SEARCH

SEARCH BY CITATION

Keywords:

  • Mash1;
  • NeuroD;
  • mouse embryos;
  • thyroid C cells;
  • ultimobranchial body;
  • calcitonin;
  • CGRP

Abstract

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

In mammals, the ultimobranchial body derived from the fourth pharyngeal pouch gives rise to thyroid C cells. The C cells of newborn mice are immunoreactive for calcitonin, calcitonin gene–related peptide (CGRP), protein gene product (PGP) 9.5 and NeuroD, and transiently exhibit the neuronal markers TuJ1 and somatostatin during fetal development. The basic helix-loop-helix (bHLH) transcription factor Mash1 plays a role in the differentiation of autonomic neurons. We show that in wild-type mouse embryos, Mash1 is expressed in the ultimobranchial body at embryonic day (E) 12.5, when the body is located close to the great arch arteries. It is also expressed in the ultimobanchial body fused with the thyroid lobe at E 13.5. Targeted disruption of Mash1 resulted in the absence of C cells in the mouse thyroid glands, since cells displaying the C-cell markers and expressing NeuroD were not detected during fetal development or at birth. The failure of C-cell formation in the null mutant thyroids was also confirmed by electron microscopy. While the formation and migration of the ultimobranchial body were not affected in the Mash1 null mutants, at E 12.5–E 13.5 both the ultimobranchial body located close to the arteries and the organ populating the thyroid lobe exhibited a marked increase in apoptotic cell numbers. Thus, in the mutant mice, the ultimobranchial body fails to complete its differentiation program and finally dies. These results indicate that Mash1 enhances survival of the C-cell progenitors by inhibiting apoptosis. Developmental Dynamics 236:262–270, 2007. © 2006 Wiley-Liss, Inc.


INTRODUCTION

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

The thyroid gland of mammals consists of two cell types, namely, follicular cells and C cells. These two cell types are of distinct embryonic origins. During development, the thyroid diverticulum, which is derived from the endodermal epithelium of the ventral pharyngeal floor, moves caudally down along the midline and forms two lateral lobes, giving rise to follicular cells. In contrast, the ultimobranchial body develops from the fourth pharyngeal pouch and migrates to its final place of residence, the thyroid gland, to give rise to C cells. It has been shown with Connexin (Cx) 43-lacZ transgenic mice, in which β-galactosidase expression is restricted to the derivatives of neural crest, that the ectomesenchymal neural crest cells populating the pharyngeal arches surround but never enter the ultimobranchial body during its initial formation and migration (Kameda et al.,2004; Chisaka and Kameda,2005). The ultimobranchial body may interact with the ectomesenchymal neural crest cells to ensure its normal development. After fusion with the thyroid lobes, the ultimobranchial body disperses as C cells throughout the thyroid parenchyma. This developmental pattern of the mammalian ultimobranchial body differs from that in lower vertebrates, including avian species, in which the organ is situated independently of the thyroid gland for life.

The C cells synthesize and secrete calcitonin, a serum calcium-lowering hormone. In addition, during fetal development, the cells are immunoreactive for calcitonin gene–related peptide (CGRP) and somatostatin (Kameda et al.,1984; Kameda,1988). The immunoreactivity of the C cells for these neuropeptides is retained in some animal species even after birth.

Several genes appear to affect the development of the ultimobranchial body. One of these is the paired box transcription factor Pax9, which is expressed in the entire pharyngeal pouch epithelium. Targeted disruption of Pax9 results in early failure of ultimobranchial body formation in combination with the absence of the thymus and parathyroid, which are both derived from the third pharyngeal pouch (Peters et al.,1998). Thus, Pax9 seems to control the early development of the pharyngeal pouch–derived organs. Another important gene is Pax3. Mutation of Pax3, as occurs in the Splotch mouse, exhibits variable defects of the thymus, parathyroid, ultimobranchial and thyroid glands (Franz,1989). Pax3 is involved in the migration of cardiac neural crest cells that populate the pharyngeal arches 3, 4, and 6; these arches are small in Splotch embryos (Epstein et al.,2000). Thus, the Pax3 null mutation may impair the interaction between the ectomesenchymal neural crest cells and pharyngeal pouch–derived organs. Several homeobox genes, Pbx1 and Hox3 paralogs including Hoxa3, Hoxb3, and Hoxd3, have effects on the development and migration of the ultimobranchial body. In Pbx1 knockout mice, the fourth pouch and the fourth pouch–derived ultimobranchial bodies are usually absent and Pbx1 has been suggested to act together with multiple Hox proteins in the development of the caudal pharyngeal region (Manley et al.,2004). Another gene that may affect the development of the ultimobranchial body is Eyes absent(Eya)1. Eya1 null mutant mice show thyroid hypoplasia with a reduction in the number of C cells and the failure of the ultimobranchial body and thyroid lobe to fuse, suggesting that Eya1 may play a role in regulating thyroid gland maturation (Xu et al.,2002).

With regard to genes that affect the development of the follicular cells, the paired domain transcription factor Pax8 is known to be required for the formation of these cells. In Pax8 null mutant mice, follicular cell development is abolished, whereas C cells exhibiting calcitonin immunoreactivity are present as clusters (Mansouri et al.,1998). The follicular cells are responsible for thyroid hormone synthesis, and the thyroglobulin and thyroid peroxidase genes, which are essential for thyroid hormone biosynthesis, are exclusively expressed in the follicular cells. An important follicular cell gene is thyroid-specific transcription factor 1 (TTF-1) (also called Nkx2.1, Titf1, or T/ebp), which plays a role in the expression of the thyroglobulin- and thyroid peroxidase-encoding genes. In TTF-1/Nkx2.1 null mutant embryos at embryonic day (E) 12–13, both the thyroid lobe and ultimobranchial body are absent (Kimura et al.,1996).

The basic helix-loop-helix (bHLH) transcription factor Mash1 is a mammalian homologue of the achaete-scute complex (asc) gene in Drosophila. It plays a key role in the differentiation of autonomic neurons from uncommitted neural crest cells (Johnson et al.,1990; Guillemot et al.,1993). Mash1 null mutant mice lack sympathetic ganglia, noradrenergic neurons in the brain, olfactory receptor neurons, lung neuroendocrine cells, and carotid body glomus cells (Guillemot et al.,1993; Cau et al.,1997; Hirsch et al.,1998; Borges et al.,1997; Kameda,2005). Moreover, the adrenal medullary cells, which are derived from the neural crest, display an immature neuroblast-like phenotype in Mash1 null mutant mice, indicating that the cell differentiation is arrested (Huber et al.,2002).

At present, the effect of the Mash1 null mutation on the formation of the ultimobranchial body in embryos is not known. Furthermore, little is known about the molecular mechanisms involved in the differentiation of C cells from their ultimobranchial progenitors. To address these questions, we here assessed the effect of defective Mash1 function on the development of the ultimobranchial body and the formation of C cells. Thus, we subjected the thyroid glands of Mash1 homozygous null mutant mouse neonates and embryos and their wild-type and heterozygous littermates to immunocytochemical analysis using various C cell–labeling antibodies. Electron microscopy was also used. Mash1 expression was detected in the wild-type ultimobranchial body, suggesting that the Mash1-dependent signaling pathway plays an important role in C-cell development. The mutant embryos showed normal ultimobranchial body formation and migration but lacked C cells at birth.

RESULTS

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

Absence of C Cells in Newborn Mash1 Null Mutant Mice

In mammals, C cells are abundantly distributed in the thyroid glands, although the shape, size, and distribution pattern of these cells differ from species to species (Kameda,1988). In mouse thyroid glands, the cells are usually oval in shape and located in sub- and inter-follicular positions as solitary cells or small groups consisting of a few cells. They are concentrated in the central part of the thyroid lobe near the parathyroid gland. We could reliably identify the C cells in wild-type mice at birth by staining for calcitonin (Fig. 1A). Furthermore, the cells exhibited a neuroendocrine phenotype as they were intensely immunoreactive for CGRP and PGP 9.5 (Fig. 1B,C) and moderately to intensely immunoreactive for TuJ1 (Fig. 1D). Immunoreactivity for the transcription factor NeuroD was also detected (Fig. 1E). The thyroid follicles of newborn animals stored a considerable amount of colloid in the follicular lumen. Both the follicular cells and colloid were immunoreactive for thyroglobulin (Fig. 1F).

thumbnail image

Figure 1. Thyroid glands of newborn wild-type (+/+; AF) and Mash1 homozygous null mutant mice (−/−; GL) were immunostained with antibodies specific for calcitonin (CT; A, G), CGRP (B, H), PGP 9.5 (C, I), TuJ1 (D, J), NeuroD (E, K), and 19S-thyroglobulin (Tg; F, L). In the wild-type thyroid, a large number of C cells immunoreactive for calcitonin, CGRP, PGP 9.5, TuJ1, and NeuroD are observed. In the null mutant, however, cells displaying C-cell markers are absent. The follicular cells and colloid are intensely labeled by the 19S-thyroglobulin antiserum in the null mutant as well as wild-type mouse. Arrows indicate nerve fibers immunoreactive for PGP 9.5 or TuJ1. F, follicle. Scale bars = 60 μm (A–E, G–K) and 30 μm (F, L).

Download figure to PowerPoint

To determine whether the Mash1 defect influences the development of C cells, the thyroid glands of newborn Mash1 homozygous null mutant mice were immunostained with the antibodies to calcitonin, CGRP, PGP 9.5, TuJ1, and NeuroD (Fig. 1G–K). No immunoreactive cells were detectable in the mutant thyroid glands, indicating that the C cells were absent in these mice. On the other hand, the follicular cells and colloid of the mutant mice displayed the same immunoreactivity for thyroglobulin as those of the controls (Fig. 1L).

C cells of newborn wild-type mice can be clearly identified by electron microscopy because they possess numerous secretory granules (90–220 nm in diameter) that distinguish these cells from follicular cells (Fig. 2A). However, the thyroid glands of Mash1 null mutants were devoid of C cells, although the follicular cells were normal in their ultrastructural features and revealed a well-developed rough endoplasmic reticulum and many mitochondria (Fig. 2B). Thus, normal Mash1 function is necessary for the formation of C cells in mouse thyroid glands.

thumbnail image

Figure 2. Electron micrographs of the thyroid glands of litter-mate wild-type (+/+; A) and Mash1 null mutant mice (−/−; B) at birth. The wild-type thyroid shows a C cell (CC) containing a large number of secretory granules (arrows) among the follicular cells (FC). The C cell exhibits well-developed Golgi complexes (G) and has many mitochondria. The null mutant thyroid gland consists of follicular cells filled with well-developed rough endoplasmic reticulum and mitochondria, but is devoid of C cells. BC, blood capillary; F, colloid-containing follicle. Scale bars = 1.3 μm.

Download figure to PowerPoint

Mash1 Expression in the Wild-Type Ultimobranchial Body

To determine the etiology of the C-cell defect in the Mash1 null mutants, we examined the formation and development of ultimobranchial body, which arises from the fourth pharyngeal pouch and eventually combines with the thyroid lobe, after which it differentiates into C cells. In both the null mutant and wild-type embryos, an ultimobranchial rudiment displaying a follicular structure and lying medial to the fourth arch artery was observed at E 11.5 (Fig. 3A,B). It has been shown previously that the endodermal epithelium of the pharyngeal pouches is immunoreactive for PGP 9.5 (Kameda et al.,2004). We found here that the ultimobranchial rudiment (which is derived from the fourth pharyngeal pouch) also exhibits PGP 9.5 immunoreactivity at E 11.5 (Fig. 3A,B). Neuronal progenitors immunoreactive for TuJ1 and NF-160, which are markers for neuronal neural crest cells, were not detected in the ultimobranchial rudiment of either genotype at E 11.5 (Fig. 3C,D). On the other hand, a few nuclei expressing the transcription factor Mash1 began to appear in the rudiment of wild-type embryos at E 11.5 (Fig. 3E).

thumbnail image

Figure 3. Frontal sections of the pharyngeal region in litter-mate wild-type (+/+; A, CE) and Mash1 homozygous null mutant (−/−; B) embryos at E 11.5. The sections were immunostained with antibodies specific for PGP 9.5 (A, B), TuJ1 (C), NF-160 (D), or Mash1 (E). The ultimobranchial rudiment (UB) in both the null mutant and wild-type embryos is located medially, close to the fourth arch artery (A4). Both nerve fibers and endodermal epithelium, including the ultimobranchial rudiment, are immunoreactive for PGP 9.5. TuJ1- and NF-160-immunoreactive cells are not observed in the ultimobranchial rudiment. A few nuclei (arrow) immunoreactive for Mash1 begin to appear in the rudiment. A3, third arch artery; A6, sixth arch artery; DA, dorsal aorta; P, pharynx; P3, third pharyngeal pouch; T, thyroid rudiment; X, vagus nerve. Scale bars = 160 μm (A, B), 110 μm (C, D), and 60 μm (E).

Download figure to PowerPoint

At E 12.5, the ultimobranchial body in both genotypes contacted the arteries, i.e., the aortic arch or basal portion of the common carotid artery (Figs. 4A, 5A–C). At this point, a vast majority of ultimobranchial cells in the control embryos expressed immunocytochemically detectable Mash1 (Fig. 4B), unlike the cells of the homozygous null embryos (data not shown). Thus, the ultimobranchial body becomes closely associated with the great arch arteries during the period when Mash1 expression is induced vigorously.

thumbnail image

Figure 4. Frontal sections of ultimobranchial body at E 12.5 (+/+; A, B) and thyroid gland (+/+; C, D) at E 13.5, respectively, in the wild-type embryos, and the wild-type thyroid glands at E 14.5 (+/+; E) and E 15.5 (+/+, F). A, C: Hematoxylin-eosin staining. B–F: Immunostaining with the Mash1 monoclonal antibody. The ultimobranchial body (UB) is in contact with the aortic arch (AA) at E 12.5. Intense immunoreactivity for Mash1 is found in the nuclei of the ultimobranchial cells at E 12.5 and E 13.5. However, this immunoreactivity weakens at E 14.5, when the cells begin to disperse within the thyroid lobe, and disappears at E 15.5. Arrows indicate the ultimobranchial cells populating the thyroid lobes. CCA, common carotid artery; P, pharynx; PT, parathyroid gland; TYM, thymus. Scale bars = 110 μm (A), 90 μm (B, E, F), and 80 μm (C, D).

Download figure to PowerPoint

thumbnail image

Figure 5. Frontal (A) or consecutive transverse (B, C) sections of the Mash1 null mutant embryos (−/−) at E 12.5, immunostained with the PGP 9.5 or TuJ1 antibody. D, E: Consecutive frontal sections of thyroid lobe in the null mutant embryo (−/−) at E 13.0, stained with hematoxylin-eosin (D) or anti-Mash1 antibody (E). The ultimobranchial body (UB) is apposed to the artery in the E 12.5 null mutant embryos. In the E 13.0 mutant embryo, the ultimobranchial body joins with the thyroid lobe (TL) but fails to express Mash1. CCA, common carotid artery; DVG, distal vagal ganglion; E, esophagus; P, pharynx; PT, parathyroid gland; TC, trachea; TYM, thymus; X, vagus nerve. Scale bars = 140 μm (A, D, E) and 110 μm (B, C).

Download figure to PowerPoint

By E 13.0, the ultimobranchial body had become attached to the thyroid lobe in both genotypes (Fig. 5D). At E 13.5, the organ entered and became part of the portion of the thyroid parenchyma. The Mash1 expression in the wild-type ultimobranchial body was still intense at this stage (Fig. 4C,D). As expected, the mutant organ did not express Mash1 (Fig. 5D,E).

Stepwise Differentiation of Wild-Type Ultimobranchial Progenitors and Their Loss in Mash1 Null Mutants

Subsequently, in E 14.5 wild-type embryos, the ultimobranchial body began to disperse within the thyroid parenchyma as C cells. While the C cells began to express Mash1 more weakly at this stage (Fig. 4E), they also began to express the C-cell markers, CGRP, and somatostatin in addition to PGP9.5 (Fig. 6A,B). Furthermore, TuJ1 immunoreactivity was now detectable in the C cells (Fig. 6C). In contrast, the Mash1 null mutant thyroid glands were devoid of CGRP-, somatostatin-, PGP 9.5-, and TuJ1-immunoreactive cells at E 14.5 (Fig. 6E–G). At E 14.5, the primitive and small follicles storing colloid were detected in the thyroid lobes of both genotypes. In either genotype, the colloid was intensely immunoreactive for thyroglobulin (Fig. 6D,H).

thumbnail image

Figure 6. Frontal sections of the thyroid glands in wild-type (+/+; A–D) and Mash1 null mutant (−/−; E–H) embryos at E 14.5, immunostained with the CGRP (A, E), somatostatin (Som; B, F), TuJ1 (C, G), or 19S-thyroglobulin antibody (Tg; D, H). C cells immunoreactive for the neuronal markers CGRP, somatostatin, and TuJ1 appear in the wild-type thyroid lobe, whereas the cells are absent in the null mutant. Primitive follicles (arrows) storing colloid immunoreactive for 19S-thyroglobulin are vigorously formed in both genotypes. Nerve fibers (arrowhead) are intensely immunoreactive for TuJ1. Scale bars = 60 μm (A–C, E–G) and 30 μm (D, H).

Download figure to PowerPoint

At E 15.5, Mash1 expression almost disappeared (Fig. 4F) and the number of CGRP-, somatostatin-, and TuJ1-immunoreactive cells increased markedly in the wild-type thyroid glands (Fig. 7A,B). NeuroD- and calcitonin-immunoreactive cells also appeared (Fig. 7C,D). NF-160 immunoreactivity was never detected in the C cells throughout their development in the wild-type mouse (Fig. 7E). At the late stages of embryonic development, somatostatin-immunoreactive cells decreased in number and TuJ1 immunoreactivity weakened in the C cells of some embryos (data not shown). In the Mash1 null mutant thyroids, cells exhibiting the C-cell markers were not detected throughout embryonic development (Fig. 7F and data not shown).

thumbnail image

Figure 7. Frontal sections of the thyroid glands in wild-type (+/+; AE) and Mash1 null mutant (−/−; F) embryos at E 15.5, immunostained with the CGRP (A), TuJ1 (B), NeuroD (C, F), calcitonin (CT; D), or NF-160 antibody (E). In the wild-type embryo, CGRP- and TuJ1-immunoreactive cells are prominent and the cells also exhibit immunoreactivities for NeuroD and calcitonin. NF-160 immunoreactivity is never detected in the C cells. In the null mutant embryo, NeuroD-immunoreactive cells are absent. Scale bar = 60 μm.

Download figure to PowerPoint

Enhanced Apoptosis of Ultimobranchial Progenitors in Mash1 Null Mutants

In the Mash1 null mutants as well as wild-type embryos, the ultimobranchial body colonized the thyroid gland. Cells exhibiting the C-cell markers, however, did not form in the mutant thyroid glands. To test whether the lack of C cells in the null mutants is due to cell death of the progenitors, we analyzed the ultimobranchial body and thyroid gland by the TUNEL procedure. At E 11.5 and earlier stages in even wild-type embryos, many apoptotic cells were observed in the ultimobranchial body and epithelium of the fourth pharyngeal pouch (Chisaka and Kameda,2005). In both E 12.5 and E 13.0 wild-type embryos, however, few apoptotic cells were present in the ultimobranchial bodies (Fig. 8A,C). In contrast, in the Mash1 null mutant embryos at the same stages, large numbers of apoptotic cells appeared in the organs (Fig. 8B,D). Furthermore, at E 13.5, the Mash1 null mutant thyroids exhibited many apoptotic cells in the region where the ultimobranchial body invaded, whereas the wild-type thyroids did not (Fig. 8E,F).

thumbnail image

Figure 8. Frontal sections of litter-mate wild type (+/+, A, C, E) and Mash1 null mutant (−/−, B, D, F) embryos at E 12.5, E 13.0, and E 13.5, respectively, subjected to the TUNEL assay. In the null mutant embryos, extensive apoptosis is seen in the ultimobranchial bodies (UB) at E 12.5 and E 13.0, and in the thyroid region (arrows) that is colonized by the ultimobranchial body at E 13.5. In the wild-type embryos, however, only a few apoptotic cells are seen in the equivalent areas. CCA, common carotid artery; L, larynx; PT, parathyroid gland; TL, thyroid lobe; TYM, thymus. Scale bars = 80 μm.

Download figure to PowerPoint

Taken together, these results indicate that the absence of C cells in the Mash1 null mutants does not result from a defect of the ultimobranchial body formation; rather, it is due to an arrest of C-cell differentiation. Moreover, it appears that the undifferentiated ultimobranchial cells may be eliminated by apoptosis.

DISCUSSION

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

C-Cell Development Depends on Mash1 Function

In mammals, the ultimobranchial body colonizes the thyroid lobe early during embryonic development and then gives rise to endocrine C cells. Our study here showed that Mash1 function is necessary for the formation of C cells from the ultimobranchial body within the thyroid parenchyma. We found that in the Mash1 homozygous null mutant mice, the initial development of the ultimobranchial body proceeded normally as it developed from the fourth pharyngeal pouch and then joined the thyroid rudiment. Subsequently, however, the mutant thyroid glands lacked cells bearing C-cell markers, namely, calcitonin, CGRP, somatostatin, PGP 9.5, TuJ1, and NeuroD. An electron microscopic study also revealed that the thyroid glands of newborn Mash1-deficient mice lacked cells containing secretory granules, which is a characteristic for C cells. Taken together, these results suggest that the ultimobranchial cells of Mash1 null mutants are unable to progress in their differentiation program up to the level of neuroendocrine phenotype acquisition. That Mash1 is needed for the C-cell development is consistent with the intense Mash1 signal in the ultimobranchial body of wild-type embryos at E 12.5–E 13.5. This marked Mash1 expression decreased at E 14.5, which is when the ultimobranchial progenitors began to differentiate into C cells. At E 15.5, the Mash1 expression of C cells almost disappeared. This indicates that Mash1 is expressed in the C-cell precursors prior to their differentiation, which supports the notion that this gene is needed at an early stage in the C-cell lineage to initiate a differentiation program.

Mash1 Is Required for the Differentiation of Neuronal and Neuroendocrine Lineages

In the peripheral nervous system, Mash1 mRNA is transiently expressed in sympathetic, parasympathetic, and enteric precursors, and all of these autonomic sublineages are affected by the Mash1 mutation (Lo et al.,1991; Guillemot et al.,1993; Hirsch et al.,1998). In mice lacking Mash1, the localization of neural crest cells at the dorsal aorta is not blocked, but their neuronal differentiation is prevented, resulting in the absence of the sympathetic ganglia (Guillemot et al.,1993; Sommer et al.,1995). In the olfactory epithelium of the Mash1 null mutants, neuronal progenitors die at an early stage, whereas the nonneuronal supporting or glial cells are present (Guillemot et al.,1993).

A close connection between Mash1 and neuroendocrine cell differentiation has also been demonstrated. Thus, not only neural crest–derived cells but also some endodermal endocrine cells require Mash1 function for their differentiation. For example, pulmonary neuroendocrine cells express Mash1 mRNA and their formation is impaired when Mash1 is defective (Borges et al.,1997; Ito et al.,2000). Moreover, in the carotid body of the Mash1 null mutant, glomus cell formation fails, although the development of sustentacular cells is stimulated (Kameda,2005). In addition, in the adrenal medulla of Mash1 null mutants, most cells do not contain chromaffin granules and retain a very immature neuroblast-like phenotype, suggesting the arrest of chromaffin-cell development at their precursor stage (Huber et al.,2002). Thus, Mash1 may be required for the differentiation, proliferation, or survival of neuroendocrine cells during their development from precursors.

In the thyroid gland of Mash1 null mutants, a small number of calcitonin-immunoreactive cells and calcitonin mRNA expression have been reported (Lanigan et al.,1998). (Like us, they have also obtained the mutants from Dr. F. Guillemot.) This is a great contrast to our results showing no C cells and also to data of other neuroendocrine cells in the Mash1 null mutants. The pulmonary neuroendocrine cells and carotid body glomus cells are absent, and adrenal chromaffin cells are retained as an immature neuroblast-like phenotype in the Mash1 mutants (Borges et al.,1997; Huber et al.,2002; Kameda,2005). Calcitonin is an endocrine C-cell marker and its immunoreactivity appeared at E 15.5, later than the appearance of neuronal markers TuJ1, CGRP, and somatostatin, in wild-type mouse embryos. Furthermore, many apoptotic cells were detected in the ultimobranchial cells at E 12.5–E 13.5. Therefore, in the Mash1 null mutants, the C-cell progenitors may degenerate before they become differentiated endocrine cells. Mash1 mRNA expression has been reported in the adult rat thyroid gland (Clark et al.,1995). In the wild-type mouse embryos, however, Mash1 signals were transiently expressed in the ultimobranchial cells at E 12.5–E 13.5 and disappeared at E 15.5.

Cells Expressing Mash1 Acquire Neuronal Traits

Regardless of the origins of endocrine cells (i.e., whether they are from the endodermal epithelium or the neural crest), Mash1 expression appears to be associated with the induction of neuronal traits in the cells. For example, the pulmonary neuroendocrine cells derived from the endodermal epithelium express the neuronal markers CGRP, serotonin, and synaptic vesicle protein 2 and are densely innervated (Borges et al.,1997; Pan et al.,2004). Moreover, adrenal chromaffin cells are derived from the neural crest and share many neuronal properties with sympathetic neurons. In addition, like the chromaffin cells, the carotid body glomus cells are immunoreactive for TuJ1, PGP 9.5, serotonin, tyrosine hydroxylase, and neuropeptide Y (Kameda,2005). Similarly, as we showed here, the C cells in the mouse thyroid glands have a neuronal phenotype, as they are immunoreactive for TuJ1, PGP 9.5, CGRP, and somatostatin, in addition to the endocrine C-cell marker calcitonin. Furthermore, immunoreactivity for NeuroD (the bHLH transcription factor involved in the terminal differentiation of neurons) is localized in the C cells. Thus, cells expressing Mash1 acquire neuronal traits.

Mash1 Defect Does Not Influence Follicular Cell Development

In Mash1 null mutant mice, the development of follicular cells was the same as that in controls, and the colloid-containing follicles immunoreactive for thyroglobulin were normally organized. Thus, follicular cell differentiation is not affected by the C-cell defects.

NeuroD Expression of C Cells

Members of the bHLH transcription factor family play important roles in cell-fate determination and differentiation. C-cell differentiation may utilize at least two bHLH transcription factors, namely, Mash1 and NeuroD. NeuroD was first identified in neurons of the developing nervous system at the time of their terminal differentiation into mature neurons (Lee et al.,1995). It was also identified in pancreatic islet cells as an activator of insulin gene transcription (Naya et al.,1995). NeuroD is known to coordinate terminal differentiation of enteroendocrine cells by inducing cell-cycle arrest and by activating the transcription of hormone genes (Mutoh et al.,1997). It also regulates the differentiation of the POMC-expressing corticotrophs in the anterior pituitary (Poulin et al.,1997). Thus, NeuroD is an important transcription factor in the differentiation of a variety of endocrine cells.

Studies of the rodent olfactory epithelium have demonstrated the relationship between Mash1 and NeuroD in the regulation of cell differentiation. Mash1 null mutant embryos fail to express NeuroD in the olfactory epithelium (Cau et al.,1997). With regard to the mouse thyroid C cells, NeuroD expression was first detectable from E 15.5, at which point TuJ1 immunoreactivity was already intense; the NeuroD expression then persisted for life. This expression pattern suggests that NeuroD may be involved in the terminal differentiation of the C-cell precursors into endocrine cells that secrete calcitonin. NeuroD expression was lost in the thyroid gland of Mash1 null mutant mice.

Conclusions

Our study of the Mash1 null mutant thyroid glands reveals an overt deficiency in TuJ1-, CGRP-, somatostatin-, and PGP 9.5-immunoreactive cells; this deficiency is first detected at E 14.5. This shows that Mash1 is essential for C-cell development. Moreover, we found that the loss of Mash1 signaling accelerates apoptosis of the ultimobranchial progenitors populating the thyroid lobes, which may be responsible for the absence of C cells in the Mash1 null mutant thyroid glands. Thus, Mash1 enhances survival of C-cell progenitors by inhibiting apoptosis.

EXPERIMENTAL PROCEDURES

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

Animals

Wild-type, heterozygous, and homozygous Mash1 mutants were obtained from intercrosses of Mash1+/− mice that have been maintained on a CD1 background (Guillemot et al.,1993). Noon on the day of which a copulation plug was found was designated as E 0.5. The animals were genotyped by PCR analysis using the previously described primers that are specific for the wild-type and mutant alleles (Blaugrund et al.,1996). The Mash1 null mutants die at or shortly after birth.

For the histological studies, the specimens were fixed in Bouin's solution or 8% paraformaldehyde (PFA) in phosphate buffer (PB) for 24–48 hr, embedded in paraffin, and then serially sectioned along the cross, frontal, or sagittal planes at a thickness of 5 μm. Selected sections were stained with hematoxylin-eosin to help determine the morphological orientation.

Immunohistochemistry

Immunohistochemical staining was carried out by the streptavidin-biotin-peroxidase or peroxidase-anti-peroxidase (PAP) methods as described previously (Kameda,1995). The following primary antibodies were employed: rabbit polyclonal antibodies to human calcitonin, somatostatin, calcitonin gene-related peptide (CGRP), protein gene product (PGP) 9.5, NeuroD, and 19S-thyroglobulin; the monoclonal antibody TuJ1 that recognizes the neuron-specific class III β-tubulin isotype; a monoclonal antibody specific for neurofilament (NF) 160; and an anti-Mash1 monoclonal antibody. The human calcitonin, somatostatin, and canine 19S-thyroglobulin antisera were produced by our laboratory as described previously (Kameda et al.,1984; Kameda,1988,1995). Furthermore, the predilute human calcitonin antiserum was purchased from Dako (Carpinteria, CA). The TuJ1 antibody was purchased from Berkeley antibody company (Richmond, CA) and used at a dilution of 1:500. The Mash1 antibody (clone, 24B72D11.1) was purchased from BD Biosciences and was used at a dilution of 1:100. The NF-160 monoclonal antibody (clone NN18) and the anti-NeuroD antiserum (which was raised against synthetic human NeuroD1 peptide) were both purchased from Sigma (Saint Louis, MO) and used at dilutions of 1:80 and 1:200, respectively. The human PGP 9.5 antiserum was purchased from Dako and used at a dilution of 1:200. Anti-CGRP antiserum was purchased from Milab (Malmö, Sweden) and used at a dilution of 1:1,000.

Electron Microscopy

The thyroid glands of newborn mice were dissected out, placed in fixative (2.5% glutaraldehyde, 0.2 mM CaCl2, and 6% sucrose in 0.1 M cacodylate buffer, pH 7.4), and then immediately irradiated for 30–40 sec in a commercial microwave oven (500 W), in which a 300-ml water bath was located (the final solution temperature was 40 ± 2°C). The specimens were subsequently fixed for an additional 2 hr in the same fixative at 4°C, after which they were postfixed in 1% osmium tetroxide in cacodylate buffer for 2 hrs. After dehydration in ethanol and propylene oxide, the specimens were embedded in Epon by standard methods. Thin sections were made with a diamond knife and then stained with uranyl acetate and lead citrate.

TUNEL Assay for Apoptosis

To visualize apoptotic nuclei, sections were stained by using a terminal transferase dUTP-biotin nick-end labeling (TUNEL) kit according to the manufacturer's instructions (ApopDETEK in situ Cell Death Assay kit; Dako).

Acknowledgements

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

This work was supported in part by a grant (no. 17590174) from the Ministry of Education of Japan to Y.K. We thank the members of the Electron Microscope Laboratory Center, Kitasato University School of Medicine, for their expert technical contributions.

REFERENCES

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