 and pV0 domains are not affected. A–F: Immunohistochemistry (A,B) and in situ hybridizations (C–F) were performed on transverse sections of the neural tube of E11.5 embryos. A,B: The V2 domain, marked by Chx10 is reduced in Ft/Ft embryos, as is the ventral expression of Mash1. C,D: Expression of Dbx1, which marks the p0 progenitor domain, was unaffected in Ft/Ft embryos. E,F: Expression of Dbx2, which marks the pV0 and pV1 domain was also found unaffected.

Download figure to PowerPoint

Consistent with the dorsal expansion of Olig2, analysis of the HB9 expression indicated a dorsal expansion in the generation of MNs in Ft/Ft embryos (Fig. 3G,H). Moreover, staining with anti-Chx10 antibodies, a marker of V2 neurons, indicated a marked reduction in V2 neuron generation concomitant with the expansion of MN production. In addition to the reduction of the Chx10 domain, we found a small number of ectopic Chx10-expressing cells in the dorsal neural tube (Figs. 3H, 4B), which may be a consequence of the dorsal expansion of Nkx6.1.

The ventral limit of Dbx1 expression defines the pV0/pV1 boundary (Pierani et al., 1999). For this marker, we found no alteration in Ft/Ft embryos (Fig. 4C,D). Furthermore, this finding was confirmed by the unaltered expression of Dbx2 (Fig. 4E,F), which defines the pV2/pV1 boundary (Briscoe et al., 2000). Thus, the expansion of the pMN and MN domains occurs concurrently to the reduction or the loss of the adjacent pV2/pV3 domains, but did not appear to affect the pV0/pV1 domain.

Irx1 Expression Is Expanded in Ft/Ft Embryos

The dorsal expansion of Olig2 expression, resulting in a new sharply defined boundary, raised the possibility that other genes are able to repress Olig2 expression in Ft/Ft mice. Previous expression analyses of Irx genes showed partially identical and overlapping expression patterns. In the developing neural tube, the expression domains of Irx1/Irx2 and Irx3/Irx5 are almost identical (Houweling et al., 2001). Irx3 and Irx5 are widely expressed in the progenitor cells with ventral limits at the pV2/pMN boundary, while Irx1/Irx2 have distinct expression domains in the intermediate region of the neural tube and an overall low expression along the rest of the dorsal–ventral axis. Moreover, comparative sequence analysis have shown a high degree of conservation between paralogous Irx genes, namely Irx1/Irx3, Irx2/Irx5, and Irx4/Irx6 (Peters et al., 2000).

In the neural tube, Iroquois expression appeared strongest for paralogous Irx1 and Irx3. Therefore, we examined whether the expression of Irx1 was altered in Ft/Ft embryos, and observed a dramatic up-regulation and ventral expansion of Irx1 expression (Fig. 5D). The same was found for expression of Irx2, which also showed a similar expanded expression pattern in Ft/Ft embryos (data not shown). Comparison of the ventral border of Irx1 expression and the dorsal border of Olig2 expression suggested a common identical border (Fig. 5D,E). Thus, in Ft/Ft embryos, Irx1 and/or Irx2 may have acquired the function of Irx3 in the regulation of Olig2.

thumbnail image

Figure 5. Irx1 and Olig2 expression domains are expanded in Ft/Ft embryos. Shown are in situ hybridizations on sections of embryonic day 11.5 wild-type (+/+) and Ft/Ft embryos. AC: In the wild-type situation, Olig2 and Irx3 repress each other and share a common border (compare B, C). DF: In Ft/Ft embryos, Irx3 expression is lost (F) and Olig2 expression is found to be expanded dorsally (E). The new dorsal border of Olig2 is now identical to the border of the Irx1 expression, which is found to be up-regulated (compare D,E).

Download figure to PowerPoint

DISCUSSION

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

Floor Plate Development in Ft/Ft Embryos

Analysis of neural tube development in Ft/Ft embryos indicated a defect in the maintenance or maturation of the floor plate. Floor plate induction appears to be normally initiated in Ft/Ft embryos, but commencing at E10.5, the expression of Shh started to decrease, and by E11.5, expression of floor plate markers was absent along the entire length of the spinal cord. Surprisingly, we found Shh RNA to be detectable longer than the Shh protein. This finding might be due to a different half-lives of RNA and protein or the mechanism that leads to the loss of Shh expression.

The generation of the floor plate in the mouse depends on Shh signalling from the notochord (Roelink et al., 1995), because mutant mice with defects in the notochord lack the floor plate (Ang and Rossant, 1994; Weinstein et al., 1994; Ma et al., 2002). In Ft/Ft embryos, the expression of Shh and FoxA2 in the notochord appears normal and consistent with this finding; floor plate induction occurs (see Results section). Moreover, many of the ventral cell types that are dependent on Shh signaling for their expression are induced in Ft/Ft embryos. Thus, the subsequent loss of floor plate identity does not appear to be due to a defect in Shh signalling from the notochord. However, in Ft/Ft embryos, the notochord fails to be become displaced ventrally from the neural tube. These aspects of the Ft/Ft phenotype are similar to those described for Gli2 mutant embryos (Ding et al., 1998). Both mutants develop a morphologically normal notochord, lack the floor plate, and the notochord fails to displace from the neural tube. Because Gli2 expression was found to be normal in Ft/Ft embryos (data not shown), the Ft mutation may interfere downstream of Gli2 or in a completely different way. Together, these data suggest that, in Ft/Ft embryos, at least one of the genes within the Ft region is required to maintain floor plate identity.

Motor Neuron Generation Expands in Ft Mutants

Previous studies have indicated that specification of motor neuron identity requires an extended period of Shh signalling (Ericson et al., 1996). Because the notochord, which is the initial source of Shh signalling undergoes a ventral displacement, the floor plate appears to be the source of Shh signalling in the second, later period. Although our analyses show a degeneration of the floor plate in Ft/Ft embryos, we found the motor neurons continued to be generated in Ft/Ft embryos. This finding is similar to the situation in Gli2 mutants, where the floor plate is lost but motor neuron development continues (Ding et al., 1998). These findings indicate that, in the absence of floor plate signalling, Shh signalling from the notochord is presumably sufficient to induce motor neuron identity. This finding may, in part, be a consequence of the lack of ventral displacement of the notochord and the proximity of notochord to the neural tube providing sufficient Shh signalling to direct MN generation.

In Ft/Ft embryos, the generation of MNs is altered. The ventral limit of the pMN marker Olig2 expanded ventrally correlating with the reduction in Nkx2.2 expression. Expression of Nkx2.2 in the pV3 domain is required to suppress motor neuron cell fate (Briscoe et al., 1999; Novitch et al., 2001) and Olig2 expression. Furthermore, Nkx2.2 is a member of the class II genes, which are positively regulated by Shh signaling. Because Nkx2.2 requires high concentrations of Shh to be activated, the decrease in Nkx2.2 expression in Ft/Ft mutants is consistent with the loss of the floor plate and the decrease of Shh expression in Ft/Ft embryos. In the absence of Nkx2.2 expression in the pV3 domain, Olig2 expression is no longer blocked and expands ventrally promoting the generation of MNs. Concomitantly, V3 neuron generation is disrupted, as assessed by the marker Sim1.

Analysis of Olig2 expression as well as markers of postmitotic MNs indicated that the MN domain also expanded dorsally in Ft/Ft embryos. In wild-type embryos, the dorsal limit of MN generation is defined by the ventral limit of Irx3 expression, and ectopic expression of Irx3 leads to a reduction of motor neurons (Briscoe et al., 2000). Thus, in Ft/Ft mice, the dorsal expansion of the Olig2 domain is most likely due to the loss of Irx3 expression. Nevertheless, the domain of Olig2 expression in Ft/Ft embryos exhibited a sharp dorsal boundary. This finding suggested that other factors are able to delimit the dorsal extent of Olig2 expression. Analyses of the expression of Irx1 and Irx2 demonstrated an up-regulation of their expression in Ft/Ft embryos. Furthermore, the ventral boundaries of Irx1 and Irx2 expression correspond to the dorsal boundary of the Olig2 expression domain, suggesting that Irx1 and/or Irx2 may compensate for the loss of Irx3 function and act to restrict the dorsal expansion of MN generation. Additionally, the up-regulation of Irx1/2 in Ft/Ft embryos suggested a regulation of Irx genes among each other.

Changes in V2 Neuron Generation in Ft/Ft Embryos

The observed reduction in V2 neuron generation in Ft/Ft embryos may be explained by the expansion of the MN domain. The V2 progenitor domain is defined by the overlapping expression domains of Nkx6.1 and Irx3 and generates Chx10-expressing V2 neurons (Briscoe et al., 2000). Concomitant with the expansion of Olig2 expression and the reduction in V2 neuron generation, we found a reduction in the ventral domain of Mash1 expression. Double staining of Mash1 together with Chx10 suggested that Mash1 is expressed in the pV2 domain. Studies from Drosophila have shown that the Drosophila Mash homologs acheate and scute are positively regulated by the Drosophila Iro-C complex (Gomez-Skarmeta and Modollel, 1996). These findings raise the possibility that a similar regulatory influence of Irx genes on the expression of Mash1 exists in vertebrates.

In Ft/Ft embryos, a small number of Nkx6.1-expressing cells were observed dorsal to the normal boundary of the Nkx6.1 expression domain. This finding may reflect a role for Ft genes in restraining Nkx6.1 expression in the dorsal neural tube. Alternatively, it is possible that Ft genes control aspects of cell affinity that restrict cell mixing so that, in Ft/Ft embryos, Nkx6.1-expressing cells are able to mix with more dorsally located progenitors.

In conclusion, we found that Ft/Ft embryos are defective in neural tube development, with regard to floor plate maintenance and the failure of notochord displacement from the neural tube. Furthermore, the generation of ventral cell types is affected; in particular, the motor neuron domain expands at the expense of V2 and V3 neurons. These findings suggest multiple roles for Ft genes in neural tube patterning. Because all affected Irx genes are expressed in the neural tube and two of them, Irx3 and Irx5, are additionally expressed in the notochord, they might be potential candidates for regulating floor plate maintenance and ventral cell fate. We cannot exclude, however, that one of the other deleted genes Fts, Ftm, or Fto or combinations of all six genes play critical roles in neural tube specification. Individual inactivation experiments of these genes will help to investigate their involvement in the observed phenotypes of the Ft mutant.

EXPERIMENTAL PROCEDURES

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

Embryos

The Ft mutation was generated by a transgene integration using (C57Bl/6 X SJL)F2 embryos (van der Hoeven et al., 1994). Embryos from timed matings were isolated at the desired stages. Homozygous embryos were readily identified by their phenotype. Additionally, polymerase chain reaction on yolk sac DNA was carried out by using the following oligonucleotides: Ft, 5-GTCCTTTCTCCATGGGTATG-3; Wt, 5-GTGGAACCCTTCTGTACATG-3; Ft/Wt, 5-C TGAAAGGTTGTACTGAGCC-3, which allowed to discriminate between wild-type, Ft/+, and Ft/Ft embryos.

In Situ Hybridization

After dissection in ice-cold phosphate buffered saline (PBS), embryos were fixed in 4% paraformaldehyde for 2 to 4 hr, washed in PBS, and incubated in 70%, 80%, 90%, and 100% ethanol for each 2 hr. Finally, they were incubated in 1-butanol overnight and transferred in paraffin for embedding.

In situ hybridizations on paraffin sections were carried out as previously described (Moormann et al., 2001). The following probes were used: Lh2b, Lmx1b (gifts from J.C. Izpisua-Belmonte), Irx1 (gift from P. Gruss), Olig2 (gift from Johan Ericson), Shh, Bmp6, Dbx1, Dbx2, Irx3, Wnt3a, Pax2 and Pax6.

Immunohistochemistry

Immunohistochemistry stainings were performed according to Briscoe et al. (1999, 2000), using antibodies against Shh, Hnf3β, Nkx2.2, Nkx6.1, Irx3, Olig2, Mash1, Chx10, Mnr2, Hb9, Dbx1, and Pax7 (Briscoe et al., 1999, 2000).

Acknowledgements

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

We thank Thomas Theil and Renate Dildrop for critical reading of the manuscript. J.B. is supported in part by EC network grants.

REFERENCES

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