Calcium flashes orchestrate the wound inflammatory response through DUOX activation and hydrogen peroxide release

Abstract

A crucial early wound response is the recruitment of inflammatory cells drawn by danger cues released by the damaged tissue. Hydrogen peroxide (H2O2) has recently been identified as the earliest wound attractant in Drosophila embryos and zebrafish larvae. The H2O2 signal is generated by activation of an NADPH oxidase, DUOX, and as a consequence, the first inflammatory cells are recruited to the wound within minutes. To date, nothing is known about how wounding activates DUOX. Here, we show that laser wounding of the Drosophila embryo epidermis triggers an instantaneous calcium flash, which travels as a wave via gap junctions several cell rows back from the wound edge. Blocking this calcium flash inhibits H2O2 release at the wound site and leads to a reduction in the number of immune cells migrating to the wound. We suggest that the wound-induced calcium flash activates DUOX via an EF hand calcium-binding motif and thus triggers the production of the attractant damage cue H2O2. Therefore, calcium represents the earliest signal in the wound inflammatory response.

Figure 1

Wounding of Drosophila Embryos Induces an Immediate Epidermal Calcium Wave (A) mCherry-moesin (red) and GCaMP3 (green) fluorescence before and after wounding of the Drosophila embryonic epidermis revealed a calcium wave rapidly spreading outward from the wound margin. The scale bar denotes 50 μm, and time is in seconds. See also Movie S1. (B) In (Bi), a wound made on the dorsal epidermis in an embryo coexpressing GCaMP3 and mCherry-moesin and adjacent to the zipper front of dorsal closure shows no spread of the calcium wave from dorsal epithelium onto amnioserosa (As) or across the seam (the wound position is marked by a white star; the arrow indicates the zippering front where the calcium wave terminates). In (Bii) are stills from movies of wound-induced calcium waves in control and inx2^G0118 and inx2^G0036 mutant embryos; time is in seconds. Graphs in (Biii) and (Biv) reveal that the distance traveled by calcium waves, but not the initial calcium intensity (F/F0) at the wound edge, was significantly lower in inx2 mutants than in control embryos. (A one-way ANOVA was used with a Bonferroni posttest; n ≥ 18 embryos per genotype). Scale bars in (Bi) and (Bii) denote 50 μm. Error bars in (Biii) and (Biv) represent the SEM. ^∗∗∗∗p < 0.0001; ns = nonsignificant. (C) Resolution of the calcium wave is revealed by GCaMP3-expressing embryos (see also Movie S2). The plot shows fluorescence-intensity change normalized to background fluorescence (F1/F0) for all cells (thick red line) and cells categorized according to their position relative to the wound edge (see schematic inset for the position of cells). The scale bar denotes 20 μm, and time is in seconds.

Figure 2

Wound-Induced Calcium Waves Activate the Inflammatory Response (A) Representative images of epithelial GCaMP3 fluorescence immediately after wounding in embryos treated with 1 μm thapsigargin or 5 mM EGTA reveal that interfering with calcium signaling impaired wound-induced calcium waves. The bar graph shows the mean integrated density of the GCaMP3 signal per embryo (n ≥ 6 embryos per treatment). The scale bars depict 25 μm. (B) Dampening calcium responses reduced recruitment of red-stinger-labeled hemocytes to wounds (wound edges are indicated by white ellipses). The scatter plot shows quantification of hemocytes per wound (the lines show the mean for ≥13 embryos per treatment). The scale bars depict 20 μm. (C) The mean integrated density of GCaMP3 fluorescence per embryo immediately after wounding of trpm² mutants or embryos expressing TRPM RNAi specifically in the epithelium is lower than that for wild-type controls, indicating an epithelial role for TRPM in the generation of a calcium wave (n > 12 embryos per genotype). (D) Impairment of the calcium wave correlated with a decrease in mean numbers of red-stinger-labeled hemocytes recruited to wounds in trpm² mutants (n ≥ 10 embryos per genotype). All error bars represent the SEM, and asterisks denote significance values of p < 0.05 (^∗), p < 0.01 (^∗∗), and p < 0.001 (^∗∗∗) via a Student’s t test.

Figure 3

DUOX Interprets the Calcium Wave via Its EF Hand Domain to Drive H₂O₂ Production and Recruitment of Macrophages (A) Single confocal slices depicting GMA (actin, green) and Amplex Ultrared (H₂O₂, red) show that embryos ubiquitously expressing DUOX RNAi and GMA produced less H₂O₂ at wounds than did wild-type controls, as assayed via Amplex Ultrared fluorescence. The scale bars represent 20 μm. (B) trpm² and inx2 mutant embryos with a reduction in their wound-induced calcium responses displayed impaired H₂O₂ production via Amplex Ultrared (see Experimental Procedures for details of F-F0 quantification). The graphs show mean ± SEM of at least 20 (trpm²) and 7 (Inx2) embryos per genotype. (C and D) Re-expression of full-length dDUOX, but not a truncated form specifically lacking the EF hand motif (dDUOXΔEF), in embryos with ubiquitous expression of dDUOX RNAi restored wound-induced H₂O₂ production (C) and hemocyte recruitment (D) to the levels of wild-type controls. The graphs show mean ± SEM of at least 21 (C) and 11 (D) embryos per genotype. See Movie S3 for a typical example of this data. Hemocyte numbers at wounds were determined from images of hemocytes labeled independently of Gal4 with srp-GMA. The p values were generated with a Student’s t test at the final time point (B [trpm²]), a two-way ANOVA with a Bonferroni posttest (B [inx2] and C), or a one-way ANOVA with a Bonferroni posttest (D). Asterisks denote p < 0.05 (^∗), p < 0.01 (^∗∗), and p < 0.001 (^∗∗∗); ns = not significant.

Figure 4

NIP Is Required for DUOX Maturation and Activation after Wounding (A) Tracking red-stinger-labeled hemocytes for 20 min after wounding revealed that fewer hemocytes migrated to wounds (indicated by white ellipses; scale bar shows 40 μm) in nip mutant embryos than in controls. The upper panel shows the final frame of a hemocyte movie at 20 min; the lower panel displays hemocyte tracks from the movies above and indicates that hemocytes tend to ignore wounds in nip mutant embryos and instead remain in their developmental positions. The wound is highlighted by the red dashed line (representative of at least three movies per genotype). (B) Quantification of mean hemocyte numbers at wounds after 20 min in wild-type versus nip mutant embryos; n ≥ 24 embryos per genotype. Error bars show the SEM; asterisks denote p < 0.001 (^∗∗∗) and p < 0.01 (^∗∗), generated via a Student’s t test. (C) Single confocal slice images of Amplex Ultrared at wounds and quantification of mean levels of fluorescence demonstrate compromised H₂O₂ production in nip mutant embryos compared to wild-type controls (see Experimental Procedures for details of quantification; n ≥ 11 embryos per genotype; scale bar shows 50 μm; time is in seconds). Error bars show the SEM; asterisks denote p < 0.001 (^∗∗∗) and p < 0.01 (^∗∗), generated via a two-way ANOVA with a Bonferroni posttest.