Leukocytes Are Required for the Initial Decline in Wound Margin H₂O₂ Concentrations(A) (i) Mean HyPer ratios reflecting the wound zone [H₂O₂] profile of WT (injured and uninjured), cloche mutant (clo) and spi1/csf3r morphant embryos. Note the significantly sustained, elevated HyPer ratio/[H₂O₂] in leukocyte-deficient clo and spi1/csf3r morphants. Insets (ii, iii) show the individual median HyPer data and p values (natural permutation test). The n values are as follows (embryos/independent days/scans): injured WT, 5/2/4; spi1/csf3r morphants, 4/2/3; clo 6/3/5; uninjured WT, 5/1/5.(B) HyPer heat maps depicting [H₂O₂] across the wounded tail fin at five sequential time points. Accompanying overlaid transmission (Trans) and DsRED2 images demonstrate number and position of Tg(lyz:DsRED2) neutrophils. Note the single DsRED2-positive cell in the spi1/csf3r morphant (black arrow head). clo did not carry the lyz:DsRED2 transgene. Scale bar (applies to all panels) represents 200 μm. Heat map legend applies to all figures. The stills are from Movie S1.

Myeloperoxidase (mpx) Deficiency in the Zebrafish Mutant durif (drf)(A) mpx expression by whole-mount in situ hybridization (WISH) in WT and drf embryos.(B and C) Loss of cells carrying Mpx activity in drf demonstrated by loss of Mpx-dependent histochemical staining (B) and stable sudanophilia (C) (arrowheads indicate positive cells in WT).(D) Cytospun adult kidney hematopoietic cells, stained for Mpx by diaminobenzidine (DAB) histochemistry with May-Grunwald/Giemsa counterstain. WT but not drf neutrophils show Mpx staining (black triangles). Providing negative and positive internal staining controls, both genotypes have Mpx-negative eosinophils (black arrow) and weak erythrocyte staining due to the pseudoperoxidase Lepehne reaction (open triangles). Scale bar represents 10 μm.(E) Ultrastructure of adult WT and drf neutrophils, showing normal primary granules. Scale bars represent left, 2 μm; right, 100 nm.(F) drf^-/- Tg(mpx:EGFP) hemizygotes have normal numbers of EGFP-expressing cells (mean ± SE: WT, 120 ± 6, n = 16; drf, 116 ± 5, n = 15).(G) Positional cloning placed drf on chromosome 10 within 0.08 cM of RFLP-356k located ~1.5 kb 3′ from the mpx gene.(H) Schematic of the mpx locus (black, exon; white, UTR; based on GenBank BC068379.1). Left sequencing chromatograms of complementary DNA shows the exon 11/12 junction: a 7 nt insertion in drf introduces a premature stop codon (^*). Right sequencing chromatograms of genomic DNA show the drf mutation (red arrow), a T→A transversion 9 nt 52 of the intron 12/exon 13 junction.

Neutrophil-Delivered Myeloperoxidase Is Required for the Initial Decline in Wound Margin H₂O₂ Concentrations(A) (i) Mean HyPer ratios reflecting the wound zone [H₂O₂] data for injured and uninjured drf embryos. Note the sustained, elevated HyPer ratio/[H₂O₂] in mpx-deficient drf mutants. These drf data are directly comparable with data of Figure 1A and share its WT control group. Inset (ii) shows the individual median HyPer data and p value compared to WT (natural permutation test). The n values are as follows (embryos/independent days/scans): injured drf, 6/5/6; uninjured drf, 4/1/4. (iii) Representative HyPer heat maps depicting [H₂O₂] across the wounded tail fin of an mpx-deficient drf embryo with overlaid transmission (Trans) and DsRED2 images demonstrating the proximity of neutrophils to the wound margin. Scale bar represents 200 μm. Stills are from Movie S1 at same time points as in Figure 1B.(B) (i) Transplantation strategy to generate HyPer-mRNA-loaded, Mpx-deficient, drf embryos selectively populated with DsRED2-expressing WT neutrophils. (ii) A successful transplant outcome. Brightfield image of 3 dpf embryo with superimposed green and red fluorescence images, showing a single residual recipient-derived EGFP-expressing neutrophil (green arrow) and scattered DsRED2-expressing donor neutrophils (red arrows).(C) (i) Mean HyPer ratios reflecting the wound zone [H₂O₂] profile of drf embryos carrying WT neutrophils, which arrive (pink) or do not arrive (orange) at the wound. WT neutrophil arrival recapitulates the WT HyPer ratio profile; WT cell nonarrival serves as a negative control demonstrating the sustained [H₂O₂] elevation of a nonrescued Mpx-deficient drf profile. Inset (ii) shows the individual median HyPer data. The n values are as follows (embryos/independent days/scans): WT cells arrived, 3/3/3;WT cell did not arrive, 3/3/3.(D) HyPer heat maps depicting [H₂O₂] across the wounded tail fin of a transplanted drf embryo with arriving WT cells, demonstrating lower [H₂O₂] levels in the tail half where WT DsRED2-expressing cells arrive. Overlaid transmission (Trans) and DsRED2 images demonstrate the proximity of Tg(lyz:DsRED2) neutrophils to the wound margin. Scale bar (applies to all panels) represents 200 μm.

Myeloperoxidase and H₂O₂ Interact Directly within Neutrophils at the Wound Zone(A) HyPer₄₈₀ images showing neutrophil-specific HyPer expression in Tg(lyz:HyPer) embryos, with HyPer ratio heat maps reflecting the corresponding intraneutrophil [H₂O₂] levels at four time points after wounding. Stills from Movie S3. Corresponding graphs show single neutrophil mean HyPer ratios plotted against distance from wound for four individual embryos, with individual animal regression lines (individual animals color-coded). The negative gradient at 30–50 min after wounding contrasts with the uniform intraneutrophil [H₂O₂] levels prior and later, and indicates higher internal [H₂O₂] within wound-proximate neutrophils at these time points.(B–F) Demonstration of a direct enzymatic Mpx/H₂O₂ interaction within the wound zone by histochemical staining for Mpx activity. For each reaction condition (B–F), the three panels are (i) an overview, (ii) details of cells in the caudal hematopoietic tissue, and (iii) wound zone. Stained cells (white arrows) indicate Mpx-dependent catalysis of p-phenylenediamine and catechol. When H₂O₂ is supplied as a reagent, cells stain throughout the embryo (B), but without supplemented H₂O₂, cells stain only at the wound margin (C). (D)–(F) are negative controls: (D) and (E) drf controls show that staining is mpx-dependent regardless of the H₂O₂ source; (F) WT uninjured control showing that detecting staining of wound-proximate cells is injury-dependent. x/y n values are as follows: (embryos with e1 positive cell in region)/(embryos scored).(G–I) The relative [H₂O₂] within and surrounding wound zone neutrophils, demonstrated using HyPer-mRNA-loaded Tg(lyz:HyPer) embryos. (G) Groups of neutrophils 29 and 85 min after wounding displayed by their HyPer₄₈₀ signal, with corresponding HyPer ratio images portraying intra- and extracellular [H₂O₂]. At 29 min, all neutrophils have higher [H₂O₂] internally than externally, but at 85 min, a cluster of 3 neutrophils displays lower [H₂O₂] internally than externally. Stills from Movie S4 are shown. (H) Tracks of nine neutrophils plotted in space (x,y) and time (z), with tracks colored in categories to indicate when the neutrophil internal [H₂O₂] exceeds (green), equals (blue), or is less than (red) the external [H₂O₂]. Legend gives percentage of exterior HyPer ratio category cut-offs. The nine cells are three neutrophils (C1-3) selected from three wounded embryos (W1–W3). (I) The difference between internal and external [H₂O₂] in time for these same nine cells expressed as the difference between internal and external mean HyPer ratios. Six out of nine cells in (H, I) are net [H₂O₂] consumers, and in 5/6 cases this phenotype is accentuated as they approach the wound margin (one cell does not migrate). Scale bars in (A) and (G) represent 50 μm. Unlabelled perspective-shortened y axis marks (H, left upper corner) indicate 50 μm steps.

Hematopoietic Gene and Marker Gene Expression in the myeloperoxidase-Deficient Zebrafish Mutant durif, Related to Figure 2
Gene expression by whole mount in situ hybridization (A-P) and transgene (Q,R) expression in wildtype (WT) and durif (drf) embryos at the ages indicated. drf shows normal expression of all these genes at the ages indicated.
All panels are direct lateral views except B,D,F,H which are ventral views corresponding to the panels to their left. drf embryos result from homozygote incrossing.

Zonal Variations in Tissue [H₂O₂] Related to the Path of an Individual Neutrophil Migrating towards the Wound Zone, Related to Figure 4, and a Model of the H₂O₂/Myeloperoxidase Interaction after Wounding
(A) Still images over 1 h showing path of a single lyz:DsRED2-marked neutrophil (upper row) and associated HyPer ratiometic data reflecting the [H₂O₂] data for the wound zone. Note the asymmetrical regional reduction on the dorsal (upper) half of the fin where the DsRED2-marked neutrophil is located, most evident by the loss of extent of yellow signal on the heat map. Stills are from Movie S5. Scale bar, 100 μm.
(B) (i) Merged transmission and red fluorescence image at 98 min after injury, overlayed with the migratory paths of lyz:DsRED2 neutrophils (4 green paths, only one of which is relevant to regions 1 and 2) and lyz:DsRED2-negative macrophages (purple, blue and cyan paths, all relevant to region 3). Regions 1-4 are: (1) and (2), regions where the single DsRED2-marked neutrophil passes in sequence; (3) region where three DsRED2-negative macrophages pass; (4) control region, distal to wound. Scale bar, 100 μm. (ii) HyPer ratiometic data reflecting the temporal [H₂O₂] data for the 4 regions of the injured zebrafish larvae detailed in (i); data are median HyPer ratio for each of the regions shown. Note that the median [H₂O₂] decreases sequentially in region 1 then 2 as the neutrophil passes through them, but remains elevated in region 3 where only macrophages pass. The [H₂O₂] remains at background in the control region (4).
(C) Representative HyPer heat maps depicting [H₂O₂] across the tail fin of a WT embryo 30 and 60 min after embedding only (i.e. without wounding by tail transection) showing resolution of small focal H₂O₂ bursts at minor nick injuries that were sustained during embedding (white arrowheads). Adjacent overlaid transmission (Trans) and DsRED images document that this resolution occurred without the arrival of any lyz:DsRED2-marked neutrophils. Scale bar, 200 m.
(D) Table showing the extent of overlap of leukocyte marker transgene expression in compound transgenic zebrafish embryos at 2-3 dpf. * = actual genotype Tg(mpeg1:GAL4)x(10xUAS:Kaede) na = not applicable ¶ = Wound margin (WM) corresponds to cell within 100 μm of the wound edge 2 h after amputation.
(E) Model of the H₂O₂/myeloperoxidase interaction in the wound zone. Schematic diagrams at three timepoints before (i), just after (ii) and 60-90 min after (iii) wounding, showing for each time point the stage in the post-wounding H₂O₂ burst (upper panels), a whole tail overview (middle panels), and a detail of a neutrophil (lower panels). Prior to wounding (i), intra-neutrophil and surrounding tissue [H₂O₂] is low. Neutrophils carry the Src-family kinase Lyn capable of H₂O₂ sensing (green), and myeloperoxidase (Mpx, blue) packaged in primary granules.
Immediately after wounding (ii), a tissue-scale [H₂O₂] gradient (red, low-high [H₂O₂] as indicated in scale) is generated by Dual oxidase (Duox, yellow). H₂O₂ is sensed by Lyn in neutrophils, initiating their migration. Initially, within migrating neutrophils the sum of their own endogenous H2O2 production (possibly generated by NADPH oxidase 2 (Nox2) and superoxide dismutase (SOD), and/or as a by-product of mitochondrial respiration) plus the diffusion of tissue-sourced H₂O₂ results in the net outcome of higher intracellular [H₂O₂] than surrounds.
The decline from peak tissue gradient H₂O₂ levels (iii) coincides with neutrophil arrival and is dependent on their delivery of enzymatically-active Mpx which interacts intracellularly with H₂O₂. Some migrating neutrophils approaching the wound develop and sustain an intracellular [H₂O₂] that is lower than their surrounds – they are persistent net consumers of endogenous and tissue-sourced H₂O₂. These neutrophils function as intracellular H₂O₂ sumps. At the same time, some other wound-proximate neutrophils display higher internal [H₂O₂] than their surrounds. However, the net effect of Mpx-dependent intracellular enzymatic HH₂O₂ consumption is dampening of the overall wound-stimulated tissue H₂O₂ burst.