Association of LTF, ENAM, and AMELX polymorphisms with dental caries susceptibility: a meta-analysis

Background This meta-analysis evaluated the association of LTF, ENAM, and AMELX polymorphisms with dental caries susceptibility. Methods We searched the Scopus, PubMed/Medline, Web of Science, and Cochrane Library databases to retrieve articles published by October 2019. Review Manager 5.3 software was used to estimate the odds ratios (ORs) and 95% confidence intervals (CIs). The results of publication bias tests were retrieved by Comprehensive Meta-Analysis 2.0 software. Results A total of 150 relevant records were identified; out of which, 16 were entered into the analysis (4 studies assessed LTF, 11 ENAM, and 11 AMELX polymorphisms). Of all polymorphisms, there was a significant association only between ENAM rs3796704 polymorphism and dental caries susceptibility. Both ENAM rs3796704 and AMELX rs17878486 polymorphisms had a significant association with dental caries risk in the Caucasian ethnicity and the studies including caries-free control group. Conclusions The results of this meta-analysis showed that the G allele and the GG genotype of ENAM rs3796704 were associated with an increased risk of caries in the case group compared with the control group. But there was no association between LTF rs1126478, ENAM (rs1264848 and rs3796703), and AMELX (rs946252, rs17878486, and rs2106416) polymorphisms and dental caries susceptibility.


Introduction
Dental caries can significantly affect the general health and quality of life in the modern world [1]. Dental caries develops following demineralization of tooth structure and often results in pulpal and periapical inflammation, and subsequent pain, infection and tooth loss [1,2]. Some environmental conditions and population groups may contribute to higher incidence of dental caries [2]. Environmental risk factors may also affect dental caries development [3]. Although exposed to the same environmental conditions, some patients may be more sensitive or more resistant to dental caries than others; such differences may be due to genetic factors in dental caries etiopathogenesis [4]. New findings raised possibilities of presence of associations between genetic factors and dental caries [5]. The etiology of dental caries involves complex interactions between genetic and environmental factors. Three prevalent genes are reportedly involved in development of dental caries namely the lactotransferrin (LTF), enamelin (ENAM), and amelogenin X (AMELX). The rs1126478 polymorphism of LTF (a saliva protein gene) produces a shift from arginine to lysine at amino acid position 47 in the antimicrobial region, and presents transcriptional activation activity [6]. LTF can decrease the lipopolysaccharide-activated innate immune response, regulate the adaptive immune system [7], and play a significant role in physiological homoeostasis, which is in turn related to disease development [8]. ENAM (a member of P/Q-rich secretory calcium-binding phosphoprotein cluster genes) is located on chromosome 4q 13.3 [9]. The gene encodes the protein enamelin, which is the largest protein found in the enamel matrix and is involved in dental enamel mineralization and its structural organization [10]. AMELX is an essential gene that produces amelogenin as the main protein of dental enamel during the secretion stage of amelogenesis [11]. AMELX polymorphisms result in distinct alternations in enamel microstructure [12]. Therefore, these polymorphisms play a critical role in regulation of mineralization and enamel thickness [13]. The association between the mutations of LTF, ENAM, and AMELX genes and dental caries susceptibility has been shown in some studies [5,14,15]. Therefore, we aimed to assess the association of polymorphisms of these genes and the risk of dental caries in a metaanalysis of case-control studies and therefore evaluating only the genetic influences for dental caries.

Search strategy and study selection
One author systematically searched the PubMed/Medline, Web of Science, Cochrane Library, and Scopus databases to retrieve articles published by October 2019 without publication period, language, and patient's age restrictions. The search terms or keywords were ("lactotransferrin" or "lactoferrin" or "LTF" or "enamelin" or "ENAM" or "amelogenin X" or "AMELX") and ("dental caries" or "caries" or "decay") and ("gene" or "polymorphism" or "variant" or "genetic"). In addition, the references of the retrieved articles related to the topic including original and review articles were searched to make sure that no study was missed. After article retrieval, another author assessed the titles and abstracts of the articles related to the topic; subsequently, the full-texts of the articles that met our eligibility criteria were downloaded and screened. After screening, the exclusion reason was recorded for any study removed, and the disagreements between the authors were resolved by another author.

Eligibility criteria
The inclusion criteria were as follows: (I) studies including two independent groups (case group with caries or high caries and caries-free control group or with low/ very low caries) without age restriction, (II) studies with any defined Decayed, Missing, and Filled Teeth (DMFT) score for the two groups, (III) studies including one or more polymorphisms of LTF, ENAM, and AMELX genes with a minimum of two relevant studies for the analysis; for example, four studies assessed LTF rs1126478, ENAM rs1264848, ENAM rs3796704, ENAM rs3796703, AMELX rs946252, AMELX rs17878486, AMELX rs6639060, and AMELX rs2106416 polymorphisms; and (IV) patients and controls had to have no genetic diseases, chronic illnesses, or other disorders. We excluded irrelevant studies, studies without sufficient data for analysis, studies without a control group, studies including less than 20 individuals in each group, duplicate studies, animal studies, case reports, conference papers, reviews, and systematic reviews.

Data abstraction
Two authors independently abstracted the data of the studies analyzed in the meta-analysis. The data from each study, including first author, publication year, country of residence of the included individuals, ethnicity, mean age of individuals in the two groups, age group of individuals in each study, genotyping method, DMFT score of the two groups, and type of reported polymorphism (s) in each study, were extracted and analyzed.

Statistical analysis
Review Manager 5.3 (RevMan 5.3) software was applied to compute the odds ratios (ORs) and 95% confidence intervals (CIs). To estimate the significance of the pooled OR by the Z test, a p-value (two-sided) < 0.05 was considered significant. The I 2 statistic was used to estimate heterogeneity. A p < 0.1 or I 2 > 50% indicated a significant heterogeneity and we used the randomeffects model for such cases; if not, the fixed-effects model was used. The publication bias across the studies was assessed using the Egger's and Begg's tests. If p < 0.05 (two-sided) for both tests or one, there was a significant degree of publication bias. In order to evaluate the stability/consistency of the results, the sensitivity analysis with both "the removal of one study" and "cumulative analysis" was performed. The results of these tests were retrieved by Comprehensive Meta-Analysis 2.0 (CMA 2.0) software. All authors revised the extracted data and the analyses and the disagreement between them was resolved by a discussion.

Study selection
A total of 150 records were identified in the databases; after removing the duplicates and irrelevant records, 29 full-text articles were evaluated for eligibility (Fig. 1). Next, 13 articles were excluded with reasons: two studies were systematic reviews, two studies were reviews, three studies lacked sufficient data, two studies had no control group, three studies did not report any of the mentioned polymorphisms in this meta-analysis and did not have any known polymorphism either, and one study reported rs1126478 polymorphism with less than 20 individuals in each group. Finally, 16 studies were entered into the analysis.

Study characteristics
The characteristics of 16 studies included in this metaanalysis are presented in Table 1. The articles had been published from 2008 to 2019. Eleven studies had been conducted on Caucasians [12, 14-16, 18-23, 25], four studies had been conducted on Asians [5,13,17,24], and one on mixed [11] ethnicities. Eleven studies evaluated children [5, 11, 13-16, 18-20, 22, 24], three studies evaluated adults [17,21,23], and two studies evaluated both adults and children [12,25]. Four studies had assessed LTF rs1126478 in 1556 cases and 1106 controls [5,16,20,24]. Five studies had assessed ENAM rs1264848 in 934 cases and 600 controls [11,14,19,22,25]. Four studies had assessed ENAM rs3796704 in 574 cases and 533 controls [11,19,23,25]. Two studies had assessed ENAM rs3796703 in 585 cases and 567 controls [5,13]. Three studies had assessed AMELX rs946252 in 151 cases and 147 controls [11,12,19]. Four studies had assessed AMELX rs17878486 in 249 cases and 193 controls [11,15,17,19]. Two studies had assessed AMELX rs6639060 in 157 cases and 144 controls [13,21], and two studies had assessed AMELX rs2106416 in 175 cases and 110 controls [13,18]. One study [18] included two subsets. The genotyping method in all studies was based on polymerase chain reaction (PCR). The controls in six studies [11,17,18,[21][22][23] had low/very low rate of dental caries and others were caries-free. In addition, the two groups in the studies were introduced as the caries  The genotype prevalence of each polymorphism included in each study in both case and control groups and the p-value for the Hardy-Weinberg equilibrium (HWE) for the controls are shown in Table 2. The controls in three studies [11,12,21] had a deviation from the HWE (P < 0.05). Table 3 shows the results of pooled analysis of each polymorphism based on five genetic models. Among the polymorphisms, only the G allele [OR = 1.38; 95%CI: 1.08, 1.76; P = 0.009; I 2 = 27% (P heterogeneity or P h = 0.25)] and the GG genotype [OR = 1.41; 95%CI: 1.06, 1.87; P = 0.02; I 2 = 18, (P h = 0.30)]) of ENAM rs3796704 had an elevated risk in the case group compared with the control group. In addition, the funnel plots of each polymorphism are presented in the Supplementary file. Therefore, there was a significant association between ENAM rs3796704 polymorphism and dental caries susceptibility. Table 4 identifies the subgroup analyses based on the ethnicity, age group, and the control group for each polymorphism and in five genetic models. The results showed that the G allele and the GG and AG genotypes of ENAM rs3796704 had an elevated risk in the case group compared with the control group in the Caucasian ethnicity; whereas in studies with a caries-free control group, the G allele and GG genotype had an elevated risk in the case group compared with the control group. In addition, the T allele, and TT and CT genotypes of AMELX rs17878486 polymorphism had an elevated risk in the case group compared with the control group in mixed ethnicity and studies with a caries-free control group. Therefore, there was a significant association between both polymorphisms of ENAM rs3796704 and AMELX rs17878486 and dental caries susceptibility in the Caucasian ethnicity and studies including caries-free individuals as the control group.

Sensitivity analysis
One study [19] was omitted from the analysis of AMELX rs17878486 because the outlier data and the results illustrated that by deleting this study, the CT [OR = 3.07; 95%CI: 1.36, 6.94; P = 0.007; I 2 = 0% (P h = 0.37)] and CT + TT [OR = 5.72; 95%CI: 2.83, 11.59; P = < 0.00001; I 2 = 21% (P h = 0.0.28)] genotypes in dental caries patients were significantly superior to controls and with a low heterogeneity, respectively (Table 3). In addition, other sensitivity analyses including "one study excluded" and "cumulative analysis" were performed and the previous results did not change qualitatively. Although the genotype distribution of the controls in three studies [11,12,21] did not follow the HWE, these analyses reported that the pooled ORs based on all genetic models were steady.

Publication bias
Both Egger's and Begg's tests were done on the previous pooled analyses with a minimum of three studies (Fig. 2). The results revealed a publication bias regarding GG vs. AA, AG vs. AA, and AG + GG vs. AA models of LTF rs1126478 polymorphism (Begg's test: P < 0.05) and also T vs. C and CT + TT vs. CC models of AMELX rs2106416 polymorphism (Begg's test: P < 0.05).

Discussion
The effect of environmental risk factors [26] and genetic predisposition [27] on development of caries has been well identified. This meta-analysis assessed the association between the common polymorphisms of LTF, ENAM, and AMELX and the risk of dental caries. The findings showed that ENAM rs3796704 polymorphism had an increased risk in the case group compared with the control group especially in the Caucasian ethnicity and studies with caries-free individuals as the control group. The prevalence of LTF polymorphisms differs among various ethnicities [28]. The genotypes related to LTF level are associated with decreased function and this may lead to decreased defense against infection with stronger stimuli for granulocyte activation and desorption, leading to greater LTF release [28]. Four studies [5,16,20,24] included in this meta-analysis assessed LTF rs1126478 polymorphism. However, none of them found any association between this polymorphism and dental caries susceptibility. In addition, the subgroup analysis in this meta-analysis did not find any association between  AMELX is the most significant factor for development of normal enamel [17] and AMELX aberration predominantly causes mineralization defects and congenital disorders such as amelogenesis imperfecta [29]. Therefore, some researchers suggest that genetic variations contribute to structural changes of the enamel and may create more levels of mineral loss, bacterial extension or biofilm deposition [30]. In our meta-analysis on AMELX rs946252, AMELX rs17878486, AMELX rs6639060, and AMELX rs2106416 polymorphisms, none of them was associated with the risk of dental caries. However, the sensitivity analysis revealed that AMELX rs17878486 polymorphism could be a risk factor for dental caries but the ethnicity and type of selected controls were the effective factors on the association between AMELX rs17878486 polymorphism and risk of dental caries. In addition, two studies included in our meta-analysis [11,15] showed an association between AMELX rs17878486 polymorphism and dental caries susceptibility. Therefore, it may point to the role of AMELX rs17878486 polymorphism in development of dental caries more than others. ENAM gene may also play a role in enamel formation [31] and any change in genes that encode enamel proteins may lead to enamel malformation [32]. Among studies on ENAM polymorphisms included in this meta-analysis, one study on Saudi patients [25] reported an elevated risk of ENAM rs1264848 polymorphism and another study on Polish children [14] showed the protective role of ENAM rs1264848 polymorphism in dental caries, which was confirmed by Abbasoğlu et al., in their study on Turkish children [33]. Also, one study [25] reported an elevated risk of ENAM rs3796704 polymorphism and another study [5] showed an elevated risk of ENAM rs3796703 polymorphism in dental caries patients compared with controls. One research [22] reported that there were significant differences in the minor allele frequency between the Poles and Czechs populations. In addition, the sensitivity analysis confirmed the effect of ethnicity and type of control group on the association of ENAM rs3796704 and risk of dental caries. Bayram and colleagues [34] showed that ENAM polymorphisms may affect the development of enamel and these effects may be different between primary and permanent dentitions.
Small number of studies included in each analysis and different DMFT scores for selecting the cases and controls between the studies were two important limitations of the present meta-analysis. However, in most analyses, the heterogeneity was low and there was no publication bias.

Conclusions
The findings of this meta-analysis confirmed that the G allele and the GG genotype of ENAM rs3796704 polymorphism were associated with an elevated risk of caries in the case group compared with the control group. But there was no association between LTF rs1126478, ENAM rs1264848, ENAM rs3796703, AMELX rs946252, AMELX rs17878486, and AMELX rs2106416 polymorphisms and dental caries susceptibility. However, subgroup analysis showed an association between ENAM rs3796704 and AMELX rs17878486 polymorphisms and dental caries susceptibility in the Caucasian ethnicity and studies including caries-free individuals as the control group. In addition, sensitivity analysis showed an increased risk of AMELX rs17878486 polymorphism in the case group compared with the control group. The future analyses with more cases in various areas should have focused on possible effects of gene-environmental interactions on caries experience.