Introduction

There can be few dental surgeons who have not pondered the mysteries surrounding congenital tooth absence at some point in their career. A genetic basis for the embryonic mechanisms underlying tooth formation is clear, even at an anecdotal level. A wide variety of animals have highly adapted and species-specific dentitions; in human populations defects of tooth development often affect particular teeth and these anomalies frequently run in families, while defects in tooth number are often associated with other anomalies of dental development (Table 1). The genetic mechanisms responsible for generating such a regionally diverse but homologous structure as the mature human dentition are still poorly understood, but progress has been made over the last decade, largely with the use of mouse models.1,2,3 In recent years, homologues of some candidate mouse genes have been identified as having a role in human dental development, particularly in the aetiology of congenitally absent teeth.4,5 However, given the large number of candidates and the prevalence of this condition, it is surprising that more genes have not been identified to date.

Table 1 Dental anomalies seen in association with hypodontia
Full size table

Clinical genetics

Hypodontia is often used as a collective term for congenitally missing teeth, although specifically it describes the absence of one to six teeth, excluding third molars (Table 2). Oligodontia refers to the absence of more than six teeth, excluding third molars, while anodontia represents a complete failure of one or both dentitions to develop.6 Hypodontia can either occur with accompanying genetic disease as part of a recognised clinical syndrome, or as a non-syndromic, familial form, which occurs as an isolated trait, affects variable numbers of teeth and appears either sporadically or in a familial fashion within a family pedigree7 (Fig. 1).

Table 2 Classification of inherited tooth loss
Full size table
Figure 1: Non-syndromic hypodontia and oligodontia
figure 1

In the upper panel there is hypodontia with UR2, UL2, LL8, LL5, LR5 and LR8 absent. In the middle panel there is more severe hypodontia, with UR8, UR5, UR4, UL4, UL5, UL8, LR8, LR5, LL5 and LL8 absent. In the lower panel there is oligodontia, with the UR8, UR5, UR4, UR2, UL2, UL5, UL8, LL8, LL5, LL1, LR4, LR5, LR8 absent

Full size image

Online Mendelian Inheritance in Man (OMIM) lists over 60 different syndromic conditions that include hypodontia as part of their phenotypic spectrum of anomalies8 and candidate genes have been identified for many of these conditions (Table 3). However, possibly of more relevance to the general dental practitioner is the more common non-syndromic or familial form of hypodontia (Fig. 1). This condition can follow autosomal dominant,9,10,11,12 autosomal recessive13,14 or sex-linked15 patterns of inheritance, with considerable variation in both penetrance and expressivity. Indeed, a multifactorial model has been suggested to explain the inheritance of anomalies in both tooth number and size with the phenotypic effect being related to certain thresholds, themselves influenced by both genetic and environmental factors.16,17 Clearly, within this model, the mutation of a major gene may be a significant enough event to result in inherited tooth loss.

Table 3 Syndromic conditions associated with hypodontia
Full size table

It is important for the general dental practitioner to fully assess any patient presenting with hypodontia and referral to a specialist clinic is often desirable. In some cases this condition can be indicative of underlying genetic disease and further referral for genetic testing might be desirable. Treatment of this condition aims to improve both aesthetics and function, and is often facilitated through a multidisciplinary approach.18,19

Non-syndromic hypodontia

Non-syndromic hypodontia is by far the most common form of congenital tooth absence and can involve variable numbers of teeth. It is more commonly seen in the secondary dentition, but in the rare cases of missing primary teeth that do occur, there is often a strong tendency towards further tooth absence in the secondary teeth. Anodontia (OMIM #206780) represents the most severe form of non-syndromic hypodontia, but is extremely rare in the absence of accompanying genetic disease,20 while oligodontia (OMIM #604625) is only seen at a level of around 0.25% within European populations.21,22 The more localised incisor-premolar type of hypodontia affects only one or a few teeth (OMIM #106600), but occurs more commonly in around 8% of the population.7 Within these clinical entities, certain teeth fail to develop more often than others. Third molars are the most commonly absent tooth in the dentition, with at least one being absent in anything up to 20-30% of the population. This is followed in Europeans by the mandibular second premolar, maxillary lateral incisor and premolars (around 2%) and the mandibular central incisor (0.2%).23 The absence of canine teeth, first molars and second molars is extremely rare in hypodontia;24 if these teeth are missing, it is usually seen in association with severe forms of syndromic oligodontia.

Candidate genes

If genes are so important in controlling tooth development, what do we know about potential candidates within the human genome? As with many aspects of mammalian development, the mouse has become one of the principle model organisms for the study of these embryonic processes and a host of genes, encoding members of numerous protein families, are expressed during development of the mouse tooth.1,3,25 Targeted deletion in many such genes within knockout mice can disrupt tooth formation. These data have provided a reference point in the search for candidate genes that may play a role in the aetiology of human forms of hypodontia.26 In particular, two genes that encode members of transcription factor families have attracted considerable attention because of their role in murine tooth development.

Msx1 (Muscle segment homeobox) is a member of a distinct sub-family of homeobox genes, which is expressed in spatially restricted regions of the head during early development, localising to regions of condensing embryonic connective tissue or ectomesenchyme in the tooth germ27,28 (Fig. 2). Furthermore, analysis of mice lacking a functional Msx1 gene reveals that all tooth development arrests at the bud stage.29 These findings demonstrate that in the mouse at least, Msx1 is essential for normal odontogenesis. Pax9 encodes a member of another transcription factor protein family, characterised by the presence of a DNA-binding paired-box domain. In the mouse embryo, Pax9 is also expressed in the prospective mesenchymal compartment of developing teeth30 (Fig. 2) and is essential during later stages of tooth development; mice with targeted mutations in Pax9 also exhibit tooth arrest at the bud stage.31 These two genes are therefore excellent candidates for human forms of hypodontia and have been the subject of intense scrutiny within human pedigrees affected by non-syndromic tooth loss.

Figure 2: Expression of Pax9 and Msx1 in the developing tooth
figure 2

At the bud stage of dental development both Pax9 and Msx1 are expressed in the ectomesenchymal component of the tooth germ, the dental papilla and follicle. These corresponding expression domains are consistent with biochemical evidence of interaction between these two proteins in the developing tooth

Full size image

MSX1

Consistent with the mouse phenotype, mutations in the human MSX1 gene have been associated with familial oligodontia12,32 and certain forms of syndromic hypodontia;33,34 however, associations with the more common incisor-premolar form of familial hypodontia are less common.7,35 The relationship of MSX1 to familial incisor-premolar hypodontia was originally investigated in five Finnish families, with a total of 20 affected individuals; but no linkage was identified.7 However, these findings did not rule out a defect in MSX1 being associated with other forms of hypodontia and analysis of a family affected with oligodontia identified a causative locus on chromosome 4p where the MSX1 gene resides.12 Sequence analysis demonstrated a missense mutation within a critical region of the MSX1 protein in all affected family members. This protein was subsequently found to be inactive in vivo and haploinsufficiency concluded to be the probable cause of the phenotype.36 A frameshift mutation in MSX1 has been identified in a family demonstrating non-syndromic hypodontia with absence of all second premolars and mandibular central incisors.37 Further studies have also demonstrated a role for MSX1 in the aetiology of some forms of syndromic hypodontia. A Dutch family showing various combinations of cleft lip, cleft palate and tooth agenesis were identified with a nonsense mutation in exon 133 and a further nonsense mutation has been shown to be responsible for Witkop syndrome (OMIM #189500), an autosomal dominant form of ectodermal dysplasia involving nail dysplasia and variable numbers of congenitally missing permanent and/or primary teeth.34

PAX9

A number of mutations38,39,40,41,42,43 and polymorphisms in the upstream promoter region44 of the human PAX9 gene have been identified in association with variable forms of oligodontia, that particularly affect the molar dentition. A family exhibiting hypodontia of most permanent molars and variable absence of second premolars and mandibular incisors were originally identified with a single base insertion that produced a frame-shift mutation and premature termination of the PAX9 protein.38 Significantly, this mutation alters the amino acid sequence within the highly conserved (paired box) region of the gene, producing reduced DNA binding of the mutant protein.45 However, another frame-shift insertional mutation outside this region can also produce hypodontia.40 Further single basepair mutations in PAX9 have since been identified in association with molar hypodontia, including nonsense39 and missense;41 in addition to a large 288 basepair insertion.41 Interestingly, while molar tooth development does seem to be particularly sensitive to alterations in PAX9 function, a PAX9 mutation has also been associated with a non-familial form of oligodontia affecting third molars, premolars and some incisor teeth.42 Haploinsufficiency of PAX9 seems to be the underlying cause of the hypodontia in these affected pedigrees, a finding reinforced by the identification of a rare father and daughter kindred affected by complete primary and permanent molar hypodontia with a deletion of one copy of their PAX9 gene.46

MSX1 and PAX9?

In addition to mutational analysis and the identification of candidate genes, biologists are now attempting to understand some of the molecular interactions underlying tooth development failure. There is now evidence to suggest that PAX9 and MSX1 interact during odontogenesis at both the gene and protein levels. Clues to this relationship are present in mice; expression of these genes co-localise in the developing tooth (Fig. 2), odontogenesis arrests at the bud stage in both mouse knockouts29,31 and this phenotype is also accompanied by a marked reduction in expression of the gene encoding Bone morphogenetic protein 4 (Bmp4) in both mouse lines. Bmp4 encodes a signalling molecule with a key role during transition of the tooth germ from bud to cap stage.28,47 Initial evidence of an interaction in humans came from a genetic epidemiological study48 and has since been confirmed by biochemical analyses; mammalian Pax9 is able to form a physical association with Msx1.49 This interaction takes the form of a heterodimeric protein complex, which enhances the ability of Pax9 to activate both Msx1 and Bmp4 gene expression during tooth development. Importantly, simulation of a known Pax9 mutation has also been shown to lack the ability to activate transcription of these target genes, even though a physical interaction with Msx1 still occurs.49

AXIN2

The identification of a four-generation Finnish family affected by autosomal dominant oligodontia has recently provided a rather unexpected further insight into the genetics of inherited tooth loss. Within this family, 11 members were identified as lacking at least eight permanent teeth and rather surprisingly, further investigation of this pedigree suggested that among these individuals affected by oligodontia, a significant risk of developing colorectal neoplasia was also present.50 Linkage analysis of this pedigree identified a candidate region on chromosome 17, which contained approximately 80 genes, among which was a gene called AXIN2 (Axis inhibition protein-2). AXIN2 was selected as a strong candidate gene for this condition for several reasons: its position within this particular chromosomal region, a previously identified association with colorectal carcinoma and the fact that AXIN2 is also a known regulator of the Wnt signalling pathway. The Wnt family of secreted proteins form part of a large family of signalling molecules that have a wide-ranging role during embryonic development and demonstrate regionally restricted expression in the tooth.51 Suppression of Wnt signal transduction in mutant mice or over-expression in wild type jaw explants can inhibit tooth development.52,53 Crucially, when further sequence analysis was carried out, all the affected family members within this pedigree had a nucleotide transition within exon 7 of AXIN2, which produced a nonsense mutation and premature termination of the encoded protein.50 Several novel polymorphisms or variants of AXIN2 have since been identified, which when present, also carry an increased risk of tooth agenesis to the individual.54

Conclusions

Given the large number of candidate genes that have been provided from studies in the mouse,25 it is perhaps surprising that mutations in so few have been identified in human family pedigrees affected by hypodontia.26 This suggests that in many cases, familial human hypodontia may represent a more complex, multifactorial condition. A number of subtle traits are apparent within human pedigrees possessing identifiable gene mutations in association with hypodontia. MSX1 function is predominantly associated with premolar and occasionally molar agenesis, while PAX9 mutations are almost always associated with missing molar and only occasionally premolar teeth. In contrast, hypodontia associated with AXIN2 mutations involves a wider range of tooth types. These observations suggest that the combined and overlapping expression domains of genes expressed within the tooth-forming regions are important for normal development of the dentition, a finding consistent with observations in the mouse.55,56 This threshold model is useful in explaining the lack of phenotypic penetrance and variability found in many pedigrees affected by hypodontia. However, it also provides some explanation for the observations that teeth along the margins of each field are more commonly absent (lateral incisors, second premolars, third molars).57 It is the peripheral domains of gene expression that are most important in delineating tooth class, but also the most susceptible to perturbations in function.

see Appendix

Table 4 Appendix 1 Glossary of terms
Full size table