Calcium Dynamics in Neutrophils during Swarming

(A) Schematic of a 3-day-post-fertilization (dpf) zebrafish larva showing the area of two-photon laser wound damage and imaging.

(B) Construct expressing GCamp6F under the control of the lysozyme C promoter (lyz). Ca²⁺ binding to the calmodulin (Cam) domain of GCamp6F increases EGFP fluorescence.

(C) Time-lapse sequence of two-photon confocal image projections showing neutrophils (color-coded for GCamp6F intensity) migrating from the caudal hematopoietic tissue (CHT) toward a laser wound (LW) (dotted line) at the ventral fin-CHT boundary (VF/CHT) in a Tg(lyz:GCamp6F) larva. The calcium wave (CW) is indicated in the second panel and the neutrophil cluster with a white arrow. Scale bar represents 50 μm.

(D) Quantification of mean GCamp6F intensity over time in all neutrophils in the field of view in the first 2 min post-wound. Intensity values were normalized to the mean intensity of segmented neutrophils prior to wound. Each line represents mean intensity per experimental larva.

(E) Quantification of GCamp6F intensity over time in neutrophils clustering at the wound versus migrating neutrophils beyond the wound (dotted line in C). Intensity values were normalized as in (D). n = 8 larvae in 8 experiments. Dotted square denotes the time-bin corresponding to analysis in (D).

(F) Normalized GCamp6F intensity in relation to the surface area of segmented neutrophils. Individual dots represent single neutrophils or clustered neutrophils at the wound (red) or migrating cells beyond the wound (blue). The maximum surface area for single neutrophils is indicated with dotted line. Data are from 8 larvae in 8 experiments.

(G) Comparison of GCamp6F intensity in single (<600 μm²) versus clustering neutrophils (>600 μm²). n = 924 single cells and n = 347 clustered cells from 8 larvae in 8 experiments; Mann-Whitney test.

In all panels, error bars represent SEM. ^∗∗∗p < 0.0002. See also Figure S1 and Videos S1 and S2.

Activation of LTB4 Biosynthesis Is Favored in Clustering Neutrophils

(A) Schematic of LTB4 biosynthesis. cPLA2 (calcium-dependent phospholipase A2) and 5-LO are recruited to the nuclear membrane and produce arachidonic acid (AA) and LTA4, respectively. LTA4 is metabolized into LTB4 by LTA4 hydrolase.

(B) Constructs for transgenic expression of a fluorescent fusion of 5-LO with tRFP in neutrophils (below). Schematic of neutrophil with 5-LO nuclear translocation is shown.

(C and D) Spinning-disk confocal projections of neutrophils in 3-dpf double-transgenic Tg(lyz:GCamp6F)xTg(lyz:tRFP-5LO) zebrafish larvae after two-photon LW in the ventral fin-CHT boundary (C) or mechanical wound (MW) in the ventral fin (D). Blue dotted lines indicate the wound area occupied by clustering neutrophils. Zoomed images of examples of neutrophils with 5-LO translocation are shown. Time in relation to translocation is indicated in minutes in the first example. Examples are from three (LW) or two (MW) different larvae. Scale bars represent 50 μm and 5 μm, respectively.

(E and F) Quantification of mean distance from the wound center (x; left) and normalized GCamp6F fluorescence intensity (right) for 5-LO-translocating cells versus non-translocating cells in laser wounds (E) and mechanical fin wounds (F). GCamp6F fluorescence intensity was normalized to the most fluorescent cell in the corresponding frame.

(E) n = 41 cells (for translocating cells, each dot is a cell; for non-translocating cells, each dot represents the mean of all non-translocating cells in the same field of view; left) and n = 31 cells from 8 larvae in 5 different experiments (right).

(F) n = 17 (left) and n = 16 (right) cells from 5 larvae in 3 different experiments. Colored dots represent examples shown in individual images in (D).

Wilcoxon matched-pairs signed rank test. Error bar represents 95% confidence intervals of medians. ^∗∗∗p < 0.0002. See also Figure S2 and Video S2.

5-LO-Associated Calcium Fluxes Are Triggered upon Contact with Necrotic Cells or Neutrophils with Ongoing Fluxes

(A) Schematic of two-photon laser wounding in the presence of propidium iodide (PI).

(B) Time-lapse, two-photon confocal projection images of a GCamp6F-expressing (white) neutrophil (indicated with an arrow) entering a contact with PI⁺ cells/tissue (red) in a Tg(lyz:GCamp6F) larva; time in relation to the first frame is indicated in seconds. Scale bar represents 10 μm.

(C) Quantification of speed (blue) and normalized GCamp6F (gray) in neutrophils before and after contact with PI⁺ tissue. GCamp6F intensity was normalized as in Figures 1D and 1E. Dotted red line indicates time of contact. Pooled cell data from n = 23 cells in 7 larvae and 4 experiments are shown.

(D) Examples of cell contacts transmitting calcium fluxes. Each case is represented by time-lapse images of a non-calcium-fluxing neutrophil (arrow) contacting a fluxing neutrophil. Time in minutes is indicated relative to cell-cell contact. Scale bar represents 10 μm. The quantification of neutrophil transmission of calcium fluxes is indicated in Figure 6.

Error bars represent SEM. See also Figure S2 and Video S3.

ATP-Gated Calcium Channels and Extracellular Calcium Entry Promote 5-LO-Capacitating Calcium Fluxes in Neutrophils In Vivo

(A) Spinning-disk confocal projections from neutrophils in Tg(lyz:GCamp6F)xTg(lyz:tRFP-5LO) larvae responding to MW, before and after addition of 50 μM calcium ionophore (A23187). Time post-wounding is indicated inside the images. Arrows indicate translocation events. Red dotted line indicates area occupied by clustering neutrophils. Scale bar represents 50 μm. Cartoon indicates area of wounding.

(B) Normalized GCamp6F intensity over time in clustering versus migrating cells (cells within or beyond denoted red line in A). The time in relation to calcium ionophore addition is shown. n = 768–2,140 cells per bin (migrating) and n = 116–461 cells per bin (clustering) from 8 larvae in 3 experiments.

(C) Percentage of 5-LO-translocating neutrophils out of all neutrophils visible in the field of view. Each line represents an individual larva; 8 larvae from 3 different experiments.

(D) Normalized radial speed over time for migrating cells. n = 743–1,851 cells per bin from 8 larvae in 3 experiments.

(E) Schematic of mechanical ventral fin wounding in the presence of NF279. Blue indicates vessels of the caudal vein plexus within the caudal hematopoietic site.

(F) Spinning-disk confocal projection images of neutrophils in Tg(lyz:GCamp6F)xTg(lyz:tRFP-5LO) larvae 120 min after MW in the presence (right) or absence (left) of 10 μM NF279. Red dotted line indicates area occupied by clustering neutrophils. Scale bar represents 25 μm.

(G) Mean normalized GCamp6F intensity larvae treated or not with NF279. GCamp6F intensity was normalized as in Figure 1D. n = 9 control larvae and n = 3 NF279-treated larvae from 3 and 2 experiments, respectively; Mann-Whitney test.

(H) Percentage of translocating neutrophils out of all neutrophils recruited into the fin over 2 h. n = 7 control and n = 3 NF279-treated larvae from 3 and 2 experiments, respectively; Mann-Whitney test.

Error bars represent SEM. ^∗p < 0.03, ^∗∗p < 0.002. See also Figures S3 and S4 and Videos S4 and S5.

Neutrophil Cx43 Is Required for Intracluster Calcium Fluxes and Swarm Initiation

(A) Time-lapse sequence of two-photon confocal image projections showing neutrophils in Tg(lyz:GCamp6F) larvae in the presence of PI, without treatment (ctr), with 50 μM CBX, or with morpholinos against cx43/cx43.4 (cx43 MO). Scale bars represent 50 μm and 10 μm for zoomed-out and zoomed-in images, respectively. Time after LW is shown in minutes. Arrows indicate neutrophils in contact with the wound.

(B) Normalized GCamp6F levels in control (n = 8), CBX-treated larvae (n = 5), and cx43 MO-treated (n = 7) from 8, 2, and 3 experiments, respectively. One-way ANOVA with Dunnett’s post test is shown. GCamp6F intensity was normalized as in Figure 1D. Data are from Tg(lyz:GCamp6F) and Tg(lyz:GCamp6F)xTg(lyz:tRFP-5LO) larvae evenly distributed across the groups.

(C) Neutrophil radial speed over time post-wounding for neutrophils in control, cx43 MO-treated, and CBX-treated larvae. Data are pooled from Tg(lyz:GCamp6F), Tg(lyz:GCamp6F)xTg(lyz:tRFP-5LO), and Tg(mpx:GFP)ⁱ¹¹⁴ zebrafish larvae. n = 1,201–1,719 cell steps per bin from 12 control, n = 1,268–1,535 cell steps per bin from 8 CBX-treated, and n = 1,308–1,554 cell steps per bin from 11 cx43 MO-injected larvae from 8, 3, and 4 experiments, respectively. Kruskal-Wallis with Dunn’s post test results indicating significance found between ctr and MO and between ctr and CBX (gray) or only between ctr and CBX (black) are shown.

(D) Time-lapse sequence of two-photon confocal image projections showing neutrophils in Tg(lyz:GCamp6F)xTg(lyz:cx43DN-T2A-mCherry) zebrafish larvae, positive (cx43 DN) or negative for the Cx43 DN-T2A-mCherry transgene (control siblings: Ctr sib). Annotations are as in (A).

(E) Normalized GCamp6F levels in neutrophils positive (cx43 DN) or negative for the Cx43DN-T2A-mCherry transgene (Ctr sib). n = 7 cx43 DN transgenics and n = 5 control siblings from 3 experiments; Mann-Whitney test.

(F) Neutrophil radial speed over time post-wounding for neutrophils positive (cx43 DN) or negative for the Cx43DN-T2A-mCherry transgene (Ctr sib). n = 1,330–1,602 cell steps per bin from 5 control siblings and n = 2,573–3,050 cell steps per bin from 7 Cx43 DN larvae from 3 experiments; Mann-Whitney test.

Error bars represent SEM. ^∗p < 0.03, ^∗∗p < 0.002, ^∗∗∗p < 0.0002. See also Figures S4–S6 and Video S6.

Neutrophil Cx43 Is Required for Autonomous and Cooperative Neutrophil Calcium Fluxes

(A) Time-lapse images showing calcium fluxes in neutrophils (arrow) contacting other neutrophils or necrotic cells in control untreated, CBX-treated, or cx43 MO-treated Tg(lyz:GCamp6F) larvae or in Tg(lyz:GCamp6F)xTg(lyz:cx43DN-T2A-mCherry) larvae. Time in minutes is indicated relative to the start of the cell-cell contact. Scale bar represents 15 μm.

(B) Cartoon illustrates contacts between neutrophils and necrotic PI⁺ cells or between fluxing and non-fluxing neutrophils resulting or not in calcium flux transmission.

(C and D) Percentage of neutrophil-neutrophil contacts (C) or neutrophil-PI⁺ cell contacts (D) resulting in transmission of calcium fluxes. Data are from Tg(lyz:GCamp6F)xTg(lyz:tRFP-5LO) (C and D) or Tg(lyz:GCamp6F) (D) larvae. Contacts in which none of the cells is initially fluxing are not included. n = 8 control, 4 CBX-treated larvae, and 7 Cx43 morphants from 8, 2, and 3 experiments, respectively (C). n = 5 control, 5 CBX-treated larvae, and 5 Cx43 morphants from 5, 2, and 2 experiments, respectively (D). One-way ANOVA, Tukey’s multiple comparisons test, is shown.

(E) Percentage of contacts resulting in transmission of calcium fluxes in Tg(lyz:GCamp6F)xTg(lyz:cx43DN-T2A-mCherry) zebrafish larvae, positive (cx43 DN) or negative for the Cx43DN-T2A-mCherry transgene (control siblings: Ctr sib). n = 7 cx43 DN and n = 5 control siblings from 3 experiments; Mann-Whitney test.

(F) Time point of arrival of first neutrophil at the wound in neutrophils positive (cx43 DN) or negative for the Cx43DN-T2A-mCherry transgene (control siblings: Ctr sib). n = 7 cx43 DN and n = 5 control siblings from 3 experiments; Mann Whitney test.

Error bars represent SEM. ^∗∗p < 0.002, ^∗∗∗p < 0.0002. See also Video S6.

Cx43 Is Required for Maximal Wound Defense from Bacterial Invasion

(A) Schematic of tail amputation and infection by PAO1 P. aeruginosa.

(B) Colony-forming units (CFUs) per larva in control wild-type (AB strain), non-injected larvae or cx43 MO-injected larvae. Time after wounding is in hours (h). n = 4 experiments, with 5 larvae per group; Mann-Whitney test.

(C) Survival over time in control wild-type, non-injected larvae or cx43 MO-injected larvae, wounded (W) or not (NW) in the presence of PAO1. n = 4 experiments, with 20 larvae per group; log rank (Mantel-Cox) test.

(D) Survival over time in Tg(lyz:GCamp6F)xTg(lyz:cx43DN-T2A-mCherry) zebrafish larvae, positive (cx43 DN) or negative for the Cx43DN-T2A-mCherry transgene (control siblings: Ctr sib), wounded (W) or not (NW) in the presence of PAO1. n = 4 experiments, with 20 larvae per group; log rank (Mantel-Cox) test.

(E) Schematic of imaging wound infection. Annotations are as in Figure 4E.

(F) Time-lapse sequence of two-photon confocal projections showing neutrophils in zebrafish larvae, positive (cx43 DN) or negative for the Cx43DN-T2A-mCherry transgene (control siblings; Ctr sib), in the presence of Syto62-labeled PAO1. Scale bars represent 50 μm and 10 μm. CW, CHT, and VF are as in Figure 1C. Arrows indicate neutrophils in contact with the wound. Dotted lines outline neutrophil clusters.

(G) Neutrophil cluster size over time post-wounding in cx43 DN or control larvae (Ctr) (includes negative siblings and single Tg(lyz:GCamp6F) transgenics). n = 5 cx43 DN and 7 control larvae from 6 and 5 experiments, respectively; Mann-Whitney test.

(H) Neutrophil radial speed over time post-wounding in cx43 DN or control larvae. n = 1,290–1,537 cell steps per bin from 7 control larvae and n = 1,527–2,079 cell-steps per bin from 8 cx43 DN larvae imaged in 6 and 5 experiments, respectively; Mann-Whitney test.

(I) GCamp6F levels normalized as in Figure 1D. n = 5 cx43DN transgenics and n = 7 control larvae from 6 and 5 experiments, respectively; Mann-Whitney test.

(J) Images of the wound (dotted outline) pseudocolored for fluorescence intensity of Syto62-labeled PAO1 bacteria. Time post-wounding is indicated in minutes. Scale bar represents 25 μm.

(K) Fluorescence intensity of bacteria at the wound relative to maximal initial intensity in this area. n = 5 Cx43 DN and n = 8 control larvae from 6 and 5 experiments, respectively; Mann-Whitney test.

Error bars represent SEM. ^∗p < 0.03, ^∗∗p < 0.002, ^∗∗∗p < 0.0002. See also Figure S7 and Video S7.