| Sign In to gain access to subscriptions and/or personal tools. |
The Growth Hormone Receptor Gene is Associated with Mandibular Height in a Chinese Population
1 Department of Orthodontics Faculty of Stomatology Capital, University of Medical Sciences, Beijing, 100050 P.R. China Correspondence: * corresponding authors: bangkangwang{at}hotmail.comsheny{at}ms.imicams.ac.cn
Genetic influences are important in the determination of mandibular morphology, and growth hormone receptor (GHR) is believed to have an important influence on the growth of craniofacial bone. In this study, we used quantitative trait locus methods to evaluate the relationship between craniofacial morphology and single-nucleotide polymorphisms (SNPs) in GHR in an unselected healthy Chinese population. We systematically screened the 10 exons and nearby introns of GHR and identified 6 SNPs. Using 4 SNPs as markers, we studied the relationships between genotypes and craniofacial linear measurements. Individuals with the genotype CC of polymorphism I526L had a significantly greater mandibular ramus length (condylion-gonion/ articulare-gonion) than those with genotype AC or AA. Haplotype analysis showed that there were also significant differences between the long and short mandibular height groups in an extreme population. Our results indicate that the GHR gene polymorphism I526L is associated with mandibular height in the Chinese population.
Key Words: mandibular height QTL analysis single-nucleotide polymorphism growth hormone receptor gene.
Because multiple genes and environmental influences are involved in the growth of the mandible, it is difficult to determine the relationship between mandibular shape and a particular gene. At the cellular level, transcription factors can switch genes on and off by activating or repressing gene expression. This control is regulated through two main groups of regulatory proteins, the growth factor family and the steroid/thyroid/retinoic acid superfamily (Evans, 1988). Among these, it is recognized that the GH/GHR/IGF-1 system plays an important role in skeletal growth and development (Sjogren et al., 2000). The growth hormone receptor gene (GHR) belongs to the superfamily of cytokine receptors. GHR is located on chromosome 5p13.1-p12 and is 87 Kb long, with 10 exons encoding 620 amino acids (Leung et al., 1987). Immunohistochemical localization has shown that GHR is present in the mandibular condyle (Lewinson et al., 1994). GHR mutations can cause many diseases, such as idiopathic short stature and Laron syndrome, in which most patients exhibit smaller linear measurements of the facial structures, facial retrognathia, and a steep vertical inclination of the mandible (Kjellberg et al., 2000; Russell, 2001). We initially selected growth hormone 1 (GH1), growth hormone 2 (GH2), GHR, insulin-like growth factor I (IGF-1), and insulin-like growth factor receptor 1 (IGF-1R) etc., as candidate genes. From the single-nucleotide polymorphism (SNP) database, we found that, in the exon area, GH1 has 4 SNPs, and that GH2, IGF-1, and IGF-1R each has a single SNP. The heterozygosity of the GH1, GH2, and IGF-1 SNP is lower than 10%. While the heterozygosity of the IGF-1R SNP is higher than 10%, it is synonymous with SNP. Because of the lack of variation in the final coding sequence, GH1, GH2, IGF-1, and IGF-1R genes are not valuable for quantitative trait locus analysis (QTL). GHR has 4 non-synonymous SNPs, however, and the heterozygosity of most of them is higher than 10%, which suggests a probable functionally relevant candidate gene. Spatial patterning of genes is important in determining mandibular morphology (Lingenberg et al., 2001), and quantitative analysis of the effects of individual genes on mandibular shape should advance our understanding of gene action in the development of craniofacial form. QTL analysis has been successfully applied to various measurements of dimension in mouse mandibles and skulls (Cheverud et al., 1996; Leamy et al., 2000). An association study of gene polymorphisms in the GH/GHR/IGF-1 system would be useful in understanding genetic influences on craniofacial morphological determinants on a molecular level. Here, we investigate polymorphisms in GHR, examining the relationship between GHR SNPs and mandibular morphology.
Subjects This study presents data collected from two major populations. The subjects in the first population (referred to hereafter as the unselected population) were 95 Chinese of Han ethnicity (age range, 20–35 yrs; average age, 28 ± 3.6 yrs). All subjects were volunteers recruited from among university students. The second population (referred to hereafter as the extreme population) consisted of 50 Chinese Han patients from Beijing Stomatology Hospital. Of these, 22 exhibited short mandibular ramus height (more than 2 SD shorter than average), and 28 exhibited long mandibular ramus height (more than 2 SD longer than average) (Fu and Tian, 1992). Both sides of all faces were symmetrical. All individuals were unrelated, healthy, and lacked congenital disorders or craniofacial trauma; none had received orthodontic or orthopedic therapy. Human subjects participated in a protocol that was reviewed and approved by the Ethics Review Board of the Capital University of Medical Sciences, and informed consent was obtained from all participating subjects. All samples were collected from peripheral blood, with EDTA as the anticoagulant.
Genomic DNA Extraction
PCR Amplification and Sequence
Craniofacial Measurements Measurements were made on tracings of lateral cephalograms by a standard protocol (Fig. 2 d2/2n) (Dahlberg, 1940), and the errors ranged from 0.2 to 0.5 mm for the linear measurements.
The Kolmogorov-Smirnov test was performed for every group, and the means, medians, SD, skewness, and kurtosis of ANB angle and height were calculated. The difference in each measurement among SNP genotypes was tested for significance by an independent-sample t test or one-way analysis of variance (ANOVA), with SPSS 10.0 for Windows. The tests for Hardy-Weinberg equilibrium, allele and genotype frequencies, and chi-square were all performed with SPSS 10.0 for Windows.
Haplotype Analysis and Linkage Disequilibrium
Kolmogorov-Smirnov test analysis of body height and the ANB angle of the 95 subjects showed that the distribution pattern of measurements from the unselected population was reasonably close to normal (data not shown). Therefore, the unselected subjects represent a normally distributed population for craniofacial morphology and height in the Chinese Han population. Sequence analysis showed that 6 SNPs were identifiable in the GHR gene in this population. G168G was located within the exon 6 extracellular domains, and another 5 SNPs were found in the exon 10 intracellular domain-coding region in a 500-bp nucleic acid interval from 30281 to 30700 (GenBank: NT_006702). C422F, P477T, I526L, and P561T are synonymous cSNPs; S473S is a non-synonymous cSNP. I526L exhibited the highest heterozygosity of the SNPs (47.6%), and the SNPs C422F, P477T, and P561T resulted in polarity changes. The allele distribution of SNPs was in Hardy-Weinberg equilibrium (data not shown).
We selected 4 cSNPs (G168G, C422F, I526L, and P561T), each with heterozygosity higher than 10%, as markers. One-way ANOVA showed that 3 craniofacial measurements—Co-Go, Ar-Go, and S-Go—are correlated with the I526L variant (Table 1
The linkage disequilibrium (LD) measure D' was used to quantify the pairwise LD level. Linkage disequilibrium analysis showed that 3 SNPs—C422F, I526L, and P561T—are in complete LD (see online Appendix). There is low haplotype diversity in this region, in which only 6 of 16 haplotypes were identified; these results are consistent with the high level of linkage disequilibrium. The results also showed a significant difference in haplotype distribution between these two groups in the extreme population (p = 0.0048) (Table 2
Previous studies have suggested that the effects of genes on the mandible should be spatially patterned. Signaling interactions coordinate the outgrowth of the facial primordium. Many genes are regarded as master genes for head and face development, controlling pattern, induction, and epithelial-mesenchymal interactions during development of the craniofacial complex (Klingenberg et al., 2001). Malocclusion should be regarded not as abnormal or as a disease, but as a variation of occlusion in a continuous, multi-factorial trait (Mossey, 1999). Most genetic studies of shape characterize in terms of the relative sizes of parts and use a set of linear distances for measurement (Klingenberg et al., 2001). In this study, we used QTL methods to investigate the relationship between mandibular shape and the GHR gene. We found that GHR I526L was correlated with mandibular ramus length in the unselected population; individuals with the genotype CC have a longer ramus than those with the genotype AC or AA. To strengthen the finding further, we used haplotype analysis in another extreme population (long/short mandibular height patients). The result above was replicated in the extreme population (p = 0.0048), which is a very strong indication of its validitiy. I526L resulted in a substitution of leucine for isoleucine (CTC to ATC) in the intracellular domain of GHR. GHR has 2 cytoplasmic regions that are important for signal transduction: One is the box1 region, and the other is the 184-amino-acid region at the C-terminus (Billestrup et al., 1995). Three of the 4 cSNP markers in our study—P477T, P561T, and I526L—are in the C-terminal 184-amino-acid region. Because this area is associated with signal transduction in the intracellular domain of GHR, these polymorphisms might affect signal transduction, resulting in a change in IGF-1 expression. GHR plays an important role in cartilage growth (Visnapuu et al., 2001), which directly affects other features of growth and development, such as body height and profile. We suspect that these changes could influence the mandibular condylar growth and, thus, mandibular ramus length, resulting in individual differences. Perhaps this scenario can partly explain the phenomenon of differing morphological characteristics among different ethnicities.
Of course, there are numerous ways in which the I526L variant could affect GHR activity. In addition to potentially causing direct changes in protein function, the variant could affect regulation of GHR, and in either case the variant could act singly or in combination with other SNPs. The effect of this SNP on GHR function and downstream gene expression should be clarified by further study. In the search for candidate genes involved in maxillary or mandibular dysmorphogenesis, polymorphisms in the candidate genes and the genes of the molecules they regulate are prime targets. The haplotype analysis from this study is consistent with the pairwise LD analysis, indicating that I526L is at least a QTL marker and may also be acting as a causative locus. Table 2
Studies have identified a correlation between the P56IT of GHR and mandibular ramus length in a Japanese population; the researchers used a DNA digestion method to study the single SNP site, P56IT, without LD or haplotype analysis (Yamaguchi et al., 2001). We did not replicate this result, but found that SNP I526L is associated with mandibular ramus length in a Chinese population. The findings from both our and the Japanese study do show a relationship between mandibular ramus length and GHR. The fact that we did not identify an association with the marker P561T could reflect either a lack of power in our experiments or an actual difference between populations. The former possibility is considerably more likely, especially because all 3 markers in strong LD with I526L were also less informative than I526L (see Table 2 The phenotype is not associated with a Class III malocclusion; the mandibular morphological contribution to a Class III malocclusion remains unclear (Singh, 1999). Conventional bivariate tests indicated that the mandibular body lengths (Go-Gn and Go-Me) were longer in the patients with Class III malocclusion, compared with the "normal" patients (Singh et al., 1998). In this study, we did not find any SNPs that have a relationship with mandibular corpus length (Go-Gn and Go-Pog) or overall mandibular length (Co-Gn, Ar-Gn, and Ar'-Pog'), and neither relationship was found in the Japanese population (Yamaguchi et al., 2001). From these data, perhaps GHR can affect only the longitudinal development of the mandible, i.e., mandibular ramus length. In conclusion, we have discovered that Chinese Han individuals with genotype CC of polymorphism I526L have a greater mandibular ramus length (Co-Go/Ar-Go) than those with genotype AC or AA. This finding reinforces the idea that variations in GHR may be associated with differences in mandibular morphology. The haplotype study also supports the idea that GHR is associated with mandibular ramus length. GHR might be a candidate gene for mandibular morphology, at least in the Chinese Han population.
We thank all the patients for their participation in this study. We also express our gratitude to Professor Songling Wang (CUMS); Drs. Lu Lu, Jeremy Peirce, and George Albert Cook (Department of Medicine) and Dr. James Vaden (Department of Orthodontics), University of Tennessee, Memphis, USA, for their helpful suggestions for and revisions of the manuscript. This study was supported by the Beijing Health Bureau (Grant 199904).
A supplemental appendix to this article is published electronically only at http://www.dentalresearch.org. Received for publication November 5, 2003. Revision received July 13, 2005. Accepted for publication July 14, 2005.
Journal of Dental Research, Vol. 84, No. 11,
1052-1056 (2005) This article has been cited by other articles:
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

d2/2n) (


