Skip to main content
NCBI home page
Genes & Development logoLink to Genes & Development
. 2004 Jul 1;18(13):1539–1552. doi: 10.1101/gad.1207204

Hoxb1 functions in both motoneurons and in tissues of the periphery to establish and maintain the proper neuronal circuitry

Benjamin R Arenkiel 1, Petr Tvrdik 1, Gary O Gaufo 1, Mario R Capecchi 1,1
PMCID: PMC443517  PMID: 15198977

Abstract

Formation of neuronal circuits in the head requires the coordinated development of neurons within the central nervous system (CNS) and neural crest-derived peripheral target tissues. Hoxb1, which is expressed throughout rhombomere 4 (r4), has been shown to be required for the specification of facial branchiomotor neuron progenitors that are programmed to innervate the muscles of facial expression. In this study, we have uncovered additional roles for Hoxb1-expressing cells in the formation and maintenance of the VIIth cranial nerve circuitry. By conditionally deleting the Hoxb1 locus in neural crest, we demonstrate that Hoxb1 is also required in r4-derived neural crest to facilitate and maintain formation of the VIIth nerve circuitry. Genetic lineage analysis revealed that a significant population of r4-derived neural crest is fated to generate glia that myelinate the VIIth cranial nerve. Neural crest cultures show that the absence of Hoxb1 function does not appear to affect overall glial progenitor specification, suggesting that a later glial function is critical for maintenance of the VIIth nerve. Taken together, these results suggest that the molecular program governing the development and maintenance of the VIIth cranial nerve is dependent upon Hoxb1, both in the neural crest-derived glia and in the facial branchiomotor neurons.

Keywords: Neural crest, hoxb1, glia, Cre/loxP, hindbrain, rhombomere

The vertebrate cranial neural crest cells give rise to many derivatives of the head, face, and neck, including neuronal and glial cells that must act in concert for proper development of the rostral, central, and peripheral nervous systems. The neural crest originates from a migratory population of cells that delaminate from the dorsal neuroepithelium of the neural tube. In the hindbrain, neural crest cells in even-numbered rhombomeres are destined to sequentially populate the corresponding branchial arch tissues (Martin and Jessell 1991; Niederlander and Lumsden 1996; Kulesa and Fraser 1998). The coordinated interactions between the outgrowing motoneurons in the developing brainstem nuclei and the flanking, neural crest-derived branchial arch tissues may provide the means for the establishment of these neuronal circuitries that govern movement and sensory perception of the neck and face (Martin and Jessell 1991).

During formation of rhombomeres in the hindbrain, each segment adopts a unique identity, which can be characterized by distinct patterns of gene expression (for review, see Wilkenson 1993). Among the differentially expressed genes found in the hindbrain, the family of homeodomain containing transcription factors encoded by the Hox genes display overlapping patterns of rhombomere-restricted expression. Thus, each rhombomere expresses a specific combination of Hox genes, generating a Hox code thought to be in part responsible for determining the identities of individual rhombomeric segments and the tissues originating from them. Several lines of evidence have implicated the dynamic patterns of Hox gene activity within the hindbrain as important modulators of cell fates in the CNS, as well as in peripheral tissues derived from the neural crest (Davenne et al. 1999; Rossel and Capecchi 1999; Gaufo et al. 2000; Trainor and Krumlauf 2001; Pattyn et al. 2003). However, due to the pleiotropic and interdependent roles of Hox gene function within the vertebrate head, it has been difficult to separate these potentially independent roles in establishing and maintaining the neuronal circuitry between the rhombomere-born nuclei and their respective neural crest-derived target tissues.

Through a series of genetic manipulations in the mouse, evidence has accumulated that the Hox genes, in addition to their role in specifying cells of the CNS, may also be involved in establishing and maintaining the proper circuitry between neurons and their peripheral target tissues (Tiret et al. 1998; del Toro et al. 2001; Watari et al. 2001). Mice harboring both individual and compound mutations of the anteriorly expressed Hox genes exhibit overt developmental defects in the anterior CNS and/or craniofacial structures. Hoxb1 null mutants exhibit distinct defects on the developing facial nerve originating from the r4 domain of the developing hindbrain (Goddard et al. 1996; Studer et al. 1996). Characterized by significant muscular atrophy and facial paralysis, the Hoxb1 null phenotype has been attributed to the selective loss of the facial brachiomotor neurons (FBMs) that normally innervate the muscles of facial expression. However, compound mutations in Hoxa1 and Hoxb1 have also demonstrated a functional role for Hoxb1 in the normal development of tissues derived from r4 neural crest (Gavalas et al. 1998; Rossel and Capecchi 1999). In these compound mutants, the r4-derived neural crest cells fail to develop and migrate correctly, consequently, all second-arch derivatives are missing. Taken together, these loss-of-function mutations provide genetic evidence that Hoxb1 is required for the development of both central and peripheral components of the developing nervous system at the level of r4.

To elucidate the roles for Hoxb1 in these complex processes, we have utilized classical genetic mosaic analysis in addition to Cre/LoxP conditional technology in the mouse. Through genetic mosaic analysis, we show that Hoxb1 is required for the specification of facial branchiomotor progenitors in a cell-autonomous fashion. By conditionally inactivating Hoxb1 specifically in the neural crest cell population, we have also uncovered a requirement for Hoxb1 in the r4-derived peripheral tissues for maintenance and continued development of the VIIth nerve. Moreover, using different temporally restricted neural crest-specific Cre drivers, we have found that this Hoxb1 function is required early in the neural crest lineage, prior to its migration. Finally, we provide evidence that Hoxb1 functions in the glial derivatives of r4-born neural crest cells.

Results

Hoxb1 expression in rhombomere 4 is required for facial branchiomotor neuron development

The facial branchiomotor neuron (FBM) populations arise from neuronal progenitors that are born in the ventral neuroepithelium of the vertebrate hindbrain. Specified in a segment-specific manner, these maturing motoneurons organize into specific clusters that give rise to the motor pools that will form the various brainstem motor nuclei. Coincident with the rhombomeric domain in which motoneurons arise, the different brainstem nuclei adopt unique fates that are destined to innervate specific targets originating from the corresponding flanking branchial arch tissues (Martin and Jessell 1991; Niederlander and Lumsden 1996). Hoxb1, the most 3′ Hoxb gene encoded on the murine B cluster, is restricted in both expression and function to the cells and tissues originating from r4 (Fig. 1A). Utilizing a knock-in loss-of-function allele in which GFP has been inserted into the Hoxb1 locus, Hoxb1+/GFP expression can be mointored throughout embryogenesis (Gaufo et al. 2000). The earliest wave of Hoxb1 expression observed in r4 corresponds to a migratory population of neural crest cells that are entering the 2nd branchial arch (Fig. 1B). By embryonic day 10.5 (E10.5), Hoxb1 is down-regulated in the neural crest derivatives, while being maintained in the motoneuron progenitors that are differentiating in the ventral neuroepithelium. Specifically, strong Hoxb1+/GFP expression can be seen in the outgrowing axons of the FBMs that will make up the VIIth motor nucleus. At this time, extensive axonal outgrowth has already facilitated the innervation of the 2nd arch tissues (Fig. 1C).

Figure 1.

Figure 1.

Open in a new tab

Hoxb1 expression domains and chimeric analysis. (A) Scanning electron micrograph of an E10.5 mouse embryo, highlighting Hoxb1-expression domain by green pseudocoloring. (B) Direct imaging of Hoxb1+/GFP expression at E9.75 in r4 neural crest cells. (C) Hoxb1+/GFP expression at E10.5 in motoneurons of VIIth cranial nerve. Hoxb1 protein in cross-section through r4 of E11.5 chimeraLow embryo (D) and chimera High embryo (E). (F) Differentiation of Hoxb1-expressing motoneuron progenitors (green) in r4 of E11.5 chimeraLow embryo by colocalization with the general post-mitotic motoneuron marker Isl1 (red). (G) Cell autonomous loss of post-mitotic motoneurons, as shown by down-regulation of IslI in Hoxb1-/- cells in r4 of E11.5 chimeraHigh embryo. (H) Maintenance of Nkx2.2 (red) in r4 progenitors and Phox2b (green) expression in post-mitotic FBMs in chimeraLow embryos. (I) Loss of Nkx2.2 and Phox2b expression in visceral motoneuron progenitors in chimeraHigh embryos. Arrows in F and G demarcate nonvisceral motoneurons, which are more numerous in the null mutant. Arrowhead in I points to the expanded population of Phox2b-expressing cells that arise in the absence of Hoxb1.

Adult mice bearing null mutations of Hoxb1 are completely lacking the motoneurons of the facial nucleus and, consequently, exhibit facial paralysis (Goddard et al. 1996; Studer et al. 1996). Embryological analysis revealed misspecification of r4 progenitors that normally give rise to FBMs (Gaufo et al. 2000). Complete absence of Hoxb1 thus precludes the analysis of later stages of FBM differentiation and development. As a first attempt to discern the differential roles for Hoxb1 in the motoneurons versus the neural crest-derived tissues in forming a functional facial branchiomotor circuit, we conducted a genetic chimera analysis. By forming morula aggregates comprised of cells from both wild-type and Hoxb1-/- embryos, we were able to generate midgestation embryos and adults that exhibited variable levels of mutant mosaicism within the hindbrain (Fig. 1D,E; data not shown), which was also reflective of the overall levels of chimerism throughout the embryo. Offspring with high numbers of Hoxb1-/- cells (ChimeraHigh) exhibited variable levels of facial paralysis and muscular atrophy, as seen in the null mutants. Offspring containing low numbers of Hoxb1-/- cells (ChimeraLow) largely resembled wild-type mice (data not shown). The colocalization of Hoxb1 and Isl1 proteins in the motoneuron precursors allowed us to follow the clonal differentiation of Hoxb1+/+ or Hoxb1-/- cell types in a mosaic background (Fig. 1F,G). Isl1, as a general motoneuron marker, identifies all motoneurons present in the developing hindbrain. Rhombomere 4 gives rise to predominantly visceral motoneurons, which have been shown to require Hoxb1 activity during neuronal progenitor specification and FBM differentiation. In ChimeraLow embryos, Hoxb1-dependent FBM progenitors can be seen differentiating in a ventriculopial manner by the concomitant expression of Isl1 and the homeobox-containing transcription factor, Phox2b (Pattyn et al. 2000; Fig. 1F; data not shown). Whereas the Hoxb1-/- mutant cells lose all of the FBM properties, such as caudal migration and axonal extension to VIIth nerve targets (data not shown). In ChimeraHigh embryos, very few cells express the Hoxb1 protein. As a result, neuronal progenitors fail to maintain Nkx2.2, Phox2b, and Isl1 in the domain that normally gives rise to FBMs (Fig. 1G,I; data not shown). However, the few remaining wild-type cells that express Hoxb1 maintain a normal developmental program, resulting in proper differentiation and a mosaic behavioral output of VIIth nerve targets. Interestingly, in the absence of developing FBM pools, r4 neuronal specification is generally compromised, resulting in the ectopic expansion of an undefined Isl1 population of cells (Fig. 1G), and presumably a transformation into 5HT-expressing serotonergic cell fates (Pattyn et al. 2003; data not shown). Consistent with previously published work (Gaufo et al. 2000; Cooper et al. 2003), we show that Hoxb1 is required cell autonomously for early neuronal specification.

However, due to the disparity of timing between Hoxb1 expression in neural crest cells (E8.0–E9.5) and r4 motoneurons (E10.5–E12.5), in addition to the complexity of axon–2nd arch interactions, germ-line loss-of-function analysis is unable to resolve peripheral tissue-specific roles for Hoxb1 in forming the VIIth nerve, facial branchiomotor circuit. To address this issue, we turned to Cre/LoxP-mediated conditional mutagenesis.

The conditional allele of Hoxb1 (b1C) is illustrated in Figure 2A. Two mutant LoxP511–Cre recognition sites were targeted into the Hoxb1 locus so as to flank the entire Hoxb1 coding sequence. Mice heterozygous for the conditional allele were identified by Southern transfer analysis (Fig. 2B). The NeomycinR selection cassette, flanked by FRT sites and used to obtain the ES cell line containing the b1C allele, was removed from the germ line of these mice by breeding to a Flp deleter strain (Rodriguez et al. 2000) and verified by Southern transfer analysis (Fig. 2C). In the absence of Cre, mice homozygous for this conditional allele are fertile and indistinguishable from the control wild-type mice.

Figure 2.

Figure 2.

Open in a new tab

Hoxb1 conditional allele. (A) Schematic representation of the Hoxb1 conditional allele. In the targeting vector, the entire Hoxb1-coding region (solid boxes) was flanked with two lox511 sites and the GFPneo fusion selection marker was inserted in the 3′ UTR (hatched box). Thymidine kinase (TK1) negative selection marker was placed 3′ of the gene. The GFPneo marker was engineered for removal using the flanking FRT sites and a Flp-expressing mouse deleter strain. (A) AccI; (B) BsrGI; (E) EcoRI, (N) NdeI; (S) Sau3AI converted to SalI. Roman numerals indicate the position of hybridization probes. (B) Southern transfer analysis of the targeted ES clones. Genomic DNA was codigested with NdeI and BsrGI and probed with external hybridization probe I (see A). In this section, lanes 2 and 4 are correctly targeted. The presence of both lox511 sites was examined by digesting with AccI, and hybridizing to probe II (data not shown). (C) Removal of the positive selection marker with the Flp transgene. Mouse-tail genomic DNA was digested with EcoRI and hybridized with probes II and III simultaneously. Animals harboring the conditional Hoxb1 allele, but lacking the GFPneo selection marker (in this example, lanes 3,4), were subsequently bred to homozygocity, omitting the Flp transgene.

For somatic removal of the Hoxb1C allele in a tissue-specific manner, we took advantage of both new and previously described Cre drivers. For removal of Hoxb1C in premigratory neural crest, and all of its derivatives, we utilized the Wnt1–Cre transgenic driver (Chai et al. 2000). Lineage analysis of the Wnt1–Cre driver in a ROSA26 background (Soriano 1999) confirmed expression in the dorsal neural tube, in addition to all neural crest derivatives (Fig. 3C,D). To target neural crest after delamination from the neural tube for the expression of Cre, and concomitant removal of Hoxb1 function, we generated an AP2–Cre driver that functions specifically in the post-migratory neural crest cells (Macatee et al. 2003). ROSA26 lineage analysis revealed the lack of X-Gal staining in the dorsalmost neural tube, with strong staining observed in all neural crest derivatives (Fig. 3A,B). The differential activity of the Wnt1–Cre versus the AP2–Cre driver afforded us the ability to define the temporal requirement for Hoxb1 in the neural crest and its derivatives. To demonstrate the temporal difference in expression patterns between AP2 and Wnt1, we analyzed AP2 expression in the presence of a ROSA26 allele activated in the Wnt1 lineage domain at stage E10.5 (Fig. 3E–G). Although most of the cells marked by the Wnt1 lineage colocalize with AP2, a subset of tissues exhibit unique patterns of differential expression, mainly AP2 in the ectoderm, Wnt1 in the dorsalmost neural tube, and a few cells deep in the arch that have down-regulated AP2 expression. As a control to test the efficacy of the conditional system, we utilized the Deleter–Cre (Schwenk et al. 1995), which is expressed from the two-cell stage in all somatic tissues, resulting in complete recombination throughout the embryo (data not shown).

Figure 3.

Figure 3.

Open in a new tab

Cre drivers used for tissue-specific ablation of Hoxb1. (A) Lateral view of E10.5 embryo harboring the AP2–Cre and ROSA26 alleles reacted with X-Gal. (B) Dorsal view of A. (C) Lateral view of E10.5 embryo harboring the Wnt1–Cre and ROSA26 alleles reacted with X-Gal. (D) Dorsal view of C. (E–G) Colocalization of the AP2 protein (demarcating presumptive neural crest cells) with the Wnt1 lineage in r4 of E9.5 embryos. (H–J) Immunohistochemistry monitoring loss of Hoxb1 protein in the differential domains of E9 embryos targeted for conditional inactivation of the Hoxb1C allele. (H) Normal domains of Hoxb1 protein in r4 of Hoxb1C/- controls. (I) Loss of Hoxb1 protein specifically from migratory neural crest cells in embryos harboring the AP2–Cre driver. (J) Loss of Hoxb1 protein in premigratory neural crest cells and the corresponding 2nd branchial arch derivatives. (ba2) 2nd branchial arch, (nc) neural crest cells.

Conditional inactivation of Hoxb1 in premigratory neural crest phenocopies null mutants

Crossing the various Cre drivers into the conditional Hoxb1C background generated conditional mutants; crosses were set up as (male Hoxb1-/-, Cre+/-) X (female Hoxb1C/C). We established that the Hoxb1C allele efficiently recombines in vivo by detecting conditional mutations using Southern transfer analysis from genomic DNA isolated from all driver lines analyzed (data not shown). Using immunohisochemistry directed against the Hoxb1 protein, efficiency of Hoxb1C recombination was verified in situ at stage E9 using the different Cre-driver lines. Serving as a control, Hoxb1C/- embryos, devoid of Cre, exhibit the normal domains of Hoxb1 expression in both the neuroepithelium and r4-derived neural crest (Fig. 3H). Hoxb1C/- embryos inheriting the AP2–Cre driver lack Hoxb1 expression specifically in the migratory neural crest (Fig. 3I), whereas embryos harboring the Wnt1–Cre driver exhibit a broader domain of Hoxb1C/- recombination, encompassing cells of the dorsalmost neuroepithelium and all of the neural crest derivatives. (Fig. 3J). When crossed to the Deleter–Cre line, the Hoxb1C allele is recombined in all cells, resulting in the complete absence of Hoxb1 protein throughout the embryo (data not shown). Taken together, these data verify that the Hoxb1C allele is differentially recombined in a tissue/cell-specific manner in the presence of the different Cre drivers.

To better understand the roles for Hoxb1 in the pre-sumptive neural crest derivatives in establishing the proper FBM circuitry, we analyzed the adult phenotypes of the conditional mutants. Interestingly, by removing Hoxb1 in the premigratory neural crest, we were able to phenocopy the null mutant behavior. Hoxb1 null mutants are unable to blink their eyes, retract their ears, or move their whiskers, a phenotype consistent with defects in FBM development. However, the expressivity of the conditional mutant phenotype was variable with incomplete penetrance. Of the mice that harbored the appropriate alleles (Hoxb1C/-, Wnt1–Cre+/-) 35% (14/40) were completely unable to respond to forced air directed at their face, whereas 10% (4/40) exhibited partial facial paralysis. Control littermates (Hoxb1C/-) reacted adversely to the forced air by squinting their eyes and retracting their ears and whiskers (Fig. 4A,B). Upon dissecting the skin from the face of the Hoxb1C/-; Wnt1–Cre+/- mutants, we observed the loss of VIIth cranial nerve branches that normally innervate the muscles of facial expression (Fig. 4C,D). Hemotoxylin and eosin staining (H&E) of paraffin sections taken through the brainstem of adult controls and Hoxb1C/-; Wnt1–Cre+/- mice revealed the specific loss of basophilic-staining motoneuron cell bodies that normally contribute to the facial nucleus (Fig. 4E,F).

Figure 4.

Figure 4.

Open in a new tab

Conditional loss of Hoxb1 in premigratory neural crest (Wnt1–Cre domain) phenocopies adult null mutant. (A) Wild-type behavior in response to forced air blown in the face. (B) Wnt1–Cre; Hoxb1 conditional mutant behavioral phenotype in response to forced air blown in the face. (C) Nerve branches that make up the VIIth cranial nerve in control animals (arrow). (D) Conditional ablation of Hoxb1 in the Wnt1–Cre expression domain results in the loss of VIIth cranial nerve. (E) H&E staining of VIIth nucleus motoneurons in control animals. (F) Loss of motoneurons in the adult CNS of Wnt1–Cre; Hoxb1 conditional mutants.

Hoxb1 is only expressed transiently in r4 neural crest cells, prior to, and during the earliest stages of migration. To address potential roles for Hoxb1 in later stages of neural crest cell development, we inactivated its function following delamination from the neural tube using a migratory neural crest driver, AP2–Cre. Surprisingly, we did not observe any behavioral phenotypes in adult Hoxb1C/-; AP2–Cre+/- conditional mutants. The marked differences in mutant phenotypes resulting from the use of the two Cre–drivers, demonstrates that whatever role Hoxb1 confers on the r4 neural crest to maintain VIIth nerve function is conferred early, prior to the delamination and migration from the neural tube.

Early motoneuron specification and development is intact in the conditional mutants

Hoxb1 protein is required for maintaining Hoxb1 expression in r4 (Popperl et al. 1995). Taking advantage of the autoregulated nature of the endogenous Hoxb1 locus, we used a GFP reporter to assist in determining the efficacy and domains of Hoxb1C inactivation. By generating conditional mutant offspring that are Hoxb1C/GFP; Cre+/-, in which (GFP) represents the GFP-knock-in null allele, we can detect the conditional inactivation of Hoxb1 through the concomitant loss of GFP transcription (Fig. 5D).

Figure 5.

Figure 5.

Open in a new tab

Conditional loss of Hoxb1 in premigratory neural crest does not affect initial motoneuron specification. Hoxb1 autoregulation in the hindbrain, as shown by GFP expression, can be used to monitor Hoxb1 ablation. (A) Dorsal view of an 11.5 embryo harboring the Cre–Deleter, Hoxb1 conditional, and Hoxb1 GFP null alleles. (B) Dorsal view of E11.5 control embryo harboring the Hoxb1 conditional allele and null-GFP allele in the absence of Cre. (C) Dorsal view of an E11.5 embryo in which Hoxb1 has been ablated in the Wnt1 domain. (B,C, inset) GFP expression in cross-section through the VIIth nucleus. (D) Model of Hoxb1 autoregulation, in which Hoxb1 transcription requires the presence of its gene product. (E,F) Isl1 immunoreactivity in motoneurons of the VIIth nucleus, control and Wnt1–Cre conditional mutant, respectively.

As a positive control, demonstrating the efficacy of the conditional allele through the lack of GFP activity, we generated Hoxb1C/GFP; Cre–deleter+/- embryos. In these animals, no GFP signal is present in r4 or its derivatives (Fig. 5A). In a similar approach, we analyzed Hoxb1C/GFP, Wnt1–Cre+/- embryos and found that GFP expression was comparable in r4 to that of Hoxb1C/GFP controls, confirming proper expression of Hoxb1 in the CNS component of conditional mutants (Fig. 5B,C). Maintenance of Hoxb1 expression in r4 of Hoxb1C/GFP; Wnt1–Cre+/- embryos correlated with proper development and specification of motoneuron progenitors through E12.5, as shown by the presence of Hoxb1 (GFP), Isl1, and Phox2b (Fig. 5E,F, insets; data not shown), differentiation markers of post-mitotic FBMs (Pfaff et al. 1996). These data demonstrate that the loss of Hoxb1 in the periphery does not affect the early developmental programs governing FBM differentiation. Also, Hoxb1 is not required in the periphery for maintaining its own transcription in r4.

To determine the developmental time point at which the loss of Hoxb1 in the neural crest derivatives affects the development of the FBMs, we compared the numbers of Isl1-positive cells in the facial nuclei of Hoxb1C/-, Wnt1–Cre+/- mutants to that of control littermates at different stages of embryogenesis (Fig. 6A). At E12.5, there were no differences in motoneuron numbers between conditional mutants and control littermates. We first observe a loss of motoneurons in Hoxb1C/-, Wnt1–Cre+/- mutants at E14.5, in which the average neuron number was 152 ± 29 (sd), compared with 182 ± 21 (sd) in control littermates (p = 0.02). Curiously, this is the time point in which normal neuronal pruning, through apoptosis, is initiated during development. By E16.5, the majority of cell loss is complete (Fig. 6B,C). At this time, Hoxb1C/-, Wnt1–Cre+/- mutants are reduced to an average number of 59 ± 40 (sd), whereas control littermates retained a constant number of 174 ± 7 (sd; p = 0.002). Of particular interest was the observation that at E16.5 Hoxb1C/-; AP2–Cre+/- conditional mutants exhibited a loss of motoneurons significantly different (p = 0.01) from that of controls, 147 ± 8 (sd) versus 174 ± 7 (sd), respectively (Fig. 5C,D), revealing a cellular phenotype undetectable through adult behavioral phenotyping.

Figure 6.

Figure 6.

Open in a new tab

Loss of motoneurons occurs during later stages of embryogenesis in the Hoxb1, neural crest-specific conditional mutants. (A) Graph illustrating the developmental loss of VIIth nucleus motoneurons in the CNS of conditional mutants. (B) Cross section through VIIth nucleus of Hoxb1 c/- control embryo immunoreacted with Isl1 antibody. (C) Cross-section through VIIth nucleus of Wnt1–Cre; Hoxb1 c/- embryo immunoreacted with Isl1 antibody. (D) Cross-section through VIIth nucleus of AP2–Cre; Hoxb1 c/- embryo immunoreacted with Isl1 antibody. (B–D) Sections of E16.5 embryos.

In all tissues analyzed, the loss of Isl1 expression coincided with the positive reaction to TUNEL (data not shown), indicating that the primary mechanism of motoneuron loss is through programmed cell death. Taken together, these data suggest a role for Hoxb1 in the premigratory neural crest population required later during the development of CNS-born FBMs.

Conditional loss of Hoxb1 in the periphery results in axon outgrowth defects

Having determined the time window of FBM apoptosis due to conditional loss of Hoxb1 in the neural crest derived tissues, we were able to focus our studies on the causative defects originating in the periphery. Because we did not observe any difference in motoneuron loss between mutants and control littermates until E14.5, we directed our peripheral analysis to earlier developmental stages. Taking advantage of the autoregulated GFP reporter system, which transiently expresses Hoxb1 (GFP) in the r4 neural crest and maintains high levels of expression in the FBMs, we evaluated the extent of recombination in the 2nd arch tissues and imaged the outgrowing VIIth nerve axons as they extended into the periphery. Through this analysis, it became evident that (1) excision of the Hoxb1 conditional allele by Wnt1–Cre was largely complete (>90% of the cells that would normally express Hoxb1–GFP lacked GFP expression) in the 2nd arch derivatives (data not shown); (2) the early steps in FBM specification and neural tube exit are not affected; and (3) the first noticeable phenotype in the periphery was at E12.5. At this time, Hoxb1C/-; Wnt1–Cre+/- mutants reproducibly exhibited defects in axonal branching and/or fasciculation, with an average number of branch points being significantly lower in mutants compared with controls (Fig. 7A–D; data not shown). These results allowed us to identify the first time point that Hoxb1 programmed cells may act in forming the peripheral FBM circuitry. However, it was still unclear as to what types of cells or tissues Hoxb1-expressing neural crest cells formed and how they affected this biological process.

Figure 7.

Figure 7.

Open in a new tab

VIIth nerve axonal outgrowth is compromised by E12.5 in conditional mutants. Autoregulation of the GFP (null) allele can be used to monitor the integrity of VIIth nerve out-growth. (A) Hoxb1–GFP expression in Hoxb1 c/- control animals (5×). (B) A 25× magnification of inset shown in A. (C) Hoxb1–GFP expression in Wnt1–Cre; Hoxb c/- conditional animals (5×). (D) A 25× magnification of inset shown in C, highlighting the onset of axonal defects that arise in the periphery of conditional Hoxb1 mutants.

Hoxb1-expressing neural crest cells give rise to glial cells of the PNS

To address the fates of Hoxb1-expressing neural crest cells, we turned to genetic lineage analysis using the ROSA26 reporter system (Soriano 1999). We mapped the Hoxb1-expressing neural crest derivatives originating from r4 using a Hoxb1–Cre mouse line (Arenkiel et al. 2003).

By crossing the Hoxb1–Cre driver to ROSA–GFP mice (Mao et al. 2001), we were able to identify the Hoxb1 lineage that gives rise to the peripheral components of the developing VIIth cranial nerve at E12.5, the stage in which we first observe the axonal defects in the periphery. At this point in development, the VIIth nerve has bifurcated into two main branches; one that dives deep into the muscle precursors of the neck, as the other navigates rostrally to form terminal synapses with the muscles of facial expression. We were interested in determining the tissue types derived from the Hoxb1 expression domain that may influence VIIth nerve development in the periphery at the time of axonal defects. Transverse sections taken through the anterior region of the embryonic face revealed that none of the target or support tissues associated with the VIIth nerve branches, other than glial progenitors, were descendents from the Hoxb1 expression domain (Fig. 8A–D). Colocalization of GFP-positive cells with multiple molecular markers demonstrated that peripheral Hoxb1 derivatives involved in the FBM circuit give rise primarily to glia. An example of this colocalization is the overlapping expression of GFP with the HMG containing transcription factor Sox10 (Britsch et al. 2001; Paratore et al. 2001; Fig. 8D). We followed the Hoxb1 lineage to adulthood, and found that these cells localize to the main fascicle of the VIIth cranial nerve that is innervating the muscles of facial expression (Fig. 8E,F). In addition to the FBMs that make up the neuronal component of the VIIth nerve, as shown by colocalization with the cytoskeletal neuronal marker β-III-Tubulin (Tuj1; Fig. 8G,H), Hoxb1-expressing neural crest also give rise to a significant percentage of glial cells associated with the VIIth nerve. More specifically, we were able to colocalize GFP expression with the Schwann cell marker Myelin Basic Protein (MBP; Fig. 8I,J). Interestingly, it appears as if the r4-born FBMs are closely associated with neural crest-derived glia that also expressed Hoxb1. In fact, <5% of the FBM axons are ensheathed by glia that are not expressing the GFP lineage marker. This correlation suggests a paradigm in which proper FBM circuitry is established early in embryogenesis through the coordinated programs of r4-born neurons and their r4-derived glial support cells.

Figure 8.

Figure 8.

Open in a new tab

Genetic lineage analysis reveals that Hoxb1 neural crest gives rise to myelinating cells of the VIIth cranial nerve. (A–D) Lineage analysis using ROSA–GFP and Hoxb1–Cre in transverse sections through anterior facial structures to identify Hoxb1 derivatives and potential VIIth nerve interactions. (A) Colocalization of β-III-Tubulin with Hoxb1-derived VIIth nerve bundles traversing through first arch derivatives. (B) PECAM expression in blood vessels associated with, but nonoverlapping with the Hoxb1 lineage. (C) Nonoverlapping expression of Myogenin-expressing facial muscle precursors and the VIIth nerve. (D) Sox10 expression in the first arch derivatives shows colocalization with VIIth nerve glia, and nonoverlapping expression in cartilage. This particular antibody cross-reacts with both Sox9 and Sox10 proteins, facilitating detection of both glial and chondrocyte progenitors (M. Wegner, pers. comm.). (E) GFP and DAPI expression in cross-section through adult peripheral nerve highlighting Hoxb1 lineage. (F) A 100× magnification of E. (G) Cross-section through the VIIth peripheral nerve immunoreacted with an antibody against the neuronal marker Tuj1, colocalizing GFP activity with axons extending from the motoneurons of the facial nucleus. (H) A 100× magnification of G. (I) Cross-section through the VIIth peripheral nerve immunoreacted with an antibody against myelin basic protein (MBP), colocalizing MBP-expressing glia reactivity to the Hoxb1 lineage. (VIIn) Seventh cranial nerve, (Vg) fifth cranial ganglia, (ba1) first branchial arch, (S) sensory tract, arrow-Hoxb1-derived Schwanns cell, arrowhead-FBM axon associated with a non-Hoxb1-derived Schwanns cell.

Loss of Hoxb1 function does not affect glial progenitor specification

Several studies have described genetic mechanisms by which Hox genes function to control cell-fate decisions along the axis of the developing embryo, primarily through influencing progenitor pools for proper specification. We have previously shown that the initial allocation and overall number of neural crest cells derived from r4 is normal in Hoxb1-/- mutants (Goddard et al. 1996). This was verified by immunohistochemical analysis directed against the general neural crest markers, transcription factor AP2 (Schorle et al. 1996), and the low-affinity neurotrophin receptor P75 (Morrison et al. 1999; data not shown). However, we wished to determine whether neural crest cell fate specification is compromised in Hoxb1-/- mutant mice. To determine whether Hoxb1 functions in the r4-derived neural crest lineage to specify glial progenitor cell fates, we turned to an in vitro analysis using primary neural crest cell cultures.

E8.5 embryos (6–9 somites) were removed from pregnant females, and segments from presumptive r3–r5 were dissected from the hindbrain. Mesenchymal and endodermal tissues were removed from explants, and the remaining neural tubes were transferred to laminincoated culture dishes containing a defined growth medium. Over a period of 72 h, neural crest cells were allowed to delaminate from the neural tube explant, migrate onto the laminin substrate, and express early differentiation markers.

Neural crest populations were identified first by morphological criteria, such as location relative to explant, cell shape, and clonal expansion (Fig. 9A). Identity was then verified by immunohistochemistry using the neural crest-specific marker AP2 (Leask et al. 1991; Schorle et al. 1996; Fig. 9B). To determine whether Hoxb1 influences the early steps in allocating peripheral glial progenitors, we assayed for the glial progenitor marker Sox10 in Hoxb1-/- versus wild-type neural crest cell cultures. We did not observe a significant difference in the number of Sox10-expressing neural crest cells between mutant and wild-type cell cultures (Fig. 9C–E). Consistent with these results, we also did not find a significant difference in expression of Sox10, GFAP, and PLP through E14.5 between mutant and wild-type embryos (Fig. 9F,G; data not shown). However, fewer overall numbers of glial cells are observed with the loss or absence of FBMs at later stages in the mutant mice. Taken together, these data suggest that the initial glial specification program is intact in the Hoxb1 mutants.

Figure 9.

Figure 9.

Open in a new tab

Glial progenitor specification is not affected in Hoxb1-/- neural crest cells. (A) Brightfield DIC image of 72-h neural crest culture. (B) Immunohistochemical identification of neural crest cells by expression of the migratory crest marker AP2 (green). (C) A 63× magnification of wild-type neural crest cells expressing the glial progenitor marker Sox10 (yellow). (D) A 63× magnification of Hoxb1-/- neural crest cells expressing Sox10 (yellow). In all panels, nuclei are marked with DAPI (blue). (E) Graph representing relative fraction of Sox10-expressing cells in 72-h neural crest cultures. Data reflects results from 12 independent samples for each genotype. (F) Immunohistochemistry directed toward Sox10 (glia) and Tuj1 (neurons) expression in an E11.5 wild-type embryo. (G) Sox10 and Tuj1 expression in an E11.5 Hoxb1-/- embryo.

Discussion

It has been postulated that the principal role for the Hox genes is to provide positional information to cells along the body axes of the developing embryo, so as to, in part, coordinate the formation of tissue at a given axial level. Such a need is particularly apparent for the formation of neural circuits along the rostro caudal axis of the body. For this reason, the observation that Hox genes, such as Hoxb1, were found to be expressed at a specific axial level both in the central nervous system as well as in the neural crest emanating from that level, was particularly satisfying (Wilkenson et al. 1989; Hunt et al. 1991). This observation leaves open the question of how Hox genes function at the cellular and molecular level to contribute to the coordinated formation of neural circuits involving, for example, motoneurons generated within the CNS and the peripheral target tissues. Herein, we have explored this question with respect to the role of Hoxb1 in the formation and/or maintenance of the facial motoneuron circuit (i.e., that involving the FBMs).

In addition to expression of Hoxb1 in the FBMs and r4-neural crest, loss-of-function mutations in Hoxb1, and in its paralogous family member, Hoxa1, also support a pleiotropic role for these Hox genes in the formation of the facial motoneuron circuit. Hoxb1 null mutants exhibit severe facial paralysis and muscular atrophy, a phenotype that has been shown to result from the loss of FBMs (Goddard et al. 1996; Studer et al. 1996). In addition, studies of compound mutant mice harboring both Hoxb1 and Hoxa1 loss-of-function alleles have revealed a specific role for Hoxb1 in tissues derived from the neural crest originating from r4 (Gavalas et al. 1998; Rossel and Capecchi 1999). Through expression analysis using a Hoxb1–GFP reporter allele, we elucidated the dynamic, differential expression pattern of Hoxb1 in cells and tissues derived from r4. Interestingly, following the strong expression at E9.5 in r4-neural crest, within 24 h Hoxb1 is rapidly down-regulated in the crest-derived tissues and strongly up-regulated in the FBMs that are programmed to innervate these same tissues. This is surprisingly similar to the scenario that has been described for the ETS-containing transcription factors, in which it has been proposed that the sensory-motor system is formed through the coordinate expression of a transcription factor in both the central and peripheral components of the nervous system in order to establish and maintain proper neuronal circuitry (Lin et al. 1998; Arber et al. 2000; Livet et al. 2002).

As a first attempt to address this issue, we utilized a genetic mosaic approach in which we generated chimeric animals comprised of both wild-type and Hoxb1-/- cells. Through this analysis, we were able to define a specific role for Hoxb1 in the initial specification of FBM progenitors, independent of influence from neural crest-derived peripheral tissues, confirming previous work conducted in the background of the germ-line null mutant and ruling out an early non-cell autonomous role for Hoxb1 in motoneuron specification. Unlike the observation that peripheral cues are required for the induction and maintenance of the ETS gene Pea3 in motor neurons of the CNS, resulting in proper cell body localization and axonal arborization (Livet et al. 2002), Hoxb1 initiates its autoregulatory expression in the CNS independent of peripheral cues. Moreover, in contrast to Pea3 not being required for general MN specification, Hoxb1, acting cell-autonomously in the CNS, is absolutely necessary to initiate the appropriate FBM differentiation program. We next turned our attention toward understanding the roles for Hoxb1 in the neural crest lineage during the formation of the FBM circuit.

Using a conditional mutagenesis approach, uncoupling the functions of Hoxb1 in the neural crest from that of the developing motoneurons, we have uncovered a novel function for Hoxb1 in maintaining the FBM circuitry at later developmental stages in the periphery. Further, we show that conditional loss of Hoxb1 in the periphery does not influence the normal developmental program governing the proper specification of FBMs that are born in the CNS. By somatic inactivation of Hoxb1 specifically in the neural crest, we generated conditional mutants that phenocopied the Hoxb1 germ-line null mutant phenotype, although with incomplete penetrance and variable expressivity.

There are a few potential explanations for the incomplete penetrance and variable expressivity. First, coordinated development of multiple tissues involving complex circuitry is biologically plastic (Trainor and Krumlauf 2000a,b). Mutant cells, or tissues, may be replaced, displaced, or reprogrammed by ones that can act in their place for proper function. For example, Hoxa1 may compensate for the loss of Hoxb1 in r4 neural crest tissues. This hypothesis is testable by generating double conditional mutants. Secondly, a specific temporal requirement for Hoxb1 may exist in the developing neural crest population originating from r4. Interestingly, in this study, we are only able to phenocopy the null mutants if we remove Hoxb1 from the premigratory neural crest population using the Wnt1–Cre driver. Although we observe a phenotypic difference at the cellular level, we do not observe a behavioral phenotype if we remove Hoxb1 function in the postmigratory crest population using AP2–Cre. Although we cannot rule out the possibility that Hoxb1 has an additional role in this circuitry through an unknown activity in the interneuron population arising from the dorsal neural tube, another domain subject to inactivation by Wnt1–Cre, we favor a temporal requirement for Hoxb1 early in crest development. Ultimately, this would have to be tested by inactivating Hoxb1 specifically in the dorsal region of the neural tube, while sparing the premigratory neural crest pool, or rescuing Hoxb1 function in these same cells in a Hoxb1c/-; Wnt1–Cre background. Finally, incomplete penetrance and variable expressivity of the conditional phenotype could arise from the inherent technical imprecisions of the Cre/loxP conditional mutagenesis system as applied to the mouse. This may result in Wnt1–Cre not triggering excision of Hoxb1 sufficiently early, or at variable extents, in the premigratory neural crest to consistently abolish all Hoxb1 activity in these cells. Nevertheless, our results demonstrate a nonautonomous role for Hoxb1 in maintaining proper FBM circuitry, beyond that of specification, in the ventral neural tube.

Having evaluated the loss of CNS-born motoneurons as a bioassay to determine when conditional ablation of Hoxb1 in the periphery manifested its phenotype, we were able to delineate that peripheral defects begin prior to E14.5. Utilizing a GFP reporter assay, driven by auto-regulation of the Hoxb1 locus, we showed that the first discernable phenotype of the VIIth nerve in the periphery occurred at E12.5, which was defective in axonal branching and outgrowth into 2nd arch derived tissues. A similar phenotype has also been described for the IXth nerve in Hoxa3 mutants (Manley and Capecchi 1997; Watari et al. 2001). This observation supports the hypothesis that loss of Hox gene function in the neural crest lineage may generally affect mechanisms governing motoneuron axonal path finding and/or growth.

To determine the cellular identity of the r4-derived neural crest tissue that directly influences the development of the VIIth cranial nerve, we used a genetic lineage approach that utilized a previously described Hoxb1–Cre driver (Arenkiel et al. 2003) and ROSA–GFP reporter. This proved to be very informative, because it identified the cells and tissues that initially associate with the out-growing VIIth nerve axons in addition to the terminally differentiated cells and tissues that make up the mature VIIth nerve. At mid-embryogenesis, when axons first encounter the periphery, they associate with the cranial sensory ganglia, coalescing with axons from the sensory neurons and glia to form a fascicle as the nerve grows into the 2nd arch. Cells of the ganglia are derived from both neural crest and placodal origin (D'Amico-Martel and Noden 1983; Stark et al. 1997), giving rise to a complex structure that relays environmental information from the periphery back to the motoneurons in the CNS to complete a functional neuronal circuit. When we conducted the lineage analysis of r4 neural crest cells, we found that the Hoxb1 expressing neural crest preferentially give rise to glial progenitors (Arenkiel et al. 2003). At E10.5, most of the Hoxb1-derived neural crest cells associated with VIIth/VIIIth ganglia express the glial progenitor marker Sox10. In the adult, this lineage persisted within the VIIth nerve branch as nonneuronal Hoxb1-derived cells and colocalized with the Schwann cell marker MBP. By E12, the anterior branch of the VIIth nerve that is destined to innervate the muscles of facial expression turns rostrally, exiting the tissues derived from the second branchial arch and entering first arch derivatives. Unable to colocalize GFP expression in the Hoxb1 lineage with any other tissues types thought to be important for peripheral circuitry, we concentrated on a role for Hoxb1 in the glial lineage. Consistent with the proposed functional register between neurons and neural crest-derived, flanking-arch tissues that express the same Hox code (Niederlander and Lumsden 1996), we observed a strong correlation between neurons born in r4 and the subsequent ensheathment by glia derived from that same segment. Albeit r4 neural crest give rise to many other cell and tissue types derived from the 2nd branchial arch, such as cartilage, muscle, and connective tissue, lineage analysis in the adult (Arenkiel et al. 2003) has yet to identify these other tissues in association with the FBM circuitry as originating from the Hoxb1 domain. It is possible that transient interactions between the Hoxb1-expressing neural crest cells and other 2nd-arch derivatives, such as muscle precursors, at an earlier developmental time point somehow program the out-growing FBMs for proper pathfinding and synapse maintenance. However, we have been unable to detect any defects in these tissues in either the germ-line or conditional Hoxb1 mutants. Moreover, our data support glial progenitor cells as the significant derivative of the Hoxb1-expressing neural crest associated with this circuit, and suggests mechanisms whereby these cells participate in the maintenance and outgrowth of the VIIth nerve.

Evaluating the potential for neural crest cultures to give rise to Sox10-expressing cells in vitro, no significant difference in glial specification between wild-type and mutant neural crest cultures was discerned. Although the overall specification program appears to be intact, by the time FBMs begin to die in our conditional model, glial cells are compromised in a concordant fashion. This may be an indication of other problems that arise later in the glial specification program, migration, or even within a subset of misspecified glial progenitors that would go undetected through our in vitro analysis. However, these observations are consistent with the idea that initial glial fate decisions are not compromised in Hoxb1 mutants, and that Hoxb1 neural crest derivatives are acting later in the overall maintenance of the r4 circuitry. In this scenario, Hoxb1 in the neural crest-derived glial cells is initiating a molecular program, independent of specification, required later for direct communication with the maturing FBMs. Compromising the regulation of these downstream targets results in loss of FBMs through programmed cell death. This is a compelling idea, as in our system, glial cells are initially intact, but are inevitably lost upon prolonged interaction and failure to support the proper branching and fasciculation of the cognate FBMs. In fact, it may be worth speculating that Hoxb1 expression in the VIIth nerve glial cells may play a much more active role in FBM branching and/or target recognition rather than, or in addition to, a general support role. For example, in Sox10-/- and ErbB3-/- mutant mice, glial cells are invariably missing, resulting in general axonal growth and branching defects prior to early neuronal cell death (Riethmacher et al. 1997; Britsch et al. 2001; Paratore et al. 2001). This is in contrast to our phenotype in which FBMs are correctly specified, send out initial fascicles that undergo a main bifurcation, but are ultimately lost near the time at which they normally make connections with their terminal targets. This may be an indication of molecular defects in the glial cells, resulting in failure to sustain guidance or provide general support. Potential candidates for regulation by Hoxb1 include axon guidance molecules, such as the Ephs and Ephrins (Wang and Anderson 1997), or growth factors belonging to the neurotrophin family (Oorschot and McLennan 1998; Oppenheim et al. 2000).

Taken together, the data presented in this study are consistent with a functional role for Hoxb1 in both the developing FBMs and neural crest-derived peripheral glia toward the establishment and maintenance of a functional neuronal circuit. Specification of neural crest-derived structures and associated nervous tissue has been postulated to be a developmentally coordinated program, in which the proper formation of one tissue is directly dependent upon the other. This study provides insight at the cellular level into how the vertebrate embryo properly coordinates such complex developmental programs through the dynamic regulation of Hox genes in multiple cell and tissue types at a given axial level.

By undertaking a detailed genetic and molecular analysis of Hoxb1, we have uncovered novel functions for Hoxb1 in FBM development and identified the cells and tissues in which it acts. Although we have yet to determine the specific molecular mechanisms by which Hoxb1 governs the formation and maintenance of FBM circuitry, we are now poised to begin to identify specific downstream targets involved in this process. Through the use of this system, in conjunction with microarray analysis, attempts at identifying downstream targets for Hoxb1 are underway.

Materials and methods

Generation of genetic chimeras

Chimeric embryos and adult mice were generated from morula derived from intercrosses between wild-type BL6 to wild-type BL6, and Hoxb1-/- to Hoxb1-/- mice. Wild-type and Hoxb1 mutant morulas were aggregated and transferred into pseudo-pregnant BL6 females. Chimeras were harvested at E11 or allowed to develop to adulthood.

Construction of the Hoxb1 conditional allele

Hoxb1 genomic DNA was isolated from a 129Sv mouse genomic library (Stratagene). To introduce the upstream lox511 sequence, an EcoRV site was engineered in the 5′UTR of the Hoxb1 gene, followed by insertion of an oligonucleotide to produce the following sequence context: GGGataacttcgtatagtata cattatacgaagttatcTGTCTCCCCCAAACGGCCCGACCCTCC TTCGGCCTCTACATG (the newly introduced sequence is in lowercase, the Hoxb1 start codon is in bold). Similarly, the downstream lox511 site along with the EGFPneo fusion gene (flanked with FRT sites) was inserted in an artificial AscI site placed 37 bp from the Hoxb1 stop codon. A total of 10.8 kb of genomic sequence was included in the targeting vector (Fig. 2A), which was linearized with SalI, electroporated in ES cells, and selected on G418 and FIAU. Positive ES clones were identified by Southern blotting, using a 0.7-kb 5′ external probe on NdeI + BsrGI-digested genomic DNA. Mice derived from one positive ES clone were mated with Flpe-expressing mice (Rodriguez et al. 2000), and the progeny were screened for the absence of the EGFPneo selection cassette (Fig. 1C).

X-gal staining

Embryos harboring both the ROSA26 reporter and either the Wnt1–Cre, AP2–IRES–Cre, or Deleter–Cre driver alleles were removed from pregnant females in PBS with 2 mM MgCl2. Embryos were fixed (2% paraformaldehyde, 1.25 mM EGTA, 2 mM MgCl2, 0.1 M PIPES at pH 6.9), washed in PBS, 2 mM MgCl2, and immersed into X-gal staining solution [15 mM K3Fe (CN)6, 15 mM K4Fe (CN)6-3H20, 2 mM MgCl2, 0.01% Na Deoxycholate, 0.02% NP40 in PBS at pH 7.2)]. Staining was carried out overnight at room temperature with rocking.

Adult phenotyping analysis

Hoxb1 conditional models were phenotypically characterized by two criteria. First, the ability or inability to respond to forced air blown in the face. Behavior responses included closure of eyelids, pinning back of ears, and retraction of whiskers. Secondly, the mice were examined for compromised integrity of the VIIth cranial nerve. Analysis included the visualization of branch loss in addition to motoneuron loss within the facial nucleus. H&E staining was done by standard protocols on paraffin-embedded brainstems cut at 10 μm.

Immunohistochemistry

Embryonic and adult tissues were dissected in PBS and fixed for 3 h at 4°C with 4% paraformaldehyde in PBS (pH 7.2). Samples were washed in PBS, equilibrated to 30% sucrose, and embedded in O.C.T. Serial sections were cut at 10 μm, washed in PBS, and preincubated in blocking solution (2% BSA, 5% NGS, 0.1% triton in PBS, at pH 7.2). Primary antibodies were diluted in blocking solution and applied overnight at 4°C. The primary antibodies used in this study were as follows: α-GFP polyclonal (Molecular probes; 1:1000), α-Nkx2.2 monoclonal (Developmental Hybridoma; 1:50), α-Islet1 monoclonal (Developmental Hybridoma; 1:50), α-AP2 monoclonal (1:50), α-Phox2b polyclonal (C. Goridis, CNRS, Paris, France; 1:1000), α-PECAM monoclonal (Pharmacia; 1:100), α-Myogenin (Developmental Hybridoma; 1:50), α-Tuj1 monoclonal (Covance; 1:1000), α-Sox10 monoclonal (Dr. Michael Wegner, Institute fur Biochemie, Universitat Erlangen, Erlangen, Germany; 1:50), α-Hoxb1 polyclonal (Covance; 1:300), α-MBP monoclonal (Chemicon; 1: 75). Alexafluor-conjugated secondary antibodies (α-Rb or α-Ms; Molecular probes) were used at a dilution of 1:500 in blocking solution and incubated at 4°C for 4 h.

GFP expression in Hoxb1gfp/+ reporter embryos was analyzed through in situ enhancement of the GFP signal. Embryos were fixed as above, washed, and preincubated in blocking solution (2% BSA, 2% milk, 0.5% triton in PBS at pH 7.2). Whole-mount immunohistochemistry was carried out using the α-GFP polyclonal (Molecular probes; 1:1500) diluted in blocking solution, followed by secondary enhancement using the Alexafluor FITC-conjugated α-Rb antibody (Molecular probes; 1:500). Finally, tissues were washed 5 × 1 h each in PBS/.1% Triton X-100, cleared by equilibration into 95% glycerol, and imaged by confocal microscopy.

For motoneuron cell counts, entire brainstems were cryo-sectioned at 10 μm and the number of IslI-positive cells present in the facial nuclei were counted. The five centralmost sections, containing the highest number of motoneuron cell bodies, were used to determine an average number of neurons per section in each nucleus. Facial nuclei were treated independently in all animals analyzed.

Immunodetection and fluorescence confocal microscopy was carried out using Bio-Rad MRC1024 instrumentation and software.

Neural crest cultures

E8.5 embryos (6–9 somites) were dissected into PBS (Mg+2, Ca+2, at pH 7.2) from timed pregnant females. Neural tube explants spanning r3–r5 were isolated, and flanking mesodermal and ventral endodermal tissues were removed with a tungsten needle. Clean neural-tube explants were transferred to laminin-coated culture dishes containing a defined neural crest growth medium. Culture dishes and medium was prepared as previously described (Strachan and Condic 2003). Laminin was bound to culture dishes at a concentration of 20 μg/mL. Cultures were grownfor 72 h at 37°C at 7.5% CO2, changing medium every 24 h. Immunohistochemistry was carried out by rinsing cells in PBS, fixing in 4% paraformaldehyde in PBS for 15 min, washing 3 × 5 min each in PBS, preincubation in blocking solution (10% NGS, 0.1% BSA, 0.3% Triton X-100 in PBS) for 15 min at room temperature, and incubation of primary antibodies for 3 h at room temperature in blocking solution. Primary antibodies were used at the following concentrations: α-Sox10 monoclonal (Dr. Michael Wegner; 1:25), α-AP2 (3B5) monoclonal (Developmental Hybridoma; 1:25). Secondary detection was carried out using the Alexafluor FITC-conjugated α-Ms antibody (Molecular probes; 1:300) after incubation for 2 h in blocking solution at room temperature, and washing 4 × 10 min each in PBS. Visualization and cell counts were done on a Zeiss inverted epifluorescent microscope using 3-Eye imaging software. Statistitical analysis was carried out on a sample size of 12 independent cultures for each genotype.

Acknowledgments

We thank Carol Lenz, Sheila Barnett, and Julie Tomlin for their technical expertise and advice on ES cell culturing conditions and early embryo manipulation. Also, we are grateful for the helpful input and advice provided by Dr. Maureen Condic and Lauren Strachan on neural crest cell culture technique. Finally, we thank Anne Boulet, Shannon Odelberg, and Matthew Hockin for critical reading and helpful discussion on this manuscript.

The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 USC section 1734 solely to indicate this fact.

Article published online ahead of print. Article and publication date are at http://www.genesdev.org/cgi/doi/10.1101/gad.1207204.

References

  1. Arenkiel B.R., Gaufo, G.O., and Capecchi, M.R. 2003. Hoxb1 neural crest preferentially form glia of the PNS. Dev. Dyn. 227: 379-386. [DOI] [PubMed] [Google Scholar]
  2. Arber S., Ladle, D.R., Lin, J.H., Frank, E., and Jessell, T.M. 2000. ETS gene Er81 controls the formation of functional connections between group Ia sensory afferents and motor neurons. Cell 101: 485-498. [DOI] [PubMed] [Google Scholar]
  3. Britsch S., Goerich, D.E., Riethmacher, D., Peirano, R.I., Rossner, M., Nave, K.A., Birchmeier, C., and Wegner, M. 2001. The transcription factor Sox10 is a key regulator of peripheral glial development. Genes & Dev. 15: 66-78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chai Y., Jiang, X., Ito, Y., Bringas Jr., P., Han, J., Rowitch, D.H., Soriano, P., McMahon, A.P., and Sucov, H.M. 2000. Fate of the mammalian cranial neural crest during tooth and mandibular morphogenesis. Development 127: 1671-1679. [DOI] [PubMed] [Google Scholar]
  5. Cooper K.L., Leisenring, W.M., and Capecchi, M.R. 2003. Autonomous and nonautonomous functions for Hox/Pbx in branchiomotor neuron development. Dev. Biol. 253: 200-213. [DOI] [PubMed] [Google Scholar]
  6. D'Amico-Martel A. and Noden, D.M. 1983. Contributions of placodal and neural crest cells to avian cranial peripheral ganglia. Am. J. Anat. 166: 445-468. [DOI] [PubMed] [Google Scholar]
  7. Davenne M., Maconochie, M.K., Neun, R., Pattyn, A., Chambon, P., Krumlauf, R., and Rijli, F.M. 1999. Hoxa2 and Hoxb2 control dorsoventral patterns of neuronal development in the rostral hindbrain. Neuron 22: 677-691. [DOI] [PubMed] [Google Scholar]
  8. del Toro E.D., Borday, V., Davenne, M., Neun, R., Rijli, F.M., and Champagnat, J. 2001. Generation of a novel functional neuronal circuit in Hoxa1 mutant mice. J. Neurosci. 21: 5637-5642. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Gaufo G.O., Flodby, P., and Capecchi, M.R. 2000. Hoxb1 controls effectors of sonic hedgehog and Mash1 signaling pathways. Development 127: 5343-5354. [DOI] [PubMed] [Google Scholar]
  10. Gavalas A., Studer, M., Lumsden, A., Rijli, F.M., Krumlauf, R., and Chambon, P. 1998. Hoxa1 and Hoxb1 synergize in patterning the hindbrain, cranial nerves and second pharyngeal arch. Development 125: 1123-1136. [DOI] [PubMed] [Google Scholar]
  11. Goddard J.M., Rossel, M., Manley, N.R., and Capecchi, M.R. 1996. Mice with targeted disruption of Hoxb-1 fail to form the motor nucleus of the VIIth nerve. Development 122: 3217-3228. [DOI] [PubMed] [Google Scholar]
  12. Hunt P., Gulisano, M., Cook, M., Sham, M.H., Faiella, A., Wilkinson, D., Boncinelli, E., and Krumlauf, R. 1991. A distinct code for the branchial region of the vertebrate head. Nature 353: 861-864. [DOI] [PubMed] [Google Scholar]
  13. Kulesa P.M. and Fraser, S.E. 1998. Neural crest cell dynamics revealed by time-lapse video microscopy of whole embryo chick explant cultures. Dev. Biol. 204: 327-344. [DOI] [PubMed] [Google Scholar]
  14. Leask A., Byrne, C., and Fuchs, E. 1991. Transcription factor AP2 and its role in epidermal-specific gene expression. Proc. Natl. Acad. Sci. 88: 7948-7952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Lin J.H., Saito, T., Anderson, D.J., Lance-Jones, C., Jessell, T.M., and Arbor, S. 1998. Functionally related motor neuron pool and muscle sensory afferent subtypes defined by coordinate ETS gene expression. Cell 95: 393-407. [DOI] [PubMed] [Google Scholar]
  16. Livet J., Sigrist, M., Stroebel, S., De Paola, V., Price, S.R., Henderson, C.E., Jessell, T.M., and Arber, S. 2002. ETS gene Pea3 controls the cental position and terminal arborization of specific motor neuron pools. Neuron 35: 877-892. [DOI] [PubMed] [Google Scholar]
  17. Macatee T.L., Hammond, B.P., Arenkiel, B.R., Francis, L., Frank, D.U., and Moon, A.M. 2003. Ablation of specific expression domains reveals discrete functions of ectoderm and endoder-derived FGF8 during cardiovascular and pharyngeal development. Development 130: 6361-6374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Manley N.R. and Capecchi, M.R. 1997. Hox group 3 paralogous genes act synergistically in the formation of somitic and neural crest-derived structures. Dev. Biol. 192: 274-288. [DOI] [PubMed] [Google Scholar]
  19. Mao X., Fujiwara, Y., Chapdelaine, A., Yang, H., and Orkin, S.H. 2001. Activation of EGFP expression by Cre-mediated excision in a new ROSA26 reporter mouse strain. Blood 97: 324-326. [DOI] [PubMed] [Google Scholar]
  20. Martin J.H. and Jessell, T.M. 1991. Development as a guide to the regional anatomy of the brain. In Principles of neuroscience, 3rd edition (eds. E.R. Kandel et al.), pp. 296-308. Elsevier, New York.
  21. Morrison S.J., White, P.M., Zock, C., and Anderson, D.J. 1999. Prospective identification, isolation by flow cytometry, and in vivo self-renewal of multipotent mammalian neural crest stem cells. Cell 96: 737-749. [DOI] [PubMed] [Google Scholar]
  22. Niederlander C. and Lumsden, A. 1996. Late emigrating neural crest cells migrate specifically to the exit points of cranial branchiomotor nerves. Development 122: 2367-2374. [DOI] [PubMed] [Google Scholar]
  23. Oorschot D.E. and McLennan, I.S. 1998. The trophic requirements of mature motoneurons. Brain Res. 789: 315-321. [DOI] [PubMed] [Google Scholar]
  24. Oppenheim R.W., Houenou, L.J., Parsadanian, A.S., Prevette, D., Snider, W.D., and Shen, L. 2000. Glial cell line-derived neurotrophic factor and developing mammalian motoneurons: Regulation of programmed cell death among motoneuron subtypes. J. Neurosci. 20: 5001-5011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Paratore C., Goerich, D.E., Suter, U., Wegner, M., and Sommer, L. 2001. Survival and glial fate acquisition of neural crest cells are regulated by an interplay between the transcription factor Sox10 and extrinsic combinatorial signaling. Development 128: 3949-3961. [DOI] [PubMed] [Google Scholar]
  26. Pattyn A., Hirsch, M., Goridis, C., and Brunet, J.F. 2000. Control of hindbrain motor neuron differentiation by the homeobox gene Phox2b. Development 127: 1349-1358. [DOI] [PubMed] [Google Scholar]
  27. Pattyn A., Vallstedt, A., Dias, J.M., Samad, O.A., Krumlauf, R., Rijli, F.M., Brunet, J.F., and Ericson, J. 2003. Coordinated temporal and spatial control of motor neuron and serotonergic neuron generation from a common pool of CNS progenitors. Genes & Dev. 17: 729-737. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Pfaff S.L., Mendelsohn, M., Stewart, C.L., Edlund, T., and Jessell, T.M. 1996. Requirement for LIM homeobox gene Isl1 in motor neuron generation reveals a motor neuron-dependent step in interneuron differentiation. Cell 84: 309-320. [DOI] [PubMed] [Google Scholar]
  29. Popperl H., Bienz, M., Studer, M., Chan, S.K., Aparicio, S., Brenner, S., Mann, R.S., and Krumlauf, R. 1995. Segmental expression of Hoxb-1 is controlled by a highly conserved auto-regulatory loop dependent upon exd/pbx. Cell 81: 1031-1042. [DOI] [PubMed] [Google Scholar]
  30. Riethmacher D., Sonnenberg-Riethmacher, E., Brinkmann, V., Yamaai, T., Lewin, G.R., and Birchmeier, C. 1997. Severe neuropathies in mice with targeted mutations in the ErbB3 receptor. Nature 389: 725-730. [DOI] [PubMed] [Google Scholar]
  31. Rodriguez C.I., Buchholz, F., Galloway, J., Sequerra, R., Kasper, J., Ayala, R., Stewart, A.F., and Dymecki, S.M. 2000. High-efficiency deleter mice show that FLPe is an alternative to Cre–loxP. Nat. Genet. 25: 139-140. [DOI] [PubMed] [Google Scholar]
  32. Rossel M. and Capecchi, M.R. 1999. Mice mutant for both Hoxa1 and Hoxb1 show extensive remodeling of the hindbrain and defects in craniofacial development. Development 126: 5027-5040. [DOI] [PubMed] [Google Scholar]
  33. Schorle H., Meier, P., Buchert, M., Jaenisch, R., and Mitchell, P.J. 1996. Transcription factor AP-2 essential for cranial closure and craniofacial development. Nature 381: 235-238. [DOI] [PubMed] [Google Scholar]
  34. Schwenk F., Baron, U., and Rajewsky, K. 1995. A cre-transgenic mouse strain for the ubiquitous deletion of loxP-flanked gene segments including deletion in germ cells. Nucleic Acids Res. 23: 5080-5081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Soriano P. 1999. Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat. Genet. 21: 70-71. [DOI] [PubMed] [Google Scholar]
  36. Stark M.R., Sechrist, J., Bronner-Fraser, M., and Marcelle, C. 1997. Neural tube–ectoderm interactions are required for trigeminal placode formation. Development 124: 4287-4295. [DOI] [PubMed] [Google Scholar]
  37. Strachan L.R. and Condic, M.L. 2003. Neural crest motility and integrin regulation are distinct in cranial and trunk populations. Dev. Biol. 259: 288-302. [DOI] [PubMed] [Google Scholar]
  38. Studer M., Lumsden, A., Ariza-McNaughton, L., Bradley, A., and Krumlauf, R. 1996. Altered segmental identity and abnormal migration of motor neurons in mice lacking Hoxb-1. Nature 384: 630-634. [DOI] [PubMed] [Google Scholar]
  39. Tiret L., Le Mouellic, H., Maury, M., and Brulet, P. 1998. Increased apoptosis of motoneurons and altered somatotopic maps in the brachial spinal cord of Hoxc-8-deficient mice. Development 125: 279-291. [DOI] [PubMed] [Google Scholar]
  40. Trainor P. and Krumlauf, R. 2000a. Plasticity in mouse neural crest cells reveals a new patterning role for cranial mesoderm. Nat. Cell Biol. 2: 96-102. [DOI] [PubMed] [Google Scholar]
  41. ____. 2000b. Patterning the cranial neural crest: Hindbrain segmentation and Hox gene plasticity. Nat. Rev. Neurosci. 1: 116-124. [DOI] [PubMed] [Google Scholar]
  42. ____. 2001. Hox genes, neural crest cells and branchial arch patterning. Curr. Opin. Cell. Biol. 13: 698-705. [DOI] [PubMed] [Google Scholar]
  43. Wang H.U. and Anderson, D.J. 1997. Eph family transmembrane ligands can mediate repulsive guidance of trunk neural crest migration and motor axon outgrowth. Neuron 18: 383-396. [DOI] [PubMed] [Google Scholar]
  44. Watari N., Kameda, Y., Takeichi, M., and Chisaka, O. 2001. Hoxa3 regulates integration of glossopharyngeal nerve precursor cells. Dev. Biol. 240: 15-31. [DOI] [PubMed] [Google Scholar]
  45. Wilkinson D.G. 1993. Molecular mechanisms of segmental patterning in the vertebrate hindbrain. Perspect. Dev. Neurobiol. 1: 117-125. [PubMed] [Google Scholar]
  46. Wilkinson D.G., Bhatt, S., Cook, M., Boncinelli, E., and Krumlauf, R. 1989. Segmental expression of Hox-2 homoeobox-containing genes in the developing mouse hindbrain. Nature 341: 405-409. [DOI] [PubMed] [Google Scholar]

Articles from Genes & Development are provided here courtesy of Cold Spring Harbor Laboratory Press

RESOURCES