Myeloid-derived growth factor regulates neutrophil motility in interstitial tissue damage

Neutrophils are the first responders to tissue damage and are critical for mediating host defense to infection. However, excess neutrophil inflammation can also impair tissue repair. A key challenge is understanding how the migration and activation of neutrophils is controlled in sterile damage while maintaining critical host defense responses to microbial injury.

Acute tissue injury induces an array of signals that regulate neutrophil recruitment to damage in the absence of infection (de Oliveira et al., 2016). However, it is unclear how neutrophils differentially respond to tissue damage and microbial cues. Recent evidence suggests that tissue damage signals via phospholipase A2 are necessary for both sterile and microbial neutrophil recruitment (Huang and Niethammer, 2018). It remains unclear what distinguishes neutrophil responses to tissue damage signaling and microbial cues.

To address this question, we sought to identify factors that are present in sterile injury, but not infection (Houseright et al., 2020). We found that myeloid-derived growth factor (MYDGF) is up-regulated in leukocytes in response to wounding in zebrafish larvae. MYDGF has been reported to promote healing in a mouse model of myocardial infarction through effects on cell proliferation (Korf-Klingebiel et al., 2015; Wang et al., 2020). However, the role of MYDGF in regulating inflammation remains unclear.

Zebrafish larvae represent a powerful system to uncover new mechanisms that regulate inflammation due to their optical transparency, which enables noninvasive, real-time imaging of inflammation. Here, using real-time imaging, we show that MYDGF is a damage signal that influences neutrophil behavior in response to tissue damage signaling, but not microbial cues. The findings suggest that MYDGF functions as a newly identified inhibitor of neutrophils that promotes wound healing in larval zebrafish by providing a brake on neutrophil inflammation.

We sought to identify factors that are increased by tissue damage, but not infection. Using translating ribosomal affinity purification gene expression analysis (Houseright et al., 2020), we found that mydgf is up-regulated in both neutrophils and macrophages in response to tail fin wounding (Fig. S1 A). In addition, mydgf expression was increased approximately fourfold in tail fins following two different types of tail fin injury, tail transection and thermal injury (Fig. S1 B). However, we detected no difference in mydgf expression in otic injection of Pseudomonas aeruginosa (Fig. S1 C), suggesting that mydgf expression is regulated by sterile injury, but not infection. These results, together with previous reports of increased mydgf expression in ischemia-reperfusion injury in mouse models (Korf-Klingebiel et al., 2015), suggest that up-regulation of mydgf is a common response to sterile tissue injury.

MYDGF is highly conserved (53% amino acid identity and 79% similarity) between zebrafish and humans. Mapping the conserved residues onto the nuclear magnetic resonance (NMR) structure of human MYDGF (PDB accession no. 6O6W) reveals a near-complete conservation of the protein face opposite the N and C termini (Fig. 1 A). To characterize the role of MYDGF during wounding, we generated a zebrafish mydgf mutant using CRISPR-Cas9 gene editing, targeting exon 5 of mydgf and resulting in a 12-bp deletion (Fig. 1, B and C). mydgf homozygous mutants (mydgf^−/−) displayed loss of MYDGF protein by immunoblotting at 3 d postfertilization (dpf; Fig. 1 D). In agreement with previous reports in mice, we found that MYDGF-deficient zebrafish had no gross morphological defects and homozygous incrosses yielded normal larval growth (data not shown), suggesting that MYDGF is dispensable for normal development.

Next, we tested whether otic injury or administration of exogenous MYDGF altered neutrophil recruitment in WT and MYDGF-deficient larvae. Media from HEK293 cells transfected with zebrafish mydgf (mydgf-pCS2) showed secretion of MYDGF into the cell-conditioned media (CCM), as confirmed by immunoblotting (Fig. 1 E). Injection of control CCM showed increased neutrophil recruitment to the wound in the otic space in mydgf^−/− larvae compared with control (Fig. 1, F and G), suggesting that endogenous MYDGF inhibits neutrophil infiltration. Injection of zebrafish MYDGF into the otic vesicle of WT larvae decreased neutrophil recruitment to the injury compared with larvae injected with control CCM, indicating that exogenous MYDGF is sufficient to inhibit neutrophil recruitment. Moreover, injection of MYDGF into the otic space of mutant larvae rescued the increased neutrophil accumulation (Fig. 1, F and G). Together, these findings show that depletion of MYDGF increases neutrophil recruitment to damaged tissues and that addition of exogenous MYDGF is sufficient to rescue the phenotype.

To determine if MYDGF regulates neutrophil recruitment to infection, we injected P. aeruginosa into the otic space, an established model of localized infection, and quantified neutrophil recruitment using Sudan Black staining (Deng et al., 2011). We found that depletion of MYDGF had no effect on neutrophil recruitment to P. aeruginosa (Fig. 1, H and I). Taken together, our findings suggest that MYDGF limits neutrophil recruitment to tissue damage, but not infection.

Neutrophils impair wound healing in some contexts (Barros-Becker et al., 2020; Brubaker et al., 2013; Ebaid, 2014). Therefore, next we asked whether mydgf^−/− larvae have altered tail fin regeneration in response to thermal injury, a model where the presence of neutrophils impairs wound healing (Barros-Becker et al., 2020). We found that MYDGF-deficient larvae had increased neutrophil recruitment to thermal injury at both 3 and 24 h postburn (hpb; Fig. 2, A and B), consistent with our findings with otic injury (Fig. 1 G). We also observed impaired wound healing of the tail fin following thermal injury at 24, 48, and 72 hpb (Fig. 2, C and D). This change is a specific defect in wound healing, as mydgf^−/− larvae are developmentally similar to WT larvae and do not show a decrease in tail length in the absence of injury (Fig. S2 A). To determine if neutrophils mediate the impaired regeneration in MYDGF-deficient larvae, we depleted neutrophils from the wound using an established transgenic zebrafish expressing a dominant inhibitory Rac2 mutation (Rac2D57N) in neutrophils that renders neutrophils migration impaired (Deng et al., 2011). Control and Rac2D57N larvae were injected with either control morpholino oligonucleotides (mo) or mo targeting mydgf. Transient depletion of mydgf (Fig. S2 B) resulted in a 35–45% decrease in mydgf expression (Fig. S2, C and D) and increased neutrophil infiltration at wounds compared with control (Fig. S2, E–G), with no effect on total neutrophil numbers (Fig. S2 H). mo-mediated depletion of mydgf impaired regeneration in control larvae but had no effect on tail fin regeneration in the Rac2D57N larvae (Fig. 2, E and F). Together, these data show that depletion of MYDGF impairs regeneration and that this defect is due, at least in part, to the presence of neutrophils at the wound.

We imaged neutrophil recruitment in response to tissue damage to determine the effects of MYDGF on neutrophil behavior in the wound microenvironment. Neutrophil infiltration was also increased at tail transection wounds at 1 and 6 h postwound (hpw) in MYDGF-deficient Tg(mpx:mCherry) larvae, with no change in total neutrophils (Fig. 3, A–C). To characterize the mechanism of persistent neutrophil inflammation, we imaged neutrophil behavior in the wound microenvironment and found that mydgf^−/− larvae recruited more neutrophils to tissue damage, and these neutrophils remained in the wound microenvironment for a longer duration (Fig. 3, D and E). We analyzed the instantaneous speed of each neutrophil as they entered and left the wound microenvironment and found that after entering the wound, neutrophils slowed down in both WT and mutants; however, neutrophils in WT, but not mutants, increased their speed at later time points (Video 1). Within an hour of reaching the wound, neutrophils in WT larvae moved faster, with speeds approaching those at their entry into the wound, while neutrophils in mydgf^−/− larvae maintained slow speeds (Fig. 3 F). These findings suggest that neutrophil behavior is different in MYDGF-deficient larvae, resulting in the retention of neutrophils in the wound microenvironment.

Neutrophil reverse migration is an important mechanism of inflammation resolution at sterile injury (Mathias et al., 2006; Wang et al., 2017). To determine whether the defect in inflammatory resolution in mydgf^−/− larvae was due to impaired neutrophil reverse migration, we used the dendra2 photoconversion system to track neutrophil fate (Yoo and Huttenlocher, 2011). Neutrophils at the wound were photoconverted at 2 hpw, and the photoconverted neutrophils at the wound were quantified at 6 hpw (Fig. 4, A and B). We found that mydgf^−/− larvae had increased numbers of both newly arrived neutrophils and lingering photoconverted neutrophils at the wound compared with WT larvae (Fig. 4, A and B). Furthermore, mydgf^−/− larvae had significantly decreased resolution of neutrophils from the wound compared with WT larvae (Fig. 4 C). This defect in inflammatory resolution is also evident following the more robust thermal injury, with decreased resolution of neutrophils from the burn wound at 24 hpb (Fig. 4, D–F). Importantly, we did not observe a change in neutrophil apoptosis at the wound using caspase-3 staining (Fig. 4, G–I). Although macrophages mediate regeneration (Fig. S3, A and B), we did not observe altered macrophage recruitment or neutrophil-macrophage interactions (Tauzin et al., 2014) at wounds in mydgf mutants (Fig. S3, C–F; and Videos 1 and 2). Together, these data suggest that the accumulation of neutrophils in the wound in the MYDGF-deficient larvae is due to both increased recruitment of neutrophils and impaired neutrophil reverse migration.

The HIF-1α pathway is an established regulator of neutrophil accumulation and resolution by reverse migration (Elks et al., 2011). Previous studies have shown that HIF-1α regulates resolution of inflammation by neutrophil reverse migration and apoptosis (Fig. 5 A). Given that this well-defined pathway produces a neutrophil recruitment phenotype similar to the one we observed in mydgf^−/− larvae (Elks et al., 2011), we asked whether MYDGF exerts its function by modulating the HIF-1α pathway. We measured prolyl hydroxylase 3 (phd3) transcription using reverse transcription quantitative PCR (RT-qPCR) as a surrogate measure of HIF-1α activity (Walmsley et al., 2011). While expression did not differ between unwounded WT and mydgf^−/− larvae, we detected a significant increase in phd3 mRNA expression in mydgf^−/− larvae following wounding (Fig. 5 B).

Next, we tested whether disrupting the HIF-1α pathway would reduce neutrophil accumulation in wounded mydgf^−/− larvae using an established mo targeting arnt-1, a critical binding partner of HIF-1α (Fig. 5 A; Elks et al., 2011). We found that depletion of arnt-1 in mydgf^−/− larvae significantly decreased neutrophil numbers at the wound at both 1 and 6 hpw, similar to those observed in WT larvae (Fig. 5, C and D). These findings suggest that in the absence of MYDGF, HIF-1α activity is increased in response to tissue damage and leads to neutrophil accumulation at sites of tissue injury in a HIF-1α–dependent manner.

Our findings suggest that MYDGF functions as an endogenous inhibitor of neutrophil inflammation. Given its widespread expression (Bortnov et al., 2018), MYDGF may have a similar role in the regulation of inflammation in other tissues. Indeed, MYDGF expression is increased in a number of disease contexts in which the innate immune system is a major player, including diabetic nephropathy (He et al., 2020), rheumatoid arthritis (Weiler et al., 2007), and hepatocellular carcinoma (Sunagozaka et al., 2011).

The mechanisms by which MYDGF is secreted and exerts its effects are incompletely understood. MYDGF is a resident protein in the ER, and in silico modeling suggests docking of MYDGF with KDEL receptors (Bortnov et al., 2019). One of the consequences of ER stress is the release of resident ER proteins, including MYDGF (Trychta et al., 2018). Thus, MYDGF may be broadly released by stressed cells after tissue damage, and provide a beneficial effect by limiting neutrophil inflammation. Persistent neutrophil inflammation in tissues in mydgf mutants was dependent on the HIF-1α pathway, suggesting that MYDGF works with HIF-1α signaling to affect neutrophil accumulation in response to tissue damage.

In summary, we identified a new biological function for the paracrine factor MYDGF. We found that MYDGF is induced by sterile injuries and limits neutrophil inflammation in response to tissue damage, but not microbial cues. These findings provide new insight into an endogenous damage signal that puts the brakes on inflammation and neutrophil-mediated tissue injury.

Animal care and use was approved by the Institutional Animal Care and Use Committee of University of Wisconsin and strictly followed guidelines set by the federal Health Research Extension Act and the Public Health Service Policy on the Humane Care and Use of Laboratory Animal, administered by the National Institutes of Health Office of Laboratory Animal Welfare. All protocols using zebrafish in this study were approved by the University of Wisconsin-Madison Research Animals Resource Center (protocol M005405-A02). AB and NHGRI-1 WT lines (LaFave et al., 2014) were used, as well as the previously published transgenic lines mpx:mCherry (Yoo et al., 2010) mpeg1:GFP (Ellett et al., 2011), mpx:dendra2 (Yoo and Huttenlocher, 2011), mpx:mCherry-2A-rac2 (Rosowski et al., 2016), and mpx:mCherry-2A-rac2D57N (Rosowski et al., 2016). Adult fish were maintained on a 14-h/10-h light/dark schedule. Following breeding, fertilized embryos were transferred to E3 medium (5 mM NaCl, 0.17 mM KCl, 0.44 mM CaCl₂, 0.33 mM MgSO₄, 0.025 mM NaOH, and 0.0003% Methylene Blue) and maintained at 28.5°C. Larval zebrafish were anesthetized using 0.2 mg/ml tricaine (ethyl 3-aminobenzoate; Sigma-Aldrich) before any experimentation or live imaging.

Tail tissue distal to the caudal yolk extension was dissected from 3 dpf larvae following a 15-min incubation over ice. Approximately 50 tails per condition were pooled, and RNA was extracted using TRIZOL reagent (Invitrogen), according to the manufacturer’s instructions. cDNA was synthesized from the RNA using the SuperScript III FirstStrand Synthesis System with Oligo(dT) (Thermo Fisher Scientific). qPCR was performed using FastStart Essential Green Master (Roche) and a LightCycler 96 (Roche). Data were normalized to ef1α using the ΔΔCq method (Livak and Schmittgen, 2001) and expressed as fold change (2^−ΔΔCq) over pooled WT or unwounded larvae. The following primers were used: mydgf (forward, 5′-CCC​AAG​GGA​AGT​CCT​ACT​TG-3′; reverse, 5′-AGC​AAC​ATC​CCT​CTG​TCC​AC-3′), phd3 (forward, 5′-CGC​TGC​GTC​ACC​TGT​ATT-3′; reverse, 5′-TAG​CAT​ACG​ACG​GCT​GAA​CT-3′), and ef1α (forward, 5′-TGC​CTT​CGT​CCC​AAT​TTC​AG-3′; reverse, 5′-TAC​CCT​CCT​TGC​GCT​CAA​TC-3′). All PCR reactions were performed in triplicate.

A MUSCLE alignment was generated in Jalview using the protein sequences of mature human MYDGF (UniProtKB accession no. Q969H8; residues V32-L173), mature zebrafish MYDGF (UniProtKB accession no. Q6DGL1; E27-L164; The UniProt Consortium, 2017), and the 22 seed sequences of MYDGF homologues provided in Pfam (Pfam accession no. PF10572). The 80 residues of human MYDGF that were identical to zebrafish MYDGF in the sequence alignment (58% sequence identity) were mapped onto the NMR solution structure of human MYDGF. The structure presented herein was solved using NMR data (Bortnov et al., 2019) to obtain a refined solution structure of human MYDGF that has been subsequently reconciled with the crystal structure (Ebenhoch et al., 2019) as described in (Bortnov, 2020) and deposited under PDB accession no. 6O6W.

Splice-blocking morpholinos were designed to the exon 3/intron 3 border and exon 5/intron 5 border of mydgf. Morpholinos were obtained from GeneTools and resuspended in water to a stock concentration of 1 mM. Morpholinos were further diluted to a final concentration of 350 µM, and 3 nl of this injection mix was injected into the yolk sac of zebrafish embryos at the one-cell stage. Morpholino sequences are as follows: mydgf E3I3 (#1), 5′-GAG​TGT​ATA​AGT​TAC​CTC​CAT​ACT​G-3′; mydgf E5I5 (#2), 5′-GGA​ATA​GCA​TAT​ACG​CAT​GCT​CAC​C-3′;mismatch control, 5′-CCT​CTT​ACC​TCA​GTT​ACA​ATT​TAT​A-3′. A previously published morpholino targeting arnt-1, 5′-GGA​TTA​GCT​GAT​GTC​ATG​TCC​GAC​A-3′ (Prasch et al., 2006), was obtained from GeneTools and resuspended in water to a stock concentration of 1 mM. Morpholinos were further diluted to a final concentration of 500 µM and injected as above.

Short guide RNA targeting zebrafish mydgf (ENSDARG00000071679) was designed using CHOPCHOP (Montague et al., 2014). The exon 5 target sequence was 5′-GGC​CGT​GGG​TGG​ACA​GG-3′. CRISPR-Cas9 gRNA synthesis and injections were performed as previously described (Gagnon et al., 2014). NHGRI-1 embryos were injected in the yolk sac at the one-cell stage with 150 pg short guide RNA and 300 pg Cas9 protein (#CP01-50; PNA Bio). To confirm genome editing, genomic DNA was extracted from 2 dpf larvae and amplified using the following primers: forward, 5′-GTA​AAA​CGA​CGG​CCA​GTG​TGA​CAT​TGA​CAT​TTG​CCG​CA-3′; reverse, 5′-AGA​CTC​TGA​CAC​TAT​GTA​CTC​T-3′. The PCR fragments were separated on a 3% agarose gel.

Sequences of F0 mosaic cuts were confirmed by topoisomerase-based cloning (Zero Blunt TOPO PCR Cloning Kit; Thermo Fisher Scientific) and sequencing. Clutches of larvae with confirmed CRISPR cuts were grown to adulthood. Adult F0 CRISPR-injected fish were screened for germline mutations by DNA extraction and topoisomerase-based cloning and sequencing of sperm or eggs from individual fish.

Heterozygous mydgf mutants were obtained by outcrossing the CRISPR mutants to AB WT zebrafish. Offspring were genotyped using genomic DNA obtained from fin clips and amplified using the primers listed above. The PCR product was separated on a 3% agarose gel to determine individual fish genotypes. F2 and/or F3 heterozygotes were in-crossed to generate homozygous mydgf^+/+ (WT) and mydgf^−/− siblings. For experimental purposes, these fish were grown to adulthood and in-crossed to produce clutches of WT and mydgf^−/− cousins. These cousins were used only for experiments and not for line maintenance or the production of future generations.

The insert encoding mature zebrafish MYDGF (UniProtKB accession no. Q6DGL1; E27-L164) was amplified by PCR from cloned Danio rerio using primers (5′-CTG​TGT​GGT​ACC​GAA​AGG​ACC​AAA​ACA​CTG​GAC​TTC​G; 3′-TAG​AGG​GCT​AGC​TCA​GAG​TTC​ATC​ATG​CCG​TAT​TCG​AC), which introduced restriction sites for cloning (5′ Kpn1 and 3′ Nhe1) into pET.ELMER (Maurer et al., 2010). Sequences were verified before expression in bacteria.

BL21 (DE3) competent cells (#69450; MilliporeSigma) were transformed with the pET.ELMER-zebrafish MYDGF plasmid and protein expression was induced using 1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) as previously described (Bortnov et al., 2019). The recombinant protein was expressed with an N-terminal polyhistidine tag and thrombin cleavage site (residues MGGSHHHHHHGSLVPRGSKGT) preceding the mature zebrafish MYDGF sequence. The recombinant protein was extracted in an 8 M urea cell lysis solution and purified by immobilized metal affinity chromatography using nickel-nitrilotriacetic acid resin (#30230; Qiagen) as detailed previously (Bortnov et al., 2019). The protein was then refolded through dialysis against dilute acetic acid (100 mM, then 1 mM; pH 3.7), which removed components of the cell lysis solution and maximized protein solubility. After screening a variety of buffers and pH conditions, we determined that this protein construct was especially insoluble in buffers containing NaCl concentrations above 100 mM and in pH environments above pH 6. The protein was ultimately dialyzed against a pH 6.0 buffer solution containing 10 mM sodium phosphate, 100 mM NaCl, and 0.5 mM dithiothreitol and purified by size-exclusion chromatography on a HiLoad 16/600 Superdex 75 prep grade column (#28-9893-33; Sigma-Aldrich) equilibrated in and eluted with the same buffer. After verification by SDS-PAGE, fractions containing only monomeric zebrafish MYDGF were combined, and the final sample was concentrated using an Amicon Ultra-4 centrifugal filter (#UFC801024; MilliporeSigma).

The zebrafish MYDGF immunogen described above was submitted to Covance Research Products Immunology Services for generation of rabbit anti-zebrafish MYDGF antiserum. Antiserum was exposed to Protein A Sepharose beads overnight at 4°C to absorb total IgG and then washed in 50 times the bead volume PBS plus 0.1% tween and 0.5% NP-40, and in two times the bead volume PBS only. IgG was eluted with 2 × 500 µl 100 mM glycine, pH 3.0, and neutralized with 1/10th volume 1 M Tris, pH 8.0. OD at 280 nm was taken for IgG concentration.

For Western blotting, 50–100 2 dpf larvae were pooled and deyolked in calcium-free Ringer’s solution with gentle disruption from a p200 pipette tip. Larvae were washed twice with PBS and stored at −80°C until samples were lysed by sonication in 20 mM Tris, pH 7.6, 0.1% Triton X-100, 0.2 mM PMSF, 1 µg/ml Pepstatin, 2 µg/ml Aprotinin, and 1 µg/ml Leupeptin at 3 µl per larvae while on ice and clarified by centrifugation. Protein concentrations were determined using a bicinchoninic acid protein assay kit (Thermo Fisher Scientific), according to the manufacturer’s instructions. Equal amounts of total protein were loaded on 12% SDS-polyacrylamide gels and transferred to nitrocellulose membrane. Affinity-purified total rabbit anti-zebrafish MYDGF was used at 1:500 dilution. Actin (monoclonal actin AC15; #A5441; Sigma-Aldrich) or β-tubulin (mouse β-tubulin [D3U1W]; #86298; Cell Signaling Technology) was detected for loading control. Western blots were imaged and quantified by normalizing to loading controls with an Odyssey Infrared Imaging System (LI-COR Biosciences). The antibody did not work for immunofluorescence.

HEK293 cells were transfected with zebrafish mydgf-pCS2 or empty pCS2 constructs using Lipofectamine 3000 reagent (#L3000008; Invitrogen) following the manufacturer’s protocol. 16–20 h after transfection, cell media was changed to SFM4HEK293 media (HyClone; Thermo Fisher Scientific). 72 h later, media was collected, clarified by centrifugation, and aliquoted and frozen at −80°C. Samples of media were run on 12% SDS-polyacrylamide gels and checked by Western blot and InstantBlue staining (Abcam).

CCM with and without zebrafish MYDGF were mixed at two parts media to one part phenol red. Tricaine-anesthetized 3 dpf WT or mydgf^−/− mpx:mCherry zebrafish larvae were injected with 1 nl of the injection mix in the left otic vesicle. At 2 h after injection, larvae were fixed in 4% PFA at 4°C, and the area of the otic vesicle was imaged as described below.

Dechorionated, tricaine-anesthetized 3 dpf larvae were transferred to milk-coated 35-mm plates, washed twice in E3 medium, and wounded in 0.2 mg/ml tricaine/E3 solution. Tail transection was performed using a #10 scalpel blade just distal to the tip of the notochord. Larvae were transferred to zebrafish Wounding and Entrapment Device for Growth and Imaging (ZWEDGI) devices (Huemer et al., 2017), or washed 3 times in E3 medium and allowed to recover at 28.5°C until live imaging or fixation at indicated times.

Thermal injury to the caudal fin was performed as previously described (Miskolci et al., 2019). Briefly, a fine tip (type E) of a line-powered thermal cautery instrument (Stoelting) was applied to the distal tip of the tail fin of tricaine-anesthetized larvae in the E3 medium for 1–2 s until the tail fin tissue curled but before injury to the notochord was observed. Following injury, the larvae were transferred to ZWEDGI devices or washed with E3 medium and allowed to recover at 28.5°C until live imaging or fixation at indicated times.

5,000 colony-forming units (CFUs) of P. aeruginosa were microinjected in 1 nl volume as previously described (Harvie and Huttenlocher, 2015). Bacterial suspension (10,000 CFU/nl) was further diluted 1:1 with PBS containing 10% glycerol, 2% PVP-40 (polyvinylpyrrolidine; Sigma-Aldrich), and 0.33% phenol red, to a final concentration of 5,000 CFU/nl. PVP-40 was used to prevent bacterial clumping and phenol red was used to visualize injection success. Tricaine-anesthetized larvae positioned laterally on an angled injection ramp made of 2% agarose in E3 medium (in 10-cm Petri dish) and pretreated with 1 ml filter-sterilized 2% BSA to prevent larval abrasions. 1 nl volume was microinjected into otic vesicle using a thin-walled glass capillary injection needle, with time range set to “millisecond” and pressure set to ∼20 PSI on the microinjector. 2 h after injection, larvae were fixed in 4% PFA at 4°C. Larvae were stained by Sudan Black to visualize neutrophils.

GFP-expressing P. aeruginosa (pMF230) was prepared for microinjection as previously described (Rosowski et al., 2016). A single colony was grown overnight in Miller LB medium (#DSL24400; Dot Scientific) without antibiotics at 37°C with shaking. In the morning, the culture was diluted 1:5 and grown for an additional 1 h 15 min at 37°C with shaking. OD at 600 nm was measured to calculate CFUs/nl (OD 1 = ∼2.6 × 10⁵ CFUs/ml). Bacterial suspension was pelleted by centrifugation for 30 s at ∼21,000 ×g, washed three times, and resuspended in sterile PBS at 10,000 CFUs/nl.

Larvae were fixed in 4% PFA at 4°C until staining. Larvae were washed three times in PBS and then incubated for 30 min in Sudan Black B working solution (0.18% stock diluted 1:5 in 70% ethanol; 0.1% phenol). Larvae were washed twice in 70% ethanol. Depigmentation was performed by incubating the larvae in a solution of 1% potassium hydroxide and 1% H₂O₂ for 10 min. The larvae were washed and then stored at 4°C in a solution of 0.1% Tween-20 in PBS.

2 dpf morpholino-injected WT AB and mpx:mCherry or 3 dpf WT and mydgf^−/− mpx:mCherry zebrafish larvae were wounded by tail transection or thermal injury, or injected with zebrafish MYDGF protein in the otic vesicle, as described above. Larvae were fixed in 4% PFA at 4°C at 1 and 6 hpw (following tail transection), 3 and 24 hpb (following burn wound) or 2 h after otic vesicle injection. Caudal fins were imaged in PBS at room temperature on a Zeiss Zoomscope (EMS3/SyCoP3; 1× Plan-NeoFluar Z objective; Zeiss) with an Axiocam Mrm charge-coupled device camera using ZenPro 2012 software (Zeiss). For protein injection assays, images of the otic vesicle region were acquired in PBS at room temperature on a spinning-disk confocal (CSU-X; Yokogawa) on a Zeiss Observer Z.1 inverted microscope and an electron-multiplying charge-coupled device Evolve 512 camera (Photometrics), with a Plan-Apochromat 20×/NA 0.8 air objective (5-µm optical sections, 2,355 × 512 resolution) using ZenPro 2012 software (Zeiss). Neutrophil numbers were counted manually in z-projected images using Zen 2.3 Lite software (Zeiss). Neutrophil numbers were counted in the area distal to the tip of the notochord (tail transections), the area distal to the caudal vessel loop (burn wounds), or within the confines of the otic vesicle (protein injections).

2 dpf morpholino-injected WT AB or 3 dpf WT and mydgf^−/−mpx:mCherry unwounded zebrafish larvae were fixed in 1.5% formaldehyde (FA; Polysciences) in 0.1 M Pipes (Sigma-Aldrich), 1.0 mM MgSO₄ (Sigma-Aldrich), and 2 mM EGTA (1.5% FA solution; Sigma-Aldrich) overnight at 4°C. Images were acquired at room temperature using a spinning-disk confocal microscope as described above (20×/NA 0.8 air objective, 10-µm optical sections, 2 × 6 tiles, 2,355 × 512 resolution) using ZenPro 2012 software (Zeiss). Tiles were stitched together, and neutrophils were counted manually in z-projected images using Zen 2.3 Lite software (Zeiss).

Tricaine-anesthetized 3 dpf WT and mydgf^−/− larvae were wounded by tail transection or thermal injury, as described above. Larvae were washed three times with E3 medium and allowed to regenerate for 3 d after injury at 28.5°C. During regeneration, larvae were fixed in 4% PFA at 4°C at 24, 48, and 72 h after injury. Fins were imaged in PBS at room temperature using a Zeiss Zoomscope, as described above. Unwounded, age-matched larval fins were collected and imaged as a developmental control. Regenerate or developmental area of the tail fins was measured from the distal tip of the notochord (tail transection) or the caudal vessel loop (burns) using FIJI image analysis software (Rueden et al., 2017).

To quantify neutrophil contributions to burn regeneration with or without MYDGF, mpx:mCherry-2A-rac2 or mpx:mCherry-2A-rac2D57N larvae (Rosowski et al., 2016) were injected with mydgf-targeting or mismatch control morpholinos as described above and incubated to 2 dpf at 28.5°C. The larvae were then wounded by thermal injury at the caudal fin, washed three times with E3 medium, allowed to regenerate for 24 hpb, and fixed in 4% PFA. Fins were imaged and measured as described above.

3 dpf WT and mydgf^−/−mpx:mCherry/mpeg1:GFP larvae were wounded by tail transection or thermal injury, as described above. Larvae were mounted in ZWEDGI devices (in 0.2 mg/ml tricaine/E3 solution and held in place using 1% low-melting-point agarose, applied at the head) or embedded in 1% low-gelling agarose (Sigma-Aldrich) containing 0.2 mg/ml tricaine in an Ibidi µ-slide two-well glass bottom chamber (Ibidi). Images of the caudal fin region were acquired at room temperature using a spinning-disk confocal microscope (10×/NA 0.3 or 20×/NA 0.8 air objective, 5-µm optical sections, one z-stack every 2.5 min for 3 or 6 h at indicated times following injury). 3D cell tracking was performed using Imaris image analysis software (Bitplane). Leukocytes were tracked using the “Spots” function, where spot size was defined as a range of 12–15 µm. Individual tracks were curated after software detection. Tracks were then filtered according to the following criteria to eliminate false tracks and tracks showing only vibrational or nonprogressive movement; short tracks comprising fewer than five spots for neutrophils, tracks with a total displacement length of less than 14 µm (neutrophils), and tracks with a mean track speed of 0 µm/s were eliminated. For experiments comparing neutrophil instantaneous speed with position relative to the notochord, the x-position of each spot over the course of the track is normalized to the x-position of the tip of the notochord for the individual larva being measured. For experiments quantifying neutrophil-macrophage contact, the absolute number of neutrophils and macrophages for each frame was normalized using the formula # of contacts × number of neutrophils/number of macrophages.

3 dpf WT or mydgf^−/− mpx:dendra larvae were wounded by tail transection or thermal injury, as described above. Tricaine-anesthetized larvae were mounted in ZWEDGI devices and held in place using 1% low melting point agarose, applied to the head. An imaging sequence was performed on each larva comprising an initial z-stack of the caudal fin area, followed by photoconversion of the neutrophils caudal to the tip of the notochord (tail transection) or vessel loop (burn wounds) and second z-stack of the caudal fin area after photoconversion. The photoconversion and associated imaging sequences were performed at room temperature using a laser-scanning confocal microscope (FluoView FV-1000, 20×/NA 0.75 air objective; Olympus). The following stimulation settings were used for photoconversion: 40% 405 nm laser transmissivity, 10 µs/pixel dwell time, and 45 s total stimulation time. Photoconversion was performed at 2 hpw for tail transections and 4 hpb for burn wounds. Larvae were maintained in E3 medium in individual wells of a 96-well plate at 28.5°C until final imaging.

At 6 hpw (tail transections) or 24 hpb (burn wounds), larvae were fixed in 1.5% FA solution overnight at 4°C. Larvae were then mounted in 1% low-melting-point agarose, and images of the tail region were acquired using a spinning-disk confocal microscope, as described above for neutrophil recruitment assays. Photoconverted (red) and nonphotoconverted (green) neutrophils in the area distal to the tip of the notochord (tail transections) or distal to the vessel loop (burn wounds) were counted manually using FIJI image analysis software (Rueden et al., 2017).

Immunostaining to identify apoptotic (active caspase-3 positive) neutrophils was performed using 3 dpf WT or mydgf^−/−mpx:mCherry larvae. Larvae wounded by tail transection were fixed at 6 hpw in 4% PFA at 4°C overnight and then stored in methanol at −20°C until immunostaining. Larvae were incubated with monoclonal rabbit anti-active caspase-3 antibody (#559565; BD Biosciences) at 1:200 in block (PBS, 1% DMSO, 1% BSA, 0.05% Triton X-100, and 1.5% goat serum), followed by incubation with Alexa Fluor 488 donkey anti-rabbit secondary antibody (Thermo Fisher Scientific). The caudal fin area was then imaged at room temperature in PBS using a spinning-disk confocal microscope, as described above (20×/NA 0.8 air objective, 5-µm optical sections). Apoptotic neutrophils were defined as discrete spots in the area distal to the tip of the notochord expressing both red (neutrophil) and green (active caspase-3) fluorescence. Apoptotic neutrophils were counted manually using Zen 2.3 Lite software (Zeiss).

Macrophage depletion was performed by injection of clodronate liposomes. At 2 dpf, anesthetized (Tg(mpeg1:GFP)) larvae were injected intravenously via the posterior caudal vein with 1 nl liposome-encapsulated clodronate (www.clodronateliposomes.org) or PBS. The clodronate liposome–injected fish were sorted for GFP-negative (macrophage depleted) larvae. Thermal injury of the caudal fin was performed on GFP-negative larvae, along with PBS liposome–injected controls at 3 dpf. The wound healing of larvae was monitored over time by live imaging at room temperature in 0.2 mg/ml tricaine/E3 medium at 1, 2, 3, and 4 d after burn on a Zeiss Zoomscope (EMS3/SyCoP3; 1× Plan-NeoFluar Z objective; Zeiss) with an Axiocam Mrm charge-coupled device camera using ZenPro 2012 software (Zeiss). Tail fin regrowth area was then measured for each larva as described above.

Independent biological replicate is defined as a separate clutch of larvae spawned on different days. RT-qPCR gene expression analyses comprised three to five independent biological replicates, and reactions were performed in three technical replicates. Statistical significance was determined by comparing the calculated ΔCq of the experimental conditions using the nonparametric Wilcoxon two-group test. Fold change in ΔCq was calculated and plotted in terms of mean with 95% confidence interval (CI). Neutrophil quantification and migration analyses and regeneration assays comprised three or four biological replicates. Replicate numbers are noted in the figure legends. Experimental conditions were compared using analysis of variance. Comparisons between two groups were performed using an unpaired, two-tailed ttest. Data distribution was assumed to be normal, but this was not formally tested. For quantification of neutrophil instantaneous speed over position, a linear mixed effect regression model was used. Genotype and position were treated as fixed effects, with experimental replicate, fish, and neutrophil (within fish) treated as random effects. Statistical analyses were performed in R version 3.5.1 (www.r-project.org) using the associated lme4 package. Reported P values are two sided, and the level of statistical significance was preset to 0.05, with no adjustment for multiplicity.

Fig. S1 displays mydgf expression in response to sterile wounds and infection as measured by translating ribosome affinity purification RNA sequencing and qPCR. Fig. S2 shows efficiency of morpholino-mediated depletion of mydgf and recruitment of neutrophils to tail transection and total number of neutrophils in morpholino-depleted zebrafish larvae. Fig. S3 depicts the characterization of macrophage response and neutrophil–macrophage interaction at sterile wounds. Video 1 shows neutrophil and macrophage response following tail transection of the caudal fin. Video 2 shows neutrophil and macrophage response following thermal injury of the caudal fin.

We thank members of the Huttenlocher laboratory for valuable discussions.

This work was supported by the National Institutes of Health (grant R35GM118027 to A. Huttenlocher and grant T32HL07899 to R.A. Houseright) and the American Heart Association (fellowship 17POST33410970 to V. Miskolci).

The authors declare no competing financial interests.

Author contributions: R.A. Houseright and A. Huttenlocher conceived and designed experiments. R.A. Houseright, V. Miskolci, O. Mulvaney, D.A. Bennin, and J. Rindy conducted the experiments and analysis. R.A. Houseright, V. Miskolci, D.F. Mosher, V. Bortnov, and A. Huttenlocher prepared the figures and wrote the manuscript.