|Home | About | Journals | Submit | Contact Us | Français|
TBX1 is a principal candidate gene for DiGeorge syndrome, a developmental anomaly that affects the heart, thymus, parathyroid, face, and teeth. A mouse model carrying a deletion in a functional region of the Tbx1 gene has been extensively used to study anomalies related to this syndrome. We have used the Tbx1 null mouse to understand the tooth phenotype reported in patients afflicted by DiGeorge syndrome. Because of the early lethality of the Tbx1−/− mice, we used long-term culture techniques that allow the unharmed growth of incisors until their full maturity. All cultured incisors of Tbx1−/− mice were hypoplastic and lacked enamel, while thorough histological examinations demonstrated the complete absence of ameloblasts. The absence of enamel is preceded by a decrease in proliferation of the ameloblast precursor cells and a reduction in amelogenin gene expression. The cervical loop area of the incisor, which contains the niche for the epithelial stem cells, was either severely reduced or completely missing in mutant incisors. In contrast, ectopic expression of Tbx1 was observed in incisors from mice with upregulated Fibroblast Growth Factor signalling and was closely linked to ectopic enamel formation and deposition in these incisors. These results demonstrate that Tbx1 is essential for the maintenance of ameloblast progenitor cells in rodent incisors and that its deletion results in the absence of enamel formation.
DiGeorge syndrome (DGS) or velocardiofacial syndrome, also known as 22q11 deletion syndrome (22q11DS), is a common haploinsufficiency developmental disorder affecting approximately 1 in 4000 live births. The great majority of DGS cases result from de novo mutations, but this syndrome can also be inherited in an autosomal dominant fashion. Patients with DGS display a variety of clinical manifestations including abnormalities of the cardiac outflow tract and aortic arch, thymus aplasia or hypoplasia, hypoparathyroidism, learning and/or behavioral disorders, and a range of malformations within the craniofacial region that includes an abnormal face, mandibular retrognathia and cleft secondary palate (Hammond et al., 2005; Kochilas et al., 2002; Vitelli et al., 2002).
The 22q11 deletion region comprises more than 30 known genes. Although synteny is not completely conserved between human and mouse, a region of murine chromosome 16 contains a significant number of genes found within the human 22q11 deletion region. This observation prompted the development of mouse models in an attempt to identify the gene or genes responsible for DGS (Scambler, 2000). TBX1, a member of the Brachyury-related (T-box) family of transcription factors, has emerged as a good candidate for the DGS because of its chromosomal location (Chieffo et al., 1997) and expression domain (Chapman et al., 1996a; Kochilas et al., 2003; Sauka-Spengler et al., 2002), and because mice generated with a targeted deletion of Tbx1 have a phenotype resembling that of DGS patients (Jerome and Papaioannou, 2001; Lindsay et al., 2001; Merscher et al., 2001). Supporting this hypothesis, a mutation in Tbx1 was identified in a case of sporadic DGS (Yagi et al., 2003).
The dentition of DGS subjects is significantly abnormal, and the presence of both hypodontia and enamel defects has been described in several clinical case series (Borglum Jensen et al., 1983; Fukui et al., 2000; Klingberg et al., 2002). The enamel defects range from a delay in mineralization to a more severe and generalized hypomineralization. These anomalies have often been regarded as secondary to the pre-term birth and early systemic insults resulting from congenital heart disease and hypocalcemia that are common in DGS patients. A direct link between impaired Tbx1 function and enamel defects has not yet been established (Fukui et al., 2000; Klingberg et al., 2002).
Enamel forms via the mineralization of specific enamel matrix proteins that are secreted by highly differentiated dental epithelial cells called ameloblasts. T-box genes are involved in the specification of many different cell populations (Papaioannou and Silver, 1998), and Tbx1 is a candidate regulator of amelogenesis since its expression in the developing mouse molars is restricted to ameloblast progenitors (Mitsiadis et al., 2008c; Zoupa et al., 2006). In addition, forced expression of Tbx1 in dental explants is capable to activate the expression of the enamel specific amelogenin gene (Mitsiadis et al., 2008c). Tbx1 expression in dental epithelium correlates with cell proliferation events that are under control of FGF signalling (Mitsiadis et al., 2008c). Absence of the Tbx1 gene in the mouse leads to death at birth (Jerome and Papaioannou, 2001), thus hampering the investigation of the role of this gene in amelogenesis, which occurs mainly at postnatal stages. In addition, the period of tooth development in humans is much longer than in rodents and the effects of a small change in Tbx1 dosage is more noticeable in humans than in rodents. To overcome these difficulties, we studied the continuously growing mouse incisors which show distinct morphological organization along their anterior–posterior axis with well-defined zones of cell proliferation, cell differentiation and mineral deposition (Mitsiadis et al., 2007). Furthermore, rodent incisors exhibit a characteristic asymmetric deposition of enamel that covers only their labial side, which contains a stem cell niche that constantly supplies ameloblast progenitors (Mitsiadis et al., 2007; Bluteau et al., 2008). In the mature incisors, all ameloblasts that form enamel are derived from cells of this niche, which is located at the cervical loop area. Differential expression of FGF molecules and the Follistatin and Sprouty genes is associated with the ability of the labial but not the lingual cervical loop to give rise to differentiated ameloblasts (Klein et al., 2008; Wang et al., 2007). Thus to determine the long-term effects of Tbx1 on cell proliferation, ameloblast differentiation and enamel formation incisor germs were removed from wild-type and Tbx1 mutant mouse embryos and transplanted into kidney capsules, which offer an ideal environment for the unimpaired development and growth of organs and tissues (Berard, 1975; Morio, 1985).
Here, we provide evidence that Tbx1 plays a pivotal role in amelogenesis and in the maintenance of epithelial stem cells that are responsible for the continuous supply of ameloblasts in rodent incisors.
Tbx1 mutant mice used in this report have been previously described (Lindsay et al., 2001). Tbx1 heterozygote mice were crossed and embryos collected at the specified embryonic (E) stage and genotyped. Mice carrying mutant alleles of Spry2 (Shim et al., 2005) and Spry4 (Klein et al., 2006) were maintained and genotyped as reported. Age-matched CD1 embryos were used as controls. Swiss and C57Bl/6 wild type mice were also used at embryonic and early postnatal (P) stages (E10 to P2). The age of the mouse embryos was determined according to the appearance of the vaginal plug (day 0) and confirmed by morphological criteria. Animals were killed by cervical dislocation and the embryos were surgically removed in Dulbecco’s phosphate-buffered saline (PBS). Dissected heads were fixed overnight at 4 °C in 4% paraformaldehyde (PFA) in PBS. For kidney capsule experiments the incisor germs of E13 to E14 embryos were dissected from the rest of the jaws into Dulbecco’s phosphate-buffered saline (PBS) prior transplantation.
All transplantation operations were performed under the appropriate UK Home Office animal licensing procedures. The host animals were anesthetized using pentobarbital (0.8 mg/10 g body weight). Isolated E13–E14 incisor germs were cultured for 2 days in Petri dishes containing media consisting of Dulbecco’s Modified Eagle Medium (DMEM; GibcoBRL), 20% foetal calf serum (FCS; GibcoBRL) and 20 U/ml penicillin/streptomycin, as previously reported (Mitsiadis and Drouin, 2008; Mitsiadis et al., 2003, 2008a). After culture the explants were grafted under the kidney capsules of host mice. The transplanted kidneys of the host mice were removed after 28 days and the presence of mature, fully mineralized incisors was determined with radio-graphs (Faxitron, Hewlett Packard). The transplanted tissues were then processed for histological preparation, sectioning, and staining.
For histological examination the E18.5–E18.75 embryonic tissues were fixed overnight at 4 °C in 4% PFA in PBS. Tissues cultured in kidneys for 28 days were first demineralized in a buffered solution of 4% EDTA, 0.2% PFA at 4 °C for 2 weeks, then dehydrated, embedded in paraffin. 5 μm serial sections were performed. Sirius red and hematoxylin-eosin stainings were performed every other section of non-decalcified (E18.5–E18.75) and decalcified (28 days culture) teeth.
Digoxigenin-labeled (Boehringer Mannheim) sense and antisense riboprobes for mouse Tbx1 (Chapman et al., 1996b; Mitsiadis and Drouin, 2008; Mitsiadis et al., 2008c), and amelogenin (Mitsiadis et al., 2008c) were used. Whole mount in situ hybridization on explants and in situ hybridization on sections were performed as previously described (Mitsiadis et al., 1995, 2003; Mitsiadis and Drouin, 2008).
Tbx1 mutant and wild type embryos were collected for skeletal analysis at E18.5 and fixed in 95% ethanol for two weeks. Samples were stained for cartilage with 200 μg/ml alcian blue 8 GX (Sigma-Aldrich) in 80% ethanol and 20% glacial acetic acid (BDH) for 3 days at RT. Thereafter the samples were placed in 95% ethanol for one week, cleared in 1% potassium hydroxide (KOH), and finally stained in 0.1% aqueous alizarin 15 red-s solution (Sigma-Aldrich) for 4 days at RT. Samples were then washed for 30 min in running tap water, decolorized in 20% glycerol (BDH) in 1% KOH, and finally prepared for storage in increasing concentrations of glycerol in 70% ethanol to a final concentration of 100% glycerol.
Mandibular and maxillary incisors were dissected out of the skeletal preparations and photographed in a bright-field Leica stereomicroscope. These incisors were then embedded in paraffin, sectioned and mounted in slides for microscopic examination.
Proliferating cells were detected using the Proliferating Cell Nuclear Antigen (PCNA) in paraffin sections of incisors from mutant and wild type mice. Immunohistochemistry was carried out essentially according to the manufacturer’s instructions (Zymed Laboratories, Hannover, Germany), using the anti-PCNA rabbit polyclonal antibody (1:200; Santa Cruz Biotechnology, USA). Biotinylated secondary antibody against rabbit IgG (1:200) was applied and detected by means of a Vectastain ABC kit (Vector) and diamino-benzidine (DAB; Sigma). Negative control was obtained by omitting the primary antibody.
Apoptosis was detected using a mouse anti-caspase-3 antibody (Transduction Laboratories, Interchim, Montluçon, France). Immunohistochemistry was performed as described previously (Mitsiadis et al., 2008b). Briefly, the anti-caspase-3 antibody (1 μg/ml) was applied on the sections for 30 min at 37 °C. After incubation with the biotinylated secondary antibody, the sections were incubated with peroxidase-conjugated streptavidine for 10 min. Peroxidase was visualized by incubation with 3-amino-9-ethylcarbazole (AEC) reaction solution. In control sections the primary antibody was omitted.
One mutant incisor grown under kidney capsule, that seemed to have some enamel, as well as two heterozygous incisors and one wild type tooth were selected for TEM examination. Following fixation, incisors were demineralized in 4.13% EDTA (pH 7.2–7.4) at 4 °C for 4 weeks and proceeded for TEM examination. Transplants were, postfixed in 2% Os-tetroxide in 0.2 M Na-cacodylate buffer (pH 7.2–7.4) at room temperature for 45 min and washed thoroughly in the same buffer. For embedding, specimens were dehydrated in ascending grades of alcohol, then transferred to propylene oxide and finally embedded in Epon 812. From the polymerized blocks, sections of about 1 to 2 μm thickness were prepared using histodiamond knives (Diatome, Biel, Switzerland) in a Reichert Ultracut E ultramicrotome (Leica Microsystems, Herbrugg, Switzerland). The sections were stained with periodic acid-Schiff (PAS) and methylene blue-Azur II to select a site for TEM examination. These sites were thin sectioned at a thickness of 60–80 nm using diamond knives (Diatome) in the same Reichert Ultracut E ultramicrotome. The thin sections were collected on copper grids, double contrasted with uranyl acetate and lead citrate, and examined in a Philips EM 400T transmission electron microscope (FEI, Eindhoven, The Netherlands) at an accelerating voltage of 60 kV. Digital micrographs were captured at a resolution of 2272×2272 pixels using a Hamamatsu ORCA-HR camera (Hamamatsu Photonics, Hamamatsu City, Japan) and the program AMT Image Capture Engine 5.42 (Advanced Microscopy Technics, Danvers, MA, USA).
It was previously reported that Tbx1−/− mice exhibit an incisor phenotype (Jerome and Papaioannou, 2001). The continuously erupting incisor of rodents displays distinct morphological organization along both the labial–lingual and the anterior–posterior axes (Bluteau et al., 2008; Mitsiadis et al., 2007). To better understand and clearly determine the function of Tbx1 in incisor formation and homeostasis, we first analyzed the expression pattern of the Tbx1 gene during mouse incisor development.
Initiation of incisor development starts at E10.5–E11 as a local thickening of the oral epithelium that invaginates into the underlying neural crest-derived mesenchyme and progressively acquires the characteristic bud, cap, and bell configurations. Whole-mount in situ hybridization analysis showed that Tbx1 mRNA was present in the oral epithelium from E10. The hybridization signal was weak at E10 in both presumptive incisor and molar areas (Fig. 1A) and became stronger at E10.5–E11, especially in the most distal areas where incisors develop (Figs. 1B and C). No hybridization signal was detected with the sense probe at this or subsequent developmental stages (data not shown).
At E13 (bud stage), Tbx1 was expressed in epithelial cells in proximity to the dental papilla mesenchyme, while the gene was not detected in cells located in the central part and flanks of the maxillary and mandibular incisor germs (Figs. 1D and E). This restricted expression pattern was maintained in the epithelium during the rotation of the maxillary and mandibular incisors at E14–E14.5 (Figs. 1F–I). At E18.5, Tbx1 transcripts were detected at the posterior end of the incisor in cells of the inner dental epithelium at the labial side (Fig. 1J). Interestingly, Tbx1 transcripts were also observed in cells of the outer dental epithelium at the very posterior end of the incisor. The gene was absent from the rest of the dental epithelium extending anteriorly (i.e. preameloblasts, ameloblasts, outer dental epithelial cells) on the labial side, and from the dental epithelium (i.e. inner and outer dental epithelia) on the lingual side (Fig. 1J).
The expression pattern of Tbx1 in the incisor suggests that this transcription factor has an important role in the developing incisor, and more precisely in amelogenesis. Indeed, the incisors of Tbx1−/− mice were significantly shorter than in wild-type mice, and furthermore, the enamel was absent in most of these mutant incisors (Figs. (Figs.22–5).
At E18.5, the incisor is at the bell stage and epithelial cells (inner dental epithelium) facing the dental papilla mesenchyme differentiate into preameloblasts. Incisors were carefully extracted from the mandible and maxilla of skeletal preparations from E18.5 Tbx1−/− mice and were compared to incisors from E18.5 wild type littermates, to evaluate their size and level of mineralization macroscopically. This examination revealed that both the growth and mineral deposition processes of the mutant incisors were altered. The Tbx1−/− teeth were smaller and less mineralized than those of the wild-type mice (Fig. 2A), and the epithelial layer responsible for enamel secretion (Fig. 2A, green arrows) was always significantly thinner in the mutant incisors when compared to that of the normal incisors. However, length reduction/hypoplasticity and mineral deposition were more prominent in the maxillary mutant incisors (Fig. 2A, orange arrows). Furthermore, mineral deposition ranged from reduced/low levels to nonexistent in the maxillary mutant incisors (Fig. 2A, black arrows). It is noteworthy that the labial–lingual asymmetry of hard tissue deposition was also altered in the Tbx1−/− incisors. In the developing incisors of wild-type mouse embryos, the process of hard tissue formation is more advanced on the labial side when compared to the lingual side (Fig. 2A, green line). This asymmetry was not evident in the mutant incisors, where often the deposition of minerals was identical on both lingual and labial sides. Longitudinal sections of the mutant maxillary incisors (Fig. 2B) showed that minerals are deposited in a non-structured, sporadic way, leading thus to un-organized mineral structure in the labial side (Figs. 2D–F) as compared to the lingual side (Fig. 2C).
The histological examination of E18.5–E18.75 Tbx1−/− incisors showed a phenotypic variability ranging from severe to mild manifestations (Fig. 3), due to the different degrees of penetrance. Although odontoblast differentiation and dentin matrix secretion occur normally, the differentiation of preameloblasts into ameloblasts appears to be defective: we were not able to detect functional ameloblasts and enamel in the Tbx1−/− incisors (Figs. 3A–C, F–J). In some Tbx1−/− incisors inner dental epithelial cells of the labial side at the median and anterior parts were tiny and exhibited a spherical shape (Figs. 3H, J), resembling the epithelial cells of the lingual side (Fig. 3H). A characteristic feature of the maxillary Tbx1−/− incisors was that the cervical loop on the labial side regressed considerably, while the cervical loop at the lingual side expanded significantly (Figs. 3A, B), when compared with the equivalent areas of the maxillary Tbx1+/− incisors (Fig. 3D). A more detailed observation showed that the structure of the cervical loop at the labial side was severely disturbed in the Tbx1−/− incisors (Fig. 3L), but not the Tbx1+/− incisors (Fig. 3K).
In order to further understand the functional role of Tbx1 in incisor development, and to overcome the problem of the early lethality of the Tbx1−/− mice, we used the kidney transplantation technique. This technique allowed a long-term culture (i.e. 28 days) of isolated incisor germs from E13–E14 Tbx1−/−, Tbx1+/−, and wild-type mouse embryos. These incisor germs were transferred to kidney capsules and their development assayed (Figs. 4A–C). From a total of 5 experiments, 3 wild-type, 4 Tbx1+/−, and 8 Tbx1−/− incisors were grown to full maturity under these conditions and recovered for further analysis. X-ray imaging was used to determine their degree of mineralization (Figs. 4D–G). It was evident from these radiographs that the radio-opacity of the mineralized tissues was more pronounced in the wild-type and Tbx1+/− incisors (Fig. 4G) than in the Tbx1−/− incisors (Figs. 4D–F). Increased radio-opacity was observed on the labial side of the wild type and Tbx1+/− incisors (Fig. 4G), while radio-opacity was equal on the lingual and labial sides of the Tbx1−/− incisors (Figs. 4D–F). Identical radio-opacity indicates that there is only dentin deposition on the lingual and labial sides of the mutant incisors, and thus represents the absence of enamel on their labial side. This was confirmed by the observation that the characteristic yellow color of rodent incisors enamel was not detected in the Tbx1−/− teeth grown in kidneys (Figs. 4H–J).
Histological examination demonstrated that wild-type and Tbx1+/− incisors develop normally in kidney-capsule explants and maintain their capacity to form enamel on their labial side (Fig. 5A, red boxes). The shape of the functional ameloblasts followed the various stages of enamel secretion and exhibited either a tall columnar shape or a low cubical profile. The development of the lingual side of these incisors was not disturbed as well (Fig. 5A, green box), and as expected, ameloblasts and enamel were absent on the lingual side. In contrast, we were not able to detect preameloblasts, ameloblasts, and enamel in the Tbx1−/− incisors (Fig. 5B). Epithelial cells of the labial side were tiny and exhibited a flat or spherical shape (Fig. 5B, red boxes), closely resembling the cells overlying the hard tissue (i.e. cementum) on the lingual side (Fig. 5B, green boxes). In addition, ameloblast precursors, which are normally detected in the cervical loop on the labial side, were not present in the mutant incisors (Fig. 5B, red box in left). A characteristic feature of the Tbx1−/− incisors was that the cervical loop on the labial side regressed considerably. This had as a consequence the reversal of the normal incisor morphology, with a lingual side longer than the labial side (compare Fig. 5A with Fig. 5B). Finally, the mutant incisors showed the same degree of dentin deposition on both labial and lingual sides, while in the wild-type dentin was thicker on the labial side when compared to the lingual side (Fig. 5A). In contrast, the dentin in the Tbx1−/− teeth was much thicker than in the wild type and Tbx1+/− incisors (compare Fig. 5A with Fig. 5B).
Only one out of the eight Tbx1 mutant incisors grown under kidney capsule seemed to have enamel deposition, and it was used to examine the ultrastructural properties of the enamel. When ameloblasts were present in mature mutant incisors these contained abnormal vacuoles that were concentrated in the secreting front and that were not present in other cell layers (Fig. 6, red arrows). The organelles and Tomes’ processes of the ameloblasts were very disorganized, partly disrupted and in a process of apparent disintegration, giving the impression of dying cells. Cytoplasmic vacuoles were also present in ameloblasts of the heterozygote incisors, but they were much less prominent and occurred only occasionally in wild type specimens.
Since incisor development and morphogenesis require a tight balance between cell proliferation and apoptosis, we asked whether these two events were abnormal in the incisors of Tbx1 null mutants. Despite the absence of ameloblasts and the reduced length of the Tbx1−/− incisors, apoptosis was not altered in incisors of E18.5 embryos. Apoptotic cells were scarcely observed in the dental epithelium (Fig. 7), and few apoptotic cells were detected in the alveolar bone (Fig. 7). No apoptosis was detected in Tbx1−/− incisors cultured in kidneys, as well as in control sections (data not shown).
In contrast, cell proliferation was abnormal in the epithelium of the Tbx1−/−incisors. This was more pronounced in the cervical loop area on the labial side, which contains a population of actively proliferating cells (Figs. 8A and B). In the mutant incisors, the total number of proliferating cells was decreased dramatically (by approximately 60%) in the cervical loop. A more severe effect was observed in the layer of presumptive inner dental epithelium, which contains the ameloblast progenitors, where cell proliferation was undetectable (Figs. 8A and B). In contrast, intense cell proliferation was detected in both the cervical loop area and the inner dental epithelium of wild type incisors (Figs. 8C and D).
Amelogenin is a marker of ameloblast differentiation and initiation of enamel formation. In our previous studies we showed that Tbx1 misexpression induced amelogenin expression in the mouse mandible (Mitsiadis et al., 2008c), thus showing that a close link exists between the Tbx1 and amelogenin genes. Thus, we asked whether Tbx1 deletion could affect amelogenin expression in incisors.
In situ hybridization in sections of E18.5–E18.75 wild type and Tbx+/− mouse embryos showed that amelogenin was strongly expressed in cells of the inner dental epithelium that differentiate into ameloblasts (Fig. 9A, and data not shown). In contrast, the amelogenin gene was either absent or weakly expressed in incisors of Tbx1 null mutants (Fig. 9A). It is noteworthy that weaker expression was observed in the maxillary incisors (Fig. 9A).
In teeth that were cultured under kidney capsule for 28 days, amelogenin transcripts were detected in ameloblasts of the wild type and Tbx1+/− incisors, while transcripts were not found in mature Tbx1−/− incisors (Fig. 9B, and data not shown).
Sprouty genes encode antagonists of receptor tyrosine kinase signalling and have been shown to prevent the generation of ameloblasts in the lingual epithelium of incisors by antagonizing an epithelial–mesenchymal FGF signalling loop (Klein et al., 2008). Loss of Sprouty gene function leads to bilateral enamel deposition as a result of ectopic generation of ameloblasts from stem cells in the lingual cervical loop. In our previous work we showed that Tbx1 is expressed in response to FGF signals in ameloblast progenitor cells (Mitsiadis et al., 2008c). To further analyze the involvement of FGF signalling in Tbx1 induction and/or maintenance, we studied and compared the expression of Tbx1 in incisors from wild type embryos with those from embryos carrying three Sprouty null alleles (Spry2+/−;Spry4−/− mice). In E16.5 wild type incisors, Tbx1 was only expressed in the labial epithelium and cervical loop (Fig. 10A). In the E16.5 Spry2+/−;Spry4−/− mutant incisors, Tbx1 was ectopically expressed on the lingual side, but was restricted to the posterior end of the tooth (Fig. 10B). To determine if this ectopic Tbx1 expression was linked to the differentiation of inner dental epithelium cells into ameloblasts we examined the expression of amelogenin in the incisors of the Sprouty mutants. Amelogenin expression was detected in both labial and lingual sides of the postnatal day 2 incisors of the Sprouty mutants, thus confirming the presence of mature ameloblasts on the lingual side (Fig. 10C).
Tooth development depends on sequential and reciprocal interactions between cells of the oral epithelium and cranial neural crest-derived mesenchyme (Cobourne and Mitsiadis, 2006; Lumsden, 1988; Mitsiadis, 2001). Cells from the dental epithelium differentiate into ameloblasts and secrete the organic matrix of enamel, which is a highly mineralized tissue unique to teeth. Specific secreted signalling molecules and transcription factors are required for odontogenesis at each stage of tooth development (reviewed by Mitsiadis, 2001; Tucker and Sharpe, 2004). Our previous studies have shown that expression of the transcription factor Tbx1 is correlated with epithelial cell proliferation and ameloblast differentiation, and, furthermore, activates amelogenin expression in the developing mouse molars (Mitsiadis et al., 2008c).
In humans and most mammals, whose teeth do not undergo continuous eruption, ameloblasts synthesize, deposit, mineralize and mature the enamel matrix, and finally regress and disappear when the tooth crowns emerge into the oral cavity. In contrast, rodent incisors erupt continuously, a process that involves the generation of new ameloblasts for the de novo formation and deposition of enamel. This continuous process is due to the presence of a population of epithelial stem cells that are located in the apical part of the incisor, the cervical loop area (Harada et al., 1999, 2002; Mitsiadis et al., 2007). Stem cells from the cervical loop generate transit amplifying progenitor cells that differentiate into all epithelial cell types of the tooth, including the ameloblasts (Bluteau et al., 2008; Mitsiadis et al., 2007). Disruption or damage of this area results in the arrest of new enamel formation (Merzel and Novaes, 2006). Therefore, the rodent incisor constitutes a unique model of amelogenesis involving well-defined compartments of cell proliferation, differentiation and maturation.
Another intriguing aspect of the rodent incisor is that enamel is present only on the labial side, while dental epithelial cells of the lingual side do not differentiate into ameloblasts and thus cannot synthesize enamel matrix. Ameloblast progenitors are morphologically indistinguishable from all other immature dental epithelial cells. Our studies suggest that Tbx1 expression is required in the progenitors of ameloblasts as opposed to the rest of the dental epithelium. Indeed, Tbx1 is expressed in cells of the cervical loop and inner enamel epithelium (i.e. the ameloblast precursors) on the labial side of the incisor. In contrast, Tbx1 is not expressed in epithelial cells of the lingual side where enamel is not formed. Therefore, it is quite possible that Tbx1 plays a pivotal role in the specification and proliferation of the restricted dental cell population that is responsible for the generation of ameloblasts.
Epithelial cells such as the ameloblast progenitors actively proliferate under the control of Tbx1 (Mitsiadis et al., 2008c; Xu et al., 2007). Indeed, proliferation of epithelial cells was greatly reduced in the Tbx1−/− incisors when compared to the incisors of wild type animals. This disturbance of cell proliferation was restricted to ameloblast precursors that normally express Tbx1, and thus is possibly responsible for the enamel defects in the Tbx1 mutant incisors. Mutant incisors grown to maturity in kidney capsules were hypoplastic and exhibited a deficiency in mineral deposition. Although a thin layer of enamel and some ameloblasts were present in very few mutant incisors, the majority of these teeth lacked enamel, ameloblasts, and their precursors. In the rare mutant incisors that exhibited enamel formation, electron microscopy revealed the presence of uncharacteristic vacuoles at the secreting poles of the ameloblasts and a defective enamel structure. Vacuoles were also present in the ameloblasts of the heterozygote incisors but their size and number were smaller than in the Tbx1−/− incisors, thus indicating that the heterozygote mice might also exhibit a mild enamel phenotype. Taken together these findings clearly show that Tbx1 is necessary for the proliferation, maintenance and homeostasis of the ameloblastic lineage.
Large variations in both tooth size and degree of mineralization have been observed in DGS patients (Borglum Jensen et al., 1983; Fukui et al., 2000; Klingberg et al., 2002). In addition to the visible enamel defects, histological examinations have demonstrated a significant enhancement in the deposition of dentin (Fukui et al., 2000), which is synthesized by the mesenchyme-derived odontoblasts. Similarly, the anomalies in the incisors of the Tbx1−/− mice were not limited to their epithelial component, as indicated by the increased thickness of dentin. However, dental abnormalities in DGS patients are less severe than those observed in the teeth of the Tbx1−/− mice. This could be explained by the fact that DGS is a haploinsufficiency condition, whereas we studied mice that were lacking all Tbx1 function.
Incisors from the Tbx1−/− mice showed variable degrees of hypomineralization and hypoplasticity. This phenotypic variability could be the result of incomplete penetrance. Alternatively, other transcription factors may partly compensate for the loss of Tbx1. An additional phenotypic variability was noticed with respect to the maxillary Tbx1−/− incisors, which consistently were more severely affected than those of the mandible. These jaw-specific differences are consistent with the observation that the development of the maxillary teeth is delayed when compared to the development of the mandibular teeth. For example, E18.5 mandibular incisors are longer and contain more hard tissue deposits than the maxillary incisors, which are shorter and less mineralized.
In the incisors of the Tbx1−/− mice the labial–lingual asymmetry of hard tissue formation was also affected. Normally, the processes of cytodifferentiation and mineralized tissue deposition are more advanced on the labial as compared to the lingual side of the developing mouse incisor. However, this difference was not evident in the incisors of the mutant mice, where the deposition of minerals was symmetric on the lingual and labial side. This modification implies that Tbx1 deletion can cause a deposition imbalance, presumably due to a cellular disruption and/or dysfunction in the epithelium of the labial side, where Tbx1 is normally expressed.
The process of enamel formation involves the synthesis and secretion of several extracellular enamel matrix proteins. The most abundant of these proteins is amelogenin (Zeichner-David et al., 1997), which is considered as the best marker for ameloblast differentiation and initiation of enamel formation. We have recently shown that forced Tbx1 expression is sufficient to induce amelogenin expression in the odontogenic epithelium (Mitsiadis et al., 2008c), indicating that amelogenin expression is under the control of Tbx1. However, in the present study we detected low levels of amelogenin in a few of the E18.5 Tbx1−/− incisors. While the levels of amelogenin expression were similar in the developing maxillary and mandibular incisors of the wild type mice, in the developing incisors of the Tbx1 mutants expression ranged from low levels in the mandibular incisors to non-detectable levels in most of the maxillary incisors. This could be the result of variability in penetrance, which may also affect the severity of the tooth phenotype. Phenotype variability in teeth or other organs, in the Tbx1 mutants and the DGS patients, may suggest that there is a threshold level associated with unstable balance, which can be perturbed by chance events (Zhang and Baldini, 2008). Alternatively, it is also possible that Tbx1 affects amelogenin expression indirectly by playing a role in maintenance and induction of differentiation of the ameloblast precursors (i.e. inner dental epithelial cells). Amelogenin was not expressed at all in the mature Tbx1−/− incisors that were grown in kidney capsules. This is reflected by the absence of enamel in these teeth, and suggests that cell fate determination in the incisors of the mutant mice was compromised. The absence of amelogenin from the Tbx1−/− incisors cultured in kidney capsules could be explained by the fact that during the culture period (i.e. 28 days) the very first formed preameloblasts, which expressed weakly the amelogenin gene at E18.5, have either already disappeared by apoptosis or changed their fate. In contrast to the wild-type incisors, cells from the cervical loop of the Tbx1−/− incisors are not able to differentiate into preameloblasts/ameloblasts and therefore do not express amelogenin.
Several studies indicate that FGF and BMP signalling are required for the maintenance of ameloblast precursors in the incisors (Klein et al., 2008; Mitsiadis et al., 2008a; Wang et al., 2007; Wang et al., 2004). We recently demonstrated that mesenchyme-derived FGF molecules are important for the maintenance of Tbx1 expression in the dental epithelium (Mitsiadis et al., 2008c). Furthermore, studies in other developing organs have shown that FGF molecules form a regulatory loop with Tbx1 (Vitelli et al., 2002). It was therefore important for us to determine whether FGF signalling induced Tbx1 expression in the epithelial progenitor cells of incisors. We therefore examined the incisors of mice carrying null alleles of Sprouty genes, which are antagonists of FGF signalling. These mice are hypersensitive to FGF and other receptor-tyrosine kinase signalling because they have decreased dosage of the FGF antagonists. We have previously shown that the incisors of Spry4−/−;Spry2+/− mice are covered by enamel on both labial and lingual sides (Klein et al., 2008). In the present study, we found that Tbx1 and amelogenin were ectopically expressed on the lingual side of these mutant incisors, indicating that upregulation of FGF signalling leads to ectopic Tbx1 expression and confirming our previous findings that Tbx1 is a target of FGF signalling. It is worth noting that the cervical loop on the lingual side of the Spry4−/−;Spry2+/− incisors was larger than that of the normal incisors. This difference in size may result in part from increased cell proliferation owing to the ectopic expression of Tbx1 on the lingual side of the mutant incisors. Thus, the results from our studies on the Spry4−/−;Spry2+/− mice reinforce our assumption that Tbx1 plays an important role in the proliferation and maintenance of the ameloblast precursors by mediating the function of FGF molecules.
In conclusion, the incisors of the Tbx1−/− mice showed diverse anomalies, with the most striking being the lack of enamel. These phenotypes resulted from decreased cell proliferation and amelogenin expression, as well as from various defects in cytodifferentiation and mineralization (Fig. 11). Hence, the phenotypes of the Tbx1−/− incisors might help to better understand dental anomalies observed in DGS patients. However, penetrance variability makes the phenotypic determination elusive and the responsible mechanisms complicated to determine.
We would like to thank Professor Antonio Baldini from the Institute of Genetics and Biophysics of Napoli, Professor Peter Scambler and Miss Catherine Roberts from UCL London, for making available the Tbx1 mutant mice and their help in starting the Tbx1 colony. We are also grateful to Professor Gail Martin for providing the Sprouty mutant mice. This work was supported by grants from the University of Zurich (T.M.), Swiss National Foundation (T.M.), Guy’s and St Thomas’ Charity Foundation (T.M., J.C., A.S.T.), and NIH (K08-DE017654) (O.K.).