The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.
CD4+FoxP3+ Treg cells suppress effector T cells and prevent autoimmune disease. Treg cell function is deficient in active rheumatoid arthritis (RA), a loss which may play a role in the pathogenesis of this disease. We previously showed that a single-nucleotide polymorphism in the FCRL3 gene led to higher expression of Fc receptor–like 3 (FcRL3) on Treg cells and that FcRL3+ Treg cells are functionally deficient in comparison to FcRL3− Treg cells. This study was undertaken to investigate the potential role of FcRL3 in RA.
A cross-sectional study was performed to evaluate the FCRL3 −169 genotype and FcRL3 expression on T cell subsets, including Treg cells, in peripheral blood samples from 51 patients with RA enrolled in the University of California, San Francisco (UCSF) RA Cohort. Clinical data were obtained from the UCSF RA Cohort database.
Patients with the FCRL3 −169C allele (genotype C/C or C/T) expressed higher levels of FcRL3 on Treg cells, and on CD8+ and γ/δ T cells, in comparison to RA patients with the T/T genotype. Higher FcRL3 expression on these T cell subpopulations correlated with RA disease activity in patients harboring the FCRL3 −169C allele. Furthermore, FcRL3 expression on Treg cells was higher in patients with erosive RA, and the FCRL3 −169C allele was overrepresented in patients with erosive RA.
Our findings indicate that FcRL3 expression, which is strongly associated with the presence of the FCRL3 −169C allele, may serve as a biomarker for RA disease activity.
Rheumatoid arthritis (RA) is a debilitating inflammatory arthritis affecting 1% of the world's population. The etiology of RA is multifactorial, and our knowledge of the specific environmental and genetic factors leading to and sustaining aberrant immune activation in the disease is limited. Treg cells are a subset of T cells that are critical in maintaining immune self tolerance and preventing autoimmune disease (1); as such, they may play a key role in RA pathogenesis. Treg cells mediate an inhibitory effect on immune activity by suppressing the proliferation and function of effector T cells. In mouse models of RA, adoptive transfer of Treg cells results in the resolution of arthritis (2). Studies using peripheral blood samples from patients with RA have shown that Treg cells are functionally deficient and, moreover, anti–tumor necrosis factor (anti-TNF) treatment restores Treg cell function in these patients (3–6).
Recent work from our group and from others has identified a transmembrane cell surface receptor, Fc receptor–like 3 (FcRL3), that is highly expressed on CD4+ Treg cells but not on conventional CD4+ T cells (7, 8). FcRL3 is also expressed on other T cell subsets, B cells, and natural killer cells (7, 9, 10). FcRL3 is part of a genetically conserved gene family bearing high structural homology to classic Fc receptors, with multiple extracellular Ig domains and with intracellular domains that carry either immunoreceptor tyrosine–based activation motifs (ITAMs), immunoreceptor tyrosine–based inhibition motifs (ITIMs), or both (11). Given these signaling domains and expression on multiple immune cell types, the FcRL family members likely modulate immune cell functions by affecting signaling pathways (11–13).
No physiologic function has been ascribed to FcRL3, and its ligand is unknown. The presence of both ITAMs and ITIMs in the FcRL3 intracellular domain suggests that engagement of it by cognate ligand might enhance or inhibit cell function. In vitro studies using the FcRL3 intracellular domain have shown that it may act as a negative regulator of B cell receptor signaling (14). Our group has demonstrated that FcRL3+ Treg cells are less capable of suppressing effector T cell proliferation in vitro than their FcRL3− Treg cell counterparts (8). Taken together, these data suggest that FcRL3 may also function as a negative regulator of Treg cell function.
A single-nucleotide polymorphism (SNP) in FCRL3 located in the promoter (−169 T→C, FCRL3_3, rs7528684) leads to enhanced NF-κB binding and to increased FCRL3 promoter activity (9). The FCRL3 −169C variant is associated with increased expression of FcRL3 on Treg cells and B cells (8, 9, 15) and has been identified as a potential genetic risk factor in multiple autoimmune diseases, including RA, autoimmune pancreatitis, systemic lupus erythematosus, and autoimmune thyroid disease (9, 16, 17). FcRL3 may therefore lie within a common pathway in autoimmune disease pathogenesis. In this study, we investigated the relevance of FcRL3 to RA by examining associations between the FCRL3 −169C variant, FcRL3 expression on FcRL3–expressing T cell subsets and B cells, RA disease activity, and erosive RA.
PATIENTS AND METHODS
A total of 51 patients were recruited from the University of California, San Francisco (UCSF) RA Cohort. All patients gave written informed consent using protocols approved by the UCSF Committee on Human Research. Patients who were seropositive for human immunodeficiency virus and/or hepatitis C, older than 65 years, or had a history of an infection in the prior month were excluded. All patients met the American College of Rheumatology 1987 revised criteria for RA (18). All clinical, laboratory, and radiographic parameters were ascertained through the UCSF RA Cohort database. Radiographs of the hands and feet were scored as erosive or nonerosive by the patient's attending rheumatologist. A summary of the clinical characteristics of the patients is shown in Table 1.
Table 1. Clinical characteristics of the 51 RA patients*
Except where indicated otherwise, values are the number (%) of patients. RA = rheumatoid arthritis; DAS28 = Disease Activity Score in 28 joints; ESR = erythrocyte sedimentation rate; anti-TNF = anti–tumor necrosis factor; DMARDs = disease-modifying antirheumatic drugs; CCP = cyclic citrullinated peptide; RF = rheumatoid factor.
46 (90)/5 (10)
Age, mean ± SD years
47 ± 12
FCRL3 −169 genotype
DAS28, mean ± SD
ESR, mean ± SD mm/hour
34 ± 24
No anti-TNF or DMARDs
CCP antibody positive
FCRL3 −169 C/C or C/T
FCRL3 −169 T/T
FCRL3 −169 C/C or C/T
FCRL3 −169 T/T
Tissue and cell isolation.
Peripheral blood mononuclear cells (PBMCs) were isolated from the blood by density-gradient centrifugation using Ficoll-Hypaque Plus (Amersham Biosciences). All washes were performed with 2% fetal bovine serum (FBS) in phosphate buffered saline (PBS).
Antibodies and flow cytometric phenotyping.
All phenotyping analyses were performed on fresh samples. PBMCs from RA patients were stained with either anti-FcRL3 antibody (kindly provided by Genentech) (10) or with an irrelevant protein control (human serum albumin; Sigma-Aldrich) that had previously been biotinylated using the FluoReporter Mini-Biotin Protein Labeling Kit (Invitrogen). Secondary detection of FcRL3 or the control was performed with a streptavidin–Qdot 655 conjugate (Invitrogen). FcRL3+ cells were defined by the negative control, since there was not a separate population of FcRL3+ cells but rather a continuum of expression. For additional cell surface phenotype analysis, the following antibodies were used: Alexa Fluor 700–conjugated anti-CD3, phycoerythrin (PE)–Cy7–conjugated anti-CD25, and PE-conjugated anti-CD127 (all from BD Biosciences); anti–γ/δ T cells (eBioscience); and Texas Red–conjugated anti-CD4 and PE–Cy5.5–conjugated anti-CD8 (Caltag). Dead cells were excluded from the analysis by cell surface staining with the Aqua Live/Dead amine-reactive dead cell stain reagent (Invitrogen).
For phenotyping, cells were incubated with the relevant antibodies diluted in PBS/2% FBS for 30 minutes on ice, followed by 3 washes with PBS/2% FBS. Secondary staining for FcRL3 detection with streptavidin–Qdot 655 was then performed, followed by an additional 3 washes in PBS/2% FBS. Cells were then fixed in 1% paraformaldehyde (Sigma) for flow cytometric analysis. Intracellular detection of FoxP3 was performed using anti-FoxP3 Pacific Blue (clone PCH101); the accompanying staining kit was provided by eBioscience and was used according to the recommendations of the manufacturer. Samples were acquired on an LSRII flow cytometer (BD Biosciences), and all data were analyzed using FlowJo software. Doublet discrimination based on forward scatter height versus area was performed to eliminate cellular aggregates.
PBMCs (1 × 106 from each patient sample) were stored as dry pellets at −80°C until analyzed. DNA was isolated from the cell pellets using the DNeasy Blood and Tissue kit, according to the recommendations of the manufacturer (Qiagen). The −169C and −169T FCRL3 alleles were discriminated from one another with a polymerase chain reaction (PCR)–based 5′-nuclease genotyping assay that used unlabeled forward and reverse PCR primers (900 nM final concentration) and 2 allele-specific probes labeled with either VIC or FAM reporter dye (200 mM final concentration; Applied Biosystems). Assay components were added to 20 ng DNA in a 20-ml reaction containing TaqMan Universal PCR Mix. An AB StepOnePlus instrument was used for amplification and detection, and AB system software was used for analysis.
The significance of differences in FcRL3 expression between groups was assessed using Prism software and the Mann-Whitney unpaired single comparison test. P values less than 0.05 (by Mann-Whitney test) were considered significant. Linear regression analysis for scatterplot data was performed using Prism software, and P values and r2 values are shown. Fisher's exact test was performed using Prism software to determine the significance of differences in allele frequency between groups. A biostatistician from the UCSF Clinical and Translational Science Institute performed all regression analyses using the R programming language (www.r-project.org). For quantitative outcomes (erythrocyte sedimentation rate [ESR] and Disease Activity Score in 28 joints [DAS28] ), we used linear regression analysis to assess the association between outcomes and predictors.
Correlation of FcRL3 expression on T cell subsets in RA patients with the presence of the FCRL3 −169C allele.
Previous work from our laboratory showed that higher FcRL3 expression on Treg cells from healthy donors was linked to the presence of the FCRL3 −169C allele (C/T or C/C genotype), but not to the T/T genotype (8). In this study, we first determined whether FcRL3 expression on T cell subsets was affected by the FCRL3 −169C genotype in RA patients as well. This question was addressed by performing flow cytometric analysis of PBMCs from 51 patients with RA who were enrolled in the UCSF RA Cohort.
The baseline characteristics of the RA patients included in this study are shown in Table 1. The study population was predominantly Hispanic (61%) and Asian (27%). Of note, 21% of the patients had the FCRL3 −169C/C genotype, 41% had the FCRL3 −169C/T genotype, and 38% were homozygous for FCRL3 −169T. This distribution was similar to the expected genotype percentages for a known C allele frequency of 0.43 in Asian populations and 0.44 in European populations (17). The mean DAS28 was 4.44 (range 1.63–7.92). All patients were rheumatoid factor positive, and 90% had antibodies against cyclic citrullinated peptides.
Since expression of FcRL3 on Treg cells was similar irrespective of whether they were defined as CD4+CD25+FoxP3+ cells or CD4+CD25+CD127low cells (Figure 1A), we defined Treg cells as CD4+ CD25+FoxP3+ for the purposes of the data analysis. Consistent with the results of our previous work in healthy individuals, we found that FcRL3 expression on Treg cells from the peripheral blood of RA patients was associated with the FCRL3 −169C allele in a dose-dependent manner (Figure 2A). The percentage of FcRL3+ Treg cells and the mean FcRL3 expression on FcRL3+ Treg cells in RA patients were also associated with the FCRL3 −169C allele (data not shown). We also confirmed that, like FcRL3+ Treg cells in healthy individuals, FcRL3+ Treg cells in RA patients express low levels of CD45RA and CCR7 and high levels of PD-1 and CD62L. (Results are available from the corresponding author upon request.) In addition, there was no statistically significant difference in the mean expression of FcRL3 on Treg cells or the percentage of FcRL3+ Treg cells between healthy controls and RA patients. (Results are available from the corresponding author upon request.) Differences in medications received, ethnicity, and age did not result in significant changes in FcRL3 expression on Treg cells. (Results are available from the corresponding author upon request.)
Since FcRL3 is expressed on CD8+ T cells, γ/δ T cells (Figures 1B and C), and B cells (7), we also examined FcRL3 expression on these subpopulations in relation to the FCRL3 −169C variant. Consistent with the results reported by Gibson et al (15), we found that the FCRL3 −169C variant was associated with higher FcRL3 expression on CD19+ B cells. (Results are available from the corresponding author upon request.) As was the case for FcRL3 expression on Treg cells, FcRL3 expression on CD8+ T cells and γ/δ T cells was associated with the presence of the −169C FCRL3 allele (Figures 2B and C). Of the three T cell subsets examined, FcRL3 expression on Treg cells was significantly higher than on CD8+ and γ/δ T cells (P < 0.001) (Figure 2D). Of note, the patients with the highest FcRL3 expression on Treg cells also had the highest FcRL3 expression on CD8+ and γ/δ T cells (data not shown). Taken together, these results show that the −169C FCRL3 allele is associated with increased expression of FcRL3 on T cell subsets.
Correlation of FcRL3 expression on T cell subsets with RA disease activity, especially in patients with the FCRL3 −169C allele.
Treg cell function is deficient in RA patients (3, 4, 6). Because FcRL3+ Treg cells are less suppressive than FcRL3− Treg cells, we hypothesized that high levels of FcRL3 expression on Treg cells (resulting in less functional Treg cells) would be associated with more pronounced RA disease activity (8). We first examined the study patients with RA as a single group and found that the levels of FcRL3 expression on Treg cells were correlated with a known marker of RA disease activity (20), the ESR (r2 = 0.10, P = 0.03). FcRL3 expression on Treg cells also correlated with a validated, composite measure of RA disease activity, the DAS28 (r2 = 0.09, P = 0.04). (Results are available from the corresponding author upon request.)
Since the FCRL3 −169C allele confers enhanced NF-κB transcription factor binding and FCRL3 promoter activity (9), we hypothesized that the associations between FcRL3 and ESR and DAS28 described above would be especially evident in those RA patients with the −169C FCRL3 allele. Indeed, examination of the subset of RA patients harboring one or two FCRL3 −169C alleles in comparison to all patients revealed a stronger association between FcRL3 expression on Treg cells and ESR (r2 = 0.25, P = 0.003) (Figure 3A) and between FcRL3 expression on Treg cells and DAS28 (r2 = 0.16, P = 0.002) (Figure 3B) in the patients with the FCRL3 −169C allele. Reciprocally, the association between FcRL3 expression on Treg cells and RA disease activity was not observed in RA patients who had the FCRL3 −169T/T genotype (Figures 3C and D).
Furthermore, the percentage of FcRL3+ Treg cells was associated with RA disease activity, and this relationship was more evident in the RA patients harboring one or two FCRL3 −169C alleles. Again, the association between the percentage of FcRL3+ Treg cells and RA disease activity was not observed in patients with the FCRL3 −169T/T genotype. (Results are available from the corresponding author upon request.) To assess whether FcRL3 expression on Treg cells and/or the FCRL3 −169 genotype were independent predictors of RA disease activity, we performed multivariate regression analysis with both genotype and expression as predictors. We found that, in this study of 51 patients, FcRL3 expression on Treg cells was an independent predictor of RA disease activity (as assessed by DAS [P = 0.02] or ESR [P = 0.04]), while the FCRL3 −169 genotype was not (P > 0.4).
We next investigated whether FcRL3 expression on other FcRL3–expressing immune cell subsets correlated with RA disease activity. When all RA patients were assessed as a single group, an association was observed between FcRL3 expression on CD8+ T cells and RA disease activity, as measured by ESR (r2 = 0.19, P = 0.002) and DAS28 (r2 = 0.26, P = 0.0003). (Results are available from the corresponding author upon request.) As was the case for FcRL3 expression on Treg cells, RA patients harboring an FCRL3 −169C allele showed a stronger association between FcRL3 expression on CD8+ T cells and ESR (r2 = 0.29, P = 0.002) (Figure 4A) and between FcRL3 expression on CD8+ T cells and DAS28 (r2 = 0.34, P = 0.0004) (Figure 4B) compared with all patients combined. Reciprocally, this association was not observed in RA patients who had the FCRL3 −169T/T genotype (data not shown).
Examination of the γ/δ T cell subset in RA patients as a single group showed a similar association between FcRL3 expression and the RA disease activity measurements ESR (r2 = 0.16, P = 0.06) and DAS28 (r2 = 0.40, P = 0.002). (Results are available from the corresponding author upon request.) As seen with Treg cells and CD8+ T cells, this association between FcRL3 expression and RA disease activity was more evident in RA patients with one or two FCRL3 −169C alleles (Figures 4C and D). Interestingly, we did not find an association between FcRL3 expression on CD19+ B cells and RA disease activity. (Results are available from the corresponding author upon request.) Collectively, these data show that higher levels of FcRL3 expression on Treg cells, CD8+ T cells, and γ/δ T cells are associated with greater degrees of RA disease activity, and that this association is more evident in RA patients who possess an FCRL3 −169C allele.
Association of erosive RA with the FCRL3 −169C allele and higher FcRL3 expression on Treg cells.
Erosive RA identifies a subset of RA patients who have more destructive and debilitating RA, and is defined by visible bone erosions on radiographs. Given the association between RA disease activity and FcRL3 expression on Treg cells, we hypothesized that the FCRL3 −169C allele would be more predominant in RA patients with erosive disease than in those without erosive disease. The data in Table 1 show that there was a trend toward a higher frequency of FCRL3 −169C alleles in the subset of RA patients with erosive disease (44% of patients with nonerosive disease versus 71% of patients with erosive disease; P = 0.057 by Fisher's exact test). In support of these data, the expression of FcRL3 on Treg cells was found to be significantly higher in RA patients with erosive disease than in those with nonerosive disease (P = 0.03) (Figure 5A). A trend toward an association between erosive RA and FcRL3 expression on either CD8+ or γ/δ T cells (Figures 5B and C) was also found, although it did not reach statistical significance. These data implicate FcRL3+ Treg cells as a relevant T cell subset associated with erosive RA.
The results of this study demonstrate that, among RA patients who harbor one or two FCRL3 −169C alleles, FcRL3 expression on T cell subsets is associated with RA disease activity. Thus, we found that FcRL3 expression was higher on Treg cells, CD8+ T cells, and γ/δ T cells in the presence of the FCRL3 −169C allele, and that such expression was correlated with RA disease activity as measured by ESR and DAS28. We also found that RA patients with erosive disease expressed higher levels of FcRL3 on Treg cells and exhibited an increased frequency of the FCRL3 −169C allele. These observations suggest that FcRL3 expression, which is strongly associated with the FCRL3 −169C variant, may serve as a biomarker for RA disease activity.
Although several earlier studies identified the FCRL3 −169C variant as a genetic risk factor for RA, genome-wide association studies did not identify FCRL3 as a risk allele (9, 21, 22). Our results suggest that this variant, through its association with FcRL3 expression, may be a genetic risk factor for development of recalcitrant RA disease activity and progression to erosive disease. Our findings support the results of a recent study by Chen et al (23) that showed an association between the FCRL3 −169 genotype and erosive disease in Taiwanese patients with RA. It is possible that the variability in identifying the FCRL3 −169C variant as a genetic risk factor for developing RA is dependent on the severity of disease and/or the ethnicity of the examined study population. Interestingly, a recent study by Maehlen et al (24) showed that the FCRL3 −169 C/C genotype was associated with erosive RA in a large Scandinavian cohort, suggesting that our findings in a predominantly Hispanic and Asian RA cohort may apply more broadly to European populations.
It remains to be determined whether FcRL3 cell surface expression contributes to or is a consequence of RA disease activity. These are not mutually exclusive hypotheses, since basal levels of FcRL3 expression may be set by the FCRL3 −169 SNP, while inducible FcRL3 expression may be driven by the FCRL3 −169 SNP and proinflammatory NF-κB–inducing cytokines that are produced during active RA. An individual with the FCRL3 −169C allele might accordingly exhibit higher basal levels of FcRL3, and FcRL3 expression may be up-regulated to a greater extent under inflammatory conditions. Furthermore, the regulation of FcRL3 expression likely involves additional factors, including other genetic factors, that are yet to be identified. Longitudinal studies of a large cohort will be necessary to elucidate a more definitive role of FcRL3 expression as a predictive biomarker for RA disease activity.
Previous work from our laboratory showed that FcRL3+ Treg cells are less suppressive than FcRL3− Treg cells and led us to postulate that FcRL3 may function as a negative regulator of Treg cell function (8). Although the mechanism by which this occurs is unknown, it is likely that, similar to classic Fc receptors, FcRL3 can modulate immunoreceptor signaling through its ITIM/ITAM. There is in vitro evidence that B cell receptor signaling is down-modulated by the intracellular domain of FcRL3. While the results of the present study show that FcRL3 is most highly expressed on Treg cells, it is also expressed on CD8+ and γ/δ T cells and, as seen with Treg cells, expression on the latter cell types also correlates with RA disease activity. Furthermore, we have found that within CD8+ T cells, FcRL3 is expressed on those that are CD8+CD28− (Swainson LA and Bajpai UD, et al: unpublished observations), a phenotype that has previously been reported to demonstrate suppressive function (25). In addition, subpopulations of γ/δ T cells are thought to have suppressive function (26, 27). It is interesting to speculate that FcRL3 may specifically function as a regulator of T suppressor cells and, by extension, that the high levels of FcRL3 found on CD8+ and γ/δ T cells are relegated primarily to CD8+ and γ/δ T cells with a suppressive function. Alternatively, given the presence of both ITAM and ITIM on its intracellular domain, FcRL3 may function as a positive or negative regulator in different T cell subsets or in response to different stimuli.
FcRL3 may play a pivotal role in a common pathway leading to autoimmune disease pathogenesis (16, 17). Recent studies have demonstrated that NF-κB activity is associated with decreased Treg cell function and RA disease activity (4, 28). It is possible that NF-κB mediates this Treg cell dysfunction and consequent unbridled immune activation, at least in part, through up-regulation of FcRL3 expression. If so, FcRL3 may represent a target for therapeutic intervention in the context of RA: its down-regulation or interruption would predictably lead to the emergence of a Treg cell pool that could more efficiently suppress the inflammatory state in RA. Further studies will be needed to characterize the role of FcRL3 on all FcRL3-expressing immune cells, including B cells and CD8+ and γ/δ T cells, and to better assess its potential role as a therapeutic target in the setting of RA and other autoimmune diseases.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. McCune had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Bajpai, Swainson, Mold, Graf, Imboden, McCune.
Acquisition of data. Bajpai, Swainson.
Analysis and interpretation of data. Bajpai, Swainson, Mold, Graf, Imboden, McCune.
We would like to thank Dr. Andy Polson and Genentech for providing the anti-FcRL3 antibody, which was crucial for these experiments. We would like to thank biostatistician Dr. Saunak Sen. We are thankful to Terence Ho at the Division of Experimental Medicine Flow Cytometry Core (UCSF) for technical assistance, Dr. Satish Pillai for statistical assistance, and the McCune Laboratory members for helpful comments. We are especially grateful to the UCSF RA Cohort staff and patients.