| Sign In to gain access to subscriptions and/or personal tools. |
ENAM Mutations in Autosomal-dominant Amelogenesis Imperfecta
1 Department of Orthodontics and Pediatric Dentistry, University of Michigan Dental Research Lab, 1210 Eisenhower Place, Ann Arbor, MI 48108, USA; Correspondence: * corresponding author, janhu{at}umich.edu
To date, 4 unique enamelin gene (ENAM) defects have been identified in kindreds with amelogenesis imperfecta. To improve our understanding of the roles of enamelin in normal enamel formation, and to gain information related to possible genotype/phenotype correlations, we have identified 2 ENAM mutations in kindreds with hypoplastic ADAI, 1 novel (g.4806A>C, IVS6-2A>C) and 1 previously identified (g.8344delG), and have characterized the resulting enamel phenotypes. The IVS6-2A>C mutation caused a severe enamel phenotype in the proband, exhibiting horizontal grooves of severely hypoplastic enamel. The affected mother had several shallow hypoplastic horizontal grooves in the lower anterior teeth. In the case of the g.8344delG mutation, the phenotype was generalized hypoplastic enamel with shallow horizontal grooves in the middle 1/3 of the anterior teeth. In general, mutations in the human enamelin gene cause hypoplastic enamel, often with horizontal grooves, but the severity of the enamel defects is variable, even among individuals with the same mutation.
Key Words: enamel enamelin amelogenesis imperfecta hypoplastic AI incomplete penetrance
Amelogenesis imperfecta (AI) is a collection of inherited diseases that exhibit tooth enamel defects in the absence of systemic manifestations (Witkop and Sauk, 1976). The enamel malformations in AI patients can be categorized as hypoplastic, hypocalcified, or hypomaturation forms, and sometimes as a combination of these types (Witkop and Sauk, 1976). Based upon the enamel phenotype and mode of inheritance, 14 clinical subtypes of amelogenesis imperfecta are recognized (Witkop, 1989). The complexity of the disease pattern suggests that mutations in many different genes cause AI. Alterations in the human amelogenin gene (AMELX) are responsible for X-linked AI, but only 5% of families with AI show an X-linked pattern of inheritance (Bäckman and Holmgren, 1988). To date, 14 AMELX mutations have been reported (Kim et al., 2004). The various enamel phenotypes observed in families with X-linked AI correlate with the location of the mutation in the amelogenin coding region (i.e., signal peptide, N-terminus, C-terminus) (Wright et al., 2003). It is hoped that characterization of the genes and mutations that cause autosomal forms of AI will reveal correlations between the genotype and phenotype and between genotype and treatment outcome, and lead to improvements in treatment provision. The candidate or known genes for autosomal forms of amelogenesis imperfecta are the genes encoding enamel matrix proteins: enamelin and ameloblastin (4q11-q21), tuftelin (1q21-31), MMP-20 (11q22), and kallikrein-4 (19q13.3-q13.4). The first linkage of an autosomal-dominant form of AI was to chromosome 4q (Forsman et al., 1994), in a region since shown to contain the ameloblastin (Kärrman et al., 1997; MacDougall et al., 1997) and enamelin genes (Hu et al., 2000). At present, mutational analyses of ADAI families have detected mutations only in the enamelin gene (Rajpar et al., 2001; Kida et al., 2002; Mårdh et al., 2002; Hart et al., 2003a; Hart et al., 2003), but other genes are certain to participate, since linkage outside of the 4q region (Kärrman et al., 1996), as well as outside of the loci for all of the known AI candidate genes (Hart et al., 2003b), has been established.
Enamelin mutations cause hypoplastic forms of autosomal-dominant and -recessive AI, with the phenotype ranging from relatively minor localized enamel pitting to severely hypoplastic enamel (Hart et al., 2003; Hu and Yamakoshi, 2003). The standardized nomenclature for ENAM mutations uses GenBank reference sequences AF125373 for the cDNA and protein, and AY167999 for the gene (Hart et al., 2003a). The enamelin gene has 10 exons; 8 are coding (Hu et al., 2001b). The structure of the enamelin gene and the positions of known ENAM mutations associated with AI are shown in Fig. 1
Identification of Kindreds and Enrollment of Human Subjects The study protocol and patient consents were reviewed and approved by the Institution Review Boards at the University of Texas Health Science Center at San Antonio, the University of Istanbul, and the University of Michigan.
DNA Sequence Analysis
Two distinct enamelin gene sequence variations were identified in two independent and unrelated kindreds with autosomal-dominant amelogenesis imperfecta. In both kindreds, the sequence variations were near intron-exon borders and were predicted to cause major alterations in enamel gene expression or protein structure.
Family 1 was of Iranian descent and consisted of the proband (III-1) and his unaffected mother (II-2), who were given oral and radiographic examinations. Knowledge of the other family members and their affection status was obtained through an interview with the mother. The proband of family 1 was an 11-year-old male (Fig. 3
The proband of family 2 was a 12-year-old female patient in the Pediatric Dentistry Clinic at the University of Istanbul (Fig. 4
Establishing correlations between enamelin gene mutations and their phenotypes may provide insights into the role of enamelin in dental enamel formation. Next, we briefly review the 5 known ENAM mutations and their associated dental phenotypes, and discuss the implications of these findings with respect to the expression and function of enamelin in normal enamel formation. The first reported enamelin gene mutation (g.6395G>A, IVS8+1G>A) spoiled the intron 8 splice donor site, causing a severe form of autosomal-dominant smooth hypoplastic AI (Rajpar et al., 2001). The enamel layer was very thin, making the teeth appear small and yellow. All affected members of the family exhibited enamel hypoplasia in both the deciduous and permanent dentitions. The second reported enamelin gene mutation (g.2382A>T, p.K53X) introduced an upstream translation termination codon in exon 5 (Mårdh et al., 2002). Clinically, this mutation was manifested as autosomal-dominant local hypoplastic AI with horizontal rows of pits, grooves, or large hypoplastic areas in the enamel. Even though there were variations in number and localization of the hypoplastic defects, the phenotype was consistent within families and also between families. The third ENAM mutation (g.8344delG) was a spontaneous mutation in a Japanese family (Kida et al., 2002). One of 6 Gs at the end of exon 9 was deleted, shifting the reading frame after Gly196, and causing the synthesis of a chimeric protein predicted to have 196 amino acids of the wild-type protein (normally 1142 amino acids), followed by 80 novel amino acids, for a total of 276 amino acids (p.N197fsX277) (Hart et al., 2003a). The proband and his younger brother showed hypoplastic enamel in both their deciduous and permanent teeth and hypersensitivity to cold stimuli. Both affected children had an anterior open bite. The affected father showed a local hypoplastic enamel defect demonstrating a horizontal lesion involving primarily the middle 1/3 of the permanent teeth, without hypersensitivity to cold stimuli. This mutation was subsequently identified in a Lebanese family (Hart et al., 2003a), having generalized thin hypoplastic enamel with a variable surface texture ranging from rough, fine, horizontal furrows to surfaces that appeared to have been worn smooth, possibly by toothbrushing. Our proband (family 1) with this mutation exhibited generalized hypoplastic enamel, and the anterior teeth had shallow horizontal grooves in the middle 1/3 of the anterior crowns. The common finding from 3 families with this mutation is a generalized thin hypoplastic phenotype, with surface texture variations from smooth with or without shallow grooves, to rough with numerous shallow pits in parallel horizontal rows. The fourth ENAM mutation (g.13185_13186insAG, p.P422fsX448) was identified in consanguineous kindreds in Turkey (Hart et al., 2003). The phenotype associated with the mutation was dose-dependent: ARAI with an open-bite malocclusion segregated as an autosomal-recessive trait, and enamel pitting as a dominant trait. Homozygotes for the mutation consistently displayed a severe generalized AI phenotype that was clinically hypoplastic and radiographically under-mineralized. All heterozygous carriers of the mutation had localized hypoplastic enamel-pitting defects on the buccal surfaces of the anterior teeth and the occlusal surfaces and cusps of the canines, premolars, and molars. We report the fifth enamelin gene mutation (g.4806A>C, IVS6-2A>C), which alters the intron 6 splice acceptor site causing ADAI. Because enamelin gene expression is restricted to developing teeth (Hu et al., 1997; Hu et al., 2001a; Hu and Yamakoshi, 2003), the effect of a human ENAM mutation on enamelin protein expression and structure must be predicted, since obtaining developing enamel epithelium from a patient for experimental purposes would result in the loss of the developing tooth. Two defective splicing outcomes are probable for this mutation. The first is the inclusion of intron 6 (1615 bp). This would insert multiple, in-frame stop codons preceding the most 3' exon. Translation of this transcript would add 8 novel amino acids to 70 amino acids of the wild-type protein, with the first 39 amino acids constituting the signal peptide. This mRNA would tend to be degraded by the nonsense-mediated decay (NMD) system, or spliced in an unpredictable way by the nonsense-associated altered splicing (NAS) system (Maquat, 2002; Moore, 2002; Wagner and Lykke-Andersen, 2002). A second defective splicing outcome would be the skipping of exon 7 (i.e., deleting intron 6, exon 7, and intron 7 as a single intron). Skipping exon 7 would maintain the reading frame, but would delete 87 amino acids (71–157) from the N-terminal side of the enamelin protein. The mutant protein would still contain segments that are normally generated by proteolytic cleavages, such as the 25-, 32-, and 34-kDa (Tanabe et al., 1990; Uchida et al., 1991) enamelin fragments, but the intact enamelin (186 kDa) and the large enamelin cleavage products (155, 142, 89 kDa) that localize near the mineralization front would be significantly altered (Fukae et al., 1996; Hu et al., 1998). The IVS6-2A>C mutation is predicted to have a significant impact on enamelin expression and protein structure, which increases our confidence that this mutation causes the enamel phenotype in this small kindred. Analysis of the resulting enamel phenotypes and the likely effect of each mutation on enamelin structure and expression have provided insights into how enamelin contributes to normal enamel formation. Enamelin is normally expressed in approximately the amount required for proper dental enamel formation. When the quantity of enamelin is reduced due to haploinsufficiency, a localized hypoplastic enamel phenotype is observed, featuring horizontal rows of pits, grooves, or large hypoplastic areas. When there is a substantial alteration of the enamelin protein structure, such as occurs by the deletion of an exon, a more severe phenotype results from the dominant-negative effect of poisoning the matrix with a defective protein (Hart et al., 2003a). In such cases, horizontal pitting and furrows might be observed, but these grooves are not as obvious because of the generalized thinness of the enamel layer. A reduction in enamel thickness is presumably caused by a failure in crystallite elongation. A direct role for enamelin in crystal elongation was initially proposed when it was learned that intact (uncleaved) enamelin is restricted in its localization to the mineralization front, where crystal elongation occurs (Hu et al., 1997). It is anticipated that further characterization of the genes and mutations that cause amelogenesis imperfecta will provide important insights into the roles of enamel matrix proteins in normal and pathological enamel formation. Mutations in the enamelin gene are predominantly autosomal-dominant in their pattern of inheritance. However, the p.P422fsX448 enamelin mutation caused a severe defect in homozygous (–/–) family members, but only enamel pitting in the heterozygotes (+/–) (Hart et al., 2003). Prior to this finding, minor enamel pitting was not recognized as a clinical form of amelogenesis imperfecta. In our family 2, with the intron 6 splice junction mutation, the three heterozygous (+/–) family members presented with enamel phenotypes that varied more in their severity than any of the previously reported dominant forms. A better understanding of the variations in the penetrance and expressivity of mutations in the genes implicated in the etiology of amelogenesis imperfecta may lead to a broader definition of AI phenotypes (that would include enamel pitting, and other subtle phenotypes such as localized surface roughness or under-mineralization) and provide insights into how genetic variations relate to disparities in caries risk in the general population. Can sequence variations in the genes underlying amelogenesis imperfecta cause subtle enamel defects, such as microscopic pitting or undermineralization, that cannot be detected clinically but, in the right genetic backgrounds, can lead to the formation of dental enamel that is more susceptible to dental caries?
This investigation was supported by the Foundation of the American Academy of Pediatric Dentistry, and by USPHS Research Grants DE12769, DE15846, and DE11301 from the National Institute of Dental and Craniofacial Research. Received for publication April 27, 2004. Revision received January 3, 2005. Accepted for publication January 4, 2005.
Journal of Dental Research, Vol. 84, No. 3,
278-282 (2005) This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||







