Investigation of a genetic association of a class III skeletal pattern

From a morphological perspective, a Class III skeletal pattern can be associated with maxillary deficiency, mandibular prognathism (MP) or a combination of both conditions.1,2 The roles and contribution of the maxilla and mandible in the development of a Class III skeletal pattern have been reported and many studies have observed maxillary retrognathism (MR) as frequently as MP.2–5 Genetics is a major determining factor in the aetiology of a Class III skeletal pattern, which invariably has a multifactorial aetiology.6 Genetic factors can cause over-development of the mandible or under-development of the maxilla and so may be involved in all combinations of a Class III skeletal pattern.1

Current management options range from orthopaedic treatment at an early developmental age to orthognathic surgery performed after dento-facial growth and development have been completed. In the future, the identification of candidate genes will enable the determination of individuals with a relatively high likelihood of developing a Class III skeletal pattern. This approach will allow early diagnosis, interceptive treatment and interventions targeting dento-facial anomalies.7

The collagen type II alpha 1 chain (COL2A1) gene is located in the long arm of chromosome 12 (12q13.11) and is encoded by 54 exons. The COL2A1 gene exons, including genomic DNA, are responsible for generating the pro-alpha₁ (II) chain in the type II collagen helix.8,9COL2A1 is a large gene of about 41 kb which covers almost six exons of genomic DNA, but only the rs1793953 region has been associated with a Class III malocclusion.9,10 For this reason, only single nucleotide polymorphisms (SNPs) belonging to this region of the COL2A1 gene were examined in the present study. In a study of a Chinese population, Xue et al.9 found marked differences in genotype distribution and allele frequency for SNP rs1793953 in the COL2A1 gene between case and control groups, suggesting that the COL2A1 gene could be a new susceptibility gene and that SNP rs1793953 in COL2A1 is a genetic risk factor for MP. To date, no prior study has investigated the genetic association of the COL2A1 gene associated with a Class III malocclusion in a Turkish population.

The human growth hormone receptor (GHR) gene (87 kb) is located on 5p13.1-p12 of chromosome 5 and carries 10 exons. The GHR gene primarily encodes the human growth hormone receptor, which is a transmembrane protein.11 A potential ethnic difference has been found in the association of GHR with mandibular ramus height.9,12–14 Zhou et al.12 compared GHR polymorphism I526L with mandibular ramus height in a Chinese population. Subsequent studies found a positive correlation between polymorphisms P561T and C422F and mandibular ramus height in a Japanese13 and a Korean population.14 In addition, Bayram et al.15 noticed that patients with the genotype CA of polymorphism P561T had greater effective mandibular height and a lower face height than those with genotype CC in a Class III Turkish population. However, these data did not include patients with a Class III maxillary deficiency. Genetic loading analyses can help determine which genes are responsible for a skeletal Class III malocclusion. Two SNPs in the coding region of GHR, including C422F (dbSNP number rs6182) and P561T (dbSNP number rs6184), are of high relevance in the aetiology of a skeletal Class III and so it was planned to examine these regions of the GHR gene. The dbSNP numbers were identified in the National Center for Biotechnology Information database (http://www.ncbi.nlm.nih.gov/SNP/). It was hypothesised that the COL2A1 and GHR genes were associated with specific growth patterns of either MR or MP. Accordingly, this hypothesis was tested by the investigation of three SNPs near or within the COL2A1 and GHR genes associated with specific growth patterns of either MR or MP in a Class III Turkish population.

The present study was approved by the Ethics Committee of Ege University, School of Medicine (16-6/17). Based on the inclusion criteria, two hundred and fifty-five patients (85 with MP, 85 with MR and 85 control subjects) were selected from individuals who presented to the Orthodontic Department of Aydın Adnan Menderes University, Faculty of Dentistry between September 2017 and May 2018. All patients and legal guardians provided written informed consent before a lateral cephalometric radiograph and a blood sample were taken. A CONSORT diagram showing the patient flow throughout the study is shown in Figure 1. No patients nor samples were lost during the study.

Figure 1.

The skeletal malocclusions were identified using the SNA, SNB and ANB angles, plus the Wits value. Patients with an ANB angle <0° and a Wits value of <0 mm were considered to have a Class III skeletal pattern. The study group included patients with a Class III skeletal pattern due to two distinct aetiologies. The MR group (n = 54 female; n = 31 male; ages, 16–56 years; average age, 19.4 years) included patients with a Class III skeletal pattern due to maxillary deficiency (SNA <79°; SNB = 79.92 ± 3.44°) and an MP group (n = 49 female; n = 36 male; ages, 16–50 years; average age, 18.2 years) included patients with a Class III skeletal pattern due to MP (SNA = 82.57 ± 3.55°; SNB >83°). A control group (n = 52 female; n = 33 male; ages, 16–49 years; average age, 17.7 years) included patients with a skeletal Class I pattern (2°< ANB <4°; 0 < Wits value <2 mm). The inclusion criteria were patients of age ≥16 years, with no previous orthodontic or interceptive treatment, no systemic disorder, no positive family history of inherited disease, no congenital anomaly, such as cleft palate or cleft lip, nor an endocrine disturbance. The demographic characteristics of the groups are presented in Table I.

Cephalometric analyses were performed for the 255 patients by a single researcher (B.T) using Dolphin Imaging 11.9 software (Dolphin Imaging and Management Solutions, Chatsworth, CA, USA). The linear cephalometric measurements used are shown in Figure 2. Thirty radiographs were randomly selected and cephalometric analyses were repeated 2 weeks later to determine intra-operator tracing errors. A paired t test was applied to the first and second measurements and no error was found (t values, 0.085–1.740).

Figure 2.

The SNPs were selected using the NCBI dbSNP databases and chosen on the basis of their minor allele frequency >10%. The characteristics of the studied SNPs are shown in Table II. The blood samples were drawn and stored at −80°C. When required for the genetic assays, the frozen samples were simultaneously thawed and assayed by DNA isolation, followed by PCR and pyro-sequencing conducted at the Molecular Biology Laboratory in the Pathology Department of Aydın Adnan Menderes University School of Medicine. Genomic DNA was extracted using the QIAamp DNA Blood Mini Kit (Qiagen GmbH, Hilden, Germany). For each DNA sample, three separate PCR reactions were performed using a SensoQuest LabCycler. The mixture of PCR reaction products was combined with the sequencing primers in each well of the Q24 plate (specific to the device) for pyro-sequencing. Sequences belonging to rs6182 and rs6184 loci of the GHR gene and the rs1793953 locus of the COL2A1 gene were analysed using PyroMark Q24, v.2.0.8 software (Qiagen, Valencia, CA, USA).

Table III presents the results of the lateral cephalometric radiographs of patients in the three groups. A difference between the groups was anticipated because the skeletal measurements were used to define the groups.

All SNPs were consistent with HWE. Table IV shows the allele and genotype distribution of 3 SNPs. In the GHR gene SNP rs6182, no individual carrying the TT genotype was found and in the GHR gene SNP rs6184, no individual carrying the AA genotype was found in any group. The genotype and allele frequency distribution of the GHR gene SNP rs6182, GHR gene SNP rs6184 and COL2A1 gene SNP rs1793963 did not significantly differ between the MP and the control group and between the MR and the control group (p > 0.05).

Genetic models for the COL2A1 gene SNP rs1793963 as dominant (major homozygous vs. heterozygous + minor homozygous) and recessive (major homozygous + heterozygous vs. minor homozygous) were applied. In the COL2A1 gene SNP rs1793963, no significant differences in the dominant model (C/C genotype vs. C/T genotype + T/T genotype) were found between the MP and the control group (p = 0.113; OR, 1.65; 95% CI, 0.88–3.09) and the MR and the control group (p = 0.155; OR, 1.56; 95% CI, 0.84−2.92). The recessive model (C/C genotype + C/T genotype vs. T/T genotype) of the COL2A1 gene SNP rs1793963 did not show any statistically significant association between the MP and the control group (p = 0.665; OR, 0.75; 95% CI, 0.32−1.77) and the MR and the control group (p = 0.841; OR, 1.17; 95% CI, 0.53−2.59) (Table IV).

Associations were detected when specific cephalometric measurements were assessed. In the statistical comparison between genotypes in SNP rs6182 using cephalometric measurements, the GT genotype correlated with ramus height (Ar–Go) in the total sample (p < 0.001). The individuals with the GT genotype had a 6.3 mm longer ramus height when compared with the GG genotype (Table V). For SNP rs6184 in the total sample, the AC genotype correlated with ramus height (Ar–Go) (p < 0.01). The individuals with the CC genotype had a 6.49 mm longer ramus height when compared with the AC genotype (Table V). No significant difference was detected between the genotypes of SNP rs1793653 in the total sample using the cephalometric measurements (p > 0.05; Table VI).

Of the aetiological factors in determining a Class III skeletal pattern, genetics plays a major role and genetic inheritance may guide the determination of the treatment of choice.16,17 The COL2A1 SNP rs1793953 and GHR SNPs rs6182 and s6184 genes were evaluated in the present study because of their suspected involvement in condylar cartilage and craniofacial development.9,18

The candidate genes selected for the present study were separately investigated in MR and MP cases as common features associated with a Class III skeletal pattern.2,4,19,20 In addition, there has been no prior genetic study investigating the COL2A1 gene or patients with a Class III skeletal pattern (either MR or MP) in a Turkish population.

In the MP group, a positive correlation was found between statural height and all measurements, but particularly with ramus height. Height was correlated with mandibular length measurements rather than maxillary length measurements. Unlike the long bones, the mandible has no epiphysial cartilage; however, the growth pattern and morphological characteristics are comparable. In addition, the growth process is not dependent on sutures, unlike in the maxilla. Thus, statural height seemed to be mainly correlated with mandibular measurements.21

No significant difference was found in the cephalometric comparisons between the genotypes of the COL2A1 gene SNP rs1793953. It has been suggested that the COL2A1 gene could be involved in the Class III skeletal pattern of individuals with MR and MP. Xue et al.9 indicated that the SNP rs1703953 of the gene was implied in MP. However, there was no significant difference in craniofacial aspects between genotypes related to this locus in the present study, possibly because of differences in ethnicity. However, distinct genes might be responsible for the aetiology of a Class III malocclusion in populations with different genetic backgrounds. Genetic factors vary among societies, and so it is important to investigate subjects within different populations. Another possible reason may be the lack of statistical power due to a relatively small study sample size leading to an inability to detect smaller differences between the groups. Studies with a larger sample size are recommended to confirm the current findings.

Tomoyasu et al.13 noted a correlation between ramus height and the GHR gene SNPs rs6182 and rs6184 in Japanese individuals. These findings are in agreement with the present study. In addition, there was a G allele frequency of 94.1% and a T allele frequency of 5.9% in Japanese subjects.13 In a Chinese population, the corresponding frequencies were 79.4% and 20.6%, respectively. It is reported that the T and A alleles have not been observed in American, European-American and Spanish subjects, while all the individuals had the G and C alleles.13 The G and T alleles identified in the present study occurred at 97.1% and 2.9% in group MP and 99.4% and 0.6% in group MR, respectively. When assessing the GHR gene SNP rs6184, the C and A alleles occurred at 94.7% and 5.2% in Japanese individuals and 80% and 20% in Chinese individuals. In the present study, the C and A alleles occurred at 97.6% and 2.4% in the MP group and 99.4% and 0.6% in the MR group. Based on these frequency data, Turkish and Japanese individuals had similar allele frequencies for SNPs rs6182 and rs6184.

Bayram et al.15 investigated the P561T and C422F variants of the GHR gene in Turkish individuals. The results revealed a correlation between the P561T variant and mandibular length, while no comparison was performed for C422F polymorphism because this variant was detected in only one individual. Although these outcomes are inconsistent with the present findings, both studies indicated that the GHR gene is important in the aetiology of a skeletal Class III malocclusion.

When the cephalometric characteristics of genotypes were compared, the result was significant for the GHR-related domains, indicating that the genotype is a determinant of a measurement’s significance. In the total sample, the individuals with the GT genotype of the GHR gene SNP rs6182 had, on average, a 6.3 mm longer ramus height relative to those with the GG genotype. In the statistical comparison, the GT genotype correlated with ramus height (Ar–Go; p < 0.001). This finding concurs with Tomoyasu et al.13 but is inconsistent with Zhou et al.12 The discrepancy might be explained by ethnic variation.

Within the total sample, ramus height was significantly increased in individuals who possessed the GHR gene SNP rs6184 (p < 0.01). In addition, individuals with the AC genotype had, on average, a 6.49 mm longer ramus height relative to those with the CC genotype. A study of 60 children, aged 3–13 years, revealed 2 of 33 individuals with a normal occlusion and 5 of 27 with mandibular protrusion were heterozygous for the P561T mutation in the GHR gene. This suggests that the mutation affects mandibular growth during early childhood.22 However, these individuals were still growing and developing. More samples and additional data on the relevant morphological and functional factors are recommended to help clarify the effects of the mutation on occlusion type. Of 96 Japanese individuals, the frequency of P561T heterozygosity was approximately 15%.23 In the present study, the GHR rs6184 (P561T) variant correlated with ramus height, in agreement with Yamaguchi et al.24 By contrast, Bayram et al.15 found that this variant was correlated with mandibular length. The findings of the present study suggested that the GHR gene may aid in the decision-making process in the treatment of a borderline skeletal Class III anomaly in the examined Turkish population.

The major limitation of the present study was the sample size because of the low frequency of the chosen genetic markers. In the current research of a population of Turkish individuals, an association study was conducted while investigating the genetic background of a skeletal Class III malocclusion to evaluate the status of genes previously identified as candidates in other ethnic groups. However, some researchers have indicated the importance of epigenetic and other mechanisms that change gene expression.25 These are considered important factors in dento-facial growth and facial phenotype and should be identified in the parameters that influence malocclusions.25 Linkage analyses conducted in future investigations will help clarify how genetic factors influence craniofacial structure at the molecular level. By examining different candidate genes and different domains of the genes evaluated in the present study, a possible genetic aetiology will likely become clearer.

Within the limitations of the present study, the GHR gene polymorphisms of the SNPs rs6182 and rs6184 were associated with ramus height in the MP group. In addition, the COL2A1 gene SNP rs1793953, evaluated for the first time in a Turkish population, was found to have no association with sub-types of a Class III malocclusion.

The authors declare that there is no conflict of interest.

This work was supported by Aydın Adnan Menderes Research Projects [DHF-15011].

All procedures performed in studies of human participants were in accordance with the ethical standards of the institutional and/or national research committee (the Ethics Committee of Ege University, Medicine School [16-6/17]) and with the 1924 Helsinki Declaration and its later amendments or comparable ethical standards. All patients and legal guardians gave written informed consent before participation.

The data underlying this article will be shared on reasonable request by the corresponding author.

The full protocol of this study can be accessed from the National Thesis Database of Turkey (tez.yok.gov.tr).