Draper/CED-1 mediates an ancient damage response to control inflammatory blood cell migration in vivo

Abstract

Tissue damage leads to a robust and rapid inflammatory response whereby leukocytes are actively drawn toward the wound. Hydrogen peroxide (H2O2) has been shown to be an immediate damage signal essential for the recruitment of these inflammatory blood cells to wound sites in both Drosophila and vertebrates [1, 2]. Recent studies in zebrafish have shown that wound-induced H2O2 is detected by the redox-sensitive Src family kinase (SFK) Lyn within the responding blood cells [3]. Here, we show the same signaling occurs in Drosophila inflammatory cells in response to wound-induced H2O2 with mutants for the Lyn homolog Src42A displaying impaired inflammatory migration to wounds. We go on to show that activation of Src42A is necessary to trigger a signaling cascade within the inflammatory cells involving the ITAM domain-containing protein Draper-I (a member of the CED-1 family of apoptotic cell clearance receptors) and a downstream kinase, Shark, that is required for migration to wounds. The Src42A-Draper-Shark-mediated signaling axis is homologous to the well-established SFK-ITAM-Syk-signaling pathway used in vertebrate adaptive immune responses. Consequently, our results suggest that adaptive immunoreceptor-signaling pathways important in distinguishing self from non-self appear to have evolved from a more-ancient damage response. Furthermore, this changes the role of H2O2 from an inflammatory chemoattractant to an activator signal that primes immune cells to respond to damage cues via the activation of damage receptors such as Draper.

Figure 1

src42A Is Required Specifically and Autonomously for Macrophage Wound Responses in Drosophila Embryos (A) Stills and trajectories of red-stinger-labeled macrophages taken at 20 min after wounding from movies of inflammatory responses to wounds in control and src42A^E1 mutant embryos. (B) Scatterplots of directionality toward the center of wounds and average speed (per macrophage, per embryo) during wound responses show that macrophages in src42A^E1 mutants essentially ignore wounds (B′) but that their ability to move at normal speeds is unaffected (B′′). (C) Representative stills of GFP-labeled macrophages (green) at wound sites at 20 and 60 min after wounding in control embryos and embryos expressing a dominant-negative version of Src42A in macrophages. (D) Scatterplot of wound responses shows numbers of macrophages per μm² of wound area normalized according to control averages. Scale bars represent 20 μm. Central lines and error bars on scatterplots represent mean and SD, respectively; ns, not significant; ^∗p < 0.05 and ^∗∗∗p < 0.001 via Mann-Whitney test (B) or one-way ANOVA followed by Sidak’s multiple comparisons test (D); Mϕ, macrophages; white ovals depict wound edges. See also Figures S1 and S3 and Movie S1.

Figure 2

Cell-Autonomous Requirement for draper and shark and Genetic Interaction between src42A and draper during Macrophage Recruitment to Wounds (A) Representative stills of GFP-labeled macrophages (green) at wound sites at 60 min after wounding in control and draper^Δ5 mutant embryos or embryos in which macrophages express a draper RNAi construct. (B) Scatterplot of wound responses shows numbers of macrophages per μm² of wound area at 60 min normalized according to control averages. (C) Representative stills of nuclear-red-stinger-labeled macrophages at 60 min after wounding reveal that embryos with a null mutation in shark have a strong reduction in macrophage recruitment to wounds compared to controls. (D) Scatterplot of wound responses at 60 min reveals defective macrophage recruitment in shark¹ and shark¹/Df embryos. (E) Representative stills of GFP-labeled macrophages (green) at 60 min post-wounding reveal that overexpression of shark in macrophages rescues shark¹ mutant wound response defects. (F) Scatterplot of wound responses shows rescue of macrophage responses at 60 min with macrophage-specific expression of shark in shark¹ mutant background, compared to shark¹ mutants. (G) Scatterplot of wound responses in src42A^E1/draper^Δ5 transheterozygotes reveals a genetic interaction of these genes during recruitment of macrophages to wounds. White ovals display wound margin; central lines and error bars on scatterplots represent mean and SD, respectively; ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, and ^∗∗∗∗p < 0.0001 according to Mann-Whitney test (B, D, and G) or one-way ANOVA with Sidak’s multiple comparisons test (F); scale bars represent 20 μm (A). See also Figure S2 and S3.

Figure 3

Draper Signaling Is Necessary in Macrophages for Normal Wound Responses, Apoptotic Cell Processing, and Basal Motility (A) Representative stills of GFP-labeled macrophages (green) at wound sites at 60 min after wounding showing a reduction in immune cell recruitment in embryos in which macrophages express Draper-II, compared to controls; white ovals denote wound outlines. (B) Scatterplot of wound responses shows defective recruitment of macrophages upon Draper-II expression at 60 min post-wounding. (C) Stills of GFP-labeled macrophages following macrophage-specific RNAi-mediated knockdown of Draper results in increased vacuolation of these cells, consistent with apoptotic corpse processing defects. (D) Scatterplot showing increase in number of vacuoles per macrophage on RNAi-mediated knockdown of Draper in macrophages; >30 macrophages from greater than ten embryos analyzed. (E) Representative macrophage tracks taken from 30-min movies of macrophages migrating in control and draper mutant embryos at stage 15, showing a reduction in basal motility in the latter; dots indicate final position of each macrophage—macrophages that leave the plane of focus during the movie terminate without a dot. (F) Scatterplot of basal motility speeds per macrophage from tracks taken from greater than four movies per genotype. Loss of draper function or macrophage-specific expression of an RNAi construct targeting Draper or overexpression of Draper-II reduces the speed of macrophage basal motility at stage 15. Central lines and error bars on scatterplots represent mean and SD, respectively; ^∗∗∗∗p < 0.0001 via the Mann-Whitney test; scale bars represent 20 μm (A and C) or 25 μm (E). See also Figure S2 and S3.

Figure 4

Draper’s ITAM Domain Is Specifically Required for Macrophage Chemotactic Migration toward Wounds (A and B) Representative stills of GFP-labeled macrophages (green) at wound sites at 60 min after wounding in control and draper mutant embryos (A) or draper mutants with macrophage-specific re-expression of either wild-type Draper-I (MΦ + Drpr-I^WT;drpr^Δ5) or a mutant form of Draper-I lacking a crucial tyrosine residue in its ITAM domain (MΦ + Drpr-I^Y949F;drpr^Δ5; B); white ovals denote wound outlines. (C) Scatterplots showing rescue of draper mutant wound responses on macrophage-specific re-expression of Draper-I^WT and a failure of Draper-I^Y949F to rescue these responses at 60 min post-wounding. (D and E) In contrast, the number of vacuoles per macrophage (D; >30 macrophages at the midline analyzed in greater than ten embryos) and the wandering macrophage migration speeds at stage 15 (E; greater than or equal to five movies per genotype) were rescued by expression of either Draper-I^WT or Draper-I^Y949F. (F) Schematic illustrating the role of the Src-Draper-Shark-signaling axis in directing inflammatory responses in the Drosophila embryo. Central lines and error bars on scatterplots represent mean and SD, respectively; ^∗∗p < 0.01 and ^∗∗∗∗p < 0.0001 according to one-way ANOVA with Sidak’s multiple comparisons test; scale bars represent 20 μm (A). See also Figures S3 and S4.