|Home | About | Journals | Submit | Contact Us | Français|
Organogenesis depends upon a well-ordered series of inductive events involving coordination of molecular pathways that regulate the generation and patterning of specific cell types. Key questions in organogenesis involve the identification of the molecular mechanisms by which proteins interact to organize distinct pattern formation and cell fate determination. Tooth development is an excellent context for investigating this complex problem because of the wealth of information emerging from studies of model organisms and human mutations. Since there are no obvious sources of stem cells in adult human teeth, any attempt to create teeth de novo will likely require the re-programming of other cell types. Thus, the fundamental understanding of the control mechanisms responsible for normal tooth patterning in the embryo will help us understand cell fate specificity and may provide valuable information towards tooth organ regeneration.
Teeth, like all epithelial appendages, form via a sequential and reciprocal series of inductive signals transmitted between the epithelium and neural crest derived mesenchyme. Each tissue layer instructs the other to differentiate in a precisely determined manner leading to the formation of highly specialized structures, such as incisors, canines, premolars and molars. Each of these groups of teeth derives from different parts of the oral epithelium and, depending on the species, teeth can be formed from both endoderm and ectoderm or from ectoderm only [1, 2••].
Morphologically, tooth development commences with a thickening of the dental epithelium to form a structure known as the dental lamina (Figure 1). Within this band of thickened epithelium the cells start to proliferate and to invaginate in certain positions to form the placodes. After this fundamental step in development, further epithelial invagination and convolution form the bud, cap and bell stages of tooth morphogenesis (Figure 1). During these stages, the constant interplay of inductive signals between the epithelium and mesenchyme (i) gives rise to distinct anatomical and functional parts of the tooth and (ii) mediates the differentiation of the epithelium into enamel-secreting ameloblasts and that of the mesenchyme into dentin-secreting odontoblasts (Figure 1).
Animal and human studies that employ the tools of contemporary molecular genetics have identified a number of genes that act at specific stages of tooth development and regulate its patterning and differentiation process (Figure 1; Tables 1–2; http://bite_it.helsinki.fi). The purpose of this review is to discuss in general terms some recent findings regarding genes and pathways that control tooth development and to provide new perspectives on the potential molecular mechanisms that coordinate the process of odontogenesis and tooth regeneration.
The conserved signaling pathways of BMP, FGF, SHH and WNT ligands and their receptors constitute the key pathways that are used reiteratively during tooth development and mediate the epithelial–mesenchymal interactions [3, 4]. Over the past 15 years, studies using transgenic animals provided functional data showing that, in most cases, disruption of genes that are part of these signaling pathways results in severe aberrations of tooth development, such as complete tooth agenesis or arrest of tooth development at early stages of development (the lamina or bud stage of development), leading to anodontia (lack of teeth) [reviewed in 5, 6•, 7, 8•; Tables 1–2 and Figure 1]. For example, conditional inactivation of FGF8 in the dental epithelium results in arrest of tooth development at the lamina stage. Over-expression of BMPR1a in transgenic mice, or functional inactivation of FGFR2b or SHH results in arrest of tooth development at the bud stage [reviewed in 4, 5].
Recently, however, it was realized that the inhibitors of these signaling pathways also contribute to control tooth development. In most cases, when the inhibitors or mediators of these signaling pathways are perturbed, more teeth are formed with abnormal shape, ameloblast or odontoblast differentation defects and reduced matrix deposition [9, 10, 11, 12, 13•, 14••; reviewed in 8, Tables 1–2].
For example, loss of Ectodin leads to supernumerary teeth through inhibition of BMP signaling [10•]. Ectodysplasin (Eda), a WNT signaling mediator, when overexpressed can lead to supernumerary teeth [15, 16]. Apc (Adenomatous polyposis coli), another WNT modifier that organizes the complex that degrades β-catenin, results in multiple tooth buds when conditionally knocked out in the oral epithelium,. Consistently, when β-catenin is overexpressed results in supernumerary teeth. These results suggest that overexpression of the canonical WNT signaling, either through loss of function of its inhibitors or by overexpression of its effectors leads to supernumerary teeth [11, 17, 18]. The importance of the tooth-inductive potential of WNT signaling manipulation is further demonstrated by the recent discovery that WNT pathway activation, even postnatally, lead to formation of extra teeth [14••]. Moreover, a member of the low-density receptor-related protein family, Lrp4, that modulates and integrates both the BMP and the canonical WNT signalling by binding the secreted BMP antagonist protein Ectodin, when mutated in mice results in supernumerary incisors and molars as well as fused molars, a phenotype identical to that of Ectodin mouse mutant .
As in the case of BMP and WNT pathways, the mediators and/or inhibitors of the SHH and FGF signaling lead to supernumerary teeth, when mutated. Primary cilia mediate SHH signaling, since mutations in their protein components affect SHH activity. Mice mutant for a cilia intraflagellar transport (IFT) protein, IFT88/polaris, result in the formation of ectopic teeth, through increase of Shh activity in the toothless region of the embryonic jaw primordia, the diastema region [20, 21, 22••]. Consistently, upregulation of Shh activity in mice mutant for Gas1, a Shh protein antagonist, results in ectopic diastema teeth [22••]. Finally, inactivation of either Sprouty2 (Spry2) and/or Sprouty 4 (Spry4), the inhibitors of FGF signaling, leads to supernumerary teeth in the diastema [12, 13].
These studies demonstrate that tooth formation is the result of a tight control between networks of activators and inhibitors, and that any modification of these networks leads to abnormalities in either number or patterning. Interestingly, the transcription factors that mediate such signaling networks are indispensable for early and late tooth development as well. Genetic experiments where the function of transcription factors such as Msx1, 2, Dlx1, 2, 5, Runx2, Pax9, Pitx2, Lef1, Gli1, 2, 3, Lhx6, 7, 8, Prx1, 2 and others (http://bite-it.helsinki.fi) is eliminated in mice or humans results in most cases in an arrest of tooth development at the bud stage or before during the lamina stage, leading to anodontia [Tables 1–2; 5, 6•, 8•; Figure 1]. Exception to the rule is the case of Sp6, a zinc finger transcription factor known as Epfn. Sp6 null mice develop numerous teeth, up to 50 incisors and 8 molars, and that would be a surprise, if we didn’t know that Sp6 functions through upregulation of Lef1, a target, again, of WNT signaling, whose activation leads to extra teeth, as mentioned above .
Another transcription factor that leads to supernumerary teeth upon mutation is Odd-skipped related −2 (Osr2) [24••]. The study by Zhang et al. regarding the role of Osr2 transcription factor in tooth development stands alone, since most of the mouse mutants that develop extra buds/teeth do so in the toothless diastema region along the already formed single row of teeth. In contrast, Osr2 deletion in mice leads to supernumerary teeth lingual to their molars, thus forming a second row of molars, through up-regulation and expansion of the odontogenic field that is driven by the BMP4-Msx1-BMP4 pathway in the mesenchyme . Thus, normally, Osr2 suppresses this pathway along the buccolingual axis to restrict molar development to one tooth row in mice [24••].
The studies mentioned above indicate the importance of keeping a fine balance between signaling ligands, their receptors, inhibitors and transcription factors in regulating all aspects of tooth development, including the patterning, the size, the number and the shape. Since perturbations of these pathways lead to such fundamental changes in patterning and number, one could hypothesize that evolution favored certain pathways versus others in promoting certain changes in the dentition of vertebrate species.
For example, many non-mammalian vertebrates, such as fish or reptiles, replace their teeth throughout their life, have multi-rowed dentition and their teeth are all of simple shape, while mammalian vertebrates develop teeth in a single row, replace their teeth once or not at all, and the teeth acquire different shapes and forms such as incisors, canines, pre-molars and molars [1, 26, 27]. Mice are examples of mammalian vertebrates that possess molars and incisors only, they are monophyodonts (one set of teeth) and they develop their teeth in a single raw. The ancestors of mice, however, the Glires, a clade including rodents and lagomorphs, possessed premolars and canines similar to the dentition observed in non-rodent species . Interestingly, some rodents, such as squirrels and guinea pigs, still have premolar teeth, suggesting that the genetic information specifying premolar and canine tooth shapes or tooth replacement processes is still in place.
The phenotypes of numerous knock out and transgenic mice that form either additional teeth in the diastema region, a region where normally premolars and canines would have exist, or multiple de novo teeth, or multi-rowed dentition just by perturbing a signaling pathway, support such a hypothesis (Tables 1–2; this review).
Despite the progress made from genetic studies described above, the processes involved in the formation of extra teeth and tooth replacement are still not known. Some studies suggest that Pitx2 and Bmp4 are key molecules associated with continuous tooth replacement in fish . Mammals, such as mice, that do not replace their teeth, form de novo teeth when the WNT signaling is over-expressed [11, 14••, 17, 18]. Could any of these pathways be the key to regenerate teeth in humans? Can we use the pathway network knowledge to drive stem cell differentiation processes towards de novo tooth formation?
Currently, efforts towards that goal are concentrated in an attempt to discover adult stem cells in human or mouse teeth. Mesenchymal stem cells have been identified in adult human teeth in the dental pulp (DPSCs) and in the dental follicle (DFSCs) [29, 30]. These cells have stem cell properties, can be cultured as stem cells in vitro, can form colonies and differentiate in vivo into odontoblasts, cementoblasts and periodontal ligament cells . Dental epithelial cells, from the quiescent Epithelial Cell Rests of Malassez (ERM)-the only dental epithelium remaining after root formation, located within the periodontal ligament (PDL)-were isolated from pigs and differentiated into ameloblast-like cells producing enamel in vivo, when co-seeded with dental pulp cells . Using the continuously growing incisor of the mouse as a model for the study of adult epithelial stem cells, researchers have shown that label-retaining cells were localized in the epithelially derived cells of the cervical loop, and this population of cells has been proposed to constitute the mouse incisor stem cell niche . Further studies indicated that members of the FGF family of ligands, namely FGF3 and FGF10, derived from mesenchyme, promote the proliferation and survival of the incisor epithelial stem cell niche [33, 34, 35••]. Consistently, FGF10 is downregulated in teeth that do not grow continuously, such as the mouse molar, while the addition of FGF10 to cultured mouse molars promotes the maintenance of their cervical loops .
Although these studies have provided some insight on potential sources of tooth stem cells in pigs and mice, the fact remains that adult human teeth do not grow continuously and no human stem cell niches have yet been identified. The epithelial cell rests of Malassez (the only remaining epithelial cells after the tooth completes its development), the dental pulp and dental follicle stem cells that have been isolated from humans are promising, but their stemness is not yet well established and, more importantly, there is no evidence yet that they are capable to direct tooth morphogenesis.
Recent advances in adult cell reprogramming through the creation of induced pluripotent stem (iPS) cell lines from adult differentiated cells offer the possibility to produce pluripotent stem cells from patient’s own tissue [37•, 38]. IPS cells are created by forced expression of defined transcription factors as Oct4, Sox2, cmyc, and Klf4, which have been shown to induce pluripotency in somatic fibroblast cells [39••]. In another study, iPS cells were generated from human fibroblasts using Oct4, Sox2, Nanog, and Lin28 . Although the four reprogramming factors were different in these two studies, all of them were transferred into the cells by means of retrovirus gene transfer, which holds some risk to cause insertion mutations. In addition, since some of the reprogramming factors are oncogenes as well, the risk of tumor induction is another potential limitation to be considered. A recent breakthrough study by the group of Doug Melton succeeded to directly reprogram pancreatic exocrine cells to insulin producing beta cells [41••]. This method is more advanced to the reprogramming of any adult differentiated cells to iPS cells, because the pancreatic exocrine cells share a common genomic and epigenomic identity with those of insulin producing beta cells. The latter approaches could be promising and suggest that tooth-specific cell types, such as ameloblasts or dental follicle cells, with matrix producing capability could be produced through reprogramming of adult cell types. Whether reprogramming of adult cells to iPS and then programming of iPS can be used to generate, instead of specific cell types, a rather complex three-dimensional organ like the tooth, is not yet known.
The tooth is a complex organ and its development a process controlled by a sequence of cellular and molecular networks that act at particular places and times to guide pluripotent cells to restricted dental cell fates. The mechanisms that control and determine the history of the cells so that the differentiation program is properly executed during embryonic development are not known. Despite the fact that genetics and epistasis analysis have led to the discovery of numerous genes and pathways involved during different stages of tooth development, the same genes are required for the development of other tissues, especially that of other ectodermal organs such as hairs, nails and glands [Figure 1 and Tables 1–2]. In addition, perturbations of these genes affect not only tooth development but, in most cases, the development of other ectodermal organs, thus leaving open the obvious question of cell fate specificity [8, 42].
What it is known, however, is that transcription factors control cell fate through a selective regulation of target genes, and that the target gene specificity is achieved through context-dependent selective protein interactions. Recent studies suggest that in early tooth morphogenesis, a network of transcription factors operate in the dental mesenchyme to regulate a specific transcriptional output and that their combinatorial action, along with their target specificity is further modified by epigenetic mechanisms, such as sumoylation (Bei unpublished). The need of such a comprehensive analysis at the molecular level, along with extensive whole-genome data sets -including ChIP-chip and ChIP-Seq of multiple transcription factors-, maps of epigenetic states, expression profiling of genetically manipulated cells, would help determine the molecular identity of the tooth and permit controlling of the iPS programming towards the generation of de novo teeth.
This work was supported by grants from the NIH (RO3 DE 018415; RO1 DE 19226), Harvard Skin Disease Research Center (Pilot Fund Award) and Harvard Medical School (Milton Fund Award) to MB. We thank Dr. Intan Ruspita for her help in preparing the tables and figure.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Papers of particular interest, some published within the period of review, have been highlighted as:
• of special interest
•• of outstanding interest