Anti-Citrullinated Protein Antibody Reactivity towards Neutrophil-Derived Antigens: Clonal Diversity and Inter-Individual Variation

Abstract

Background: Why the adaptive immune system turns against citrullinated antigens in rheumatoid arthritis (RA) and whether anti-citrullinated protein antibodies (ACPAs) contribute to pathogenesis are questions that have triggered intense research, but still are not fully answered. Neutrophils may be crucial in this context, both as sources of citrullinated antigens and also as targets of ACPAs. To better understand how ACPAs and neutrophils contribute to RA, we studied the reactivity of a broad spectrum of RA patient-derived ACPA clones to activated or resting neutrophils, and we also compared neutrophil binding using polyclonal ACPAs from different patients.

Methods: Neutrophils were activated by Ca²⁺ ionophore, PMA, nigericin, zymosan or IL-8, and ACPA binding was studied using flow cytometry and confocal microscopy. The roles of PAD2 and PAD4 were studied using PAD-deficient mice or the PAD4 inhibitor BMS-P5.

Results: ACPAs broadly targeted NET-like structures, but did not bind to intact cells or influence NETosis. We observed high clonal diversity in ACPA binding to neutrophil-derived antigens. PAD2 was dispensable, but most ACPA clones required PAD4 for neutrophil binding. Using ACPA preparations from different patients, we observed high patient-to-patient variability in targeting neutrophil-derived antigens and similarly in another cellular effect of ACPAs, the stimulation of osteoclast differentiation.

Conclusions: Neutrophils can be important sources of citrullinated antigens under conditions that lead to PAD4 activation, NETosis and the extrusion of intracellular material. A substantial clonal diversity in targeting neutrophils and a high variability among individuals in neutrophil binding and osteoclast stimulation suggest that ACPAs may influence RA-related symptoms with high patient-to-patient variability.

Keywords: ACPA; NETosis; citrullination; neutrophil; rheumatoid arthritis.

Figure 1

ACPA binding to neutrophil-derived antigen is associated with NETosis. Neutrophils were exposed to different stimuli and NETosis was quantified using IncuCyte (A). Granularity and cell size were analyzed by flow cytometry (B). Neutrophils were activated by using different stimuli for 3 h, and binding of ACPA or control antibodies was studied by confocal microscopy. DNA was stained by DAPI (C). ACPA binding was also analyzed by flow cytometry under the same conditions, using monoclonal and polyclonal ACPA and control antibody staining on fixed and permeabilized samples. Dashed lines, placed with the help of the control antibody-stained samples, represent the cutoff for stained events (D). Representative results of three independent experiments are shown. A23187^high, A23187^med and A23187^low concentrations correspond to 25, 5 and 1 μM, respectively.

Figure 2

ACPA binding to intracellular epitopes in activated neutrophils. ACPA binding was studied to intact or fixed/permeabilized neutrophils that were either non-activated or exposed to A23187 for 45 min. Representative contour plots are shown with gated positive events (A) or cumulative data from three independent experiments, with staining detectable in activated and fixed/permeabilized (F/P) neutrophils only (B). Fine-specificity of the tested clones was analyzed to citrullinated (Cit), carbamylated (Carb) and acetylated (Acet) autoantigens (C). ACPA binding was analyzed by confocal microscopy to A23187-activated neutrophils in fixed/permeabilized or non-fixed samples (D). The effect of ACPA and control antibodies on NETosis was analyzed using IncuCyte; results are expressed after normalization to non-treated samples ((E), upper panel). Alternatively, the cells were exposed to a moderate dose (5 μM) of A23187 with or without the different antibodies, and values were normalized to samples treated with A23187 only ((E), lower panel). Average values calculated from three independent experiments are show.

Figure 3

ACPA binding to PAD2- or PAD4-deficient neutrophils. Neutrophils were isolated from wild-type, PAD2- or PAD4-deficient mice, activated with A23187 and then fixed/permeabilized and stained with different ACPA and control antibody preparations. Representative stainings are shown with contour plots, with antibodies characterized by more or less PAD4-independent binding capacity (A). Average values are shown from at least three independent experiments (B). PAD4 deficiency resulted in a significant decrease in neutrophil binding (p < 0.001 for all clones, except for C04 where p < 0.01).

Figure 4

The PAD4 inhibitor BMS-5 reduces ACPA binding to neutrophil-derived antigens Human peripheral blood neutrophils were pre-treated with the BMS-5 or DMSO for 30 min at room temperature, followed by their activation with A23187. The cells were thereafter fixed, permeabilized and stained with different monoclonal or polyclonal ACPA preparations. Average values are shown, calculated from at least three independent experiments. *** p < 0.001, ** p < 0.01, * p < 0.05.

Figure 5

Individual variability in ACPA cellular targets. ACPAs were purified from plasma samples obtained from eight individual RA patients. Fine-specificity patterns of the original plasma samples, ACPAs purified with CCP affinity chromatography and the ACPA-depleted IgG fractions (flowthrough, FT) were studied using a multiplex solid-phase peptide array platform (A). Peripheral blood neutrophils were activated with A23187 for 45 min, then fixed and permeabilized and stained with the different ACPA or the corresponding control (ACPA-depleted) IgG preparations (B). The same antibodies were also added to cultures of developing OCs. Osteoclast numbers were normalized to control samples cultured without antibodies; average values were calculated from three independent experiments (C). **** p < 0.0001, ** p < 0.01, * p < 0.05.