Association of rheumatoid factor production with FcγRIIIa polymorphism in Taiwanese rheumatoid arthritis

Abstract

Fcγ receptors (FcγR) impact upon the development of inflammatory arthritis through immune complex stimulation and proinflammatory cytokine production. FcγRIIa, FcγRΙΙΙa and FcRγIIIb polymorphisms were genotyped in 212 rheumatoid arthritis (RA) patients and 371 healthy control subjects using an allelic-specific polymerase chain reaction (PCR). No significant skewing in the distribution of FcγRIIa H/R131, FcγRIIIa F/V158 and FcγRIIIb NA1/NA2 was found between RA patients and healthy control subjects. However, a significant skewing distribution of the FcγRIIIa F/V158 polymorphism was observed between rheumatoid factor (RF)-positive versus RF-negative RA patients (P = 0·01). The low-affinity FcγRIIIa F158 allele seems to have a protective role in RF production, in comparison with the FcγRIIIa V158 allele (P = 0·004; OR = 0·485; 95% CI: 0·293–0·803). A high frequency of FcγRIIIa F/F158 was identified in RA patients with negative RF compared with RF-positive patients (for FF158 versus FV158 + VV158; P = 0·002; OR = 0·372; 95% CI: 0·194–0·713). In addition, no association was found between FcγRIIa H/R131, FcγΡIIIa F/V158 and FcγRIIIb NA1/NA2 polymorphisms and other clinical parameters. The results of this study suggest that three activating FcγRs polymorphisms lack association with RA but FcγIIIa F/V158 polymorphism may influence RF production and IgG RF immune complex handling in Taiwanese RA patients.

Introduction

Rheumatoid arthritis (RA) is a chronic inflammatory autoimmune disease characterized by synovial cell overproliferation, mononuclear cell infiltration and pannus formation, which results finally in the destruction of cartilage and bone. The precise aetiology and pathogenesis of RA remain for future study. Genetic background and environmental triggers are both important in the disease determination of RA [1,2]. Human leucocyte antigen (HLA) class II with a shared epitope is a major genetic risk factor but cumulative studies suggest that non-HLA genes may also contribute to the disease process [3–5].

Rheumatoid factors (RFs) are produced through a bystander effect of non-specific polyclonal activation of B cells or an antigen-driven specific subset of B cells. IgG immune complexes (ICs) of RFs and other autoantibodies may play an important role in triggering inflammatory arthritis [6]. These RF complexes are present in the circulation as well as synovial fluids and joints, and thus may be disease process mediators. RF titres are correlated with the severity of joint destruction and extra-articular manifestations [6–9].

Interaction between the IgG ICs and Fcγ receptors (FcγRs) affects critically the cellular biological functions of immune response. FcγRs comprise several activation receptors and one major inhibitory receptor. Two specialized motifs in the cytoplasmic tail of the γ chain interact with distinct tyrosine kinases, transmitting activating signals using the immunoreceptor tyrosine-based activation motif (ITAM), or inhibiting signals using the immunoreceptor tyrosine-based inhibition motif (ITIM) [10,11]. FcγR knock-out mice, with the deletion of the common γ chain and lacking activating FcγRI and FcγRIII functions, exhibited protective effects against collagen, experimental antigen and immune complex-induced arthritis [12–16]. In contrast, FcRIIb-deficient mice without regulatory inhibitory modulation were susceptible to collagen-induced arthritis [12,16].

In human RA studies, expression levels of FcγRII and FcγRIII up-regulated in monocytes and macrophages resulted in proinflammatory cytokine production and arthritis inflammation [17]. FcγRIIa expression levels were also susceptible to down-regulation following disease-modifying anti-rheumatic drug (DMARD) therapy [18]. Thus, FcγR activation may be essential in the pathogenesis of RA clinical disease activity and sequential synovial changes. Single nucleotide polymorphisms (SNPs) in three activating FcγR genes, FcγRIIa H/R131, FcγRIIIa F/V158 and FcγRIIIb NA1/NA2, exhibit biological functions that differ among the distinct FcγR genotypes, especially IgG subtypes binding activity and receptor-mediated phagocytosis of ICs [19]. Burn et al. reported that the FcγRIIa R/R131 genotype was less effective in the processing of circulating ICs and was associated with RA disease severity [20]. Notably, FcγRIIIa F/V158 polymorphism has been studied in different ethnic groups of RA patients for disease susceptibility and severity, but the results were inconsistent [21–25].

In this study, we examined the relationship between FcγR polymorphisms and RA susceptibility its clinical implications. The aims of this study are to identify whether SNPs in FcγRs are genetic risk factors for RA in the Taiwanese population. The information may be valuable in predicting disease severity and in providing guidelines for early interventions to prevent further destruction of cartilage and bony structure.

Materials and methods

Patients

A case–control study was conducted to determine the genetic association of rheumatoid arthritis and FcγR IIa, IIIa and IIIb polymorphisms. For this study, we enrolled 212 patients with RA in Taiwan. The sample population comprised patients with RA treated in the rheumatology clinics of the Chang Gung Memorial Hospital. All patients were assessed by rheumatology specialists and fulfilled the 1987 American Rheumatism Association criteria for RA. The study was approved by the local ethics committee of Chang Gung Memorial Hospital. Meanwhile, 371 healthy control subjects (75·2% female) with a mean age of 34·0 ± 10·3 years were selected from blood donors, who each completed a questionnaire survey to exclude autoimmune diseases. Laboratory examination results were defined as positive when anti-nuclear antibody (ANA) ≥ 1 : 80 dilution titres using Hep-2 cell assay and rheumatoid factor (RF) were ≥ 20 IU/ml using a nepholometric assay. Rheumatology specialists graded serial damage of the X-ray films of both hands as follows: soft tissue swelling and peri-articular osteoporosis (stage I), bony erosion (stage II), joint subluxation (stage III) and joint ankylosis (stage IV). Two investigators confirmed the X-ray changes and staging of each patient.

Nucleic acid extraction and allele-specific polymerase chain reaction (PCR) for FcγR genotyping

Genomic DNA was extracted from ethylenediamine tetraacetic acid (EDTA)-anti-coagulated peripheral blood using the Purgene DNA isolation kit (Bioman, Taipei, Taiwan). The allele-specific PCR assays were developed and used for the genotyping of SNPs in the FcγRIIa (SNP H/R131), FcγRIIIa (SNP V/F158) and FcγRIIIb (SNP NA1/NA2), as described previously [26]. The PCR reactions were performed in a GeneAmp PCR system (ABI Biosystems, Foster City, CA, USA) with 200 ng of genomic DNA in a 50 µl reaction. The allele-specific PCR product was assayed on agarose gel staining with ethidium bromide. The appearance of the PCR product in the oligonucleotide allele-specific reaction indicated the presence of that allele. Several control donors with known genotypes for each gene were included for each PCR setup to safeguard the authenticity of genotyping assays. To avoid genotyping errors resulting from poor-quality DNA, this study included only those donors who had unambiguous genotypes on all three FcγR genes for analysis.

Statistical analysis

Genotype frequencies of the three FcγR polymorphisms in the RA patients and control subjects were compared using the χ² test. Data were analysed using the spss version 10 statistical package for Windows (spss, Gorinchem, the Netherlands). Variant genotype distributions were also compared among RA patients according to age of onset, X-ray severity and between positive versus negative for RF and ANA. A P-value < 0·05 was considered significant.

Results

Clinical characteristics of RA in Taiwan

Two hundred and twelve RA patients and 371 healthy blood donors were recruited for this study. The study cohort comprised 177 female and 35 male RA patients. The RA patients were followed up for at least 1 year to evaluate the disease course. The mean disease duration was 9·6 ± 7·8 years and the mean age at onset was 46·2 ± 13·1 years. One hundred and sixty-two of a total of 212 RA patients were positive for RF ≥ 20 IU/ml (76·4%) and 108 of 200 RA patients had ANA test titres ≥ 1 : 80 (54%). Regarding X-ray findings of hands, 144 patients exhibited bony erosion (≥ stage II), 83 patients demonstrated destruction of joint subluxation and/or ankylosis (≥ stage III) and 38 patients had cervical spine involvement. The patients in this study displayed low prevalence of rheumatoid nodule presentation.

Distribution of FcγRIIa H/R131, FcγRIIIa F/V158 and FcγIIIb NA1/NA2 genotypes and associations with RA

No significant skewing in the distributions of FcγRIIa H/R131, FcγRIIIa F/V158 and FcγRIIIb NA1/NA2 was found between the RA patients and the healthy control subjects in Taiwan. The genotype frequencies of those SNPs in the RA patients and healthy control subjects were both within Hardy–Weinberg equilibrium. To assess the correlation between severity and genotypes of RA patients, FcγRIIa H/R131, FcγRIIIa F/V158 and FcγRIIIb NA1/NA2 genotypes of RA patients with erosion (n = 144) or severe destruction (n = 83) were compared with healthy controls. The distributions of FcγRIIa H/R131, FcγRIIIa F/V158 and FcγRIIIb NA1/NA2 genotypes of the RA patients still displayed no significant skewing in RA patients with erosion or severe destruction. Moreover, no significant differences in FcγRIIa H/R131, FcγRIIIa F/V158 and FcγRIIIb NA1/NA2 genotypes distribution were found among the five different age groups at onset and between the ANA positive versus negative RA patients (Table 1). However, a significant difference in the genotype distributions of FcγRIIIa F/V158 polymorphisms was observed between RF-positive and -negative RA patients (P = 0·01). The distributions of FcγRIIIa allele frequencies and genotypes were analysed further in RF-positive and -negative RA patients. Table 2 shows that the FcγRIIIa F158 allele had a protective role in RF production in comparison with the FcγIIIa V158 allele (P = 0·004, OR 0·485, 95% CI 0·293–0·803). Furthermore, we observed a significant enrichment of the low binding allele (FcγRIIIa F158) homozygotes in RF-negative RA patients (30 of 50) compared with RF-positive RA patients (58 of 162) (P = 0·002, OR 0·372, 95% CI 0·194–0·713). Additionally, two locus-combined genotype analyses of Fcγ IIa H/R131, FcγRIIIa F/V158 and FcγRIIIb NA1/NA2 showed no significant association with RA disease susceptibility and clinical parameters. FcγRIIIa F/V158 polymorphisms were also compared in previous case–control association studies and displayed similar genotype distribution among various ethnic groups (Table 3).

Table 1.

Frequencies of FcγRIIa, FcγRIIIa and FcγRIIIb alleles for various clinical manifestations and autoantibodies positive versus negative of RA in Taiwan.

		FcγRIIa			FcγRIIIa			FcγRIIIb
Characteristics	(no.)	131HH (%)	131H (%)	131RR (%)	158FF (%)	158FV (%)	158VV (%)	NA1/1 (%)	NA1/2 (%)	NA2/2 (%)
Normal	371	41·2	46·9	11·9	41·8	45·8	12·4	40·2	44·7	15·1
RA	212	42·5	49·5	8·00	41·5	42·9	15·6	42·5	42·5	15·1
RA erosion	144	44·4	45·8	9·70	43·1	43·8	13·2	41·0	43·0	16·0
RA destruction	83	41·0	47·0	12·0	45·8	41·0	13·3	49·4	37·3	13·3
Age onset (years)
16–29	23	47·8	39·1	13·1	26·1	52·2	21·7	56·5	21·7	21·7
30–39	41	46·3	46·3	7·30	53·7	34·1	12·2	31·7	58·5	9·80
40–49	67	41·8	49·2	9·00	37·3	41·8	20·9	43·3	38·8	17·9
50–59	44	43·2	50·0	6·80	36·4	50·0	13·6	47·7	38·6	13·6
≥ 60	37	35·1	59·5	5·40	51·4	40·5	8·10	37·8	48·6	13·5
RF
(+)	162	41·4	50·0	8·60	35·8	46·9	17·3	45·7	40·7	13·6
(−)	50	46·0	48·0	6·00	60·0*	30·0	10·0	32·0	48·0	20·0
ANA
(+)	108	48·2	43·5	8·30	36·1	46·3	17·6	39·8	45·4	14·8
(−)	92	35·9	55·4	8·70	48·9	38·0	13·1	44·6	42·4	13·0

^*P-value = 0·01. RA, rheumatoid arthritis; RF, rheumatoid factor; ANA, anti-nuclear antibody.

Table 2.

Frequencies of FcγRIIIa F/V158 alleles and genotypes in rheumatoid factor (RF) positive versus negative of rheumatoid arthritis (RA) in Taiwan.

Patients group (no.)	RF-positive (n = 162)	RF-negative (n = 50)	Odds ratio (95% CI)	P-value
F allele	192 (0·59)	75 (0·75)	0·485 (0·293–0·803)	0·004
V allele	132 (0·41)	25 (0·25)
FF	58 (0·36)	30 (0·60)	0·372 (0·194–0·713)	0·002
FV + VV	104 (0·64)	20 (0·40)
VV	28 (0·17)	5 (0·10)	1·881 (0·685–5·162)	0·214
FV + FF	134 (0·83)	45 (0·90)

Table 3.

Comparison of frequencies of FcγRIIIa genotypes among rheumatoid arthritis (RA) and control subjects in studies involving various ethnic groups.

		RA			Normal
Author (ref. no.)	Ethnic group/patient no. (RA : normal)		158FF (%)	158FV (%)	158VV (%)	158FF (%)	158FV (%)	158VV (%)
Nieto et al. [24]	Caucasian*	(142 : 117)	48·7	38·5	12·8	32·4	53·5	14·1
Morgan et al. [23]	Caucasian*	(141 : 124)	40	45	15	52	39	9
Milicic et al. [22]	Caucasian	(401 : 420)	41	47	12	41	51	8
Brun et al. [20]	Caucasian	(112 : 89)	39·3	45·5	15·2	48·3	36	15·7
Morgan et al. [21]	Caucasian*	(828 : 581)	43	44	14	47	43	10
Morgan et al. [23]	India*	(108 : 113)	37	54	9	52	41	7
Milicic et al. [22]	India*	(63 : 93)	57	40	3	47	38	15
Kyogoku et al. [45]	Japanese	(382 : 303)	53·7	39	7·3	47·8	43·6	8·6
Chen et al. [this study]	Taiwanese	(212 : 371)	41·5	42·9	15·6	41·8	45·8	12·4

^*Positive association with susceptibility to RA.

Discussion

RA is characterized by persistent inflammatory responses that result in progressive joint and bone destruction. The severity of RA may influence physical, psychological and social functions as well as quality of life. The genetic associations of RA susceptibility and severity remain to be clarified [27–32]. In RA development, IgG containing immune complexes may be crucial in the initiation and persistence of the inflammatory cascade in the local joints. The receptors for IgG play important roles in the regulation of immune responses. The functional coordination of activating and inhibitory Fcγ receptors is thought to determine the severity of IC-mediated inflammation in vivo [33,34]. Accordingly, the functional roles of FcγRs in inflammatory diseases have been studied extensively in animals and humans. Fcγ chain knock-out mice (Fcγ^–/–) were able to avoid severe chronic inflammation and cartilage destruction, whereas the inhibitory Fc receptor-deficient mice are susceptible to collagen-induced arthritis [13,16,34]. Furthermore, the balance of activating and inhibitory FcγR, which regulate by Th1 and Th2 cytokines, plays an important role in the activation of monocytes and production of inflammatory cytokines [35,36]. Therefore, it is believed that the functional SNPs in FcγR genes will influence IC-mediated autoimmune diseases [19]. Expressions of FcγRIIa and FcγRIIIa were enhanced on monocytes and macrophages, and the high levels of activating FcγR expression may contribute to increased activation of monocyte and macrophages, resuling ultimately in high tumour necrosis factor, interleukin 1 and matrix metalloproteinase production [17,37].

RF complex plays many physiological and functional roles in RA. Therefore, the receptors for IgG are reasonable candidate genes in the study of RA susceptibility. Our data indicate that activating FcγRs may not influence the immune tolerance or may not be risk factors for RA in general. Neverthless, the SNPs in the FcγRs could be disease-modulating factors. Interestingly, we observed a significant enrichment of the FcγRIIIa low-affinity allele (F158 allele) in RF-negative RA patients. Homozygotes of the low-affinity allele (FcγRIIIa F/F158 genotype) were also increased significantly in RF-negative Taiwanese RA patients. Our data suggest that the low-affinity allele of FcγRIIIa (F158) may have a protective role in RF production and SNP in FcγRIIIa may, after all, influence RF immune complex handling. FcγRIIIa is expressed mainly on NK cells, macrophages, γδ T cells, subsets of monocytes and dendritic cells and mast cells in humans [38–42]. These cells may play important roles and serve as critical effector cells in RA pathogenesis. Therefore, FcγRIIIa-expressing cells may have a key role in the perpetuation of inflammatory responses to IgG immune complexes. Consequently, functional SNP in FcγRIIIa may influence the disease process. It has been postulated that macrophages expressing the high-affinity FcγRIIIa allele (V158) could enhance capture of IgG opsonized pathogens or IgG immune complexes, feeding them directly into the antigen-processing pathway and resulting in a more efficient presentation of arthritogenic peptides [21,23,43]. This process could lead to enhanced RF production. Increased immune complex-binding and subsequently enhancing downstream effector functions in phagocytes could play a significant role in sustaining inflammation and RF production in local joints. On the other hand, the low-affinity FcγRIIIa allele (F158) will bind less IC or RF complexes and could cause the diminished inflammatory responses.

It is reasonable to assume that functional SNP in FcγRIIIa could be associated with RA susceptibility and severity. Nevertheless, genetic studies of FcγRIIIa SNP (F/V158) exhibit inconsistent findings in RA patients. It has been shown that the SNP in FcγRIIIa was associated with RA phenotypes, but there were reports that also demonstrated the lack of SNP association with RA [20–24,44,45]. Table 3 lists the published studies that showed both positive and negative associations of the SNP with RA phenotypes. There are several possible explanations for the inconsistent results. First, the genetic contribution of the FcγRIIIa SNP to geologically different populations or ethnicities may not be the same, which might explain the association of the SNP with RA in certain Caucasian groups but lack of association in Japanese and Taiwanese. Secondly, the sample size is critical in association studies, especially in the situation of a weak association. The small sample size might result in either false positive or false negative associations. Therefore, studies involving small sample populations should be interpreted cautiously. Thirdly, the modest association of FcγR polymorphisms with RA susceptibility found in some studies may result from the linkage disequilibrium with the true disease susceptibility response genes located in the neighbourhood of FcγR genes.

We are aware that our study groups comprised modest sample sizes both for RA patients and normal controls. After further stratification, we could have missed the weak association of certain FcγR SNP genotypes with some clinical phenotypes of RA. On the other hand, we found that the FcγIIIa F158 allele has a protective role in RF production in RA patients. Enrichments of the F158 allele and F158 homozygotes in RF-negative RA patients were still statistically significant (P < 0·01) after adjusting for multiple testing. Even so, our finding of the protective effect of the FcγRIIIa F158 allele against RF production needs to be replicated independently in a large cohort.

The FcγRIIA R/H131 polymorphism has been implicated in lupus nephritis and other autoimmune diseases owing to the different IgG bingeing avidity and phagocytosis activity [19]. Burn et al. claimed that the FcγRIIA R/R131 genotype displayed low IC clearance and was predisposed to RA disease severity [20]. Their study samples were also limited and require further confirmation. FcγRIIIb, expressed in neutrophils, has been correlated with vasculitis with a difference in phagocytosis between FcγRIIIb NA1/NA2 genotypes [46]. However, FcγRIIIb NA1/NA2 polymorphism showed no significant skewing of genotype distribution in any previous studies involving RA. Our data also confirmed the lack of association of FcγRIIIb SNP with RA phenotypes.

FcγR polymorphisms are gene clusters located on chromosome 1q22–24. FcγR polymorphisms may share the linkage disequilibrium effect with each other. However, two-locus analysis did not find an additive effect of FcγRIIa H/R131, FcγRIIIa F/V158 and FcγRIIIb NA1/NA2 in the present study. In a previous study, the knock-out mice model demonstrated that the inhibitory receptors FcγRIIb play a significant role in inflammatory arthritis [16]. Recently, Kyogoku et al. reported no significant association of a potential FcγRIIb polymorphism over the coding region with Japanese RA [45]. These findings require further confirmation.

Regarding RA severity, the present study compared patients exhibiting bony erosion and destructive changes with healthy subjects, because the disease duration of non-destructive RA patients may not be long enough and substantial individual treatment bias existed. However, we found no significant differences in genotype distribution among X-ray films that have shown bony erosion and/or destruction. As a polygenic determinant disease, RA requires the sequential effector response genes to follow continuous pathological pathways [3].

In summary, we found that three activating FcγRs polymorphisms are not associated with RA susceptibility and severity, but the FcγIIIa F/V158 polymorphism may be implicated in the immunopathogenesis of RF production and IgG RF immune complex handling in Taiwanese RA patients. Genetic variations associated with RA susceptibility or severity remains for further investigation.