| Sign In to gain access to subscriptions and/or personal tools. |
A Mutation in the Enamelin Gene in a Mouse Model
1 Department of Prosthetic Dentistry, University Medical Center, Hamburg-Eppendorf, Martinistr. 52, D-20246 Hamburg, Germany; Correspondence: * corresponding author, seedorf{at}uke.uni-hamburg.de
Amelogenesis imperfecta is an inherited disorder affecting tooth enamel formation. We previously isolated a mouse strain with an amelogenesis imperfecta phenotype (ATE1 mice) from a dominant ethylnitrosourea screen and mapped the disease-causing defect to a 9-cM region of mouse chromosome 5. In the current study, we tested the hypothesis that there is a mutation in enamelin (ENAM) or ameloblastin (AMBN), both of which are located wihin the linkage region, by sequencing these two candidate genes. Analysis of our data shows that the amelogenesis imperfecta phenotype is linked to a C > T transition in exon 8 of the enamelin gene. The mutation predicts a C826T transition, which is present in the enamelin transcript and changes the glutamine (Gln) codon at position 176 into a premature stop codon (Gln176X). Conversely, no mutation could be detected in the ameloblastin gene. These results define the ATE1 mice as a model for local hypoplastic autosomal-dominant amelogenesis imperfecta (AIH2), which is caused by enamelin truncation mutations in humans.
Key Words: amelogenesis imperfecta • ethylnitrosourea-induced mutagenesis • mutational analysis • mouse disease models • dental enamelin proteins
Tooth enamel, the exterior tissue of vertebrate teeth, is a biomineral with remarkable hardness and resistance to physical and biochemical attack (Schroeder, 1992). Its unique properties are related to large, highly organized hydroxyapatite crystals that contribute more than 90% to the total weight of mature enamel (Brudevold and Störenmark, 1967). The hydroxyapatite crystals’ high level of organization is the result of a complex biomineralization process, which is controlled by a regulated expression of multiple genes (Paine et al., 2001). Disorders in this biomineralization process of enamel can lead to diseases, one of which is amelogenesis imperfecta. To date, mutations in 5 genes [amelogenin (AMELX), enamelin (ENAM), kallikrein-4 (KLK4), enamelysin (MMP-20), and DLX3 (Dong et al., 2005)] have been found to cause amelogenesis imperfecta. Ameloblastin, which has been reported to have a modulating effect on calcium phosphate crystal morphology (Moradian-Oldak et al., 2003), is another candidate gene. Ameloblastin-null mice show severe enamel hypoplasia and develop odontogenic tumors (Fukumoto et al., 2004). However, the involvement of ameloblastin in amelogenesis imperfecta has not yet been fully understood (Stephanopoulos et al., 2005). We previously introduced a mouse strain with an amelogenesis imperfecta phenotype, which can serve as a new mouse model for hypoplastic autosomal-co-dominant amelogenesis imperfecta (ATE1 mice) (Seedorf et al., 2004). The phenotype was visible to the naked eye in homozygotic and heterozygotic animals. Microradiographs and toluidine-blue staining of non-decalcified sections showed complete loss of enamel in the front and molar teeth of homozygotic ATE mutants, cracked enamel of reduced width (approx. 50%) in incisor and molar teeth of heterozygotic ATE mutants, and regular structure of incisor and molar teeth of wild-type mice. Scanning electron micrographs of teeth of homozygotic ATE mutants showed loss of enamel and exposed dentinal tubules (Seedorf et al., 2004). We mapped the defect to a 9-cM region of mouse chromosome 5, corresponding to human chromosome 4q21, the location of the human ameloblastin and enamelin genes) (Hu et al., 2000), and hence considered ameloblastin and enamelin to be the strongest candidates within this region to cause the disease in ATE1 mice (Seedorf et al., 2004). Ameloblastin (also called amelin or sheathlin) is part of the organic enamel matrix. It is expressed at high levels by ameloblasts during enamel formation (Fong et al., 1996a), and also by odontoblasts (Nagano et al., 2003), in Hertwig’s epithelial root sheath (Fong et al., 1996b) and in odontogenic tumors (Toyosawa et al., 2000). The mature protein is composed of 447 amino acids. Soon after secretion, the initial cleavages generate small polypeptides containing the N-terminus and relatively large polypeptides containing the C-terminus. The role of the protein in mineralization remains obscure. The murine gene consists of 11 exons and 10 introns. The putative start codon is at 82 base pairs in exon 1. The transcript is alternatively spliced in humans (Mardh et al., 2001) and mice (Simmons et al., 1998). No amelogenesis-imperfectacausing mutations in ameloblastin have been reported (Stephanopoulos et al., 2005), while several mutations causing an amelogenesis imperfecta phenotype have been described for enamelin (Stephanopoulos et al., 2005). Some of these mutations were induced by N-ethyl-N-nitrosourea (ENU) (Masuya et al., 2005). Enamelin is the largest and least abundant of the 3 enamel matrix proteins (enamelin, amelogenin, and ameloblastin), representing about 5% of the total proteins found in developing enamel (Uchida et al., 1991). It is a tooth-specific gene expressed by the enamel organ and, at a low level, in odontoblasts (Nagano et al., 2003). The mouse gene consists of 10 exons, while the human enamelin gene contains 9 exons, because the sequence corresponding to mouse exon 2 is absent in humans, though the intron between the first and second exons in humans shows a sequence homologous to mouse exon 2. This region is flanked by splice junctions, so that alternative splicing may occur (Hu et al., 2001). Like all enamel matrix proteins, enamelin goes through proteolytic processing that leads to cleavage products of 155 kDa, 142 kDa, 89 kDa, 34 kDa, 32 kDa, and 25 kDa (Fukae et al., 1996). This processing is believed to be of great importance for proper enamel biomineralization (Stephanopoulos et al., 2005). In this study, by sequencing the enamelin and ameloblastin gene, we tested the hypothesis that a mutation in the enamelin or ameloblastin gene is the cause of the disease in ATE1 mice.
Animals The ATE1 mutants were obtained by intraperitoneal injection of ethylnitrosourea in wild-type C3HeB/FeJ animals at the National Research Center for Environment and Health GmbH, Helmholtz Association, according to the German law governing the protection of animals. Isolation of the ATE1 strain and breeding conditions were described previously (Seedorf et al., 2004).
Genome-wide Linkage Analysis For linkage analysis, a genome-wide mapping panel consisting of 157 single-nucleotide polymorphism (SNP) markers was applied. Genotyping of this panel was performed with MassEXTEND, a MALDI-TOF high-throughput genotyping system supplied by Sequenom (San Diego, CA, USA).
Mutation Screening and Genotyping
Heterozygotes and homozygotes were categorized according to their phenotypes (see Fig. 1
Genome-wide linkage analysis revealed a strong linkage [  2 = 57, –log10 (P) = 13.36] on chromosome 5 with SNP marker rs31585424 (Mouse Genome Assembly [MGA] 36.1: 80.03 Mb). Haplotype analysis of this region showed no recombinant allele at marker rs31585424, one recombinant allele at rs13478429, and 6 recombinant alleles at marker rs29635956 (Fig. 2 ). The maximum 95% confidence interval of rs31585424 and the putative mutated gene is 6.2 cM, either proximal or distal to that marker. This can be narrowed down by calculation of the minimum confidence intervals of the surrounding markers. At a 95% level, these are 5 cM for rs29635956 and 0.4 cM for rs13478429. Considering the genetic distances of 10.7 cM between rs29635956 and rs31585424 and 1.8 cM between rs31585424 and rs13478429, this means that the target region must start 5.7 cM proximal and end 1.4 cM distal to rs31585424. Roughly translated to physical distances, the linkage region spans an area from 73 Mb to 100 Mb on chromosome 5.
Linkage with markers on other chromosomes could not be established with a probability P  0.01 or –log10 (P) 2 (Table).
Two candidate genes, enamelin (Enam, MGA 36.1: 89.58 Mb) and ameloblastin (Ambn, MGA 26.1: 89.53 Mb), are located next to each other within the linkage region. Hence, we conducted mutation screening of the enamelin and ameloblastin genes, but not of Dmp1, osteopontin, or integrin-binding sialoprotein.
To screen for amelogenesis-imperfecta-causing mutations, we sequenced all exons, including flanking intron sequences and two overlapping DNA fragments of ca. 500 bp, located upstream of the transcription initiation sites, which harbor the putative core promoters of the two genes. In the case of the ameloblastin gene, identical sequences were obtained from homozygous amelogenesis imperfecta mice and mice from the C57BL/6 reference strain (data not shown). However, a sequence deviation could be detected in the 8th exon of the enamelin gene (Fig. 3a
In this manuscript, we present the positional cloning of the disease-causing gene in the ATE1 mutant mouse strain, which resulted from a dominant ENU mutagenesis screen (Seedorf et al., 2004). Our results demonstrated that abnormal tooth enamel formation is linked to a mutation in exon 8 of the enamelin gene. The mutation, a C826T transition which is predicted to be present in the enamelin transcript, converts the Gln codon at position 176 into a premature stop codon (Gln176X). ENU is a powerful alkylating mutagene acting on spermatogonial stem cells in mice. Although the most commonly reported ENU-induced mutations are AT-to-TA transversions or AT-to-GC transitions, our finding of a GC-to-AT transition is consistent with the ENU mutagenesis mechanism, and ENU-induced GC-to-AT transitions have been described previously in several cases (Jansen et al., 1994; Masuya et al., 2005). In a study of ENU-induced mutations, Noveroske et al.(2000) found that 63% of mutations were missense mutations, 26% caused abnormal splicing, and 10% resulted in nonsense mutations. Like all enamel matrix proteins, enamelin goes through a complex pathway of proteolytic processing that involves multiple cleavage products (Fukae et al., 1996). This processing is believed to be crucial for proper enamel biomineralization (Stephanopoulos et al., 2005). Intact enamelin and the cleavage products containing the C-terminus are exclusively present at the superficial layers of the developing enamel matrix, whereas the downstream cleavage products accumulate predominantly in the deeper layers (Hu et al., 1997). In porcine enamel, enamelin is secreted by ameloblasts as a 186-kDa precursor, which accumulates along the secretory face of the ameloblast Tomes’ process. Proteolytic processing toward the C-terminus via a 155-kDa intermediate leads to the 142-kDa form of enamelin. Subsequent cleavage steps result in the 89-kDa fragment consisting of the first 627 amino acids of the enamelin precursor and a 34-kDa polypeptide starting at residue 632. Further processing of the 89-kDa fragment results in polypeptides of 32 kDa (containing amino acids 136–238) and 25 kDa (containing amino acids 477–638) (Fukae et al., 1996). Since the Gln176X mutation leads to a truncated peptide containing only the first 175 amino acids of the 1142-amino-acid enamelin precursor, all functionally important enamelin-processing products would either be absent or truncated significantly in the ATE1 mice. Thus, it seems plausible to us that the Gln176X variant would not be able to mediate proper enamel biomineralization, and that the variant is essentially nonfunctional. The amelogenesis imperfecta phenotype has previously been mapped to a ~ 9-cM region between markers D5Mit10 and D5Mit18 at map position 45–54 cM, according to the Mouse Genome Informatics (MGI©) database of the Jackson Laboratory© gene map, version 3.5 (Seedorf et al., 2004). Since ameloblastin and enamelin are located next to each other at map position 46 cM in this map, our previous linkage data are consistent with our current finding of a nonsense mutation in enamelin. In contrast, the most recent version of the mouse Ensemble gene map assigns the two markers to map positions 94.4 and 104.5 Mb, while the ameloblastin-enamelin gene cluster is placed outside the linkage region at map positions 89.53 to 89.58 Mb. At first glance, the findings of the previous study would therefore be in contradiction to the data presented here. However, the assembly and annotation of the physical map of the mouse genome are still incomplete, and contigs may be relocated in the future. Additionally, analysis of the linkage data based on a set of markers with a proven physical location showed strong linkage with the region where the enamelin/ameloblastin cluster is located. The phenotype of the C826T is consistent with that reported for the EnamRgsc521 by Masuya et al.(2005). Both mutants have a truncated enamelin-protein, due to a premature stop codon, and, for both, homozygotes differed from heterozygotes based on their enamel abnormalities. Whereas the teeth in heterozygotes had a whitish, chalk-like appearance, and microradiographs and toluidine-blue staining of non-decalcified sections showed cracked enamel of reduced width (approx. 50%), homozygote C826T and EnamRgsc521 showed complete loss of enamel. Thus, the mutation is inherited as a co-dominant trait. In summary, our results provide further support for enamelin as a prominent factor involved in the pathogenesis of amelogenesis imperfecta, and define ATE1 mice as a model for local hypoplastic autosomal-dominant amelogenesis imperfecta (AIH2), which is caused by enamelin truncation mutations in humans as well.
The authors thank Helga Reschke for expert technical assistance regarding PCR amplifications and DNA sequencing. Financial support was provided by each of the institutions involved in the study.
A supplemental appendix to this article is published electronically only at http://www.dentalresearch.org. Received for publication July 13, 2006. Revision received February 9, 2007. Accepted for publication April 16, 2007.
Journal of Dental Research, Vol. 86, No. 8,
764-768 (2007) This article has been cited by other articles:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
