1. INTRODUCTION
Autoantibodies (antibodies produced by the immune system) directed against immunoglobulin G (IgG) Fc fragments are a primary characteristic of rheumatoid arthritis (RA). However, whether IgM, IgG, or IgA rheumatoid factors (RFs) correlate with RA disease activity remains controversial,1–4 5 , 6
A previous study reported lower levels of IgG-linked galactose in RA patients.7 , 8 9 10 2 -3, α2-6, or α2-8 linkages.11 12 13 14
The carbohydrate moiety of IgG is vital for certain antibody effector functions.15 , 16 17 18 19
Hence, we hypothesized that α2,6-SIA contents on IgG, IgM, and autoantibodies might correlate with the occurrence of arthritis and RA disease activity. These perspectives will add to the currently used laboratory parameters for determining inflammation, such as CRP level and erythrocyte sedimentation rate (ESR), to improve the measures of disease activity in RA. Accordingly, we analyzed the α2,6-SIA contents on IgG, IgM, and autoantibodies (including anti-CCP and RF antibodies), as well as their correlations with the occurrence of mouse arthritis and clinical activity in RA patients.
2. METHODS
2.1. CIA in mice
Six- to 8-year-old C57B6 mice were obtained from the National Animal Center (Taipei, Taiwan) and housed in a temperature-controlled, specific pathogen-free room with a 12-hour day/night cycle at the Animal Center of Chang Gung Memorial Hospital at Linkou (Taoyuan, Taiwan). We followed the same method as described previously to elicit CIA.20 , 21 Fig. 1B ] or with the forelimb swelling together) as CIA mice. Moreover, the mice in the experimental group with no occurrence of gross arthritis were designated as non-CIA mice, and the mice injected with PBS were designated as control mice (Table 1 ).
Table 1 -
Contrasting Ig-α2,6-SIA/Ig ratios in different mouse groups
Groups
Day 0
Week 2
Week 4
Week 6
SIA/IgG ratios
CIA mice
1.631 ± 0.670
0.741 ± 0.237
0.559 ± 0.118
0.463 ± 0.125* (4)
Non-CIA mice
1.629 ± 0.376**
0.906 ± 0.351
0.830 ± 0.365
0.811 ± 0.476**,*** (6)
Control mice
2.022 ± 0.322
1.640 ± 0.084
1.545 ± 0.144
1.536 ± 0.349*,*** (4)
SIA/IgM ratios
CIA mice
0.706 ± 0.103****
0.720 ± 0.121
0.916 ± 0.056
0.887 ± 0.036**** (4)
Non-CIA mice
0.910 ± 0.372*****
0.737 ± 0.064
0.822 ± 0.076
0.761 ± 0.083*****,****** (6)
Control mice
0.923 ± 0.155
0.847 ± 0.153
0.875 ± 0.103
0.860 ± 0.082****** (4)
Blood samples were obtained on day 0 (the same day when collagen was injected), and in weeks 2, 4, and 6 as described in Methods. Parentheses indicate the number of mice used in individual treatment group. Upper half: blood samples were directly tested for mouse IgG α2,6-SIA contents and the results were divided by IgG amounts to give final ratios. No difference in day 0 data of the IgG α2,6-SIA level among three groups was observed (compared using the t test: control mice day 0 vs CIA mice day 0: p = 0.348 [95% confidence interval, –0.612 to 1.393]; control mice day 0 vs non-CIA mice day 0: p = 0.108 [–0.111 to 0.897]); control mice between day 0 and week 2 were not different (p = 0.095 [–0.114 to 0.877]), between day 0 and week 4 were comparable (p = 0.052 [–0.006 to 0.959]) and between day 0 and week 6 were similar (p = 0.087 [–0.096 to 1.067]). Lower half: mouse IgM α2,6-SIA contents tested in blood were divided by IgM amounts to give final ratios. Day 0 IgM SIA levels among three groups were not different (CIA mice day 0 vs non-CIA mice day 0 and vs control mice day 0 provided p = 0.213 [−0.554 to 0.146] and 0.065 [−0.452 to 0.019], respectively). This is representative of two experiments.
CIA mice = mice with collagen-induced arthritis; control mice = mice injected with phosphate buffered saline without arthritis; Ig = immunoglobulin; non-CIA mice = mice injected with collagen and did not develop arthritis; SIA = sialic acid.
*p = 0.005 [0.545 to 1.600].
**p = 0.007 [0.278 to 1.358].
***p = 0.025 [0.121 to 1.329].
****p = 0.016 [−0.313 to −0.047].
*****p = 0.338 [−0.196 to 0.495].
******p = 0.107 [−0.028 to 0.226].
Fig. 1: The whole course comparison of IgG and IgM alpha-2,6-sialic acid (SIA) ratios between collagen-induced arthritis (CIA) mice and control mice. A, The IgG alpha-2,6-SIA/IgG ratios expressed as mean ± SD over a 6-wk period. Comparison between week 0 and week 6 in CIA mice gave p < 0.05 (95% confidence interval (CI): 0.123 to 2.213); that between week 0 and week 6 in control mice gave p > 0.05 (95% CI: −0.957 to 1.067); that between day 0 and week 2 in control mice were not different (p = 0.095, 95% CI:–0.114-0.877). (B) Hind leg arthritis in CIA mice vs no arthritis in control mice. (C) IgM-α2,6-SIA/IgM ratios are expressed as mean ± SD during the 6-wk period. Comparison between week 0 and week 6 in non-CIA mice yielded p > 0.05 (95% CI: −0.194 to 0.493); in control mice p > 0.05 (95% CI: −0.151 to 0.277). D, IgG and IgM α2,6-SIA ratio changes (expressed as mean ± SD) in CIA mice during a 6-wk period. The comparison of IgG α2,6-SIA ratios between week 0 and week 6 in CIA mice yielded p < 0.05 (95% CI: 0.123 to 2.213) and that of IgM α2,6-SIA ratios surrendered p < 0.05 (95% CI: −0.336 to −0.025). This is representative of two experiments.
2.2. Patients and clinical and laboratory parameters
In total, 108 volunteer RA patients (female to male ratio, 85:23) provided their informed consent before their inclusion in this study. All patients fulfilled the 1987 ACR criteria for classification of RA. Blood samples were collected from 30 newly diagnosed/untreated and 78 regularly treated RA patients. None of the recruited patients had secondary Sjogren’s syndrome with systemic lupus, tuberculosis, Crohn’s disease,22 , 23
2.3. Study approval
The Institutional Animal Care and Use Committee of Chang Gung Memorial Hospital at Linkou (Taoyuan, Taiwan) approved the protocol for mouse experiments. The “Principles of Laboratory Animal Care” (NIH Publication Vol 25, No. 28 revised 1996), as well as specific national laws where applicable, were followed. The Chang Gung Institutional Review Board (Taoyuan, Taiwan) approved the human study, and the subjects’ written consent was obtained in accordance with the ethical standards of the 1975 Declaration of Helsinki (as revised in Brazil, 2013).
2.4. Isolation of immunoglobulins
Plasma was obtained from blood collected in ethylenediaminetetraacetic acid tubes undergoing centrifugation and separation as previously described.24
2.5. Preparation of F9.4 mouse IgM and AF1.9-L2 mouse IgG
Mouse IgM from the F9-4 cell line 25
2.6. ELISA for quantification of human or mouse IgG and IgM
The ELISA assays were performed as described previously, with modifications.26 2 fragment of a goat anti–human IgG, Fc fragment–specific antibody (Jackson ImmunoResearch Inc., West Grove, PA). Diluted isolated IgGs or plasma samples were then added and followed by the horseradish peroxidase (HRP)–conjugated F(ab′)2 fragment of a goat anti–human IgG to capture the target antibodies. Human IgG (range, 100-0.048 ng/well, 2-fold serial dilutions; contains little or no α2,6-SIA content; Sigma-Aldrich) served as a standard.
Similarly, human IgM quantification was measured with human IgM (range, 200-0.195 ng/well, 2-fold serial dilutions; Sigma-Aldrich) as a standard. Analogous procedures were followed for quantifying mouse IgG and mouse IgM. AF1.9-L2 IgG (range, 200-0.195 ng/well, 2-fold serial dilutions) and F9-4 IgM (range, 200-0.195 ng/well, 2-fold serial dilutions) were used as standards for mouse IgG and IgM, respectively.
2.7. ELISA for quantification of α2,6-SIA contents of human IgG and IgM
ELISA microtiter plates were coated with Sambucus nigra lectin (SNA-1, which binds α2,6-SIA; EY Laboratories, Inc., San Mateo, CA). After washing, isolated IgGs (or plasma) were added at appropriate dilutions. Incubation with secondary antibodies, with either a biotin-conjugated anti–human IgG antibody or a biotin-conjugated anti–mouse IgG antibody (Jackson ImmunoResearch Inc.), followed by incubation with an HRP–streptavidin detector. This was used to quantify the α2,6-SIA contents in either human IgG or AF1.9-L2 mouse IgG (range, 400-25 ng/well, 2-fold serial dilutions, used as a standard for detecting human IgG α2,6-SIA), respectively. Thereafter, ELISA results for α2,6-SIA were correlated with results assayed by the lectin blot method.27
Human IgM α2,6-SIA was quantified with human IgM (range, 200-0.195 ng/well, 2-fold serial dilutions; carrying a high α2,6-SIA content; Sigma-Aldrich) as a standard.
2.8. Lectin blotting
As described previously,28
2.9. ELISA for quantification of α2,6-SIA contents of mouse IgG and IgM
The same procedure as above was used for quantifying mouse IgG and IgM α2,6-SIA contents. AF1.9-L2 (range, 200-0.195 ng/well, 2-fold serial dilutions) was used as the standard for α2,6-SIA of IgG. Mouse IgM SIAs were measured with F9-4 IgM (range, 200-0.195 ng/well, 2-fold serial dilutions) as the standard.
2.10. ELISA for plasma IgMRF and for IgMRF α2,6-SIA contents
ELISA plates were coated with human IgG Fc fragment (Jackson ImmunoResearch Inc.) at 10 μg/mL for IgMRF detection.29 , 30 Sambucus nigra )–HRP solution (EY Laboratories, Inc.) was coated on the plates for measurement of the α2,6-SIA content of IgM (designated as the IgMRF-α2,6-SIA).30 29 , 30
2.11. Quantification of plasma anti–CCP antibodies and of their SIA contents
Plasma was analyzed for IgG anti-CCP antibodies (Quantalite CCP3 IgG ELISA; Inova Diagnostics, Inc.). For α2,6-SIA quantification, plates were incubated with 1:100 dilutions of isolated human IgG samples for 1 hour. Next, an SNA–HRP solution (EY Laboratories Inc.) was added. Furthermore, isolated IgG-anti-CCP-α2,6-SIA was assayed with the standard using the serum containing high IgG-anti-CCP antibody levels (at 250 units, measured as above) with a relatively high α2,6-SIA content.
2.12. Definition of immunoglobulin SIA contents and ratios
Plasma IgG-α2,6-SIA/IgG ratios were defined as the α2,6-SIA contents of plasma IgG divided by the plasma IgG levels (ng/mL:ng/mL). Plasma IgM-α2,6-SIA/IgM ratios were delineated as the α2,6-SIA contents of plasma IgM divided by the plasma IgM levels.
Isolated IgG-α2,6-SIA/IgG ratios were defined as the α2,6-SIA contents of isolated IgG (from a protein G column) divided by the plasma IgG levels. Isolated IgM-α2,6-SIA/IgM ratios were delineated as the α2,6-SIA contents of isolated IgM (from a protein A column together with a protein G column) divided by the plasma IgM levels.
Similarly, isolated IgG anti-CCP-α2,6-SIA/plasma IgG-anti-CCP ratios were defined as the α2,6-SIA contents of isolated IgG-anti-CCP divided by the plasma IgG-anti-CCP antibody levels. Plasma IgMRF-α2,6-SIA/IgMRF ratios were delineated as the α2,6-SIA contents of plasma IgMRF divided by the plasma IgMRF levels. Plasma IgGRF-α2,6-SIA/IgGRF ratios were similarly defined.
2.13. Statistical analyses
The correlation between groups was analyzed by means of Pearson’s correlation coefficient (r ) and linear regression when outcome variables (DAS28) were normally distributed. Otherwise, the correlation was assessed using Spearman’s correlation coefficient (rho, ρ). The comparison in mouse experiments was conducted by two-sample t or paired t test. All analyses were performed using SPSS 16.0 software (SPSS Inc, Chicago, IL, USA). A value of p < 0.05 was considered statistically significant. A trend was considered for 0.05 ≤ p < 0.10 for the correlation of variables in RA patients because patient characteristics were diversely distributed31
3. RESULTS
3.1. Changes in blood IgG or IgM α2,6-SIA contents in CIA
No difference was observed in the plasma IgG α2,6-SIA/IgG ratios among three groups (CIA mice, non-CIA mice, and control mice as described in Section 2.1.) on day 0 (Table 1 ). Plasma IgG-α2,6-SIA/IgG ratios of CIA mice were significantly lower than the control mice at week 6 (Table 1 ; Fig. 1A ). Plasma IgG-α2,6-SIA/IgG ratios of non-CIA mice were also lower compared with control mice at week 6 (Table 1 ). In addition, plasma IgG-α2,6-SIA/IgG ratios of CIA mice at week 6 were significantly lower than those at week 0, but no difference was observed in control mice between week 6 and week 0 (Fig. 1A ). Similarly, no difference was observed in plasma IgG-α2,6-SIA/IgG ratios of control mice between day 0 and week 2 (p = 0.095, 95% confidence interval:–0.114 to 0.877) (Fig. 1A ; Table 1 ). Thus, both CIA and non-CIA mice had decreased plasma IgG-α2,6-SIA/IgG ratios in week 6, which is indicative of inflammation. Note, such an observation was not seen in control mice (Table 1 ). This phenomenon raises the question regarding what other factors might be associated with arthritis in CIA mice but not in non-CIA mice.
Furthermore, we examined plasma IgM-α2,6-SIA/IgM ratios in mice. We found that there was no change in the plasma IgM-α2,6-SIA/IgM ratio in non-CIA and control mice between week 0 and week 6 (Table 1 ; Fig. 1C ). In contrast, the IgM-α2,6-SIA/IgM ratio in week 6 was significantly higher in CIA mice than that in week 0 (Table 1 ; Fig. 1D ). Moreover, there was a notable difference between the decreasing IgG-α2,6-SIA/IgG ratios over time and the increasing IgM-α2,6-SIA/IgM ratios over time in CIA mice (Fig. 1D ). Hence, α2,6-SIA ratio changes occurred in an inverse proportionality for IgG and IgM in CIA mice. In contrast, α2,6-SIA ratio changes affected only IgG in non-CIA mice (Table 1 ). We then tested if similar phenomena existed in human patients with RA.
3.2. Correlation between α2,6-SIA ratios of isolated IgG and DAS28 in RA patients
The female to male ratio of the 108 RA patients was 85:23, and the average age was 55.5 ± 11.7 years (mean ± SD). Mean CRP levels were 14.90 ± 27.66 mg/L (range, 0.21-196.45); mean ESR levels were 31.64 ± 28.81 mm/h (range, 3-134); mean RF titers were 80.73 ± 198.42 IU (range, 0-943); mean SJC was 2.31 ± 4.13 (range, 0-26); mean TJC was 6.56 ± 6.90 (range, 0-28); and mean DAS28 were 3.88 ± 1.40 (range, 1.46-9.18). In newly diagnosed/untreated RA patients, isolated IgG-α2,6-SIA/IgG ratios significantly correlated inversely with the DAS28 (Fig. 2A ).32 t =–2.120 and p = 0.043. Additionally, as determined by ELISA, the α2,6-SIA contents of isolated IgG correlated modestly with those measured by the lectin blot (n = 30, ρ = 0.331, and p = 0.074).
Fig. 2: Correlation between the IgG or IgM α2,6-sialic acid (SIA) ratios and disease activity score 28 (DAS28). A, IgG α2,6-SIA/IgG ratios in 30 newly diagnosed/untreated rheumatoid arthritis (RA) patients. Isolated IgG-α2,6-SIA/IgG ratios yielded a significant and inverse correlation with the DAS28: r =–0.383 and p = 0.037. B, Correlation of isolated IgM-α2,6-SIA/IgM ratios with DAS28 in 54 RA patients: r = 0.351, p = 0.009.
3.3. Correlation between α2,6-SIA ratios of isolated IgM and DAS28 in RA patients
The isolated IgM-α2,6-SIA/IgM ratio from RA patients exhibited a positive correlation with the DAS28 (Fig. 2B ). This result was also supported by the univariable linear regression result: isolated IgM-α2,6-SIA/IgM ratios against DAS28 gave t = 2.391 and p = 0.021. These data suggest that high IgM-α2,6-SIA/IgM ratios correlate with the high arthritic activity scores in RA patients, similar to the occurrence of arthritis in CIA mice at week 6, when compared with week 0 (Table 1 ; Fig. 1D ).
3.4. The ability of SIA ratios of anti-CCP antibodies to differentiate different RA disease activities
Isolated IgG-anti-CCP-α2,6-SIA/plasma IgG-anti-CCP ratios differed, with a trend between remission (DAS28 < 2.6) and nonremission (DAS28 ≥ 2.6), as was observed for CRP but not for ESR (Table 2 ). The ratios differed significantly between moderate and high disease activity (but not CRP and ESR, Table 2 ). In other words, lower isolated IgG-anti-CCP-α2,6-SIA/plasma IgG-anti-CCP ratios were seen in RA patients with higher disease activity (Table 2 ). Consequently, isolated IgG-anti-CCP-α2,6-SIA/plasma IgG-anti-CCP ratios may be useful in differentiating between various disease activities in RA patients.
Table 2 -
Laboratory data to differentiate disease activity score 28 disease activities in rheumatoid arthritis patients
Differentiation of disease activities
ESR
CRP
SIA/anti-CCP ratios
Remission vs non-remission
0.421
0.041
0.061
Low/moderate disease activity vs high disease activity
0.071
0.090
0.057
Moderate disease activity vs high disease activity
0.132
0.164
0.045
Data shown are p values. SIA/anti-CCP ratios indicate the SIA contents of isolated IgG anti-CCP antibodies divided by plasma IgG anti-CCP antibody levels. Remission means rheumatoid arthritis patients with disease activity score 28 (DAS28) < 2.6; nonremission means DAS28 ≥ 2.6; low disease activity, 2.6 ≤ DAS28 ≤ 3.2; moderate disease activity, 3.2 < DAS28 ≤ 5.1; high disease activity, DAS28 > 5.1. Comparisons were done using t -test or Mann-Whitney U test where appropriate.
CCP = cyclic citrullinated peptide; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; SIA = sialic acid.
3.5. The ability of SIA ratios of plasma IgM RF to differentiate diverse RA disease activities
By using multivariable linear regression, plasma IgMRF-α2,6-SIA/IgMRF ratios significantly correlated with the DAS28 (as an outcome variable, p = 0.006), similar to ESR (p = 0.005), but different from CRP (p = 0.222). That is, CRP was not as useful as IgMRF-α2,6-SIA/IgMRF ratios and ESR in predicting DAS28 in multivariable analysis.
Moreover, plasma IgGRF-α2,6-SIA/IgGRF ratios (r =–0.012, p = 0.906) did not correlate with DAS28. Nevertheless, plasma IgGRF-α2,6-SIA/IgGRF ratios were (not significantly) inversely associated with DAS28, which is contrary to the positive correlation of plasma IgMRF-α2,6-SIA/IgMRF ratios with DAS28.
3.6. AUCs of various laboratory parameters against different DAS28 disease activities with sex distinction
The ROC curves of ESR, CRP, and anti-CCP against DAS28 ≥ 2.60 (nonremission) yielded meaningful AUCs (Table 3 ). It was demonstrated that ESR and CRP had significantly higher AUCs than isolated IgG-anti-CCP-α2,6-SIA/plasma IgG-anti-CCP ratios (Table 3 ). In contrast, isolated IgG-anti-CCP-α2,6-SIA/plasma IgG-anti-CCP ratios were inclined to predict a DAS28 < 2.60 better than ESR and CRP (Table 3 ). Intriguingly, isolated IgG-anti-CCP-α2,6-SIA/plasma IgG-anti-CCP ratios also predicted a DAS28 ≦ 3.2 significantly better than ESR and CRP (Table 3 ).
Table 3 -
AUC for inflammatory markers and SIA ratios of autoantibodies against DAS28 in patients with rheumatoid arthritis
Laboratory data
DAS28 cut-off with < 2.6
DAS28 cut-off with ≧ 2.6
DAS28 cut-off with ≦ 3.2
DAS28 cut-off with > 3.2
ESR
0.321*1 (0.177-0.465)
0.679*1 (0.535-0.823)
0.241*2 (0.137-0.346)
0.759*2 (0.654-0.863)
CRP
0.266**1 (0.120-0.441)
0.734**1 (0.589-0.880)
0.349**2 (0.217-0.482)
0.651**2 (0.518-0.783)
Anti-CCP
0.335 (0.167-0.503)
0.665 (0.497-0.833)
0.452 (0.311-0.592)
0.548 (0.408-0.689)
SIA/anti-CCP ratios
0.666***1 (0.503-0.828)
0.334***1 (0.172-0.497)
0.649 (0.495-0.802)
0.351 (0.198-0.505)
IgM
0.465 (0.305-0.625)
0.535 (0.375-0.695)
0.482 (0.354-0.613)
0.517 (0.387-0.646)
IgMRFSIA/IgMRF ratios
0.505 (0.334-0.676)
0.495 (0.324-0.666)
0.492 (0.361-0.622)
0.508 (0.378-0.639)
Data shown are AUC values with (95% confidence interval [CI]). DAS28 with inclusion of ESR. DAS28 ≥ 2.6 indicates rheumatoid arthritis (RA) patients with nonremission; DAS28 < 2.6 designates RA patients with remission. SIA/anti-CCP ratios indicates the SIA content of isolated IgG anti-CCP divided by the amount of plasma IgG anti-CCP. IgMRF SIA/IgMRF ratios indicates the SIA content of IgMRFs divided by the amount of IgMRFs in plasma.
CCP = cyclic citrullinated peptide; CRP = C-reactive protein; DAS28 = disease activity score 28; ESR = erythrocyte sedimentation rate; SIA = sialic acid.
**,***Significant difference of AUCs was impressed when 95% CI were compared between AUCs of two continuous categories (*1 vs *1; *2 vs *2; and so on) were nonoverlapped. In contrast, nonsignificant difference of AUCs was impressed when 95% CI between compared AUCs of two continuous categories were overlapped. ESR against DAS28 > 5.1 provided AUC (0.688, 95% CI: 0.515-0.822), not different from that AUC (0.373, 0.197-0.548) of isolated IgG-anti-CCP-α2,6-SIA/plasma IgG-anti-CCP ratios against DAS28 > 5.1.
Moreover, when we classified RA patients into male and female subgroups, ESR and CRP of male patients predicted DAS28 ≦ 5.1 much better than DAS28 > 5.1 (Table 4 ). In contrast, ESR and CRP of female patients predicted DAS28 > 5.1 better than DAS28 ≦ 5.1 (Table 4 ). However, isolated IgG-anti-CCP-α2,6-SIA/plasma IgG-anti-CCP ratios did not distinguish between different DAS28; especially between gender.
Table 4 -
Different tendencies of area under the curve represented by key laboratory markers in male and female rheumatoid arthritis patients
Laboratory markers
DAS28 ≦ 5.1
DAS28 > 5.1
Male RA (n = 23)
ESR
0.872* (0.711 to 1.032)
0.032* (–0.102 to 0.289)
CRP
0.846** (0.591 to 1.101)
0.154** (–0.101 to 0.409)
SIA/anti-CCP ratios
0.667 (0.400 to 0.933)
0.333 (0.067 to 0.600)
Female RA (n = 85)
ESR
0.281*** (0.128 to 0.434)
0.719*** (0.566 to 0.872)
CRP
0.390 (0.234 to 0.547)
0.610 (0.453 to 0.766)
SIA/anti-CCP ratios
0.514 (0.345 to 0.684)
0.486 (0.316 to 0.655)
DAS28 with inclusion of ESR; SIA/anti-CCP ratios = ratios of sialic acid contents of isolated IgG anti-CCP divided by plasma IgG anti-CCP levels. Data shown are area under the curve values with (95% confidence interval [CI]).
CCP = cyclic citrullinated peptide; CRP = C-reactive protein; DAS28 = disease activity score 28; ESR = erythrocyte sedimentation rate; RA = rheumatoid arthritis; SIA = sialic acid.
*,**,***Significant difference of AUCs was impressed when 95% CI between compared area under the curves of two continuous categories were nonoverlapped.
4. DISCUSSION
CIA (Table 1 ; Fig. 1 ) demonstrated the following: (1) collagen immunization induced significantly lower plasma IgG-α2,6-SIA/IgG ratios compared with the PBS control; (2) collagen immunization significantly increased plasma IgM-α2,6-SIA/IgM ratios in arthritic mice, but not in collagen-injected mice without gross arthritis; (3) plasma IgG-α2,6-SIA/IgG ratios and plasma IgM-α2,6-SIA/IgM ratios displayed an inverse relationship in CIA mice. The third trend was also observed in RA patients (Fig. 2A , 2B). A similar tendency for IgG SIA was noted in a previous report.17
However, whether high-α2,6-SIA-carrying IgMs in CIA mice and RA patients are anti-inflammatory or proinflammatory is an interesting question that has not yet been answered (Figs. 1D and 2B and Section 3). Nevertheless, mouse and human data tentatively suggest that high-α2,6-SIA-carrying IgMs were associated with the occurrence of arthritis and higher arthritis activity in addition to the anti-inflammatory role of α2,6-SIA–rich IgGs in serum-induced arthritis and ADCC.14 , 17
No change was found in the α2,6-SIA content of whole IgG in pregnant RA patients,33 34 Fig. 2A ). This viewpoint was supported by studies in which T-cell–dependent and T-cell–independent antigens both induced tolerant immunosuppressive sialylated IgG antibodies.35 , 36 37 37 35 , 36 37 , 38
Therefore, the strength of our study was observed through the inverse correlation between the α2,6-SIA ratios in IgG and IgM in CIA mice and human RA, regardless of the dispute on whether an SNA-column preferably selects pathogenic autoantibodies.17 , 39 Table 2 ) and a better predictor for RA remission than ESR (Table 3 ). Third, ESR and CRP in male RA patients predicted a DAS28 ≦ 5.1 (ESR of female patients predicted a DAS28 > 5.1) (Table 4 ), which was not reported in the past.40 41 42 , 43
Moreover, whether ESR and CRP play different roles in shaping disease activity between male and female RA patients (Table 4 ) is an ideal topic to investigate for future studies. In terms of why only IgMRFSIA/IgMRF ratios, but not IgGRF-α2,6-SIA/IgGRF ratios, correlated with DAS28 scores is currently unknown. However, it may be related to the nature of IgGRF to self-associate, hence, making measurements of IgGRF-α2,6-SIA and IgGRF less accurate.44 , 45
The primary limitation of the current study is that the correlation of laboratory measurements with DAS28 was mostly modest to reasonable (Fig. 2 ).32 Table 2 ). Second, no anti-collagen IgG and IgM α2,6-SIA were measured, due to the unavailability of blood for testing (Fig. 1 ). Yet, our utmost important goal for clinical human application has been accomplished by the conclusions drawn from Fig. 2 and Tables 2 and 3 . Third, only isolated IgG and IgM α2,6-SIA ratios, but not human blood IgG and IgM α2,6-SIA ratios, correlated with DAS28 (Fig. 2 ). Thus, these limitations could constrain the convenient use of this method for clinical use. Nevertheless, these findings open up a new direction for potential biomarkers to evaluate RA disease activity. Therefore, further investigation is warranted by enrolling a much larger population for study to possibly circumvent these limitations.
New biomarkers for RA are still desired for clinical practices,31 46 , 47
In conclusion, isolated IgG-α2,6-SIA/IgG and IgM-α2,6-SIA/IgM ratios correlated negatively and positively with DAS28 scores in RA patients, respectively, similar to those in CIA. The α2,6-SIA ratio of IgG-anti-CCP antibodies and IgMRF are potential biomarkers for evaluating RA disease activities. Altogether, we described a new potential way to explore concurrent SIA changes in blood IgG and IgM, especially in autoantibodies, to differentiate RA disease activities.
ACKNOWLEDGMENTS
This work was supported by Chang Gung Medical foundation, Taiwan (CMRPG3E1533 and CORPG2H0011). The funding source has no role in study design, in data collection and analysis, preparation of the manuscript, or decision to publish.
REFERENCES
1. Wernick R, Merryman P, Jaffe I, Ziff M. IgG and IgM rheumatoid factors in rheumatoid arthritis. Quantitative response to penicillamine therapy and relationship to disease activity. Arthritis Rheum. 1983; 26:593–8.
2. Tarkowski A, Nilsson LA. Isotype-specific measurement of rheumatoid factor with reference to clinical features of rheumatoid arthritis. J Clin Lab Immunol. 1983; 12:129–35.
3. Eberhardt KB, Svensson B, Truedsson L, Wollheim FA. The occurrence of rheumatoid factor isotypes in early definite rheumatoid arthritis–no relationship with erosions or disease activity. J Rheumatol. 1988; 15:1070–4.
4. Silvestris F, Goodwin JS, Williams RC Jr. IgM, IgA and IgG rheumatoid factors in patients with rheumatoid arthritis and normal donors. Clin Rheumatol. 1985; 4:392–8.
5. Vencovský J, Machácek S, Sedová L, Kafková J, Gatterová J, Pesáková V, et al. Autoantibodies can be prognostic markers of an erosive disease in early rheumatoid arthritis. Ann Rheum Dis. 2003; 62:427–30.
6. Bongi SM, Manetti R, Melchiorre D, Turchini S, Boccaccini P, Vanni L, et al. Anti-cyclic citrullinated peptide antibodies are highly associated with severe bone lesions in rheumatoid arthritis. Autoimmunity. 2004; 37:495–501.
7. Mullinax F, Hymes AJ, Mullinax GL. Molecular site and enzymatic origin of IgG galactose deficiency in rheumatoid arthritis and SLE. Arthritis Rheum. 1976; 19:813–9.
8. Parekh RB, Dwek RA, Sutton BJ, Fernandes DL, Leung A, Stanworth D, et al. Association of rheumatoid arthritis and primary osteoarthritis with changes in the glycosylation pattern of total serum IgG. Nature. 1985; 316:452–7.
9. Youings A, Chang SC, Dwek RA, Scragg IG. Site-specific glycosylation of human immunoglobulin G is altered in four rheumatoid arthritis patients. Biochem J. 1996; 314(pt 2):621–30.
10. Adebajo AO, Hazelman BL. IgG glycosylation in association with tropical infections and rheumatoid arthritis in the tropics. Clin Exp Rheumatol. 1995; 13:737–40.
11. McMillan SJ, Crocker PR. CD33-related sialic-acid-binding immunoglobulin-like lectins in health and disease. Carbohydr Res. 2008; 343:2050–6.
12. Casburn-Budd R, Youinou P, Hager H, Elkington D, Baxter I, Berthelot JM, et al. Asialylated IgG in the serum and synovial fluid of patients with rheumatoid arthritis. J Rheumatol. 1992; 19:1070–4.
13. Matsumoto A, Shikata K, Takeuchi F, Kojima N, Mizuochi T. Autoantibody activity of IgG rheumatoid factor increases with decreasing levels of galactosylation and sialylation. J Biochem. 2000; 128:621–8.
14. Scallon BJ, Tam SH, McCarthy SG, Cai AN, Raju TS. Higher levels of sialylated Fc glycans in immunoglobulin G molecules can adversely impact functionality. Mol Immunol. 2007; 44:1524–34.
15. Heyman B, Nose M, Weigle WO. Carbohydrate chains on IgG2b: a requirement for efficient feedback immunosuppression. J Immunol. 1985; 134:4018–23.
16. Nose M, Wigzell H. Biological significance of carbohydrate chains on monoclonal antibodies. Proc Natl Acad Sci U S A. 1983; 80:6632–6.
17. Kaneko Y, Nimmerjahn F, Ravetch JV. Anti-inflammatory activity of immunoglobulin G resulting from Fc sialylation. Science. 2006; 313:670–3.
18. Ohmi Y, Ise W, Harazono A, Takakura D, Fukuyama H, Baba Y, et al. Sialylation converts arthritogenic IgG into inhibitors of collagen-induced arthritis. Nat Commun. 2016; 7:11205.
19. Bartsch YC, Rahmöller J, Mertes MMM, Eiglmeier S, Lorenz FKM, Stoehr AD, et al. Sialylated autoantigen-reactive IgG antibodies attenuate disease development in autoimmune mouse models of lupus nephritis and rheumatoid arthritis. Front Immunol. 2018; 9:1183.
20. Inglis JJ, Criado G, Medghalchi M, Andrews M, Sandison A, Feldmann M, et al. Collagen-induced arthritis in C57BL/6 mice is associated with a robust and sustained T-cell response to type II collagen. Arthritis Res Ther. 2007; 9:R113.
21. Williams RO. Collagen-induced arthritis as a model for rheumatoid arthritis. Methods Mol Med. 2004; 98:207–16.
22. Parekh R, Isenberg D, Rook G, Roitt I, Dwek R, Rademacher T. A comparative analysis of disease-associated changes in the galactosylation of serum IgG. J Autoimmun. 1989; 2:101–14.
23. Bond A, Alavi A, Hay FC, Youinou P, Axford JS. Immunoglobulin G glycosylation changes occur in rheumatoid arthritis and primary Sjogren syndrome, but due to different mechanisms. Arthritis Rheum. 1993; 36(9 Suppl):S242.
24. Rajasekhar L, Liou LB, Chan CY, Tsai WP, Cheng CY. Matrix metalloproteinase-8 in sera and from polymorphonuclear leucocytes in rheumatoid arthritis: in vitro characterization and correlation with disease activity. Clin Exp Rheumatol. 2004; 22:597–602.
25. Anderson SJ, Scott DW. Generation of antigen-responsive B-cell line by somatic cell hybridization. Hybridoma. 1983; 2:211–8.
26. Ezaki I, Shingu M, Hashimoto M, Isayama T, Tohmatsu J, Kanda H, et al. Analysis of the genes encoding the variable regions of human IgG rheumatoid factor. J Rheumatol. 1994; 21:2005–10.
27. Dalziel M, McFarlane I, Axford JS. Lectin analysis of human immunoglobulin G N-glycan sialylation. Glycoconj J. 1999; 16:801–7.
28. Albert H, Collin M, Dudziak D, Ravetch JV, Nimmerjahn F. In vivo enzymatic modulation of IgG glycosylation inhibits autoimmune disease in an IgG subclass-dependent manner. Proc Natl Acad Sci U S A. 2008; 105:15005–9.
29. Espersen GT, Ernst E, Vestergaard M, Grunnet N. ELISA estimations of rheumatoid factor IgM, IgA and IgG in sera from RA patients with high disease activity. DTT treatment studies. Scand J Rheumatol Suppl. 1988; 75:40–5.
30. van Leeuwen MA, Westra J, Limburg PC, de Jong HJ, Marrink J, van Rijswijk MH. Quantitation of IgM, IgA and IgG rheumatoid factors by ELISA in rheumatoid arthritis and other rheumatic disorders. Scand J Rheumatol Suppl. 1988; 75:25–31.
31. Liou LB, Tsai WP, Chang CJ, Chao WJ, Chen MH. Blood monocyte chemotactic protein-1 (MCP-1) and adapted disease activity score28-MCP-1: favorable indicators for rheumatoid arthritis activity. PLoS One. 2013; 8:e55346.
32. Harris M, Taylor G. eds. Medical Statistics Made Easy. 2nd ed. Bloxham, UK: Scion Publishing Ltd; 2008.
33. Alavi A, Arden N, Spector TD, Axford JS. Immunoglobulin G glycosylation and clinical outcome in rheumatoid arthritis during pregnancy. J Rheumatol. 2000; 27:1379–85.
34. van de Geijn FE, Wuhrer M, Selman MH, Willemsen SP, de Man YA, Deelder AM, et al. Immunoglobulin G galactosylation and sialylation are associated with pregnancy-induced improvement of rheumatoid arthritis and the postpartum flare: results from a large prospective cohort study. Arthritis Res Ther. 2009; 11:R193.
35. Oefner CM, Winkler A, Hess C, Lorenz AK, Holecska V, Huxdorf M, et al. Tolerance induction with T cell-dependent protein antigens induces regulatory sialylated IgGs. J Allergy Clin Immunol. 2012; 129:1647–55.e13.
36. Hess C, Winkler A, Lorenz AK, Holecska V, Blanchard V, Eiglmeier S, et al. T cell-independent B cell activation induces immunosuppressive sialylated IgG antibodies. J Clin Invest. 2013; 123:3788–96.
37. Stümer J, Biermann MHC, Knopf J, Magorivska I, Kastbom A, Svärd A, et al. Altered glycan accessibility on native immunoglobulin G complexes in early rheumatoid arthritis and its changes during therapy. Clin Exp Immunol. 2017; 189:372–82.
38. Magorivska I, Muñoz LE, Janko C, Dumych T, Rech J, Schett G, et al. Sialylation of anti-histone immunoglobulin G autoantibodies determines their capabilities to participate in the clearance of late apoptotic cells. Clin Exp Immunol. 2016; 184:110–7.
39. Guhr T, Bloem J, Derksen NI, Wuhrer M, Koenderman AH, Aalberse RC, et al. Enrichment of sialylated IgG by lectin fractionation does not enhance the efficacy of immunoglobulin G in a murine model of immune thrombocytopenia. PLoS One. 2011; 6:e21246.
40. Sokka T, Toloza S, Cutolo M, Kautiainen H, Makinen H, Gogus F, et al.; QUEST-RA Group. Women, men, and rheumatoid arthritis: analyses of disease activity, disease characteristics, and treatments in the QUEST-RA study. Arthritis Res Ther. 2009; 11:R7.
41. Witte T. IgM antibodies against dsDNA in SLE. Clin Rev Allergy Immunol. 2008; 34:345–7.
42. Nowak UM, Newkirk MM. Rheumatoid factors: good or bad for you? Int Arch Allergy Immunol. 2005; 138:180–8.
43. Choe JY, Bae J, Lee H, Bae SC, Kim SK. Relation of rheumatoid factor and anti-cyclic citrullinated peptide antibody with disease activity in rheumatoid arthritis: cross-sectional study. Rheumatol Int. 2013; 33:2373–9.
44. Pope RM, Tellert DC, Mannik M. The molecular basis of self-association of antibodies to IgG (rheumatoid factors) in rheumatoid arthritis (IgG-rheumatoid factor/ultracentrifugal analysis). Proc Nat Acad Sci USA. 1974; 71:517–21.
45. Nardella FA, Teller DC, Mannik M. Studies on the antigenic determinants in the self-association of IgG rheumatoid factor. J Exp Med. 1981; 154:112–25.
46. Wolfe F. The many myths of erythrocyte sedimentation rate and C-reactive protein. J Rheumatol. 2009; 36:1568–9.
47. Pope JE, Thorne C, Haraoui BP, Truong D, Psaradellis E, Sampalis JS. Does C-reactive protein add value in active rheumatoid arthritis? Results from the Optimization of Humira Trial. Scand J Rheumatol. 2011; 40:232–3.
