Introduction

Uveitis is an umbrella term for inflammatory diseases involving the uveal tract, retina, and vitreous, which can lead to significant visual impairment.1 Pediatric uveitis (PU) accounts for 2.2–13.8% of all uveitis cases and has a high risk of developing band keratopathy, cataract, secondary glaucoma, amblyopia, and other ocular complications.2 Almost all uveitis entities occurring in adults may also be seen in children. However, the most common subtype is pediatric idiopathic uveitis (PIU), and the second is juvenile idiopathic arthritis (JIA)-associated uveitis.3 Although the etiology of uveitis in children is not completely clarified, aberrant immune reactivity and genetic predisposition have been shown to be involved in its pathogenesis.4

Apoptosis is a mechanism that precisely eliminates unwanted cells and is essential for normal physiological transformation of cells as well as in various pathological processes.5 Previous studies have shown that the nuclear factor-kB signaling pathway is closely related to apoptosis and is involved in the transcriptional regulation of various apoptosis-related genes, which is an important pathway regulating immunity.6,7 There are four major genes in this pathway, including programmed cell death 1 (PDCD1), PDCD1LG2, FAS, and FASLG.8,9,10 Aberrant expression of these genes has been shown to be involved in certain autoimmune diseases.11,12,13,14 Polymorphisms of these genes have also been shown to be associated with a variety of autoimmune diseases, such as rheumatoid arthritis (RA),15 Graves’ disease,16 ankylosing spondylitis (AS),17 and systemic lupus erythematosus (SLE).18 Furthermore, their polymorphisms have also been shown to be associated with certain cancers,19,20 which might be related to its role in the process of cancer immunology. Whether these polymorphisms are also associated with PIU is not known and was therefore the subject of the present study.

Materials and methods

Study population

A total of 1238 patients with PIU, 128 JIA-associated uveitis patients, and 1114 healthy controls were included in this study. All the enrolled patients and healthy controls were Han Chinese. PIU was identified as uveitis in children whose age at onset was <16 years and not belonging to a well-defined uveitis entity. Patients with definite posterior involvement or with panuveitis were excluded from the study. These PIU patients all showed anterior uveitis evidenced by keratic precipitates, aqueous flare, and cells. Vitreous cells and opacities were observed in 148 cases. There was no obvious fundus abnormalities in our patients. Children with infectious uveitis entities and specific entities such as Vogt–Koyanagi–Harada syndrome diagnosed according to both sets of criteria presumed by the International Committee on Vogt–Koyanagi–Harada Disease Nomenclature21 and Yang et al.,22 Behcet’s disease (BD) diagnosed by the diagnostic criteria of the International Study Group for BD,23 Fuchs syndrome, or other defined uveitis types were also excluded from this study. According to the International Union of Rheumatology Association standards, JIA was defined as arthritis, without any other identifiable etiology in children under the age of 16 years, and existing for at least 6 weeks.24 All JIA patients were diagnosed by Rheumatologists and uveitis was diagnosed at the uveitis center of our hospital. All uveitis patients were seen by the same senior uveitis expert (P.Y.). Blood samples of all controls and patients were obtained from the uveitis center of the Department of Ophthalmology of the First Affiliated Hospital of Chongqing Medical University (Chongqing, China from May 2009 to May 2018) and the Department of Ophthalmology of the First Affiliated Hospital of Zhengzhou University (Zhengzhou, China from May 2017 to May 2018).

Ethical considerations

The study was approved by the Medical Ethics Committee of The First Affiliated Hospital of Zhengzhou University and the Clinical Research Ethics Committee of the First Affiliated Hospital of Chongqing Medical University, and each participant offered their informed consent before enrolment in the study. For minors, the informed consent was obtained from their parents or guardians. All procedures in this project followed the principles of the Helsinki Declaration.

Single-nucleotide polymorphism (SNP) choice and genotypes

Fourteen SNPs were selected from PDCD1, PDCD1LG2, FAS, and FASLG genes as candidates in this study, which were earlier shown to be significantly (P < 0.05) associated with a variety of autoimmune diseases or certain cancers.15,17,18,19,20,25,26,27,28 Candidate SNPs were screened on HaploView 4.2 with an r2 threshold of 0.8 and a minor allele frequency >0.05. The data of linkage disequilibrium from the Han Chinese Hap Map database were also taken into account. After using these criteria, the following SNPs were chosen for the study: four SNPs (rs7421861, rs2227982, rs6710479, rs7565639) of PDCD1, three SNPs (rs12001295, rs7852996, rs16923189) of PDCD1LG2, three SNPs (rs2234978, rs2234767, rs1468063) of FAS, and four SNPs (rs763110, rs5030772, rs859637, rs9286879) of FASLG.

The extraction of DNA and genotyping

Genomic DNA was obtained from peripheral blood using the QIAamp DNA Blood Mini Kit (QIAGEN, Valencia, CA). DNA was stored at −80 °C. The primers used to genotype were made according to the MassARRAY Assay software. Genotyping of these 14 selected SNPs were ascertained by iPLEX Gold Assay and the MassARRAY platform (Sequenom, San Diego, CA) using standard procedures. Data were analyzed by the MassARRAY Typer software (version 4.0).

Statistical analysis

The Hardy–Weinberg equilibrium of all candidate SNPs was calculated by χ2 test. Differences in allele and genotype frequencies were assessed by χ2 test or Fisher’s exact correction using the SPSS software (version 21.0, Chicago, IL), as well as calculation of odds ratios (ORs) and 95% confidence intervals. The P value was corrected by the Bonferroni method for multiple comparisons (Pc). Pc < 0.05 was defined as statistically significant.

Results

Clinical features

The clinical characteristics, family history of autoimmune diseases, laterality, gender, and age distribution of the PIU and JIA-associated uveitis patients as well as healthy controls are shown in Table 1. The PIU patients consisted of 1238 subjects (603 males and 635 females) with an average age of 10.5 ± 4.3 years at disease onset. JIA-associated uveitis patients included 47 males and 81 females with an average age of 9.6 ± 7.2 years at first disease onset. The mean age of healthy controls (including 516 males and 598 females) was 39.5 ± 10.4 years. In the PIU patients, family history of autoimmune diseases was noted in 40 cases, including RA (15), AS (12), SLE (2), vitiligo (3), psoriasis (4), scleroderma (1), polymyositis (1), eczema (1), and ichthyosis (1). In the JIA-associated uveitis patients, family history of autoimmune diseases was present in 9 cases, including RA (5), AS (3), and psoriasis (1).

Table 1 Clinical characteristics, family history of autoimmune diseases, laterality, sex, and age of participants
Full size table

Genotype and allele frequencies of SNPs in PIU

Fourteen SNPs in PDCD1, PDCD1LG2, FAS, and FASLG were genotyped and statistically analyzed in 1238 PIU patients and 1114 healthy controls. Significant differences were observed for SNP rs6710479 and rs7421861 in PDCD1 between patients and controls (Table 2). The frequency of the CC genotype of rs6710479 in PIU cases was significantly lower (P = 7.13 × 10−4; Pc = 3.42 × 10−3; OR = 0.466) than that in normal controls. Analysis of rs7421861 showed a lower frequency of the G allele (P = 1.01 × 10−4; Pc = 4.85 × 10−3; OR = 0.741) and higher frequencies of the AA genotype (P = 1.64 × 10−4; Pc = 7.87 × 10−3; OR = 1.407) and A allele (P = 1.01 × 10−4; Pc = 4.85 × 10−3; OR = 1.350) in PIU patients. No association could be detected between PIU and the other SNPs examined (Appendix 1).

Table 2 Genotype and allele frequency analysis of PDCD1/rs6710479 and rs7421861 polymorphisms in PIU
Full size table

Genotype and allele frequency of SNPs in the stratified analysis

Band keratopathy and cataract were the two most common complications in the PIU patients. To assess the association of selected SNPs with these two clinical features, we made a stratified analysis. A higher frequency of the PDCD1/rs7565639 T allele and CT genotype (P = 7.83 × 10−4, Pc = 3.76 × 10−2, OR = 1.732; P = 2.18 × 10−4, Pc = 1.05 × 10−2, OR = 1.879, respectively) and a lower frequency of the C allele (P = 7.83 × 10−4; Pc = 3.76 × 10−2; OR = 0.577) was found in patients with band keratopathy as compared to healthy controls (Table 3). There was no association of the 14 SNPs with the presence of complicated cataract (Appendix 2).

Table 3 Genotype and allele frequency analysis of PDCD1/rs7565639 polymorphisms in PIU with band keratopathy
Full size table

Genotype and allele frequencies of SNPs in JIA-associated uveitis

Fourteen SNPs in PDCD1, PDCD1LG2, FAS, and FASLG were also genotyped and statistically analyzed in 128 JIA-associated uveitis patients and 1114 healthy controls. The results showed a tendency of an association between three SNPs (rs2227982, rs6710479 and rs7421861) of the PDCD1 gene and JIA-associated uveitis (Table 4). Significance was, however, lost after Bonferroni correction (Appendix 3).

Table 4 Genotype and allele frequency analysis of PDCD1/rs2227982, rs6710479, and rs7421861 polymorphisms in JIA-associated uveitis
Full size table

Discussion

The present study explored the genetic susceptibility of PIU in Han Chinese with apoptosis-related genes and showed a significant association with PDCD1 but not with PDCD1LG2, FAS, and FASLG. Polymorphisms of rs6710479 and rs7421861 in the PDCD1 gene were associated with PIU in Han Chinese, whereby individuals carrying the PDCD1/rs6710479 CC genotype were protected against PIU. Individuals with the PDCD1/ rs7421861 AA genotype or A allele were at a higher risk to develop this disease. Stratification analysis showed an association between band keratopathy and PDCD1/rs7565639, whereby the T allele or CT genotype was a risk factor and the CC genotype was protective.

Although the true pathogenesis of PIU remains unclear, immunological and genetic factors are likely to be involved.4 Previous studies on the genetic susceptibility of PIU have demonstrated that genetic polymorphisms of certain genes may play a role in the pathogenesis of PIU.29,30 However, its genetic background is not exactly clear yet, mainly due to the difficulty in collecting enough samples for such studies. For this purpose, we continuously collected the samples from PU patients over the past 9 years. In the patients with PU included in the present study, 98% (1213) were being referred to our uveitis center by doctors from local hospitals due to recurrent or refractory uveitis. The patients with one or two attacks were always treated at local hospitals. To ensure patient consistency and diagnostic accuracy, we made a large effort to only include those PIU patients and JIA patients who fulfilled strict diagnostic criteria. The 14 SNPs we selected from the apoptosis-related genes PDCD1, PDCD1LG2, FAS, and FASLG were based on previous studies and all the included SNPs had earlier been shown to be associated with at least one autoimmune disease or certain cancers.15,17,18,19,20,25,26,27,28

To the best of our knowledge, an association of PIU with a genetic variant of PDCD1 has not been previously reported. PDCD1 is a member of the cluster of differentiation 28 (CD28)/B7 family.9 PDCD1 is also expressed on ocular tissues and is involved in the induction of T lymphocyte apoptosis and inhibition of lymphocyte proliferation and cytokine secretion. It is thought to play an important role in the induction and/or maintenance of peripheral tolerance and autoimmune disease.31 PIU-associated SNPs of PDCD1 (rs7565639, rs6710479, and rs7421861) identified in the present study are located in intron 1, which may be related to transcriptional regulation and gene splicing.32 Owing to proximity to the promoter region, it may affect the expression of the PDCD1 gene and contribute to the development of PIU by controlling the balance of T lymphocyte apoptosis. More studies are, however, needed to clarify the biological mechanisms whereby these SNPs may affect the development of PIU.

The polymorphisms of rs7421861 and rs6710479 in the PDCD1 gene found to be important in our study have also been shown to be involved in the pathogenesis of colorectal cancer, in which PDCD1 plays a negative regulation of T cell activity,19 suggesting that genetic susceptibility to PIU may share biological mechanisms with cancer immunity. The OR value of the rs7421861 AA genotype and A allele in our study were 1.40 and 1.35, while in colorectal cancer, the AG genotype showed an OR value of 1.314. SNP 7565639, which was associated with band keratopathy in our patients, has earlier been shown to have a weak association with autoimmune myasthenia gravis.28 This suggests that rs7565639 may be a common susceptibility site for certain autoimmune diseases, although it did not show a correlation with Wegener’s granulomatosis. More autoimmune diseases need to be tested for this SNP to clarify this discrepancy.

JIA-associated uveitis shows clinical features similar to PIU, including location and nature of the intraocular inflammation. Complicated cataract and band keratopathy are commonly observed in both diseases.33 In this study, we also investigated the association of the tested SNPs with JIA-associated uveitis in Han Chinese but did not detect any significant associations following Bonferroni correction for multiple comparisons. These findings suggest that the two diseases do not share the same genetic background and may indeed be two separate uveitis entities.

This study has several limitations. First, our research only showed a significant association between rs7421861, rs6710479, rs7565639, and the risk of PIU but we did not provide the biological mechanisms explaining these observations. Second, owing to the relatively small number of samples from JIA-associated uveitis patients, we might have missed some associations and further studies using a larger sample size are needed. Third, our population was restricted to Han Chinese and studies in other ethnic groups are needed to confirm our findings. Another limitation is the fact that we only tested SNPs that were earlier shown to be associated with cancer or autoimmune disease and we may have missed associations with other unknown SNPs.

In conclusion, the results of this study showed associations of PDCD1/rs6710479 and rs7421861 with PIU. In addition, a stratified analysis according to the clinical features showed that PDCD1/rs7565639 is associated with band keratopathy in PIU. The results support a genetic association between PDCD1 polymorphisms and susceptibility to PIU in Han Chinese and provides new insights into the pathogenesis of this disease.