This manuscript presents data acquired from subjects enrolled into the Brain Biomarkers after Trauma Study, an ongoing prospective longitudinal observational study of adult trauma patients conducted at a single Major Trauma Center site in the UK (University Hospitals Birmingham NHS Foundation Trust, Birmingham). Ethical approval for the study was granted by the North Wales Research Ethics Committee–West (REC reference: 13/WA/0399, Protocol Number: RG_13-164).

Patient enrolment began in the pre-hospital setting, where on a 24/7 basis between March 2016 and October 2018, emergency care teams acquired blood samples from adult trauma patients (≥18 years) with a suspected injury severity score (ISS) ≥8 within 1 h of injury (defined as the time of phone call to emergency services). In the pre-hospital setting, blood samples were not taken from patients who were deemed unlikely to survive transportation to hospital. Post admission, patients were excluded if they were aged <18 years, if pre-hospital blood samples had been acquired >1 h post-injury and if clinical assessments confirmed either an ISS <8 or a previous diagnosis of neuro-degenerative disease. No patients received blood products in the pre-hospital setting.

Due to the nature of injuries sustained, patients were unlikely to provide informed consent for their participation at the time of study enrolment. Consequently, patient recruitment was performed under the guidance of the Mental Health Capacity Act 2005 for research in emergency situations and the Declaration of Helsinki. For patients who lacked capacity, an agreement for study participation was sought from a legal consultee (family member or clinician not directly involved in the study), with written consent obtained from the patient once they regained capacity. In instances where the patient did not regain capacity, data were retained in accordance with the agreement of the legal consultee.

In the pre-hospital environment, peripheral venous blood samples were acquired during the intravenous cannulation of patients or by venepuncture. Once taken, blood tubes were stored at room temperature (RT) until arrival at hospital, where analysis began within 1 h by a single laboratory researcher on a 24/7 basis. Additional blood samples were acquired 4–12 and 48–72 h post-injury. At all three time points, blood samples were collected into BD Vacutainers® (BD Biosciences, Oxford, UK) containing ethylenediaminetetraacetic acid, z-serum clotting activator or 1/10 volume of 3.2% trisodium citrate. Full blood counts were performed using a Sysmex XN-1000 hematology analyser (Sysmex UK, Milton Keynes, UK) that measures a white cell differential and IGs, which are defined as promyelocytes, myelocytes, and metamyelocytes. The analyser uses fluorescence dyes that label intracellular DNA and RNA, with the intensity of the fluorescence signal directly proportional to the nucleic acid content of the cell. Due to their higher RNA content, IGs are discriminated from mature neutrophils via their stronger fluorescence signal. Daily internal quality control measurements (XN check, Sysmex UK) and monthly external quality control samples (UKNEQAS, Watford, UK) ensured instrument performance.

Sixty-seven adults (mean age 31 years, range 18–80) served as a cohort of healthy controls (HCs). HCs were volunteers who were not taking any regular medication for a diagnosed illness and did not have an acute episode of infection prior to the time of sampling. The recruitment of HCs was carried out in accordance with the ethical approval granted by the University of Birmingham Research Ethics Committee (Ref: ERN_12-1184) with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki.

MtDNA and mtDAMPs were prepared from mitochondria isolated from the K562 tumor cell line (ATCC®, Teddington, Middlesex, UK) as described previously (24). MtDNA concentration and protein content within mtDAMPs were determined by spectrophotometry (Nanodrop 2000; Thermo Fisher Scientific, Paisley, UK) and preparations stored at −80°C prior to use.

Neutrophils were isolated by Percoll density gradient centrifugation (Scientific Lab Supplies, Nottingham, UK) with cell purity, which was routinely ≥99%, determined using a Sysmex XN-1000 hematology analyser. Neutrophils were re-suspended at concentrations of 1-10 × 10⁶/ml in phenol red free or phenol red containing RPMI-1640 media supplemented with 2 mM L-glutamine, 100 U/ml penicillin and 100 μg/ml streptomycin (GPS; Sigma-Aldrich, Dorset, UK), phenol red free or phenol red containing RPMI-1640 media supplemented with GPS and 10% heat-inactivated fetal calf serum [HI-FCS; hereafter referred to as complete medium (CM); Sigma-Aldrich], glucose free RPMI-1640 media supplemented with GPS (Gibco, Fisher Scientific UK Ltd, Loughborough, UK), Hank's balanced salt solution (HBSS) supplemented with calcium and magnesium (hereafter referred to as HBSS^+/+; Gibco, Life Technologies, Cheshire, UK) or HEPES buffer containing 1 mM Ca²⁺.

For mtDAMPs and mtDNA experiments, neutrophils were pre-treated for 15 min (37°C/5% CO₂) with 40 or 100 μg/ml mtDAMPs or mtDNA prior to secondary stimulation. Prior to inclusion in transmigration, lactate and ROS assays, neutrophils were pelleted, supernatants removed, and cells resuspended in specified media. To inhibit FPR-1 signaling, neutrophils were treated for 60 min (37°C/5% CO₂) with 2.5 μM cyclosporin H (CsH; Abcam, Cambridge, UK) or vehicle control, prior to mtDAMP stimulation. For compound C experiments, neutrophils were treated for 60 min with 200 μM compound C (Sigma-Aldrich) or vehicle control prior to mtDAMP and PMA treatment. To inhibit calcium-calmodulin-dependent protein kinase kinases (CaMKKs), neutrophils were incubated for 60 min (37°C/5% CO₂) with 2.5 μM STO-609 (Sigma-Aldrich) or vehicle control prior to mtDAMP treatment. To induce AMP-activated protein kinase (AMPK) signaling, neutrophils were treated for 60 min with 1 mM AICAR (Sigma-Aldrich) prior to PMA stimulation.

Neutrophils (2 × 10⁵ in phenol red free or phenol red containing RPMI + GPS or glucose free RPMI-1640 media supplemented with GPS) were stimulated with 25 nM PMA (Sigma-Aldrich) for 3 h at 37 °C/5% CO₂. Post-stimulation, supernatants were collected and centrifuged at 2,200 x g for 10 min at 4°C, after which the DNA content of cell-free supernatants was analyzed. Briefly, 100 μl aliquots of cell-free supernatant were incubated with 1 μM SYTOX Green dye (Life Technologies) for 10 min at RT. Fluorescence was measured using a BioTek Synergy 2 fluorometric plate reader (NorthStar Scientific Ltd, Sandy, UK) with excitation and emission set at 485 and 528 nm respectively. In our trauma-based studies, DNA quantification was performed using a λ-DNA standard curve (Fisher Scientific) with PMA-induced NET generation presented as DNA concentration after subtracting the readings obtained from untreated controls. For mtDAMP experiments, background fluorescence values acquired from SYTOX Green staining of mtDAMPs in the absence of neutrophils were subtracted from test readings, with NET production expressed as a fold increase above untreated controls.

2 × 10⁵ neutrophils in phenol red free or phenol red containing RPMI + GPS or glucose free RPMI-1640 media supplemented with GPS were seeded onto glass coverslips and incubated for 30 min at 37°C/5% CO2 to allow for cell adherence. Following a 3 h stimulation with 25 nM PMA (37°C, 5% CO2), samples were fixed for 30 min with 4% paraformaldehyde (37°C, 5% CO2), washed three times in phosphate buffered saline (PBS) and permeabilised with 0.1% Triton X-100 (Sigma-Aldrich). DNA was then stained with 1 μM SYTOX Green dye for 5 min, after which slides were washed once in PBS, mounted in fluoromount medium and visualized using a LEICA DMI 6000 B microscope (LEICA, Milton Keynes, UK) at x20 or x40 objective.

For ex vivo analysis of neutrophils isolated from trauma patients, ROS generation was assessed by lucigenin-amplified chemiluminescence. The effect of mtDAMP pre-treatment on ROS production was examined using luminol-amplified chemiluminescence. In both instances, 100 μl aliquots of neutrophils (1 × 10⁶/ml in HBSS^+/+) were dispensed into wells of a 96-well white-bottomed flat plate (BD Biosciences), pre-coated with PBS/2% BSA, that contained 25 μl of luminol (pH 7.3; final concentration 100 μM; Sigma-Aldrich) or lucigenin (final concentration 200 μM; Sigma-Aldrich) and 50 μl HBSS^+/+. Neutrophils were then stimulated with 25 nM PMA or vehicle control, after which ROS generation was assessed at 1 min intervals for 180 min using a Berthold Centro LB 960 luminometer (Berthold Technologies, Hertfordshire, UK). Experiments were performed in quadruplicate, with ROS production measured as relative light units and calculated as area under the curve (AUC).

Neutrophils (2 × 10⁶ in phenol red free RPMI + GPS) were stimulated for 1, 2, or 3 h (37°C, 5% CO₂) with 25 nM PMA or vehicle control. At each time-point, cell-free supernatants were harvested (800 x g, 5 min, 4°C) and samples stored at −80°C prior to analysis. Lactate concentration in 25 μl aliquots of supernatant was determined using a commercially available lactate assay kit according to manufacturer's instructions (Sigma-Aldrich).

Following a 15 min rest period at 37°C/5% CO₂, neutrophils (1 × 10⁶ in RPMI-1640 media without glucose) were stimulated for 60 min (37°C/5% CO₂) with 25 nM PMA or vehicle. With 10 min of the stimulation period remaining, the fluorescent glucose analog 2-N-7-nitrobenzen-2oxa-1,3-diazol-4-yl amino-2-deoxyglucose (2-NBDG; Thermo Fisher) at a final concentration of 100 μM was added. Post-incubation, samples were washed and cells re-suspended in glucose free RPMI in preparation for flow cytometric analysis, which was performed on a CyAnADPTM bench top cytometer (Dako, Cambridgeshire, UK). Ten thousand neutrophils were collected and FL1 mean fluorescence intensity values recorded.

Neutrophils (1 × 10⁷/ml) in HEPES buffer containing 1 mM Ca²⁺ were incubated for 30 min in a 37°C water bath with 3 μg/ml calcein-acetoxmethyl ester (calcein-AM, Fisher Scientific), after which cells were pelleted, supernatants removed and neutrophils re-suspended at 1 × 10⁷/ml in phenol red free CM. A total of 1 × 10⁶ neutrophils were dispensed into the upper chambers of polycarbonate membrane cell culture inserts with 3 μM pores (Corning, New York, USA) that had been pre-loaded into wells of a 24-well flat bottomed plate (BD Biosciences) containing pre-warmed phenol red free CM and 1 nM LTB₄ (R and D Systems, Abingdon, UK). Following a 90 min incubation at 37°C, cell culture inserts were removed and plates read immediately for calcein fluorescence using a BioTek Synergy 2 fluorometric plate reader with excitation and emission set at 485 and 528 nm respectively. Fluorescence intensities were converted into neutrophil numbers via the use of a standard curve that was generated from calcein-AM loaded neutrophils that had been incubated alongside the test samples in the conditions described above. The number of neutrophils measured in media in which no chemokine was added was subtracted from the numbers calculated for wells that contained 1 ng/ml LTB₄ in order to determine specific chemokine-mediated migration.

Freshly isolated neutrophils (1 × 10⁵ in CM) were stimulated with 100 μg/ml mtDAMPs or vehicle control for 15 min at 37°C in a humidified 5% CO₂ atmosphere. Post–treatment, samples were stained on ice for 20 min with the following mouse anti-human monoclonal antibodies or their concentration-matched isotype controls: 2 μg/ml fluorescein isothiocyanate (FITC)-labeled CD62L (clone DREG56; eBioscience, Hatfield, UK); 1 μg/ml CXCR1-FITC (clone eBIO8F1-1-4; eBioscience); 0.5 μg/ml R-phycoerythrin (PE)-labeled CXCR2-PE (clone eBio5E8-C7-F10; eBioscience) or 2.5 μg/ml allophycocyanin (APC)-labeled CD11b (clone ICRF44, BioLegend, London, UK). Post incubation, cells were pelleted (250 x g, 5 min, 4°C), supernatants discarded and neutrophils washed once in PBS/1%BSA. Following resuspension in PBS, samples were transferred to polypropylene FACS tubes for flow cytometric analysis, which was performed on an AccuriC6^TM bench top cytometer (BD Biosciences). Ten thousand neutrophils, gated according to their forward scatter (FS)/sideward scatter (SS) properties, were acquired for analysis, where receptor expression was measured as median fluorescence intensity (MedFI).

To determine signaling through AMPK and MAPK pathways, cell lysates prepared from 2 × 10⁶ resting neutrophils, 1 × 10⁶ neutrophils stimulated with either 25 nM PMA or 100 μg/ml mtDAMPs for 2–90 min (37°C/5% CO₂), or 2 × 10⁶ neutrophils stimulated with 100 μg/ml mtDAMPs for 5 min following 1 h pre-treatment with 2.5 μM STO-609 or 2.5 μM CsH were separated on 10 or 12% SDS-polyacrylamide gels. Following protein transfer to polyvinylidene difluoride membranes (Bio-Rad, Hertfordshire, UK), blots were probed overnight at 4°C with rabbit anti-human antibodies (Cell Signaling Technology, Massachusetts, USA) directed against phosphorylated AMPK (pAMPK), phosphorylated ERK1/2 (pERK1/2), phosphorylated P38 (pP38), lactate dehydrogenase A (LDHA), or pyruvate kinase (PKM2). Post incubation, membranes were washed in tris-buffered saline containing 0.001% tween (TBST) and incubated for 1 h at RT with a goat anti-rabbit secondary antibody conjugated to horse radish peroxidase (HRP; diluted 1:4000 in TBST; GE Healthcare, Buckinghamshire, UK). HRP activity was detected using enhanced chemiluminescence (Bio-Rad). To confirm equal loading of proteins, blots were probed with antibodies against total ERK 1/2, total P38 (1:1000; Cell Signaling Technology), or β-actin (1:5000, GeneTex, California, USA). Densitometry analysis was performed using Image J software (National Institutes of Health, Bethesda, MD, USA).

Serum was prepared from blood collected into BD vacutainers containing z-serum clotting activator. Following a 30 min incubation at RT, blood samples were centrifuged at 1,620 x g for 10 min at 4°C, after which serum was aliquoted and stored at −80°C until analyzed. ELISAs to measure serum concentrations of HMGB-1 (IBL International, Hamburg, Germany), mitochondrial encoded NADH dehydrogenase 6 (ND6; MyBioSource, San Diego, California, USA) and IL-33 (R and D Systems) were performed in accordance with manufacturer's instructions.

Statistical analyses were performed using GraphPad Prism® software (GraphPad Software Ltd, California, USA). Data distribution was examined using the Kolmogorov-Smirnov or Shapiro-Wilk normality test. For data that followed a normal distribution, paired student t-tests, a repeated measures ANOVA with Bonferroni multiple comparison post hoc test or a one way ANOVA with Dunnett's multiple comparison post hoc test were performed. For non-normally distributed data, a Wilcoxon matched-pairs signed rank test, a Friedman test with Dunn's multiple comparison post hoc test or a Kruskal-Wallis with Dunn's multiple comparison post hoc test was performed. For box and whisker plots, whiskers represent minimum and maximum values. Statistical significance was accepted at p ≤ 0.05.

1,070 adult trauma patients were screened for study inclusion, with 87 subjects enrolled into the study (Supplementary Figure 1). Of these, 62 patients with a mean age of 44 years (range 19–95 years) and mean injury severity score of 26 (range 9–57) had their immune function analyzed (Table 1). The mean time of pre-hospital blood sampling was 39 min post-injury (range 13–59 min).

Compared to neutrophils isolated from HCs, neutrophils acquired from trauma patients within 1 h of injury exhibited significantly enhanced basal NET generation (Figure 1A), a hyperactivity that was accompanied by significantly elevated serum concentrations of HMGB-1 (Figure 1B) and IL-33 (Figure 1C). By the 4–12 and 48–72 h post-injury time points, a significant reduction in basal NET production was observed (Figure 1A). In response to stimulation with PMA, patient neutrophils released significantly less DNA at all three sampling time-points when compared to HCs (Figure 1D). Fluorescence microscopy confirmed the impairment in NET generation (Figure 1E).

ROS generation is a non-redundant event in NET formation (17). Having observed trauma-induced alterations in both basal and stimulated NET formation, we examined the effect of injury on ROS production. In the absence of stimulation, patient neutrophils isolated 48–72 h post-injury exhibited significantly enhanced ROS production when compared to the response of neutrophils from HCs (Figure 2A). No difference in basal ROS generation was seen between HCs and patient neutrophils acquired ≤ 1 h or 4–12 h post-injury (Figure 2A). In response to PMA stimulation, there was a significant reduction in ROS production, relative to HCs, for neutrophils isolated from patients only at the 48–72 h post-injury time point (Figure 2B).

Compared to their mature counterparts, immature neutrophils exhibit impaired ex vivo NET production and reduced ROS production upon stimulation with inflammatory agonists (1, 27). Relative to the values recorded for HCs, trauma patients presented, at all sampling time points, with a significantly elevated frequency (Figure 2C) and absolute number (Figure 2D) of circulating IGs.

MAPK signaling is a prerequisite for PMA-induced NET production (20). Due to the significant lymphocytosis that occurs within minutes of traumatic injury (3), and the small blood volume collected from patients at the scene of injury, we were unable to isolate a sufficient number of neutrophils from pre-hospital blood samples to examine MAPK signaling. However, we found neutrophils isolated from patients 4–12 and 48–72 h post-injury exhibited significantly increased phosphorylation of P38 MAPK (Figure 3A) but not ERK1/2 (Figure 4A) in the absence of exogenous stimulation.

In response to treatment with PMA, neutrophils obtained from HCs exhibited a significant increase in P38 phosphorylation (Figures 3B–D). In contrast, no significant PMA-induced increase in P38 phosphorylation was observed for neutrophils isolated from trauma patients 4–12 or 48–72 hours post-injury (Figures 3B–D). Compared to untreated cells, neutrophils isolated from HCs and trauma patients at the 4–12 and 48–72 h post-injury time points displayed a significant increase in ERK1/2 phosphorylation following 5, 10, and 15 min of PMA stimulation (Figures 4B–D). However, across these three stimulation time points, the degree of ERK1/2 phosphorylation was significantly greater in neutrophils isolated from HCs (Figures 4B–D).

Confirming the results of a recent study that demonstrated a necessity for exogenous glucose in PMA-induced NET production (21), we found neutrophils cultured in glucose free media released significantly less DNA upon PMA stimulation than neutrophils stimulated in glucose containing media (Supplementary Figure 2). Based on our observation of a trauma-induced impairment in ex vivo NET generation following PMA treatment, we investigated the effect of injury on neutrophil glucose uptake. Using the fluorescent glucose analog 2-NBDG, enhanced basal glucose uptake was recorded for neutrophils isolated from trauma patients within 1 h of injury (Figure 5A), but in response to PMA stimulation, a significant trauma-induced impairment in neutrophil glucose uptake was seen at all sampling time points (Figure 5B).

We next examined whether injury impacted upon glucose metabolism, a non-redundant step in NET formation triggered by PMA stimulation (21). Using lactate production as a marker of neutrophil glycolytic activity, we measured lactate concentrations in supernatants collected from resting and PMA-stimulated neutrophils following a 3 h in vitro culture. Compared to HCs, neutrophils isolated from trauma patients at the 48–72 h post-injury time-point exhibited enhanced basal (Figure 5C) but impaired PMA-induced lactate production (Figure 5D). The increase in basal lactate generation was accompanied by a significant up-regulation in the expression of the glycolytic enzymes pyruvate kinase and lactate dehydrogenase A (Figures 5E,F).

Compared to the levels measured in samples from HCs, serum concentrations of the mitochondrial-derived protein ND6 were significantly increased in patients at all post-injury time points, confirming the release of mtDAMPs after trauma (Figure 6A). Demonstrating the immune stimulatory properties of mtDAMPs, we measured significantly reduced CD62L, CXCR1, and CXCR2 expression as well as increased CD11b density on the surface of mtDAMP treated neutrophils (Supplementary Table 1). These changes in neutrophil surface phenotype were accompanied by activation of ERK 1/2 MAPK signaling (Supplementary Figure 3A). The emerging concept of mtDAMP-induced tolerance of neutrophil function is based in part on experimental data that has shown prior activation of neutrophils with bacterial-derived N-formylated peptides results in impaired migration upon secondary stimulation (10). Confirming these findings, we found that neutrophils pre-treated with 40 or 100 μg/ml preparations of whole mtDAMPs exhibited significantly reduced transmigration toward the chemokine LTB₄ (Supplementary Figure 3B). In contrast, no impairment in migration was witnessed for neutrophils pre-treated with 100 μg/ml of purified mtDNA (Supplementary Figure 3C).

To determine whether prior mtDAMP treatment influenced PMA-induced NET production, fluorometric analysis was performed on cell-free supernatants collected from cultures of PMA stimulated neutrophils that had been pre-treated with mtDAMPs or vehicle control. Analysis revealed neutrophils pre-exposed to 40 or 100 μg/ml mtDAMPs released significantly less DNA following a 3 h stimulation with PMA than vehicle-treated controls (Figure 6B). Fluorescence microscopy confirmed this mtDAMP-induced inhibition of PMA-induced NET generation (Figure 6C). Interestingly, focusing upon neutrophils pre-treated with 100 μg/ml mtDAMPs, images revealed that despite a significant reduction in NET formation, the cells had lost their multi-lobed nuclear morphology following stimulation with PMA, presenting with decondensed nuclear material (Figure 6C). In contrast to whole mtDAMP preparations, neutrophils pre-treated with 40 μg/ml of purified mtDNA prior to PMA stimulation showed no impairment in NET production (data not shown). Interestingly, ROS production, which is a prerequisite for NET formation, was significantly higher upon secondary PMA stimulation for neutrophils pre-treated with 100 μg/ml mtDAMPs (Figure 6D).

In stimulated T cells, elevated intracellular calcium levels activate AMPK, a recently described negative regulator of PMA-induced NET formation (28, 29). As raised intracellular calcium levels are a feature of mtDAMP treated neutrophils (23), we determined the activation status of AMPK in neutrophils following mtDAMP stimulation. To do this, cell lysates, prepared from neutrophils stimulated for 2, 5, 10, and 15 min with 100 μg/ml mtDAMPs, were probed with a phospho-specific antibody directed against Thr172, a residue within the activation loop of AMPK. As shown in Figure 7A, mtDAMP treatment resulted in an immediate and persistent phosphorylation of residue Thr172. Treating neutrophils with the FPR-1 antagonist CsH prior to mtDAMP stimulation resulted in a significant reduction in AMPK phosphorylation, suggesting that N-formyl peptides drive mtDAMP-induced activation of AMPK (Figure 7B). In antigen challenged T cells, phosphorylation of AMPK requires the activation of calcium-calmodulin-dependent protein kinase kinases (CaMKKs), a class of serine/threonine protein kinases activated by increases in intracellular calcium (28). To investigate whether CaMKKs were involved in mtDAMP-induced phosphorylation of AMPK in neutrophils, we treated neutrophils with the CaMKK selective inhibitor STO-609 prior to mtDAMP stimulation. Compared to vehicle control, a significant impairment in mtDAMP-induced activation of AMPK was detected in neutrophils pre-treated with STO-609 (Figure 7C).

AMPK has recently been shown to be a negative regulator of PMA-induced NET formation (29, 30). To investigate whether AMPK signaling was involved in mtDAMP-mediated suppression of NET formation, we treated neutrophils with compound C, an inhibitor of AMPK, prior to mtDAMP exposure. Compared to vehicle control, significantly greater NET production in response to PMA stimulation was recorded for neutrophils pre-treated with compound C (Figure 7D).

Confirming results of previous studies that had shown AMPK to be a negative regulator of aerobic glycolysis (31), we measured significantly lower concentrations of lactate in supernatants collected from PMA stimulated neutrophils that had been pre-treated with the AMP mimetic AICAR when compared to vehicle control (Figure 8A). Given that aerobic glycolysis is a key metabolic event in PMA-induced NET formation (21) and our observation of reduced NET generation following PMA stimulation for neutrophils pre-treated with mtDAMPs (Figures 6B,C), we investigated whether AMPK activation triggered by mtDAMP exposure was associated with an impairment in neutrophil glycolysis. Following 1, 2, or 3 h stimulation with PMA, significantly lower lactate concentrations were measured in supernatants collected from neutrophils pre-treated with 100 μg/ml mtDAMPs (Figure 8B).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.