Perivascular Macrophages Convert Physical Wound Signals Into Rapid Vascular Responses

Abstract

Leukocytes detect distant wounds within seconds to minutes, which is essential for effective pathogen defense, tissue healing, and regeneration. Blood vessels must detect distant wounds just as rapidly to initiate local leukocyte extravasation, but the mechanism behind this immediate vascular response remains unclear. Using high-speed imaging of live zebrafish larvae, we investigated how blood vessels achieve rapid wound detection. We monitored two hallmark vascular responses: vessel dilation and serum exudation. Our experiments-including genetic, pharmacologic, and osmotic perturbations, along with chemogenetic leukocyte depletion-revealed that the cPla₂ nuclear shape sensing pathway in perivascular macrophages converts a fast (~50 μm/s) osmotic wound signal into a vessel-permeabilizing, 5-lipoxygenase (Alox5a) derived lipid within seconds of injury. These findings demonstrate that perivascular macrophages act as physicochemical relays, bridging osmotic wound signals and vascular responses. By uncovering this novel type of communication, we provide new insights into the coordination of immune and vascular responses to injury.

Keywords: ER lipid flow; cPla2; epithelial wounding; macrophage; mechanotransduction; nuclear membrane tension; vessel dilation; vessel osmotic surveillance; vessel permeabilization.

Figure 1.. Osmotic surveillance mediates rapid wound detection by blood vessels.

(a) Upper panel, cartoon scheme representing the experimental workflow of fluorescence microangiography, tailfin wounding, and ionic/osmotic bath treatment. HYPO, regular E3 solution. ISO_(NaCl), E3 adjusted to interstitial osmolarity by supplementing E3 with 135 mM NaCl. ISO*, E3 adjusted to isotonicity with other salts/osmolytes. Dotted box indicates time of solution switch (shift). (b) Cartoon schemes depicting the regions of interest for measuring the dextran permeability of vessels (Iv), wounds (Ib), as well as vessel dilation (Dv). Left panel, confocal maximal intensity projection (MIP) of wounded endothelial reporter larvae (Tg(kdrl:eGFP)) before (T= 270 s) and after (T= 3600 s) switch of bathing solutions. Green, kdrl:eGFP fluorescence. Magenta, pseudo-coloured 70 kDa dextran fluorescence. Right panel, kymographs of vessel diameter before and after switch of bathing solution. (c) Top panel, rate (dI dt⁻¹) of vessel leakage (blue) and wound leakage (red) versus time upon switching from ISO_(NaCl) to HYPO. Arrows indicate the maximal rate of change for vessel permeability (cyan) or wound permeability (black). Bottom panel, rate (dI dt⁻¹) of vessel leakage versus time upon switching from ISO_(NaCl) to HYPO (blue) or ISO_(NaCl) to ISO_(NaCl). Shaded error bars, SD. Note, the blue curves represent the same, replotted dataset. (d) Quantification of normalized, integrated (between T= 0–3600 s) dextran leakage (Iv_tot/norm) or (e) or normalized dextran leakage from the wound (Ib_norm) measured at T= 3600 s. (f) Maximal, normalized endothelial diameter (max(Dv_norm)), obtained from kymograph). If indicated, the raw data was normalized by the mean of the first 10 frames (T= 0–270 s, i.e., preceding solution switching). Note, ISO* contains pooled results from different isotonic salt/osmolyte treatments. Error bars, SD. White numbers, animals. Black numbers, mean of dataset. P values, one-way analysis of variance (ANOVA) with Dunn’s post-hoc test (in d, f, e). All P values are listed in the Supplementary Excel File 1 (F1_T1). *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. Scale bars, 50 μm.

Figure 2.. Osmotic control of vessel permeability requires Alox5a.

(a) simplified scheme of enzymatic derivatives of AA, the pathways enzymes tested in the study are highlighted in red. Representative confocal MIP images of pseudo-coloured 70 kDa dextran fluorescence before (T=270 s, left panel) and after (T-3600 s, right panel) ISO_(NaCl) → HYPO shift in wild-type animals (Casper), (b) alox12^mk215/mk215 mutants, (c) alox5a^mk211/mk211 mutants, (d) and lta₄h^mk216/mk216 mutants. 3 dpf larvae were compared to their matched wild-type siblings. (e-g) Quantification of integrated (T= 0–3600 s), normalized vessel leakage and wound permeability (T= 3600 s). The P values in (e) and (g) were calculated using a non-paired, two-tailed student’s t-test. The data in (f) were not normally distributed, and a two-sided Mann–Whitney U test was used instead. (h) Integrated vessel leakage after pretreatment of wild-type (Casper) larvae with 130 nM Diclofenac, 50 μM Licofelone and 10 μM MK886, or DMSO (vehicle) and ISO_(NaCl) → HYPO shifting. P values, one-way ANOVA with Dunn’s post-hoc test. White numbers, animals. Black numbers, mean of dataset. Error bars, SD. All figure source data and numerical P values are listed in the Supplementary Excel File 1. **P ≤ 0.01, ****P ≤ 0.0001. ns, not significant (P > 0.05). Timestamp, hh: mm: ss. Scale bars, 50 μm.

Figure 3.. Macrophages mediate rapid osmotic vessel permeabilization.

(a) Experimental timeline for metronidazole-induced depletion of macrophages or neutrophils in 3 dpf Tg(mpeg1.1: YFP-NTR2.0) or Tg(lyz: YFP-NTR2.0) larvae, respectively. Blue arrow, start of the vehicle treatment (0.15% DMSO). Black arrow, start of the metronidazole (150 μM MTZ) treatment. Magenta arrow, dextran injection. Red arrow, start of imaging. (b) Representative confocal MIPs of leukocytes and tissue dextran fluorescence before (T= 270 s) and after (T= 3600 s) ISO_(NaCl) → HYPO shifting. Green, pseudo-coloured lyz: YFP-NTR2.0 (upper panel) or mpeg1.1: YFP-NTR2.0 (lower panel) fluorescence. Magenta, pseudo-coloured 70 kDa dextran fluorescence. (c, d) Quantification of vessel and wound leakage in neutrophil- or macrophage-depleted animals. P values, non-paired, two-tailed Student’s t-test. (e) Quantification of vessel leakage after ISO_(NaCl) → HYPO shifting in macrophage-containing (DMSO + DMSO, DMSO + Licofelone) or -depleted (MTZ + Licofelone) larvae ± 50 μM Licofelone or 150 μM MTZ, respectively. P value, one-way ANOVA with Dunn’s post-hoc test. White numbers, animals. Black numbers, mean of dataset. Error bars, SD. The figure source data and numerical P values are listed in the Supplementary Excel File 1. ****P ≤ 0.0001. ns, not significant (P > 0.05) (c-e). Timestamp, hh:mm:ss. Scale bars, 50 μm.

Figure 4.. Nuclear shape sensing by cPla₂ relays rapid osmotic wound cues to vessels.

(a) Quantification of vessel and wound leakage in homozygous (pla2g4aa^mk217/mk217) and heterozygous (pla2g4aa^mk217/wt as control) cPla₂ mutant animals. (b-c) Representative MIPs of Tg(mpeg1.1: cPla₂-mKate2) larvae before and after laser injury (dashed, red circle) in HYPO (10 mOsm, upper panel) or ISO (280 mOsm, lower panel) bathing solution. Magenta pseudo-colour, mpeg1.1: cPla₂-mKate2. Green dashed line, vessel outline. Insets, correspond to the cells within the numbered ROIs depicted in the middle panel. Time stamp, hh:mm:ss. Scale bars, 50 μm (d) Line profiles of cPla₂-mKate2 fluorescence in the ROI-labelled cells show translocation after laser injury in HYPO but not in (e) ISO bathing solution. (f) Left panel, cPla₂-mKate2-INM binding dynamics. UV laser injury is induced at T= 40 s. INM-binding of cPla₂-mKate2 is quantified as a ratio of nucleoplasmic to perinuclear fluorescence signal and normalized to its initial (T= 0 s) value. Right panel, comparison of peak translocation. N, number of analysed nuclei from 53 and 59 perivascular macrophages for ISO_(NaCl) and HYPO laser wounds, respectively. Note, the panel depicts two different representations of the same dataset. (g) Simplified cartoon scheme. Perivascular macrophage nuclei are reversibly stretched by osmotic wound signals. In the presence of Ca²⁺, nuclear membrane tension causes arachidonic acid (AA) release by cPla₂. AA is metabolized into a vessel permeabilizing lipid mediator. P values, non-paired two-tailed Student’s t-test (a) and non-paired two-tailed Mann Whitney test (f). White numbers, number of analysed nuclei pooled from 16–17 animals. Black numbers, mean of dataset. Error bars, SD. The figure source data and numerical P values are listed in the Supplementary Excel File 1. ****P ≤ 0.0001. ns, not significant (P > 0.05).