Differentiation of HL-60 promyeloblast cell line into granulocyte-like cells by ATRA

The HL-60 promyeloblast cell line (CCL240, ATCC, Rockville, MD, USA) was grown in RPMI-1640 containing 10% heat-inactivated FBS and antibiotics at 37 °C in 5% CO2-95% air [25]. The HL-60 cells were differentiated to granulocyte-like cells by 10 µM ATRA (Sigma-Aldrich, St. Louis, MO, USA) and 25 ng/mL G-CSF (PeproTech, Rocky Hill, NJ, USA) in 24-well plates for 5 days [26]. The differentiated HL-60 cells (dHL-60) were confirmed by staining with PerCP-Cy5.5-anti-CD11b antibody (BD Biosciences, UK) as described in our previous report [27].

Identification of NET-MSU aggregate formation by different immunofluorescence antibody staining

For simulating the low protein content in the pre-attack joint fluid, the dHL-60 cells were previously cultured in 2% FBS in RPMI-1640 for 1 h before interaction with MSU crystals. The dHL-60 cells (1.5 × 106 cells/mL) were then cultured on 12-well chamber slides pre-coated with 0.01% L-lysine (Ibidi, Gräfelfing, Germany). Commercially available MSU crystals (InvivoGen, San Diego, CA, USA) are firstly diluted in RPMI (5 mg/ml) to form a uniform suspension after pipetting. Subsequently, they are further diluted in the medium to the required concentration according to experimental conditions. After interaction with MSU crystals from 12.5 to 400 µg/mL for 1–22 h, the NET-entrapped MSU crystals were firstly confirmed by their needle-shaped microcrystals with negative birefringence under compensatory polarized light microscopic observation (CPLM). The NET components in the aggregates were then identified by stain with DAPI (1:10,000, Invitrogen, Carlsbad, CA, USA) for DNA, Alexa Fluor 488-conjugated anti-neutrophil elastase (anti-NE) antibody (1:100, Paso Robles, California, USA), and rabbit anti-citrullinated histone H3 (CitH3, citrulline R2 + R8 + R17, 1:1,000, Abcam, Cambridge, MA, USA) overnight, followed by Alexa Fluor 594-conjugated goat anti-rabbit antibody (1:1,000, Invitrogen, Carlsbad, CA, USA) as secondary antibodies. After washing, the NET-MSU aggregates were observed under a fluorescence microscope.

Quantitation of NET area by SYTOX green stain after dHL-60 and MSU crystal interaction for 1–22 h

The viable dHL-60 cells were previously labeled with 1 µM CellTracker™ Red CMTPX probe (Thermo Fisher Scientific, Waltham, MA, USA) at 37 °C for 30 min. The stained cells (3.0 × 105 cells/0.2 mL) were then interaction with MSU crystals in micro-wells for 1–22 h. SYTOX Green (250 nM) stain was used for NET area measurement. The △NET area denoting the velocity of NET formation was also calculated. Time-lapse imaging (ImageXpress, Molecular Devices, CA, USA) was used for dynamically capturing the dHL-60-MSU interactions. The areas with green fluorescence (three or nine fields per well at 10X or 20X magnification) were analyzed with Fiji software for measuring the NET area (mm3/field or % field per well) [28,29,30].

Quantitation of pro-inflammatory IL-8 and TNF-α and anti-inflammatory IL-1RA cytokines in the culture supernatants for calculating the estimate inflammation score (EIS)

dHL-60 cells (3.0 × 105 cells/0.2 mL) were incubated with or without MSU (200 µg/mL) for various hours. The amounts of IL-8, TNF-α, and IL-1RA in the culture supernatants were measured by the respective ELISA kit (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions. The detection limit is 7.5 pg/mL for IL-8, 6.23 pg/mL for TNF-α, and 18.3 pg/mL for IL-1RA.

Estimate Inflammation Score (EIS) was developed as a simplified metric to estimate changes in inflammation [31, 32]. In this cell-based study, EIS was calculated by the following equation:

$$}\left\}\left( } \right)} \over }\left( } \right)}} + }\left( } \right)} \over }\left( } \right)}} - }\left( } \right)} \over }\left( } \right)}}} \right\}\,\,},$$

T (treatment): the mean production of cytokine by dHL-60 + MSU crystals (200 µg/mL).

C (control): the mean production of the same cytokine by dHL-60 + medium.

Disruption of DNA scaffold in the NET-MSU aggregates by treatment with DNase I

To disrupt the DNA scaffold in the NET-MSU aggregates, recombinant human DNase I (30 IU/mL, Abcam, Cambridge, MA, USA) was simultaneously added to the mixture of MSU (200 µg/mL) and dHL-60 (3.0 × 105 cells/0.2 mL), followed by incubation at 37 °C for 4–8 h. The structure changes of NET-MSU aggregates were observed and NET areas were measured after SYTOX Green stain (250nM, Invitrogen, Carlsbad, CA, USA) as described above.

Addition of new MSU and new dHL-60 into the 4 h-incubated dHL-60 + MSU mixture for another 18 h incubation mimicking the clinical scenario

After 4 h of incubation, the old supernatant was replaced with the same amount of new MSU and dHL-60 cells for an additional 18 h of incubation. After incubation, both NET area and the amount of pro-inflammatory and anti-inflammatory cytokines were measured for calculating EIS as described in the above paragraph.

Detection of the cellular oxidation-reduction (redox) state

The dHL-60 cells were previously stained with a FRET-based redox sensor (10µM, StressMarq Biosciences, Victoria, Canada) and were incubated with or without MSU (200 µg/mL) in micro-wells (3.0 × 105 cells/0.2 mL) for 4 h. The cellular redox state was monitored by a high-content confocal system (ImageXpress, Molecular Devices, CA, USA) in that blue fluorescence reflects redox reductive, whereas the green fluorescence denotes redox oxidative state in the cells. Nine images per well were analyzed by Fiji software for the average blue/green fluorescence intensity ratio. For more contrast, the SYTOX Red (2 nM, Invitrogen, Carlsbad, CA, USA) instead of SYTOX Green, was used for staining the NET area.

Detection of phosphorylated intracellular cytokine signaling regulators by Western blot

dHL-60 cells (3.0 × 105 cells/0.2 mL) were incubated with or without MSU (200 µg/mL) for 2–4 h followed by lysis in RIPA buffer containing protease inhibitor cocktail (Roche, Penzberg, Germany). The cell lysates were electrophoresed in 10% SDS-PAGE and were then transferred to PVDF membranes (Millipore, Billerica, MA, USA) and were probed by rabbit antibodies against ERK1/2, phospho-ERK1/2 (Thr202/Tyr204), SHP-1, phospho-SHP-1 (Tyr564), SHIP1, or phospho-SHIP1 (Tyr1020) (Cell Signaling Technology, Danvers, MA, USA), followed by HRP-conjugated anti-rabbit IgGs. After washes, the relative protein amounts were analyzed by densitometry.

Quantitation of the intracellular CISH and SOCSs mRNA expression by RT-PCR

Total RNAs were extracted from dHL-60 cells ± MSU after 1–4 h incubation by using TRIzol kit (Invitrogen, Carlsbad, CA, USA). These RNAs were then reversely transcribed to cDNA using iScript™ cDNA synthesis kit (Bio-Rad, Berkeley, CA, USA). qPCR was performed by a QuantStudio System (Applied Biosystems, Foster City, CA, USA) for 40 cycles in duplicate. The mRNA amounts were normalized to cyclophilin A. The primer pair sequences are shown in Table S1.

Identification of N1/N2 phenotype by flow cytometry after stain with surface marker CD54 for N1 and CD182 for N2

dHL-60 cells (3.0 × 105 cells/0.2 mL) ± MSU crystals (200 µg/mL) were incubated for 0.5–2 h. The MSU crystals were dissolved and the cells were concomitantly fixed in 4% formaldehyde. The fixed cells were washed with PBS, and then were stained with BV421-conjugated anti-CD54 (for N1), PE-conjugated anti-CD182 (for N2), PerCP-Cy5.5-conjugated anti-CD11b, and fixable viability dye eFluor™ 780 (BD Biosciences, Oxford, UK). The stained cells were analyzed by a Cytek Aurora flow cytometer (Cytek Biosciences, Fremont, CA, USA) for measuring the median fluorescence intensity of N1 and N2 polarized cells.

Statistical analysis

Data were statistically analyzed by Mann–Whitney U test using GraphPad Prism 8.0.2 (GraphPad Software Inc. San Diego, CA, USA). The data are displayed in bar graphs as means ± SEM. All experiments were repeated for more than three times. Statistical significance was defined as p < 0.05.