Anti-dsDNA antibodies bind to TLR4 and activate NLRP3 inflammasome in lupus monocytes/macrophages

In this study, 72 patients from the First Affiliated Hospital, Sun Yat-sen University who fulfilled the American College of Rheumatology criteria for the classification of SLE [21] and 36 age and sex matched healthy donors were enrolled. Patients with comorbidities of cancers or infections were excluded. Disease activity was scored using SLE Disease Activity Index (SLEDAI) scoring system [22]. Only patients with moderate to severe disease activity were included in the study (defined as SLEDAI > 4) [23]. Demographic and clinical characteristics of the SLE patients are shown in Table 1.

Human peripheral blood mononuclear cells (PBMCs) were isolated from SLE patients or from healthy controls with Ficoll-Hypaque by density gradient centrifugation. Cells were cultured in antibiotic-free RPMI 1640 medium containing 10 % FCS, 2 mM glutamate, 50 mM 2-ME, and 10 mM HEPES buffer in 24-well plates. To obtain monocytes, PBMCs were seeded to the culture dishes for 3 h to let the monocytes to adhere. Three hours later, non-adherent cells were washed away with PBS. To obtain macrophages, isolated monocytes were cultured with culture medium with 10 ng/ml of M-CSF for 7 days as previously described [24]. Culture medium was changed every 2 days.

Anti-dsDNA antibody-positive sera were collected from active SLE patients. Healthy sera were used as controls. Polyclonal anti-dsDNA antibodies were isolated from sera of SLE patients and control IgG from healthy subjects by affinity chromatography as we previously described [25]. Immune complexes were precipitated with polyethylene glycol (3.5 % [wt/vol]). Serum samples were then diluted with PBS and added to native DNA-cellulose column (GE Biotech, USA). Non-DNA-binding fractions were flushed. DNA binding fractions were eluted with a linear NaCl gradient. IgG was isolated with protein G Sepharose affinity chromatography kits (GE Biotech, USA) and the purities of eluted IgG were confirmed by 10 % sodium dodecyl sulfate–polyacrylamide gel electrophoresis (≥90 %).

PBMCs from SLE patients or healthy controls were isolated and 1 × 10⁶ cells were seeded to 48-well culture plate. Non-adherent cells were washed away and cells were maintained at 37 °C in 5 % CO₂ in a humidified cell culture incubator. Monocytes isolated from PBMCs were stimulated with healthy serum, rheumatoid factor (RF)-positive serum from rheumatoid arthritis (RA) patients, anti-dsDNA antibody-positive serum from SLE patients, or purified anti-dsDNA antibodies for 16 h. In addition, monocytes from SLE patients or healthy controls were stimulated with anti-dsDNA antibodies or control IgG for 16 h. Macrophages differentiated from SLE monocytes were stimulated with anti-dsDNA antibodies or control IgG for 16 h. Unless otherwise stated, drug concentrations were used as the following: Mito-TEMPO ((2-(2,2,6,6-tetramethylpiperidin-1-oxyl-4-ylamino)-2-oxoethyl) triphenylphosphonium chloride) (100 μM) (Sigma, Shanghai, China), glyburide (20 μM) (Sigma, Shanghai, China), Ac-YVAD-CMK (50 μM) (Calbiochem, Hong Kong, China) or IκB kinase inhibitor peptide (200 μM) (Calbiochem, Hong Kong, China). Cells were stimulated with 100 ng/ml of LPS for 6 h and 5 μM of ATP were included for the last 2 h, which was used as positive control. Supernatant was collected and frozen at −80 °C until tested.

Monocytes/macrophages from SLE patients were acquired as described above. Cells were collected and incubated with anti-CD16/CD32/CD64 antibodies (Biolegend, San Diego, CA, USA) to block Fc receptor before incubated with control IgG or anti-dsDNA antibodies. Briefly, cells were incubated with control IgG (5 μg/ml) or different concentrations of anti-dsDNA antibodies at 4 °C for 1 h. Cells were washed and stained with APC-anti-Toll-like receptor-4 (TLR4) antibody (Biolegend, San Diego, CA, USA) at 4 °C for 30 min. Data were acquired by flow cytometer FACS Aria (BD Bioscience, San Jose, CA, USA).

One million THP-1 cells (a human monocytic cell line, ATCC, USA) were seeded to 6-well culture plates. Cells were stimulated with 100 nM of PMA for 18 h. Briefly, cells were washed and transfected with 50 nM TLR4 siRNA (Ruibo-bio, Guangzhou, China) together with Lipofectamine 2000 (Invitrogen, New York, USA) in 500 μl Opti-MEM I Reduced-Serum Medium (Invitrogen, New York, USA) at 37 °C in a CO₂ incubator. Cells were also transfected with 50 nM of scramble siRNA complexed with Lipofectamine 2000. The RNAi-Lipofectamine complex was removed after incubated for 6 h, and the cells were cultured overnight in RPMI plus 10 % FBS. Twenty-four hours after transfection, cells were maintained with serum-free medium for 24 h prior to use. Then THP-1 cells were incubated with 10 μg/ml of anti-dsDNA antibodies or IgG control for 16 h.

To determine the activation of caspase-1 intracellularly, a fluorescence-labeled inhibitor specific for caspase-1 (FLICA) probe (ImmunoChemistry Technologies, Bloomington, MN, USA) was used as previously described [26]. Cells were collected and incubated with FLICA in 37 °C for 30 min. Cells were washed and analyzed by flow cytometry. For measuring mitochondria membrane potential (ΔΨm), Rhodamine 123 (Sigma, Shanghai, China) was used as previously described [27]. 2,7-dihydrochlorofluorescein (DHCF) (Sigma, Shanghai, Chiba) was used to measure the production of ROS in the cells [28]. To detect the frequency of Th17 cells in the spleen, single cell suspension was prepared from the spleen and 1 × 10⁶ cells was stimulated with 50 ng/ml phorbolmyristate acetate (PMA) plus 500 ng/ml ionomycin (Sigma, Shanghai, China) for 5 h, with inclusion of 10 μg/ml of brebeldin A (eBiosciences, San Diego, CA, USA) in the last 2 h. Cells were fixed, permeabilized, and stained with PE-anti-IL-17A (BD Pharmingen, San Jose, CA, USA) after stained with FITC-anti-CD4 antibodies (BD Pharmingen, San Jose, CA, USA). To analyze the frequency of Foxp3⁺ Treg in the spleen, 1 × 10⁶ cells were stained with anti-CD4 antibody (eBioscience, San Diego, CA, USA). Cells were fixed, permeabilized, and stained with PE-Cy5-anti-Foxp3 antibody (eBioscience, San Diego, CA, USA). Data was analyzed by Flowjo software (Tree Star Inc, Ashland, OR, USA).

Cells were collected and protein was extracted. Briefly, extracted cellular proteins were loaded to SDS–polyacrylamide gels. Protein was electrotransferred onto polyvinylidinedifluoride membranes. The membranes were blocked with 5 % bovine serum albumin in TBST and incubated with anti-ASC (Novus, Littleton, CO, USA), anti-caspase-1 p10, anti-P2X7 receptor (P2X7R), anti-Cathepsin-B (Abcam, Hong Kong, China), anti-pro-IL-1 (Santa, Cruz Biotechnology, Dallas, TX, USA), or anti-GAPDH (Kangcheng, Shanghai, China) primary antibodies at 4 °C overnight. The membranes were then washed and incubated with horseradish peroxidase conjugated anti-rabbit IgG (Cell Signaling Technology, Beverly, MA, USA) at room temperature for 60 min, and the signals were detected by enhanced chemiluminescence.

Human serum collected from SLE patients or healthy controls were frozen at −80 °C until analyzed by Luminex assay (Invitrogen, New York, USA), as measured on a Bio-Plex system (Bio-Rad, Hercules, USA). IL-1β and IL-17A were included in the multiplex analysis kit. For the measurement of IL-1β and IL-17A in mouse serum, commercial ELISA kits (eBioscience, San Diego, USA) were used according to the manufacturer’s instructions.

The data was expressed as the mean ± SD. Statistical analysis was performed using SPSS 13.0 (SPSS Inc, Chicago, USA). The differences were assessed by t test, or one way ANOVA with or without repeated measurements followed by Bonferroni’s multiple comparison post test as appropriate. Correlation analyses were done by Spearman’s rank correlation test. Two-tailed p <0.05 was considered statistically significant.

Our previous study has shown that caspase-1 was activated in a mouse model of SLE [8]. To measure the activity of caspase-1 in active SLE patients, we used a fluorescence-labeled inhibitor probe (FLICA), which binds to intracellular active caspase-1 specifically. The medium fluorescence intensity (MFI) of active capase-1 in monocytes of active SLE patients was significantly higher than that of the healthy controls (Fig. 1a, b). Since the production of IL-1β and IL-17A is increased following NLRP3 activation, we then measured serum concentration of IL-1β and IL-17A in these patients. Serum levels of IL-1β and IL-17A in SLE patients were significantly higher than those of healthy controls (Fig. 1c, d). Serum level of IL-1β was correlated with capase-1 activities in active SLE patients (Fig. 1e). Interestingly, the MFI of active capase-1 was also correlated with serum anti-dsDNA antibody level (Fig. 1f), suggesting the possibility of anti-dsDNA antibodies in triggering NLRP3 inflamasome activation. The MFI of caspase-1 was also correlated with disease activity index, the SLEDAI (Fig. 1g).

Monocytes isolated from SLE patients were stimulated with different stimuli (serum from healthy controls, RF-positive serum from RA patients, anti-dsDNA antibody-positive serum from SLE patients and LPS + ATP). Anti-dsDNA antibody-positive serum stimulation resulted in the activation of inflammasome in monocytes. However, healthy control serum or RF-positive serum did not activate inflammasome as measured by flow cytometry by using FLICA (Fig. 2a). Previous study showed that anti-dsDNA antibodies from SLE patients stimulated the overproduction of IL-1 from mononuclear cells [17]. To study the mechanism of anti-dsDNA antibodies in the production of IL-1, anti-dsDNA antibodies isolated from active SLE patients were used to stimulate monocytes. There was marked activation of inflammasome in monocytes stimulated with anti-dsDNA antibodies as measured by FLICA (Fig. 2b). On the other hand, anti-dsDNA antibodies also activated inflammasome in monocytes from healthy controls (Fig. 2c). Monocytes from active SLE patients had higher activation level of NLRP3 inflammasome following anti-dsDNA antibody stimulation than that from healthy controls (Fig. 2c). Furthermore, anti-dsDNA antibodies stimulated the activation of NLRP3 inflammasome in monocyte-derived macrophages (Fig. 2d), resulting in the production of IL-1β (Fig. 2e). Glyburide, which specifically inhibits NLPR3 inflammasome activation [10], was used to inhibit NLRP3, and capase-1 inhibitor YVAD was used to inhibit caspase-1. Glyburide and YVAD significantly inhibited the activation of NLRP3 inflammasome and the production of IL-1β stimulated by anti-dsDNA antibodies (Fig. 2d, e). It implied that anti-dsDNA antibodies activated NLRP3 inflammasome in monocytes/macrophages from SLE patients.

Monocyte-derived macrophages from SLE patients were stimulated with anti-dsDNA antibodies for different duration. Macrophages were incubated with DCHF-DA and analyzed by flow cytometry. The production of ROS started to increase 1 h after anti-dsDNA antibody stimulation and continued to increase up to 4 h (Fig. 3a). In addition, anti-dsDNA antibodies from active SLE patients affected the function of mitochondria as evident by decreased mitochondria membrane potential (ΔΨm) (Fig. 3b). However, there was no increase in the expression of P2X7R and cathepsin-B, the upstream molecules of NLRP3 inflammasome (Fig. 3c). The activation of NLRP3 inflammasome stimulated by anti-dsDNA antibodies was inhibited by mitochondria-targeting antioxidant Mito-TEMPO. The expression of ASC and caspase-1 p10 and the production of IL-1β were significantly reduced by pre-treatment with Mito-TEMPO in the culture system (Fig. 3d, e).

Monocytes/macrophages are rich in Fc receptor (FcR). IgG can bind to cell surface by interacting with FcR through Fc region. To rule out the binding of IgG to FcR, FcR was blocked by using anti-CD16/CD32/CD64 antibodies before incubation with anti-TLR4 antibody. Cells were then pre-treated with anti-dsDNA antibodies before incubation with anti-TLR4 antibody. The binding of anti-TLR4 antibody to monocyte-derived macrophages was completely blocked by pre-treatment with anti-dsDNA antibodies but not with control IgG (Fig. 4a). The activation of NLRP3 inflammasome stimulated by anti-dsDNA antibodies was further confirmed in THP-1 cell line. TLR4-specific siRNA was used to deplete TLR4 in THP-1 cells and gene silencing was confirmed by flow cytometry (Fig. 4b). Anti-dsDNA antibodies stimulated the activation of NLRP3 inflammasome as evident by increased expression of Caspase-1 p10 in THP-1 cell line (Fig. 4c). The expression of caspase-1 p10 was significantly inhibited by TLR4-specific siRNA (Fig. 4c). The production of IL-1β was also significantly decreased (Fig. 4d). Interestingly, the expression of pro-IL-1 and ASC was also significantly inhibited by TLR4-specific siRNA. Furthermore, IκB kinase inhibitor peptide significantly decreased the expression of pro-IL-1 and the activation of NLRP3 inflammasome (Fig. 4c, d). These data implied that anti-dsDNA antibodies activated NLRP3 inflammasome by binding to TLR4 in macrophages.

Anti-dsDNA antibodies were injected into female (NZB × NZW) F1 mice. Healthy control IgG or vehicle was used as controls. Anti-dsDNA antibodies stimulated the activation of inflammasome in monocytes of mice receiving anti-dsDNA antibody injection (Fig. 5a). In the anti-dsDNA antibody-injected mice, serum concentration of IL-1β and IL-17A was significantly higher as compared to control IgG or vehicle injected mice (Fig. 5b, c). The frequency of Th17 cells significantly increased in anti-dsDNA antibody-injected mice (Fig. 5d, e). However, the frequency of CD4⁺Foxp3⁺ Treg cells significantly decreased in anti-dsDNA antibody-injected mice compared to control IgG or vehicle-injected mice (Fig. 5f, g).