Introduction
Helicobacter pylori (H. pylori), a prevalent gastric pathogen, affects over fifty percent of the global population, with more than 10 million individuals newly infected.1 The occurrence and progression of stomach pathologies, such as peptic ulcer disease, gastric cancer, and mucosa-associated lymphoid tissue lymphoma, have been linked to this infection.2,3 Interestingly, some individuals have never been infected with H. pylori during their lifetime.4 Studies have shown that susceptibility to infection depends on a combination of H. pylori virulence factors, environmental factors, genetic susceptibility of the host, and the effectiveness of the host immune system.5,6
Toll-like receptors (TLRs) belong to the pathogen-associated molecular pattern family, which specifically recognizes various pathogenic microorganisms, including H. pylori, and activates both specific and non-specific immune responses.7 TLRs’ specific recognition of ligands triggers NF-κB activation, which subsequently promotes the production of inflammation-related cytokines and chemokines.8,9 TLRs play an essential role in the recognition of H. pylori and the subsequent innate and adaptive immune responses.10,11H. pylori expresses various pathogen-related molecular pattern antigens, including lipopolysaccharide (LPS) and flagellin.12 TLR10, as a functional receptor, is involved in the innate immune response to H. pylori infection. TLR10 can form a heterodimer with TLR2, called the TLR2/TLR10 heterodimer, which functions in H. pylori LPS recognition.13–15 A study that employed real-time quantitative polymerase chain reaction and immunohistochemical examination on gastric biopsy specimens from both H. pylori-infected patients and uninfected individuals revealed a significant increase in the expression of TLR10 mRNA and TLR10 in the gastric epithelial cells of infected patients. Upon exposure to heat-killed H. pylori or its lipopolysaccharide, the TLR2/TLR10 heterodimer, among the TLR2 subfamily heterodimers, elicited the strongest NF-κB activation. These observations underscore the pivotal role of TLR10 in the immunological defense against H. pylori infection.16
The presence of genetic variants in TLRs is hypothesized to significantly influence an individual’s predisposition to H. pylori infection.17 Accumulating evidence indicates that genetic polymorphisms and expression levels of TLR2, TLR4, TLR5, and TLR9 can modulate susceptibility to H. pylori infection.18TLR10 rs10004195 is a polymorphic site located within the promoter region of the TLR10 gene, where the nucleotide changes from T to A.19 Recently, numerous case-control studies have found a correlation between the TLR10 rs10004195 gene polymorphism and susceptibility to H. pylori infection. However, due to limited statistical certainty in studies with small sample sizes, results have indicated that certain TLR10 rs10004195 genotypes may increase, decrease, or have no effect on susceptibility to H. pylori infection. To further explore the role of TLR10 rs10004195 in the risk of H. pylori infection, we conducted a quantitative analysis of high-quality research findings from these studies, aiming to eliminate the impact of factors such as insufficient sample sizes.20
Materials and methods
Literature search
Our meta-analysis adhered to the guidelines in the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 Checklist (File S1).21–24 The databases, containing PubMed, EMBASE, Web of Science, and China National Knowledge Infrastructure, were independently searched by two authors, including literature from database establishment to February 2024. The search keywords were: “Polymorphism” or “Polymorphisms” or “Variant”, “Helicobacter pylori infection” and “Toll-like receptor”. No restrictions were placed on publication dates or language.
Criteria for inclusion and exclusion
The inclusion criteria were as follows: (a) A case-control study that assesses the relationship between TLR10 rs10004195 polymorphism and the risk of H. pylori infection; (b) Availability of exact genotype frequencies, either directly obtained or indirectly calculated from the literature; and (c) Genetic loci in the control group assessed for Hardy-Weinberg equilibrium conformity. The exclusion criteria were: (a) Studies that did not adhere to a case-control design; (b) Literature with missing or duplicate data; and (c) Among similar studies published by the same author, the research with the largest sample size was selected.
Data extraction and research quality evaluation
Two researchers independently retrieved and input information based on the predefined inclusion and exclusion criteria, followed by data comparison. In cases of disputed literature data, discussions were held with the corresponding author to reach a consensus. Extracted data included the first author, publication year, continent (country), age, gender, sample size, Hardy-Weinberg equilibrium conformity, and genotype distribution. The quality assessment of the articles was conducted utilizing the Newcastle-Ottawa scale.25 The criteria encompassed three key dimensions: the method of selecting cases and controls (0–4); the degree of comparability between groups (0–2); and the assessment of exposure outcomes and relevant factors (0–3).
Statistical analysis
Based on the genotype distribution information extracted from each included study, SPSSAU was used to calculate the OR and 95% CI for each genetic model. Heterogeneity testing and meta-analysis were performed using Stata 17.0 software. To account for heterogeneity, a random effects model was initially applied to aggregate and contrast outcomes. When p < 0.10 or I2 > 50%, heterogeneity was indicated, and the random effects model was used for analysis. Otherwise, a fixed effects model was applied. Publication bias evaluation and sensitivity analysis were conducted using SPSSAU.
Results
Literature retrieval and evaluation results
Figure 1 illustrates the flowchart of the search process for the current meta-analysis. In total, eight studies investigating TLR10 rs10004195 were included, most of which were conducted in Asia. The total number of subjects was 5,144, including 3,004 individuals in the H. pylori-positive group and 2,140 in the negative control group. Table 1 presents a comprehensive overview of the fundamental details and genotype distribution in the literature (Table S1).10,26–32 The Newcastle-Ottawa scale assessment revealed consistently high quality across all studies, with an average score of 8.00, as shown in Table 2.10,26–32
Table 1Attributes of all case-control investigations included in the meta-analysis
| Literature | Year | Continent (Country) | Age
| Gender
| Size of the sample | HWE | Genotype type frequency (Case)
| Genotype type frequency (Control)
|
|---|
| HP (+) | HP (–) |
|---|
| HP (+)/HP (–) | M/F | M/F | HP (+)/HP (–)* | AA | AG | GG | AA | AG | GG |
|---|
| Emad M. Eed10 | 2020 | Aisa (Arabia) | 45±17.7/42±22.3 | 117/93 | 41/39 | 210/80 | Yes | 124 | 61 | 25 | 33 | 30 | 17 |
| Fu-bing Tang26 | 2015 | Aisa (China) | 48.9±6.4/49.5±6.8 | 731/780 | 473/569 | 1,486/1,008 | Yes | 498 | 712 | 276 | 308 | 493 | 207 |
| Laith AL-Eitan27 | 2021 | Aisa (Jordan) | – | – | – | 223/217 | Yes | 11 | 8 | 204 | 28 | 7 | 182 |
| Sevgi Kalkanli Tas28 | 2020 | Aisa-Europe (Turkey) | 47.7±12/51.2±12 | 83/122 | 84/111 | 205/195 | Yes | 54 | 37 | 114 | 7 | 38 | 150 |
| Taweesak Tongtawee29 | 2018 | Aisa (Thailand) | 46±1.5/42±2.5 | 71/133 | 65/131 | 204/196 | Yes | 59 | 10 | 135 | 22 | 123 | 51 |
| Ying Su30 | 2016 | Aisa (China) | – | – | – | 418/234 | Yes | 130 | 190 | 98 | 72 | 125 | 37 |
| Xin-juan Yu31 | 2014 | Aisa (China) | 48.7±10.8/49.4±12.1 | 142/59 | 130/52 | 201/182 | Yes | 54 | 87 | 60 | 33 | 93 | 56 |
| M. Ravishankar Ram32 | 2015 | Asia (Malaysia) | – | – | – | 57/28 | Yes | 24 | 18 | 15 | 9 | 11 | 8 |
Table 2Evaluation of the quality of eight case-control studies using the Newcastle-Ottawa scale criteria
| Literature | Selection of enrolled study subjects | Between-group comparability | Exposure outcomes and factors | Total |
|---|
| Emad M. Eed10 | 3 | 2 | 3 | 8 |
| Fu-bing Tang26 | 3 | 2 | 3 | 8 |
| Laith AL-Eitan27 | 4 | 2 | 3 | 9 |
| Sevgi Kalkanli Tas28 | 4 | 2 | 3 | 9 |
| Taweesak Tongtawee29 | 3 | 2 | 2 | 7 |
| Ying Su30 | 3 | 2 | 3 | 8 |
| Xin-juan Yu31 | 3 | 2 | 3 | 8 |
| M. Ravishankar Ram32 | 3 | 2 | 2 | 7 |
| Average | 3.25 | 2 | 2.75 | 8.00 |
Allele and genotype-wide meta-analysis
After conducting a meta-analysis on each of the five genetic models from the eight investigations included in this meta-analysis, the results revealed the following: The overall detection rate of the recessive homozygote AA genotype was 28.5%. In the H. pylori-positive group, the detection rate of the recessive homozygote AA genotype was 31.8%, while in the negative control group, this proportion was only 23.9%. Using the recessive genetic model (AA vs. AT+TT), individuals with the TLR10 rs10004195 AA genotype were found to have a significantly elevated risk of contracting H. pylori infection compared to those with the other two genotypes (OR = 1.64, 95% CI: 1.04–2.58, p = 0.034), as shown in Table 3 and Figure 2. No statistically significant relationships were found in any of the other genetic models (Figs. S1–S4).
Table 3Meta-analysis results of the relationship between TLR10 rs10004195 gene polymorphism and the risk of H. pylori infection
| Comparison | N | Association analysis
| Mode | Heterogeneity analysis
|
|---|
| OR | 95%CI | p | p | I 2(%) |
|---|
| A vs. T | 8 | 1.12 | (0.79, 1.59) | 0.512 | Random | <0.001 | 91.90 |
| AA vs. AT+TT | 8 | 1.64 | (1.04, 2.58) | 0.034 | Random | <0.001 | 87.20 |
| AA vs. TT | 8 | 1.37 | (0.82, 2.29) | 0.235 | Random | <0.001 | 85.30 |
| TT vs. AT+AA | 8 | 1.14 | (0.65, 2.03) | 0.646 | Random | <0.001 | 92.70 |
| TA vs. TT | 8 | 0.64 | (0.32, 1.25) | 0.191 | Random | <0.001 | 92.50 |
Between-study publication bias and sensitivity analysis
We conducted a publication bias analysis on the literature included in this study regarding the TLR10 rs10004195. The Egger’s test results showed no statistically significant asymmetry (p: 0.308–0.883), indicating no significant publication bias (Figs. S5–S9). Sensitivity analysis results indicated that the current results were stable; regardless of which study was excluded, no significant change was observed in the combined effect size.
Discussion
Genetic variations in TLR10 have been linked to various infectious diseases, such as complex skin and subcutaneous tissue infections, tuberculosis, bacterial meningitis, and Congo hemorrhagic fever.33–36 Recent studies have also found that the TLR10 rs10004195 gene polymorphism may influence the risk of host susceptibility to H. pylori infection to a certain extent.28 However, the conclusions drawn by different studies are not entirely consistent. Our findings reveal a notably elevated detection frequency of the recessive homozygous AA genotype in the H. pylori-positive cohort compared to the negative control group. Furthermore, carrying the TLR10 rs10004195 AA genotype significantly increases the risk of H. pylori infection (OR = 1.64) when analyzed using a recessive genetic model.
Previous studies conducted by Taweesak Tongtawee and Xin-juan Yu also support our conclusion. Interestingly, research by M. Ravishankar Ram indicated no correlation between TLR10 rs10004195 polymorphism and H. pylori infection among Malaysian individuals.29,31,32 We attribute this discrepancy to the relatively small sample size in their study and to Malaysia’s status as a country of multi-ethnic immigrants. Our research, which integrates all currently published high-quality studies, establishes a substantial association between TLR10 rs10004195 polymorphism and the susceptibility to H. pylori infection in human hosts, with stable results and no evident publication bias. However, the study participants were predominantly from Asian populations, and thus no stratified analysis by ethnicity was conducted. If further research provides sufficient data, meta-analyses with ethnic stratification could yield more valuable results.
TLR10 has been suggested to be a primary receptor involved in the innate immune response elicited by H. pylori infection.16 Previous studies have shown that the expression of TLR10 and its heterodimer on gastric mucosal epithelium can contribute to the immune reaction following H. pylori infection by augmenting NF-κB activation and promoting interleukin-1β production.37,38TLR10 gene polymorphisms can modulate the balance between pro-inflammatory and anti-inflammatory reactions, thereby regulating susceptibility to infections and autoimmune diseases.39 The rs10004195 SNP in TLR10’s promoter may affect TLR2/10 heterodimer-mediated immune responses.40 This variant in TLR10 could impair the ability of the TLR2/10 heterodimer on the gastric mucus layer to recognize H. pylori LPS. Another study found that, compared to other genotypes at the TLR10 rs10004195 locus, the AA homozygous genotype is associated with a heightened degree of inflammation in Thai patients with H. pylori-related gastritis.15 Therefore, the evidence suggests that TLR10 rs10004195 can influence the host’s ability to recognize H. pylori and modulate the functionality of the TLR2/10 heterodimer, thereby affecting the immune response to H. pylori infection.
Limitations and future directions
Our meta-analysis did not conduct a stratified analysis based on geographic regions, which is a major limitation since the study populations we included were primarily from Asian regions. Additionally, all the studies in the meta-analysis were case-control studies. Although we retrieved a population-based study conducted by Julia Mayerle,41 we were unable to extract the genotype frequencies necessary for our analysis from their article.
Our current analysis suggests that the TLR10 rs10004195 gene polymorphism can serve as an important reference indicator for predicting the risk of H. pylori infection in healthy individuals. In the future, it is anticipated that a polygenic risk score model for H. pylori infection prediction will be developed, incorporating information from multiple loci, including this one. Of course, establishing such a prediction model will require a large amount of high-quality data for support. Therefore, future research efforts are essential to explore the polymorphism of this gene locus and the risk of H. pylori infection in non-Asian populations. Additionally, cohort studies need to be conducted in multiple regions.
Conclusions
By conducting a meta-analysis of high-quality studies on this relevant topic, our research has unified the previous discrepancies in studies examining the relationship between TLR10 rs10004195 gene polymorphism and susceptibility to H. pylori infection, arriving at a conclusion that is more reliable than those drawn from individual studies. Individuals carrying the TLR10 rs10004195 AA genotype have a significantly higher risk of H. pylori infection. Therefore, in clinical applications, patients can be classified based on their genetic characteristics, including the findings from this study, to assess their genetic susceptibility to H. pylori infection. Incorporating individual gene polymorphisms enhances the precision of H. pylori infection risk assessment, allowing for more personalized health recommendations for patients.
Supporting information
Supplementary material for this article is available at https://doi.org/10.14218/ERHM.2024.00023 .
File S1
PRISMA 2020 Checklist.
(DOCX)
Table S1
Additional information for the literature incorporated in the meta-analysis.
(DOCX)
Fig. S1
Forest plot for the associations between TLR10 rs10004195 polymorphism and H.pylori infection risk through allelic model (A vs.T).
(DOCX)
Fig. S2
Forest plot for the associations between TLR10 rs10004195 polymorphism and H.pylori infection risk through homozygous model (AA vs.TT).
(DOCX)
Fig. S3
Forest plot for the associations between TLR10 rs10004195 polymorphism and H.pylori infection risk through dominant genetic model (TT vs.AT+AA).
(DOCX)
Fig. S4
Forest plot for the associations between TLR10 rs10004195 polymorphism and H.pylori infection risk through heterozygous model (AT vs.TT).
(DOCX)
Fig. S5
Funnel plot of allelic model (A vs. T).
(DOCX)
Fig. S6
Funnel plot of recessive genetic model (AA vs. AT + TT).
(DOCX)
Fig. S7
Funnel plot of homozygous genetic model (AA vs. TT).
(DOCX)
Fig. S8
Funnel plot of dominant genetic model (TT vs. AT + AA).
(DOCX)
Fig. S9
Funnel plot of heterozygous genetic model (AT vs. TT).
(DOCX)
Declarations
Data sharing statement
No additional data are available.
Funding
This study did not receive any funding support.
Conflict of interest
QJD has been an editorial board member of Exploratory Research and Hypothesis in Medicine since December 2018. The authors have no other conflicts of interest regarding the publication of this article.
Authors’ contributions
Design of the study (QJD, LLW), selection and data collection process (ZJX, WL, WLL), evaluation of methodological criteria (ZJX, DLJ), disagreement consultation (QJD), and corrections on formulations and illustrations (QJD, LLW). All authors approved the final manuscript.