Malondialdehyde-acetaldehyde adducts and anti-malondialdehyde-acetaldehyde antibodies in rheumatoid arthritis

Abstract

Objective: Malondialdehyde-acetaldehyde (MAA) adducts are a product of oxidative stress associated with tolerance loss in several disease states. This study was undertaken to investigate the presence of MAA adducts and circulating anti-MAA antibodies in patients with rheumatoid arthritis (RA).

Methods: Synovial tissue from patients with RA and patients with osteoarthritis (OA) were examined for the presence of MAA-modified and citrullinated proteins. Anti-MAA antibody isotypes were measured in RA patients (n = 1,720) and healthy controls (n = 80) by enzyme-linked immunosorbent assay. Antigen-specific anti-citrullinated protein antibodies (ACPAs) were measured in RA patients using a multiplex antigen array. Anti-MAA isotype concentrations were compared in a subset of RA patients (n = 80) and matched healthy controls (n = 80). Associations of anti-MAA antibody isotypes with disease characteristics, including ACPA positivity, were examined in all RA patients.

Results: Expression of MAA adducts was increased in RA synovial tissue compared to OA synovial tissue, and colocalization with citrullinated proteins was found. Increased levels of anti-MAA antibody isotypes were observed in RA patients compared to controls (P < 0.001). Among RA patients, anti-MAA antibody isotypes were associated with seropositivity for ACPAs and rheumatoid factor (P < 0.001) in addition to select measures of disease activity. Higher anti-MAA antibody concentrations were associated with a greater number of positive antigen-specific ACPA analytes (expressed at high titer) (P < 0.001) and a higher ACPA score (P < 0.001), independent of other covariates.

Conclusion: MAA adduct formation is increased in RA and appears to result in robust antibody responses that are strongly associated with ACPAs. These results support speculation that MAA formation may be a cofactor that drives tolerance loss, resulting in the autoimmune responses characteristic of RA.

Figure 1

Confocal colocalization images of MAA and citrullinated proteins in rheumatoid arthritis (RA) and osteoarthritis (OA) joint tissues. Frozen joint sections from 3 RA and 3 OA patients were stained for the presence of MAA, citrullinated proteins, and CD45 using IgG polyclonal rabbit anti-MAA antibody, IgM monoclonal mouse anti-citrulline antibody, and IgG2b anti-rat CD45: (A) RA sample stained for MAA proteins (B) RA sample stained for citrullinated protein (C) RA sample stained for CD45 (D) merged RA image of MAA proteins, citrullinated proteins and CD45 (E) OA sample stained for MAA proteins (F) OA sample stained for citrullinated protein (G) OA sample stained for CD45 (H) merged OA image of MAA proteins, citrullinated proteins and CD45. Images are at 63x power and created using a Zeiss 510 Meta Confocal Laser Scanning Microscope and anayized using ZEN 2012 software. Images from 1 RA and 1 OA patient shown; images from additional 2 RA patients and 2 OA patients shown in Supplemental Figure 1.

Figure 2

Box plots demonstrating RA case-control differences in anti-MAA antibody isotype concentrations; controls (CTRL; n = 80) matched to cases (n = 80), shown separately based on anti-CCP antibody status (CCP− and CCP+); RA-control matching based on age (mean of 51 years in both groups), gender (78% women), race (90% Caucasian), and smoking status (41% ever smokers)

Figure 3

Heatmap demonstrating relative expression of 19 antigen-specific ACPA and antibody to unmodified fibrinogen among anti-CCP positive RA patients; results reflect Statistical Analysis of Microarray (SAM) scores among those with highest relative to lowest tertile of anti-MAA antibody isotype (IgG, IgM, IgA); no increased expression observed for any isotype referent to unmodified fibrinogen, fibrinogen-cit, H2A/a-2 1–20 cit, H2B/A 62–81 cit cyclic, ApoE cit, biglycan 247–266 cot sm1 cyclic, or ApoE 277–296 cit2 sm2 cyclic; remaining ACPA significantly increased for ≥ 1 anti-MAA antibody isotype. Heatmap represents mean antigen-specific ACPA expression from all 1,720 RA cases that were analyzed separately for each anti-MAA antibody isotype.

Figure 4

Number of ACPA analytes positive (left) and total ACPA score (right) based on the quintile of circulating IgG anti-MAA antibody; positive threshold defined as two standard deviations (S.D.) above the mean value for RA patients; ACPA score defined as the sum of normalized fluorescent values divided by the number of analytes examined; points shown reflect means and bars reflect S.D.