Association of Toll-like receptor signaling molecule gene polymorphisms with risk of cervical lesions in HPV-infected women Authors: Sutong Wang, Chunlin Zhang, Ting Li, Ping Luo, Yang Yang, Zicong Lin, Bing Mei Published: 2025 ABSTRACT BackgroundToll-like receptor signaling molecules are involved in tumor associated inflammatory response. This study aims to investigate the role of Toll-like receptor signaling molecule gene polymorphisms in the risk of HPV-related cervical lesions.MethodsA total of 352 study subjects, 178 women with cervical lesions and 174 women without cervical lesion were involved in a case-control study. Real-time polymerase chain reaction (real-time PCR) was used to detect HR-HPV genotypes. Genotyping of SNPs for MYD88 gene (rs7744) and TRAF3 gene (rs1131877, rs12435483, rs3803286, rs12147254) were performed by Sanger sequencing. The MYD88 and TRAF3 haplotype frequencies were analyzed using the HaploView 4.2 program. Gene-gene interactions were analyzed using the multifactor dimensionality reduction (MDR).ResultsThe TRAF3 rs3803286 (A) variant allele (OR = 0.724, p = 0.045) and TRAF3 rs1131877 (TC) variant genotype (OR = 0.563, p = 0.023) were found to be associated with a significantly reduced risk of cervical lesions. rs3803286-HPV-16/18 was the best interaction model for the risk of cervical lesions (p = 0.0004).ConclusionTRAF3 rs3803286 and rs1131877 are potential biomarkers of susceptibility to cervical lesions.Clinical trial numberNot applicable.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12879-025-12133-5. Background Cervical cancer (CC) is the leading cause of cancer-related deaths worldwide and the fourth most common cancer among women. According to the World Health Organization, in 2022, there were 660,000 newly diagnosed cases, out of which 350,000 women died from this disease [1]. According to the latest Global Burden of Cancer Study, China has the second highest incidence (i.e., 110,000 new cases) and mortality (i.e., 59,000 deaths) of cervical cancer in the world [2]. Persistent infection of human papillomavirus (HPV) is the main cause of cervical cancer [3]. Almost all cervical cancers are caused by infection with some types of high-risk human papillomavirus (HR-HPV), among which HPV16 and HPV18 account for 71% of global cervical cancer cases [4]. The development of cervical cancer is a multi-step process. The cervical intraepithelial neoplasia (CIN) is routinely classified into three levels: low-grade squamous intraepithelial lesions (LSILs), high-grade squamous intraepithelial lesions (HSILs), and invasive cervical cancer [5]. The transition from CIN to invasive cancer may take several years or even decades, providing numerous opportunities for intervention. Only a small number of infected individuals will develop invasive cervical cancer at last [6]. It is known that approximately 6.7 to 14.9% of HR-HPV infected patients had a cervical lesion [7]. The development of cervical cancer is not solely caused by HPV, also involved other factors, including environmental and genetic factors [8]. Toll-like receptors (TLRs) are the major pattern recognition receptors (PRRs) in the innate immune system, and they are essential for host defense and tumor immune surveillance [9]. TLRs signaling pathway play an important role in HPV clearance by stimulating the innate immune response [10]. It has been reported that the inactivation of the TLR pathway may result in evasion of the innate immune system and increase the survival probability of HPV-positive tumor cells [11]. TLR family in the human consists of 10 receptors (TLR1-10) [12, 13]. TLRs recognize a broad spectrum of microbial components, with each TLR recognizing a specific pathogen-associated molecular patterns (PAMPs) [14]. After recognizing PAMPs, TLRs recruit specific adaptor proteins, including myeloid differentiation factor 88 (MYD88) and Toll/interleukin-1 receptor domain–containing adaptor inducing interferon-β (TRIF), and sequentially stimulate multiple signaling pathways to induce target gene expression [15]. The signaling pathway mediated by MYD88 is referred to as the MYD88-dependent pathway, which is a common pathway for the signaling of almost all TLRs (except for TLR3). The signaling pathway connected by TRIF is known as the MYD88-independent pathway [16]. TNF receptor associated factor 3 (TRAF3) plays a very important role in the regulation of MYD88-dependent and TRIF-dependent signal transduction through its own degradation, which can activate MYD88-dependent signal transduction and inhibit TRIF-dependent signal transduction [17,18]. Single nucleotide polymorphisms (SNPs) may affect the transcription, translation and functional expression of genes. Current studies indicated that the CT genotype of the rs12435483 polymorphism in the TRAF3 gene reduces the risk of cervical precancerous lesions [30]. Other loci of the TRAF3 gene have not been reported in relation to cervical lesions. Additionally, MYD88 is also an important molecule in the TLRs signaling pathway, and its polymorphisms in relation to cervical lesions have not yet been studied. Therefore, it is necessary to conduct such research. SNPs in the TLR4 and TLR9 genes have been linked to susceptibility to viral infections and related cancers [19-21]. At present, there are few studies on the correlation between the downstream molecular gene polymorphism of TLRs signaling pathway and HR-HPV infection and cervical lesions, and difference exists in gene polymorphism of TLRs signaling pathway downstream molecules between different ethnic and regional populations. The purpose of this study was to evaluate the possible association between polymorphisms of MYD88 and TRAF3 genes associated with TLRs signaling pathway with the risk of HPV infection and cervical lesions in central China, and to provide a scientific basis for disease prevention and treatment. Materials and methods Study subjects A case-control study recruited 352 women from 3721 female patients with positive HR-HPV typing results at Jingzhou Hospital affiliated to Yangtze University from April 2019 to October 2023. The inclusion criteria for the study subjects were HR-HPV positive and underwent cervical lesion screening through cytology and/or histopathology simultaneously. The exclusion criteria for the study subjects were: previous chemotherapy or radiotherapy treatment, other cancers, and autoimmune diseases. The patients without cervical lesions were designated as the control group (174 cases), and the patients with cervical lesions were designated as the case group (178 cases). The flow chart for study subject selection was shown in Fig. 1. The pathological diagnosis methods for cervical lesions include cervical thin-layer liquid-based cytology examination, colposcopy, and cervical biopsy, with histopathological diagnosis of biopsied tissue being the ultimate standard. The diagnostic criteria for cervical liquid-based cytology adopt the TBS classification system launched by the International Cancer Association, and the main diagnostic results include: (1) negative; (2) atypical squamous cell of undetermined significance; (3) atypical squamous cells cannot exclude high-grade intraepithelial lesions; (4) LSIL; (5) HSIL; (6) squamous cell carcinoma. The histological status, including cervical intraepithelial neoplasia (CIN) I, CIN II, CIN III and CC, was confirmed by pathological diagnosis of surgical specimens (CIN I was considered LSIL; CIN II and CIN III were considered HSIL). Fig. 1Flow chart of participants’ inclusion and exclusion Flow chart of participants’ inclusion and exclusion Specimen collection In this study, cervical exfoliated cells collected from research subjects served as experimental specimens of HPV testing and SNP genotyping, sampled using cervical swabs. The procedure was carried out by a qualified gynecologist, and the cervical exfoliated cells were preserved in sample tubes containing cell preservation solution. Samples were required to be sent for testing as soon as possible; if immediate analysis was not feasible, they should be stored in a refrigerator at 4 °C and tested within 48 h. DNA extraction and HPV testing DNA was extracted from cervical exfoliated cells of the above samples using nucleic acid extraction solution (Shanghai ZJ Bio-Tech Co. Ltd) according to the manufacturer’s instructions. Then, extracted DNA was used for HPV testing. At the same time, the extracted DNA were collected and stored at − 80 °C until using for SNP genotyping. The HPV test was performed using the HR-HPV typing test kit (Shanghai ZJ Bio-Tech Co. Ltd) according to the manufacturer’s instruction, including 15 types of HR-HPV (HPV-16, -18, -52, -58, -68, -66, -56, -31, -33, -35, -45, -82, -39, -51, -59). The HPV detection kit used in this study was approved by the National Medical Products Administration of China (IVD accredited number: 20153400044). SNP selection and genotyping SNPs were selected utilizing the SNP database from the National Center for Biotechnology Information (NCBI). Based on preliminary studies on a subset of samples, five SNPs were ultimately identified; one from the MYD88 gene (rs7744) and four from the TRAF3 gene (rs1131877, rs12435483, rs3803286, rs12147254). Real-time polymerase chain reaction (real-time PCR) was used to characterize the gene polymorphisms. Primers were designed using Primer Premier 6 and synthesized by Sangon Biotech. PCR was conducted in a total reaction volume of 25 µl with 15.875 µl ddH2O, 2.5 µl 10× buffer (Mg2+) (Takara), 2.0 µl dNTP (Takara), 1.25 µl of forward and reverse primers (20 µM) (Sangon Biotech), 0.125 µl Taq DNA polymerase (5 U/L) (Takara), and 2.0 µl DNA extract. The PCR amplification parameters were set as follows: preheating at 95 °C for 3 min, denaturation at 98 °C for 10 s, annealing at 59.8 °C to 63 °C for 30 s, extension at 72 °C for 1 min, with a total of 25 cycles. The reaction system and amplification parameters of PCR are consistent with our previous research methods [21]. Supplementary Table 1 provides detailed information on the primers, annealing temperatures, and extension lengths for genotyping the five SNPs. The PCR products were subsequently analyzed by 3% agarose gel electrophoresis and stained with GelRed nucleic acid dye (Biotium). Supplementary Fig. 1 provides the gel electrophoresis band images for these five SNPs. Sanger sequencing (Sangon Biotech) was employed to genotype these five SNPs. Statistical analysis In this study, quantitative data of the study subjects were presented as mean ± standard deviation (mean ± SD), while count data were presented as percentages. Differences in quantitative data and count data were compared using t-test and chi-square test or Fisher’s exact probability test, respectively. Logistic regression models were used for the analysis of genotypes and alleles, and odds ratios (ORs) with 95% confidence intervals (CI) were calculated to represent the risk factors. All statistical analyses were performed using SPSS software, version 25.0. The SNPStats online website (https://www.snpstats.net/preproc.php) was used to test whether the SNP conform to the Hardy-Weinberg equilibrium law in the general population. When p > 0.05, it indicates genetic equilibrium in the population, and the study subjects are representative of the population. Haploview 4.2 software was used to measure linkage disequilibrium and allele frequencies. Multifactor Dimensionality Reduction software was utilized to assess the impact of gene-environment interactions. The best model was chosen with the maximation of testing balanced accuracy. Significance was considered when the p-value was less than 0.05. Results Characteristics of study subjects The general characteristics of the study population were summarized in Table 1. Among the 352 HR-HPV-positive women, 178 had cervical lesions (LSIL, n = 64; HSIL, n = 75; CC, n = 39), and 174 were control subjects. The average ages (mean ± standard deviation) of cervical lesion Individuals (49.235 ± 11.727) and controls (41.810 ± 9.694) were significantly different (p = 0.029). The distributions of genotypes and allele frequencies of MYD88 and TRAF3 gene polymorphisms between case and control were all in accordance with Hardy–Weinberg equilibrium (Supplementary Table 2). Table 1General characteristics for the individuals with cervical lesions and controlVariableCervical lesion patients (n = 178)controls (n = 174)p-value Age (years) – Mean ± standard deviation49.235 ± 11.72741.81 ± 9.694 0.029 – Range20–7619–67 Histology (n %) – Low-grade squamous intraepithelial lesions64(36.0) – High-grade squamous intraepithelial lesions75(42.1) – Cervical cancer39(21.9) HPV infection (n %) – HPV-16/18 single infections43(24.2)20(11.5) – Other high-risk-HPV single infection†89(50)124(71.3) – High-risk-HPV coinfection46(25.8)30(17.2)Bold value is statistically significant†Including HPV-52, -58, -68, -66, -56, -31, -33, -35, -45, -82, -39, -51, -59 Table 1General characteristics for the individuals with cervical lesions and controlVariableCervical lesion patients (n = 178)controls (n = 174)p-value Age (years) – Mean ± standard deviation49.235 ± 11.72741.81 ± 9.694 0.029 – Range20–7619–67 Histology (n %) – Low-grade squamous intraepithelial lesions64(36.0) – High-grade squamous intraepithelial lesions75(42.1) – Cervical cancer39(21.9) HPV infection (n %) – HPV-16/18 single infections43(24.2)20(11.5) – Other high-risk-HPV single infection†89(50)124(71.3) – High-risk-HPV coinfection46(25.8)30(17.2)Bold value is statistically significant†Including HPV-52, -58, -68, -66, -56, -31, -33, -35, -45, -82, -39, -51, -59 General characteristics for the individuals with cervical lesions and control Bold value is statistically significant †Including HPV-52, -58, -68, -66, -56, -31, -33, -35, -45, -82, -39, -51, -59 Association of MYD88 and TRAF3 gene polymorphisms and haplotypes with the risk of cervical lesions Distinct genotype and allele distributions of MYD88 and TRAF3 SNPs among the case and control groups were presented in Table 2. Logistic regression analysis showed that, compared with wild pure genotype rs1131877 TT and rs3803286 GG carriers, rs1131877 TC genotype (TC: OR = 0.563, 95%CI = 0.342–0.925, p = 0.023), rs3803286 GA genotype (GA: OR = 0.463, 95%CI = 0.275–0.777, p = 0.004) had a significantly reduced risk of cervical lesions. In the dominant model, rs3803286 GA + AA genotype (GA + AA: OR = 0.493, 95%CI = 0.302–0.805, p = 0.005) was associated with a decreased risk of cervical lesions compared with wild homozygous genotype. Polymorphism loci alleles results showed the correlation between the lower risk of cervical lesions and TRAF3 rs3803286 mutant allele (A) (A: OR = 0.724, 95% CI = 0.527–0.993, p = 0.045). Table 2Distribution of TRAF3 and MYD88 gene polymorphisms among cervical lesion subjects and control group with their correlations to cervical lesion riskGeneSingle-nucleotide polymorphismGenotype/alleleControls, n = 174 (n %)Cases, n = 178 (n %)Odds ratio (95% CI)p-value*χ2MYD88rs7744AA73(41.95)79(44.38)1.00AG81(46.55)76(42.70)0.877(0.546–1.407)0.5860.529GG20(11.50)23(12.92)1.108(0.542–2.265)0.7790.167AA/AG + GG73(41.95)/101 (58.05)79(44.38)/99(55.62)0.922(0.589–1.443)0.7210.211A227(65.23)234(65.73)1.00G121(34.77)122(34.27)0.997(0.719–1.383)0.9870.019TRAF3rs12435483CC71(40.80)86(48.31)1.00CT81(46.55)70(39.33)0.654(0.405–1.055)0.0821.876TT22(12.64)22(12.36)1.042(0.515–2.109)0.9090.006CC/CT + TT71(40.80)/103(59.20)86(48.31)/92(51.69)0.729(0.467–1.140)0.1662.008C223(64.08)242(67.98)1.00T125(35.92)114(21.02)0.893(0.642–1.242)0.5001.192rs1131877TT60(34.48)71(39.89)1.00TC89(51.15)74(41.57)0.563(0.342–0.925) 0.023 3.245CC25(14.37)33(18.54)0.978(0.504–1.899)0.9491.113TT/TC + CC60(34.48)/114(65.52)71(39.89)/107(60.11)0.652(0.410–1.038)0.0711.100T209(60.06)216(60.67)1.00C139(39.94)140(39.33)0.884(0.642–1.219)0.4530.028rs3803286GG43(24.71)65(36.52)1.00GA98(56.32)81(45.50)0.463(0.275–0.777) 0.004 4.119AA33(18.97)32(17.98)0.582(0.302–1.123)0.1070.057GG/GA + AA43(24.71)/131(75.29)65(36.52)/113(63.48)0.493(0.302–0.805) 0.005 5.765G184(52.87)211(59.27)1.00A164(47.13)145(40.73)0.724(0.527–0.993) 0.045 2.923rs12147254GG63(36.21)62(34.83)1.00GA81(46.55)77(43.26)0.974(0.591–1.605)0.9190.386AA30(17.24)39(21.91)1.649(0.878–3.097)0.1201.217GG/GA + AA63(36.21)/111(63.79)62(34.83)/116(65.17)1.142(0.719–1.816)0.5740.073G207(59.48)201(56.46)1.00A141(40.52)155(43.54)1.257(0.914–1.728)0.1590.660* Adjusted for age Table 2Distribution of TRAF3 and MYD88 gene polymorphisms among cervical lesion subjects and control group with their correlations to cervical lesion riskGeneSingle-nucleotide polymorphismGenotype/alleleControls, n = 174 (n %)Cases, n = 178 (n %)Odds ratio (95% CI)p-value*χ2MYD88rs7744AA73(41.95)79(44.38)1.00AG81(46.55)76(42.70)0.877(0.546–1.407)0.5860.529GG20(11.50)23(12.92)1.108(0.542–2.265)0.7790.167AA/AG + GG73(41.95)/101 (58.05)79(44.38)/99(55.62)0.922(0.589–1.443)0.7210.211A227(65.23)234(65.73)1.00G121(34.77)122(34.27)0.997(0.719–1.383)0.9870.019TRAF3rs12435483CC71(40.80)86(48.31)1.00CT81(46.55)70(39.33)0.654(0.405–1.055)0.0821.876TT22(12.64)22(12.36)1.042(0.515–2.109)0.9090.006CC/CT + TT71(40.80)/103(59.20)86(48.31)/92(51.69)0.729(0.467–1.140)0.1662.008C223(64.08)242(67.98)1.00T125(35.92)114(21.02)0.893(0.642–1.242)0.5001.192rs1131877TT60(34.48)71(39.89)1.00TC89(51.15)74(41.57)0.563(0.342–0.925) 0.023 3.245CC25(14.37)33(18.54)0.978(0.504–1.899)0.9491.113TT/TC + CC60(34.48)/114(65.52)71(39.89)/107(60.11)0.652(0.410–1.038)0.0711.100T209(60.06)216(60.67)1.00C139(39.94)140(39.33)0.884(0.642–1.219)0.4530.028rs3803286GG43(24.71)65(36.52)1.00GA98(56.32)81(45.50)0.463(0.275–0.777) 0.004 4.119AA33(18.97)32(17.98)0.582(0.302–1.123)0.1070.057GG/GA + AA43(24.71)/131(75.29)65(36.52)/113(63.48)0.493(0.302–0.805) 0.005 5.765G184(52.87)211(59.27)1.00A164(47.13)145(40.73)0.724(0.527–0.993) 0.045 2.923rs12147254GG63(36.21)62(34.83)1.00GA81(46.55)77(43.26)0.974(0.591–1.605)0.9190.386AA30(17.24)39(21.91)1.649(0.878–3.097)0.1201.217GG/GA + AA63(36.21)/111(63.79)62(34.83)/116(65.17)1.142(0.719–1.816)0.5740.073G207(59.48)201(56.46)1.00A141(40.52)155(43.54)1.257(0.914–1.728)0.1590.660* Adjusted for age Distribution of TRAF3 and MYD88 gene polymorphisms among cervical lesion subjects and control group with their correlations to cervical lesion risk * Adjusted for age TRAF3 (rs3803286, rs12147254) and (rs12435483, rs1131877) SNPs were in strong linkage disequilibrium (Fig. 2). Three TRAF3 (rs3803286, rs12147254) haplotypes (A-G: 43.4%; G-A: 41.6%; G-G: 14.6%) and three TRAF3 (rs12435483, rs1131877) haplotypes (C-C: 38.9%; T-T: 33.2%; C-T: 27.2%) were found to be common in the Chinese population, showing a frequency > 0.01 (Table 3). No association was found between TRAF3 haplotypes and cervical lesion risk. Fig. 2Linkage disequilibrium pattern for the TRAF3 genes. The numbers inside the diamond represent the calculated D’, showing strong linkage disequilibrium between the TRAF3 (rs3803286, rs12147254) and TRAF3 (rs12435483, rs1131877). Haplotype frequencies in the haplotype blocks are shown under the matrix Linkage disequilibrium pattern for the TRAF3 genes. The numbers inside the diamond represent the calculated D’, showing strong linkage disequilibrium between the TRAF3 (rs3803286, rs12147254) and TRAF3 (rs12435483, rs1131877). Haplotype frequencies in the haplotype blocks are shown under the matrix Table 3Distribution of TRAF3 haplotypes among cervical lesion patients and controls and their associations with cervical lesion riskSNPHaplotypes†FrequenciesOdds ratio (95% CI)p-value‡Overall (%)Cases (%)Controls (%)rs3803286-rs12147254A-G43.440.146.81G-A41.642.940.20.79 (0.58–1.09)0.15G-G14.616.412.70.62 (0.38–1.01)0.058rs12435483-rs1131877C-C38.93938.81T-T33.231.734.71.10 (0.78–1.55)0.57C-T27.22925.30.88 (0.61–1.28)0.5Bold values are statistically significant†Adjusted for age‡The haplotype with a frequency ≥ 0.01 is shown in the table Table 3Distribution of TRAF3 haplotypes among cervical lesion patients and controls and their associations with cervical lesion riskSNPHaplotypes†FrequenciesOdds ratio (95% CI)p-value‡Overall (%)Cases (%)Controls (%)rs3803286-rs12147254A-G43.440.146.81G-A41.642.940.20.79 (0.58–1.09)0.15G-G14.616.412.70.62 (0.38–1.01)0.058rs12435483-rs1131877C-C38.93938.81T-T33.231.734.71.10 (0.78–1.55)0.57C-T27.22925.30.88 (0.61–1.28)0.5Bold values are statistically significant†Adjusted for age‡The haplotype with a frequency ≥ 0.01 is shown in the table Distribution of TRAF3 haplotypes among cervical lesion patients and controls and their associations with cervical lesion risk Bold values are statistically significant †Adjusted for age ‡The haplotype with a frequency ≥ 0.01 is shown in the table Association of MYD88 and TRAF3 gene polymorphisms with the risk of precancerous lesions and cervical cancer Cervical lesion group can be divided into subgroups as precancerous lesions (LSIL and HSIL) and cervical cancer. The data showed that heterozygous genotypes TRAF3 rs1131877 TC (p1 = 0.035), rs3803286 GA (p1 = 0.004) and GA + AA (p1 = 0.005) were associated with a decreased risk of precancerous lesions. In addition, heterozygous genotypes TRAF3 rs12435483 CT (p2 = 0.044) and CT + TT (p2 = 0.043) were associated with a decreased risk of cervical cancer. The frequency distribution of TRAF3 genotypes in precancerous lesions and cervical cancer subgroups did not differ significantly. Additionally, there was no discernible variation in the MYD88 polymorphism genotype frequency between the cervical cancer, precancerous lesion, and control groups (Table 4). Table 4MYD88 and TRAF3 gene polymorphism distribution in cervical cancer patients and precancerous lesions (low-grade squamous intraepithelial lesions + high-grade squamous intraepithelial lesions)GeneSingle-nucleotide polymorphismGenotypeControls, n = 174(n %)Precancerous lesions, n = 139(%)Cervical cancer, n = 39(%)p1*p2*p3*MYD88rs7744AA73(41.95)60(43.17)19(48.72)1.001.001.00AG81(46.55)61(43.88)15(38.46)0.7880.3800.680GG20(11.50)18(12.95)5(12.82)0.7150.8740.893AA/AG + GG73(41.95)/101(58.05)60(43.17)/79(56.83)19(48.72)/20 (51.28)0.9210.4930.699TRAF3rs12435483CC71(40.80)64(46.04)22(56.41)1.001.001.00CT81(46.55)56(40.29)14(35.90)0.198 0.044 0.323TT22(12.64)19(13.67)3(7.69)0.7250.3690.522CC/CT + TT71(40.80)/103(59.20)64(46.04)/75(53.96)22(56.41)/17 (43.59)0.356 0.043 0.280rs1131877TT60(34.48)57(41.01)14(35.90)1.001.001.00TC89(51.15)56(40.29)18(46.15) 0.035 0.2030.943CC25(14.37)26(18.70)7(17.95)0.9880.6600.951TT/TC + CC60(34.48)/114(65.52)57(41.01)/82(58.99)14(35.90)/25 (64.10)0.0990.2370.978rs3803286GG43(24.71)53(38.13)12(30.77)1.001.001.00GA98(56.32)62(44.60)19(48.72) 0.004 0.2840.952AA33(18.97)24(17.27)8(20.51)0.0970.7660.678GG/GA + AA43(24.71)/131(75.29)53(38.13)/86(61.87)12(30.77)/27 (69.23) 0.005 0.3460.834rs12147254GG63(36.21)46(33.10)16(41.03)1.001.001.00GA81(46.55)61(43.88)16(41.03)0.8830.6200.494AA30(17.24)32(23.02)7(17.95)0.1000.9200.754GG/GA + AA63(36.21)/111(63.79)46(33.10)/93(66.90)16(41.03)/23 (58.97)0.4350.7340.515*Adjusted for agep1: Precancerous lesions/Controls; p2: Cervical cancer/Controls; p3: Precancerous lesions/Cervical cancerPrecancerous lesions: low-grade squamous intraepithelial lesions + high-grade squamous intraepithelial lesions Table 4MYD88 and TRAF3 gene polymorphism distribution in cervical cancer patients and precancerous lesions (low-grade squamous intraepithelial lesions + high-grade squamous intraepithelial lesions)GeneSingle-nucleotide polymorphismGenotypeControls, n = 174(n %)Precancerous lesions, n = 139(%)Cervical cancer, n = 39(%)p1*p2*p3*MYD88rs7744AA73(41.95)60(43.17)19(48.72)1.001.001.00AG81(46.55)61(43.88)15(38.46)0.7880.3800.680GG20(11.50)18(12.95)5(12.82)0.7150.8740.893AA/AG + GG73(41.95)/101(58.05)60(43.17)/79(56.83)19(48.72)/20 (51.28)0.9210.4930.699TRAF3rs12435483CC71(40.80)64(46.04)22(56.41)1.001.001.00CT81(46.55)56(40.29)14(35.90)0.198 0.044 0.323TT22(12.64)19(13.67)3(7.69)0.7250.3690.522CC/CT + TT71(40.80)/103(59.20)64(46.04)/75(53.96)22(56.41)/17 (43.59)0.356 0.043 0.280rs1131877TT60(34.48)57(41.01)14(35.90)1.001.001.00TC89(51.15)56(40.29)18(46.15) 0.035 0.2030.943CC25(14.37)26(18.70)7(17.95)0.9880.6600.951TT/TC + CC60(34.48)/114(65.52)57(41.01)/82(58.99)14(35.90)/25 (64.10)0.0990.2370.978rs3803286GG43(24.71)53(38.13)12(30.77)1.001.001.00GA98(56.32)62(44.60)19(48.72) 0.004 0.2840.952AA33(18.97)24(17.27)8(20.51)0.0970.7660.678GG/GA + AA43(24.71)/131(75.29)53(38.13)/86(61.87)12(30.77)/27 (69.23) 0.005 0.3460.834rs12147254GG63(36.21)46(33.10)16(41.03)1.001.001.00GA81(46.55)61(43.88)16(41.03)0.8830.6200.494AA30(17.24)32(23.02)7(17.95)0.1000.9200.754GG/GA + AA63(36.21)/111(63.79)46(33.10)/93(66.90)16(41.03)/23 (58.97)0.4350.7340.515*Adjusted for agep1: Precancerous lesions/Controls; p2: Cervical cancer/Controls; p3: Precancerous lesions/Cervical cancerPrecancerous lesions: low-grade squamous intraepithelial lesions + high-grade squamous intraepithelial lesions MYD88 and TRAF3 gene polymorphism distribution in cervical cancer patients and precancerous lesions (low-grade squamous intraepithelial lesions + high-grade squamous intraepithelial lesions) *Adjusted for age p1: Precancerous lesions/Controls; p2: Cervical cancer/Controls; p3: Precancerous lesions/Cervical cancer Precancerous lesions: low-grade squamous intraepithelial lesions + high-grade squamous intraepithelial lesions Association and interaction of MYD88 and TRAF3 gene polymorphisms with HPV-16/18 infection We conducted gene-environment interaction analysis in the two-locus to four-locus models by Multifactor Dimensionality Reduction software. The interaction of TRAF3 gene polymorphisms with HPV-16/18 infection was observed to be related to cervical lesion risk. The rs3803286-HPV-16/18 two-locus model was the best model, and the test precision was 0.599 (p = 0.0004), with a good cross-validation consistency (CVC) (9/10). (Table 5) Table 5Multifactor dimensionality reduction analysis for gene–HPV-16/18 with TRAF3 gene polymorphisms interaction modelsLocus nModelsCross-validation consistencyTesting balanced accuracyp-value‡2rs3803286×HPV16/18 interaction†0.5999/10 0.0004 3rs12435483×rs3803286×HPV16/18 interaction0.57559/10 0.0001 4rs12435483× rs1131877× rs3803286×HPV16/18 interaction0.543210/10 0.0001 †The best interaction models‡p-value of each best model was calculated after a 1000-fold permutation test Table 5Multifactor dimensionality reduction analysis for gene–HPV-16/18 with TRAF3 gene polymorphisms interaction modelsLocus nModelsCross-validation consistencyTesting balanced accuracyp-value‡2rs3803286×HPV16/18 interaction†0.5999/10 0.0004 3rs12435483×rs3803286×HPV16/18 interaction0.57559/10 0.0001 4rs12435483× rs1131877× rs3803286×HPV16/18 interaction0.543210/10 0.0001 †The best interaction models‡p-value of each best model was calculated after a 1000-fold permutation test Multifactor dimensionality reduction analysis for gene–HPV-16/18 with TRAF3 gene polymorphisms interaction models †The best interaction models ‡p-value of each best model was calculated after a 1000-fold permutation test Discussion CC is a complex multifactorial disease, and there is increasing evidence that host genetic factors also play an important role in HPV infection. Our previous study has explored the relationship between TLR4/TLR9 gene polymorphisms and cervical lesions, and found that TLR4 rs10116253, rs1927911, and rs10759931 mutant genotypes have a protective effect on cervical lesions in HPV-infected women in central China [21]. However, the study never investigates the relationship between Toll-like signaling molecules gene polymorphisms and the risk of cervical lesions induced by HPV infection. Therefore, the aim of this study was to analyze the potential association between gene polymorphisms of molecules involved in TLRs signaling pathway and cervical lesions. Recent studies have shown that TLR4 can bind with MYD88 adapters, leading to activation of NF-κB, which results in the pro-inflammatory response [22]. However, it has been found that active infection of HR-HPV blocks the transmission of downstream signals from PRRs to the nucleus, thereby affecting the production of type I interferons and proinflammatory cytokines and chemokines [23]. As an important molecule in the Toll-like receptor signaling pathway, MYD88 has been repeatedly studied for its correlation with a variety of diseases, such as viral infections, cancer and autoimmune diseases [24-26], but its SNP rs7744 is rarely reported. MYD88 is located on chromosome 3p22, and its SNP rs7744 is located in the 3 ‘-untranslated region [27]. However, if MYD88’s expression and function are affected by its polymorphism, then rs7744, the only polymorphism in MYD88, may possibly be associated with susceptibility or pathogenesis of cervical lesions. As far as we know, currently, there are only a few studies on MYD88 polymorphism and its relationship with human diseases. Matsunaga Kazuhiro et al. [28] reported that rs7744 allele variants were significantly associated with susceptibility to ulcerative colitis. Li Yaping et al. [29] reported that MYD88 rs7744 gene polymorphism was not associated with the susceptibility and severity of enterovirus 71 hand-foot-and mouth disease in children. Additionally, there is few evidence regarding the clinical significance of rs7744. This study found no significantly correlation between MYD88 SNPs and risk of cervical lesions. TRAF3 rs12435483 is located in the intron region. Some studies have reported that TRAF3 rs12435483 CT genotype significantly reduced the risk of cervical precancerous lesions [30]. However, in this study, only the CT and CT + TT genotypes of TRAF3 rs12435483 were associated with a reduced risk of cervical cancer, without showing a significant correlation with the reduction of risk for cervical precancerous lesions. The research on relationship between cervical lesions and TRAF3 rs3803286, rs12147254, rs1131877 have not yet been reported. In the results of the present study, it was shown that individuals carrying TRAF3 rs1131877 (TC) and rs3803286 (GA and GA + AA) mutant alleles had a lower risk of cervical lesions than wild-type genotype carriers. The rs12147254 is located in the intronic region of the TRAF3 gene. TRAF3 rs12147254 has been found to affect the occurrence and progress of acute myeloid leukemia [31]. TRAF3 can inhibit NF-κB signaling by directly binding to the NF-κB-inducing kinase and degrading this kinase or by affecting other TRAF family members [32, 33]. Du et al. reported that the TRAF3 rs12147254 A allele was associated with a reduced risk of multiple myeloma, possibly because the TRAF3 gene polymorphism is responsible for NF-κB activation [34]. In addition, the intronic rs12147254 in the TRAF3 gene may affect the function of TRAF3 through the promoter region, splice site, or intronic microRNA. Zhou Zixing et al. [35] found that GA genotype and GA + AA genotype of TRAF3 rs12147254 increased the risk of coronary heart disease in patients with type 2 diabetes mellitus. For TRAF3 rs1131877, two researches have analyzed the association between TRAF3 rs1131877 and the disease [36, 37]. TRAF3 rs3803286 has not been reported, but it is interesting to note that our results proved that the TRAF3 rs3803286 (GA + AA) and GA genotype and A allele was associated with lower risk of cervical lesions. To the best of our knowledge, this is the first study to indicate an association existed between TRAF3 rs3803286 and cervical lesions. CC originates from precancerous lesions of SILs, including LSIL and HSIL [38]. Persistent HPV infection may induce TLR expression to increase, resulting in chronic inflammation and promoting the progress of cervical lesions. The role of MYD88 and TRAF3 gene polymorphisms in the progression of cervical lesions remains unknown to some extent. The subgroup analysis of this study revealed a certain correlation between TRAF3 gene polymorphisms and the occurrence of precancerous lesions and CC. Among the 352 patients in this study, 178 had cervical lesions and 174 did not. It is well known approximately 6.7%-14.9% of patients with HR-HPV infection exhibited lesions [7]. The reason for this difference may be age. A study has shown that the incidence of cervical lesions varies with age, with higher ages being more prone to precancerous lesions [39]. The data from this study indicated that the average age of the cervical lesion group was higher than that of the healthy group, and this difference was statistically significant. Another reason may be the sample size, so we will continue to increase the sample size in the later stage to verify in the future. Genetic susceptibility can influence the HPV infection status and alter the risk of cervical lesions. This study provides more comprehensive information for immunogenetic research on HPV-infected women in central China. However, as the patients’ clinical data were obtained from outpatient electronic medical records, and relevant information was not recorded by clinicians during the visit, we were unable to collect additional information such as smoking status, alcohol abuse, number of pregnancies, number of sexual partners and so on. These factors were reported to be associated with cervical lesions, and failure to consider these factors will weaken the power of the analysis. Additionally, the sample size was limited to some extent, and selection bias cannot be completely ruled out. Therefore, further large-scale studies involving diverse populations are still needed to validate our findings. Conclusion Our study suggested that TRAF3 rs3803286(GA and GA + AA) and rs1131877(TC) mutant genotypes were protective against cervical lesions in HPV-infected women in central China. The interaction between TRAF3 SNPs and HPV-16/18 infection has an impact on the risk of cervical lesions. This information may be helpful in the investigation of the etiology of cervical lesions and the assessment of the risk of cervical lesions. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3 Supplementary Material 4: Gel electrophoresis images of MYD88 and TRAF3 SNPs (A: rs7744; B: rs12435483; C: rs1131877; D: rs3803286; E: rs12147254). Supplementary Material 4: Gel electrophoresis images of MYD88 and TRAF3 SNPs (A: rs7744; B: rs12435483; C: rs1131877; D: rs3803286; E: rs12147254).