Formation of cartilaginous foci in the central fibrous body of the heart in Syrian hamsters (Mesocricetus auratus)

Abstract

The formation of cartilage in the mammalian heart has been studied in the aortic and pulmonary valves. The chondrogenetic process that takes place in the cardiac skeleton is still unknown. The present study was designed to illustrate the ontogeny of cartilaginous foci occurring in the central fibrous body of the Syrian hamster (Mesocricetus auratus) heart. Hearts from 472 animals aged 0–708 days were examined using histological, histochemical and immunohistochemical techniques. Cartilage was present in the central fibrous body of 118 (25%) specimens. A further 104 hamsters were used for the detection of calcific deposits in the central fibrous body. Six (5.8%) showed calcified cartilage. The first sign related to chondrogenesis was the presence of small groups of cells embedded in a type II collagen-positive extracellular matrix. These cellular groups, which can appear as early as 2 days after birth, differentiate into hyaline cartilage or, less frequently, into fibrocartilage. The highest production of cartilaginous foci takes place between days 40 and 80. Thereafter, formation of new foci is uncommon. This indicates that appearance of cartilage in the central fibrous body of the heart is not a consequence of cardiac aging. The cartilaginous foci seem to act as pivots resisting mechanical tensions generated during the cardiac cycle. Deposition of calcium in the extracellular matrix of the foci can be regarded as a reinforcement of the cartilaginous tissue.

Introduction

Cartilaginous foci regularly form in cardiac fibrous areas of several mammalian species. The presence of such foci has been reported in rodents, lagomorphs, carnivores, proboscideans, perissodactyls and artiodactyls (see Matumoto, 1938; Sans-Coma et al. 1994; Egerbacher et al. 2000 for extensive reviews of the literature). The cartilaginous tissue mainly appears in the right fibrous trigone and along the attachments of the aortic valve leaflets to their supporting sinuses (Matumoto, 1938; Hueper, 1939; Hollander, 1968; Kelsall & Visci, 1970; Barone, 1972; Sans-Coma et al. 1994; López et al. 2004). Less frequent is the formation of cartilage in the left fibrous trigone, membranous interventricular septum and base of the pulmonary valve (Matumoto, 1938; López et al. 2001).

The occurrence of cardiac bone has been reported in the otter (Egerbacher et al. 2000), dog (James & Drake, 1968), wolf (Matumoto, 1938), elephant (see Matumoto, 1938), horse (Matumoto, 1938) and several artiodactyls (Retterer & Lelièvre, 1912a, b; Matumoto, 1938; Bescol-Liversac, 1947;James, 1965; Malik et al. 1972; Mia, 1973; Frink & Merrick, 1974). Bones are located in the right fibrous trigone and, less frequently, in the left fibrous trigone and membranous portions of the interatrial and interventricular septa. Occasionally, small bony structures have been detected in the mitral, tricuspid and aortic valves (Matumoto, 1938; James, 1965; Frink & Merrick, 1974). The cardiac bones are believed to form by endochondral ossification (Matumoto, 1938; Bescol-Liversac, 1947; Frink & Merrick, 1974; Egerbacher et al. 2000), although intramembranous ossification has also been proposed as a possible differentiation mechanism (Retterer & Lelièvre, 1912a). Chondro-osteoid metaplasia has been described in sheep hearts (Bhagawan et al. 1978).

Much of the early work on cardiac cartilage was concerned with the location, incidence, size, shape, histological features and function of the cartilaginous deposits occurring in adult animals. Chondrogenesis in the mammalian heart has been studied in the aortic (López et al. 2004) and pulmonary (López et al. 2001) valves of Syrian hamsters (Mesocricetus auratus), whereas the formation of cartilage in the so-called cardiac fibrous skeleton remains unknown. This is a complex framework of dense connective tissue, consisting of (1) the area of aortic-to-mitral fibrous continuity, (2) the left fibrous trigone and (3) the central fibrous body of the heart, which is composed of the right fibrous trigone and the membranous interventricular septum.

An ongoing investigation of the histological features of the cardiac skeleton in Syrian hamsters showed that a considerable proportion of individuals displayed cartilaginous deposits in the central fibrous body of the heart. On this basis, we conducted a study to illustrate the ontogeny of such deposits. Our ultimate goals were (1) to gather information about the appearance, location and structural features of the cartilage, (2) to gain insight into the factor or factors implicated in their formation and (3) to compare the chondrogenetic process that takes place in the cardiac skeleton with that occurring in the cardiac semilunar valves. The study was done in young and adult hamsters, which were examined using histological, histochemical and immunohistochemical techniques.

Materials and methods

The sample studied consisted of 576 Syrian hamsters (270 male, 306 female) aged 0–708 days. They were housed in polypropylene cages in a room in which both temperature and photoperiod were controlled. Commercial mouse food (UAR/Panlab s.l. A.04) and water were given ad libitum, starting at weaning. There was no known exposure of the animals to teratogenic agents. All hamsters were handled in compliance with international policies for animal care and welfare. They were killed by overdosing with chloroform or with carbon dioxide at a concentration of 75%, delivered into a chamber. Hearts (n = 272) were examined using histological, histochemical and immunohistochemical techniques for light microscopy. A whole-mount immunostaining technique to visualize type II collagen was applied in a further 200 specimens. The remaining 104 hearts were used for the detection of calcific deposits.

Histological and histochemical techniques for light microscopy

The hearts were removed after perfusion with 0.02 m phosphate-buffered saline (pH 7.3),fixed by immersion in Bouin's solution (ratio of fixative to tissue volume = 80 : 1) and embedded in Paraplast (Sigma Chemical Co., UK). Transverse, longitudinal or sagital sections, serially cut at 10 µm for light microscopy, were stained with haematoxylin–eosin or Mallory's trichrome stain for a general assessment of the cardiac tissue structure, with Weigert–van Gieson stain for the detection of elastin, or with 0.05% alcian blue 8GX in 0.05 m acetate buffer (pH 5.8) plus 0.65 m magnesium to identify sulphated glycosaminoglycans (Scott & Dorling, 1965). Another stain applied was picrosirius for the detection of collagen (Junqueira et al. 1979).

Immunohistochemical techniques for light microscopy

The removed hearts were washed in phosphate-buffered saline and fixed by immersion in MAW fixative (methanol–acetone–water, 2 : 2 : 1) or in Bouin's solution (ratio of fixative to tissue volume = 80 : 1).The specimens were embedded in Paraplast, and transversely cut at 10 µm. Sections were stained with monoclonal antibody to proliferating cell nuclear antigen (PCNA) (Sigma clon PC10) to detect proliferating cells (Hall et al. 1990; Waseem & Lane, 1990), or with monoclonal antibody CIIC1 (Developmental Studies Hybridoma Bank, University of Iowa), which is specific for type II collagen. This latter technique was used because expression of type II collagen is cartilage-specific (Miller & Matukas, 1969; Miller, 1976; Kosher, 1983; Hall & Miyake, 1992, 1995 ), even though type II collagen is also synthesized by several non-chondrogenic cell types (see Kosher, 1983; Swiderski et al. 1994 for reviews of the literature).

The sections were dewaxed in xylene, dehydrated in an ethanolic series, and washed in Tris-phosphate-buffered saline (TPBS, pH 7.8). For the detection of type II collagen, the tissues were digested for 30 min – 1 h at 37 °C with 0.5% papain in phosphate buffer (pH 4.7).

Endogenous peroxidase activity was quenched by incubation with 3% hydrogen peroxide in TPBS for 30 min. After washing in TPBS, non-specific binding sites were saturated for 30 min with 10% sheep serum and 1% bovine serum albumine in TPBS (SB) for staining with CIIC1, or with the same solution plus 0.5% Triton X-100 (SBT) for staining with PCNA. Sections were then incubated overnight at 4 °C in the primary antibody diluted in SB when staining with CIIC1 or in SBT when staining with PCNA. Control slides were incubated in SB or SBT only.

After incubation, the sections were washed in TPBS (3 × 5 min), incubated for 1 h at room temperature in biotin-conjugated anti-mouse IgG (Sigma) diluted 1 : 100 in TPBS, washed again and incubated for 1 h in ExtrAvidin-peroxidase conjugate (Sigma) diluted 1 : 150 in TPBS. Peroxidase activity was developed with Sigma Fast® 3,3′-diaminobenzidine tablets, according to the supplier's recommendations. In several cases, the sections were counterstained with haematoxylin.

Type II collagen whole mount immunostaining technique

Heart portions containing the semilunar valves and the whole cardiac skeleton were transferred to Cornwell centrifuge tubes, fixed by immersion in Bouin's solution, washed with TPBS and permeated for 15 min in acetone at −20 °C. After washing in TPBS, the specimens were immersed for 30 min in a 3% Triton X-1000 solution in TPBS, and washed again in TPBS. Thereafter, the tissues were digested with 10 µg mL⁻¹ proteinase K for 15 min and papain for 15 min, washed in TPBS and disgested with 0.5% papain in phosphate buffer (pH 4.7) for 4 h at 37 °C. Endogenous peroxidase activity was quenched for 1 h by incubation in 3% hydrogen peroxidase in TPBS. After washing with TPBS, non-specific binding sites were saturated for 2 h with SB. Finally, the specimens were processed following the protocol used for the detection of type II collagen in tissue sections, starting from incubation with the primary antibody.

Alizarin red S in toto staining technique

This technique was used for the specific detection of calcific deposits in the cardiac skeleton,based on the previous confirmation, by means of X-ray diffraction,that calcium was present in Alizarin red S-positive cartilage located in aortic valves of Syrian hamsters (data not shown).The heart was removed and transferred to Ringer's solution.Heart portions containing the semilunar valves and the whole cardiac skeleton were fixed by immersion in 10% neutral formalin buffered with magnesium carbonate (ratio of fixative to tissue volume = 80 : 1) and stained with Alizarin red S, according to the protocol described by Richmond & Bennet (1938). Whenever calcium deposition was detected, the specimen was embedded in Paraplast and transversely cut for examination with a light microscope.

Statistical methods

The the χ²-test was used. A probability of 0.05 or less was required as evidence of a significant difference.

Results

Cartilaginous foci were present in the right fibrous trigone and/or membranous interventricular septum of 118 (25%) of the 472 hamsters examined by using histological, histochemical and/or immunohistochemical techniques for light microscopy or by means of the type II collagen whole mount immunostaining technique. Some foci were unequivocally located in one of these two components of the cardiac skeleton. The other foci extended along both components. Hence, we use the term central fibrous body to designate the whole cardiac region in which we detected the cartilaginous tissue.

In the hamsters used in this study, no sex differences were observed with regard to the occurrence of cartilage in the central fibrous body. Therefore, male and female data were pooled.

The first sign of cartilage formation was the presence of a small group of cells embedded in a type II collagen-positive extracellular matrix (Fig. 1A). This occurred in two (5%) of the 40 hamsters aged 2 days (Table 1). In both cases, the cellular group was located in the right fibrous trigone. Type II collagen-positive cellular groups of similar or slightly larger size were observed in one (2.9%) of 34 animals aged 3 days and two (6.7%) of 30 animals aged 4–10 days.

Fig. 1.

Cartilaginous foci in the fibrous central body of Syrian hamster hearts.(A) Transverse section of the heart from a hamster aged 2 days. Type II collagen immunostaining, counterstained with haematoxylin–eosin. The arrows point to a small cellular group, located in the right fibrous trigone, surrounded by a type II collagen-positive extracellular matrix. (B) Transverse section of the heart from a hamster aged 23 days. Type II collagen immunostaining, counterstained with haematoxylin–eosin. A type II collagen-positive cellular group is present in the right fibrous trigone. The chondrocytes of the central part are larger than those placed in the periphery of the cartilaginous deposit.(C) Transverse section of the heart from a hamster aged 20 days. PCNA immunostaining. The arrows indicate proliferating chondrocytes. (D) Type II collagen whole mount immunostaining of the aortic valve and central fibrous body of the heart of a hamster aged 121 days. A cartilaginous focus (arrow) is placed in the central fibrous body. Several cartilaginous deposits (arrowheads) can be seen along the attachments of the dorsal and right aortic valve leaflets to their supporting aortic sinuses. (E) Type II collagen whole mount immunostaining of the aortic valve and central fibrous body of the heart of a hamster aged 67 days. The white arrow points to a cartilaginous focus located in the membranous interventricular septum. The black arrow indicates another focus placed in the right fibrous trigone. Cartilaginous deposits are also present along the attachments of the dorsal and right aortic valve leaflets to their supporting aortic sinuses (arrowheads). (F) Transverse section of the heart from a hamster aged 156 days. Picrosirius stain. A nodular-shaped hyaline cartilage is present in the right fibrous trigone. DS, dorsal aortic sinus; RS, right aortic sinus. Scale bars = A, B, 50 µm; C, F, 100 µm; D, E, 300 µm.

Table 1.

Cartilaginous deposits in the fibrous central body of the heart of Syrian hamsters, according to the age of the animals

Age (days)	n	n cart	%
0	0	40	0
1	0	34	0
2	2	40	5.0
3	1	34	2.9
4–10	2	30	6.7
11–40	11	60	18.3
41–80	32	58	55.2
> 80	70	176	39.8

n= number of specimens examined; n cart = number of specimens with cartilaginous deposits in the central fibrous body of the heart.

Eleven (18.3%) of the 60 hamsters aged 11–40 days possessed cartilaginous deposits in the central fibrous body (Table 1). In one case there were two deposits; in all other cases a sole deposit was present. The deposits varied in size and histological condition. Some were composed of a limited number of cells surrounded by a type II collagen-positive matrix. They were embedded in a remarkably well-developed fibrous cellular matrix, a fact that hampered their detection by means of conventional histological techniques. The other cartilaginous deposits consisted of a higher amount of chondrocytes, surrounded by an extracellular matrix rich in glycosaminoglycans. In most cases, the chondrocytes located in the central part were obviously larger than those placed in the periphery of the deposit (Fig. 1B). Immunolabelling with PCNA antibody demonstrated that chondrocytes were proliferating (Fig. 1C).

Overall, 102 (43.6%) of 234 hamsters aged more than 40 days had cartilage in the central fibrous body. The cartilaginous tissue was detected by means of the type II collagen whole mount immunostaining technique in 85 specimens. In the remaining 17, it was observed in tissue sections. As shown in Table 1, the cartilaginous deposits occurred in 32 (55.1%) of 58 hamsters aged 41–80 days and in 70 (39.7%) of 176 aged 81 days and older. However, this difference was not statistically significant (P > 0.20; χ²-test).

In the specimens examined there was usually a single cartilaginous focus in the central fibrous body (Fig. 1D). Existence of two foci (Fig. 1E) must be regarded as an uncommon event. As revealed by the tissue sections, most cartilages were hyaline (Fig. 1F). They showed a nodular (Fig. 1F), elipsoidal or conal shape, and were composed of chondrocytes surrounded by a type II collagen-positive matrix that stained metacromatically with haematoxilyn. In some cases, the chondrocytes were of similar size. In other cases, the central part of the cartilaginous focus was occupied by hyperthrophic chondrocytes. Several cartilages had a thin perichondrium composed of collagen fibres that ran in a circumferential direction and contained flattened cells. In most cases, however, the perichondrium was lacking, and the cartilaginous tissue merged with the surrounding connective tissue.

Fibrocartilage was present in the central fibrous body of a few specimens. The tissue consisted of a series of cells embedded in a type II collagen-positive matrix and were contained within a meshwork of collagen fibres and elastin.

Calcific deposits occurred in six (5.8%) of the 104 specimens stained in toto with Alizarin red S. These animals were aged 5 months and older. In all cases, a single calcified deposit was located in the right fibrous trigone (Fig. 2A). As revealed by the tissue sections, the mineral had precipitated in the matrix of cartilaginous tissue (Fig. 2B).

Fig. 2.

Calcification of cardiac cartilage located in the central fibrous body of the Syrian hamster heart. (A) Alizarin red S in toto staining of the aortic valve and fibrous central body of the heart of a hamster aged 520 days. An Alizarin red S-positive focus (arrow) is present in the right fibrous trigone. The arrowhead indicates the ostium of the right coronary artery. (B) Transverse section of the Alizarin red S-positive focus shown in A. Note that the calcium is deposited in the extracellular matrix of cartilaginous tissue. DS, dorsal aortic sinus; RS, right aortic sinus. Scale bars = A, 500 µm; B, 50 µm.

Discussion

Matumoto (1938) was the first to state that the cartilage occurring in mammalian heart develops after birth. He based his assumption on the fact that he detected no cartilaginous tissue in embryonic hearts from laboratory rats. Hollander (1968) reported that in R-Amsterdam rats, cartilaginous tissue appears in the bottom of the dorsal aortic sinus from the second week of life. More recently, it has been shown that in the aortic (López et al. 2004) and pulmonary valves (López et al. 2001) of Syrian hamsters, the development of cartilage can start from the first day of life. The present findings prove that in the central fibrous body of the Syrian hamster heart, chondrogenesis can begin from the second day after birth. As in the cardiac semilunar valves (López et al. 2001, 2004), the first evidence of the development of a cartilaginous focus is the appearance of small groups of cells embedded in a type II collagen-positive extracellular matrix. Thereafter, these cellular groups increase in size, due in part to cell proliferation, and differentiate into hyaline cartilage or, less frequently, into fibrocartilage.

The presence of cartilage or bone in the left fibrous trigone has been reported in rabbits and horses (Matumoto, 1938). The absence of cartilaginous deposits in this portion of the cardiac fibrous skeleton in the specimens examined remains unexplained.

In reptilian (López et al. 2003) and avian (Stiefel, 1926; Matumoto, 1938; Tsusaki et al. 1956; Sumida et al. 1989; López et al. 2000) hearts, chondrogenesis starts, during embryonic life, with the formation of prechondrogenic condensations, composed of a considerable number of loosely packed mesenchymal cells, which can be well recognized in tissue sections by means of conventional histological techniques. The appearance of such cellular aggregations suggests the establishment of cell to cell contacts, a fact which is considered to be the first step of the process that regulates the early stages of cartilage differentation (Thorgood & Hinchliffe, 1975; Ede, 1983; Tachetti et al. 1992; Cancedda et al. 1995). In contrast, we detected no conspicuous prechondrogenic condensation in the central fibrous body of the heart of the present Syrian hamsters, as is also the case for the aortic (López et al. 2004) and pulmonary (López et al. 2001) valves of this rodent species. This points to the possibility of a differention process in chondrocyte precursors independent of cell aggregation. Further studies, using approriate techniques, are needed to verify this supposition.

Most of the cartilaginous foci occurring in aortic valves of Syrian hamsters form within the first 40 days of life, i.e. when the histogenesis of the valves takes place (López et al. 2004). In contrast, the highest production of cartilaginous deposits in the central fibrous body of the heart corresponds to the period between days 40 and 80. As suggested by the statistical results, appearance of new deposits after this interval of time seems to be an uncommon event. In spite of this timing difference between the formation of cartilage in the aortic valve and fibrous central body, it can be concluded that in the Syrian hamster, formation of cardiac cartilage is not a consequence of cardiac aging.

Cartilages occurring in the cardiac semilunar valves of birds originate from neural crest cells. This assumption relies on the findings of Sumida et al. (1989) in quail-chick chimera embryos belonging to the developmental stage 45 of Hamburger & Hamilton (1951). In such embryos, Sumida et al. (1989) detected a cluster of neural crest-derived cells between the proximal aorta and the pulmonary trunk which had differentiated into cartilage and surrounding connective tissue. Currently, this is the sole experimental evidence that the precursors of cardiac chondrocytes are neural crest-derived elements. Any other conjecture in the literature about such an origin of the cardiac cartilages is based on the fact that the cartilaginous tissue appears in the heart where neural crest cells are present during embryonic life or later, then as undifferentiated cells. This is the case for the cartilage that forms (1) in the cardiac outflow tract and fibrous portion of the horizontal ventricular septum in turtles (López et al. 2003), (2) in the attachments of the conal valves to the longitudinal ridges of the distal portion of the cardiac outflow tract, the bulbus arteriosus, in teleost fishes (Blanco et al. 2001) and (3) in the attachments of the semilunar valve leaflets to their supporting sinuses in mammals (López et al. 2001, 2004). In this context, it should be noted that in teleosts, cells from the neural crest migrate through the arterial pole of the heart, reaching the ventricle and atrium where a certain number of such crest cells adopt a cardiomyocyte cell lineage (Sato & Yost, 2003). The cardiac areas reached by the migrating neural crest cells in most vertebrate groups are still uncertain. However, the observations in fish strongly suggest that the neural crest has decreased its contribution to the formation of the cardiac chambers during vertebrate evolution. In mammals, undifferentiated cells, presumably from the neural crest, have been recently observed in the developing fibrous cardiac skeleton after birth, i.e. in the heart of a 9-week-old adult mouse (Jiang et al. 2000). Moreover, it has been shown that a subpopulation of crest cells takes part in the closure of the embryonic ventricular septum (Waldo et al. 1999). The data reported herein allow no conclusion to be made on the morphogenetic origin of the cartilaginous foci that appear in the fibrous central body of the mammalian heart. Yet, the presence of cartilage in this cardiac area, together with the findings in mice cited above, suggest an area for further investigation, namely the possible population of the presumptive fibrous skeleton of the mammalian heart by neural crest cells through the arterial pole.

The functional significance of cartilaginous and/or osseus deposits in the vertebrate heart is still an open question. In teleost fishes, the cartilaginous foci located in the bulbus arteriosus seem to act as supports in the context of the mechanical forces to which the conal valves are subjected during the cardiac cycle (Blanco et al. 2001). In amphibians, cartilage forms in different parts of the cardiac outflow tract subject to intense mechanical action (Matumoto, 1938). In several reptilian species, cartilaginous deposits are regularly present in the interaortic and inteventricular septa and/or in the cardiac outflow tract, i.e. at sites where the cartilages act as competent pivots resisting mechanical tension (Matumoto, 1938; White, 1956, 1959; Kashyap, 1959; Webb, 1979; Young, 1994; López et al. 2003). In birds and mammals, cartilaginous deposits appear in the sinus boundaries of the cardiac valves (see the literature cited above) to which large stresses, generated in the leaflets or cusps during the cardiac cycle, are distributed (Broom, 1988). Moreover, cartilage and/or bones develop in the fibrous trigones and/or in the membranous components of the interatrial and interventricular septa, all of which provide insertion for the myocardium and rigidity during heart performance. In this context, the deposition of calcium in the extracellular matrix of the cartilage located in the central fibrous body of the heart of the present hamsters can be regarded as a reinforcement of the cartilaginous tissue, and not as a degenerative process. This agrees with the viewpoint of Egerbacher et al. (2000) concerning the calcification of cartilage matrix material located in the heart skeleton of the otter.

In the chick, the topographic distribution of the cartilaginous tissue in the cardiac semilunar valves is very similar to that of the tenascin during valvular morphogenesis (García Martínez et al. 1990), and the distribution of tenascin during valvulogenesis is associated with zones specialized in bearing mechanical loads (García Martínez et al. 1990). This, together with the preceding data, suggest that mechanical action may be a primary factor in the formation of cardiac cartilage. However, this needs further verification. In any case, we presume that mechanical stimulation plays at least an essential role in the differentiation process of the anticipated cardiac cartilaginous tissue into hyaline cartilage. Subsequent ossification of the cartilaginous tissue may be also mediated by mechanical action. Indeed, mechanical stimulation is an important factor in shaping the architecture of bone (Duncan & Turner, 1995; Turner & Pavalko, 1998; Martin, 2000; Weyts et al. 2003).

Nonetheless, the presence of cartilage or bone is not a requisite for normal cardiac performance in vertebrates. Cartilage seems to be lacking in the hearts of cyclostomes and elasmobranchs (Matumoto, 1938). The occurrence of cardiac cartilage in actinopterygians is an unusual event, and its incidence in amphibians is relatively low (Matumoto, 1938). In reptiles, birds and mammals, the presence of cardiac cartilage, with or without subsequent ossification, varies widely between species and between members of the same species (see López et al. 2000, 2004). In humans, cartilage-like structures or fibrocartilages have been reported in the right fibrous trigone of the heart (Mönckeberg, 1902). However, deposits of hyaline cartilage in the cardiac skeleton have only been recorded in certain pathological conditions (Topham, 1906; Henke & Lubarsch, 1924;McConell, 1970; Ferris & Aherne, 1971; Gittenberger-de Groot, 1972; Seemayer et al. 1973; Bell & Greco, 1981; Groom & Starke, 1990).

The preceding facts lead to the question as to why some of the cardiac cartilage precursors (i.e. cells from the neural crest) differentiate into chondrocytes, whereas others do not. This obviously needs further investigation, including a thorough study of the mechanisms implicated in the differentiation of neural crest-derived cells into chondrocytes. Recently, it has been reported that Sox9 is required for the determination of the chondrogenic lineage in neural crest cells (Mori-Akiyama et al. 2003). However, most molecular mechanisms underlying the determination of cell fate in neural crest cells remain unclear (Mori-Akiyama et al. 2003).