Epitope-Specific Anti-C1q Autoantibodies in Systemic Lupus Erythematosus

Abstract

Objective: In patients with systemic lupus erythematosus (SLE) complement C1q is frequently targeted by autoantibodies (anti-C1q), that correlate best with active renal disease. Anti-C1q bind to largely unknown epitopes on the collagen-like region (CLR) of this highly functional molecule. Here we aimed at exploring the role of epitope-specific anti-C1q in SLE patients.

Methods: First, 22 sera of SLE patients, healthy controls and anti-C1q positive patients without SLE were screened for anti-C1q epitopes by a PEPperMAP^® microarray, expressing CLR of C1q derived peptides with one amino acid (AA) shift in different lengths and conformations. Afterwards, samples of 378 SLE patients and 100 healthy blood donors were analyzed for antibodies against the identified epitopes by peptide-based ELISA. Relationships between peptide-specific autoantibodies and SLE disease manifestations were explored by logistic regression models.

Results: The epitope mapping showed increased IgG binding to three peptides of the C1q A- and three of the C1q B-chain. In subsequent peptide-based ELISAs, SLE sera showed significantly higher binding to two N-terminally located C1q A-chain peptides than controls (p < 0.0001), but not to the other peptides. While anti-C1q were associated with a broad spectrum of disease manifestations, some of the peptide-antibodies were associated with selected disease manifestations, and antibodies against the N-terminal C1q A-chain showed a stronger discrimination between SLE and controls than conventional anti-C1q.

Conclusion: In this large explorative study anti-C1q correlate with SLE overall disease activity. In contrast, peptide-antibodies are associated with specific aspects of the disease suggesting epitope-specific effects of anti-C1q in patients with SLE.

Keywords: autoantibody(ies); autoimmune diseases; complement; lupus nephritis; systemic lupus erythematosus (SLE).

Figure 1

Epitope mapping of the collagen-like region of C1q. Six patients with SLE and two healthy blood donors were screened for antibodies against peptides of the CLR of C1q (A-, B- and C-chain). The heatmap color represents the intensity of the antibody binding signal in each sample (column) to each peptide, named according to the position of their first AA on the C1q molecule (rows, left site). Patients in bold were anti-C1q positive at the time of blood collection, all others anti-C1q negative. (A) 7 AA peptides in cyclic confirmation. (B) 15 AA linear peptides.

Figure 2

Binding of IgG from SLE patients and healthy controls to candidate epitopes of the collagen-like region and correlation of autoantibodies among each other. (A–G) Graphs are named according to examined epitopes and show Tukey’s boxplots with whisker lengths of 1.5x interquartile range. Outliers are shown as dots. Since the data are markedly skewed, Y-axis is segmented. Cutoffs for positivity are indicated by dashed lines. Statistical significance was considered as *p ≤ 0.05, ****p < 0.0001 respectively, ns, not significant. (H) Correlation-plot showing spearman correlation coefficients of measured autoantibodies among each other.

Figure 3

Univariate Logistic Regression. Positivity in ELISAs as binary predictor and presence of disease manifestations as binary dependent variable. The graphs show odds-ratios and 95% confidence intervals of SLE manifestations. ESR, erythrocyte sedimentation rate; APL, antiphospholipid.

Figure 4

Comparison of the diagnostic performance between anti-C1q and anti-A09/A15 as determined by ELISA. ROC curves analyzing the diagnostic performance of anti-A09, anti-A15 and anti-C1q regarding the discrimination of SLE patients from healthy donors, SLE patients with active versus inactive disease and proteinuria versus no proteinuria. (A) ROC curves of anti-A09 and anti-A15. (B) ROC curves of anti-C1q.