SEARCH

SEARCH BY CITATION

Keywords:

  • Dentistry;
  • genetics;
  • epigenetics;
  • phenomics;
  • morphometrics

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Approaches to Phenotyping
  5. Phenotyping and Oral Diseases
  6. Phenotyping and Craniofacial Development
  7. Computer Modelling of Genotype-Phenotype Correlations
  8. Future Directions
  9. Conclusions
  10. Acknowledgements
  11. Disclosure Statement
  12. References

The field of dental phenomics provides many opportunities to elucidate the roles of genetic, epigenetic and environmental factors in craniofacial development. To date, research findings have helped to clarify the pathogenesis of many conditions, aiding diagnosis and clinical management. This paper provides an overview of dental phenomics research in some commonly encountered oral diseases in everyday clinical practice, as well as research relating to craniofacial growth and development. Clinically, advances in cariology and periodontology have led to better diagnostic capabilities and treatment provision. In the study of growth and development, important information regarding the varying clinical presentation and pathogenesis of many disorders is now apparent through the accurate quantification of phenotypes. Improvements in two-dimensional (2D) and three-dimensional (3D) imaging and analytical techniques have allowed for accurate dental phenotyping, and efforts are ongoing to apply these in vitro techniques to the in vivo setting. The field of dental phenomics represents an exciting avenue that links research findings to practical application, and collaboration between researcher and clinicians will help advance the field further.


Abbreviations and acronyms
CL/P

cleft lip and/or palate

CPP-ACP

casein phosphopeptide-amorphous calcium phosphate

DP

digital photography

EDI

Enamel Defects Index

GIS

geographic information system

n-HAp

nano-hydroxyapatite

OPC

orientation patch count

OPCR

orientation patch count rotated

QLF

quantitative light-induced fluorescence

SMMCI

solitary maxillary median central incisor

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Approaches to Phenotyping
  5. Phenotyping and Oral Diseases
  6. Phenotyping and Craniofacial Development
  7. Computer Modelling of Genotype-Phenotype Correlations
  8. Future Directions
  9. Conclusions
  10. Acknowledgements
  11. Disclosure Statement
  12. References

The Human Genome Project has led to major advances in genomics and revealed new frontiers for research and collaboration.[1] However, genetic input alone does not determine all phenotypic traits; rather, the final phenotype of an organism is established via the dynamic interactions between genetic, epigenetic and environmental factors. In dental development, these interactions account for variations in tooth number, size, shape and mineralization within and between species.[2] Advances in phenotyping are crucial in elucidating the roles of these factors,[3-6] and the interactions between them, as well as to further understand the development of the dentition as a Complex Adaptive System.[7]

Phenomics is defined as ‘the acquisition of high-dimensional phenotypic data on an organism-wide scale’[6] and has progressed significantly in various fields over the past decade. Studies of phenotypic traits in human diseases, such as neuropsychiatric syndromes[8] and cardiovascular complications,[9] have provided pre-symptomatic biomarkers and insights into disease risk factors. Further understanding of disease pathology and future treatment trends has also been gained via animal modelling of human diseases.[10, 11] Plant phenomic research has improved both food quantity and quality by identifying desirable phenotypes and their corresponding genotypes.[12, 13]

‘Dental phenomics’ was proposed initially as a new field for quantification of phenotypic variations in the dentition,[14] and it has evolved to encompass other craniofacial structures relevant to the field of dentistry. The human dentition provides a valuable dataset, as individual teeth have varying developmental time frames. The complex development of each tooth occurs over several years, while the development of the whole human dentition extends, on average, from the sixth week in utero to 20 years of age. The teeth and jaw bones are resistant to degradation and provide an accessible record of development to highlight the effects of developmental or pathological events at specific times. Hence, advances in measuring dental phenotypes can enhance the genotype-phenotype correlation[15] to identify contributing factors to observed variations. Clinically, accurate quantification of changes in dental tissues can guide diagnosis and treatment provision, illustrating the relevance of dental phenomics to clinical practice.

In addition, dental phenomics is applicable to both dental anthropology and forensic odontology. Sophisticated imaging and analytical techniques serve as important adjuncts to crime scene investigation and victim identification. Morphometric analysis of tooth dimensions[16] and the human mandible[17] have provided useful information relating to sexual dimorphism from skeletal remains. In archaeological studies, three-dimensional (3D) comparisons of human and great ape craniofacial morphology have provided insight into evolutionary changes that account for the distinct differences between species.[18] Assessment of enamel hypoplasia from skeletal remains has also been used to assess health in the past.[19]

The aims of this paper are to evaluate approaches used to quantify dental phenotypes and to highlight advances and future opportunities in key research areas, together with their clinical relevance. Although new technologies are being applied to phenotyping in many areas of dental research and practice, this review focuses on phenotyping of common oral diseases, including dental caries and periodontal disease, as well as craniofacial development.

Approaches to Phenotyping

  1. Top of page
  2. Abstract
  3. Introduction
  4. Approaches to Phenotyping
  5. Phenotyping and Oral Diseases
  6. Phenotyping and Craniofacial Development
  7. Computer Modelling of Genotype-Phenotype Correlations
  8. Future Directions
  9. Conclusions
  10. Acknowledgements
  11. Disclosure Statement
  12. References

In dental phenomics, a decision to carry out either extensive or intensive phenotyping is made based on the research question. Extensive phenotyping involves measuring a small number of parameters in a large number of specimens or individuals,[20] while intensive phenotyping measures a large number of parameters in a limited number of specimens.[21] A comparison using examples of both methods is presented in Table 1. During the development of analytical methods, accuracy (the closeness of measured values to the true value) and precision (the closeness of repeated measurements of the same quantity) are essential components to consider.[22] In addition, the methods should be valid, i.e. the obtained measurements should provide an accurate representation of what is intended for assessment. Figure 1 provides a visual representation of the concepts of accuracy, precision and validity.

Table 1. Examples of extensive and intensive phenotyping
Extensive phenotypingIntensive phenotyping
In one study, a large number of individuals (n = 134) was examined to identify the effects of a single contributing factor (intrauterine hormone effect) on a small number of parameters (tooth dimensions).[20]In another study, a smaller number of samples (5 from each subgroup of differing genetic make-up) was utilized to study the specific roles of certain genes (Amelx and Enam). Each sample was examined for a number of parameters, such as tooth colour, surface whiteness and jaw dimensions.[21]
image

Figure 1. Precision, accuracy and validity (adapted from Harris and Smith[22]). The bull's eyes on the ‘red’ dart boards represent the intended characteristic to be tested. The yellow crosses/darts represent the results obtained. Scenario 1 demonstrates both low precision and low accuracy. Scenario 2 demonstrates high precision, but low accuracy. Scenario 3 demonstrates both high precision and high accuracy. Lastly, scenario 4 demonstrates high precision and accuracy, but poor validity as the darts have completely missed the red board (hitting the purple board instead), thus the method used does not inform on the parameters of interest.

Download figure to PowerPoint

Phenotyping and Oral Diseases

  1. Top of page
  2. Abstract
  3. Introduction
  4. Approaches to Phenotyping
  5. Phenotyping and Oral Diseases
  6. Phenotyping and Craniofacial Development
  7. Computer Modelling of Genotype-Phenotype Correlations
  8. Future Directions
  9. Conclusions
  10. Acknowledgements
  11. Disclosure Statement
  12. References

Cariology

Dental caries is one of the most prevalent chronic diseases worldwide,[23] with its prevention, diagnosis and treatment comprising a major component of everyday clinical practice. Over the years, an enhanced understanding of caries aetiology and the application of minimal intervention principles has revolutionized caries management.[24, 25] The ability to accurately quantify caries is of great importance in diagnosing and monitoring lesions, while also assessing the effectiveness of prescribed intervention.

Caries assessment

Several techniques that rely on the principles of light transmission, absorption and scattering have been described to indicate changes in the mineral content of teeth. The use of quantitative light-induced fluorescence (QLF) for caries detection[26, 27] and monitoring[28] has shown good validity and may direct a ‘non-restorative’ approach to caries management. Currently, no ‘actual’ clinical value has been established to indicate an active lesion, but comparison between initial records with subsequent readings can be used to deduce this information. A change in fluorescence indicates the progression of a lesion and may help to justify the need for restorative intervention. Digital photography (DP) has also been used to assess carious lesions.[29, 30] This technology may be feasible for surgeries that already possess photographic devices. However, variations in lighting conditions, surface reflection and tooth position can alter the image properties, thus complicating the assessment of a lesion. In addition, difficulty has also been reported in comparing two images of the same tooth at different time points,[30, 31] indicating the need for stringent standardization. In a recent study, the absence of glare and reduced variability associated with QLF indicated that it is a preferable technique in caries assessment compared to DP.[31]

In vitro, micro-CT techniques have been employed to quantify changes in the mineral content of tooth structure.[32] This technique has been used to quantitatively evaluate the efficacy of preventive and restorative intervention. The use of casein phosphopeptide-amorphous calcium phosphate (CPP-ACP) with sodium fluoride (CPP-ACPF) was found to promote higher levels of remineralization of enamel when compared to the use of CPP-ACP alone.[33] In carious dentine, an increase in radiopacity due to mineral deposition from the overlying glass ionomer cement justifies a ‘minimally invasive’ approach to caries management.[34] An increase in remineralization potential by nano-hydroxyapatite (n-HAp) containing toothpaste when compared to fluoride containing toothpaste has also been demonstrated using micro-CT technology.[35] These examples demonstrate how advances in dental phenomics can reaffirm or direct various treatment modalities based on stringent evaluative protocols.

Caries heritability

Recently, Shaffer et al.[36] have used patterns of caries distribution as phenotypes to further clarify the multifactorial nature of dental caries. Data on familial history and genetic markers were obtained and the authors noted caries patterns in the mandibular canines and premolars as being strongly heritable, suggesting a genetic aetiology.[36] The authors concluded that the use of caries distribution provides better risk assessment compared to traditional DMFS/ DMFT indices as it takes into account the site specific nature of caries risk.

Periodontology

The periodontal tissues demonstrate varying responses to disease risk factors and dental treatment (periodontal, restorative and surgical). Hence, this represents a critical consideration during treatment planning. Phenomic research is commonly carried out at both the clinical (visible) and cellular (microscopic) levels to understand the dynamic nature of the periodontal tissues.

Biotype assessment

Clinically, gingival tissue phenotypes can be classified as ‘thin’ or ‘thick’ biotypes, based on the width and thickness of gingival tissue as well as tooth form.[37] The ‘thin’ biotype has been associated with an increased risk of recession, as a result of orthodontic tooth movement[38] and restorative treatment. In addition, tissue characteristics around an implant are critical to maintain long-term aesthetics.[39, 40] Hence, different methods have been proposed to quantify variations in gingival tissues. Previously, the use of an ultrasonic device was deemed a non-invasive[41, 42] and reproducible[43] way to characterize gingival tissues. However, care must be taken to maintain the position of the transducer probe to ensure an accurate reading is obtained.

Another simple method to discriminate thin from thick gingivae is based on the transparency of the periodontal probe through the gingival margin.[44] This method involves placing a periodontal probe into the gingival sulcus and assigning a score based on the visibility of the probe through the gingival tissue. While this technique is non-invasive and cost-effective, it is prone to errors from visual inspection, potentially leading to aesthetic complications following surgical and/or restorative therapy.[45] Hence, a more accurate method to clinically assess variations in gingival tissues is required to ensure thorough preoperative planning. In addition, identification of genetic factors that influence these variations will help to explain the phenotypic features of gingival cells (i.e. cell populations, differentiation capacity and immunomodulatory properties)[46] and advance tissue engineering research (please refer to the paper by Han and colleagues[47] in this issue for further discussion on this topic).

Risk assessment

At the cellular level, variations in periodontal ligament cell[48] and fibroblast phenotypes[49] have provided insight into tissue responses to both normal and pathological processes. Various risk factors have also been established for periodontal disease. For example, a comparison between monozygotic and dizygotic twins[50] yielded a heritability estimate of 50% for adult periodontitis, which remained stable despite adjustments to behavioural variables such as smoking. Other studies reveal variations in T-cell phenotypes[51] and genetic polymorphism,[52] reinforcing the notion that periodontal disease is the consequence of complex interactions between host and environmental factors.[53, 54] Ultimately, such information helps to direct the profession in risk assessment and treatment provision.

Phenotyping and Craniofacial Development

  1. Top of page
  2. Abstract
  3. Introduction
  4. Approaches to Phenotyping
  5. Phenotyping and Oral Diseases
  6. Phenotyping and Craniofacial Development
  7. Computer Modelling of Genotype-Phenotype Correlations
  8. Future Directions
  9. Conclusions
  10. Acknowledgements
  11. Disclosure Statement
  12. References

Tooth morphology: size and shape

Traditional techniques for studying dental morphology have been observational, including the development of non-continuous indices, classifications and manual measurements of linear dimensions, i.e. mesio-distal and bucco-lingual crown dimensions. While the use of indices and classifications can have satisfactory levels of intra-operator reproducibility,[55] variations between observers may exist with this approach. In addition, linear measurements only provide limited information concerning structures of complex forms, i.e. tooth/arch/facial shape. More sophisticated analytical techniques are needed to overcome these limitations. New measurement techniques[56-58] can provide accurate and reliable information for additional parameters regarding the size, shape, structure and colour of teeth. Concurrent assessment of linear measurements, curved surfaces and the superimposition of images allows for the accurate description and quantification of variations among individuals.

Two-dimensional digital imaging

Two-dimensional (2D) digital imaging represents an important addition to the analysis of dental morphology. Studies relating to tooth dimensions have measured the phenotype of specific genetic contributions to odontogenesis. For example, in cases of hypodontia, a reduction in crown dimensions in patients with PAX9 mutation suggests a major role of this gene in odontogenesis.[59] On the other hand, larger tooth size dimensions have been observed in individuals with supernumerary teeth, suggesting a common factor which influences both the dental anomaly and the overall dentition.[60, 61] A summary of 2D imaging methodology is provided in Table 2 to highlight the advantages and limitations of this approach and serve as a basis for future research development. This technique is not only applicable to the assessment of dental morphology alone, but also useful for a wide range of clinical and research purposes.

Table 2. Summary of 2D digital imaging techniques (applicable to the study of both dental morphology and enamel defects)
2D imaging (in vitro)
EquipmentAdvantagesLimitations

An adjustable stand to mount study models OR a single tooth holder

Standardized lighting

A scale

A digital camera

Polarizing capacity

Fixed positioning/ reproducible imaging with calibration

Able to assess a variety of dimensions[59]

The specimens are not damaged

Digital images can be stored and reinvestigated in other locations for comparison between different studies[94]

Ability to take polarized images to enhance opacity imaging by removal of reflection.

Careful calibration of the digital camera is required as settings and image presentation can alter colour and surface characteristics of the specimen surface

Light reflection from surface can alter visual properties – thus, manipulation of illumination is critical in studying tooth surfaces[94]

The samples need to be re-positioned when different surfaces are studied

2D imaging (in vivo)
EquipmentAdvantagesLimitations

Mobile digital imaging apparatus including a frame to support a digital camera and lighting, which facilitates precise positioning of camera, lights and patient (and/or tooth)

Semi-automatic software calibration

Adjustable scale for image calibration

Polarization facility

The patient's head is placed on a purpose built chin rest for alignment. An adjustable scale may be moved into the plane of the image facilitating multiple positioning for variation in size of both patient head and/or tooth

The unit can be used in the field, as the components of the unit can be easily dismantled, transported and reassembled

Software facilitates calibrated colour images specific to the ambient lighting conditions and digital camera performance in relation to fixed absolute whiteness, lightness, and colour shift values

Polarization facilitates opacity imaging without reflections

Requires careful calibration of imaging technique and software

Adjustment of inbuilt lighting is also critical, especially when assessing differences in colour

Limited application for posterior teeth imaging

Three-dimensional imaging

Methods for 3D imaging of dental morphology began in the 1960s. Scanning modalities such as stereophotogrammetry,[62-64] the Optocom[65] and the Reflex Metrograph[66] were used for the assessment of teeth, dental casts and facial features. However, the implementation of these early techniques was extremely time-consuming, lacked measurement accuracy and showed substantial inter-operator variability. Early computer technology also had limited analytical and storage capabilities, thus hampering the widespread application of 3D imaging techniques. The advent of laser scanning and computer-aided tomography has overcome these shortcomings, resulting in the increased use of 3D technology.

Generally, the 3D equipment consists of an image acquisition scanner connected to appropriate computer software. Sufficient scanning resolution is important to ensure image definition, while the appropriate software must be able to clean, align and merge the separate images or models accurately. Analysis of the acquired images enables measurement of various dimensions to quantify variations in size and shape with a high degree of accuracy and reliability.[67-72] Figure 2 provides an example of surface area and angular measurements that are possible with the use of 3D imaging.

image

Figure 2. 3D model of a study model, demonstrating the angle between the buccal and palatal cusp tips (to indicate cuspal inclination) of the upper left first premolar and the occlusal surface area of the upper left second premolar. The picture on the right is an enlarged view of the area of interest.

Download figure to PowerPoint

In one study, scanned 3D models were ‘mapped’ against mathematical grids to objectively quantify the effects of MSX-1 genetic mutation on tooth morphology in patients with hypodontia.[73] The authors noted variations in central incisor size, suggesting the role of different signalling pathways and the genes involved in odontogenesis. The data on dental morphology can also supplement clinical examination[74, 75] and radiographic records[74, 75] to clarify the genetic, epigenetic and environmental factors involved in dental agenesis. Future research involving dental agenesis, both in syndromic and non-syndromic forms, will help identify specific contributing factors to the condition.[76] It is common to detect alterations in tooth number and size in syndromes such as ectodermal dysplasia, suggesting the pleiotropic nature of genes during craniofacial development.[77] Information obtained from various studies will also be applicable to the emerging fields of tooth bioengineering and regenerative dental medicine.[78]

Superimposition

2D and 3D morphometric analysis of shape can provide valuable insights into genetic and environmental contributions to observed variations between specimens. Generally, each specimen is represented by the relative positions of morphological landmarks that can be located to establish one-to-one correspondence among all tested specimens. This can be particularly challenging during the comparison of objects of dissimilar size, position and orientation. Hence, the use of Procrustes superimposition removes these differences and is at the core of geometric morphometrics,[79-81] and this approach has been applied to investigate tooth shape.[82] Essentially, landmarks were established on 2D images of teeth, and a linear model was generated for superimposition to compare the variations in tooth shape between individuals. While the results showed some promise, the authors highlighted challenges such as inconsistencies in orientation[82] and landmark establishment on curved surfaces,[83] both resulting in low inter-operator reliability and potential inaccuracies in data analysis.

On the other hand, the assessment of a solitary maxillary median central incisor (SMMCI) using the Procrustes analysis demonstrated good validity when measuring symmetry within a single tooth.[84] SMMCI may be related to holoprosencephaly, which affects midline structures. The premature fusion of the dental lamina from both sides of the maxilla results in a solitary incisor. As this tooth represents the distal halves of both upper central incisors, it demonstrates a high degree of symmetry.[84] Therefore, if a highly symmetrical lone central incisor is noted, referral to a geneticist may be warranted,[85] provided factors such as traumatic loss and impaction have been accounted for.

The use of 3D images can supplement the information provided via 2D superimposition alone. Smith et al.[86] developed a method for 3D superimposition to form a comparable interface between non-similar objects (Fig. 3). This approach demonstrated good reliability but the authors highlighted potential challenges with handling 3D images, i.e. the images are not ‘fixed’ and the orientation of the images can be a source of error. Hence, research to establish a protocol is ongoing to reduce possible analytical problems during image comparison. A study utilizing an elastic style of registration (similar mathematically to the diffusion approach used by Smith et al.[86]) has shown increased sensitivity when comparing 3D datasets.[87] These techniques have been shown to be more reliable in establishing differences during the superimposition of non-identical objects, when compared to the best-fit model, and future application to analysis of dental structures is warranted.

image

Figure 3. 3D synetic superimposition of upper permanent first molars in a pair of monozygotic twins demonstrating differences in shape and surface area. Despite monozygotic twins sharing the same genetic information, the differences indicate that other confounding factors are involved in the final tooth morphology.

Download figure to PowerPoint

Complexity

A relatively new method for quantifying morphological variation is dental complexity.[58] The objective of this method is to quantitatively characterize variation in morphology that cannot be captured using linear measurements or landmarks. Using a 3D model of the tooth surface, geographic information systems (GIS) algorithms calculate the orientation or aspect of each point on the surface as one of eight cardinal directions, e.g. north, north-east. Contiguous surface points that face the same direction are grouped into a patch, and the total number of patches over the surface is the ‘orientation patch count’ (OPC). This calculation can be repeated to give values that are independent of rotation in the x,y plane, which is then called ‘orientation patch count rotated’ (OPCR).[88] Figure 4 illustrates the ability of OPC to characterize fine differences in shape – the higher morphological complexity of the tooth on the left is quantified as an OPCR of 51.9 compared to 47.5 for the other. While this method has largely been used to investigate gross differences in dental morphology among animal species,[58, 89, 90] it can also be used to show the contribution of the dentino-enamel junction to outer enamel surface complexity[91] and shows promise as an additional measure of dental phenomics.

image

Figure 4. Topography maps (top) and orientation maps (bottom) of permanent lower right first molars from two different individuals. Each model is illustrated with orientation patches when first orientation is at 0º, with the OPCR score at bottom right, demonstrating the differing complexity of the occlusal surfaces between the individuals. Scale bars = 1 mm; height in mm.

Download figure to PowerPoint

Tooth structure: enamel

2D digital imaging

Enamel defects present a range of clinical appearances, so accurate description and recording are critical for diagnostic purposes as well as aetiological studies. Various indices have been developed in the past, but each has its limitations.[92] The Enamel Defects Index (EDI) was developed based on three core components, i.e. hypoplasia, opacity and post-eruptive breakdown,[93] with each scored on a binary yes/no basis. The index can be expanded in line with the specific aims of a study, while still allowing for inter-study comparisons, as the same core components are used for data collection and analysis. It is also easy to apply and has shown good validity and reproducibility.[93, 94] In one study, altered enamel characteristics (i.e. demarcated opacities) and surface topography were assessed to clarify the role of specific genes in amelogenesis. It was noted that Amelx and Enam gene mutations in Amelogenesis Imperfecta in experimental mice were associated with altered enamel structure and mineralization.[21] These subtle changes have been associated with an increased risk of post-eruptive breakdown,[94] demonstrating how visual changes of the enamel structure are ultimately linked to structural integrity. This information is critical for clinicians to establish caries risk and preventive care for affected patients. In addition, alterations in mandibular morphology have also been attributed to Amelx gene mutations in mice, suggesting amelogenin's role as a multifunctional protein in the development of craniofacial features.[21] This association warrants further clarification in humans and, if present, orthodontic assessment may be required as part of the overall management plan. Readers are referred to the paper by Seow[95] in this special issue for a more in-depth discussion of developmental defects of enamel and dentine.

Figure 5 provides an example on how 2D imaging helps to quantify the defects in enamel (from a study model), while Fig. 6 demonstrates how polarized images clearly demarcate enamel opacities for accurate image analysis.

image

Figure 5. Digital 2D image of hypoplastic lesions on an upper incisor (from a study model) compiled with image analysis of the surface defects.

Download figure to PowerPoint

image

Figure 6. Demonstration of 2D imaging of opacities under standard illumination (a), polarized illumination (b) and measurement area output from polarized digital images (c).

Download figure to PowerPoint

Facial morphology

The study of facial morphology is particularly important in the fields of orthodontics[96] and craniofacial growth and development.[97] Traditional techniques to discern morphological features of interest involved comparing hand tracings from radiographs[96] or 2D photographs,[97] or combining 2D images to form a 3D representation.[63, 98] While these methods have provided the building blocks to understand variations in facial morphology, they carry a risk of operator error, along with limited opportunity for analysis. Hence, 3D assessment of facial morphology represents another exciting field of research.

3D imaging

Initial images may be obtained via CT imaging, MRI scans,[99] laser scanning,[100, 101] or photogrammetric devices. Photogrammetric devices provide a non-invasive and faster alternative when compared to the other imaging techniques and thus they are the favoured option for image acquisition.[102] During the image acquisition process, several images are obtained and these need to be merged to form the final facial image. Hence, appropriate software is utilized to generate tens of thousands of corresponding points based on certain predetermined landmarks, e.g. the eyes and the corners of the mouth (a technique known as the dense surface model).[103] The generated points are subsequently used to align and merge the obtained scans, resulting in a final image that accurately replicates the facial contour and topography.

The use of 3D imaging techniques has aided in clarifying the roles of key genes in the development of facial features. In 2012, collaborative efforts by the International Visible Trait Genetics (VisiGen) Consortium identified five candidate genes (PRDM16, PAX3, TP63, C5orf50 and COL17A) that influence facial phenotypes in individuals of European ancestry via the comparison of facial morphological landmarks.[100] Among the five mentioned genes, PAX 3 showed the most correlation statistically to influence facial morphology. In another study, the effects of PAX3 variants to nasion prominence and position was demonstrated, further detailing the role this gene plays in facial development.[99]

In cases of facial dysmorphism, the final image is compared to ‘references’ to quantify the noted variations[104] and this may assist in clarifying the impact of underlying genetic influences.[105] For example, subtle changes in facial and nasal shape were noted in patients with Wolf–Hirschhorn syndrome. This information clarified how the mutation resulted in varying clinical presentations of patients.[105] In addition, facial morphometric analysis allows for the assessment of disease progression[106] and is a non-invasive method to assess the efficacy of therapeutic intervention by comparing facial dimensions before and after therapy.[107] Future research will involve the exploration of normal facial variation within different age, gender and ethnic groups. The use of this technology for disease diagnosis, monitoring, pre-surgical assessment and review of medical intervention (postoperative surgical and/or drug therapy) demonstrates its potential for a wide range of applications.[108-110]

Cleft lip and/or palate

Ultrasonography

Another important field of phenomic research has been the study of cleft lip and/or palate (CL/P). In recent years, phenotyping of CL/P has expanded understanding of the clinical spectrum of the deformity, with microform clefts being the mildest form. Subtle clinical presentations such as a small lip or alveolar arch defect can be easily missed,[111] and careful characterization of defects has provided insight into varying clinical presentations. Clinical examination alone fails to detect subepithelial changes in ‘non-visible’ cases of cleft lip. Ultrasonography has been used to demonstrate defects in the orbicularis oris muscle in non CL/P relatives of individuals with obvious CL/P,[112] highlighting links between certain genetic mutations with non-visible (submucosal) changes in microforms of cleft lip.[113] These findings add to the known aetiological factors of CL/P,[114, 115] while also improving clinical detection of the more subtle forms. In future, the use of ultrasonography for ‘at-risk’ patients, i.e. relatives of patients with obvious CL/P may help to detect minor forms of CL/P. This will reduce the risk of under-diagnosis and help to ascertain the risk of familial inheritance.

Computer Modelling of Genotype-Phenotype Correlations

  1. Top of page
  2. Abstract
  3. Introduction
  4. Approaches to Phenotyping
  5. Phenotyping and Oral Diseases
  6. Phenotyping and Craniofacial Development
  7. Computer Modelling of Genotype-Phenotype Correlations
  8. Future Directions
  9. Conclusions
  10. Acknowledgements
  11. Disclosure Statement
  12. References

Once the links between variable phenotypes and their respective contributing factors are established, a cohesive genotype to phenotype (cG-P) model can be constructed.[6] An accurate model serves as a predictive tool for phenotypic changes when influencing parameters are altered and this information is useful in both education and future research.

Examples of dental cG-P models have been described.[116-119] Kangas et al.[116] and Kassai et al.[118] described how tooth cusp patterns were altered by manipulating a targeted signalling protein. Two papers by Salazar-Ciudad and Jervall[117, 119] provide in-depth discussions of the parameters used to formulate their models. The first model[117] was termed the ‘morphodynamic model’ as it was able to reproduce the morphology of all intermediate stages of tooth development, while also considering the variations in different species. However, this model considered the effects of genetic input alone as the causal factor for the final observed trait, ignoring the role of other factors such as epigenetic and environmental influences. Later, the authors built on their earlier work to include cellular parameters in their second model to more closely resemble the effects of varying factors on tooth morphology.[119] In a recent paper, the model has been used to explore the adaptive nature of teeth to evolutionary challenges.[120]

To date, generated cG-P models are only able to predict the morphology of the dentino-enamel junction, but not the outer enamel surface or root anatomy. As they do not simulate mineralization, this prevents the investigation of the shape of mineralized structures such as enamel. Despite these limitations, the present developments provide the building blocks for more sophisticated models in the future. The constructed models should improve the predictive accuracy, ease of use and enable the complete visualization of tooth structure in relation to altered genetic, epigenetic and environmental influences.

Future Directions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Approaches to Phenotyping
  5. Phenotyping and Oral Diseases
  6. Phenotyping and Craniofacial Development
  7. Computer Modelling of Genotype-Phenotype Correlations
  8. Future Directions
  9. Conclusions
  10. Acknowledgements
  11. Disclosure Statement
  12. References

Advances in research methodology and data acquisition bring analytical challenges. Currently, most phenotypes are described by either using quantitative (i.e. measurement) or categorical (i.e. the presence or absence of a particular trait) data. However, in some cases, the answer may lie between two points or categories, leading to the loss of useful information due to the continuous nature of most phenotypes. The application of state-of-the-art analytical techniques utilized in other disciplines, including evolutionary biology[121, 122] and materials science,[123-134] may provide valuable approaches to further characterize the properties of bioceramic materials, such as enamel and dentine, from the atomic to macro-scale. For example, a combined assessment of structural analysis, mechanical properties and chemical composition of teeth will unravel the effects of developmental influences on the structure and function of teeth.[135-140] These techniques are summarized in Table 3.

Table 3. Examples of techniques for detailed characterization of structural analysis, mechanical testing and chemical composition of teeth in dental phenomic studies. Non-destructive techniques that allow longitudinal changes in the properties of teeth to be assessed, in response to external factors (such as tooth demineralization), have advantages over analytical techniques that alter the tooth surface as a result of surface coating and extensive sample preparation. The techniques outlined here are limited to in vitro applications at present, but future technological advancement could enable in vivo applications
TechniqueApplicationsSurface/sub- surface assessmentLongitudinal assessmentSample preparationKey limitations
Structural analysis
Micro computed tomography (micro-CT)

3D quantitative and qualitative assessment of surface topography and sub-surface features (e.g. DEJ and pulp chamber) by using 3D reconstructions

(Resolution = 5–50 μm)

Surface +

sub-surface

+NoneProlonged scanning and data processing
Confocal Microscope (CM)3D qualitative and quantitative assessment of surface topography on 3D reconstructions; more efficient than SEM and atomic force microscopy (Resolution = 1–2 μm)Surface+NoneUnable to analyse specimens in hydrated conditions (e.g. in water storage)
Scanning Electron Microscopy (SEM)2D qualitative assessment of surface topography; can provide limited information on chemical composition (Resolution = 2–4 nm)Surface

-

(no polishing or sectioning)

Carbon/gold coatingUnable to conduct longitudinal assessment on the original surface without polishing off carbon/gold coating
Environmental Scanning Electron Microscopy (ESEM)

2D quantitative assessment of surface topography; can provide data on chemical composition

(Resolution = 1–2 μm)

Surface+NoneReduced resolution compared with SEM
Mechanical properties
Nanohardness testingQuantitative assessment of surface hardness/porosity; provides data on elastic modulus and deformation behaviour (Indentation tip radii range from microns to nanometers)Surface + shallow sub-surface+Highly polished, flat samples 
Nanoscratch testing

Quantitative assessment of tribological (wear) properties; provides data on coefficient of friction and surface roughness

(Indentation tip radii range from microns to nanometers)

Surface + shallow subsurface+Highly polished, flat samplesScratches cannot be superimposed before and after treatment
Chemical composition
Energy Dispersive X-ray Spectroscopy (EDXS)[3]Quantitative and qualitative surface elemental analysis (with atomic number >3), usually in combination with SEM (detection limit of 0.1-0.5%wt) or transmission electron microscopy (detection limit of 0.01-0.1%wt)Surface-

Carbon/gold coating

No additional preparation for qualitative analysis; flat and polished for quantitative analysis

Lacks sensitivity for trace element analysis
Time of flight Secondary Ion Mass Spectrometry (Tof-SIMS)

Semi-quantitative chemical characterization of the tooth surface in the form of organic and inorganic mass spectral data

(Has high sensitivity in parts per billion for entire elements in the periodic table from 1–2 nm of the surface)

Surface+Highly polished, flat samplesHighly time consuming and prone to contamination (i.e. requires ultra-clean environment for sample processing)
X-ray Photoelectron Spectroscopy (XPS)

Quantitative chemical characterization of the tooth surface (as atomic percentages) using both high- and low-resolution spectra

(Spatial resolution <10 μm)

Surface+Highly polished, flat samplesHighly time consuming and prone to contamination (i.e. requires an ultra-clean environment for sample processing)

The development of more sophisticated analytical methods will be critical to further develop accurate phenotype quantification. At present, many of the techniques described are limited to in vitro use, with the exception of 2D imaging for the assessment of tooth colour,[56] enamel defects,[57] gingival health[141] and plaque scores.[142] Efforts to make other in vitro techniques applicable to the clinical setting are ongoing and represent a key area for both researchers and clinicians. In future, one aim is to develop analytical techniques to accurately assess variations of tooth morphology and surface characteristics in the dental chair. This will improve the speed of data acquisition and enable clinicians to quickly characterize variations or defects present, and provide appropriate advice and treatment for affected patients.

Dental phenomics presents many exciting opportunities. The use of twin models helps determine the cause of variations seen in dentofacial structures and oral health.[143] Cross species data-sharing using animal models continues to unravel the impact of mutations in the human dentition.[21, 144] Collaboration between various centres of excellence will be important to advance the field. The International Collaborating Network in Oro-facial Genetics and Development aims to promote international collaboration[145] by the sharing of expertise and data. Fragmented scientific research in the past, using independent samples and non-standardized analytical techniques, represents a missed opportunity to enable inter-study comparison. While widespread integrative studies will require much effort, the potential gains for doing so will greatly benefit understanding of human biology. Clinicians can also play a crucial role in gathering accurate phenotypic data for analysis, thus actively participating in ongoing research projects in their respective fields of interest.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Approaches to Phenotyping
  5. Phenotyping and Oral Diseases
  6. Phenotyping and Craniofacial Development
  7. Computer Modelling of Genotype-Phenotype Correlations
  8. Future Directions
  9. Conclusions
  10. Acknowledgements
  11. Disclosure Statement
  12. References

The craniofacial structures present important biomarkers for certain diseases and genetic mutations, providing many clues regarding growth and development. Research findings inform the profession of both normal and abnormal variations, while also directing treatment provision. In addition, the significance of accurate phenotyping methods to dental anthropology and forensic odontology has been established.

New imaging techniques allow for the preservation of studied specimens by reducing wear and tear due to manual handling. The measurements obtained also produce comparable or better reliability when compared to manual methods. Furthermore, additional parameters such as curved surfaces, surface topography and colour differences can now be assessed accurately. These new parameters clarify variations seen in the craniofacial complex and are relevant to both the clinical and research environment. In clinical practice, the detection of altered dental tissues and submucosal changes aid recognition of disease and disorders to direct future treatment plans.

Dental phenomics helps to increase the yield obtained from aetiological studies and bridges the gap between the research environment and practical application. Future development in the field of dental phenomics will provide many exciting opportunities to unravel the mysteries behind variations in both health and disease.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Approaches to Phenotyping
  5. Phenotyping and Oral Diseases
  6. Phenotyping and Craniofacial Development
  7. Computer Modelling of Genotype-Phenotype Correlations
  8. Future Directions
  9. Conclusions
  10. Acknowledgements
  11. Disclosure Statement
  12. References

We wish to acknowledge the National Health and Medical Research Council of Australia and the Australian Dental Research Foundation for supporting various ongoing research projects. We also thank Dr Atika Ashar from Universiti Kebangsaan Malaysia for the provision of Fig. 2 and advice on 3D analytical software.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Approaches to Phenotyping
  5. Phenotyping and Oral Diseases
  6. Phenotyping and Craniofacial Development
  7. Computer Modelling of Genotype-Phenotype Correlations
  8. Future Directions
  9. Conclusions
  10. Acknowledgements
  11. Disclosure Statement
  12. References