Morphological and Compositional Analysis of Neutrophil Extracellular Traps Induced by Microbial and Chemical Stimuli

Summary

Presented here is a protocol for the induction and analysis of in vitro neutrophil extracellular traps (NETs). Quantification of DNA, cathelicidin (LL37), and enzyme activity yielded data that show the variability in the composition and morphology of NETs induced by microbial and chemical stimuli under similar controlled conditions.

Abstract

Neutrophils function as the first line of cellular defense in an innate immune response by employing diverse mechanisms, such as the formation of neutrophil extracellular traps (NETs). This study analyzes the morphological and compositional changes in NETs induced by microbial and chemical stimuli using standardized in vitro methodologies for NET induction and characterization with human cells. The procedures described here allow the analysis of NET morphology (lytic or non-lytic) and composition (DNA-protein structures and enzymatic activity), and the effect of soluble factors or cellular contact on such characteristics. Additionally, the techniques described here could be modified to evaluate the effect of exogenous soluble factors or cellular contact on NET composition.

The applied techniques include the purification of polymorphonuclear cells from human peripheral blood using a double density gradient (1.079-1.098 g/mL), guaranteeing optimal purity and viability (≥ 95%) as demonstrated by Wright's staining, trypan blue exclusion, and flow cytometry, including FSC versus SSC analysis and 7AAD staining. NET formation is induced with microbial (Pseudomonas aeruginosa, Staphylococcus aureus, and Candida albicans) and chemical (phorbol myristate acetate, HOCl) stimuli, and the NETs are characterized by DNA-DAPI staining, immunostaining for the antimicrobial peptide cathelicidin (LL37), and quantification of enzymatic activity (neutrophil elastase, cathepsin G, and myeloperoxidase). The images are acquired through fluorescence microscopy and analyzed with ImageJ.

Introduction

Neutrophils are the most abundant leukocytes in the bloodstream, playing an essential role during the clearance of pathogenic agents by several mechanisms, including the release of large chromatin structures composed of DNA and several nuclear, cytoplasmic, and granular antibacterial proteins^1,2. The direct antecedent describing this antimicrobial role of neutrophils was made by Takei et al.³ in 1996. These authors reported a new form of death different from apoptosis and necroptosis in neutrophils, showed morphological changes exhibiting nuclear rupture, followed by spilling out of the nucleoplasm into the cytoplasm, and an increase in membrane permeability from 3 h of incubation with phorbol myristate acetate (PMA)^2,3. However, it was not until 2004 that the term "neutrophil extracellular traps (NETs)" was used⁴.

NET formation has been observed in various conditions, such as bacterial, fungal⁵, viral⁶, and parasitic infections, for neutralizing, killing, and preventing microbial dissemination⁷. Other studies show that it can also occur in non-pathogenic conditions by sterile stimuli, such as cytokines, monosodium uric acid or cholesterol crystals, autoantibodies, immune complexes, and activated platelets⁷. Lipopolysaccharide (LPS), interleukin-8 (IL-8), and PMA were among the first in vitro stimuli described as NET inducers, and the in vivo NET involvement in pathogenic processes was demonstrated in two models of acute inflammation: experimental dysentery and spontaneous human appendicitis⁴. DNA is an essential NET component. Its appropriate structure and composition are necessary for the sequestration and killing of microorganisms by delivering a high local concentration of antimicrobial molecules toward the caught microbes, as demonstrated by a brief deoxyribonuclease (DNase) treatment that disintegrates NETs and their microbicidal properties⁴. Besides DNA, NETs comprise attached proteins such as histones, neutrophil elastase (NE), cathepsin G (CG), proteinase 3, lactoferrin, gelatinase, myeloperoxidase (MPO), and antimicrobial peptides (AMPs) such as the cationic pro-inflammatory peptide cathelicidin LL-37 among others^8,9. Such aggregates may form larger threads with diameters up to 50 nm. These factors can disrupt the microbial virulence factors or the integrity of the pathogen cell membrane; additionally, the AMPs can stabilize the NET-derived DNA against degradation by bacterial nucleases¹⁰.

The specific mechanisms regulating NET formation have not yet been completely clarified. The best-characterized pathway leading to NET release is through ERK signaling, which leads to NADPH oxidase activation and reactive oxygen species (ROS) production, as well as increased intracellular calcium that triggers activation of the MPO pathway. This in turn transforms hydrogen peroxide into hypochlorous acid, activating NE by oxidation^11,12. NE is responsible for degrading the actin filaments of the cytoskeleton to block phagocytosis and translocating them to the nucleus for processing by proteolytic cleavage and deamination by PAD4 that drive the desensitization of chromatin fibers, which associate with granule and cytoplasmic proteins, and are then released extracellularly⁷. These proteases include those released from the azurosome complex of the azurophil granules and other proteases such as cathepsin G¹³.

Depending on the morphological changes in neutrophils, NETs are classified into two types: suicidal or lytic NET formation leading to cell death⁴, and vital or non-lytic NET formation produced by viable cells mediated by a vesicular release of nuclear or mitochondrial DNA, with a remnant of an anucleated cytoplast with phagocytic capability^14,15. Generally, NETs composed of mitochondrial DNA present an elongated fiber¹⁴ morphology, while those structured of nuclear DNA have a cloud-like appearance³. However, it is not known how the neutrophil chooses its DNA origin. Contrary to previous studies that described the canonical pathways of NETs as requiring several hours, the vital pathway is rapidly activated in just 5-60 min¹⁵.

Despite these advances, the NET composition varies depending on the stimulus; for example, different mucoid and non-mucoid strains of P. aeruginosa induce the formation of NETs containing 33 common proteins and up to 50 variable proteins⁷. Thus, it is necessary to homogenize techniques that allow the generation of objective conclusions in research groups. This paper describes a protocol with various techniques that allow comparison and evaluation of the composition, structure, and morphology of NETs induced with different microorganisms: Staphylococcus aureus (gram-positive bacterium), Pseudomonas aeruginosa (gram-negative bacterium), and Candida albicans (fungus), as well as chemical stimuli (PMA, HOCl) in human neutrophils from healthy individuals. The representative results demonstrate the heterogeneity of NETs depending on their inducing stimulus under comparable in vitro conditions, characterized by DNA-DAPI staining, immunostaining for LL37, and quantification of enzymatic activity (NE, CG, and MPO).

Protocol

The blood samples were obtained as donations from clinically healthy participants after informed consent. All experiments were performed with the permission of the Human Research Ethics Committee of the Faculty of Biochemical Sciences, Universidad Autónoma 'Benito Juárez' of Oaxaca.

NOTE: The inclusion criteria in the study were indistinct sex and age, and clinically healthy according to participant responses to a questionnaire prior to taking a blood sample. A hematological analysis was performed to determine the cell count and rule out infections or anemia, as well as the C-reactive protein test to rule out inflammation in the donor.

1. Peripheral blood collection and obtaining the erythrocyte and leukocyte package

1. Collect 10 mL of peripheral blood by venipuncture in tubes with 1.8 mg/mL of K_2·EDTA as anticoagulant (see Table of Materials) from clinically healthy individuals after obtaining informed consent. Then, perform standard blood biometry and C-reactive protein test to rule out infection or inflammation, ensuring the quality of the sample.
2. Centrifuge the peripheral blood sample at 82 x g for 15 min to remove the platelet-rich plasma, followed by a second centrifugation at 630 x g for 5 min. Discard the remaining plasma to obtain the erythrocyte and leukocyte package.
3. Dilute it in 1:1 ratio (v/v) with 1x Dulbecco's phosphate-buffered saline (DPBS).

2. Polymorphonuclear neutrophil (PMN) purification using a double-density gradient

NOTE: Perform neutrophil purification immediately after the blood is collected, because they have a limited in vitro lifetime of about 8 h.

1. Deposit the following in a sterile 10 mL glass tube (see Table of Materials) in order: 1 mL of 1.098 g/mL density solution, 1 mL of 1.079 g/mL density solution (see Table of Materials), and then 4 mL of the diluted erythrocyte and leukocyte package. Pour over the walls without breaking the surface tension between the layers to prevent them from mixing.
2. Centrifuge at 320 x g for 20 min at 4 °C, avoiding acceleration/deceleration so that the high forces of the centrifuge do not disturb the gradient.
3. Aspirate the phase that corresponds to granulocytes (Figure 1A) by pipetting, and transfer to another sterile 10 mL glass tube. Wash with 4 mL of 1x DPBS at 300 x g for 10 min at 4 °C.
4. Discard the supernatant and treat the cells with osmotic shock to remove the remaining erythrocytes. Add 4 mL of 0.2% saline solution for 2 min at 4 °C, and centrifuge at 300 x g for 10 min at 4 °C. Discard the supernatant. Then add 4 mL of the isotonic solution (0.65% saline) for 5 min at 4 °C to restore membrane integrity, and centrifuge at 300 x g for 10 min at 4 °C.
   NOTE: The 0.2% saline solution is a hypotonic medium, with a lower solute concentration relative to that of the RBC intracellular medium. Contact with the hypotonicmedium allows water to diffuse into the RBC, leading to their swelling and hemolysis. This removal of RBC from the supernatant was confirmed by microscopic observation.
5. Remove the supernatant. Resuspend the cells in 4 mL of 1x DPBS to remove cellular debris, and then centrifuge at 300 x g for 10 min at 4 °C. Finally, resuspend the cell pellet in 2 mL of cold Hank's balanced salt solution (HBSS) buffer.

3. Neutrophil morphology and viability (Figure 1B)

1. Trypan blue exclusion test
   1. Dilute 5 μL of the cell suspension in 20 μL of 0.4% trypan blue (1:5 ratio). Count the cells in a Neubauer chamber and determine cell viability using an exclusion test. Consider the cells that maintain the integrity of their membrane without permeabilizing the dye as viable.
   2. Mount 5 μL of the cell suspension on a slide; dry and stain with Wright's stain for 15 s. Immediately fix the sample with phosphate buffer pH 6.4 for 30 s. Wash with sufficient distilled water and observe the morphology under an optical microscope (100x).
2. 7AAD-staining and flow cytometry analysis
   1. Add 1 x 10⁵ cells to flow cytometry tubes, and stain with 1 μL of 7AAD in 100 μL of FACS buffer (1x DPBS, 0.1% sodium azide, and 10% autologous decomplemented plasma) for 15 min at 4 °C in the dark.
   2. Wash with 500 μL of FACS buffer at 300 x g for 10 min. Fix the cells with 500 μL of 2% paraformaldehyde, and store at 4 °C until their analysis in the flow cytometer.
   3. For a dead cell control, fix 1 x 10⁵ cells with 200 μL of 4% paraformaldehyde for 30 min, and wash with 500 μL of 1x PBS at 300 x g for 10 min at 4 °C. Draw off the supernatant and discard. Then add 200 μL of 0.1% Triton X-100 for 1 h at 4 °C. Wash with 500 μL of 1x PBS and stain with 7AAD as in step 3.2.1.
   4. Using a flow cytometer (see Table of Materials), perform FSC versus SSC analysis to analyze cell purity and SSC versus 7AAD staining to analyze the cell viability. Read 3 x 10⁴ events in 100 μL of uptake volume at medium flow (1,000 cells/s) in the polymorphonuclear settings (FSC, 400-490 and SSC, 300-320).
   5. Analyze the captured data in the flow cytometer software (see Table of Materials), and determine the percentage of purity and positive cells for 7AAD in the polymorphonuclear population, presented through dot plots and histograms.

4. CFSE staining of microorganisms

1. Add 1 x 10⁸ bacteria or 1 x 10⁶ fungal pseudohyphae in 1.5 mL microtubes, and stain with 200 µL of 5 µM carboxyfluorescein succinimidyl ester (CFSE) dissolved in 1x PBS. Mix for a few seconds, and incubate at 37 °C for 10 min in the dark.
2. Stop the reaction by adding 500 µL of decomplemented plasma, and centrifuge at 620 x g for 10 min for pseudohyphae or at 1,800 x g for 10 min for bacteria.
3. Discard the supernatants and wash the pellets with 1 mL of 1x PBS with centrifugation as in step 4.2. Finally, resuspend the microorganisms in 250 µL of 1x PBS.
4. Prepare 50 µL aliquots in microtubes of 1.5 mL with 2 x 10⁷ bacteria (MOI: 100) or 2 x 10⁵ pseudohyphae (MOI:1) for NET induction.

5. NET induction

1. Place 10 mm x 10 mm sterile glass coverslips in a 24-well plate and cover with 10 µL of 0.001% poly-L-lysine for 1 h at room temperature. Wash twice with 100 μL of 1x PBS, air dry, and irradiate with UV light for 15 min.
2. Replace the HBSS solution of the neutrophil suspension in step 2.5 with RPMI 1640 medium supplemented with 10% autologous plasma. To the 24-well plate (step 5.1), add 350 μL of this cell suspension, for a final concentration of 2 x 10⁵ neutrophils/well.
3. Allow the cells to adhere to the bottom of the wells by incubating for 20 min at 37 °C with 5% CO₂.
4. Add the stimuli to induce NET formation in 50 µL: microbial stimuli-gram-positive bacterium S. aureus (ATCC 25923), gram-negative bacterium P. aeruginosa (ATCC 10145) at MOI 100, and pseudohyphae of C. albicans (ATCC 10231) at MOI:1; biochemical stimuli-PMA (200 nM) and HOCl (4.5 mM), and control with stimulus absent (50 µL of HBSS).
5. Obtain a final volume of 400 µL per well. Mix on a plate shaker at 140 rpm for 30 s, and incubate for 4 h at 37 °C and 5% CO₂.

6. Visualization of NETs by fluorescence microscopy

1. DNA and LL37 immunostaining
   1. After NET induction, remove the supernatants from the wells by pipetting carefully, and fix the cells with 300 µL of 4% paraformaldehyde for 30 min.
   2. Wash the cells with 200 μL of 1x PBS without centrifuging, and add 200 µL of blocking buffer (10% decomplemented plasma in 1x PBS) for 30 min.
   3. For LL-37 stain, permeabilize the cells with 200 µL of 0.2% Triton X-100 in 1x PBS for 10 min to allow the antibody to enter the cells. Wash 2x carefully with 1x PBS to remove the excess detergent.
   4. Mount the coverslips on glass slides (four coverslips on each slide). DNA stain the cells with 2 µL of DAPI (see Table of Materials), seal the coverslips, and store at -20 °C until their analysis by confocal fluorescence microscopy.
2. Acquisition and analysis of fluorescent images
   1. Take NET images to quantify their components, and use the corresponding filters in the confocal fluorescence microscope (see Table of Materials) to acquire the images with the computer's software.
      NOTE: Consider that the DNA is stained with DAPI (blue color), showing excitation at 360 nm and emission at 460 nm. The microorganisms are stained with CFSE (green color), which has an excitation of 492 nm and an emission of 521 nm. The LL37 peptide is labeled with anti-LL37 Alexa Fluor 594 antibody (red color), which has an excitation of 594 nm and an emission of 614 nm.
   2. Calibrate the microscope. Place the slide and focus using differential interference contrast (DIC) with normal light on. Choose Live to project the image on the monitor.
   3. Turn off the light and select the fluorochrome corresponding channel. For example, select filter 365 nm/blue for DAPI, 43 HE DsRed for Alexa 594, or 38 HE GFP for CFSE.
   4. Adjust the settings with the isotype control antibody for LL37 and unstained cells for DAPI and CFSE. Set the same exposure time, voltage, contrast, and lens settings to capture all the images under the same conditions.
      NOTE: In this study, the exposure time, voltage, and contrast were set at 1.0 ms, 4.0 V, and 0.0, respectively, with a 40x objective. These values can be adjusted to facilitate the best image capture for the samples.
   5. Select Snap to capture the image. Save five images (four extremes and the center) per well, and of the colocalization (merge) of DNA/LL37/CFSE.
   6. Define the three classes of pixels as background with the independent images of each color and analyze the Mean Gray Signal value per area with the Image J software.

7. Enzymatic activity quantification

1. In a 96-well plate, add 90 μL of cell suspension in HBSS containing 1 x 10⁵ neutrophils for NET induction, and incubate for 20 min at 37 °C and 5% CO₂.
2. Immediately, add 10 μL of the corresponding stimuli (concentration as in step 5.4) and incubate for 4 h at 37 °C with 5% CO₂.
3. Discard the supernatants and wash the cells with 100 μL of HBSS. Treat with 1 U/mL of DNase for 10 min at 37 °C to favor the release of DNA-protein structures, and centrifuge at 1,800 x g for 10 min.
4. Recover the supernatants and evaluate the enzyme activity in the supernatant using colorimetric reactions as previously described by White et al.¹⁷.
5. Determine the maximum enzyme activity of NE, CG, and MPO in neutrophils under the same experimental conditions without adding any stimuli for NET induction. Then, freeze the cell sample at -70 °C and thaw at 37 °C in a water bath, generating a temperature shock to favor the release of intracell proteins by cell lysis. Centrifuge at 1,800 x g for 10 min and recover the supernatants.
6. Add 50 μL of the supernatant to each well in 96-well plates, and then add 50 μL of each substrate as indicated in step 7.7.
7. Add 0.5 M of N-methoxysuccinyl-Ala-Ala-Pro-Val-p-nitro aniline as the substrate for NE, and 1 mM of N-succinyl-Ala-Ala-Pro-Phe-p-nitroanilide for CG. Incubate for 3 h at room temperature. For MPO, add 1.6 mM of 3,3', 5,5'-tetramethylbenzidine (TMB) and incubate for 30 min at room temperature.
8. Post-incubation, add 50 μL of the stop solution (0.5 M H₂SO₄) for MPO and measure the absorbance at 405 nm for NE and CG and 450 nm for MPO, using a spectrophotometer.
9. Compare the values obtained with the corresponding calibration curves and show the results of each condition relative to the maximum enzyme activity (100%).

8. Statistical analysis

1. Analyze the measurement data in triplicate for each independent experiment (n = 10) and perform an ANOVA for statistical analysis by comparing groups with a 95% confidence level.

Figure 1: PMN purification and NET induction protocol. (A) Plasma was removed from the peripheral blood to obtain the erythrocyte and leukocyte package and diluted 1:1 (v/v) with 1x DPBS. Then, 4 mL of the dilution was added along the wall to the double-density gradient tube, and centrifuged at 320 x g for 20 min at 4 °C, obtaining the separation of different cell layers and recovering the one corresponding to PMN. (B) The purified cells were counted, and their morphology was analyzed by Wright's staining. Viability was determined by trypan blue exclusion and 7AAD staining using flow cytometry. Once optimal neutrophil purity and viability were verified, NET formation was induced by adding microbes (S. aureus, P. aeruginosa, and C. albicans) or chemicals (PMA, HOCl) in 24-well plates for analysis by fluorescence microscopy with DAPI-DNA, anti-LL37 Alexa Fluor 594, and microorganism-CFSE staining. For enzyme quantification, NETs were induced in 96-well plates for 3 h and treated with DNase, followed by the addition of substrates for each enzyme: NE, CG, and MPO; color changes were quantified by spectrophotometry. DPBS = Dulbecco's phosphate-buffered saline; PBMC = Peripheral blood mononuclear cells; PMN = Polymorphonuclear neutrophils; NE = Neutrophil elastase; CG = Cathepsin G; MPO = Myeloperoxidase; PMA = Phorbol myristate acetate; HOCl = Hypochlorous acid. Please click here to view a larger version of this figure.

Representative Results

Purity and viability of neutrophils
The dynamic cellular phases are visualized in the tube from the double-density gradient purification. Within these layers, the layer corresponding to granulocytes is above the 1.079 g/mL density layer, distinguished from the phases of peripheral blood mononucleocytes (PBMCs) and erythrocytes (Figure 1A). The morphology of the purified cells was verified with Wright's staining by observing cells with segmented nuclei connected with thread-like filaments corresponding to mature neutrophils (Figure 2A). Due to the reduced half-life of neutrophils, viability is a crucial point for the following NET induction experiments. Hence, cell viability was verified with trypan blue exclusion staining, evaluating the integrity of the plasma membrane when the dye does not penetrate (Figure 2B). Additionally, 7AAD intercalates in cytosine and guanine bases of DNA and emits fluorescence, which facilitates the exclusion of dead cells (Figure 2C,D). These procedures resulted in a cell purity of 98.9% ± 0.06% and a cell viability of 95.5% ± 4.2% in 10 performed experiments.

Figure 2: Morphology and viability of freshly purified neutrophils. (A) After gradient purification, the typical neutrophil morphology was confirmed by Wright's staining (100x). The cells showed a multilobed nucleus characteristic of mature neutrophils in the bloodstream and (B) optimal viability by trypan blue exclusion without membrane disruption. (C) Analysis of SSC versus FSC by flow cytometry confirms a population corresponding to PMN, obtaining cell purity of 98.9% ± 0.06% (n = 10). (D) PMN dot plots are presented (unstained cells, left panels vs. 7AAD stained cell right panels) with their respective histograms (unstained-blue vs. stained-red), indicating a high cell viability of 95.5% ± 4.2% (n = 10) in the freshly isolated neutrophils compared the control cells permeabilized with 0.1% Triton (lower panels) to validate the results. The 7AAD dye binds to DNA and is excluded from viable cells, so detergent pretreatment caused alterations in the cell membrane and allowed the entry of 7AAD. Please click here to view a larger version of this figure.

Morphology and composition of NETs induced by chemical and microbiologic stimuli
Figure 3 shows representative images proving in vitro NET induction by the action of chemical (PMA, HOCl) or microbial (C. albicans, S. aureus, and P. aeruginosa) agents after 4 h of incubation (Figure 3A). The NET composition was determined by DNA/LL37 staining, and the images were captured for quantification of mean gray values in the area with ImageJ (Figure 3B,C). Suicidal (lytic) or vital (non-lytic) NET formation was described based on their morphology. In comparison, unstimulated control neutrophils cultured in HBSS for 4 h showed LL37 localized in the cytoplasm with condensed multilobed-segmented chromatin (Figure 3A 1-3), characteristic of these resting cells.

PMA is a potent ROS inducer through PKC activation, thus functioning as a positive control to induce suicidal (lytic) NET formation. The images of PMA-induced NETs show networks with abundant chromatin decondensation forming cloud-like structures associated with the suicide pathway (Figure 3A 4-6), in which the neutrophil dies due to abundant DNA being released into the extracellular space covering large proportions of the analyzed area. Within these cloudy areas covering the DNA, LL37 was colocalized, both intra- and extracellularly in high concentrations (Figure 3A 6, B-C). In contrast, HOCl is a physiological, biochemical compound generated as a product of the oxidative activity of MPO. NETs formed with HOCl showed minor DNA dispersion in the extracellular space with filamentous morphology that appears to be derived from the cell membrane (Figure 3A 7-9). Such morphologic characteristics correspond to the vital NET formation, and they did not show colocalization of LL37 with the DNA filaments; LL37 remained localized at the nuclear level (Figure 3A 7-9,C). These results indicate that under similar controlled conditions, these chemical agents show divergences in the composition of the released NETs, both DNA and LL37.

Analysis of NETs induced by microorganisms showed that pseudohyphae of C. albicans induced networks with low concentrations of DNA released into the extracellular space with filamentous structures, highlighting the presence of anucleated cytoplast-like structures characteristic of the vital NET formation (Figure 3A 10-13), which are not observed with HOCl and S. aureus that activate this same pathway. Meanwhile, LL37 colocalizes with DNA at the cellular level, and low concentrations are associated with pseudohyphae, delimiting their propagation. Both components did not show statistically significant differences from the control with HBSS (Figure 3B,C). S. aureus is one of the most common pathogenic bacteria worldwide, with the ability to induce NETs showing a vital-like morphology as previously reported in several studies^15,19. Abundant DNA was released in statistically significant concentrations as filamentous structures that colocalized with LL37 and formed bacterial clusters, indicating that the scaffold might delimit and favor elimination of the bacteria (Figure 3A 14-17). On the other hand, the gram-negative bacterium P. aeruginosa induced the release of large concentrations of DNA with a cloudy morphology associated with the suicide NET formation, and LL37 significantly colocalized with DNA and bacteria (Figure 3A 18-21,B,C).

Visualization of NETs by fluorescence microscopy showed that the NETs induced by HOCl, S. aureus, and C. albicans present vital-like morphology with filamentous structures. However, only S. aureus led to the release of statistically significant amounts of DNA. The morphology of the other two stimuli (PMA and P. aeruginosa)-induced NETs corresponds to the type of suicide NET formation, with statistically significant amounts of DNA and LL37 compared to unstimulated neutrophils. There were abundant cloud-like DNA structures, more significant in NETs induced by P. aeruginosa than those induced by PMA.

Figure 3: DNA and LL37 composition and morphology of NETs induced by chemical and microbial stimuli. (A) Human blood-derived neutrophils were stimulated with PMA (4-6), HOCl (7-9), pseudohyphae of C. albicans (10-13), S. aureus (14-17), P. aeruginosa (18-21), and an unstimulated control (1-3) in HBSS, for 4 h. NETs were visualized using DNA-DAPI (blue), anti-LL37 Alexa Fluor 594 (red), or isotype control, and microorganisms were pre-stained with 5 µM CFSE (green). Images show representative fluorescence microscopy results of 10 independent experiments performed in triplicate. (B-C) Graphs show averages ± SD of mean gray value of the signal per area for the indicated colors of five images (four extremes and the center) per well and analyzed with the ImageJ software for (B) DAPI-DNA and (C) Alexa Fluor 594-LL37 staining. ANOVA was performed to compare groups of experimental conditions with a confidence level of 95%. * = p ≤ 0.05. HBSS = Hank's balanced salt solution; PMA = phorbol myristate acetate; HOCl = hypochlorous acid. Adapted from Sosa-Luis et al.¹⁶ with permission. Please click here to view a larger version of this figure.

Enzymatic activity of NE, CG, and MPO in NETs induced by chemical and microbiological stimuli
Another objective of this study was to quantify the activity of enzymes described as central components in NETs, such as NE, CG, and MPO, which are abundant with antimicrobial activity derived from azurophilic granules. Therefore, NETs were induced and the cell networks were treated with DNase, favoring the release of structure-associated proteins, and enzyme activity was determined using colorimetric assays. The analysis showed residual activity for NE (2.3% ± 0.6%), CG (6.2% ± 2.7%), and MPO (21.4% ± 2.0%) associated with DNA structures derived from all NETs, compared to the maximum activity obtained from lysed resting neutrophils taken as 100% (Figure 4). NETs induced by most of the stimuli showed low enzyme activity values for NE and CG, except S. aureus for NE and PMA for CG, which showed statistical significance. On the other hand, the MPO activity increased in all the stimuli without any statistically significant difference between them, but was significant between the groups of chemical stimuli versus microorganisms.

These results demonstrate the feasibility of the presented protocol, providing information on the heterogeneity in NET composition by quantifying DNA, LL37, and enzyme activity (NE, CG, and MPO) with different stimuli under the same in vitro conditions. The results reveal that the neutrophils can differentially and specifically respond to each stimulus by forming suitable DNA structures with various antimicrobial proteins during NET formation.

Figure 4: Residual enzymatic activity in NETs induced by chemical and microbial stimuli. Colorimetric assays were used to evaluate the enzymatic activity in NETs after disengaging DNA-protein structures by DNase to analyze just NET-bound proteins. The graph shows the average ± SD of the percentage of residual enzymatic activity in NETs induced by each chemical or microbial stimuli compared to the maximum enzymatic activity (in the supernatants of the neutrophils lysed by freeze-thawing at -70 °C) obtained for NE (3.54 ± 2.52 U/mL), CG (34.90 ± 25.85 U/mL), and MPO (0.29 ± 0.13 U/mL). Basal shows the enzymatic activity in the supernatants of the neutrophils kept in HBSS buffer. Statistical analysis was performed with data from 10 independent experiments by ANOVA to compare groups of experimental conditions with a confidence level of 95%. * and ** = p ≤ 0.05 compared to all other stimuli. *** = p ≤ 0.05 comparing the signaled groups of stimuli. NE = neutrophil elastase; CG = cathepsin G; MPO = myeloperoxidase; PMA = phorbol myristate acetate; HOCl = hypochlorous acid. Adapted from Sosa-Luis et al.¹⁶ with permission. Please click here to view a larger version of this figure.

Discussion

A highly pure population of viable neutrophils must be obtained to induce the release of NETs since these cells have a limited ex vivo lifetime of 8 h on average, a period within which all the experiments must be performed. To this end, the ideal methodology is the double-density gradient to optimize the purification time by isolating nonactivated cells more responsive to exogenous stimulation, in contrast to Ficoll-Histopaque gradient or Dextran sedimentation techniques¹⁷. Another advantage of the double-density gradient method is its relatively low cost compared to high purity cell enrichment techniques such as fluorescence-activated cell sorting (FACS) and magnetically-activated cell sorting (MACS) that rely on the cell surface antigens, such as CD16^hi in neutrophils. Besides a high cost, these sorting methods may also cause unintended activation of the target cells during the antibody-antigen interaction, and increase the ex vivo time of neutrophils, as these two techniques are generally used after density gradient purification. Cell morphology and viability rule out alterations in the purified neutrophils and confirm that the observed changes are specific in response to the added stimuli versus the resting neutrophils under the same experimental conditions without stimulation (Figure 2). This corroborates the efficiency of the protocol as an aseptic and reproducible method for neutrophil isolation without altering the cell viability and morphology.This contribution of eosinophils, which are commonly found in blood with counts below 2%, was ruled out and did not alter the statistical significance of the results, as eosinophils were not observed in the sample on Wright's staining when analyzing various fields (Figure 2A).This could be confirmed by flow cytometry using the CD16 marker, which is not present in eosinophils but is present in high concentrations in neutrophils.

In the study of NETs, there are consensuses and discrepancies¹⁸; various studies have reported contradictory results in response to the same stimuli, which could be due to differences in the neutrophil isolation protocols, the prolonged experimentation time that activates the neutrophils spontaneously, the buffers, and different concentrations of the same stimulus used. Therefore, it is essential to homogenize the protocols with explicit details so that the protocols can be consistently replicated to allow comparisons. The most common methods used are immunocytochemistry and immunohistochemistry, which enable the identification of structures containing extracellular DNA colocalizing with granule-derived proteins¹⁸ that usually include components with bactericidal activity able to destroy virulence factors⁸, such as NE, MPO, CG, and LL37. Additionally, recent studies for monitoring NET formation recommend analyzing this process using real-time live-cell methods such as intravital microscopy^1,18. Few studies such as this study unify the experimental conditions to compare the composition of NETs under different stimuli (bacteria, fungi, and chemicals)^1,9, using the technique of fluorescence microscopy to visualize the dispersion of DNA-DAPI, LL37-Alexa 594, and its colocalization with microorganisms-CFSE. Likewise, the treatment with DNase releases only those proteins associated with the DNA structure of the NETs, so the spectrophotometric technique ensures quantifying the enzymatic activity of MPO, NE, and CG, ruling out that they come from degranulation.

NET morphology may be defined as suicidal or vital, with the possibility of differentiating them according to the dispersion adopted by DNA in the extracellular space. The suicidal pathway is characterized by a cloud-like structure, while the vital pathway is a filamentous form^15,19,20. Based on this, Figure 3 indicates that PMA and P. aeruginosa represent a suicidal NET type. These results validate this technique by replicating under our conditions what was described by Brinckman et al. with PMA, the most commonly used compound to induce NETs by triggering the activation of c-Raf, MEK, Akt, ERK, and protein kinase C (PKC), which in turn activate NADPH oxidase¹¹ and cause a sharp increase in intracellular ROS. In addition, the results rule out the possibility that external artifacts, incubation times, or the brief mechanical agitation used in this study caused the spontaneous in vitro release of NETs with a cloud-like appearance as some authors have mentioned¹⁷, since resting neutrophils without stimulation post 4 h showed a multilobed nucleus with intact chromatin (Figure 3A 1-3).

Similarly, NETs induced with HOCl, C. albicans, and S. aureus presented filamentous morphology, corresponding to the vital NET morphology (Figure 3A 7-17). This pathway is characterized by neutrophils responding in a unique and an oxidant-independent manner without cell lysis or even rupture of the plasma membrane, only budding of vesicles with nuclear DNA¹⁵. This result again validates the potential of this technique to discern both morphological appearances of NET production as suicidal or vital. Regarding NET composition, the results obtained using this methodology showed statistical significance in the differential requirements of LL37 for PMA and P. aeruginosa stimulation (Figure 3C). NE (only for S. aureus) and MPO activity were higher in the microorganisms versus the chemical group (Figure 4). This information again proves neutrophil selectivity in responding to each stimulus, as reported by Kenny et al.⁵ when comparing different stimuli in NET production in vitro. It was described that in some cases the production of NET needs ROS or has differential requirements of specific enzymes such as PAD4, while in others the presence of these molecules is irrelevant to their production⁵.

A limitation of the technique of quantifying DNA in the extracellular space described in this study is that DNA can also come from other forms of cell death with a necrotic morphotype, and it is impossible for this technique to discern between them. However, it has been reported that NETs can be differentiated from this process by confirming the presence of granular proteins¹⁸. Likewise, the origin of the DNA is unknown, which can be confirmed by amplifying organelle-specific, mitochondrial, or nuclear genes by PCR¹⁹. Another weakness is distinguishing NETs from other fibrous structures such as fibrin; such discernment would require sophisticated techniques such as high-resolution scanning electron microscopy²⁰. In conclusion, the presented techniques allow in vitro characterization (under similar conditions) of the composition and morphology of NETs induced with different stimuli, showing the selectivity of neutrophils in the release of some of their components (DNA, LL37, NE, CG, MPO) to a specific stimulus. These techniques can be adapted to evaluate the effect of exogenous soluble factors or cellular contact on such NET characteristics.

Materials

Name	Company	Catalog Number	Comments
24 Well plate for cell culture	Corning	3526
7-aminoactinomycin D (7-AAD)	BD Pharmingen	51-668981E
96 Well plate for cell culture	Costar	3596	Flat bottom
Agitator	CRM Globe	CRM-OS1
Antibody LL37	Santa Cruz Biotechnology	sc-166770
Blood collection tubes	BD VACUTAINER	368171	K2 EDTA 7.2 mg
Carboxyfluorescein succinimidyl ester (CFSE)	Sigma-Aldrich	21878
Centrifuge	Hettich	1406-01
Coverslip	Madesa	M03-CUB-22X22	22 mm x 22 mm
Dulbecco´s phosphate-buffered saline (DPBS)	Caisson	1201022
Falcon tubes 50 mL	CORNING	430829
Flow Cytometry Tubes	Miltenyi Biotec		5 mL - Without caps
FlowJo Software	BD Biosciences		Analyze flow cytometry data
Fluorescence microscope	DM 2000	LEICA
Fluoroshield with DAPI	Sigma-Aldrich	F6057
Incubator	NUAIRE	UN-4750
MACSQuant Analyzer	Miltenyi Biotec		Flow cytometer
Microplate reader photometer	Clarkson Laboratory - CL
Microtubes 1.5 mL	Zhejiang Runlab Tech	35200N	wire snap
Minitab Software	Minitab		Statistical analysis
Needles	BD VACUTAINER	301746	Diameter 1.34 mm
Optical microscope	VELAB	VE-B50
Percoll	GE Healthcare	17-0891-01	Solution for density gradient
Phosphate Buffered Saline (10x)	Caisson	PBL07-500ML
Pyrex culture tubes	CORNING	CLS982025	N°9820
RPMI 1640 1x	Corning	10-104-CV	contains Glutagro
Slides	Madesa	PDI257550	22 mm x 75 mm
Trypan Blue solution 0.4%	SIGMA	T8154-100ML