Sepsis Induces a Dysregulated Neutrophil Phenotype That Is Associated with Increased Mortality

Background Neutrophil dysfunction in sepsis has been implicated in the pathogenesis of multiorgan failure; however, the role of neutrophil extracellular traps (NETs) remains uncertain. We aimed to determine the sequential changes in ex vivo NETosis and its relationship with mortality in patients with sepsis and severe sepsis. Methods This was a prospective observational cohort study enrolling 21 healthy age-matched controls and 39 sepsis and 60 severe sepsis patients from acute admissions to two UK hospitals. Patients had sequential bloods for the ex vivo assessment of NETosis in response to phorbol-myristate acetate (PMA) using a fluorometric technique and chemotaxis using time-lapse video microscopy. Continuous data was tested for normality, with appropriate parametric and nonparametric tests, whilst categorical data was analysed using a chi-squared test. Correlations were performed using Spearman's rho. Results Ex vivo NETosis was reduced in patients with severe sepsis, compared to patients with sepsis and controls (p = 0.002). PMA NETosis from patients with septic shock was reduced further (p < 0.001) compared to controls. The degree of metabolic acidosis correlated with reduced NETosis (p < 0.001), and this was replicated when neutrophils from healthy donors were incubated in acidotic media. Reduced NETosis at baseline was associated with an increased 30-day (p = 0.002) and 90-day mortality (p = 0.014) in sepsis patients. These findings were accompanied by defects in neutrophil migration and delayed apoptosis. Resolution of sepsis was not associated with the return to baseline levels of NETosis or migration. Conclusions Sepsis induces significant changes in neutrophil function with the degree of dysfunction corresponding to the severity of the septic insult which persists beyond physiological recovery from sepsis. The changes induced lead to the failure to effectively contain and eliminate the invading pathogens and contribute to sepsis-induced immunosuppression. For the first time, we demonstrate that reduced ex vivo NETosis is associated with poorer outcomes from sepsis.

Heparin and EDTA vacutainers (Becton Dickinson) were transferred into separate 50ml sterile Falcon TM tubes (Becton Dickinson). 2% dextran (Sigma-Aldrich) was added (1ml for every 6mls of blood) and gently mixed prior to incubation for 30minutes at room temperature to sediment the erythrocytes.
To prepare the gradients, a working stock of Percoll was made by mixing 45 ml of Percoll (Sigma-Aldrich) with 5 ml of 9% (v/v) sodium chloride (Sigma-Aldrich).
The 80% Percoll was prepared by diluting 40 ml of the Percoll stock with 10 ml 0.9% (v/v) sodium chloride (Baxter) whilst 56% Percoll comprised of 28 ml of the working Percoll stock and 22 ml of 0.9% (v/v) sodium chloride (Baxter).
Gradients were then prepared by carefully layering 5 ml of 56% Percoll on top of 2.5 ml 80% Percoll in a 15ml sterile Falcon TM tube (Becton Dickinson).
The gradients and leucocyte-rich plasma were centrifuged at 220G for 20 minutes with no acceleration and no brake at room temperature. Neutrophils were removed from the 80% and 56% gradient interface and re-suspended then washed in phosphate buffered saline (PBS; Gibco Invitrogen, Paisley, UK) at 440G for 10minutes at room temperature. Post-centrifugation the supernatant was discarded and the cells re-suspended in either RPMI 1640 (Sigma-Aldrich) or appropriate media. Purity was assessed by cytospin and staining with Giemsa stain (Diff-Qik; Gentaur Europe, Brussels, Belgium) routinely yielding neutrophil purity of greater than 95% and a viability of >98% as evaluated by tryptan blue exclusion

Neutrophil Migration
Neutrophil migration was assessed using an Insall Chamber (Weber Scientific International Ltd, Teddington). This improved chemotaxis chamber allows the migratory dynamics of individual cells to be assessed and provides greater information than traditional methods that provide only quantitative measures of cell migration.
Isolated neutrophils from Lithium-Heparin vacutainers were re-suspended in RPMI 1640 (Sigma-Aldrich) at a concentration of 5 x10 6 /ml. Bovine Albumin Fraction V (Sigma-Aldrich) was added to the neutrophils at a final concentration of 1.125% v/v. Neutrophils were then placed on a cleaned (with 0.4M H 2 SO 4 ) albumin coated coverslip (22x22mm, Surgipath Medical Industries Inc. Europe) and allowed to adhere for 20 minutes at room temperature. Once adhered the coverslip was inverted and placed on a clean Insall Chamber pre-filled with sterile RPMI 1640 (Sigma-Aldrich). The RPMI 1640 (Sigma-Aldrich) was replaced with 100nM CXCL-8 (R&D systems, Abingdon, UK) as chemoattractants. Gradients were allowed to develop for 5 minutes before assessment of migration.
Real-time video microscopy using a Leica DMI6000B with DFC360FX camera was used to capture neutrophil migration. Images were captured every 20 seconds for 12minutes producing 36 frames. This time frame was considered optimal for migration studies. Images were analysed using Image J software (Wayne Rasband, NIH, Bethesda) using vector analysis. Images were divided into ten individual segments and one cell randomly selected within each segment for analysis. Therefore, ten randomly selected cells were analysed.

NETosis Experiments
Freshly isolated neutrophils from EDTA vacutainers (Becton Dickinson) were used in the quantification of NET formation. Neutrophils were re-suspended in RPMI 1640 (Sigma-Aldrich) supplemented with 2nM L-Glutamine, 100U/ml Streptomycin and 100ug/ml Penicillin (GPS; all purchased from Sigma-Aldrich), to ensure the culture remained sterile, at a concentration of 1 x10 6 /ml. 1 x10 5 cells were placed in a 96-well flat bottomed plate (Becton Dickinson) with an additional 75µl of RPMI 1640 with GPS (Sigma-Aldrich). Cells were then stimulated with 25nM PMA (Sigma-Aldrich) or an additional 25µl of the media was added for the negative control and incubated for 3 hours at 37°C supplemented with 5% CO 2 . Experiments were performed in quadruplicate.  <0.0001

Figure 1: Severe Sepsis is associated with reduced chemotaxis
A: The difference in chemotaxis between healthy aged controls, sepsis patients and severe sepsis patient on admission to hospital. B: the sequential changes in neutrophil chemotaxis in patients with sepsis (Day 1=28, day 4 N=18 and day 7 N=14) and severe sepsis (day 1 N=42, day 4 N=32 and day 7 N=21). Bars represent the median and IQR with the error bars from a Tukey's distribution. P-values from a Kruskal-Wallis test.