Neutrophil elastase plays a non-redundant role in remodeling the venular basement membrane and neutrophil diapedesis post-ischemia/reperfusion injury

Abstract

Ischemia/reperfusion (I/R) injury is a severe inflammatory insult associated with numerous pathologies, such as myocardial infarction, stroke and acute kidney injury. I/R injury is characterized by a rapid influx of activated neutrophils secreting toxic free radical species and degrading enzymes that can irreversibly damage the tissue, thus impairing organ functions. Significant efforts have been invested in identifying therapeutic targets to suppress neutrophil recruitment and activation post-I/R injury. In this context, pharmacological targeting of neutrophil elastase (NE) has shown promising anti-inflammatory efficacy in a number of experimental and clinical settings of I/R injury and is considered a plausible clinical strategy for organ care. However, the mechanisms of action of NE, and hence its inhibitors, in this process are not fully understood. Here we conducted a comprehensive analysis of the impact of NE genetic deletion on neutrophil infiltration in four murine models of I/R injury as induced in the heart, kidneys, intestine and cremaster muscle. In all models, neutrophil migration into ischemic regions was significantly suppressed in NE^-/- mice as compared with wild-type controls. Analysis of inflamed cremaster muscle and mesenteric microvessels by intravital and confocal microscopy revealed a selective entrapment of neutrophils within venular walls, most notably at the level of the venular basement membrane (BM) following NE deletion/pharmacological blockade. This effect was associated with the suppression of NE-mediated remodeling of the low matrix protein expressing regions within the venular BM used by transmigrating neutrophils as exit portals. Furthermore, whilst NE deficiency led to reduced neutrophil activation and vascular leakage, levels of monocytes and prohealing M2 macrophages were reduced in tissues of NE^-/- mice subjected to I/R. Collectively our results identify a vital and non-redundant role for NE in supporting neutrophil breaching of the venular BM post-I/R injury but also suggest a protective role for NE in promoting tissue repair. © 2019 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.

Keywords: elastase; ischemia/reperfusion injury; neutrophil; venular basement membrane.

Figure 1

Neutrophil recruitment into the area at risk of NE‐deficient mice (NE^−/−) is impaired following MI and reperfusion injury. (A) Reconstructed tiling images of myocardial cryosections acquired by confocal microscopy from WT animals and NE^−/− mice and subjected to 25 min of MI followed by 2 h of reperfusion (I/R). Tissue sections from sham‐operated WT (left panels), I/R‐subjected WT (middle panels) and I/R‐subjected NE^−/− (right panels) animals were immunostained for collagen type IV (red). The yellow boxes exemplify the ischemic and reperfused regions of the left ventricle (area at risk). (B) Quantification of the collagen type IV mean intensity in the area at risk (yellow box from (A)). (C) Confocal images of the area at risk (delimited by the dotted line) from a WT animal subjected to myocardial I/R and immunostained for collagen type IV (red), PECAM‐1 (blue) and MRP‐14 (green) to visualize the ECM, blood vasculature and neutrophils, respectively. The image shows neutrophils infiltrating the region at risk only. (D) Magnified regions of the region of left ventricle subjected or not to I/R from WT (top panels) and NE^−/− (bottom panels) animals. Cryosections were immunostained for collagen type IV (red), PECAM‐1 (blue) and MRP‐14 (green) to visualize the ECM, blood vasculature and neutrophils, respectively. Left panels show the three channels together, whereas the right panels show the neutrophil channel only. The images exemplify the absence of neutrophil infiltration in the region at risk of the myocardium from NE^−/− mice subjected to I/R as compared with WT littermates. The top right panels (WT I/R) are magnified images of the same sample region as shown in (D). (E) Quantification of the number of neutrophils present in the area at risk. Data represent mean ± SEM from three or four mice per group. **p < 0.01 for the comparison between I/R and sham‐operated animals; #p < 0.05, ##p < 0.01 for the comparison between WT and NE^−/− mice as indicated by lines. Scale bars = 100 μm.

Figure 2

NE‐deficient mice are protected from renal I/R injury. (A) Histopathology of kidney sections from sham‐operated WT or WT and NE^−/− mouse subjected to renal I/R and stained with H&E. The bottom pictures are magnified regions (yellow box) demonstrating the modification of the architectural structure of tubules and glomeruli upon I/R injury in WT but not NE^−/− animals. (B) Histological score analysis of kidney sections. (C) Quantification of the number of PMN infiltrating the kidney. Data represent mean ± SEM from at least four animals per group. *p < 0.05 and ****p < 0.0001 for the comparison between I/R and sham‐operated animals; ####p < 0.0001 for the comparison between WT and NE^−/− mice as indicated by lines. Scale bar = 100 μm.

Figure 3

Effect of NE deficiency on leukocyte transmigration responses in post‐capillary venules in vivo following I/R injury. Leukocytes' firm adhesion and transmigration in post‐capillary venules of mouse cremaster muscles or mesentery in response to I/R were investigated using IVM. (A) The mouse cremaster muscle from WT and NE^−/− animals was surgically exteriorized, superfused with Tyrode's solution and basal leukocyte responses were quantified for 20 min prior to the induction of ischemia using a clamp, as detailed in Materials and methods. Thirty minutes later, the clamp on the exteriorized cremaster muscle was removed to induce reperfusion of the vessels. Leukocyte adhesion (left panel) and extravasation (right panel) responses were quantified at regular intervals for 120 min post‐reperfusion. For each genotype, a sham‐operated group was also analyzed (exteriorized cremaster tissues without the clamp). (B) WT and NE^−/− mice were subjected to occlusion of the superior mesenteric artery for 35 min, followed by 90 min reperfusion. For each genotype, a sham‐treatment group was also analyzed, where laparotomy was conducted without the occlusion of the mesenteric arteries. Both leukocyte adhesion (left panel) and transmigration (right panel) responses were quantified in post‐capillary venules at 90 min post‐reperfusion. (C) Bone marrow neutrophils were isolated from WT or NE^−/− mice, fluorescently labeled and injected i.v. into WT or NE−/− recipient mice prior to the induction of I/R of the cremaster muscle, as detailed in Materials and methods. At the end of the experiment, tissues where harvested, fixed and immunostained for neutrophils (MRP‐14). Data show the quantification of fluorescent donor neutrophils present in the tissue and normalized to the number of blood circulating donor cells. (D) The images are representative confocal pictures of post‐capillary venules of the cremaster muscles of WT (left panel) and NE^−/− (right panel) mice subjected to 30 min ischemia followed by 120 min reperfusion of the cremaster muscles. The images show that whereas in WT neutrophils access the interstitial tissue, NE^−/− cells are trapped within the venular BM (arrows). (E) Quantification of the number of neutrophils present within the venular BM 2 h post‐reperfusion of the cremaster muscles. (F) Quantification of the number of neutrophils present within the venular BM following 1.5 h post‐reperfusion of the mesentery. Figures are representative of four to seven animals per group. Mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 for the comparison between I/R and sham‐operated animals (or between WT and NE^−/− donor cells for the adoptive transfer experiment); #p < 0.05, ##p < 0.01, ###p < 0.001 for the comparison between WT and NE^−/− mice as indicated by lines. Bars = 20 μm.

Figure 4

NE is mobilized during the migration of neutrophils through the venular BM. The cremaster muscles and mesentery of WT and NE^−/− mice were subjected to temporary ischemia followed by reperfusion. At the end of the reperfusion period, tissues were collected, fixed and immunostained prior to visualization of the samples using confocal microscopy. (A) Representative image of a post‐capillary venule (CD31, red) of WT and NE^−/− animals subjected to I/R injury and demonstrating the association of the enzymatic activity of NE using the NE680FAST fluorescent substrate (blue) with neutrophils (green) present in the abluminal aspect (i.e. BM) of the vessel wall (plain arrows) and, to a lesser extent, interstitial neutrophils (dotted arrows). (B) Representative confocal image of a post‐capillary venule (2 μm longitudinal cross‐section from the middle of the vessel, green) from a WT mouse subjected to I/R showing NE expression (red) in the cytoplasmic compartments (blue) of neutrophils. The bottom panels are enlargements of the yellow boxed region showing neutrophils at different stages of their migration route, i.e. luminal (i), within the vascular BM (ii) or within the interstitial tissue (iii). A 5% opacity filter on the MRP‐14 channel was applied to highlight NE expression in neutrophils on the right panel. (C) Quantification of the mean fluorescence intensity of NE expression in neutrophils at the three stages of their migration (as expressed as a percentage change over the intensity of NE from luminal neutrophils). (D) Quantification of the fluorescence intensity of NE expression within the venular BM from WT and NE^−/− animals at 1 h post‐reperfusion. Figures are representative of four to seven animals per group. Mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, for the comparison of NE intensity (mean fluorescence intensity) between BM/interstitial neutrophils and luminal cells (C) or between I/R and sham‐operated animals (D); #p < 0.05 for the comparison between WT and NE^−/− mice as indicated by the line. Bars = 10 μm.

Figure 5

Venular BM is remodeled during I/R injury and transmigrated neutrophils are positive for laminin. WT and NE^−/− mice were subjected to I/R of their cremaster muscles or mesentery as described in Materials and methods. At the end of the reperfusion period, tissues were collected, fixed and whole‐mount immunostained for neutrophils (MRP‐14) and BM (laminin‐α5 or pan‐laminin) prior to visualization of the samples by confocal microscopy. (A) Representative images showing the presence of LERs (circles) within the BM (laminin‐α5) of post‐capillary venules of mouse cremaster muscles. (B) Quantification of the size of the LER within the BM (laminin‐α5) of the cremaster post‐capillary venules. (C) Representative images showing the presence and remodeling of LERs (circles) within the BM (laminin‐α5) of post‐capillary venules of mouse mesentery. (D) Quantification of the size of the LER within the BM (laminin‐α5) of the mesenteric post‐capillary venules. (E) Representative confocal image of a post‐capillary venule (from a WT mouse subjected to I/R) showing the presence of laminin‐positive neutrophils within the interstitial tissue (arrows). The bottom picture shows the staining of laminin only. (F) Quantification of the number of neutrophils present within the venular BM 1.5 h post‐reperfusion of the mesentery. Figures are representative of four to seven animals per group. Mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, for the comparison between sham and I/R groups; #p < 0.05, ###p < 0.001 for the comparison between WT and NE^−/− mice as indicated by the line. Bars = 10 μm.

Figure 6

NE genetic deficiency inhibits neutrophil activation and monocyte/macrophage recruitment. The cremaster muscles of WT and NE^−/− mice were subjected to temporary ischemia followed by reperfusion; leukocyte phenotype and vascular leakage were quantified by flow cytometry and Evan's Blue assay, respectively. (A) Quantification of the total number of neutrophils in the cremaster muscles (i.e. interstitial and vascular neutrophils included). (B) Quantification of ROS generation by isolated neutrophils from blood or cremaster muscles as measured by DHE mean fluorescence intensity (MFI). (C) Quantification of the vascular leakage into the tissue post‐reperfusion (4 h). (D) Quantification of the total number of monocytes in the cremaster muscles at 20 h post‐reperfusion. (E) Quantification of the total number of M2 (CD206+) macrophages in the cremaster muscles at 20 h post‐reperfusion. Data represent mean ± SEM from four to six mice per group (from three independent experiments). *p < 0.05, **p < 0.01, for the comparison between sham and I/R groups; #p < 0.05, ##p < 0.01 for the comparison between WT and NE^−/− mice as indicated by the line.