The Xenopus Irx genes are essential for neural patterning and define the border between prethalamus and thalamus through mutual antagonism with the anterior repressors Fezf and Arx

The Iroquois (Irx) genes encode homeoproteins conserved during evolution. Vertebrate genomes contain six Irx genes organized in two clusters, IrxA (which harbors Irx1, Irx2 and Irx4) and IrxB (which harbors Irx3, Irx5 and Irx6). To determine the precise role of these genes during development and their putative redundancies, we conducted a comparative expression analysis and a comprehensive loss-of-function study of all the early expressed Irx genes (Irx1-5) using specific morpholinos in Xenopus. We found that the five Irx genes display largely overlapping expression patterns and contribute to neural patterning. All Irx genes are required for proper formation of posterior forebrain, midbrain, hindbrain and, to a lesser an extent, spinal cord. Nevertheless, Irx1 and Irx3 seem to have a predominant role during regionalization of the neural plate. In addition, we find that the common anterior limit of Irx gene expression, which will correspond to the future border between the prethalamus and thalamus, is defined by mutual repression between Fezf and Irx proteins. This mutual repression is likely direct. Finally, we show that Arx, another anteriorly expressed repressor, also contribute to delineate the anterior border of Irx expression.     

Analysis of mouse kreisler mutants reveals new roles of hindbrain-derived signals in the establishment of the otic neurogenic domain

The inner ear, the sensory organ responsible for hearing and balance, contains specialized sensory and non-sensory epithelia arranged in a highly complex three-dimensional structure. To achieve this complexity, a tight coordination between morphogenesis and cell fate specification is essential during otic development. Tissues surrounding the otic primordium, and more particularly the adjacent segmented hindbrain, have been implicated in specifying structures along the anteroposterior and dorsoventral axes of the inner ear. In this work we have first characterized the generation and axial specification of the otic neurogenic domain, and second, we have investigated the effects of the mutation of kreisler/MafB - a gene transiently expressed in rhombomeres 5 and 6 of the developing hindbrain - in early otic patterning and cell specification. We show that kr/kr embryos display an expansion of the otic neurogenic domain, due to defects in otic patterning. Although many reports have pointed to the role of FGF3 in otic regionalisation, we provide evidence that FGF3 is not sufficient to govern this process. Neither Krox20 nor Fgf3 mutant embryos, characterized by a downregulation or absence of Fgf3 in r5 and r6, display ectopic neuroblasts in the otic primordium. However, Fgf3−/−Fgf10−/− double mutants show a phenotype very similar to kr/kr embryos: they present ectopic neuroblasts along the AP and DV otic axes. Finally, partial rescue of the kr/kr phenotype is obtained when Fgf3 or Fgf10 are ectopically expressed in the hindbrain of kr/kr embryos. These results highlight the importance of hindbrain-derived signals in the regulation of otic neurogenesis.     

Paralog group 1 hox genes regulate rhombomere 5/6 expression of vhnf1, a repressor of rostral hindbrain fates, in a meis-dependent manner

The vertebrate hindbrain is segmented into an array of rhombomeres (r), but it remains to be fully understood how segmentation is achieved. Here we report that reducing meis function transforms the caudal hindbrain to an r4-like fate, and we exploit this experimental state to explore how r4 versus r5-r6 segments are set aside. We demonstrate that r4 transformation of the caudal hindbrain is mediated by paralog group 1 (PG1) hox genes and can be repressed by vhnf1, a gene expressed in r5-r6. We further find that vhnf1 expression is regulated by PG1 hox genes in a meis-dependent manner. This implies that PG1 hox genes not only induce r4 fates throughout the caudal hindbrain, but also induce expression of vhnf1, which then represses r4 fates in the future r5-r6. Our results further indicate that r4 transformation of the caudal hindbrain occurs at intermediate levels of meis function, while extensive removal of meis function produces a hindbrain completely devoid of segments, suggesting that different hox-dependent processes may have distinct meis requirements. Notably, reductions in the function of another Hox cofactor, pbx, have not been reported to transform the caudal hindbrain, suggesting that Meis and Pbx proteins may also function differently in their roles as Hox cofactors.     

An ancient mechanism of hindbrain patterning has been conserved in vertebrate evolution

The hindbrain is a vertebrate-specific embryonic structure of the central nervous system formed by iterative transitory units called rhombomeres (r). Rhombomeric cells are segregated by interhombomeric boundaries which are prefigured by sharp gene expression borders. The positioning of the first molecular boundary within the hindbrain (the prospective r4/r5 boundary) responds to the expression of an Iroquois (Irx) gene in the anterior (r4) and the gene vHnf1 at the posterior (r5). However, while Irx3 is expressed anteriorly in amniotes, a novel Irx gene, iro7, acts in teleosts. To assess the evolutionary history of the genes responsible for the positioning of the r4/r5 boundary in vertebrates, we have stepped outside the gnathostomes to investigate these genes in the agnathans Lethenteron japonicum and Petromyzon marinus. We identified one representative of the Hnf1 family in agnathans. Its expression pattern recapitulates that of vHnf1 and Hnf1 in higher vertebrates. Our phylogenetic analysis places this gene basal to gnathostome Hnf1 and vHnf1 genes. We propose that the duplication of an ancestral hnf1 gene present in the common ancestor of agnathans and gnathostomes gave rise to the two genes found in gnathostomes. We have also amplified 3 Irx genes in L. japonicum: LjIrxA, LjIrxC, LjIrxD. The expression pattern of LjIrxA (the agnathan Irx1/3 ortholog) resembles those of Irx3 or iro7 in gnathostomes. We propose that an Irx/hnf1 pair already present in early vertebrates positioned the r4/r5 boundary and that gene duplications occurred in these gene families after the divergence of the agnathans.     

Hox gene colinear expression in the avian medulla oblongata is correlated with pseudorhombomeric domains

The medulla oblongata (or caudal hindbrain) is not overtly segmented, since it lacks observable interrhombomeric boundaries. However, quail-chick fate maps showed that it is formed by 5 pseudorhombomeres (r7-r11) which were empirically found to be delimited consistently at planes crossing through adjacent somites (Cambronero and Puelles, 2000). We aimed to reexamine the possible segmentation or rostrocaudal regionalisation of this brain region attending to molecular criteria. To this end, we studied the expression of Hox genes from groups 3 to 7 correlative to the differentiating nuclei of the medulla oblongata. Our results show that these genes are differentially expressed in the mature medulla oblongata, displaying instances of typical antero-posterior (3′ to 5′) Hox colinearity. The different sensory and motor columns, as well as the reticular formation, appear rostrocaudally regionalised according to spaced steps in their Hox expression pattern. The anterior limits of the respective expression domains largely fit boundaries defined between the experimental pseudorhombomeres. Therefore the medulla oblongata shows a Hox-related rostrocaudal molecular regionalisation comparable to that found among rhombomeres, and numerically consistent with the pseudorhombomere list. This suggests that medullary pseudorhombomeres share some AP patterning mechanisms with the rhombomeres present in the rostral, overtly-segmented hindbrain, irrespective of variant boundary properties.     

Foxh1 and Foxa2 are not required for formation of the midgut and hindgut definitive endoderm

The definitive endoderm forms during gastrulation and is rapidly transformed into the gut tube which is divided along the anterior-posterior axis into the foregut, midgut, and hindgut. Lineage tracing and genetic analysis have examined the origin of the definitive endoderm during gastrulation and demonstrated that the majority of definitive endoderm arises at the anterior end of the primitive streak (APS). Foxh1 and Foxa2 have been shown to play a role in specification of the APS and definitive endoderm. However, prior studies have focused on the role of these factors in specification of foregut definitive endoderm, while their role in the specification of midgut and hindgut definitive endoderm is less understood. Furthermore, previous analyses of these mutants have utilized definitive endoderm markers that are restricted to the anterior endoderm, expressed in extraembryonic endoderm, or present in other germ layers. Here, we characterized the expression of several novel definitive and visceral endoderm markers in Foxh1 and Foxa2 null embryos. In accordance with previous studies, we observed a deficiency of foregut definitive endoderm resulting in incorporation of visceral endoderm into the foregut. Interestingly, this analysis revealed that formation of midgut and hindgut definitive endoderm is unaffected by loss of Foxh1 or Foxa2. This finding represents a significant insight into specification and regionalization of mouse definitive endoderm.