Association of TLR2 Gene Polymorphisms With Ocular Behçet's Disease in a Chinese Han Population

Abstract

Purpose.: TLR2, TLR4, TLR8, and TLR9 have been reported to be associated with several autoimmune diseases. The current study aimed to explore whether singe nucleotide polymorphisms (SNPs) of these four genes were associated with ocular Behçet's disease (BD), Vogt-Koyanagi-Harada (VKH) syndrome, acute anterior uveitis (AAU) with or without ankylosing spondylitis (AS), or pediatric uveitis in Han Chinese.

Methods.: Genotyping was performed by PCR–restriction fragment length polymorphism. The first stage study comprised 400 ocular BD patients, 400 VKH syndrome patients, 400 AAU ± AS patients, 400 pediatric uveitis patients and 600 healthy subjects. The second stage included 438 ocular BD patients and 1000 healthy subjects. Allele and genotype frequencies were compared between patients and controls using the χ² test. Real-time PCR was used to detect mRNA expression from PBMCs obtained from healthy controls. Levels of TNF-α, IL-6, IL-10, and IL-1beta in culture supernatants were measured by ELISA.

Results.: In the first stage study, only the frequencies of the rs2289318/TLR2 genotype A and C allele and rs3804099/TLR2 genotype CT were significantly higher in ocular BD patients (P _c = 0.048; P _c = 0.008; P _c = 0.005, respectively) compared with controls. The second stage and combined studies confirmed the association (P _c = 0.001; P _c = 6.89E-06, P _c = 2.426E-06, respectively). TLR2 mRNA expression in PBMCs was increased in healthy carriers of the CC genotype of rs2289318/TLR2 and TT genotype of rs3804099/TLR2 following stimulation with peptidoglycan (PGN; P = 0.028; P = 0.004, respectively). No effect of the various TLR2 rs2289318 and rs3804099 genotypes on the release of TNF-α, IL-6, IL-10, and IL-1beta could be detected.

Conclusions.: This study provides evidence that the TLR2 gene is involved in the susceptibility to ocular BD.

Introduction

Uveitis is a serious intraocular inflammation that can lead to visual impairment. It can be caused by infectious or noninfectious mechanisms. ¹ Noninfectious uveitis entities can be mediated by autoinflammatory or autoimmune mechanisms, ² whereby infectious triggers may also play a role. ³  

Toll-like receptors (TLRs) are thought to be one of the links between infection and autoinflammatory or autoimmune disease. TLRs, together with retinoic acid-inducible gene 1–like receptors (RLRs) and nucleotide oligomerization domain–like receptors (NLRs), are members of the family of pattern recognition receptors (PRRs), which represent the main innate immune sensors. So far, 13 distinct mammalian TLRs have been identified, 10 of which are functional in humans (TLR1–10). ⁴ TLRs have been shown to play a role in the pathogenesis of various inflammatory diseases including uveitis. ^5,6 One of the approaches to study the role of TLRs is aimed at identifying disease associations with certain TLR gene polymorphisms. Abundant evidence is now available to show that several autoimmune diseases including rheumatoid arthritis (RA), systemic sclerosis (SS), Behçet's disease (BD), and systemic lupus erythematosus (SLE) are associated with certain TLR gene polymorphisms. ^7–10 The role of TLR genotypes in the pathogenesis of noninfectious uveitis has been studied in patients with anterior uveitis, Vogt-Koyanagi-Harada (VKH) syndrome, and sarcoid uveitis, but to date no significant associations have been reported. ^11–13 A positive association with TLR variants and Behçet's disease has been reported in some studies, ^9,14 but could not be seen in others. ^14–17 This may have been due to the small sample size used in these studies or to the fact that not all TLRs were studied. We repeated these studies and tested four different TLRs; and to obtain a sufficient sample size, we chose a number of frequently occurring noninfectious uveitis entities in a population of Chinese Han. 

We found a significant association with a polymorphism for the gene encoding TLR-2 in ocular BD patients, but not in the other uveitis entities studied such as VKH syndrome, acute anterior uveitis (AAU), AAU associated ankylosing spondylitis (AAU⁺AS⁺) or pediatric uveitis (PU). No association was found with polymorphisms for the genes encoding TLR4, TLR8, or TLR9. 

Materials and Methods

Patients and Healthy Controls Study Population

A two-stage case-control association study was performed. The first stage studied group comprised 400 ocular BD patients, 400 VKH syndrome patients, 400 AAU with or without AS patients, 400 PU patients and 600 healthy subjects who were referred to the Zhongshan Ophthalmic Center, Sun Yat-sen University, and the First Affiliated Hospital of Chongqing Medical University, Chongqing, China. The second stage included a total of 438 BD patients and 1000 healthy controls. All control subjects, which were collected from April 2005 to March 2013, were matched ethnically Han Chinese and geographically with the patients. The patients with ocular BD fulfilled the criteria of the International Study Group. ¹⁸ The clinical features were recorded during follow-up and summarized in Table 1. The diagnosis of VKH syndrome was based on the First International Workshop criteria for VKH syndrome. ¹⁹ The clinical characteristics of the patients are presented in Table 2. AAU patients were diagnosed principally according to clinical manifestations ^20,21 and the patients with AS fulfilled the modified New York criteria. ²² The clinical characteristics of the AAU patients with or without AS are presented in Table 3. Pediatric uveitis was defined as uveitis (including anterior uveitis, intermediate uveitis, posterior uveitis and panuveitis) first presenting at an age below 16 years. The clinical characteristics of this patient group are shown in Table 4. Pediatric patients with BD, VKH syndrome, or with definite infectious uveitis were excluded. To obtain a comparable genetic background, we strictly chose the cases from Chinese Han descendents. The study protocol was approved by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University, Chongqing, China (Permit Number: 2009-201008). All procedures followed the tenets of the Declaration of Helsinki. The patients and controls gave their written informed consent for this study. 

Table 1

 

Clinical Features of the Investigated Ocular BD Patients

Clinical Features	Patients With BD
	N = 838	%
Age at onset, y ± SD	31.1 ± 9.8
Uveitis	838	100
Oral ulcer	838	100
Genital ulcer	435	51.9
Skin lesions	552	65.9
Pathergy reaction	216	25.8
Hypopyon	191	22.8
Arthritis	140	16.7

Table 2

 

Clinical Features of the Investigated VKH Syndrome Patients

Clinical Features	Patients With VKH Syndrome
	N = 400	%
Age at onset, y ± SD	33.5 ± 8.9
Uveitis	400	100
Nuchal rigidity	52	13.0
Headache	204	51.0
Scalp allergy	40	10.0
Tinnitus	156	39.0
Alopecia	84	21.0
Gray hair	104	26.0
Vitiligo	100	25.0

Table 3

 

Clinical Features of the Investigated AAU ± AS Patients

Clinical Features	Patients With AAU Disease
	N = 400	%
Age at onset, y ± SD	39.2 ± 10.0
Age of AAU + AS at onset, y ± SD	38.9 ± 9.6
Age of AAU − AS at onset, y ± SD	39.5 ± 10.3
AAU + AS	209	52.3
AAU − AS	191	47.7
HLA-B27(+)	147	36.8
HLA-B27(−)	253	63.2

Table 4

 

Clinical Features of the Investigated PU Patients

Clinical Features	Patients With PU
	N = 400	%
Age at onset, y ± SD	14.4 ± 8.9
Uveitis	400	100
Juvenile idiopathic arthritis	29	7.3
Retinal vasculitis	28	7.0

SNP Selection and Genotyping

In view of previously reported TLR gene associations, ^9,10,23–26 we selected three SNPs (rs2289318, rs3804099, rs13150331) of TLR2; one SNP (rs7037117) of TLR4; one SNP (rs3764880) of TLR8; and three SNPs (rs187084, rs352139, rs352140) of TLR9 as our candidate SNPs to explore the possible association with BD, VKH syndrome, AAU with or without AS, and PU. Heparinized blood samples were taken from patients and controls and genomic DNA was isolated with a DNA purification kit (QIAamp DNA Blood Mini Kit; Qiagen, Valencia, CA). The extracted DNA was stored at −20°C until used. Samples were genotyped by PCR–restriction fragment length polymorphism (PCR-RFLP) for rs2289318, rs3804099, rs13150331, rs7037117, rs187084, rs352139, rs352140, or by PCR assay (TaqMan SNP Genotyping Assay; Applied Biosystems, Foster City, CA) for rs3764880. The genotyping success rate ranged between 97.9% to 98.4% for the various groups. We randomly selected 5% of the samples to undergo direct sequencing to validate the results of genotyping by PCR-RFLP and PCR assay (Life Technologies, Beijing, China). The primers of rs2289318, rs3804099, rs13150331, rs7037117, rs187084, rs352139, and rs352140 used to amplify target DNA sequences by PCR are shown in Supplementary Table S1. Digestion products were visualized on a 5% agarose gel (rs7037117, rs352139) or 4% agarose gel (rs2289318, rs3804099, rs13150331, rs187084, rs352140) and stained with a nucleic acid stain (GoldView; SBS Genetech, Beijing, China). Restriction enzymes were purchased from Thermo Fisher Scientific. Direct sequencing was performed by Majorbio Biotechnology Company (Shanghai, China) using randomly selected subjects (5% of all samples) to validate the result of the PCR-RFLP method used in our study. For rs3764880 (TaqMan assay ID: C_2183830_10), the genotype was determined by PCR assay (Applied Biosystems) on real-time PCR equipment (7500 Fast Real-Time PCR System; Applied Biosystems), according to the supplier manual. Negative controls were included in each plate. Genotype analysis was performed using a genotyping data analysis tool (TaqMan Genotyper Software; Life Technologies). 

mRNA Expression PCR

Peripheral blood mononuclear cells (PBMCs) obtained from healthy individuals were prepared from heparinized blood by Ficoll-Hypaque density-gradient centrifugation. Total RNA was extracted from PBMCs with or without stimulation by peptidoglycan (PGN, 5 μg/mL, Fluka, Buchs, Switzerland) for 72 hours at a density of 1 × 10⁶ cells/mL, using a commercial reagent (TRIzol; Life Technologies), followed by reverse transcription using a transcriptase kit (Takara Biosystems). Real-time Quantitative PCR was performed to compare the mRNA expression of the TLR2 gene (sense primer: 5′ GGAGGCTGCATATTCCAAGG 3′, antisense primer: 5′ GCCAGGCATCCTCACAGG 3′), using real-time PCR equipment with a commercial dye kit (SYBR Green I Assay kit; Applied Biosystems). Data were normalized to mRNA β-actin as reported by previous studies. ^21,22 The assays were performed on real-time PCR equipment (Applied Biosystems). Relative expression levels were calculated using the 2^−ΔΔCt method. 

Cytokine Measurements

PBMCs obtained from healthy individuals were stimulated with PGN (5 μg/mL,) or anti-CD3 (5 μg/mL; eBioscience, San Diego, CA) plus anti-CD28 (5 μg/mL; eBioscience) plus PGN (5 μg/mL) for 72 hours at a density of 1 × 10⁶ cells/mL. The concentration of TNF-α, IL-6, IL-10, and IL-1β in cell culture supernatants was measured with a commercial ELISA kit (DuoSet ELISA Development Systems; R&D Systems, Minneapolis, MN). 

Statistical Analysis

Hardy-Weinberg equilibrium was tested using the χ² test. The patterns of linkage disequilibrium (LD) of the SNPs (rs2289318, rs3804099, rs13150331) of TLR2 (rs187084, rs352139, rs352140) and of TLR9 were compared using haplotype analysis software (Haploview 4.0; Broad Institute, Cambridge, MA), respectively. Genotype frequencies of all SNPs were estimated by direct counting except one SNP (rs3764880) of TLR8. Allele and genotype frequencies were compared between patients and controls by the χ² test using statistical software (SPSS, version 17.0; SPSS, Inc., Chicago, IL). As TLR8 is located on the X chromosome, the allele and genotype frequency of the SNP (rs3764880) of TLR8 was divided into two alleles and the frequencies were calculated according to sex. P values were corrected (P _c) for multiple comparisons with the Bonferroni correction by multiplying with the number of analyses performed. The number of independent comparisons is 24. 

Results

Association of the Eight Gene Polymorphisms With Susceptibility to Ocular Behçet's Disease

A total of 400 BD patients and 600 healthy controls were genotyped for seven TLR SNPs: three SNPs (rs2289318, rs3804099, rs13150331) of TLR2; one SNP (rs7037117) of TLR4; and three SNPs (rs187084, rs352139, rs352140) of TLR9. A total of 400 BD patients (male/female = 225:175) and 600 (male/female = 300:300) healthy controls were genotyped for one SNP (rs3764880) of TLR8. These eight SNPs were successfully genotyped and conformed to Hardy-Weinberg expectation in controls. Frequencies of the rs2289318/TLR2 genotype CC and C allele and rs3804099/TLR2 genotype CT were significantly higher in BD patients (P _c = 0.048, OR = 1.537, 95% confidence interval [CI] 1.175–2.011; P _c = 0.008, OR = 1.489, 95% CI 1.180–1.878; P _c = 0.005, OR = 0.696, 95% CI 0.537–0.902, respectively) compared with controls. The P values of the rs2289318/TLR2 genotype GG or CG and rs3804099/TLR2 genotype TT lost significance following Bonferroni correction. After extending the number of BD cases (n = 838) and healthy controls (n = 1600), we confirmed the association with the rs2289318/TLR2 genotype CC and C allele and rs3804099/TLR2 genotype CT (P _c = 0.001, OR = 1.462, 95% CI 1.223–1.747; P _c = 6.89E-06, OR = 1.470, 95% CI 1.260–1.714; P _c = 2.426E-06, OR = 0.626, 95% CI 0.526–0.744, respectively) and found that the corrected P value for rs2289318 genotype GG and rs3804099 genotype CC and TT were significant (P _c = 0.001, OR = 0.363, 95% CI 0.220–0.601; P _c = 0.024, OR = 1.709, 95% CI 1.227–2.381; P _c = 0.005, OR = 1.375, 95% CI 1.162–1.627, respectively; Table 5). No significant differences were found between BD patients and controls concerning the frequencies of the other six SNPs (Table 5). 

Table 5

 

Polymorphisms of TLR2, TLR4, TLR8, and TLR9 Genes in Ocular Behçet's Disease

SNPs	Stage	Genotype Allele	Case, n (freq)	Control, n (freq)	P Value	P c Value	OR (95% CI)
TLR2 (rs2289318)	First	CC	279 (0.698)	360 (0.600)	0.002	0.048	1.537 (1.175–2.011)
		CG	113 (0.283)	210 (0.350)	0.025	0.6	0.731 (0.556–0.962)
		GG	8 (0.020)	30 (0.050)	0.015	0.36	0.388 (0.176–0.855)
		C	671 (0.839)	933 (0.778)	0.001	0.008	1.489 (1.180–1.878)
		G	129 (0.161)	267 (0.223)	0.001	0.008	0.672 (0.532–0.848)
	Replication	CC	304 (0.694)	616 (0.616)	0.005	0.03	1.414 (1.113–1.798)
		CG	123 (0.281)	324 (0.324)	0.103	0.618	0.815 (0.636–1.043)
		GG	11 (0.025)	60 (0.060)	0.005	0.03	0.404 (0.210–0.775)
		C	731 (0.834)	1556 (0.778)	0.001	0.002	1.439 (1.170–1.769)
		G	145 (0.166)	444 (0.222)	0.001	0.002	0.695 (0.565–0.855)
	Combined	CC	583 (0.696)	976 (0.610)	2.84E-05	0.001	1.462 (1.223–1.747)
		CG	236 (0.282)	534 (0.334)	0.009	0.216	0.783 (0.652–0.940)
		GG	19 (0.023)	90 (0.056)	4.04E-05	0.001	0.363 (0.220–0.601)
		C	1402 (0.837)	2486 (0.777)	8.62E-07	6.89E-06	1.470 (1.260–1.714)
		G	274 (0.163)	714 (0.223)	8.62E-07	6.89E-06	0.680 (0.583–0.794)
TLR2 (rs3804099) (TLR2/+596)	First	CC	32 (0.080)	31 (0.052)	0.071	1.704	1.596 (0.957–2.661)
		CT	135 (0.338)	273 (0.455)	2.12E-04	0.005	0.696 (0.537–0.902)
		TT	233 (0.582)	296 (0.493)	0.006	0.144	1.433 (1.110–1.849)
		C	199 (0.249)	335 (0.279)	0.132	1.056	0.855 (0.697–1.048)
		T	601 (0.751)	865 (0.721)	0.132	1.056	1.170 (0.954–1.434)
	Replication	CC	38 (0.087)	50 (0.050)	0.007	0.042	1.805 (1.165–2.796)
		CT	154 (0.352)	458 (0.458)	1.73E-04	0.001	0.642 (0.509–0.809)
		TT	246 (0.561)	492 (0.492)	0.015	0.090	1.323 (1.056–1.658)
		C	230 (0.263)	558 (0.279)	0.363	0.726	0.920 (0.769–1.101)
		T	646 (0.737)	1442 (0.721)	0.363	0.726	1.087 (0.908–1.300)
	Combined	CC	70 (0.084)	81 (0.051)	0.001	0.024	1.709 (1.227–2.381)
		CT	289 (0.345)	731 (0.457)	1.01E-07	2.426E-06	0.626 (0.526–0.744)
		TT	479 (0.571)	788 (0.492)	2.05E-04	0.005	1.375 (1.162–1.627)
		C	429 (0.256)	893 (0.279)	0.085	0.68	0.889 (0.777–1.016)
		T	1247 (0.744)	2307 (0.721)	0.085	0.68	1.125 (0.984–1.287)
TLR2 (rs13150331) (TLR2/-24438)	First	AA	102 (0.255)	157 (0.262)	0.814	19.536	0.966 (0.723–1.290)
		AG	267 (0.668)	397 (0.662)	0.848	20.352	1.027 (0.785–1.342)
		GG	31 (0.078)	46 (0.077)	0.961	23.064	1.012 (0.630–1.626)
		A	471 (0.589)	711 (0.593)	0.867	5.202	0.985 (0.821–1.181)
		G	329 (0.411)	489 (0.407)	0.867	5.202	1.016 (0.847–1.218)
TLR4 (rs7037117)	First	AA	256 (0.64)	368 (0.613)	0.394	9.456	1.121 (0.862–1.457)
		AG	128 (0.32)	206 (0.343)	0.443	10.632	0.900 (0.688–1.178)
		GG	16 (0.04)	26 (0.043)	0.797	19.128	0.920 (0.487–1.738)
		A	640 (0.8)	942 (0.785)	0.419	3.352	1.096 (0.878–1.367)
		G	160 (0.2)	258 (0.215)	0.419	3.352	0.913 (0.732–1.139)
TLR9 (rs187084)	First	CC	61 (0.153)	86 (0.143)	0.688	16.512	1.075 (0.754–1.535)
		CT	196 (0.49)	306 (0.51)	0.535	12.84	0.923 (0.717–1.189)
		TT	143 (0.358)	208 (0.347)	0.725	17.4	1.049 (0.805–1.366)
		C	318 (0.398)	478 (0.398)	0.97	7.76	0.997 (0.830–1.196)
		T	482 (0.603)	722 (0.602)	0.97	7.76	1.003 (0.836–1.205)
TLR9 (rs352139)	First	AA	145 (0.363)	211 (0.352)	0.726	17.424	1.048 (0.805–1.365)
		AG	193 (0.483)	314 (0.523)	0.206	4.944	0.849 (0.659–1.094)
		GG	62 (0.155)	75 (0.125)	0.176	4.224	1.284 (0.893–1.846)
		A	483 (0.604)	736 (0.613)	0.667	5.336	0.961 (0.800–1.154)
		G	317 (0.396)	464 (0.387)	0.667	5.336	1.041 (0.867–1.250)
TLR9 (rs352140)	First	CC	140 (0.350)	216 (0.360)	0.746	17.904	0.957 (0.735–1.247)
		CT	211 (0.528)	312 (0.520)	0.816	19.584	1.031 (0.800–1.328)
		TT	49 (0.123)	72 (0.120)	0.905	21.72	1.024 (0.695–1.508)
		C	491 (0.614)	744 (0.620)	0.778	6.224	0.974 (0.810–1.171)
		T	309 (0.386)	456 (0.380)	0.778	6.224	1.027 (0.854–1.234)
TLR8 (rs3764880) male	First	0A	40 (0.178)	50 (0.167)	0.738	17.712	1.081 (0.684–1.708)
		0G	185 (0.822)	250 (0.833)	0.738	17.712	0.925 (0.586–1.461)
TLR8 (rs3764880) female	First	AA	8 (0.046)	20 (0.067)	0.35	8.4	0.671 (0.289–1.557)
		AG	46 (0.263)	77 (0.257)	0.882	21.168	1.033 (0.675–1.579)
		GG	121 (0.691)	203 (0.677)	0.739	17.736	1.071 (0.716–1.600)
		A	62 (0.177)	117 (0.195)	0.497	3.976	0.889 (0.632–1.249)
		G	288 (0.823)	483 (0.805)	0.497	3.976	1.125 (0.800–1.582)

We also studied the association of polymorphisms of the four genes with the clinical features of the BD patients, using stratified analysis. A significantly higher frequencies of CC genotype and C allele of rs2289318 and rs3804099 were noticed in BD patients with genital ulcers (Table 6). A significantly higher frequency of CT, TT genotypes, and C allele of rs3804099 was also noticed in BD patients with pathergy reaction (Table 6). No significant statistical association was found for the other six SNP genotypes and clinical manifestations in BD. 

Table 6

 

Polymorphisms of the TLR2 Gene in Patients With Extraocular Features of BD Versus Healthy Controls

SNP	Clinical Features	Genotype	BD Patients, n (%)	Controls, n (%)	P Value	P c	OR (95% CI)
rs2289318	Genital ulcer	CC	299 (0.687)	976 (0.610)	0.003	0.045	1.406 (1.121–1.762)
		CG	121 (0.278)	534 (0.334)	0.028	0.140	0.769 (0.609–0.972)
		GG	15 (0.034)	90 (0.056)	0.399	1.995	0.786 (0.449–1.376)
		C	719 (0.826)	2486 (0.777)	0.002	0.030	1.368 (1.126–1.661)
		G	151 (0.174)	714 (0.223)	0.002	0.030	0.731 (0.602–0.888)
	Skin lesions	CC	355 (0.643)	976 (0.610)	0.167	0.835	1.152 (0.942–1.409)
		CG	181 (0.328)	534 (0.334)	0.801	4.005	0.974 (0.793–1.197)
		GG	16 (0.029)	90 (0.056)	0.011	0.055	0.501 (0.292–0.860)
		C	891 (0.807)	2486 (0.777)	0.035	0.175	1.201 (1.012–1.426)
		G	213 (0.193)	714 (0.223)	0.035	0.175	0.832 (0.701–0.988)
	Pathergy reaction	CC	141 (0.653)	976 (0.610)	0.225	1.125	1.202 (0.893–1.619)
		CG	65 (0.301)	534 (0.334)	0.335	1.675	0.859 (0.631–1.170)
		GG	10 (0.046)	90 (0.056)	0.547	2.735	0.814 (0.417–1.591)
		C	347 (0.803)	2486 (0.777)	0.214	1.070	1.172 (0.912–1.508)
		G	85 (0.197)	714 (0.223)	0.214	1.070	0.853 (0.663–1.097)
	Hypopyon	CC	122 (0.639)	976 (0.610)	0.441	2.205	1.130 (0.828–1.544)
		CG	58 (0.304)	534 (0.334)	0.403	2.015	0.871 (0.629–1.205)
		GG	11 (0.058)	90 (0.056)	0.939	4.695	1.025 (0.538–1.954)
		C	302 (0.791)	2486 (0.777)	0.542	2.710	1.084 (0.836–1.406)
		G	80 (0.209)	714 (0.223)	0.542	2.710	0.922 (0.711–1.196)
	Arthritis	CC	84 (0.600)	976 (0.610)	0.816	4.080	0.959 (0.674–1.365)
		CG	48 (0.343)	534 (0.334)	0.827	4.135	1.042 (0.724–1.499)
		GG	8 (0.057)	90 (0.056)	0.965	4.825	1.017 (0.483–2.141)
		C	216 (0.771)	2486 (0.777)	0.834	4.170	0.969 (0.725–1.297)
		G	64 (0.229)	714 (0.223)	0.834	4.170	1.032 (0.771–1.380)
rs3804099	Genital ulcer	CC	40 (0.092)	81 (0.051)	1.00E-03	0.015	1.899 (1.280–2.818)
		CT	215 (0.494)	731 (0.457)	0.166	2.490	1.162 (0.940–1.436)
		TT	180 (0.414)	788 (0.492)	4.00E-03	0.060	0.727 (0.587–0.901)
		C	295 (0.339)	893 (0.279)	1.00E-03	0.01	1.325 (1.129–1.556)
		T	575 (0.661)	2307 (0.721)	1.00E-03	0.01	0.754 (0.643–0.886)
	Skin lesions	CC	41 (0.074)	81 (0.051)	0.038	0.570	1.505 (1.020–2.220)
		CT	270 (0.489)	731 (0.457)	0.190	2.850	1.138 (0.938–1.381)
		TT	241 (0.437)	788 (0.492)	0.023	0.345	0.799 (0.657–0.970)
		C	352 (0.319)	893 (0.279)	0.012	0.06	1.209 (1.043–1.403)
		T	752 (0.681)	2307 (0.721)	0.012	0.06	0.827 (0.713–0.959)
	Pathergy reaction	CC	15 (0.069)	81 (0.051)	0.246	3.690	1.399 (0.791–2.475)
		CT	123 (0.569)	731 (0.457)	2.00E-03	0.030	1.572 (1.180–2.094)
		TT	78 (0.362)	788 (0.492)	2.85E-04	0.004	0.582 (0.434–0.782)
		C	153 (0.354)	893 (0.279)	1.00E-03	0.01	1.417 (1.146–1.751)
		T	279 (0.646)	2307 (0.721)	1.00E-03	0.01	0.706 (0.571–0.872)
	Hypopyon	CC	11 (0.079)	81 (0.051)	0.680	10.200	1.146 (0.599–2.192)
		CT	91 (0.476)	731 (0.457)	0.608	9.120	1.082 (0.801–1.461)
		TT	89 (0.466)	788 (0.492)	0.488	7.320	0.899 (0.666–1.215)
		C	113 (0.296)	893 (0.279)	0.491	2.46	1.085 (0.860–1.370)
		T	269 (0.704)	2307 (0.721)	0.491	2.46	0.921 (0.730–1.163)
	Arthritis	CC	8 (0.057)	81 (0.051)	0.737	11.055	1.137 (0.538–2.401)
		CT	61 (0.436)	731 (0.457)	0.630	9.450	0.918 (0.648–1.300)
		TT	71 (0.507)	788 (0.492)	0.740	11.100	1.060 (0.751–1.498)
		C	77 (0.275)	893 (0.279)	0.884	4.42	0.980 (0.745–1.288)
		T	203 (0.725)	2307 (0.721)	0.884	4.42	1.020 (0.776–1.341)

Association of the Eight Gene Polymorphisms With Susceptibility to VKH Syndrome, AAU With or Without Ankylosing Spondylitis or Pediatric Uveitis

A total of 400 VKH patients, 400 AAU with or without AS patients, 400 PU patients, and 600 healthy controls were genotyped for seven SNPs: three SNPs (rs2289318, rs3804099, rs13150331) of TLR2; one SNP (rs7037117) of TLR4; and three SNPs (rs187084, rs352139, rs352140) of TLR9. A total of 400 VKH patients (male/female = 214:186), 400 AAU ± AS patients (male/female = 193:207), 400 PU patients (male/female = 199:201), and 600 (male/female = 300:300) healthy controls were genotyped for one SNP (rs3764880) of TLR8. The eight SNPs were successfully genotyped and conformed to Hardy-Weinberg expectation in controls. The genotype and allele frequencies of the eight SNPs examined in VKH syndrome, AAU with or without AS, or PU patients and healthy controls are summarized in Supplementary Tables S2, S3, and S4, respectively. 

There were no statistically significant differences concerning the genotype or allele frequencies of the eight SNPs between VKH syndrome, AAU with or without AS, PU, and healthy controls following Bonferroni correction. 

Relationship Between Two SNPs (rs2289318, rs3804009) of TLR2 Genotypes and the Corresponding Gene Expression at the mRNA Level

Since we found that polymorphisms of TLR2 showed an association with susceptibility to ocular BD, we investigated whether the different genotypes studied affect the expression of TLR2 under normal or inflammatory conditions. Real-time PCR was performed to detect mRNA expression from PBMCs obtained from healthy controls. We genotyped healthy individuals for the TLR2 SNPs (rs2289318, rs3804099) and then used real-time PCR to detect TLR2 mRNA expression by PBMCs with or without stimulation. No difference in gene expression could be detected between the various genotypes when PBMCs were not stimulated (Figs. 1, 2). Following stimulation by PGN, carriers with the CC genotype in SNP rs2289318 had a higher TLR2 mRNA expression compared with individuals carrying the CG genotype (P = 0.028; Fig. 3) and carriers with the TT genotype in SNP rs3804099 had a higher TLR2 mRNA expression compared with individuals carrying the CC or CT genotype (P = 0.004; P = 0.02, respectively; Fig. 4). No significant difference was observed between CC and GG carriers of rs2289318 or CC and CT carriers of rs3804099, but this could be due to the fact that the numbers of individuals carrying the GG genotype of rs2289318 and the CC genotype of rs3804099 were very low (n = 4; n = 5, respectively). 

Figure 1

 

mRNA expression of different genotypes of rs2289318/TLR2 by nonstimulated PBMCs. TLR2 expression was not significantly different between genotypes. Data are expressed as mean ± SD.

Figure 2

 

mRNA expression of different genotypes of rs3804099/TLR2 by nonstimulated PBMCs. TLR2 expression was not significantly different between genotypes. Data are expressed as mean ± SD.

Figure 3

 

mRNA expression of different genotypes of rs2289318/TLR2 by PGN stimulated PBMCs. TLR2 expression in carriers of the CC genotype was significantly higher than in individuals carrying the CG genotype (P = 0.028). Data are expressed as mean ± SD.

Figure 4

 

mRNA expression of different genotypes of rs3804099/TLR2 by PGN stimulated PBMCs. TLR2 expression in carriers of the TT genotype was significantly higher than in individuals carrying the CC or CT genotypes (P = 0.004; P = 0.02, respectively). Data are expressed as mean ± SD.

Relationship Between the Two SNPs (rs2289318, rs3804099) of TLR2 Genotypes and Downstream Inflammatory Factors

The aforementioned result showed that different genotypes of rs2289318 and rs3804099 could affect TLR2 expression and therefore a further study was designed to investigate if different genotypes of rs2289318 or rs3804099 could also affect the cytokine response of PBMCs following PGN stimulation. We measured the production of TNF-α, IL-6, IL-10, and IL-1β, which are important TLR2 downstream factors ^27–29 in PBMC culture supernatants by ELISA. However, no effect of the various TLR2 rs2289318 or rs3804099 genotypes on the release of these four cytokines could be detected (data not shown). 

Discussion

In this study, we investigated the association of gene polymorphisms of TLRs 2, 4, 8, and 9 with the risk of developing ocular BD, VKH syndrome, AAU with or without AS, and PU in a Han Chinese population and found an association between the TLR2 genes rs2289318 and rs3804099 with susceptibility to ocular BD stratification analysis also showed an association between certain clinical features of ocular BD with the tested SNPs of TLR2 (rs2289318, rs3804099). In view of the association with TLR2, we performed further investigations into a possible functional association. This approach revealed that healthy carriers of the CC genotype in SNP rs2289318 and TT genotype in rs3804099 had a higher TLR2 mRNA expression level following stimulation by PGN, compared with individuals carrying the other genotypes. Our study is the first to report a TLR2 gene association with ocular BD and supports the absence of an association in other noninfectious uveitis entities. Earlier studies by Wakefield's group did not show a correlation between AAU and TLR2 and TLR4 genotypes. ³⁰ Another Japanese study whereby only the TLR4 gene was investigated did not find an association with sarcoid uveitis. ¹² A small study in Japanese VKH patients (n = 94) that was restricted to the TLR9 gene also did not observe a significant association. ¹³  

BD is one of the most commonly seen uveitis entities in China. ³¹ The exact etiology of ocular BD is still unknown. It is becoming clear, however, that both genetic ^32,33 as well as environmental factors play a role in the development of this disease. ³⁴ A potential relationship between bacterial infection (i.e., Streptococcus sanguinis ) and BD is based on findings of bacterial components and antibodies in the oral flora and serum of patients with BD. ³⁵ More data is becoming available concerning the genetic background of BD. In the recent years, we have reported on the association of various immune response related genes and susceptibility to BD, including signal transducer and activator of transcription (STAT)-4, STAT3, Janus kinase 2, and CD40. ^36–39  

An association between the TLR2 SNPs rs3804099 and rs2289318 as we observed in ocular BD was recently also reported in a French cohort of asthma patients. Though the significance of the rs2289318 SNP lost significance after adjusting for multiple comparisons, ²³ the association of rs3804099/TLR2 with asthma remained. We did not find an association of the two TLR2 SNPs with susceptibility to VKH, AAU with or without AS, and PU. The discrepancy between ocular BD and the other uveitis entities may be due to the fact that the pathogenetic mechanisms for BD are unique and differ from these other three intraocular inflammatory diseases. Evidence is mounting that BD is an autoinflammatory disease caused by an aberrant response against infectious agents, which might also explain a role for TLRs. ⁴⁰ We also studied the association of polymorphisms of rs2289318 and rs3804099 with the clinical features of BD patients, using stratified analysis. Significantly higher frequencies of the CC genotype and C allele of rs2289318 and rs3804099 were noticed in BD patients with genital ulcers (Table 6). A significantly higher frequency of the CT, TT genotypes, and C allele of rs3804099 was also noticed in BD patients with the pathergy reaction (Table 6). The reason for an association between the frequency of genotypes of rs2289318 or rs3804099 and BD patients with genital ulcers and (or) the pathergy reaction is not clear. Based on the above data, there is a possibility that the TLR2 associations with the BD cohorts we described are independent of uveitis. Our BD patients all have uveitis and future studies investigating the role of TLR2 in BD patients without uveitis may shed more light on this issue. 

During recent years, three other TLRs (TLR4, TLR8, and TLR9) have emerged as candidate susceptibility factors for a number of immune diseases. A role for TLR4 has been addressed in Japanese and Korean BD patients. ^9,41 Similarly, another study revealed an association between TLR9 and SLE in Chinese and Japanese patients. ^10,26 SNP (rs187084) of the TLR9 gene was shown to affect the risk of BD in Tunisian patients. We did not find an association between polymorphisms in TLR4, TLR8, and TLR9 genes in the uveitis entities included in our study. Similar to our results, no association with these three TLR related gene polymorphisms was reported in JIA (juvenile idiopathic arthritis) patients from the United Kingdom. ⁴²  

Because the two TLR2 SNPs (rs2289318, rs3804099) were shown to be associated with ocular BD, we investigated whether the different genotypes could affect its expression. Our data showed that when we stimulated PBMCs from healthy individuals with PGN, there was an increased expression of TLR2 mRNA in individuals with the rs2289318CC genotype compared with those carrying the rs2289318CG genotype. An increased expression of TLR2 mRNA in individuals with the rs3804099 TT genotype compared with those carrying the rs3804099 CC or CT genotypes was also observed when we stimulated PBMCs with PGN. We did not yet test TLR2 expression in genotyped BD patients since the degree of inflammation as well as the immunosuppressive treatment may introduce a confounding effect on gene expression. ⁴³ We recently showed that active BD patients not receiving immunosuppression had a markedly higher expression at the mRNA and protein level of TLR2, TLR3, TLR4, and TLR8 compared with healthy controls. ⁴⁴ In the current study, we did not measure TLR expression in our patients and did not observe a significant difference of the various TLR2 rs2289318 or rs3804099 genotypes on the release of the TLR downstream cytokines such as TNF-α, IL-6, IL-10, and IL-1β. These findings suggest that the mechanism explaining the role of the TLR2 polymorphisms with BD is related to the quantitative TLR2 expression which will in turn lead to a higher cytokine response. More studies on the TLR2 and cytokine response in genotyped BD patients are needed to support this hypothesis. The fact that successful anti-TNFα treatment of BD patients leads to a lower TLR2 gene expression is further evidence for a role of TLR2 in mediating the ocular inflammatory response in this disease. ⁴³  

Several authors investigated the association between polymorphisms of TLRs and their functional involvement in autoimmune or autoinflammatory diseases. Systemic sclerosis patients carrying a rare TLR2 Pro-631 His variant from a European population showed a marked increase in the production of IL-6 and TNF-α by DCs following stimulation with PGN. ⁸ However, no association was reported between TLR4 expression on PBMCs and carriers of the TLR4 Asp299Gly or Thr399Ile genotypes in rheumatoid arthritis. ⁴⁵ Similarly, no association was found between autoantibody production and rs187084/TLR9 polymorphisms in Turkish rheumatoid arthritis patients. ⁴⁶  

How TLR2 exactly affects the predisposition to ocular BD in our Chinese patients is not yet clear. It should be noted that activation of TLRs is dependent on the interaction with its ligands and that the control of the expression of these ligands in BD may depend on as yet unknown factors. Besides TLRs, many other immune related genes are also involved in the proinflammatory pathways that will finally result in the expression of clinical uveitis in BD and further studies are needed to evaluate the relative contribution of each of these factors. 

Our study was performed in Han Chinese and confirmatory studies are needed to show whether the same association can be reproduced in other ethnic populations and further functional and linkage studies are required to investigate the exact role of TLR2 gene polymorphisms in ocular BD pathogenesis. The fact that we failed to find an association with TLR4, 8, and 9 genes does not rule out the possibility that other SNPs in TLR genes can be associated with uveitis. We only studied a few uveitis entities and it is possible that TLR gene associations are present in other types of intraocular inflammation. Furthermore it is not clear whether our observed TLR2 association with ocular BD is also present in BD patients seen at other medical departments. 

Taken together, our study, for the first time, reported an association between the polymorphism of two SNPs (rs2289318 and rs3804099) of TLR2 with susceptibility to ocular BD in a Chinese Han population. 

Supplementary Materials

pdf S1 

Acknowledgments

The authors thank all donors enrolled in the present study. 

Supported by National Basic Research Program of China (973 Program; 2011CB510200); Key Project of Natural Science Foundation (81130019); Research Fund for the Doctoral Program of Higher Education of China (20115503110002); Clinic Key Project of Ministry of Health, Basic Research Program of Chongqing, Chongqing Key Laboratory of Ophthalmology (CSTC, 2008CA5003); Basic Research Program of Chongqing (CSTC, 2013JCYJC10001); Key Project of Health Bureau of Chongqing (2012-1-003); and Fund for PAR-EU Scholars Program. The authors alone are responsible for the content and writing of the paper. 

Disclosure: J. Fang, None; R. Hu, None; S. Hou, None; Z. Ye, None; Q. Xiang, None; J. Qi, None; Y. Zhou, None; A. Kijlstra, None; P. Yang, None 

Footnotes

 JF and RH contributed equally to the work presented here and should therefore be regarded as equivalent authors.