Chromatin swelling drives neutrophil extracellular trap release

Abstract

Neutrophilic granulocytes are able to release their own DNA as neutrophil extracellular traps (NETs) to capture and eliminate pathogens. DNA expulsion (NETosis) has also been documented for other cells and organisms, thus highlighting the evolutionary conservation of this process. Moreover, dysregulated NETosis has been implicated in many diseases, including cancer and inflammatory disorders. During NETosis, neutrophils undergo dynamic and dramatic alterations of their cellular as well as sub-cellular morphology whose biophysical basis is poorly understood. Here we investigate NETosis in real-time on the single-cell level using fluorescence and atomic force microscopy. Our results show that NETosis is highly organized into three distinct phases with a clear point of no return defined by chromatin status. Entropic chromatin swelling is the major physical driving force that causes cell morphology changes and the rupture of both nuclear envelope and plasma membrane. Through its material properties, chromatin thus directly orchestrates this complex biological process.

Fig. 1

Phases of NETosis. a Morphological changes of chromatin (blue) and cell membrane (red) during NETosis of human neutrophils (stimulated with 100 nM PMA) imaged by live-cell confocal laser scanning microscopy (CLSM). The lobular nucleus loses its shape and chromatin decondenses until it fills the entire cell. Finally, the cell rounds up and releases the NET (white arrow). Scale bar = 5 µm. b Corresponding chromatin area of a NETotic neutrophil (a) as a function of time reveals three distinct phases. P1: Activation, lobulated nucleus. P2: Decondensation/expansion of chromatin within the cell (t₁ = start of chromatin expansion, t₂ = maximal chromatin expansion within the cell); cell rounding. P3: Rupture of the cell membrane (arrow a) and NET release (t₃ = NET release). c Histogram of onset times of the different phases. n = 139 cells. N = 5 donors. Lines represent Gaussian distribution function fits. d Time course of chromatin area for stimulation with LPS (Lipopolysaccharide, from Pseudomonas aeruginosa, 25 µg ml⁻¹). e Time course of chromatin area for stimulation with calcium ionophore (4 µM). c–e data acquired with live-cell wide field fluorescence microscopy. f Colocalization of decondensed neutrophil chromatin (blue) and myeloperoxidase (green). Fixed cells imaged by wide field fluorescence microscopy. Scale bar = 20 µm

Fig. 2

Chromatin swelling drives morphological changes. a Live-cell CLSM side view of a neutrophil during NETosis. Chromatin (blue) decondenses/expands, reaches the membrane (red) and the cell rounds up until the membrane ruptures. z-stack depth: 1 µm. b Cell height as measured by atomic force microscopy (AFM) on life neutrophils. PMA stimulated cells adhere and flatten (compared to the control cells that stay more or less round) and then round up (>8 µm) in P2. n = 3. Mean ± SEM. c Characteristic distribution of lamin B1 (green) in the three phases, CSLM images of fixed cells. Lamin B1 first surrounds single lobuli of the nucleus/chromatin (blue). When chromatin starts to expand corresponding to the start of P2 (around t₁), the lamin B1 layer/nuclear envelope ruptures on at least on one side of the nucleus. During P2 and P3 lamin B1 further decomposes. White arrows indicate rupture sites of the lamin B1 layer. Scale bar = 5 µm. d The original shape of the nucleus remains recognizable during the expansion process, particularly in the first part of P2 (t₁ to t₂), indicating isomorphic chromatin swelling and not directional transport (Supplementary Movies 13–15). In P1 the nucleus has a lobulated structure, which is maintained (self-similarity) during P2. Finally, the membrane is reached and, for a short period of time, this barrier prevents further expansion until it burst. Scale bar = 5 µm. Live-cell CSLM images

Fig. 3

Active and passive processes during NETosis. a ATP levels in stimulated neutrophils decrease during P1 and reach a plateau in P2. Inhibition of glucose metabolism further reduces ATP levels (Supplementary Fig. 6c). N = 3. Mean ± SEM. b Metabolic inhibitors (sodium azide/3 mM, 2-deoxy-D-glucose/5 mM, 4-aminobenzoic acid hydrazide/100 µM) influence NET formation determined as relative number of decondensed nuclei after 180 min compared to activation with PMA only. All inhibitors decrease NET formation when added in P1, while P2 is not or only slightly affected, indicating a point of no return. N = 3 donors. Statistics: two-way ANOVA (Bonferroni’s multiple comparisons test; *p < 0.05; **p < 0.01; ****p < 0.0001; ns = not significant). Mean ± SEM. c Phase duration at different temperatures (23.5, 37, 40 °C). P1 is significantly prolonged at lower temperatures, whereas P2 displays no or marginal temperature dependence. N = 3 (23.5, 40 °C). N = 5 (37 °C). Statistics: Kruskal-Wallis test (Dunn’s multiple comparisons test; *p < 0.05; **p < 0.01; ****p < 0.0001; ns = not significant). Life-cell imaging. Boxplots display the 25th and 75th percentile and the horizontal line the median. Hollow squares represent the mean and whiskers the SD

Fig. 4

Entropic swelling of chromatin causes membrane rupture. a Correlation between cell area at t₃ (chromatin area at t₃ ≈ total cell area at t₃) and time span until membrane rupture occurs following maximal chromatin expansion (t₃– t₂, rupture delay time). Larger cells rupture later than smaller cells. n = 112 cells (only cells of population 1 included, see Supplementary Fig. 3a, b). Fit lines show normalized 2D-probability density function calculated by kernel density estimation (KDE). N = 5 donors. b Average cell area (at t₃) of remaining/intact cells. The area of intact cells increases from ≈151 to ≈230 µm² with time, indicating earlier rupture of small cells. n = 121 cells. N = 5 donors. Mean ± SEM. c Live-cell confocal (left row) and STED (right row) images of chromatin (SiR-DNA) of two neutrophils undergoing NETosis (late P2) show almost uniform distribution of chromatin suggesting a mesh size smaller than the resolution of the microscope (≈120 nm (xy) and ≈150 nm (z), Supplementary Fig. 8b). Scale = 2 µm. d Time delay until rupture as a function of the calculated pressure p at t₂. n = 112 cells (only cells of population 1 included, see Supplementary Fig. 3a, b). N = 5 donors. The contour fit is generated by fitting the maximal data points to an exponential decay curve. e Neutrophils undergoing NETosis exert an increasing pressure on a fixed AFM cantilever until they rupture (end point of measurement). Inset: N = 5 (n = 14 cells). Boxplots display the 25th and 75th percentile and the horizontal line the median. Hollow squares represent the mean and whiskers the SD

Fig. 5

Rearrangement of the cytoskeleton and evolution of mechanical properties. a At the beginning of NETosis F-actin (red) is laterally enriched and localizes in the lamellipodia. α-Tubulin filaments (green) are arranged originating from the microtubule organizing center (MTOC) in unstimulated cells. Within the next hours (during P1) cytoskeletal components disintegrate. Remaining F-actin accumulates at the cell margin and α-tubulin is first rearranged in centrosome-like structures which disappear at the beginning of P2. CLSM images of fixed cells. Activation = PMA (100 nM). Blue = chromatin. Scale = 10 µm. b, c Inhibition of NET formation with the F-actin polymerization inhibitors Cytochalasin D (100 nM) and Latrunculin A (1 µM), F-actin-stabilizing drug Jasplakinolide (10 µM) and the ROCK-inhibitor Y-27632 (19.2 µM) significantly reduces the formation of NETs (measured as %-relative number of decondensed nuclei after 180 min compared to activation with PMA only) in P1, while P2 depends less or not on F-actin stabilization and ROCK-inhibition. Statistics: two-way ANOVA (Bonferroni’s multiple comparisons test; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns = not significant). N = 3 donors. Mean ± SEM. d Normalized tether tension of life neutrophils (measured with AFM) decreases over the entire time course (raw data: >0.35 mN m to <0.07 mN m) of PMA-activated NETosis indicating a loss of cytoskeletal stability. Values of control cells remain stable. N = 3. Mean ± SEM. e Cell stiffness (Young’s modulus) of life neutrophils decreases from >1.5 kPa to <0.3 kPa after stimulation with PMA whereas the stiffness of control cells remains constant. N = 3. Mean ± SEM

Fig. 6

Predetermination of the membrane rupture point. a Velocity plots of chromatin swelling. Changes to darker colors indicate faster movement (shorter residence time) of chromatin. The actual rupture point (red circle) often correlates with areas of slow movement (predicted rupture point, green circle). b Live-cell CLSM images of the cell membrane (PKH26 staining) directly before NET release. The cell rounds up and ruptures when maximum circularity is reached (t = 98 min). Scale bar = 5 μm. c Schematic of the rupture point analysis. Rupture points were analyzed by (1) fitting an ellipse to the cell before it became round and determining the rupture axis between the rupture point A and the center of mass M and (2) determining the retraction speed (v_A and v_B) on both sides of the (previously elliptic) cell (see Methods). d Shrinking velocity of the two opposing cell poles (A and B). The neutrophil retracts its membrane with a significantly higher velocity at the future rupture site (A). n = 17. N = 4 independent experiments. Statistics: Mann–Whitney test, two-tailed (***p < 0.001). Boxplots display the 25th and 75th percentile and the horizontal line the median. Hollow squares represent the mean and whiskers the SD. e Angle plot shows that the membrane ruptures in proximity of the major axis. α = rupture point angle. ß = major axis angle. n = 17. N = 4 independent experiments

Fig. 7

Biophysical model of NET release NETosis can be divided into three distinct phases (according to chromatin status) that are separated by a point of no return. The major physical driving force for morphological changes and NET release after phase 1 is entropic swelling of chromatin