Simu-dependent clearance of dying cells regulates macrophage function and inflammation resolution

Abstract

Macrophages encounter and clear apoptotic cells during normal development and homeostasis, including at numerous sites of pathology. Clearance of apoptotic cells has been intensively studied, but the effects of macrophage-apoptotic cell interactions on macrophage behaviour are poorly understood. Using Drosophila embryos, we have exploited the ease of manipulating cell death and apoptotic cell clearance in this model to identify that the loss of the apoptotic cell clearance receptor Six-microns-under (Simu) leads to perturbation of macrophage migration and inflammatory responses via pathological levels of apoptotic cells. Removal of apoptosis ameliorates these phenotypes, while acute induction of apoptosis phenocopies these defects and reveals that phagocytosis of apoptotic cells is not necessary for their anti-inflammatory action. Furthermore, Simu is necessary for clearance of necrotic debris and retention of macrophages at wounds. Thus, Simu is a general detector of damaged self and represents a novel molecular player regulating macrophages during resolution of inflammation.

Fig 1. Apoptotic cell death contributes to developmental dispersal of Drosophila embryonic macrophages.

(a–b) Lateral projections (anterior is left) of control (w;srp-GAL4,UAS-GFP) and apoptosis-null Df(3L)H99 mutant embryos (w;srp-GAL4,UAS-GFP;Df(3L)H99) showing macrophage distribution at stage 13/14 of development, including along ventral midline (arrow). (c–d) Ventral projections (anterior is up) showing macrophages (anti-GFP staining, green in merge) and structure of the CNS (22c10/anti-Futch staining, purple in merge) at stage 13 of development in controls (w;;crq-GAL4,UAS-GFP) and in the absence of apoptosis (w;;Df(3L)H99,crq-GAL4,UAS-GFP). Macrophage projections constructed from z-slices corresponding to superficial macrophages on the ventral side of the VNC only, whereas Futch projection covers the entire volume of the VNC. (e–f) Ventral projections (anterior is up) showing macrophage distribution (GFP) alone (e–f), or merged with Futch staining (e’–f’) to show CNS structure in control and Df(3L)H99 mutant embryos at stage 15 (genotypes as per c–d). (g) Scatterplot showing percentage of macrophages on the midline that move laterally at stage 13 of development in control and apoptosis-null embryos (n = 9 and 10, respectively; P = 0.0006, Mann–Whitney test); lateral migration concludes around the end of stage 14. (h) Scatterplot showing speed per macrophage, per embryo (μm per min) in control and apoptosis-null embryos at stage 13 (P < 0.0001) and stage 15 (n = 18, 8, 14, 6 [left–right]; P = 0.47, one-way ANOVA with Tukey’s multiple comparison test). Genotypes in (g–h) are as per (c–f). (i) Stills from movie of macrophages phagocytosing at stage 13 on the ventral midline in a control embryo (w;srp-GAL4,UAS-2xFYVE-GFP;crq-GAL4,UAS-CD4-tdTomato). GFP and tdTomato are shown in green and purple, respectively, while asterisk shows nascent phagosome at indicated times (mins:secs) as it becomes positive for the PI3P sensor 2xFYVE-GFP post engulfment; CD4-tdTomato labels membranes. (j–m) Ventral projections of control (w;srp-GAL4,UAS-2xFYVE-GFP;crq-GAL4,UAS-tdTomato) and apoptosis-null embryos (w;srp-GAL4,UAS-2xFYVE-GFP;crq-GAL4,UAS-tdTomato;Df(3L)H99) at stage 13 (j–k) and stage 15 (l–m), showing recent phagocytic events via 2xFYVE-GFP sensor for PI3P. Asterisks show examples of 2xFYVE-GFP positive cells (j, l). (n) Scatterplot showing number of phagocytic events (via 2xFYVE-GFP positive vacuoles) per macrophage, per embryo at stages 13 and 15 (n = 19 and 15, respectively; P < 0.0001, Mann–Whitney test). Lines and error bars show mean and standard deviation (g–h, n); *** and **** denote P < 0.001 and P < 0.0001; scale bars represent 50 μm (a–b), 10 μm (c–f), 5 μm (i), and 10 μm (j–m). All data used to plot graphs may be found in Supporting information file S1 Data. CNS, central nervous system; GFP, green fluorescent protein; PI3P, phosphatidylinositol 3-phosphate; UAS, upstream activating sequence.

Fig 2. Loss of simu function results in exposure of macrophages to pathological levels of uncleared apoptotic cells.

(a–d) Maximum projections of control embryos (w;srp-GAL4,UAS-GFP/+;crq-GAL4,UAS-GFP/+) immunostained to show close contact of macrophages (anti-GFP, green), and glia (anti-Repo, purple) at stage 12 during migration along both sides of the ventral midline (a, lateral; b, ventral) and later in development at stage 15 (c, lateral; d, ventral). (a’) and (c’) are maximum projections of a smaller number of z-slices to show position of macrophages between epidermis and and VNC. Anterior is left and arrows show pioneer macrophages; arrows ‘v’ and ‘d’ in (b) label most advanced macrophages moving along the ventral and dorsal side of the VNC, respectively; arrows in (c’) show two macrophages between VNC and epidermis. (e–f) Maximum projections of stage 15 control and simu mutant embryos showing macrophages (GFP, green in merge) and apoptotic particles (cDCP-1, purple in merge), respectively. Projections are ventral views and correspond to void between epidermis and VNC. (e’–f’) Show zooms of macrophages indicated by arrows in (e–f); white line shows edge of embryos. (g) Scatterplot of untouched apoptotic punctae (cDCP-1 punctae not in contact with macrophages in the field of view) per embryo in control and simu mutant embryos at stage 15 (n = 7 per genotype; P = 0.0006, Mann–Whitney test). (h) Scatterplot of phagocytic index (cDCP-1 punctae engulfed per macrophage, per embryo) in control and simu mutant embryos at stage 15 (n = 6 and 5 embryos, respectively; >10 cells analysed per embryo, 96 and 64 macrophages analysed in total, respectively; P = 0.017, Mann–Whitney test). Genotypes in (e–h) are w;;crq-GAL4,UAS-GFP (control) and w;simu²;crq-GAL4,UAS-GFP (simu mutants). Error bars represent mean ± standard deviation; * and *** denote P < 0.05 and P < 0.001; scale bars represent 10 μm (a–d), 20 μm (e–f), and 5 μm in zoomed panels (e’–f’). All data used to plot graphs may be found in Supporting information file S1 Data. cDCP-1, cleaved death caspase 1; GFP, green fluorescent protein; UAS, upstream activating sequence; VNC, ventral nerve cord.

Fig 3. Loss of simu function leads to defects in developmental dispersal of macrophages and an early build up of apoptotic cells.

(a–b) Lateral images of control (w;srp-GAL4,UAS-red stinger) and simu mutant embryos (w;simu²,srp-GAL4,UAS-red stinger) showing migration of red stinger-labelled macrophages along the ventral midline at stage 13; arrows indicate ventral midline, anterior is left, ventral is up. (c) Bar graph showing percentage of segments with GFP-labelled macrophages on the ventral side of the VNC at stage 13 in controls (w;;crq-GAL4,UAS-GFP) and simu mutants (w;simu²;crq-GAL4,UAS-GFP) (n = 15 and 15, respectively; P > 0.999, Mann–Whitney test). (d) Total numbers of red stinger-labelled macrophages in lateral views of control and simu mutants (genotypes as per [a–b] at stage 13; n = 7 for each; P = 0.066, Mann–Whitney test). (e–f) GFP-labelled macrophages on the ventral side of the embryo in controls (e) and simu mutants (f); imaging started at the end of stage 12 with time indicating duration of imaging. (g) Numbers of GFP-labelled macrophages on the ventral surface of embryos at the indicated timepoints in control and simu mutants (n = 17 and16, respectively, P = 0.0162 and 0.046 and at 0 and 90 mins, respectively, Student t tests); genotypes as per (c) in e–g. (h–i) Ventral views showing migration of mCherry-labelled macrophages (purple in merge) down the ventral midline (anterior is up) encountering GC3ai-positive apoptotic cells (green in merge) during stage 12 in control and simu mutant embryos. (j) GC3ai fluorescence levels (arbitrary units) quantified from average projections of the ventral midline region of control and simu mutant embryos at stage 12; genotypes are w;;da-GAL4,UAS-GC3ai/srp-3x-mCherry (control) and w;simu²;da-GAL4,UAS-GC3ai/srp-3x-mCherry (simu) in (h–j) (n = 7 and 6, respectively; P = 0.0004, Student t test). Lines/bars and error bars show mean and standard deviation; ns, * and *** denote not significant, P < 0.05 and P < 0.001, respectively; scale bars represent 50 μm (a–b) and 20 μm (e–f, h–i). All data used to plot graphs may be found in Supporting information file S1 Data. GFP, green fluorescent protein; UAS, upstream activating sequence.

Fig 4. Pathological levels of apoptosis impair motility of macrophages and inflammatory migration to wounds.

(a–b) Images of GFP-labelled macrophages and their associated tracks on the ventral midline at stage 15 in control (w;;crq-GAL4,UAS-GFP, a) and simu mutant (w;simu²;crq-GAL4,UAS-GFP, b) embryos. (a’–b’) Tracks show macrophage migration over a 30-minute period; dots show starting positions of those macrophages present in the first frame of the movie; anterior is left. (c) Scatterplot of average speed per macrophage, per embryo on the ventral midline at stage 15 in μm per min in controls and simu mutants (n = 7 and 6, respectively; P = 0.035, Mann–Whitney test). (d–e) Images of GFP-labelled macrophages on the ventral midline in stage 15 control (d–d’) and simu mutant (e–e’) embryos before (d–e) and at 60-minutes post wounding (d’–e’). Dotted lines indicate wound sites. (f) Scatterplot of wound response at 60 minutes (number of macrophages at wound divided by wound area, normalised to control average) in control and simu mutant embryos (n = 23 and 16, respectively; P < 0.0001, Mann–Whitney test). (g) Scatterplot showing percentage of macrophages responding to wounds (percentage of those in field of view at t = 0 not initially in contact with the wound site that reach the wound site in a 60-minute movie) in control and simu mutant embryos (n = 7 and 9, respectively; P = 0.0006, Mann–Whitney test). All genotypes as per (a–b). Lines and error bars show mean and standard deviation; *, ***, and **** denote P < 0.05, P < 0.001, and P < 0.0001 (c, f, g); scale bars represent 20 μm (a–b) and 10 μm (d–e). All data used to plot graphs may be found in Supporting information file S1 Data. GFP, green fluorescent protein; UAS, upstream activating sequence.

Fig 5. Acute induction of apoptosis impairs inflammatory responses without the requirement for engulfment by macrophages.

(a) Ventral projections of embryos containing hs-hid (w;CyO hs-hid/+) stained for active caspases (anti-cDCP-1) following heat-shock for 0, 15, or 30 minutes; embryos fixed and stained at indicated times. (b) Schematic showing experimental design to induce exogenous apoptosis ahead of clearance by phagocytes in the embryo. (c) Scatterplot showing vacuoles per macrophage, per embryo as a read out of phagocytosis at 60 minutes post heat shock (immediately before wounding) in control and hs-hid embryos (n = 12 and 14, respectively; P = 0.829, Mann–Whitney test). (d) Prewound and 60-minute postwound ventral views of control (w;srp-GAL4,UAS-GFP/CyO) and hs-hid embryos (w;srp-GAL4,UAS-GFP/CyO hs-hid) subjected to 15-minute heat-shock; experimental design was as per (b), with wounding taking place at 60 minutes post heat shock. Dotted lines show wound edges. (e) Scatterplot of wound responses (density of macrophages at wound sites normalised to wound area and to control average) at 60 minutes post wounding of control and hs-hid embryos (n = 9 and 11, respectively; P = 0.0005, Mann–Whitney test). Line and error bars show mean and standard deviation in all scatterplots; ns and *** denote not significant and P < 0.001; scale bars represent 20 μm. All data used to plot graphs may be found in Supporting information file S1 Data. cDCP-1, cleaved death caspase 1; GFP, green fluorescent protein; UAS, upstream activating sequence.

Fig 6. Removal of apoptosis from simu mutants rescues wandering migration and initial responses to wounds.

(a–a’) Maximum projections of GFP-labelled macrophages on the ventral midline at stage 15 (a) and tracks of their migration in the subsequent 60 minutes (a’) in controls, apoptosis-null embryos (Df(3L)H99), simu mutants (simu²) and simu mutants that lack apoptosis (simu²;Df(3L)H99). (b–b’) Scatterplots of speed per macrophage (b) and average speed per embryo (b’) in μm per min at stage 15 in control and apoptosis-null embryos (n = 7 and 9, respectively; P = 0.252, Mann–Whitney test), and simu mutants and simu mutants that lack apoptosis (n = 8 and 10, respectively; P = 0.006, Mann–Whitney test, b’). (c) Maximum projections of macrophages at wounds 60-minutes post wounding in indicated embryos; dotted white ellipses show wound edges. (d) Scatterplot of wound responses (macrophage density at wounds normalised to control average) comparing control and apoptosis-null embryos (n = 22 and 26, respectively; P = 0.039, Mann–Whitney test), and simu mutants and simu mutants that lack apoptosis (n = 18 and 19, respectively; P = 0.30, Mann–Whitney test). (e) Scatterplot of percentage of responding macrophages (percentage that migrate to the wound that were not at the wound at t = 0 minutes) comparing control and apoptosis-null embryos (n = 18 and 27, respectively; P = 0.80, Mann–Whitney test), and simu mutants and simu mutants that lack apoptosis (n = 25 and 18, respectively; P = 0.019, Mann–Whitney test). Genotypes for all embryos: control (w;;crq-GAL4,UAS-GFP), Df(3L)H99 (w;;Df(3L)H99,crq-GAL4,UAS-GFP), simu² (w;simu²;crq-GAL4,UAS-GFP) and simu²;Df(3L)H99 (w;simu²;Df(3L)H99,crq-GAL4,UAS-GFP). Lines and error bars in scatterplots show mean and standard deviation; ns, *, **, **** denote not significant, P < 0.05, P < 0.001, and P < 0.0001; scale bars represent 20 μm. All data used to plot graphs may be found in Supporting information file S1 Data. GFP, green fluorescent protein; UAS, upstream activating sequence.

Fig 7. Macrophages that fail to respond to wounds in simu mutants interact with uncleared apoptotic cells.

(a–c) Ventral projections of the midline region and associated tracks of macrophages (mCherry, purple) migrating to wounds in control (a) and simu mutant embryos (b–c) in the presence of the caspase reporter GC3ai (green) to label apoptotic cells/fragments at the indicated times. Asterisks show centre of wound and dotted white lines show wound edges at 60 minutes; arrows indicate examples of regions in which interactions between GC3ai punctae and macrophages occur, disrupting wound responses. (b’–c’) Show zooms and single-channel images of the boxed regions in (b–c). (d–e) Percentage of cells responding to wounding (d) and proportions of nonresponding cells that interact with GC3ai punctae (e) in control and simu mutant embryos (n = 9 for each genotype, P < 0.0001 [d] and P = 0.0007 [e], via Mann–Whitney test). All genotypes are w;;da-GAL4,UAS-GC3ai/srp-3x-mCherry (control) or w;simu²;da-GAL4,UAS-GC3ai/srp-3x-mCherry (simu). N.b., images have not been rotated in order to provide a larger field of view to visualise a greater number of macrophage–apoptotic cell interactions; scale bars represent 20 μm; lines and error bars on (d–e) show mean and standard deviation; *** and **** denote P < 0.001 and P < 0.0001. All data used to plot graphs may be found in Supporting information file S1 Data. UAS, upstream activating sequence.

Fig 8. Simu is required to prevent precocious exit of macrophages from sites of inflammation.

(a) GFP-labelled macrophages and their associated tracks at 0, 44, or 48 and 60 minutes post wounding in controls and simu mutants lacking apoptosis (simu²;Df(3L)H99). Left panels show macrophages present in field of view at 0 minutes; asterisk marks centre of wound. Central panels show tracks of macrophages that migrate to and then leave the wound (‘leavers’). Right panels show tracks of cells present at 0 minutes that remain in the field of view until 60-minutes post wounding. (b–c) Scatterplots showing percentage of macrophages that migrate away from wounds over the course of 60-minute movies of inflammatory responses in controls, apoptosis-null embryos (Df(3L)H99), simu mutants (simu²), and simu mutants lacking apoptosis (simu²;Df(3L)H99) (b) and upon macrophage-specific re-expression of wild-type simu in a simu mutant background (c); P values compared to control in (b) are P = 0.999, 0.047, and 0.031, respectively from left to right (Kruskal–Wallis test with Dunn’s multiple comparisons post-test; n = 18, 27, 25, 18); for (c) P = 0.004 (Mann–Whitney test, n = 14 and 18). (c’) Scatterplot of percentages of cells responding to wounds in simu mutants and simu mutants with re-expression of wild-type simu in macrophages (n = 14 and 18, P = 0.045, Mann–Whitney test). Genotypes are as follows: (a–b) control (w;;crq-GAL4,UAS-GFP), Df(3L)H99 (w;;Df(3L)H99,crq-GAL4,UAS-GFP), simu² (w;simu²;crq-GAL4,UAS-GFP), and simu²;Df(3L)H99 (w;simu²;Df(3L)H99,crq-GAL4,UAS-GFP); (c) simu² (w;simu²,srp-GAL4,UAS-red stinger/simu²;crq-GAL4,UAS-GFP/+) and simu²;UAS-simu (w;simu²,srp-GAL4,UAS-red stinger/simu²;crq-GAL4,UAS-GFP/UAS-simu). Lines and error bars show mean and standard deviation in scatterplots; scale bars represent 20 μm (a). All data used to plot graphs may be found in Supporting information file S1 Data. GFP, green fluorescent protein; UAS, upstream activating sequence.

Fig 9. Simu recognises and facilitates removal of necrotic debris at wounds in a PS-dependent manner, preventing early exit of macrophages from sites of inflammation.

(a) PI staining to show necrotic cells immediately after wounding in a stage 15 control embryo (GFP-labelled macrophages, green; PI, purple). Dotted lines in left-hand panel show wound edge; right-hand panel shows zoom of wound. (b) Panels show GFP-labelled macrophages (green in merged image), PS staining (via annexin V, purple in merged image) and merged image at a wound site at 60-minutes post wounding of a stage 15 control embryo. (c) Maximum projections of GFP-labelled macrophages in apoptosis-null embryos (Df(3L)H99) and simu mutants that lack apoptosis (simu²;Df(3L)H99) at 60-minutes post wounding; note absence of vacuoles in macrophages away from the wound (indicated via dotted line). (c’) Shows zooms of wound regions taken from single z-slices from image stacks used to make projections in (c). (d) Scatterplot comparing numbers of vacuoles present in macrophages at wounds in apoptosis-null embryos (Df(3L)H99) with apoptosis-null embryos that lack simu (simu²;Df(3L)H99) (n = 11 and 9, respectively; P < 0.0001, Mann–Whitney test), and apoptosis null embryos with (Df(3L)H99 + simu RNAi) or without (Df(3L)H99) macrophage-specific RNAi-mediated knockdown of simu (n = 10 and 9, respectively; P < 0.0001, Mann–Whitney test). (e) Heat treatment of S2 cells causes all cells to undergo necrosis and become labelled with propidium iodide (purple). (f) Images showing that expression of wild-type simu but not a truncated form of simu that cannot bind PS (ΔEMI-2) increases recognition of necrotic S2 cells by larval macrophages compared to controls. hmlΔ-GAL4 used drive larval macrophage-specific expression of UAS-GFP (green) and UAS-simu constructs; S2 cells labelled using cell tracker (purple) prior to induction of necrosis; inset shows zoomed examples of macrophages binding and engulfing necrotic S2 cells (indicated by boxes in main panel). (g) Scatterplot showing quantification of binding and/or engulfment of necrotic cells by larval macrophages (n = 15 per genotype, P < 0.0001 [control versus UAS-simu], P = 0.433 [control versus UAS-ΔEMI-2] and P < 0.0001 [UAS-simu versus UAS-simuΔEMI-2], one-way ANOVA). (h) Scatterplot showing that macrophage-specific expression of simu ΔEMI-2 leads to enhanced exit of macrophages from wound sites compared to wild-type simu in a simu mutant background (w;simu²,srp-GAL4,UAS-red stinger,Ecad-mCherry/simu²;crq-GAL4,UAS-GFP/UAS-simu or UAS-simu ΔEMI-2; n = 26 and 21, respectively, P = 0.039, Mann–Whitney test). (i) Schematic summarising roles of apoptotic cells and Simu in macrophage dispersal, migration, retention at wounds, and phagocytosis. Lines and error bars on graphs show mean and standard deviation; *, ****, and ns represent P < 0.05, P < 0.0001, and not significant; scale bars represent 20 μm (a, c) and 10 μm (b) and 50 μm (e–f). All data used to plot graphs may be found in Supporting information file S1 Data. GFP, green fluorescent protein; PI, propidium iodide; PS, phosphatidylserine; RNAi, RNA interference; UAS, upstream activating sequence.