Circulating secretory component-containing anti-citrullinated protein antibodies prior to symptom onset in rheumatoid arthritis

Abstract

Introduction

Rheumatoid arthritis (RA) develops gradually over several years. This process includes a preclinical phase in which autoantibodies occur in the circulation, but symptoms and histological signs of joint inflammation are absent [1]. The mechanisms triggering arthralgia and subsequent progression into clinically manifest arthritis remain unknown. There are, however, several indications that mucosal surfaces are involved in the formation of antibody responses to citrullinated proteins (ACPA) and/or the transformation from preclinical autoimmunity into clinically evident inflammation [2]. Inflammation and mucosal breaches related to for instance smoking [3, 4] or oral bacteria like Porphyromonas gingivalis [5, 6] and Aggregatibacter actinomycetemcomitans [7] are two potential routes for both increased antigen citrullination and local mucosal ACPA production.

Secretory IgA is the dominant antibody at mucosal linings, where it is formed upon attachment of a secretory component (SC) to dimeric IgA [8]. We previously identified secretory component-containing ACPA (SC ACPA) in serum from RA patients [9] and that levels correlate with those in saliva and bronchoalveolar fluid [10, 11]. Importantly, serum SC ACPA levels are independently prognostic for arthritis development in IgG ACPA positive symptomatic at-risk patients [12], and SC ACPA also occur in RA joint fluid [13]. Thus, it can be hypothesized that mucosal ACPA production is mechanistically important in disease development, but we lack knowledge concerning in which phase of disease development these antibodies occur. In symptomatic at-risk patients, serum SC ACPA levels are surprisingly stable over time, regardless of progression to arthritis [12]. In the asymptomatic phase, however, the occurrence and trajectory of SC ACPA responses in the circulation remain unknown.

Non-symptomatic first-degree relatives (FDRs) to RA patients are frequently used to represent a pre-RA phase since they possibly share genetic and environmental risk factors. IgA ACPA is common among FDRs [14] and associate with citrullinated protein-expressing neutrophil extracellular traps [15] supporting the hypothesis of mucosal immune responses as part of the early stage of RA. We previously found that circulating SC ACPA was rare among FDRs [16], but that study was cross-sectional and did not include information concerning subsequent RA development.

Since data are lacking concerning serum SC ACPA in asymptomatic individuals subsequently developing RA, it remains unknown whether these antibodies occur prior to symptom onset, and whether they precede the IgG ACPA response. We therefore aimed to investigate occurrence and dynamics of circulating SC ACPA and total secretory IgA (TSIgA) in asymptomatic pre-RA individuals.

Material and methods

Subjects

We performed a case–control study within the Medical Biobank of northern Sweden, where recruitment, blood sampling and storage conditions (at −80°C) have been described previously [17]. Patients attending the Department of Rheumatology, Umeå, fulfilling the 1987 American College of Rheumatology classification criteria for RA [18], and with a known date for the onset of symptoms of joint disease were co-analysed with the registers of the Medical Biobank. Consequently, for this study 319 individuals [mean age (s.d.) 54.8 (8.5) years; designated as ‘pre-symptomatic individuals’, 81.8% women] were identified as having donated blood samples before the onset of any symptoms of joint disease. Up to three plasma samples were identified for each individual; 270 with one sample, 44 with two samples and five with three samples. The median predating time was 4.7 years [interquartile range (IQR) 4.4 years] until onset of symptoms. Control subjects [n = 100, age 51.8 (7.9) years, 69% females] were randomly selected from the same Biobank cohorts. Of the individuals identified as pre-symptomatic individuals, 181 had also provided blood samples when attending the clinic at the time of being diagnosed with RA [age 58.0 (8.4) years, 70.7% females]. The median time (IQR) between the onset of symptoms, and a diagnosis of RA was 7.7 (6.8) months. All cases were classified as either a ‘non-smoker’ or ‘ever-smoker’.

The Regional Ethics Committee at the University Hospital, Umeå, Sweden approved this study (Dnr 2013-347-31), and all participants gave their written informed consent when donating samples.

Secretory IgA analysis

An in-house sandwich ELISA was modified to detect TSIgA [13]. Half-area 96-well plates were coated with a monoclonal anti-secretory component antibody (Sigma-Aldrich, St Louis, MO, USA) diluted 1:1000 in phosphate-buffered saline (PBS) pH 7.4 (Medicago AB, Uppsala, Sweden) and incubated in a moist chamber at 37°C for 2 h followed by incubation for 3 h at 4°C. The plates were washed and blocked with 0.5% BSA (Sigma-Aldrich). Thereafter, plasma samples were diluted 1:100 in PBS–Tween–0.5% BSA and incubated for 1 h in a moist chamber at 37°C. Following washing, the detection antibody was added (polyclonal goat anti-human IgA α-chain-specific peroxidase conjugated antibody, Sigma-Aldrich, diluted 1:25000) and incubated for 1 h in a moist chamber at 37°C. Tetramethylbenzidine (Sigma-Aldrich) was added to the plates and incubated in the dark at room temperature for 30 min, stopped using 1.8 M sulfuric acid and read at OD 450 nm (Tecan Sunrise, Männedorf Switzerland). The intra- and inter-assay variation was 6% and 18%, respectively. A seven-step standard curve was set up using purified IgA from human colostrum (containing SIgA, Sigma-Aldrich). The cut-off value was set at the 98th percentile among controls (319 ng/ml).

SC ACPA analysis

SC ACPA was measured in plasma by modifying commercially available anti-CCP ELISA kits (anti-CCP2, Svar, Malmö, Sweden) [19]. Samples were diluted 1:25, added to pre-coated CCP microtitre plates and incubated at room temperature. Following washing, horseradish peroxidase (HRP)-conjugated polyclonal goat anti-human SC (Nordic Biosite, Täby, Sweden) diluted 1:2000 was added. Incubation was stopped and read at 450 nm (Tecan Sunrise). A seven-step serial dilution was used for standard curve calculations using patient sera with high levels of SC ACPA. Cut-off values for positive test were set at the 98th percentile among the population controls (183 AU/ml). The intra- and inter-assay variation was 4% and 8%, respectively.

IgG ACPA analysis

Detection of IgG ACPA was performed by using enzyme-linked immunoassays according to the manufacturer’s instructions (Svar), including cut-off for positivity at 25 AU/ml.

Statistics

Continuous data were compared (pre-symptomatic individuals vs controls or patients) by non-parametric analyses the Mann–Whitney test for two groups and the Kruskal–Wallis test for comparing several groups. Wilcoxon’s signed rank test and Friedman’s test were used for matched pairs (e.g. several samples from the same individual). Relationships between categorical data (positive vs negative) were compared using χ² analysis or Fisher’s exact test as appropriate. Correlation analyses were performed using Spearman’s rank correlation. The relationship between antibody levels or positivity vs negativity and time to symptom onset were calculated using linear regression analysis and logistic regression analysis, respectively, adjusted for sex, age and smoking habits.

Results

Secretory antibodies in pre-symptomatic individuals and RA patients

SC ACPA levels in pre-symptomatic individuals [median (IQR) 108 (108) AU/ml] and RA patients [179 (248) AU/ml] were significantly increased compared with controls [12.5 (537) AU/ml, P < 0.001], and SC ACPA levels in RA patients were significantly increased compared with pre-symptomatic individuals (P < 0.001) (Fig. 1A). TSIgA levels, on the other hand, did not differ significantly between the groups (Fig. 1B). SC ACPA positivity was significantly more common in the pre-symptomatic phase and after disease onset compared with controls, while no significant differences were seen for TSIgA (Supplementary Table S1, available at Rheumatology online). SC ACPA levels correlated significantly with high IgG ACPA levels (rho: 0.553, P ≤ 0.001). Neither SC ACPA nor IgG ACPA was associated with being an ever smoker (data not shown).

Figure 1.

Secretory component-containing ACPA (SC ACPA) and total secretory IgA (TsIgA) in pre-symptomatic individuals, patients with RA and controls. (A) Plasma levels of SC ACPA in pre-symptomatic individuals, patients with RA and controls. Cut-off for positivity (183 AU/ml) is marked as a dotted line. ***P < 0.001. (B) Plasma levels of TsIgA in pre-symptomatic individuals, patients with RA and controls. Cut-off for positivity (319 ng/ml) is marked as a dotted line. ns: non-significant

Open in new tabDownload slide

Secretory antibodies over time

The earliest positive SC ACPA was detected 9 years before symptom onset of RA (Fig. 2), with a frequency of positivity during predating time of 13.0% (35/270). The proportion of positivity increased significantly during the pre-symptomatic phase (from 0% to 27.6%, i.e. 8–29 individuals ≤1 year before symptom onset, P = 0.007), and further increased following onset of symptoms and subsequent diagnosis [8/29 (27.6%) vs 89/181 (49.2%), P = 0.03]. Supplementary Fig. S1 (available at Rheumatology online) illustrates plasma SC ACPA levels among the pre-symptomatic individuals with more than one sample available. Among these, SC ACPA levels increased non-significantly from the oldest sample [>−9 years; median 562 (IQR 1188) AU/ml] to the one closest to symptom onset [<−1 year; 1117 (IQR 2155) AU/ml] when including all samples (P = 0.276), but when including samples after diagnosis and one or two samples, respectively, during predating time, the increase in levels was significant [1669 (IQR 2280) and 1676 (IQR 2324) AU/ml, P < 0.001 and P < 0.01, respectively]. The increase in SC ACPA levels ≤1 year before symptom onset and after onset was significant for matched pairs (P = 0.003).

Figure 2.

Cumulative percentages of positivity over time for serum secretory component containing ACPA (SC ACPA), IgG ACPA and total secretory IgA (TsIgA) in pre-symptomatic individuals (labelled −11 to −1) and patients at RA diagnosis (labelled as year 1)

Open in new tabDownload slide

IgG ACPA positive pre-symptomatic individuals could be detected already at 11 years before symptom onset, and the cumulative proportion of positive samples showed a steeper increase than SC ACPA and TSIgA during the pre-symptomatic phase (Fig. 2). Positivity for TSIgA was detected 7.6 years before diagnosis among pre-symptomatic individuals (Fig. 2), and the levels showed a marginal and non-significant (P = 0.679) increase until diagnosis. Only two pre-symptomatic cases and one patient were SC ACPA positive without testing positive for IgG ACPA.

To further delineate the dynamics of SC ACPA during the asymptomatic phase, we performed regression analyses including time to symptom onset adjusted for age, sex and smoking. In the linear regression model, including individuals with only one predating sample, a significant association between time to symptom onset and SC ACPA levels (on average 95 AU/ml per year closer to symptom onset, P < 0.001) was found (Supplementary Table S2, available at Rheumatology online). Likewise, in a logistic regression model, the proportion of SC ACPA positivity increased closer to symptom onset [odds ratio = 1.41 (95% CI 1.19, 1.66), P < 0.001] (Supplementary Table S3, available at Rheumatology online).

Discussion

In this uniquely large cohort of pre-symptomatic individuals subsequently developing RA, we showed that SC ACPA positivity can be detected in plasma up to 9 years before RA diagnosis. We also showed that SC ACPA levels and positivity increased both during the asymptomatic phase and during the transition from preclinical autoimmunity to subsequent RA. This indicates that SC ACPA production could be a part of the mucosa-associated links for RA development following persistent inflammation in mucosal sites due to smoking or presence of bacteria [20]. Although these data show that the SC ACPA response arises and magnifies before RA symptoms develop, we found no evidence that these antibodies precede the development of an IgG ACPA response, at least not in the circulation. Instead, the results show that IgG ACPA, which does not carry a secretory component [13], appear earlier and more frequently during the pre-symptomatic phase compared with SC ACPA. In symptomatic at-risk patients, elevations of both antigen specific antibodies (SC ACPA) and non-antigen specific antibodies (TSIgA and TSIgM) can be detected prior to arthritis development [13]. In contrast, the present study shows that only antigen-specific SC ACPA, and not TSIgA, were increased in pre-symptomatic individuals.

As previously reported [17], IgG ACPA predated symptom onset by up to 11 years, i.e. they appeared earlier than the first SC ACPA positive sample. Further, only two pre-symptomatic individuals had SC ACPA without concomitant IgG ACPA. Therefore, our data do not support a role of mucosal immunization as the trigger for a systemic IgG ACPA response. It must be noted, however, that the present study shows data from the circulation as no samples from mucosal sites were available from the biobanks used. Such samples, e.g. from sputum or saliva, would potentially have yielded different results. Nevertheless, we previously studied saliva samples from asymptomatic FDRs without finding SC ACPA [16], but whether or not they progressed to RA was not known, and the markedly lower SC ACPA positivity in plasma (1% vs 13% in the current study) implies that FDRs are substantially different from the pre-symptomatic individuals in the current study.

This study does not address the question whether circulating SC-containing autoantibody responses in general are specific for RA, or more prominent in RA compared with other autoantibody-associated diseases. In a previous study, it was reported that 36% of patients with ANCA-associated vasculitis had SC-containing anti-proteinase 3 antibodies in serum [21], implying that circulating SC-containing autoantibodies are prevalent in several autoimmune diseases. However, additional diseases need to be investigated to resolve this issue.

Strengths of the current study include the large group of pre-symptomatic individuals with sampling many years before symptom onset. Limitations include the lack of mucosal samples, and the limited number of repeated samples during the asymptomatic phase. Nevertheless, the results imply that in a subgroup of RA patients, SC ACPA responses develop and magnify during the asymptomatic phase of disease development. This suggests mucosal involvement in disease development prior to both arthralgia and subsequent arthritis. However, as mirrored in the circulation, secretory ACPA responses do not seem to be the initial trigger for the IgG ACPA response.

Supplementary material

Supplementary material is available at Rheumatology online.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Funding

This study was funded by the Swedish Rheumatism Association, the King Gustav V 80-year Foundation, the Research Council of Southeast Sweden (FORSS), and ALF-grants from Region Östergötland and Västerbotten and Umeå University.

Disclosure statement: The authors have declared no conflicts of interest.

Acknowledgements

We would like to acknowledge the staff of the Department of Rheumatology and the Medical Biobank, Umeå.

References

Author notes