Introduction

Immunoglobulin A nephropathy (IgAN), characterized pathologically by deposition of IgA complexes in the glomerular mesangium, is the most common form of glomerulonephritis in children and adolescents and is considered to be an immune-complex disease [1]. Although its pathogenesis is unclear, experimental and clinical findings suggest that upper respiratory tract infections and tonsillitis are related to exacerbations of IgAN [22]. The level of IgA against Haemophilus parainfluenzae antigens was increased in tonsillar mononuclear cells from patient with IgAN [9]. In mice, Myxovirus parainfluenzae 1 produces a nice model of IgAN [12]. In patients with IgAN, infection of the tonsils with group A Streptococci is common and can cause renal deterioration [18]. These data suggest that exogenous antigens derived from pathogens could play a role in the pathogenesis of IgAN despite the fact that mechanisms through which these antigens trigger IgAN are still undefined. This fact emphasizes that the innate immune response is involved in the onset of IgAN and the production of IgA antibodies.

Toll-like receptors (TLRs) are a family of transmembrane receptors which play a fundamental role in pathogen recognition and activation of innate immunity. These receptors have been evolutionarily conserved to recognize specific microbial molecular components. Once activated, TLRs engage a signaling cascade resulting in the stimulation of innate and adaptive immune responses targeting the invading pathogen, with induction of cytokines, chemokines, cell surface adhesion molecules, and costimulatory molecules [19]. Thus, we speculate that TLRs may play an important role in pathophysiology of IgAN. Indeed, previous studies reported changed activation of TLRs in IgAN. The hyperexpression of TLRs was shown in peripheral lymphomonocytes of patients with IgAN during its acute flares [8]. An upregulation of TLR4 was also demonstrated in circulating monocytes of IgAN patients and in particular in those with proteinuria and macroscopic hematuria [7]. Moreover, the overexpression of TLR2 was shown in the kidney of patients with IgAN [16]. Suzuki et al. reported that transcriptional level of TLR9 was increased in spleens of ddY mice, which spontaneously develop IgAN [21]. Additionally, the treatment of ligands for TLR9 to ddY mice aggravated renal injury, led to strong T helper cell 1 polarization, and increased serum and mesangial IgA [21]. Furthermore, an association was observed between TLR9 gene polymorphisms and progressive IgAN, analyzed in two cohorts of patients with IgAN [21].

TLR10 is predominantly expressed in immune cell-rich tissues, including spleen, lymph node, and lung, and cell lines linked to the immune system such as promyelocytic HL-60s and the human B cell lines Ramos and Raji cells [5]. Bourke et al. showed that resting B cells stimulated with anti-mu and anti-CD40 antibodies or with Staphylococcus aureus had increased mRNA expression of TLR10 and TLR9 [3]. TLR9 and/or TLR10 were also found to be highly expressed in Epstein–Barr virus (EBV)-transformed B cell lines and in cell lines representative of mature B cell neoplasias (Burkitt lymphoma, follicular lymphoma, and multiple myeloma). Of the 13–15 TLRs described to date, no specific ligand for TLR10 has been defined so far. Therefore, its biological function is not well established yet. However, TLR10 gene has been revealed the genetic associations with infection or immune disease and has been suggested as susceptible factor [14, 15, 17, 24]. Mailaparambil et al. studied the association between 19 polymorphisms of TLR genes (TLR1, 2, 3, 5, 6, 9) and infection by respiratory syncytial virus (RSV) and suggested the association of one TLR9 promoter polymorphism and TLR10 haplotypes [17]. In addition, haplotypes of TLR10 gene were associated with nasopharyngeal carcinoma (related with EBV infection) [24]. Moreover, TLR10 gene has been revealed the association with asthma [14, 15].

Considering these findings, we expected that TLR10 may contribute to risk and/or progression of IgAN. In this study, we investigated a genetic association between TLR10 gene polymorphisms and childhood IgAN.

Materials and methods

Subjects

One hundred ninety-nine IgAN patients [median (interquartile range, IQR), 12.0 (9.5–14.3) years; male/female, 118/81] and 289 control subjects [30.0 (29.0–52.0) years; male/female, 158/131] were enrolled for this study. Demographic and clinical findings of IgAN patients show in Table 1. Each IgAN patient was diagnosed by two specialized pediatric nephrologists. All IgAN patients with prolonged hematuria or concomitant proteinuria were examined by kidney biopsy. IgAN patients were divided according to presence or absence of proteinuria at kidney biopsy (>4 mg/m2/h) [10, 11], gross hematuria episodes as an initial symptom of IgAN, and advanced pathological disease markers (interstitial fibrosis, tubular atrophy, or global sclerosis), respectively. The control group was recruited from subjects visiting the hospital for routine health checkups. Healthy adults without a history of IgAN symptoms and any other diseases were used as the control group. All IgAN patients and control subjects were recruited in Kyung Hee Medical Center in Seoul, Republic of Korea, and were of Korean background. Written informed consents were obtained from parents of IgAN patients and control subjects. The study was approved by the ethics review committee of the Medical Research Institute, Kyung Hee University Medical Center, Seoul, Republic of Korea.

Table 1 Demographic characteristics of the study subjects
Full size table

SNP selection and genotyping

We obtained the TLR10 genetic sequence and information for SNPs from the HapMap (http://www.hapmap.org/; genome build 36), SNP databases (dbSNP; www.ncbi.nlm.nih.gov/SNP; BUILD 131), and Ensemble database (http://www.ensembl.org; release 59). TLR10 gene is located on chromosome 4p14 and contains four exons. In the human TLR10 gene, theoretically six possible mRNA transcripts can be produced by alternative splicing, and so far, five splicing variants are identified (http://www.ensembl.org). The splicing variants of the TLR10 gene arise from transcripts initiated at exon 1 or 3. The full-length transcripts are 3,958 bp.

Of SNPs in TLR10 gene, the missense SNPs in the coding region and SNPs in promoter region within 1,000 bp from the transcriptional start site were selected. On TLR10 gene, there are 17 missense SNPs and 11 promoter SNPs within 1,000 bp from the transcriptional start site. The SNPs without genotype frequency data (rs61464560, rs73236610, rs62617795 in missense SNPs and rs10012016, rs10012017, rs79022084, rs10004521, rs9306966, rs10440315, rs10440316, rs75777505, rs73142650 in promoter SNPs) and with a heterozygosity ≤0.1 (rs4129008, rs11566660, rs11466658, rs11466656, rs77389171, rs11466650 in missense SNPs and rs10034903 in promoter SNPs) or a minor allele frequency ≤0.1 (rs11466657, rs11466655, rs11466653, rs11466651, rs11466649 in missense SNPs; particularly in Asian population) were excluded. As a result, we selected one promoter SNP (rs1004195, −135T/A) and three missense SNPs [rs11096957 (Asn241His), rs11096955 (Ile369Leu), and rs4129009 (Ile775Val)], which are SNPs of the TLR10 gene that were studied intensively before [14, 15, 17, 20, 24]. DNA was isolated from a peripheral blood sample using the DNA Isolation Kit for blood (Roche, Indianapolis, IN, USA). SNP genotyping was conducted using direct sequencing using the specific primers for each SNP (Table 2). The PCR products were sequenced using an ABI PRISM 3730XL analyzer (PE Applied Biosystems, Foster City, CA, USA). Sequence data were analyzed using SeqManII software (DNASTAR Inc., Madison, WI, USA).

Table 2 Sequences of primers used for each SNP in TLR10 gene
Full size table

Statistical analysis

For analysis of genetic data, HelixTree (Golden Helix Inc., Bozeman, MT, USA), SNPAnalyzer Pro (ISTECH Inc., Goyang, Republic of Korea), and SNPStats (http://bioinfo.iconcologia.net/index.php) were used. The associations between genotypes of SNPs and risk of IgAN were estimated by computing the odds ratios (ORs) and their 95% confidence intervals (CIs) with logistic regression analyses, controlling for gender as a covariable [10]. In logistic regression analysis for each SNP, models assuming either codominant inheritance (that is, the relative hazard differed between subjects with one minor allele and those with two minor alleles), dominant inheritance (that is, subjects with one or two minor alleles had the same relative hazard for the disease), or recessive inheritance (that is, only subjects with two minor alleles were at increased risk of the disease) were used. The Bonferroni correction was applied by multiplying p values by the number of SNPs analyzed (n = 4). For detecting that SNPs within TLR10 gene were linked each other, linkage disequilibrium (LD) was tested using Haploview version 4.2 (Broad Institute, Cambridge, MA, USA). The LD block was constructed using Gabriel’s method [2]. The haplotypes, combination of alleles at different loci, were reconstructed by using PL-EM algorithm in LD block. The association of haplotypes with IgAN was analyzed using HapAnalyzer version 1.0 (http://hap.ngri.go.kr/). The differences between TLR10 gene polymorphisms and the level of proteinuria (milligrams per square meter per hour) at kidney biopsy were analyzed using Statistical Package for the Social Sciences version 17.0 (SPSS Inc., Chicago, IL, USA). Statistical comparison of the levels among genotypes was assessed for each SNP by Kruskal–Wallis test for codominant model and by Mann–Whitney U test for dominant and recessive models. The p value < 0.05 was considered statistically significant.

We calculated the power of sample size to verify our data, using a genetic power calculator (http://pngu.mgh.harvard.edu/∼purcell/gpc/cc2.html). In this study, sample powers of SNPs were 0.965 (rs10004195, number of effective sample for 80% power = 114), 0.964 (rs11096957, number of effective sample for 80% power = 114), 0.964 (rs11096955, number of effective sample for 80% power = 114), and 0.965 (rs4129009, number of effective sample for 80% power = 113), respectively (α = 0.05, genotype relative risk = 2 fold). Thus, our data were thought to be acceptable.

Results

Four SNPs of TLR10 gene were polymorphic, and genotype distributions of the SNPs were in Hardy–Weinberg equilibrium (control group, p > 0.05); thus, in our study, there was no error of genotyping data, control subjects were selected randomly, and sample size was sufficient so as to minimize the effect of genetic drift.

As shown in Table 3, rs10004195 showed a significant association between IgAN patients and control subjects in codominant and dominant models. In codominant model, the frequencies of TT, TA, and AA genotypes were 31.1%, 46.7%, and 22.1% in the control group and 18.6%, 55.3%, and 26.1% in the IgAN group, respectively. Both TA heterozygote and AA homozygote of rs10004195 were associated with a significantly increased risk of IgAN (p = 0.016, OR = 1.97, 95% CI = 1.25–3.12 for TA heterozygote; p = 0.044, OR = 1.99, 95% CI = 1.17–3.38 for AA homozygote). In the dominant model, the frequencies of the genotype not containing A allele (TT) and the genotype containing A allele (TA or AA) were 31.1% and 68.9% in the control group and 18.6% and 81.4% in the IgAN group, respectively. The frequency of the genotype containing A allele was increased in IgAN patients, compared to control subjects (p = 0.0068, OR = 1.98, 95% CI = 1.28–3.06). The rest of SNPs (rs11096957, rs11096955, and rs4129009) were not statistically associated with IgAN.

Table 3 Multiple logistic regression analysis of TLR10 gene polymorphisms between immunoglobulin A nephropathy patients and control subjects
Full size table

In analysis of allele frequency by chi-square test, we also observed that allele frequency of rs1004195 was significantly different between IgAN patients and control subjects (p = 0.01, OR = 1.39, 95% CI = 1.08–1.80). The frequency of A allele was increased in IgAN group compared to control group (T and A allele, 54.5% and 45.5% in control group; 46.2% and 53.8% in IgAN group).

Four SNPs of TLR10 gene were analyzed for LD and haplotypes using Haploview 4.2 and HapAnalyzer. One LD block was constructed among rs11096957, rs11096955, and rs4129009 of SNPs of TLR10 gene (D′ ≥ 0.95, r 2 ≥ 0.8). However, haplotypes in the LD block were not associated with IgAN (data not shown).

We investigated the relationship between TLR10 gene and the level of proteinuria of IgAN. Analysis between codominant model of SNPs of TLR10 gene and the level was performed by Kruskal–Wallis test. The relationship between the dominant and recessive models of SNPs and the level was detected by Mann–Whitney U test. As shown in Table 4, the level of proteinuria was found to be significantly correlated with genotypes of rs1004195 in dominant model (TT vs. TA/AA; p = 0.033). In dominant model of rs1004195, the level of proteinuria of patients with genotype containing A allele [median, 4.01 mg/m2/h (IQR, 1.76–10.66 mg/m2/h)] was higher than that of patients with genotype not containing A allele [median, 2.00 mg/m2/h (IQR, 1.22–6.08 mg/m2/h)].

Table 4 The relationship between proteinuria level and TLR10 gene polymorphisms
Full size table

We also analyzed the association of SNPs of TLR10 gene in patient subgroups determined by the presence or absence of gross hematuria, or advanced disease marker; however, no association was observed (data not shown).

Discussion

This study evaluated genetic association between TLR10 gene polymorphisms and IgAN in a Korean population. Our results show that TA/AA genotype containing A allele is more frequent in IgAN patients as compared to control subjects and is associated with an increased risk of IgAN. We also find the significant correlation between the level of proteinuria of patients with IgAN and the genotypes of rs1004195.

TLR10 gene polymorphisms and in particular rs11096957, rs11096955, and rs4129009 analyzed in this study were reported to be associated with infections [17, 24] as well as with asthma [14, 15]. Zhou et al. reported that the haplotypes containing rs11096957 and rs11096955 were associated with EBV infection-related nasopharyngeal carcinoma [24]. Mailaparambil et al. showed the association between TLR10 haplotypes containing rs11096957, rs11096955, and rs4129009 and RSV infection [17]. It was also reported that rs4129009 was associated with asthma [14, 15]. However, in our study, these SNPs (rs11096957, rs11096955, and rs4129009) were not associated with IgAN.

In European population, minor allele is C allele in rs11096957 and also C allele in rs11096955 (dbSNP, BUILD 131). However, in our study, the minor alleles were A and A alleles. In dbSNP, the genotype frequencies of AA, AC, and CC genotypes of rs11096957 were 0.044, 0.511, and 0.444 in Chinese population and 0.111, 0.378, and 0.511 in Japanese population, respectively (dbSNP). In our study, they were 0.196 (AA), 0.498 (AC), and 0.339 (CC) in control subjects. The genotype frequencies of AA, AC, and CC of rs11096955 were 0.044, 0.511, and 0.444 in Chinese population and 0.089, 0.400, and 0.511 in Japanese population (dbSNP). In our study, they were 0.196 (AA), 0.498 (AC), and 0.339 (CC) in control subjects. Among Asian population, the genotype distributions of rs11096957 and rs11096955 may be different. However, in Asians, it is consistent that the minor allele of rs11096957 and rs11096955 are A and A alleles, respectively, different from the other ethnics such as Europeans and African Americans. Moreover, we found that rs11096957 and rs11096955 were in complete LD (D′ = 1, r 2 = 1.0). In Chinese population of dbSNP, these two SNPs were also in complete LD although the genotype distributions were different from our study (dbSNP). Additionally, Stevens et al. reported complete LD between two SNPs in their study using American Cancer Society Cancer Prevention II Nutrition Cohort [20]. However, in other populations, complete LD was not shown in dbSNP. There may be ethnic differences in rs11096957 and rs11096955. Although there were the ethnic differences on rs10004195 and rs4129009, the genotype frequencies of these SNPs in our data were similar to Japanese population. Further studies are needed to be applied to various ethnic groups.

Interestingly, in our study, rs1004195 of TLR10 gene was associated with IgAN, and the A allele of rs1004195 was more frequent in IgAN group, compared to control group. In addition, the proteinuria level of patients with genotype containing A allele of rs1004195 was higher than it of patients with genotype not containing A allele. Persistent proteinuria is the principal marker of kidney damage [13] as well as a predicting factor of progression of IgAN [6]. In a previous study, Chen et al. investigated the relationship between angiotensin-converting enzyme (ACE) gene insertion/deletion (I/D) polymorphism and clinicopathological manifestations in Chinese patients with IgAN and reported higher frequencies for the DD genotype of ACE gene in IgAN patients with heavy proteinuria, suggesting that the deletion polymorphism of the ACE gene played a role in IgAN deterioration and progression [4]. Recently, a significant correlation was shown between expression of TLR4 on circulating mononuclear cells (CD14-positive cells) and amount of proteinuria in patients with IgAN [7]. In a previous study, the association of CD14 gene-159C polymorphism with progression of IgAN was also reported [23]. Based on these findings, Coppo et al. suggested that TLR4 engagement in circulating mononuclear cells could play a role in the development of glomerular inflammation and a possible specific involvement in active and progressive IgAN [7]. Thus, our results provide the possibility that TLR10 gene and in particular A allele of rs1004195 may play a role in the progression of IgAN. Further studies about whether the alleles of rs1004195 could affect the change of TLR10 expression and whether TLR10 expression was related with the level of proteinuria in IgAN are needed.

In conclusion, we obtained the significant support for the association of TLR10 gene with IgAN in the Korean population. Also, the relationship was observed between TLR10 gene and the level of proteinuria of IgAN patients. Taken together, the results of the present study suggest that TLR10 gene may be involved in IgAN susceptibility.