Study design

In the first step of our investigation, using isolated circuiting neutrophils from healthy volunteers, we defined mechanisms involved in neutrophil TNF-pre-activation, LPS direct stimulation and IL-10-induced suppressor/tissue regeneration neutrophils. We made two assumptions: the first, that the antimicrobial and regenerative properties should be excluded as processes that do not co-occur; the second follows from the first, that the prior action of pro-inflammatory factors precludes the activity of IL-10 and vice versa – IL-10 prevents pre- and activation of neutrophils. To demonstrate this process, we stimulated neutrophils in a different order of TNF, LPS or IL-10. 2 × 106 of neutrophils were incubated in vitro without stimulation, and in the presence of 100ng/ml ultrapure LPS from E. coli (serotype R515, Alexis Biochemicals), 100ng/ml human TNF (recombinant, expressed in E. coli; Sigma-Aldrich, St. Louis, MO, USA), and 100ng/ml human IL-10 (expressed in Sf21 insect cells, Sigma-Aldrich, St. Louis, MO, USA) in RPMI 1640 for 5 h (5% CO2, 37 °C, humid atmosphere). In the second step, we verified the thesis by extrapolating obtained data to an adequate clinical status where circulating neutrophils are permanently exposed to pro-inflammatory factors and characterised by prolongated life spine: neutrophils isolated from the blood of sepsis patients as LPS-stimulated [20], Neuromyelitis Optica Spectrum Disorders as an autoimmune disease with pre-activated neutrophils [21] as the control of physiological response- neutrophils vulnerable to IL-10, circuiting neutrophils from periodontitis patients were used [7]. Isolated neutrophils were incubated for 5 h in RPMI under similar conditions as neutrophils used in the in vitro model.

There were four independent replicates for each experiment in the in vitro model. To validate results from ChIP-Seq analysis obtained from the in vitro model, three representative patients with functional test parameters closest to the mean values of their group were classified to ChIP-Seq analysis.

Assumptions for PTM selection

First, we selected modifications that seemed important in changing the nature of neutrophils from pro- to anti-inflammatory cells after IL-10 stimulation in a relatively short time (10 min). Therefore, we focused on PTMs marked active promotors and transcribed regions, assuming that neutrophils must be ready to differentiate the pro- vs. anti-inflammatory signal by triggering genes that engage chromatin-modifying enzymes. This process can be mediated by positioning the TSS region (H3K4me3) and active promotors (H3K36me3, H3K79me2) of transcribed regions. Simultaneously, we suspect that modification of transcription suppressor occurs to hide pro-inflammatory genes (H3K9me3, H3K27me3) and PTMs responsible for initiating genes within heterochromatin (H3S10p). Second, we selected modifications enhancing antimicrobial neutrophil properties observed during neutrophil TNF-preactivation. We suspect such neutrophils’ properties must be mediated by H3K27ac – PTM, described as a transcription enhancer. Third, as neutrophils leaving the bone marrow are already prepared to neutralise pathogens, we assume that LPS stimulation directly positions TSS regions by H3K4me3.

PatientsNeuromyelitis Optica Spectrum Disorders (NMOSD)

Three patients with confirmed AQP4-IgG seropositive NMOSD, fulfilling the 2015 Wingerchuk criteria, were selected for ChIP-Seq analyses [22]. All the patients were without systemic steroids or other anti-inflammatory drugs for at least three months before the study.

Sepsis

Six patients with sepsis developed at least two symptoms from the following criteria: body temperature above 38℃ or below 36℃; pulse rate > 90/min; respiratory rate > 20/min or PaCO2 < 32 mmHg; leukocyte > 12,000/µL or < 4000/µL were classified for investigations [23]. Patients who had not received antibiotics for at least three months before the study and with positive microbial tests for E. coli in the blood were selected for ChIP-Seq analyses.

Periodontitis

Twelve patients with generalised stage III and IV periodontitis were selected for this study. The criteria of the Classification of Periodontal and Peri-Implant Diseases and Conditions 2017 [24] were used for periodontitis staging and classified for further investigation. Patients who had not received antibiotics or anti-inflammatory drugs for at least three months before the study were selected for ChIP-Seq analyses.

All participants of the study were diagnosed and recruited at the Department of Neurology (patients with NMOSD), at the I Department of Urology (patients with sepsis) and at the Department of Periodontology and Oral Mucosal Diseases (patients with periodontitis and healthy controls), all from Medical University of Lodz.

Neutrophil isolation

20 mL of the whole blood on lithium heparin anticoagulant was collected from sepsis, NMOSD, periodontitis patients and HC. Neutrophils were purified by negative selection by microbeads, which allowed the removal of DCs, B cells, monocytes, macrophages, activated T cells, and activated NK cells (MACSxpress Whole Blood Neutrophil Isolation Kit, Miltenyi Biotec GmbH, Germany). Residual erythrocytes were lysed using 2mL ammonium chloride Lysing Reagent (BD Biosciences) for 5 min. The final purity of the neutrophil population was assessed by flow cytometry using CD14-PE (clone M5E2), CD15-FITC (MMA), and CD16-PECy7 (3G8, all from BD Pharmingen) mAbs. Flow cytometric analysis of the isolated cell population showed that the percentage of CD15highCD16+CD14− neutrophils was > 98%. The level of contaminating CD14+CD15+ monocytes was about 0.4%, and CD15+CD16− eosinophils was < 0.1% after isolation, as we proved in our previous studies [7, 9, 19].

Phagocytosis and reactive oxygen intermediates (ROI) production

With some modification, ready-to-use kits were used to analyse neutrophil phagocytosis and ROI production (Bursttest and Phagoburst, OrphoGen Pharma, Heidelberg, Germany). Briefly, isolated neutrophils were stimulated by different LPS, TNF and IL-10 constellations. Each stimulator was added to a 500 mL suspended cell for 10 min, 37 °C, in a water bath. The total time of neutrophil stimulation in this part of the experiment was 20 min. After incubation, samples were cooled down in an ice bath for 10 min. The ability of neutrophils to phagocytosis was performed on E. coli conjugated with FITC (3.3 × 108 bacteria/mL) and presented as the mean fluorescence intensity acquired from intercellular FITC signal. The ROI production was performed with or without restimulation by PMA (phorbol 12-myristate 13-acetate, 0.162 µM, 10 min) and presented as the mean intensity of 1,2,3-dihydrorhodamine (1,2,3-DHR), whose reduction depends on intracellular ROI concentration. Samples were analysed within 30 min using flow cytometry (BD LSRII, FACSDiva™).

Analysis of cell surface markers

The surface expression of CD11b, CD18 and CD62L were determined using BD LSRII flow cytometry and FACSDiva™ analysis software. After incubation with different stimuli combinations, the samples were washed, labelled with mAbs, fixed with 1% paraformaldehyde and analysed. For immunostaining, conjugated CD11b-PE (clone D12, BD Biosciences, CA, USA) CD18-FITC (clone 6.7, BD Biosciences, CA, USA), CD62L-FITC (clone SK11, BD Biosciences, CA, USA), antibodies and respective mouse isotype controls: IgG2a-PE (DAK-G05, Daco A/S, Denmark) and IgG1-FITC (DAK-G01, Daco A/S, Denmark) were employed.

Levels of post-translation histone modification by dot blot analysis

The stimulation of neutrophils (2 mln cells resuspended in 2mL RPMI) by different LPS, TNF and IL-10 constellations was stopped by ice-cold PBS rinsing. Subsequently, neutrophils were lysed on ice in Pierce™ RIPA buffer (ThermoScientific) containing Halt™ Protease & Phosphatase Inhibitor Cocktail (ThermoScientific) for 5 min and sonicated using Bioruptor® Pico Sonicator (Diagenode, five cycles; 30 s. “ON” 30 s. “OFF”). The lysates were collected and centrifuged at 12,000×g for 10 min at 4 °C. The protein concentrations were determined by a Nanodrop (ThermoFisher). The protein samples were transferred to 0.2 μm nitrocellulose membrane (Bio-Rad) for immunoblotting (20 µg to each reaction). The membrane was incubated for 50 min in RT with primary antibodies: anti-monomethyl-Histone H3 Lys4 (H3K4me1, rabbit, 1:100), anti-dimethyl-Histone H3 Lys79 (H3K79me2; clone NL59, 1:500), anti-phospho-Histone H3 Ser10 (H3S10p, clone MC463, 1:500), anti-trimethyl-Histone H3 Lys36 (H3K36me3, rabbit, 1:500), anti-trimethyl-Histone H3 Lys27 (H3K27me3, rabbit, 1:1000), anti-acetyl-Histone H4 Lys16 (H4K16ac, rabbit, 1:5000), anti-acetyl-Histone H3 Lys27 (H3K27ac, rabbit, 1:2000), anti-trimethyl-Histone H3 Lys4 (H3K4me3, rabbit, 1:500), anti-acetyl-Histone H3 Lys9 (H3K9ac, rabbit, 1:1000), anti-phospho-Histone H2 Thr120 (H2Ap, rabbit, 1:1000), anti-trimethyl-Histone H3 Lys9 (H3K9me3, rabbit, 1:500) and GAPDG (1:5000) as an endogenous control. All antibodies dedicated to ChIP-Seq analysis were obtained from Merck-Millipore. Subsequently, the membrane was ‘blocked’ by SuperBlock™ Blocking Buffer- Blotting in TBS. After 1 h of incubation, the RT membrane was 3x washed by TTBS buffer and secondary anti-rabbit IgG peroxidase conjugate antibodies were added for 35 min at RT (1:3000, Sigma Aldrich. The membrane was washed 2x by TTBS and 2x by TBS. Dot-blotting was performed using the ECL chemiluminescence peroxide solution (Bio-Rad). The densitometric scanning of the dot blots and analysis of the visualised spots was performed using a G-Box (Syngene) and Genesys Image Acquisition software (version 1.2.5.0). The results are presented as the intensity of optical density of the area underspot’spots’ peak (average value of the pixels enclosed after background correction) normalised to GAPDH.

Immunocytochemical analysis (ICC)

ICC analysis was performed on isolated neutrophils stimulated by LPS, TNF or IL-10 in different sequences (in vitro model) and on neutrophils incubated in RPMI and obtained from patients: NMOSD, sepsis or periodontitis (neutrophils stimulated in vivo). Neutrophils were transferred to poly-L-lysine-coated microscope slides by cytospin (300xg,10 min) and fixed with 4% formaldehyde solution for 6 min at RT. Fixed cells were washed with PBS and blocked with 10% rabbit blocking serum (Santa Cruz Biotechnology, Dallas, TX, USA) supplemented with 0.3% Triton™ X-100 (Sigma-Aldrich, St. Louis, MO, USA) for 45 min at 21℃. Next, they were washed and double-stained for SETD1A/MLL1, JMJD2A/CHD1, H3S10p, and H3K27ac. Anti-SETD1A (2 µg/ml, mouse, Proteintech, Manchester, UK), anti-MLL1 (KMT2A, 2 µg/ml, rabbit, Proteintech, Manchester, UK), anti-JMJD2 (1:100, D-9, mouse, Santa Cruz Biotechnology, Dallas, TX, USA), anti-CHD1 (2 µg/ml, mouse, Proteintech, Manchester, UK), anti-H3S10p (rabbit, clone MC463, 1:300, Merck-Millipore, Temecula, USA), anti-H3K27ac (rabbit, clone Lys27, 1:300, Merck-Millipore, Temecula, USA) and human IgG Isotype Control (Invitrogen, Cat.#31,154) as negative primary antibody control, were used. All antibodies were suspended in PBS supplemented with 1.5% blocking rabbit serum, 0.3% Triton X-100, and 0.01% sodium aide and incubated overnight at 4 °C. Cells were washed, and secondary fluorescent Abs were added one or 1 h at RT: goat pAb to mouse TR (5 µg/ml, cat. T862, Invitrogen, USA) with goat pAbs to rabbit FITC (2 µg/ml, cat. F2765, Invitrogen, USA). As isotype secondary antibody controls, goat IgG F(ab’)2 FITC (Invitrogen, Cat. #11,301 C) and rat IgG2a Texas Red (Invitrogen, cat# R2A17) were used. DAPI (1.5 µg/ml UltraCruz Mounting Medium, Santa Cruz Biotechnology, Dallas, TX, USA) was used for nuclei DNA staining. The confocal laser scanning microscopy platform TCS SP8 (Leica Microsystems, Germany) with the objective 63×/1.40 (HC PL APO CS2, Leica Microsystems, Germany) was used for microscopic imaging. Leica Application Suite X (LAS X, Leica Microsystems, Germany) was used for cell imaging. Fluorescence intensity was determined in the Region of Interest as the sum of the fluorescence from all segments (bordered by the line) divided by their number (arbitrary units- a.u.). The average fluorescence was calculated using at least 100 single cells for each sample. Nonspecific fluorescence (signal noise) was electronically diminished to the level when the nonspecific signal was undetectable (background). ICC data were additionally presented as the values of Colocalization Rate that indicate the overlap of the fluorescence signals between the channels FITC, TR and DAPI (nucleus). It was calculated as the mean value of a single Region of Interest using Leica Microsystem (LAS-X, ver. 3.7.020979 software, Leica, Germany).

H3K4me3 ChIP-Seq analysisChromatin immunoprecipitation (ChIP)

Neutrophils stimulated by LPS, TNF or IL-10 (in vitro model) and isolated from patients: NMOSD, sepsis, and periodontitis (neutrophils stimulated in vivo) were used to analyse DNA coupled with H3K4me3 histone. ChIP was carried out in neutrophils according to the manual of Magna ChIP™ A/G Chromatin Immunoprecipitation Kit (Merck Millipore). Cells were fixed with 1% formaldehyde in RPMI solution for 10 min at RT and quenched with 10x glycine in 5-minute incubation at RT to stop the fixation. After washing with cold PBS, cells were treated sequentially with 1x Protease Inhibitor Cocktail II, Lysis Buffer with Protease Inhibitor Cocktail II, and Protease Inhibitor Cocktail II with Nuclear Lysis Buffer. The supernatant was removed, and the cell pellet was resuspended in a Nuclear Lysis Buffer. Sonication (10 cycles; 30 s. “ON” 30 s. “OFF”) was done using Bioruptor® Pico Sonicator (Diagenode, Belgium). The obtained chromatin was spun at a minimum of 10,000 x g at 4 °C for 10 min to remove insoluble material. Each immunoprecipitation required the addition of Dilution Buffer and Protease Inhibitor Cocktail II. 25 µL of the diluted chromatin as ‘Input’ was saved at 4 °C for further proceeding. Chromatin immunoprecipitation was performed with the use of the set of antibodies: Normal mouse IgG (negative control), anti-RNA Polymerase II (clone CTD4H8) as positive control and anti-trimethyl-Histone H3 (Lys4) (MC315, Merck Millipore) mAbs. Both antibodies were recommended for use in the ChIP-Seq technique [25]. Immunoprecipitation reactions were incubated overnight at 4 °C with rotation. DNA was eluted and purified using spin columns. The DNA concentrations of obtained samples were measured by Qubit 4 Fluorometer (ThermoFisher Scientific).

Library preparation and NGS sequencing

Double-stranded DNA was generated from a single-stranded fraction of ChIPed DNA using NEBNext® Ultra™ II Non-Directional RNA Second Strand Synthesis Module (E6111S, New England Biolabs). The reaction was performed with random primers from NEBNext® RNA First Strand Synthesis Module (E7525, New England Biolabs). Libraries for sequencing were prepared using NEBNext® Ultra™ II DNA Library Prep Kit for Illumina® (E7645L, New England Biolabs). Single-end sequencing with a read length of 75 bases (SE75) was performed with NextSeq550 (Illumina) to obtain at least 20 million reads per sample that could be mapped to the human genome [26].

Bioinformatic methodology of the ChIP-Seq analysis

In the first stage, the quality of the raw sequence reads was checked using the FASTQC software (version: 0.11.8). Next, all reads were subjected to the adapter, and quality filtering (minimum quality (-q 25), minimum length (-m 15) using the Cutadapt tool (version: 1.18) in NextSeq reads mode. Trimmed reads were aligned to the reference genome (GRCh38) using the Bowtie2 (version: 2.2.9) in the single-end mode. Duplicated reads were located and tagged using the Picard MarkDuplicates tool (version: 2.18.4). Reads with low mapping quality score (MAPQ < 10) were removed from downstream analysis with the Samtools software (version: 1.6). Protein binding sites identification in the previously prepared BAM files was performed with the MACS2 (Model-based Analysis of ChIP-Seq) software (version: 2.1.0) in narrow peak mode [27]. Subsequently, identified peaks were annotated using annotatePeaks.pl from Homer software (version: 4.11.1, hg38 annotation library). Functional enrichment analysis for various categories (molecular function, biological process, cellular component and pathways interaction) was also executed [28]. To find enriched motifs in ChIP-Seq peaks, the findMotifsGenome.pl program from Homer software (version: 4.11.1) was used. The quantitative assessment of ChIP-Seq quality was checked by applying the ChIPQC package (version: 1.21.0) from R Bioconductor (version: 3.6.0). Differentially enriched sites between experimental conditions were identified using the DiffBind package (version: 2.12.0) from R Bioconductor (version: 3.6.0).

ChIP-Seq library quality control analysis and assessment of ChIP-Seq quality were provided in our previous studies [9].

mRNA analysis of SETD1, MLL, CHD1 and JMJD2A by real-time quantitative PCR

Total RNA was extracted from neutrophils with a mirVana™ miRNA Kit (Thermo Fisher Scientific). An Agilent small RNA kit was used to estimate purity and total RNA concentration (2100 Bioanalyzer, Agilent 2100 expert software). RNA was transcribed into cDNA using iScript Reverse Transcription Supermix (BioRad, USA). cDNA was amplified in the presence of TaqMan specific primers: JMJD2A (F 5’TGCGGCAAGTTGAGGATGGTCT3’; R 5’GCTGCTTGTTCTTCCTCCTCATC3’), CHD1 (F 5’TCTCTTCCTGCCAAGGTTGAGC3’; R 5’TGCCCTTGGAACCTTTGCTGAG3’),

MLL(KMT2A) (F 5’GTGCTTTGTGGTCAGCGGAAGT3’;

R 5’TGTGAGACAGCAACCCACGGTG3’),

SETD1A (F 5’TCGAGAGGAAGCTGTGGATACC3’; R 5’CGCCATCTGAGTCAGCATACAG3’) using 7500 Real-Time PCR System (Applied Biosystems) according to the following program: 95 °C 10 min; 40 cycles of (95 °C, 15s; 60 °C, 60s). We used RT2 Real-Time™ SYBR Green/PCR Master Mix (Qiagen, UK) that contains all reagents and buffers required for qRT-PCR. The expression levels of the β-actin were used for the normalisation of the cDNA samples (Ct value). The relative expression to non-stimulated neutrophils was calculated using the ΔΔCt method.

Statistics

Arithmetic means and standard deviations were calculated for all parameters. A statistical analysis of differences was performed using the one-way ANOVA (comparison between ‘n.s.’, LPS- and TNF-stimulated neutrophils) or the two-way ANOVA test (analysis of IL-10 on LPS stimulation or TNF-preactivation processes). Tukey’s test was used for multiple comparisons as a post hoc test when statistical significance was identified in the ANOVA test. The statistical comparison between clinical groups (sepsis, NMOSD and periodontitis patients) and healthy volunteers was performed using the Student-t test. p ≤ 0.05 was considered as the significant difference.