| Sign In to gain access to subscriptions and/or personal tools. |
Molar Intercuspal Dimensions: Genetic Input to Phenotypic VariationDental School, The University of Adelaide, South Australia 5005, Australia Correspondence: * corresponding author, grant.townsend{at}adelaide.edu.au
Molecular studies indicate that epigenetic events are important in determining how the internal enamel epithelium folds during odontogenesis. Since this process of folding leads to the subsequent arrangement of cusps on molar teeth, we hypothesized that intercuspal distances of human molar teeth would display greater phenotypic variation but lower heritabilities than overall crown diameters. Intercuspal distances and maximum crown diameters were recorded from digitized images of dental casts in 100 monozygotic and 74 dizygotic twin pairs. Intercuspal distances displayed less sexual dimorphism in mean values but greater relative variability and fluctuating asymmetry than overall crown measures. Correlations between intercuspal distances and overall crown measures were low. Models incorporating only environmental effects accounted for observed variation in several intercuspal measures. For those intercuspal variables displaying significant additive genetic variance, estimates of heritability ranged from 43 to 79%, whereas those for overall crown size were higher generally, ranging from 60 to 82%. Our finding of high phenotypic variation in intercuspal distances with only moderate genetic contribution is consistent with substantial epigenetic influence on the progressive folding of the internal enamel epithelium, following formation of the primary and secondary enamel knots.
Key Words: genetics • teeth • variability • morphology
The molecular basis of dental development has been studied extensively over the past decade (e.g., Maas and Bei, 1997; Tucker and Sharpe, 1999; Jernvall and Thesleff, 2000; Lesot et al., 2001; Sharpe, 2001). These studies have confirmed that tooth morphogenesis is regulated by a series of reciprocal interactions between epithelial and mesenchymal tissues and that occlusal morphology depends on the specific pattern of folding of the ectodermally derived internal enamel epithelium. Indeed, the basal lamina between the internal enamel epithelium and the dental papilla represents the site of the future dentino-enamel junction and is a blueprint for final crown form that becomes established once enamel deposition has been completed. The folding of the internal enamel epithelium is associated with the appearance of groups of non-dividing cells that act as signaling centers. The primary enamel knot is one such transient signaling center that seems to be an important regulator of overall tooth shape and is associated with the development of the cap stage of odontogenesis (Thesleff et al., 2001). Secondary enamel knots form subsequently at the sites of future cusp tips. The timing of their formation and the growth rates of the intercuspal areas of developing crowns determine the arrangement of the cusps (Jernvall and Jung, 2000). Intercuspal dimensions increase until the slopes between are calcified by dentin deposition, causing the distances to become fixed (Butler, 1967; Dahlberg, 1971). One of our previous studies of dental crown morphology in indigenous Australians showed that premolar intercuspal dimensions displayed no evidence of sexual dimorphism, in contrast to maximum mesiodistal and buccolingual crown diameters, where average values for males were significantly greater than those for females (Townsend, 1985). This finding had been reported previously by Garn (1977) in North American children. The premolar intercuspal distances in indigenous Australians also showed greater relative variability and greater fluctuating asymmetry than overall crown diameters. Correlations between intercuspal measures and overall crown diameters were low. In another study of Finns with Turner syndrome (45,X), premolar intercuspal distances were also found to be more variable than overall crown measures (Townsend and Alvesalo, 1995). In fact, intercuspal distances in individuals with X-chromosomal aneuploidies increased with each additional X chromosome, reflecting an associated increase in enamel thickness (Townsend and Alvesalo, 1999). By comparing overall crown dimensions and intercuspal distances of deciduous second molars and permanent first molars between monozygotic and dizygotic twin pairs, we aimed to quantify and then compare the extent of phenotypic variation between these different crown components. By comparing similarities between monozygotic and dizygotic pairs, we also aimed to partition observed variation into genetic and environmental components, and to calculate estimates of heritability. We then aimed to explain our findings in terms of the cellular and molecular events known to occur during odontogenesis. We tested the hypothesis that intercuspal distances would display greater phenotypic variation but lower heritabilities than overall crown diameters, reflecting the importance of epigenetic events in determining the folding of the inner enamel epithelium during odontogenesis and, therefore, the subsequent arrangement of cusps on molar teeth.
Study Population The study sample was comprised of 174 pairs of twins aged 6–14 years who are enrolled in a study of dento-facial development of twins and their families at The University of Adelaide. The twins in this study were of European ancestry, and their zygosities were confirmed by analysis of up to six highly variable genetic loci with multiple alleles (FES, vWA31, F13A1, THO1, D21S11, FGA) on six different chromosomes, with buccal cell DNA. The probability of dizygosity, given concordance for all systems, was less than 1%. The study included 48 monozygotic male pairs, 52 monozygotic female pairs, 32 dizygotic male pairs, 28 dizygotic female pairs, and 14 opposite-sex dizygotic pairs. The Adelaide Twin Study has been approved by the Committee on the Ethics of Human Experimentation, The University of Adelaide (Approval No. H/07/84A), and all participants provided informed consent.
Measurement Methods
Replicability studies were performed on 40 randomly selected subjects, both on pairs of scanned images and also for comparison of measurements obtained directly from casts with those determined from scanned images. Paired t tests were used to detect systematic errors. The technical error of measurement, Se, was used to determine the magnitude of the random error of measurement (Cameron, 1984). In addition, the ratio of the error variance, Se2, to the total observed variance was determined for all variables.
Statistical Analysis
Genetic Modeling
Initially, a path coefficient model incorporating unique environmental influences only (E model) was fitted to the data. Where this model failed, it was extended to include common environmental variance (a CE model) or additive genetic variance (an AE model). Path coefficients (a, c, and e) were estimated, and the chi-squared values for goodness-of-fit of the models were calculated. Akaike’s Information Criterion (AIC =
There was no indication of systematic methodological errors between direct and scanned measurements or between first and second scanned determinations, based on paired t tests. Mean differences between first and second determinations ranged from -0.12 to 0.14 mm. The technical error of measurement ranged from 0.16 to 0.19 mm, and error variances were all less than 10% of total observed variation, confirming that errors of the method were small and unlikely to bias the results. Preliminary analyses confirmed that there were no systematic differences in crown dimensions between the twin sample and a sample of 100 singletons of European ancestry, confirming that the twinning event had not affected tooth dimensions significantly and that the results of the study could therefore be extrapolated to the general population. Comparisons of mean values for monozygotic and dizygotic twins also failed to disclose any systematic differences between zygosity groups. Tests of skewness and kurtosis indicated that all variables conformed to normal distributions and could therefore be described in terms of mean values and standard deviations.
Both intercuspal distances and overall crown diameters were consistently larger in males than in females, but sexual dimorphism was more marked in the mesiodistal and buccolingual dimensions than in the intercuspal distances (Table 1
In the absence of any evidence of systematic directional asymmetry, fluctuating asymmetry was quantified for all variables. The most striking finding, for both males and females, was that the degree of fluctuating asymmetry in all of the intercuspal measures was approximately an order of magnitude greater than that for overall crown dimensions (Table 2  ).
The values of Pearson product-moment correlation coefficients, calculated between the intercuspal distances and between intercuspal distances and overall crown dimensions for each side, were consistently lower than those calculated between overall crown measures. Most of the former fell in the range 0.15 to 0.40, indicating that less than 20% of the variation of one variable could be explained by variation of the other. In contrast, the values of correlation coefficients between overall crown dimensions ranged from 0.50 to 0.70 (Appendix, www.dentalresearch.org).
The best-fitting models were determined for each of the variables, as well as heritability estimates and 95% confidence intervals for those variables with a significant contribution of additive genetic variance (Table 3
Our finding that the magnitude of sexual dimorphism in intercuspal distances was less than that for mesiodistal and buccolingual crown dimensions provides further support for the notion put forward originally by Garn (1977), that "conventional sex differences in crown dimensions are established after the cusps are laid down". It is also consistent with our earlier observations of premolar teeth in indigenous Australians (Townsend, 1985). The low values of correlations between intercuspal distances and overall crown dimensions in the twin sample also matched those obtained previously where crown dimensions were compared between Finnish individuals with Turner (45,X) syndrome and their unaffected male and female relatives (Townsend and Alvesalo, 1995). The higher level of fluctuating asymmetry in intercuspal distances compared with overall crown measures suggests that the location of cusp tips, and therefore the distances between them, are influenced to a greater extent by environmental factors during development. It is generally assumed that the genetic input to bilateral structures, such as teeth, is similar for right and left sides. The small, random deviations in size of antimeric teeth that have been reported widely (Kieser, 1990) are thought to reflect developmental instability, so our finding implies greater instability in the localization of cusp tips than in overall crown size. While the contour of the dentino-enamel junction serves as a blueprint for the final external crown form of teeth, Kraus (1952) reported that enamel growth proceeded "in such a manner that the completed enamel apices are dispersed linguo-buccally more than mesiodistally relative to their dentine analogues". Stern and Skobe (1985) also observed that the buccal cusp apex of mandibular first premolars was located buccally to the dentin buccal cusp apex. When these findings are considered together with those for individuals with Turner and Klinefelter syndromes, it seems that there is a definite trend in the relationship of the buccal and lingual cusps of premolar teeth, with intercuspal distances diverging as enamel thickness increases in association with additional X chromosomes (Townsend and Alvesalo, 1999). Three-dimensional analyses of occlusal surfaces and dentino-enamel junctions of extracted human teeth have confirmed that molar cusp tips tend to diverge as enamel thickness increases (Kanazawa et al., 1987), as have more recent computerized tomography studies of fossil hominid teeth (Smith et al., 1997). The findings of our genetic modeling approach are consistent with those of previous twin studies in which mesiodistal and buccolingual tooth-size data were analyzed in the deciduous and permanent dentitions (Hughes et al., 2000; Dempsey and Townsend, 2001). In these earlier investigations, models incorporating additive genetic and unique environmental variance (AE model) or common environmental and unique environmental variance (CE model) accounted for observed variation in all deciduous teeth. A model including additive genetic and unique environmental variance (AE model) also provided a good fit for most permanent tooth dimensions, although there was evidence of a common environmental influence on maxillary first molar dimensions. Our finding in the present study of a possible effect of common environment on deciduous second molar and first permanent molar crown size variability is not unexpected, given that the crowns of these teeth are calcifying in the peri-natal period, when maternal effects could presumably affect development. The finding that models incorporating unique environmental variance (E model) or common environmental and unique environmental variance (CE model) provided an adequate fit for several intercuspal variables, and the fact that heritability estimates for the other intercuspal measures were, with a couple of exceptions, less than those for mesiodistal and buccolingual crown dimensions provide support for our original hypothesis. It would appear that the marked phenotypic variation of intercuspal distances, as evidenced by their large coefficients of variation and fluctuating asymmetry scores, is associated with a lower genetic contribution than that for overall crown size variation. Can these results be interpreted in the light of recent experimental findings and modeling approaches? We believe that they can. Salazar-Ciudad and Jernvall (2002) have recently devised a mathematical model that does not include any implicit code for cusp position or size but still reproduces the morphology of mammalian tooth crowns by integrating experimental data on gene interactions and growth in developing tooth germs. They have shown that developing morphology seems to have a causal role in patterning of cusps, and that large morphological effects can be achieved by small changes. As they point out, no cusp-specific genes have been identified, but rather the balance between two signaling molecules, one inducing the formation of enamel knots and the other repressing their formation, determines the final cusp pattern. Our findings are consistent with the view that even though the number of secondary enamel knots in developing molar crowns is likely to be under strong genetic control, their arrangement (and hence variation in the distances between the future cusps) is not under such strong direct genetic influence. As far as we are aware, our combined phenotypic and genetic analyses of fully formed molar crowns in twins provide the first evidence in humans to support experimental findings that have shown that the locations of secondary enamel knots, and therefore subsequently formed cusp tips of teeth, are determined by a cascade of epigenetic events, rather than being under direct genetic control (Thesleff et al., 2001).
This study forms part of an ongoing investigation of the teeth and faces of Australian twins and their families that is supported by the National Health and Medical Research Council of Australia (Competing Epidemiological Grant 157904). We particularly thank the twins and their families who have agreed to participate in this research project and The Australian Twin Registry for their continuing assistance. The assistance of Sandy Pinkerton and Wendy Schwerdt is gratefully acknowledged. Preliminary findings were presented at the 80th General Session of the IADR, San Diego, California, in March, 2002.
A supplemental appendix to this article is published electronically only at http://www.dentalresearch.org. Received for publication September 6, 2002. Revision received December 6, 2002. Accepted for publication January 9, 2003.
Journal of Dental Research, Vol. 82, No. 5,
350-355 (2003)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
2 - twice the degrees of freedom) was used to indicate the parsimony of each model. The general approach was to accept a more complex model only when a simpler one had failed. Heritability estimates (h2), that can range theoretically from 0 to 100%, were calculated as the ratio of additive genetic variation to total phenotypic variation in the best-fitting model.