Interaction of the antiphospholipid syndrome autoantigen beta-2 glycoprotein I with DNA and neutrophil extracellular traps

Abstract

Beta-2 glycoprotein I (β₂GPI) is a phospholipid-binding plasma protein and prominent autoantigen in antiphospholipid syndrome (APS). Here, we tested the hypothesis that β₂GPI might bind to not only phospholipids, but also cell-free DNA and neutrophil extracellular traps (NETs). We report that β₂GPI interacts with cell-free DNA from different species, as well as NETs, in a dose-dependent manner, retarding their migration in an agarose-gel electrophoretic mobility shift assay. We confirm the direct binding interaction of β₂GPI with DNA and NETs by ELISA. We also demonstrate that β₂GPI colocalizes with NET strands by immunofluorescence microscopy. Finally, we provide evidence that β₂GPI-DNA complexes can be detected in the plasma of APS patients, where they positively correlate with an established biomarker of NET remnants. Taken together, our findings indicate that β₂GPI interacts with DNA and NETs, and suggest that this interaction may play a role as a perpetuator and/or instigator of autoimmunity in APS.

Keywords: Antiphospholipid antibodies; Antiphospholipid syndrome; DNA; Lupus; Neutrophil extracellular traps; beta-2 glycoprotein I.

Figure 1:. β₂GPI retards the migration of DNA from multiple species through an agarose gel.

A-D, Purified β₂GPI or bovine serum albumin (BSA) protein was allowed to form complexes with E. coli DNA (A), salmon sperm DNA (B), human genomic DNA (C), or NETs (D) for 1 hour at 37°C. Samples were then resolved on a 0.5% agarose gel. Quantification is derived from 3 independent experiments; *p<0.05, **p<0.01, ****p<0.0001 as compared with the BSA lane by one-way ANOVA corrected for multiple comparisons. The DNA ladder was from New England Biolabs (1 kb Plus DNA Ladder) in panels A and B, and Invitrogen (1 Kb Plus DNA Ladder) in panels C and D.

Figure 2:. β₂GPI binds directly to DNAs from different species and to NETs by ELISA.

A-B, ELISA plates were coated with isolated DNA (E. coli, calf thymus, salmon testes [native, or fragmented and denatured], or human [PBMC]), or NETs (A), or controls (cardiolipin, glu-plasminogen [Glu-P], or gelatin (B). Human β₂GPI (purified in-house) was added at the concentrations shown, followed by anti-β₂GPI, and bound antibody was detected by alkaline phosphatase-conjugated goat anti-mouse IgG. Binding of β₂GPI (OD₄₀₅) to DNA, NETs, or controls is shown as pooled data for duplicate samples from two independent experiments. Bars denote the means ± standard deviation.

Figure 3:. Colocalization of β₂GPI with NETs.

Neutrophils were stimulated with PMA to induce NETs in the presence or absence of 10% healthy human serum (a source of β₂GPI) as indicated. After fixation, NETs were stained for DNA (blue), neutrophil elastase (green), and β₂GPI (red). Overlap is shown to the right with some examples highlighted by yellow arrowheads. Scale bars=100 μm.

Figure 4:. Elevated levels of β₂GPI-DNA complexes in patients with APS.

A, β₂GPI-DNA complexes in 52 healthy controls compared with 132 patients with APS. Individual data points and the median are plotted; *p<0.05 by Mann-Whitney test. B, β₂GPI-DNA complexes in 106 patients with primary APS compared with 26 patients with APS and concomitant systemic lupus erythematosus; *p<0.05 by Mann-Whitney test. C, Correlation between β₂GPI-DNA complexes and neutrophil elastase-DNA complexes (a biomarker for NET remnants) in 132 patients with APS; correlation was determined by Spearman’s method. OD=optical density.