The Research Ethics Board (REB) of the Hospital for Sick Children has approved present study protocols. All the study participants (healthy donors) signed the informed consent. All related methods were performed in accordance with the REB guidelines and regulations.

Chemicals and reagents, including PMA, LPS (0111:B4 and O128:B12), A23187, ionomycin, NaCl, d-mannitol, d-sorbitol, and choline chloride were obtained from Sigma (St. Louis, MO, USA), unless otherwise stated. RPMI medium (Wisent Inc.) with 10 mM HEPES (pH 7.2) has been used as the standard medium for all experiments. Without modification, this medium contains 103.44 mM NaCl and 109 mM total chloride. In the clinical treatment of CF, the nebulization of up to 7% hypertonic saline results in significant mucociliary clearance (33, 34). In most of the clinical settings, approximately 3% hypertonic saline has been used in the pulmonary treatment of CF (35). In our study, 3% hypertonic saline corresponds to ~513 mM NaCl concentration. To test whether this range of NaCl influences NETosis, eight different concentrations (109, 115, 121, 134, 159, 209, 309, and 509 mM) of NaCl, or choline chloride were prepared in RPMI medium. The media with altered concentrations were used for NETosis, ROS, and different assays. To test the effects of tonicity and osmolarity on neutrophils, NETosis kinetics and ROS generation were assessed in the presence of non-ionic osmolytes d-mannitol and d-sorbitol. d-mannitol and d-sorbitol are known to increase osmolarity.

Human peripheral blood was drawn from healthy donors and collected in K2 EDTA blood collection tubes (Becton, Dickinson, Co.). Neutrophils were isolated from whole blood using PolymorphPrep (Axis-Shield) according to the manufacturer’s instructions with minor modifications as reported in previous studies (16, 21, 23, 25). After washing the neutrophils, residual red blood cells were lysed using a 0.2% (w/v) NaCl hypotonic solution for 30 s. Subsequently, an equal volume of 1.6% (w/v) NaCl solution with 20 mM HEPES buffer was added. In order to eliminate red blood cell debris, cells were washed twice with 0.85% (w/v) NaCl containing 10 mM HEPES. Then, neutrophils were resuspended in RPMI medium (Invitrogen) with 10 mM HEPES (pH 7.2) for further use. Finally, purity of the neutrophils was ascertained; cell density was quantified using a hemocytometer, cell purity by Cytospin preparations, and cell viability determined by trypan blue dye. Neutrophil preparations having >95% purity were used in all the experiments reported in this study. The normal neutrophil count was defined as 1−2 million purified neutrophils per mL of the collected blood. Donors who had high neutrophil count were excluded to avoid potential baseline activation due to any underlying inflammatory conditions.

NETosis (% DNA release) kinetics were assessed using Sytox Green, a cell impermeable DNA binding dye. The Sytox Green dye exhibits >500-fold fluorescence enhancement upon binding with nucleic acids. Therefore, a measure of the fluorescence emitted by the dye has been used as a proxy for DNA release. 50,000 cells in a volume of 50 µl cell suspension with 5 µM Sytox Green dye were seeded into 96-well plates. The salt and different non-ionic osmolytes (d-mannitol, d-sorbitol) conditions of corresponding wells were altered by adding an equal volume of RPMI media with different salt and osmolyte concentrations (109, 115, 121, 134, 159, 209, 309, and 509 mM). Immediately after changing the salt or osmolytes concentrations of the media, cells were either left untreated (−ve control) or activated by NOX2-dependent agonists (25 nM PMA; 25 µg/mL 0111:B4 LPS and 5 μg/mL O128:B12), NOX2-independent agonists (4 µM A23187; 5 µM ionomycin), or bacteria [Pseudomonas aeruginosa at 20 multiplicity of infection (MOI), Staphylococcus aureus at 20 MOI]. After cell activation, the changes in the fluorescence intensities were measured every 30 min for up to 210 min using a POLARstar OMEGA (BMG Labtech) at 504 nm excitation and 523 nm emission. To calculate the amount of NETosis as a function of percent DNA release, the fluorescence reading at time zero was subtracted from the fluorescence reading at every time point. This value was then divided by the amount of total DNA, as determined by the fluorescence values of neutrophils lysed with 0.5% (w/v) Triton X-100. These experiments were repeated replacing NaCl with, d-mannitol, d-sorbitol, or choline chloride. Assays were excluded if baseline activation was >30%.

Immunofluorescence confocal imaging was performed to confirm the induction of NETosis. 50,000 neutrophils in a volume of 100 µl were seeded into eight-well chamber slides (BD Falcon). Salt or osmolyte (d-mannitol or d-sorbitol) concentrations of the neutrophils were changed by adding equal volume of RPMI with different concentrations of salt and d-mannitol to adjust for 109, 115, 121, 134, 159, 209, 309, and 509 mM into corresponding wells. Neutrophils in different salt media were either left untreated (−ve control) or activated by PMA or LPS, and incubated at 37°C for 120 min. After the incubation time, cells were fixed with 4% (w/v) PFA overnight at 4°C. Following three washes with phosphate-buffered saline (PBS), cells were permeabilized using 0.05% (w/v) Triton X-100 for 15 min. Following another three washes with PBS, cells were blocked for 1 h using 5% (w/v) bovine serum albumin. Furthermore, the cells were washed thrice using PBS. Then, cells were incubated with a primary antibody against myeloperoxidase (MPO, ab25989, Abcam) at a 1:1,000 concentration for 1.5 h. The isotype controls, mouse IgG (eBioscience 14-4714-81; 1:100 dilution) was used to confirm the MPO signals and mouse MPO antibody specificity. After three washes, a secondary antibody and nucleic acid-binding DAPI dye (Invitrogen; at 1:100 dilution) was added for 45 min. The secondary antibody was a donkey anti-mouse (488 nm, ab98766, Abcam) at a concentration of 1:1,000. Cells were then washed three more times, and then mounted for imaging. Confocal images were taken with an Olympus IX81 inverted fluorescence microscope with a Hamamatsu C9100-13 back-thinned EM-CCD camera and Yokogawa CSU × 1 spinning disk confocal scan head with Spectral Aurora Borealis upgrade, four separate diode-pumped solid state laser lines (Spectral Applied Research, 405, 491, 561, and 642 nm). The images were analyzed using Volocity Software (Perkin-Elmer).

The effect of salt on ROS production by neutrophils was assessed by DHR123 dye (Life Technologies). ROS oxidize DHR123 into R123, which emits a green fluorescence signal (36); thus, the amount of the green fluorescence signal indicates the amount of ROS production. Cells were preincubated with 25 µM of DHR123 dye for 15 min at 37°C. These preloaded cells were seeded into 96-well plates, and salt concentrations were altered by adding equal volumes of different RPMI media with either NaCl or choline chloride (109, 115, 121, 134, 159, 209, 309, and 509 mM). These cells were either unstimulated (−ve control) or activated by PMA or LPS. Fluorescence was measured at 10-min intervals for up to 60 min using an Omega fluorescence microplate reader. For the visualization and qualitative analysis, the same DHR123 plates were used for the confocal imaging after counter staining the nuclei with DAPI.

The effect of salt on ROS generation was confirmed by using a cell permeant reagent DCFDA (also known as H2DCFDA). This fluorogenic dye measures ROS activity within the cell. After diffusion into the cell, DCFDA/H2DCFDA is deacetylated by cellular esterases to a non-fluorescent compound, which is later oxidized by ROS into 2′,7′-dichlorofluorescein (DCF). DCF emits green fluorescence which is directly proportional to the amount of oxidized DCFDA to DCF. The maximum excitation and emission spectra are 495 and 529 nm, respectively (37, 38). Neutrophils were preloaded with 5 µM of DCFDA dye and either left unstimulated (−ve control) or activated by PMA at different salt and d-mannitol concentrations. Fluorescence was measured at 10 min intervals up to 60 min using an Omega fluorescence microplate reader as previously described.

The neutrophils were prepared for immunostaining of apoptosis markers following the same procedure as described in the section titled “Immunofluorescence confocal imaging.” During immunofluorescence staining, anti-cleaved caspase-3 rabbit primary antibody (Asp175, Cell Signaling) at 1:1,000 dilution and goat anti-rabbit secondary antibody (Alexa Fluor 555, A21428, Life Technologies) at 1:1,000 dilutions were used. Further, these immunostained slides were imaged by confocal microscopy, as described previously.

A single colony of Pseudomonas aeruginosa (mPA01) and Staphylococcus aureus (RN4220) was picked from the respective agar plates and grown overnight in sterile LB-broth. The next day, the bacterial cultures were diluted 10-fold and further subcultured for three hours to eliminate the dead colonies or bacteria. Bacterial culture were harvested and washed three times in PBS (pH 7.4) by centrifugation at 5,000 g for 5 min at 4°C. After the third wash, the optical density (OD) of resuspension was determined at a wavelength of 600 nm. The bacterial concentration was determined by taking OD at 600 nm. The different MOI were determined by a formula (colony forming unit 10⁸ = OD600 × 30.88 − 99.607). NETosis was induced using either of these bacteria at an MOI of 20.

SNARF^®-4F 5-(and-6)-carboxylic acid (Thermo Fisher Scientific), dye was used to determine the intracellular pH changes in neutrophils. Carboxy SNARF-4F dye is typically used by exciting the dye at one wavelength (between 488 and 530 nm) while monitoring the fluorescence emission at two wave lengths (580 and 640 nm). The ratio of the fluorescence intensities of the two emission wavelengths 640/580 nm has been used as a proxy for intracellular pH (39). Neutrophils were preloaded with 10 µM dye for 15 min at 37°C and 5% (v/v) CO₂. After the incubation, preloaded cells were washed and resuspended in fresh RPMI medium at the concentration of 1 × 10⁶ cells/ml. 50,000 neutrophils in a volume of 50 µL were seeded into a clear bottom black 96-well plate (BD Biosciences). An equal volume (50 µL) of RPMI containing different concentrations of the NaCl was added into the corresponding wells. These neutrophils were either left unstimulated (−ve control) or activated by PMA or LPS. Then, dual emission spectra of SNARF were measured. The changes in pH were recorded every 10 mins up to 90 min. The emission spectrum of SNARF undergoes a pH-dependent wavelength shift; therefore, the ratio (640/580 nm) of the fluorescence intensities from the dye at two emission wavelengths was used to determine the intracellular pH of neutrophils.

Statistical analyses were performed using Graphpad Prism 7. Where applicable, One-way ANOVA with Dunnett’s and Tukey’s Post-tests, Two-way ANOVA with Bonferroni post-test, or t-test were used. A p-value of <0.05 was considered to represent a statistically significant difference. All data presented as mean ± SEM. The biological replicates and applied statistical tests are noted in the figure legend.

To determine the effects of hypertonic saline (NaCl) on NETosis, we resuspended the neutrophils in RPMI medium with varying concentrations of NaCl (final concentrations of 109, 115, 121, 134, 159, 209, 309, and 509 mM) and 5 µM of the Sytox Green fluorescent dye. These neutrophils were either untreated (−ve control) or activated by NOX2-dependent prototypic agonist PMA. The fluorescence emission by Sytox Green was used as a proxy for extracellular DNA formation, and the green fluorescence signal was recorded every 30 mins up to 210 min. The kinetics of DNA release showed that changing NaCl concentration from 109 to 159 mM had little effect on NETosis. Hypertonic saline from 209 mM (1.2%, w/v) to 509 mM (3%, w/v) suppressed both background NETosis and PMA-induced NETosis (Figures 1A,B). Particularly, 309 mM (1.8%, w/v) and 509 mM (3%, w/v) NaCl concentrations significantly suppressed NETosis. To confirm true NETosis, confocal images were obtained after staining the DNA with DAPI and immunostaining MPO. Nuclear morphology and the colocalization of MPO (green) with DNA (blue) clearly indicated the suppression of NETosis in both unstimulated and PMA-activated neutrophils at hypertonic NaCl concentrations (Figure 1C). In unstimulated neutrophils, a few neutrophils underwent NETosis. However, with increasing concentrations of NaCl, the nuclei of neutrophils became condensed. When cells were activated with PMA in isotonic (or physiological) conditions, most cells underwent NETosis. However, at hypertonic conditions (especially at 309 and 509 mM NaCl), few or no neutrophils underwent NETosis. Rather, the nuclei became condensed. The isotype control for MPO immunostaining (Figure S1 in Supplementary Material) and low magnification images with all NaCl concentrations (Figure S2 in Supplementary Material) further confirmed the specificity of the antibodies and the effect of hypertonic saline on neutrophils.

NOX2-mediated ROS generation is important for PMA-mediated NETosis (25, 40, 41). Therefore, we investigated the effect of NaCl concentration on ROS production with a plate reader assay using the DHR123 dye. Neutrophils in media with different NaCl concentrations (109, 115, 121, 134, 159, 309, and 509 mM) were pre-loaded with DHR123 were either left untreated (−ve control) or activated usnig PMA. Changes in green fluorescence (oxidized DHR123 or R123) were recorded every 10 min up to 60 min. The % ROS production was calculated by considering the ROS produced by PMA-activated neutrophils under normotonic conditions at 60 min as 100%. DHR123 assays showed a small amount of background ROS production in unstimulated neutrophils. This baseline ROS production was suppressed by hypertonic saline. Above 209 mM NaCl, ROS production was significantly suppressed (Figure 2A). As expected, neutrophils stimulated with PMA generated much higher amounts of ROS under normotonic conditions (Figure 2B). However, beyond 109 mM of NaCl, ROS production was significantly suppressed. Images of neutrophils stained with DHR123 and DAPI confirmed the ROS production under normotonic conditions, and the suppression of ROS production at hypertonic NaCl concentrations (Figure 2C). The low magnification images of neutrophils stained with DHR123 and DAPI for all eight salt concentrations (Figure S3 in Supplementary Material) showed the dose-dependent suppression of ROS production.

We further validated the effect of increasing NaCl concentration on ROS production in neutrophils using another dye, DCFDA (also known as H2DCFDA). DCFDA is oxidized by ROS into DCF, which emits green fluorescence. Neutrophils were preloaded with 5 µM DCFDA and were left untreated or activated by PMA at different NaCl concentrations. Fluorescence was measured every 10 mins up to 60 min. Similar to the results observed by DHR123 assay, DCFDA assays showed that hypertonic NaCl suppressed ROS production (Figures 2D,E). Therefore, hypertonic NaCl suppresses ROS production in neutrophils.

To determine whether providing oxidants to neutrophils could overcome the hypertonic NaCl-mediated suppression of ROS production, we used H₂O₂, a key oxidant generated by neutrophils to execute NOX2-dependent NETosis (42, 43). We have previously shown that 1 mM H₂O₂ induces NETosis (20). Therefore, we induced NOX2-dependent NETosis using PMA in neutrophils treated with different NaCl concentrations (109, 209, 309, 509 mM) and monitored DNA release using Sytox Green assay. After 120 min, 1 mM H₂O₂ was added and NETosis kinetics was monitored up to 210 min. Data showed that H₂O₂ treatment increased NET release by reversing the inhibitory effect of hypertonic NaCl on NETosis (Figures 2F,G). Collectively, these data (Figures 1 and 2) suggest that hypertonic NaCl suppresses NET formation by suppressing ROS production during NOX2-dependent NETosis.

Next, we determined whether the sodium or chloride ion in the compound NaCl is responsible for the suppression of ROS and NOX2-dependent NETosis. To test this, we substituted choline chloride for sodium chloride and performed NETosis experiments. The NETosis kinetics elucidated that increasing concentrations of choline chloride has the same overall effect as increasing the concentration of NaCl; choline chloride suppresses both baseline and PMA-induced NETosis (Figures 3A,B). In addition, the % ROS kinetics showed suppression of baseline and PMA-induced ROS production with increasing choline chloride concentrations (Figures 3C,D). Therefore, an increase in Na⁺ ion concentration may not be responsible for the effect of hypertonic saline on NETosis kinetics. Since an increase in equimolar concentrations of NaCl and choline chloride have comparable effects, it becomes clear that hypertonic saline-mediated suppression of NETosis could depend on the increase in chloride concentrations or osmolarity.

To determine whether osmolarity is a key factor in regulating PMA-mediated NETosis, we performed experiments in the presence of equimolar concentrations of non-ionic osmolytes such as d-mannitol and d-sorbitol. Neutrophils resuspended in RPMI media containing different concentrations of d-mannitol or d-sorbitol (109, 115, 121, 134, 159, 209, 309, and 509 mM) were either unstimulated (−ve control) or activated by PMA. The % DNA release kinetics showed that both d-mannitol (Figures 4A,B) and d-sorbitol (Figure S4 in Supplementary Material) dose-dependently suppressed PMA-mediated NETosis. Determining ROS production by DCFDA plate reader assays showed that increasing concentrations of d-mannitol also suppressed ROS production (Figures 4C,D). Collectively, these data suggest that increased osmolarity suppresses ROS production and subsequently suppresses NETosis.

Both PMA and LPS (E. coli strain 0111:B4; O128:B12) induce NOX2-dependent NETosis (25, 44). Therefore, we tested the effect of increasing NaCl concentrations in LPS-mediated NETosis. The NETosis kinetics as determined by the % DNA release showed a NaCl concentration-dependent suppression of LPS-mediated NETosis (Figure 5A). Confocal microscopy images of DAPI and MPO-immunostained neutrophils confirmed the suppression of NETosis at higher NaCl concentrations (Figure 5B).

To determine whether hypertonic NaCl also attenuates LPS-mediated ROS production, we determined ROS production in neutrophils by DHR123 and DCFDA assays. Hypertonic NaCl also suppressed ROS production in a dose-dependent manner (Figures 5C,D). Therefore, hypertonic NaCl suppresses LPS-mediated ROS production and subsequent NOX2-dependent NETosis.

To determine whether increasing osmolarity suppressed LPS-mediated ROS production and NETosis, we used the osmolytes d-mannitol and d-sorbitol. The % DNA release was suppressed at higher concentrations of d-mannitol (Figure 6A) and d-sorbitol (Figure S5 in Supplementary Material) in a concentration-dependent manner. Untreated or LPS-activated neutrophils in different d-mannitol concentrations were stained for DNA (DAPI) and immunostained for MPO. Images showed colocalization of MPO and DNA, which indicated LPS-induced NET formation. Increasing concentrations of d-mannitol also suppressed NETosis (Figure 6B).

Measuring ROS production by DCFDA assays showed that d-mannitol suppressed LPS-mediated ROS production in a concentration-dependent manner (Figure 6C). Choline chloride also suppressed LPS-mediated ROS production in a concentration-dependent manner (Figure S6 in Supplementary Material). These data suggest that increased osmolarity suppresses LPS-mediated NETosis in a dosage dependent manner through the suppression of ROS production. Collectively (Figures 1–6), an increased concentration of neither sodium nor chloride was solely responsible for the suppressive effect of hypertonic NaCl on ROS production. Increased osmolarity was sufficient to suppress ROS production and subsequent NOX2-mediated NETosis.

We and other have shown that intracellular pH is an important factor in regulating NOX2-dependent NETosis (45–47). Therefore, to verify whether hypertonic NaCl alters cellular pH, we measured the intracellular pH of neutrophils that were either untreated (−ve control) or activated by PMA or LPS in different NaCl conditions (109, 159, 209, 309, 509 mM) in pH preadjusted media (6.6 and 7.8). Ratios of SNARF duel wavelength measurements indicated that pH standard at pH 7.8 and 6.6 showed higher and lower ratios, respectively. These results reflect that the assay worked as expected. However, hypertonic saline conditions did not significantly change intracellular pH (Figure S7 in Supplementary Material). Therefore, the effect of hypertonic NaCl on suppressing NOX2-mediated NETosis does not occur by altering intracellular pH of the neutrophils.

Since neutrophils were not dying by NETosis at hypertonic NaCl concentrations, we examined whether these cells were dying via apoptosis. We conducted immunocytochemistry experiments to determine two key apoptotic makers, cleaved caspase-3 (cCasp-3) and condensed nuclei. The immunostaining experiments were performed after treating neutrophils (untreated or PMA-treated) with varying concentrations of NaCl for 210 min. Qualitative examination of nuclear morphology and cCasp-3 staining showed that higher salt concentrations suppressed NETosis by promoting apoptosis in unstimulated (−ve control) and PMA-activated neutrophils (Figure 7A). Quantitative analysis of the % total cell counts (based on the nuclear morphology, DNA decondensation, NETs release and cCasp-3 staining) confirmed the suppression of NETosis and promotion of apoptosis (Figure 7B). Hence, this dataset suggests that suppressing NOX2-dependent NETosis by hyperosmotic NaCl is associated with an increase in apoptosis.

Neutrophils could also undergo NOX2-independent NETosis. As previously discussed, calcium ionophores A23187 and ionomycininduce this type of NETosis (19, 21, 24, 48, 49). To determine the effect of higher NaCl concentration on NOX2-independent NETosis, we incubated neutrophils with A23187 or ionomycin (obtained from S. conglobatus) under different NaCl concentrations. Unlike NOX2-dependent NETosis, the suppressive effect of high NaCl concentrations on NOX2-independent NETosis was not clear. The effect of hypertonic saline was different at different time points and varied based on the agonist used. For A23187-induced NETosis, hypertonic saline increased NETosis at the early time points but not at later time points. In contrast, for ionomycin-induced NETosis, hypotonic saline suppressed NETosis at the later time points (Figures 8A,B). Furthermore, NaCl was replaced by d-mannitol or d-sorbitol to study the effect of osmolarity during NETosis mediated by A23187 and ionomycin. The Sytox green kinetics showed suppression of NETosis only for ionomycin condition at higher concentrations of d-mannitol (Figures 8C,D) and d-sorbitol (Figures 8E,F). Collectively, the NETosis kinetics data show that hypertonic saline and high concentrations of osmolytes suppressed ionomycin-mediated NETosis, but not A23187-mediated NETosis. Inconsistent effects of NaCl, d-mannitol, and d-sorbitol on calcium-ionophore-induced NETosis suggest that these molecules alter NOX2-independent NETosis in a complex manner.

To examine the relevance of hypertonic saline in bacteria-induced NETosis, we cultured neutrophils with NETosis-inducing ratios (MOI of 20) of Gram-negative (P. aeruginosa) and Gram-positive (S. aureus) bacteria in different NaCl and choline chloride concentrations. Under normotonic conditions the % DNA release analyses showed that both of these bacteria induced NETosis at 20 MOI. Increasing concentration of both NaCl and choline chloride suppressed P. aeruginosa-mediated NETosis (Figures 9A,B). This pattern becomes evident beyond 209 mM of NaCl or choline chloride concentrations. However, neither NaCl nor choline chloride suppressed S. aureus-induced NETosis (Figures 9C,D). Therefore, hypertonic saline could differentially regulate different bacterial infections.