In multicellular organisms, intercellular junctions are important structures for tissue architecture and physiological function. There are three main types of intercellular junctions; tight, adherens, and gap junctions. Tight junctions have been thought to be responsible for the intercellular sealing that plays a critical role in the movement of solutes, ions, and water. Tight junctions appear by transmission electron microscopy as a series of focal contacts between the plasma membranes of adjacent cells and as a set of continuous, circumferential networks of intramembranous particle strands. Tight junctions have been found not only in simple epithelia but also stratified epithelium. Tight junctions are also believed to play a critical role in regulating the development and normal function of cells (Balda and Matter, 1998; Brizuela et al., 2001; Hashizume et al., 2004; Saathoff et al., 2004; Haworth et al., 2005). Claudins are transmembrane proteins that have been identified as components of tight junction strands and are essential molecules for forming the tight junctions (see review; Tsukita and Furuse, 2002). There are over 20 different claudins known in mouse, and it is reported that single cells can coexpress more than two claudins (Furuse et al., 1999; Wang et al., 2003; Holmes et al., 2006). Claudins display varied tissue distribution, including coexpression, which suggests that their differential expression may explain the variable permeability observed in different tissues.
The tooth is an organ that develops as a result of sequential and reciprocal interactions between the oral epithelium and neural crest-derived mesenchyme. The first morphological sign of tooth development is an epithelial thickening. The thickened epithelium progressively takes the form of the bud, cap, and bell configurations and differentiation proceeds. Subsequently, epithelial cells and mesenchymal cells (dental papilla) differentiate into enamel-secreting ameloblasts and dentin-secreting odontoblasts, respectively.
Cell–cell contacts in developing tooth germs may play a critical role in morphogenesis and cell differentiation. Changes in the distribution of adherens junctions, desmosomes, and hemidesmosomes have been reported during tooth development (Fausser et al., 1998). Tight junctions were also observed in odontoblasts at later stages of tooth development (Arana-Chavez and Katchburian, 1997).
The expression patterns of 11 claudin genes (claudin1 – claudin11) were mapped by in situ hybridization in mouse embryonic mandibular first molar tooth germs between E12.5 and newborn. This period encompasses tooth development from the earliest formation of the epithelial thickening to the onset of cytodifferentiation. Six claudins at the thickening stage, eight claudins at the bud stage, seven claudins at the cap stage, and five claudins at the early bell stage showed spatial-specific expression in tooth development.
RESULTS AND DISCUSSION
Initiation of Tooth Development (E12.5)
Thickening of the oral epithelium takes place from embryonic day (E) 12. At E12.5, claudin4 (Fig. 1D) and claudin6 (Fig. 1F) were expressed in the thickening tooth epithelium. The expression of both claudin3 (Fig. 1C) and claudin7 (Fig. 2A) was restricted to a single cell layer facing the oral cavity in the thickening tooth epithelium. Claudin1 (Fig. 1A) and claudin10 (Fig. 2D) showed weak expression in the epithelium. Claudin5 expression was found in mesenchyme (Fig. 1E). Claudin2, claudin8, claudin9, and claudin11 showed no expression in the mandible epithelium or mesenchyme (Figs. 1B, 2B,C,E).
Bud Stage (E13.5)
By E13.5, the tooth epithelium has invaginated into underlying mesenchyme to form the epithelial bud. The bud epithelium consists of basal epithelium and stellate reticulum. The expression of claudin1 (Fig. 1G) was restricted to the buccal basal epithelium of the bud epithelium, whereas claudin10 (Fig. 2I) showed localized expression in the lingual basal epithelium of the bud epithelium. The expression of claudin6 was observed in the entire budding epithelium (Fig. 1L). The expression of claudin3 (Fig. 1I) and claudin7 (Fig. 2F) were restricted to the most orally located regions of the stellate reticulum of the bud epithelium, whereas claudin4 showed localized expression in the entire stellate reticulum (Fig. 1J). The expression of claudin5 was found in presumptive endothelial precursor cells (Fig. 1K; Morita et al., 1999). Claudin11 was expressed outside of the dental follicle and dental papilla in what are probably osteogenic centers (Fig. 2K; Bronstain et al., 2000). Claudin2, claudin8, and claudin9 expression was undetectable (Figs. 1H, 2G,H).
Cap Stage (E14.5)
By E14.5, the bud basal epithelium develops into the internal and the external (outer) enamel epithelium, and the mesenchyme develops into the dental papillae and the dental follicle. Claudin1 was expressed in the buccal basal epithelium with weaker expression in the enamel knots (Fig. 1M), while claudin10 showed restricted expression in lingual epithelium (Fig. 2N). Expression of claudin4 (Fig. 1P) was detected in most of the epithelial enamel organ but was absent from the inner enamel epithelium. Claudin6 was expressed in entire tooth epithelium (Fig. 1R), whereas claudin7 was expressed weakly in the epithelium (Fig. 2K). The expression of claudin5 (Fig. 1Q) and claudin11 (Fig. 2O) was found in mesenchyme. Claudin2, claudin3, claudin8, and claudin9 showed no expression in the tooth germ (Figs. 1N,O, 2L,M).
The terminal differentiation of dentin-forming odontoblasts from dental papilla cells and the enamel-forming ameloblasts from the internal epithelium occurred between E18 to postnatal day (P) 0. Claudin1 was expressed in the outer enamel epithelium and stratum intermedium (Fig. 1S), whereas claudin2 showed weak expression in ameloblasts (Fig. 1T). The expression of claudin4 was detected in the stratum intermedium (Fig. 1V). Claudin10 was expressed in odontoblasts and stratum intermedium (Fig. 2S). The expression of claudin11 was found in the dental follicle (Fig. 2T). Claudin3, claudin5, claudin6, claudin7, claudin8, and claudin9 were not expressed at this stage (Figs. 1U,W,X, 2P–R).
The expression of claudins during murine tooth development is summarized diagrammatically in Figures 3 and 4. All claudins in incisor development showed similar expression patterns to molar development (data not shown).
Developing tooth germs showed overlapping expression patterns of nine claudins. Tight junction strands are formed by several claudins as heteropolymers and claudin molecules adhere to each other in both a homotypic and heterotypic manner. The combination and mixing ratios of claudins within individual paired strands determines their tightness and selectivity (Tsukita and Furuse, 2002).
It has been reported that claudins are also expressed during submandibular gland development (Hashizume et al., 2004). We found expression of claudins in whisker follicles (data not shown). Of interest, there are similarities in the expression patterns of claudins between these different ectodermal appendages. For example, claudin2 expression could not be detected, whereas claudin3 and claudin7 are expressed in similar regions in tooth germs, whisker follicles, and submandibular glands. The expression of claudins has also been noted in development of other tissues which develop from epithelial–mesenchyme interactions such as kidney, lung, and otic vesicles (Reyes et al., 2002; Meyer et al., 2004; Haworth et al., 2005; Simard et al., 2005; Abuazza et al., 2006). These findings may suggest that claudins play a critical role in organogenesis through epithelial–mesenchme interactions.
Another component of tight junctions, occludin is also expressed during tooth development (Unda et al., 2003). It is believed that tight junctions participate in the generation/maintenance of cellular polarization (Anderson, 2001; Brizuela et al., 2001). At the bud stage in particular, six claudins are expressed in unique and overlapping domains (Fig. 5). Claudin1 showed localized expression in the buccal basal epithelium of the tooth bud epithelium, whereas claudin10 expression was restricted to the lingual basal epithelium of tooth bud epithelium. Claudin3, claudin4, and claudin7 showed localized expression in stellate reticulum. Other molecules show some similarities in expression patterns to the claudins in the tooth development. Bmp4 is detected in the mesenchyme around the epithelial buds, with somewhat stronger expression at the buccal side of the bud, whereas Lunatic fringe is restricted to the lingual basal epithelium of tooth bud epithelium and Notch1 is expressed in stellate reticulum (Mitsiadis et al., 1995; Aberg et al., 1997; Mustonen et al., 2002). Interactions between claudins and Bmp's have been reported in embryonic epithelium (Turksen and Troy, 2001). Thus the specific expresssion of claudins and occludins in different epithelial cells of the tooth bud may have an important role in differentiation of these cells.
Production and Analysis of Mice
CD1 mice were used. Day E0 was taken to be midnight before finding a vaginal plug. To accurately assess the age of embryos, somite pairs were counted and the stage was confirmed using morphological criteria, e.g., relative sizes of maxillary and mandibular primordia, extent of nasal placode invagination, and the size of limb buds. Embryo heads were fixed in 4% buffered paraformaldehyde, wax embedded, and serially sectioned at 7 μm. Sections were split over 5 to 10 slides and prepared for histology or radioactive in situ hybridization. Decalcification using 0.5 M ethylenediaminetetraacetic acid (EDTA, pH 7.6) was performed after fixation of newborn mice.
In Situ Hybridization
Radioactive section in situ hybridization using 35S-UTP radiolabeled riboprobes was carried out according to Tucker et al. (1998). The radioactive antisense probes were generated from mouse cDNA clones that were gifts from several laboratories: claudin1, claudin2, claudin4, and claudin6 (S. Tsukita) and claudin3, claudin5, claudin7, and claudin8 (N. Sawada; Chiba et al., 2003). For making claudin9, claudin10, and caludin11 probes, reverse transcriptase-polymerase chain reaction (RT-PCR) of total RNA from mandible primordia with submandibular glands of E16.5 embryo was carried out and RT-PCR products were cloned into pGEM-T Easy vector (Promega) and subjected to automated DNA sequence analysis. PCR primers were designed according to Turksen and Troy (2002).
We thank Kim Haworth for critically reading of the manuscript.