The transcription factor Hes genes are essential effectors of Notch signaling, which regulates the maintenance of progenitor cells and the timing of their differentiation in various tissues and organs (Kageyama and Ohtsuka, 1999). Hes genes play crucial roles in neurulation by upholding the pluripotency and self-renewal of neural stem cells until the appropriate developmental stages for their differentiation (Hatakeyama et al., 2004; Hatakeyama and Kageyama, 2006). Hes1, a member of the family of Hes genes, acts as a negative regulator of neuronal differentiation by repressing the activities of proneural bHLH factors such as Mash1, neurogenin (Ngn) 2, and Math1. The lack of Hes1 causes the severe neural hypoplasia, which is attributable to accelerated neuronal differentiation and concomitant depletion of neural precursor cells (Ishibashi et al., 1995). A closure defect of the cranial neural tube is a characteristic phenotype of Hes1 null mutant embryos; exencephaly occurs at the forebrain and midbrain boundary and dramatic brain malformations are induced (Akimoto et al., 2010a; Kameda et al., 2011). In addition, the null mutants exhibit multiple craniofacial malformations including calvaria agenesis, defective anterior cranial base, shortened maxilla and mandible, and abnormal palate (Akimoto et al., 2010b). The mesenchyme of each of these affected organs is derived from cranial neural crest cells. Thus, Hes1 plays a role in regulating the development of mesenchymal neural crest cells in tissues other than the nervous tissue in the head and neck. Hes1 is widely expressed in the neuronal epithelium and mesenchymal cells of the cranium (Kita et al., 2007). Furthermore, in the pharyngeal region of early mouse embryos, Hes1 is expressed in the endoderm, ectoderm, and mesenchyme (Van Bueren et al., 2010).
Neural crest cells are multipotent stem cells that originate from the dorsal neural tube and contribute to many structures of the head, face, neck, and cardiovascular system (Le Douarin and Kalcheim, 1999). The Wnt1-Cre/R26R (double transgenic) mice are available for the elucidation of the long-term fate of neural crest lineages (Chai et al., 2000; Jiang et al., 2000). The sympathetic ganglia are derived from the neuronal neural crest cells. During early embryonic development, neural crest cells migrate ventrally and aggregate adjacent to the dorsal aorta to form the primary sympathetic chain (Huber, 2006). The sympathetic ganglia primordia are characterized by the expression of the transcription factors Phox2b, Mash1, Phox2a, and Hand2, which are induced by bone morphogenetic proteins (BMPs) secreted by the dorsal aorta (Hirsch et al., 1998; Howard, 2005; Huber, 2006). These transcription factors initiate the expression of neuronal markers and the enzymes responsible for noradrenalin synthesis (i.e., tyrosine hydroxylase [TH] and dopamine β-hydroxylase) in the sympathetic ganglion cells, causing them to acquire a catecholaminergic neuronal phenotype. The cervical sympathetic ganglion primordium in the mouse embryos initially forms a uniform column at the all cervical levels. The superior cervical ganglion (SCG) is then separated from the satellite ganglion at embryonic day (E) 13.5 and becomes localized between cervical levels 1 and 4 (C1 and C4) at E14.5 (Nishino et al., 1999). At these stages of development, the noradrenergic neuronal progenitors are commonly denoted by their expression of TH, whereas the glial progenitors for sympathetic satellite cells are denoted by their expression of brain lipid-binding protein (BLBP; Shi et al., 2008). BLBP belongs to the family of fatty-acid binding protein and is expressed in radial glia in the developing brain.
The carotid body is a chemosensory organ that is sensitive to arterial hypoxia, hypercapnia, and acidity. The organ is critical for the regulation of the respiratory system. The carotid body is made up of two types of cells, the glomus cells (also known as chief cells or type I cells) and the sustentacular cells (also known as type II cells). Glomus cells are of neuroendocrine lineage and exhibit immunoreactivity for TuJ1 antigen (neuron-specific class III β-tubulin isotype), protein gene product (PGP) 9.5 (a marker of the nervous and neuroendocrine systems), neuropeptide Y and chromogranin A (Kameda et al., 2002). Glomus cells also produce biogenic amines (e.g., serotonin [5-HT] and noradrenalin) and are intensely labeled by antibody against TH, the rate-limiting enzyme involved in catecholamine synthesis, from E13.5 onward in the mouse. Sustentacular cells, on the other hand, are of glial lineage and contain aggregates of intermediate filaments. The cells express glial markers such as glial fibrillary acidic protein (GFAP), S100 protein, and vimentin (Kameda, 1996, 2005). The carotid body is located in the bifurcation of the carotid artery and is densely innervated by the carotid sinus nerve, a sensory branch of the glossopharyngeal nerve, and also by sympathetic nerve fibers originating from the SCG (Kameda et al., 2008). The murine carotid body connects with the SCG at the carotid bifurcation region. In mouse embryos, the carotid body rudiment is formed in the wall of the third arch artery as a condensation of mesenchymal cells at E13.0 (Kameda et al., 2008). The walls of the pharyngeal arch arteries, including the third arch artery, are derived from mesenchymal neural crest cells (Jiang et al., 2000). The carotid body rudiment is surrounded by the nerve bundles extending from the SCG. The ganglion provides glomus cell precursors which begin to invade the carotid body rudiment at E13.5 and differentiate into glomus cells within this structure (Kameda, 2005).
Mash1 is essential for the differentiation of sympathetic ganglia derived from neural crest cells (Guillemot et al., 1993; Sommer et al., 1995). In mice deficient for Mash1, neural crest cells are able to migrate to the dorsal aorta but fail to differentiate into sympathetic neurons, resulting in the absence of the sympathetic ganglia (Hirsch et al., 1998). The deficiency of Mash1 also induces the lack of glomus cells in the carotid body; as a result, the organ consists of sustentacular cells only (Kameda, 2005). As noted above, Mash1 expression is repressed by Hes1 in the nervous system (Ishibashi et al., 1995). Because the differentiation of both the sympathetic ganglia and the carotid body is regulated by Mash1, we hypothesized that the development of the SCG and carotid body is under the influence of Hes1. The results of this study demonstrated that severe hypoplasia of the SCG was induced in Hes1 homozygous null mutant mice. In addition, the carotid body was absent in a subset of null mutants in which the third arch artery was defective.
Expression of Neural Crest-Derived Cell Markers in the SCG and Carotid Body
To prove that not only the SCG but also the carotid body is derived from neural crest cells, Wnt1-Cre/R26R mice were examined by X-galactosidase (X-gal) staining for the detection of β-galactosidase. By the use of this two-component genetic system, the progeny of the neural crest cells were labeled indelibly. In the embryos at E12.5, the cervical sympathetic primordium appeared as a uniform column consisting of X-gal-positive cells (Fig. 1A). At this developmental stage, the carotid body was not yet formed. In the newborn Wnt1-Cre/R26R mice, both the SCG and the carotid body exhibited intense X-gal staining (Fig. 1B).
Development of the SCG in Hes1 Null Mutant Mice
The development of the SCG in Hes1 null mutant mice compared with wild-type mice was analyzed by immunohistochemistry. Antibodies to TH, chromogranin A, TuJ1, and PGP 9.5 were used to label sympathetic neurons, whereas antibodies to BLBP and S100 protein were used to label glial cells. In wild-type embryos at E12.5, the column-like structure of the sympathetic primordium was detected along the common carotid artery in the cervical region (Fig. 2A). In the absence of the Hes1 gene, the sympathetic primordium was normally formed and located along the artery at E12.5 (Fig. 2B). The volume of the mutant primordium, however, was already smaller than that of the wild-type (Fig. 8A).
In E13.5 wild-type embryos, the SCG was separated from the stellate ganglion, which lies at the cervical/thoracic border, began to migrate rostrally and became thick (Fig. 3A). The formation of the SCG reflects the accumulation of cells in the upper portions of the sympathetic primordium. The rostral pole of the SCG abutted to the cochlea at this time point. The sympathetic ganglion was primarily composed of noradrenergic neurons and satellite glial cells. The sympathetic neurons exhibited intense immunoreactivity for TH (Fig. 3A), in addition to immunoreactivity for the neuronal markers TuJ1 and PGP 9.5 (not shown). The glial cells in the wild-type SCG were immunoreactive for BLBP, a marker of glial progenitors (Fig. 3B). Like the SCG in the wild-type embryos, the SCG in the Hes1 null mutant embryos at E13.5 displayed intense immunoreactivity for TH (Fig. 3D) and BLBP (Fig. 3E), as well as TuJ1, PGP9.5 (not shown). However, the SCG in the null mutants had an attenuated growth relative to the SCG in wild-types (Fig. 3D–F). In 5 of 14 cases in the E13.5 null mutants, the SCG retained an immature phenotype such as that observed for E12.5 embryos, exhibiting elongated features without thickening (Fig. 4A,B). The elongated, thin ganglion was often continuous with the stellate ganglion. In some cases, the SCG in the null mutants separated into several nodular ganglia (Fig. 4C). In others, the SCG failed to undergo rostral migration, remaining in the lower cervical region (Fig. 4D). Thus, a remarkable phenotypic diversity of the SCG was encountered in the E13.5 Hes1 null mutants, and its location often deviated from that seen in wild-type embryos. In contrast, the stellate (lower cervical) ganglion showed no obvious changes in size and location between the two genotypes. The number of phospho-histone H3 immunoreactive (i.e., proliferating) cells was markedly decreased in the SCG of null mutants compared with wild-type embryos (Fig. 3C,F). Enhanced apoptotic cell death was not observed in either the SCG or the carotid body of the null mutants, as assessed by TUNEL labeling experiments (not shown).
At E14.5, the SCG of wild-type embryos was now located in its final position, between the levels of cervical vertebrae C1 and C3. The size of the SCG progressively increased with age in wild-type embryos (Fig. 8A). In the null mutants, however, the increase of SCG with age was slight and rather impaired at the latest fetal stage.
In wild-type mouse embryos at E15.5, the enlarged SCG was situated close to the carotid bifurcation in the upper cervical region (Fig. 5A,B). In the Hes1 null mutants, the SCG still exhibited an elongated, narrow phenotype, which is hypoplastic (Fig. 5C,D). In some cases, the elongated SCG occupied the entire cervical region (Fig. 5E).
At later time periods (E17.5 and E18.5), the SCG in the null mutants was markedly reduced in size, compared with that in the wild-type embryos. In most cases, the small SCG was located close to the carotid bifurcation which was located in the upper cervical region (Fig. 6A,B). In 1 of 16 cases, the SCG deviated remarkably, locating in the mid cervical region (Fig. 6C). Furthermore, in one case, the thin SCG extended over a particularly long range (Fig. 6D–G). For each genotype, the glial cells of the SCG were immunoreactive for BLBP but not for S100 protein, whereas the carotid body sustentacular cells were immunoreactive for S100 but not for BLBP (Fig. 6F,G).
The carotid sinus nerve (also known as the carotid sinus baroreceptor) arising from the glossopharyngeal nerve was densely distributed around the root of the internal carotid artery in the null mutants as well as in wild-type embryos (Fig. 7A,B). Thus, the lack of Hes1 did not appear to affect the development of this nerve.
Development of the Carotid Body in Hes1 Null Mutant Mice
In wild-type mice, the carotid body rudiment was formed in the wall of the third arch artery at E13.0. The rudiment always made contact with the SCG and was surrounded by nerve bundles extending from the SCG. At E13.5, the carotid body rudiment was round or oval in shape and was located adjacent to the upper portion of the common carotid artery, i.e., in the carotid artery bifurcation. Glomus cells that were immunoreactive for TH (Fig. 3A), chromogranin A, PGP9.5, and TuJ1 (not shown) began to appear in the peripheral portion of the carotid body rudiment at E13.5. In the Hes1 null mutants, the time course of carotid body formation was similar to that in the wild-type embryos. At E13.5, however, the carotid body rudiment in the null mutant was small in size, compared with that in the wild-type. Furthermore, the glomus cells already occupied a larger portion of the mutant carotid body at this stage (Figs. 3D, 4B,D). The common carotid artery in the null mutants frequently bifurcated at a lower cervical level. Therefore, the mutant carotid body was located at a lower level than the wild-type organ, and displayed elongated features (Fig. 4A,B). In 2 of 10 cases in E13.5 null mutant mice, the carotid body was absent as a result of the elimination of the third arch artery, i.e., the common carotid artery.
In wild-type mice from E15.5 and onward, the carotid body was packed with the small clusters of TH-positive glomus cells (Fig. 5A,B). Except for the cases in which the common carotid artery was missing, the carotid body in Hes1 null mutants compared with wild-type embryos was small in size, and the glomus cells appeared to be decreased in number (Fig. 5D,E). Furthermore, many mutant embryos displayed a short common carotid artery and therefore, as noted above, the carotid body in the mutants was situated at a lower cervical level (Fig. 5C,D).
Quantification of the Volume and Cell Number of the SCG and Carotid Body
The volume of the SCG and the number of constituent cells on both sides of each animal were measured in wild-type and Hes1 null mutant embryos at E12.5, E13.5, E15.5, and E17.5–E18.5 (Table 1; Fig. 8A). The volume of the carotid body was measured at E13.5, E15.5, and E17.5–E18.5 (Fig. 8B). Because many, but not all, of the null mutants died at E15.5, the embryos at both E17.5 and E18.5 were grouped and jointly regarded as the latest developmental stage for the procurement of quantitative data.
Table 1. Volume and Cell Number of the SCG in Wild-type and Hes1 Null Mutant Mice at Various Developmental Stagesa
No. of samples
Ganglion volume ×10−3(mm3)
Cell density (per mm2)
Total cell number (per ganglion)
Ratio of −/− to +/+ value ×100
SCG, superior cervical ganglion of the sympathetic trunk; +/+, the wild type; −/−, the null mutant. Each value is the mean ± SEM.
P < 0.05;
P < 0.01;
P < 0.001 vs. wild types assessed by Student's t test.
Quantitative data revealed that the volume of the SCG in the Hes1 null mutants was significantly smaller than that in wild-type embryos at all stages examined (Fig. 8A; Table 1). In addition, the number of cells in the SCG was significantly decreased in the null mutants; the mutant SCG became hypoplastic (Table 1). Thus, the lack of Hes1 gene caused a conspicuous reduction in SCG volume and neuron number.
The volume of the carotid body in the Hes1 null mutants was also measured and compared with the volume in wild-type embryos (Fig. 8B). In contrast to the SCG, the increase in the volume of the carotid body with age was relatively small in wild-type embryos, particularly at the later developmental stages. The carotid body was notably affected by the lack of Hes1 gene. In 4 of 30 cases examined (13.3%), the common carotid artery was missing, and the carotid body was consequently not formed. In the cases in which the common carotid artery was present, the mutant carotid body was reduced in volume, although the reduction was far less than that of the mutant SCG.
The development of the SCG in the sympathetic trunk was severely affected by the absence of Hes1 function. The volume of the SCG in Hes1 null mutant mice was reduced to 45% of its value in wild-type embryos as early as E13.5. By E17.5 and E18.5, the volume of the mutant SCG was decreased to 26.4% of the value in the wild-type controls. Similarly, an obvious reduction in the number of cells in the SCG was demonstrated for the null mutants. Thus, the SCG in the null mutants was small in size and hypoplastic, compared with that in wild-type mice, although the appearance and location of the ganglion were variable among mutant animals and even between the right and left sides of the same animal. Notably, few variations in the appearance and/or location of the SCG were found for wild-type animals.
Neural crest cells give rise to both neurons and glia of the sympathetic nervous system. The neural crest cell-derived sympathetic primordium became aligned in a uniform column along the dorsal aorta in wild-type mice. After separating from the stellate ganglion, the primordium migrated rostrally to generate the SCG and settled between the C1 and C3 vertebral levels. The SCG was gradually enlarged in volume with fetal age. This morphogenic sequence may be attributed to the marked increase of sympathetic neurons at level C1 to C3 and below C7, the sites of the SCG and stellate ganglion, respectively (Rubin, 1985). The number of cells in the SCG increases until birth in wild-type mice (Nishino et al., 1999). In Hes1 null mutants, some SCG retained an immature phenotype for extended periods of time; an elongated, narrow ganglion that failed to separate from the stellate ganglion was encountered at E13.5 and thereafter. In these cases, the rostral migration and the accumulation of cells to form the SCG appeared to be arrested. In wild-type mice at E13.5, the SCG displayed significant levels of H3-positive proliferating cells. In the null mutants at the same age, the proliferating cells in the SCG were markedly decreased in number. Apoptotic cells, however, were not increased. Taken together, it is supported that the hypoplasia of the SCG in the Hes1 null mutants is attributable to decreased cell proliferation.
The proneural gene Mash1 is transiently expressed in migrating sympathetic precursor cells (Guillemot et al., 1993). When the precursor cells are induced to differentiate, the Mash1-expression is rapidly down-regulated. The targeted disruption of Mash1 results in the elimination of the sympathetic ganglia, including the SCG (Kameda, 2005). Mash1 expression is negatively regulated by Hes1, which is required for the maintenance of neuronal progenitors and their pluripotency (see Kageyama et al., 1997, for a review). Hes1 exerts its repressive actions through direct binding to a class C site in the promoter region of the Mash1 gene (Chen et al., 1997). Up-regulation of Mash1 expression has been demonstrated in the mice deficient for Hes1 (Ishibashi et al., 1995). The present study indicated that the sympathetic primordium of the cervical region in Hes1 null mutants compared with wild-type embryos was already hypoplastic at E12.5. Therefore, the differentiation of the sympathetic progenitors may have been accelerated by increased expression of Mash1 in the null mutants at earlier developmental stages. Thus, premature neurogenesis may induce a reduction in the size of the sympathetic primordium. Progenitor maintenance and neuronal differentiation from progenitor cells are both essential for neurogenesis during sympathetic ganglion development (Tsarovina et al., 2008). In addition, cell proliferation was markedly decreased in the Hes1 null mutants, resulting in severe SCG defects at later stages.
The SCG originates in the vagal neural crest of the hindbrain (corresponding to somites 1–5). The other ganglia of the sympathetic chain, including the stellate ganglion, originate in the trunk neural crest (posterior to somite 6; Durbec et al., 1996). In addition to Mash1 and Hes1, several genes required for the development of the sympathetic ganglia have been demonstrated by loss-of-function experiments; i.e., the homeobox gene Phox2a (a determinant of the noradrenergic phenotype), ret, glial cell-line derived neurotrophic factor (GDNF), Gfrα3 (a component of the receptor for neurotrophic factor artemin), and neurotrophin-3 (Moore et al., 1996; Morin et al., 1997; Nishino et al., 1999; Francis et al., 1999; Enomoto et al., 2001). In the null mutant mice deficient for these genes, the SCG displays variable reductions in size. On the other hand, any obvious anomalies of the stellate ganglion are not detected in Phox2a−/−, c-ret−/−, and Gfrα3−/− mice (Morin et al., 1997; Durbec et al., 1996; Nishino et al., 1999). Similarly, in FRSα2F/2F (see below) and Necdin mutant mice, the SCG is affected, whereas no differences in the size and position are observed for sympathetic ganglia with more posterior locations, including the stellate ganglion (Kameda et al., 2008; Tennese et al., 2008). Thus, the SCG derived from the vagal neural crest of the hindbrain appears to have unique properties relative to the other sympathetic ganglia derived from trunk neural crest.
We previously demonstrated that three elements, i.e., the third arch artery, the carotid sinus nerve and the SCG, are required for the formation and development of the carotid body (Kameda et al., 2008). Hox genes regulate the formation of the mammalian body plan along the anteroposterior axis. In Hoxa3 null mutant mice, the thymus and the parathyroid glands, which are derived from the third pharyngeal pouch, are absent, and the common carotid artery, which is derived from the third arch artery, is missing bilaterally (Kameda et al., 2002, 2003). Because the third arch artery degenerates at E11.5, the carotid body rudiment is never formed in Hoxa3 null mutants. The docking protein FRS2α is an important mediator of fibroblast growth factor (FGF) –induced signal transduction, and functions by linking FGF receptors to a variety of intracellular signaling pathways (Hadari et al., 2001). In FRS2α2F/2F mice, in which the Shp2-binding sites of FRS2α are disrupted, the carotid sinus nerve fibers do not project properly to the third arch artery and the formation of the carotid body rudiment is arrested (Kameda et al., 2008). In Mash1 null mutants in which the SCG is not formed, the carotid body contains only the sustentacular cells and no glomus cells (Kameda, 2005).
The smooth muscle cells of the pharyngeal arch arteries are derivatives of the mesenchymal neural crest cells (Jiang et al., 2000). It has been reported in the Hes1−/− mouse embryos that hypo/aplasia of the 3rd, 4th, and 6th arch arteries are detected as a partially penetrated phenotype (35%; Van Bueren et al., 2010). The initial formation of the pharyngeal arch arteries occurs normally in the null mutants. Therefore, the defective patterning of the pharyngeal arch arteries may be due to a decreased contribution of the neural crest cells, and/or to a failure of the cells to differentiate properly. In the present study, the lack of the common carotid artery derived from the third arch artery was encountered in 4 of 30 cases (13.3%) in the Hes1 null mutants. Furthermore, a shorter than normal common carotid artery, which divided into the internal and external carotid arteries at a lower cervical level, was frequently observed in the null mutants. In the cases in which the common carotid artery was missing, the carotid body was not formed. The organ also settled at a lower level when the carotid bifurcation was located at a lower cervical level.
With the exception of cases in which the common carotid artery was missing, the carotid body rudiment was formed in the Hes1 null mutant embryos. Furthermore, the carotid sinus nerve, a sensory branch of the glossopharyngeal nerve, that is crucial for the initial formation of the carotid body rudiment, was normally distributed around the root of the internal carotid artery. The carotid body of the null mutants, however, was smaller in size than that of the wild-type embryos at all developmental stages examined, although significant differences were observed only at the latest stage (E17.5–E18.5). At E17.5 and E18.5, the volume of the carotid body in the null mutants was reduced to 52.5% of the value in wild-types. The reduction in the volume of the carotid body was far less than the reduction in the volume of the SCG. Mash1, which is negatively regulated by Hes1, is required for the genesis of glomus cells but not sustentacular cells (Kameda, 2005). The carotid body, therefore, may be less affected than the SCG by the lack of Hes1.
The sympathetic progenitors contribute to the generation of the carotid body glomus cells (Kameda, 2005). SCG-derived glomus cell precursors begin to enter the carotid body rudiment at E13.5 in wild-type mice. In the Hes1 null mutants, SCG-derived precursors were also able to colonize the rudiment and differentiate into glomus cells, despite the reduction of carotid body volume. Sympathetic neurons and glomus cells share similar neuronal properties; both cells are immunoreactive for TuJ1, PGP 9.5, NPY, TH, and chromogranin A (Kameda et al., 2002). Moreover, glomus cells exhibit 5-HT immunoreactivity. Sustentacular cells derived from the mesenchymal neural crest cells exhibited different properties from sympathetic glial cells in both wild-type and mutant embryos. The glial cells were immunoreactive for BLBP but not for S100, whereas the sustentacular cells were immunoreactive for S100 but not for BLBP.
In conclusion, the SCG, carotid body and common carotid artery in Wnt1/Cre-R26R mice were intensely labeled by X-gal staining; these organs are derived from the neural crest cells. In the Hes1 null mutant embryos, the SCG was markedly hypoplastic and was retained as an elongated, narrow ganglion for extended period of time. In the cases in which the common carotid artery was absent, the carotid body rudiment was not formed. When the artery was present, the carotid body in the null mutants was smaller than that in the wild-type embryos. Taken together, the results of the current study suggest that the lack of Hes1 affects the development of the neural crest derivatives in the cervical region.
Wild-type, heterozygous, and homozygous Hes1 mice were obtained by intercrossing Hes1+/− mice, which were maintained in a CD1 background (Ishibashi et al., 1995). The Hes1+/− mouse was supplied by Prof. R. Kageyama. E0.5 was designated as noon on the day that a copulation plug was identified. Mouse genotypes were obtained by polymerase chain reaction analysis using the previously described primers (Ishibashi et al., 1995). For the fate mapping study of neural crest cells, the transgenic mouse lines Wnt1-Cre and ROSA26 LacZ reporter (R26R) allele, the latter of which expresses β-galactosidase (β-gal) upon Cre-mediated recombination, were used. Wnt1-Cre/R26R double transgenic mice were produced as described previously (Chai et al., 2000; Jiang et al., 2000).
For histological analysis, specimens were incubated in Bouin's solution or 8% paraformaldehyde (PFA) in phosphate buffer (PB) overnight, embedded in paraffin, and then serially sectioned along the frontal and sagittal planes at a thickness of 5–6 μm. For frozen sections, specimens were fixed in 4% PFA for time periods ranging from 30 min to 2 hr. Selected sections were stained with hematoxylin-eosin to determine the morphological orientation.
Immunohistochemical staining was performed using the streptavidin-biotin-peroxidase method. The following primary antibodies were used: rabbit polyclonal antibodies to tyrosine hydroxylase (TH; Chemicon, El Segundo, CA; 1:100 dilution), 5-HT (serotonin; Chemicon; 1:1,000 dilution), protein gene product (PGP) 9.5 (Dako, Carpinteria, CA; 1:200 dilution), chromogranin A (Yanaihara Inst., Shizuoka, Japan; 1:1,000 dilution), S100 protein (Dako, prediluted) and brain lipid-binding protein (BLBP; Frontier Inst., Ishikari, Hokkaido, Japan; 1:100 dilution), and the monoclonal antibody TuJ1 that recognizes the neuron-specific class III β-tubulin isotype (Berkeley Antibody Company, Richmond, CA; 1:500 dilution). Proliferating cells were examined using an anti-phospho-histone H3 (Ser10) antiserum (Santa Cruz Biotechnology, 1:100 dilution), a mitosis marker.
X-galactosidase (X-gal) staining was performed according to standard procedures, as described previously (Chisaka and Kameda, 2005).
TUNEL Assay for Apoptosis
To visualize apoptotic nuclei, tissue sections were subjected to staining using the terminal transferase dUTP-biotin nick-end labeling (TUNEL) kit (ApoMark Apoptosis Detection Kit; Exalpha Biological, Inc.; Maynard, MA), according to the manufacturer's instructions.
The SCG and carotid body were photographed in Hes1 null mutant and wild-type littermate embryos at E12.5, E13.5, E15.5, E17.5, and E18.5. Images of the photographed organs were digitized using a digital camera system (Olympus, DP72, Tokyo) attached to an Olympus AX80 microscope. Morphometric quantification was performed using WinRoof Image software (version 6.1; Mitani Corp., Fukui, Japan). To measure the size of each organ, the outlines of the SCG and carotid body were traced by hand, and an automated system was used to assess the profile area in every fifth section in a series of consecutive sections. The volume of each organ was then estimated by adding the profile area of each section and multiplying the sum by five and by the thickness of the section. Concerning the measurements of cell number, it is difficult to discriminate clearly non-neuronal cells from neurons in SCG during fetal period. In particular, the SCG primordium mostly consists of progenitor cells. Therefore, all papers on the mouse embryonic SCG published until now have counted these cells as neuronal cells. The present study followed the usage. For determination of the number of neural cells per ganglion, the SCG in hematoxylin-eosin-stained sections was assessed automatically by color binalization, as described previously (Chisaka and Kameda, 2005). The average cell density in the SCG was quantified, and the total cell number in each section was calculated from the profile area and the average cell density. The cell counting was corrected according to the Abercrombie's formula (Scorisa et al., 2009). Because the difference in size affects cell counts, the diameters of 25 randomly picked neurons from each animal were measured, and the mean value was calculated and used for cell count correction. The total number of cells in the ganglion was then estimated by adding the numbers of cells in all of the sections and multiplying by five. At least four animals per genotype were analyzed at each stage for the quantification studies. Results were expressed as the mean ± SEM. Differences were assessed by the Student's t-test and considered statistically significant at P < 0.05.
We thank Prof. Ryoichiro Kageyama (Institute for Virus Research, Kyoto University, Kyoto, Japan) for the gift of the Hes1 mouse line.