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

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.

Figure 1.

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 chimera^Low embryo (D) and chimera ^High embryo (E). (F) Differentiation of Hoxb1-expressing motoneuron progenitors (green) in r4 of E11.5 chimera^{Low 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 chimera^High embryo. (H) Maintenance of Nkx2.2 (red) in r4 progenitors and Phox2b (green) expression in post-mitotic FBMs in chimera^Low embryos. (I) Loss of Nkx2.2 and Phox2b expression in visceral motoneuron progenitors in chimera^High 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.

Figure 2.

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.

Figure 3.

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 Hoxb1^C allele. (H) Normal domains of Hoxb1 protein in r4 of Hoxb1^C/- 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 2^nd branchial arch derivatives. (ba2) 2^nd branchial arch, (nc) neural crest cells.

Figure 4.

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.

Figure 5.

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.

Figure 6.

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.

Figure 7.

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.

Figure 8.

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.

Figure 9.

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.