Dental crowding is a problem for both adolescents and adults in modern society. The purpose of this research was to identify single nucleotide polymorphisms (SNPs) responsible for crowding in subjects with skeletal Class I relationships.
The case subjects consisted of healthy Chinese people living in Hong Kong with skeletal Class I relationships and at least 5 mm of crowding in either arch. The control subjects met the same requirements but lacked crowding or spacing. SNP genotyping was performed on the MassARRAY platform. The chi-square test was used to compare genotype and allele type distributions between the case and the control groups. Logistic regression was used to calculate odds ratios with 95% confidence intervals, and the effects of age and sex for each SNP. Analyses of linkage disequilibrium and haplotype associations between SNPs were performed with software.
Five SNPs were found to be significantly different in genotype or allele type distributions. SNP rs372024 was significantly associated with crowding ( P = 0.004). Two SNPs, rs3764746 and rs3795170, on the EDA gene were found to be associated marginally. SNPs rs1005464 and rs15705 also exhibited marginal association with crowding. The effects of associated SNPs remained significant after adjustments for age and sex factors.
This study suggests an association for the genes EDA and XEDAR in dental crowding in the Hong Kong Chinese population.
Crowding is a complex dental anomaly that affects esthetics and quality of life. Crowding is usually caused by insufficient arch space that cannot accommodate all erupting permanent teeth. Genetics is suggested to contribute to the etiology of crowding. Identification of the predisposing markers will help in predicting potential crowding for early prevention or intervention. The genetic markers could be useful when a child is identified as having “crowding potential” in a public health screening situation. These kids could be closely monitored, and then radiographic screening could be justified.
The etiology of crowding has both environmental and genetic components, in which the environmental factors might be related to impacted third molars or early loss of deciduous teeth. In a twin study conducted in the 1940s, dental arch forms, crowding and spacing, and the degree of overbite were suggested to be influenced by heredity. Similarities in dental arch form and tooth position among family members have also been reported. A number of biometric studies have concluded that tooth size (especially mesiodistal length) mainly contributes to the etiology of crowding for those with skeletal Class I malocclusion. Furthermore, it was estimated that the accumulated genetic variation of tooth sizes was as high as 80%, leaving environmental effects at 20%.
Efforts have been made to understand mammalian tooth development. Teeth develop from 2 cellular origins—oral ectoderm and mesenchyme—in which a vast number of signaling pathways are involved. Currently, many genes were discovered to be expressed during tooth development in mammals ; these findings could indicate the targeting genes specifically for tooth anomalies. Genetic polymorphisms have been successfully reported in several forms of tooth anomaly including amelogenesis imperfecta and hypodontia. However, no study has been conducted to identify the genetic associations with crowding.
The aim of this study was to determine possible genetic variations that are associated with crowding. In the search for genetic variants, single nucleotide polymorphisms (SNPs) on each “candidate gene” were compared between the case and the control groups. The advantage of an allelic association over a traditional twin study can be seen from the report that showed little or no genetic influence on temporomandibular disorders in contrast to the latter study of allelic association analysis. The twin study-based method could result in underpowered statistics. SNPs are the most frequent genetic mutations in humans, affecting protein expression and functioning, which then cause disease. Crowding is a common dental anomaly, and, from the “common disease, common variant” hypothesis, it can be speculated that SNPs are important genetic components of crowding.
Material and methods
The case subjects were recruited from the orthodontic clinic at Prince Philip Dental Hospital, Faculty of Dentistry, University of Hong Kong, between December 2007 and December 2008. The control subjects were undergraduate dental students or secondary school students in Hong Kong. Since it was not possible to obtain x-ray images from each control subject for cephalometric analysis (dental and skeletal relationship), facial profile and intraoral examinations were performed. Cross-examination of subjects was performed to minimize selection bias and error; 2 experienced orthodontists were first calibrated and participated throughout the screening sessions. The case subjects were selected according to the following criteria: (1) Chinese ancestry living in Hong Kong; (2) healthy and without orthodontic treatment before attending the orthodontic clinic; (3) full permanent dentition and age less than 27 years; (4) diagnosed with a skeletal Class I relationship according to the ANB angle, Wits appraisal, and facial profile; and (5) not less than 5 mm of crowding in at least 1 arch.
The criteria for the control subjects were similar, except that they were of skeletal Class I profile, Class I molar relationship, and without visible crowding or spacing (±2 mm). The study protocol was approved by the Institute Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster. Written informed consents were obtained from all study participants.
Saliva was used as the source of DNA from all subjects. The Oragene DNA kit (DNA Genotek; Kanata, Ontario, Canada) was adapted for saliva collection, storage, and DNA purification according to the manufacturer’s recommendations. In brief, the subjects refrained from eating for 2 hours and were asked to spit into a collection vial until the indicated mark (3 mL). The collected saliva was then stored at room temperature before DNA extraction.
The purity of each DNA sample was screened by measuring the optical-density ratio against 260/280 nm ultraviolet light absorbance from a photospectrometer; only those with optical-density values greater than 1.7 were used. Ten percent of the DNA samples were randomly selected to be checked by gel electrophoresis to make sure that the samples used were not degraded.
A total of 121 SNPs were selected from 7 candidate genes ( see Appendix ); Tag SNPs and SNPs in the coding regions and 3 and 5′ untranslated regions (UTRs) were selected based on the dbSNP database on the National Center for Biotechnology Information Web site in February 2009.
Genotyping of SNPs was done on a MassARRAY system (Sequenom, San Diego, Calif), which technology was based on the matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. The iPLEX gold assay (Sequenom) was used according to manufacturer’s standard protocol. Assays were designed and checked for accuracy using RealSNP and MassARRAY Designer (Sequenom).
Genotype call rate (≤0.1) and minor allele frequency (≥0.01) were first examined for each SNP. Only SNPs that met these criteria were included in the subsequent statistical analysis. The chi-square test was used to test for SNPs against Hardy-Weinberg equilibrium (HWE) in the control subjects and to compare differences in genotype and allele type frequencies between the case and the control groups. Logistic regression was used to calculate the odds ratios (OR) and 95% confidence intervals (CI) for each SNP. The effects of risk factors such as age and sex were also analyzed by backward stepwise logistic regression. The software package SPSS for Windows (version 16.0; SPSS, Chicago, Ill) was used for all statistical analyses. Linkage disequilibrium (LD) between SNPs and haplotype association analysis were performed by using the Haploview software package (version 4.1; Broad Institute, Cambridge, Mass).
A total of 254 subjects were recruited for this study, including 133 with crowding (mean age, 15.8 years; SD, 3.3 years) and 121 controls without crowding (mean age, 16 years; SD, 3.3 years). The percentages of males in the case and the control groups were 42% and 32%, respectively. Among the case subjects, 47% had severe crowding (>8 mm) in both arches, 42% had maxillary crowding, and the rest had mandibular crowding ( Table I ). Of 121 candidate SNPs, 112 pairs of primers were successfully designed due to technical limitations of the assay used (iPLEX gold). The genotyping success rates for each SNP were over 90%. One control subject could not be assayed, possibly because of poor DNA quality. The success rates for the duplicate check and the blank check were 99.5% and over 97%, respectively.
|Case||Control||Statistical test||P value|
|Mean age (y)||15.8 ± 3.3||16 ± 3.3||t test||NS|
|Male percentage||42%||32%||Chi-square test||NS|
|Maxillary arch||Mandibular arch||Both arches|
|Crowding in case population||42%||11%||47%|
Alleles and genotypes in 5 SNPs were found to be significantly associated with crowding ( Tables II and III ). The SNP rs372024 was significantly associated with crowding ( P = 0.004), in which the allele G could increase the risk by more than 2-fold ( P = 0.008; OR, 2.65; 95% CI, 1.295-5.406). Two SNPs, rs3764746 and rs3795170, on the EDA gene were found to be associated marginally ( P = 0.02 and 0.047, respectively); the allele G on both SNPs exhibited higher prevalences in the crowding subjects (OR, 1.79; 95% CI, 1.09-2.291; P = 0.021 for rs3764746; and OR, 1.43; 95% CI, 1.006-2.032; P = 0.046 for rs3795170). However, in the HWE test, rs3795170 was significantly different from the equilibrium ( P <0.01). The SNP rs1005464 in the BMP2 gene was also found to be significantly associated ( P = 0.017), and another SNP, rs15705, in the same gene exhibited a marginal association ( P = 0.031). Stepwise regression analysis indicated that age and sex did not influence the statistical significance in all 5 SNPs.
|SNP ID||Gene||Gene region||Variation (major/minor)||MAF||P for HWE|
|SNP ID||Genotype/allele type||Crowding (%)||Control (%)||P value ∗||OR †||95% CI||P value|
|rs372024||GG||102 (77)||108 (91)||0.004||2.89||1.375-6.064||0.005|
|GC||30 (22)||11 (9.2)||1||1||1|
|(XEDAR)||/G||234 (88.6)||227 (95.3)||0.006||1||1||1|
|/C||30 (11.3)||11 (4.6)||2.65||1.295-5.406||0.008|
|rs1005464||AA||17 (13.1)||6 (5.1)||0.043||1||1||1|
|GG||54 (41.5)||63 (53.4)||0.43||0.156-1.161||0.095|
|(BMP-2)||GA||59 (45.4)||49 (41.5)||1.41||0.831-2.375||0.205|
|/G||167 (64)||175 (74)||0.017||1||1||1|
|/A||93 (36)||61 (26)||0.63||0.425-0.921||0.017|
|rs15705||AA||38 (29)||21 (17)||0.067||1||1||1|
|CC||29 (22)||36 (30)||0.55||0.293-1.044||0.067|
|(BMP-2)||CA||64 (49)||64 (53)||1.24||0.682-2.260||0.48|
|/A||140 (53)||106 (44)||0.031||0.68||0.478-0.965||0.031|
|/C||122 (47)||134 (56)||1||1||1|
|rs3764746||CC||15 (11)||4 (3)||0.023||1||1||1|
|CG||22 (17)||21 (18)||0.27||0.085-0.833||0.023|
|(EDA)||GG||95 (72)||95 (79)||0.96||0.492-1.851||0.89|
|/G||212 (80)||211 (88)||0.02||1.79||1.09-2.291||0.021|
|/C||52 (20)||29 (12)||1||1||1|
|rs3795170||GG||38 (29)||48 (40)||0.094||1||1||1|
|GT||38 (38)||29 (24)||1.65||0.919-2.945||0.094|
|(EDA)||TT||56 (56)||43 (36)||0.99||0.532-1.858||0.985|
|/T||150 (57)||115 (48)||0.047||1||1||1|
|/G||114 (43)||125 (52)||1.43||1.006-2.032||0.046|