Mast cells mediate early neutrophil recruitment and exhibit anti‐inflammatory properties via the formyl peptide receptor 2/lipoxin A₄ receptor

Abstract

Background and Purpose

In recent years, studies have focused on the resolution of inflammation, which can be achieved by endogenous anti‐inflammatory agonists such as Annexin A1 (AnxA1). Here, we investigated the effects of mast cells (MCs) on early LPS‐induced neutrophil recruitment and the involvement of the AnxA1‐formyl peptide receptor 2/ALX (FPR2/ALX or lipoxin A₄ receptor) pathway.

Experimental Approach

Intravital microscopy (IVM) was used to visualize and quantify the effects of LPS (10 μg per mouse i.p.) on murine mesenteric cellular interactions. Furthermore, the role that MCs play in these inflammatory responses was determined in vivo and in vitro, and effects of AnxA1 mimetic peptide Ac2‐26 were assessed.

Key Results

LPS increased both neutrophil endothelial cell interactions within the mesenteric microcirculation and MC activation (determined by IVM and ruthenium red dye uptake), which in turn lead to the early stages of neutrophil recruitment. MC recruitment of neutrophils could be blocked by preventing the pro‐inflammatory activation (using cromolyn sodium) or enhancing an anti‐inflammatory phenotype (using Ac2‐26) in MCs. Furthermore, MCs induced neutrophil migration in vitro, and MC stabilization enhanced the release of AnxA1 from neutrophils. Pharmacological approaches (such as the administration of FPR pan‐antagonist Boc2, or the FPR2/ALX antagonist WRW4) revealed neutrophil FPR2/ALX to be important in this process.

Conclusions and Implications

Data presented here provide evidence for a role of MCs, which are ideally positioned in close proximity to the vasculature, to act as sentinel cells in neutrophil extravasation and resolution of inflammation via the AnxA1‐FPR2/ALX pathway.

Abbreviations

Ac2‐26

annexin A1 mimetic peptide

AnxA1

annexin A1

ATL

15‐epimer‐lipoxin A₄

Bk

background fluorescence

Boc2

N‐tert‐butoxycarbonyl‐L‐Phe‐D‐Leu‐L‐Phe‐D‐Leu‐L‐Phe

CMP 48/80

compound 48/80

FL_in

fluorescence intensity inside the vessel

FL_out

fluorescence intensity outside the vessel

fMLP

formyl‐Met‐Leu‐Phe

FPR

formyl peptide receptor

IVM

intravital microscopy

LXA₄

lipoxin A4

MC

mast cell

MCET

mast cell extracellular trap

PPE

plasma protein extravasation

SAA

serum amyloid A

TLR

toll‐like receptor

WKYMVm

Trp‐Lys‐Tyr‐Val‐D‐Met

Introduction

The innate immune system is the first line of defence against invading pathogens such as fungi or bacteria. Within the heterogenic cell populations of the innate immune system, neutrophils are the most abundant leukocyte population (50–70%) in human blood and, as such, are well recognized as a major player during acute inflammation, being quickly directed to sites of infection/injury (Nathan, 2006; Ley et al., 2007). Upon activation, neutrophils engulf bacteria or release a variety of factors, including ROS, and facilitate the formation of neutrophil extracellular traps (NETs) to capture and destroy pathogens (Segal, 2005; Serhan et al., 2008; Wright et al., 2010; Yipp and Kubes, 2013). Since neutrophil activation can mediate tissue injury and perpetuate the inflammatory response, their physiological effector functions are tightly mediated and regulated via cell‐surface receptors [e.g. toll‐like receptor family (TLR)] (Segal, 2005; Chiang et al., 2006).

Resolution of inflammation normally occurs when the invading pathogen has been neutralized and before the immune response becomes pathological. However, failure to resolve inflammation can lead to excessive or deregulated neutrophil responses, which (together with inadequate repair) contribute to the persisting tissue damage that underlies many infectious, ischaemic or inflammatory diseases, for example, stroke, sepsis and myocardial infarction (Nathan, 2006; Borregaard, 2010; Goldenberg et al., 2011). It is now appreciated that the host response to inflammatory insults involves tightly controlled and active (rather than passive) resolution programs (Chiang et al., 2006; Serhan et al., 2008), in which efficient resolution depends on inhibition of neutrophil influx, rapid clearance of infiltrating neutrophils and regeneration of disrupted tissue structures (Serhan et al., 2007).

Neutrophils undergo a sequential pattern of interaction with vascular endothelial cells, orchestrated by a well‐characterized sequence of rolling, adhesion and emigration into inflamed/infected tissue (Hughes et al., 2013). Mast cells (MCs) are an important source of many pro‐inflammatory mediators, including those that activate the expression and/or activation of adhesion molecules involved in the leukocyte recruitment cascade (Kubes and Granger, 1996, Yazid et al., 2010a,b; Yao et al., 2014). In particular, the close proximity of MC to microcirculatory vessels puts them in a prime position to activate and direct circulating neutrophils. In addition, MCs are rather promiscuous and also appear to play a protective role in regulation of innate immune responses. For example, MCs have been shown to aid the resolution of inflammation by mobilizing neutrophils to the site of infection and by producing endogenous anti‐inflammatory and pro‐resolving mediators such as annexin A1 (AnxA1) (Da Silva et al., 2011).

The 37 kDa protein AnxA1 (and its mimetic peptide Ac2‐26) is a potent inhibitor of leukocyte infiltration. It regulates leukocyte detachment from the postcapillary endothelium in both acute and chronic inflammatory states (Gavins et al., 2012 ), and deletion of the AnxA1 gene is associated with exacerbated inflammatory responses (Hannon et al., 2003). AnxA1 is widely distributed, being detected in, for example, the lung, kidney, bone marrow, intestine, spleen, thymus or brain (Fava et al., 1989; Gavins et al., 2012; Vital et al., 2016), and secreted by a variety of cell types including MCs (Kwon et al., 2012) and neutrophils (Oliani et al., 2001). The importance of AnxA1 in limiting inflammatory responses is also highlighted by findings that silencing AnxA1 (Kao et al., 2014) or using AnxA1 knockout mice (Gavins et al., 2003) exacerbates inflammation. Several groups have shown that AnxA1 exerts many of its anti‐inflammatory and pro‐resolving actions through a GPCR: formyl peptide receptor type 2/lipoxin A₄ receptor (FPR2/ALX, termed Fpr2/3 in the mouse) (Serhan et al., 2008; Perretti and D'Acquisto, 2009; Ye et al., 2009; Dufton et al., 2010; Bena et al., 2012; Cooray et al., 2013; Vital et al., 2016), which is expressed on endothelial cells and cells of myeloid lineages (Chiang et al., 2006). It is important to note that AnxA1, or a proteolytic fragment, also exerts anti‐inflammatory and pro‐resolving actions through FPR1 (Leoni et al., 2013) or possibly a heterodimer (Cooray et al., 2013). In addition, unlike the full‐length protein, peptides derived from the AnxA1 N‐terminal region have been shown to activate all three receptors of the FPR family (Perretti and D'Acquisto, 2009).

Previously, we have shown that AnxA1 plays a significant role in limiting inflammatory responses (Hughes et al., 2013, Vital et al., 2016) and that stimulation of phosphorylation and subsequent release of AnxA1 is an important component of the inhibitory actions of MC‐stabilizing drugs (Yazid et al., 2012; Yazid et al., 2013). However, the effect of MCs on early neutrophil recruitment and the role of the AnxA1‐FPR2/ALX pathway have not been extensively investigated. We used intravital microscopy (IVM), coupled with pharmacological approaches to investigate the interaction of MCs and the AnxA1‐FPR2/ALX pathway in early neutrophil recruitment.

Methods

Animals

Male mice aged 5–8 weeks were housed in controlled‐temperature/humidity environment (22 ± 1°C, 60–70% relative humidity) in individual cages (five mice per cage, with wood shaving bedding and nesting material), with a 12 h light/dark cycle (lights on at 0700 h), and had access to a standard chow pellet diet and tap water ad libitum. C57BL/6 mice were purchased from either Jackson Laboratory (Bar Harbor, ME, USA) or Charles River Limited (Margate, Kent, UK). Mice were allowed to acclimatize to their housing environment for at least 5 days prior to experimentation and to the experimental room for 1 h before experiments. Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny et al., 2010; McGrath and Lilley, 2015). All animal experiments followed the European Union Directive (2010/63/EU) or were approved by the Louisiana State University Health Sciences Centre Shreveport Institutional Animal Care and Use Committee and were in accordance with the guidelines of the American Physiological Society. All studies were performed blinded and randomized, with a key system to identify which animal/sample had undergone which treatment. Furthermore, compounds administered were made by laboratory personnel other than the one performing the experiment. Animals were deeply anaesthetized, as determined by the absence of a pedal reflex, with ketamine (Ketaset, 150 mg·kg⁻¹, Fort Dodge Animal Health, Southampton, UK) and xylazine (Rompun, 7.5 mg·kg⁻¹, Bayer Healthcare, Newbury, UK) i.p. before being killed by cervical dislocation.

Drugs

The following drugs were used at the most effective doses/concentrations based on our previous findings and dissolved in sterile saline unless otherwise stated: zymosan A (1 mg in 0.5 mL PBS; Sigma‐Aldrich, Poole, Dorset, UK) (Chatterjee et al., 2005); ammonium thioglycollate (thioglycollate, Sigma, UK) (Henderson et al., 2001); cromolyn sodium (Sigma, UK) (Yazid et al., 2013); AnxA1 mimetic peptide Ac2‐26 (AnxA1_Ac2‐26, Ac‐AMVSEFLKQAWFIENEEQEYVQTVK, Cambridge Research Biochemicals, Cleveland, UK; dissolved in PBS) (Smith et al., 2015); compound 48/80 (CMP 48/80, Sigma, UK) (Sinniah et al., 2016); Boc2 (MPBiomedicals, Cambridge, UK. Dissolved in 0.02% DMSO and sterile saline) (Vital et al., 2016); WRW4 (Trp‐Arg‐Trp‐Trp‐Trp‐NH₂, Tocris, Bristol, UK) (Smith et al., 2015); WKYMVm peptide (Trp‐Lys‐Tyr‐Val‐D‐Met, Tocris, Bristol, UK) (Heo et al., 2014); 15‐epimer‐lipoxin A₄ (ATL, 15‐epi‐LXA₄; Cayman Chemicals, Ann Arbour, Michigan, USA) (Vital et al., 2016). fMet‐Leu‐Phe (fMLP, Sigma, UK) (Smith et al., 2015); serum amyloid A (SAA; Peprotech, Rocky Hills, NJ, USA) (Dufton et al., 2010). FITC‐labelled albumin (0.25 mg·g⁻¹ body weight, Sigma, UK; Gavins et al., 2003) was administered i.v. to enable visualization of plasma leakage.

Induction of experimental endotoxaemia

Endotoxaemia was induced by i.p. injection of LPS (an endotoxin of Escherichia coli serotype 0111:B4, 10 μg per mouse, Sigma, UK) whilst controls received a corresponding volume (50 μL, i.p.) of saline vehicle. We chose to use a low dose of LPS within the study in order to cause activation of the vasculature but not provoke decreased microvascular perfusion or mortality, which occurs with higher LPS doses and would complicate interpretation of the inflammatory response (Zanetti et al., 1992; Damazo et al., 2005). To ascertain the time course of the responses to LPS, mesenteric leukocyte activity was visualized by IVM 0, 10, 20, 60 or 120 min after injection. In addition, serum pro‐inflammatory cytokines TNF‐α, IL‐1ß and IL‐6 were also determined 0, 10, 20, 60 or 120 min after injection by elisa as per the manufacturer's instructions (R&D Systems, Minneapolis, MN, USA). To ascertain the specificity of the response to LPS, zymosan (1 mg in 0.5 mL PBS, i.p.) and thioglycollate (250 μL of a 3% solution, i.p.) were injected in place of LPS. In pretreatment experiments, mice received sodium cromolyn (10 mg·kg⁻¹, i.p.), Ac2‐26 (100 μg per mouse, i.v.) or saline vehicle 15 min prior to LPS injection.

Visualization of the mesentery by IVM

Anaesthesia was induced with an i.p. injection of ketamine (Ketaset, 150 mg·kg⁻¹, Fort Dodge Animal Health) and xylazine (Rompun, 7.5 mg·kg⁻¹, Bayer Healthcare) and topped up as required (upon presence of a pedal reflex). The jugular vein was exposed and cannulated with polyethylene tubing (PE10) for drug administration. A midline incision was made, and the mesenteric vascular bed was exteriorized, gently laid across a Plexiglass viewing stage, and mounted on an Olympus BW61WI microscope with a water‐immersion objective lens (magnification of 40×, LUMPlan, FI/IR, Olympus, Tokyo, Japan). The tissue preparation was transilluminated with a 12 V, 100 W halogen light source. The mesenteric vessels were superfused with bicarbonate‐buffered solution (g·L⁻¹: NaCl, 7.71; KCl, 0.25; MgSO₄, 0.14; NaHCO₃, 1.51 and CaCl₂, 0.22, pH 7.4), and 1–5 randomly selected post‐capillary venules (20–40 μm wide and 100 μm long) were observed per mouse. Leukocyte rolling velocity (white blood cell velocoty) was measured and calculated in μm·s⁻¹. Leukocyte adhesion was measured by counting clearly visible cells that remained stationary on the vessel wall for at least 30 s, within the 100 μm stretch. Leukocyte emigration from the microcirculation into the tissue was quantified by counting the number of cells 50 μm outside the vessel wall on either side of the 100 μm stretch. Real‐time videos of the vessels were made with a black‐and‐white camera (model CoolSNAP HQ², Photometrics, Tucson, AZ, USA) coupled to a Windows XP‐based computer for recording by Slidebook 4.2 (Intelligent Imaging Innovations, Inc., Denver, CO, USA).

Tissue processing and staining

Mice were anaesthetised using ketamine (150 mg·kg⁻¹) and xylazine (7.5 mg·kg⁻¹) i.p. for transcardial perfusion with 10 mL PBS followed by 10 mL formaldehyde. Mice were killed by cervical dislocation. All efforts were made to minimize animal suffering. Tissues were fixed in formaldehyde solution overnight before routine processing to paraffin wax. Sections of 4 μm were cut on a microtome and stained with haematoxylin and eosin.

Plasma protein extravasation

FITC‐labelled albumin was injected i.v. (0.25 mg·g⁻¹ body weight) 5–10 min before the end of the experiment. A snapshot of vessel fluorescence was taken using block filter (excitation at 450–490 nm, and emission at 535–620 nm). Albumin leakage was quantified by measuring mean fluorescence intensity using ImageJ64 (National Institute of Health, USA). Average fluorescence intensity in three areas of equal size was measured: inside the vessel (Fl_in), outside the vessel (Fl_out) and background fluorescence (bk) in an area with no obvious leakage. Albumin leakage was then determined as follows: [(Fl_out × bk)/(Fl_in × bk)] × 100%.

MC visualization and activation

Mice were treated with LPS or sterile saline vehicle and after 20 min, the mesentery was exteriorised as described above and superfused with bicarbonate buffered saline containing 0.001% ruthenium red, which is selectively taken up by activated MCs (Da Silva et al., 2011). Images of the mesenteric vascular bed were recorded under a ×10 objective every 5 min over a 15 min period, maintaining the same field of view. After that time, the MC degranulating agent CMP 48/80 (1 μg·mL⁻¹) was added to the superfusion buffer to cause maximum activation of the MCs, and a series of final snapshots was taken after 1, 5 and 10 min. MC activation was analysed using ImageJ64, where images were converted to black and white and single MCs were selected for analysis. The freehand tool was used to measure the greyscale intensity at each time point, expressed as a percentage of the maximum (taken as the final snapshot 10 min after CMP48/80 superfusion). Three to five MCs were analysed per field of view and calculated as the mean.

MC isolation

Peritoneal leukocytes (2 × 10⁸ cells mL⁻¹) were fractionated on a preformed continuous gradient generated from 70% isotonic Percoll in 0.15 M NaCl (Amersham Pharmacia Biotech, NJ, USA). The gradient was calibrated between 1.018 and 1.138 g·mL⁻¹ with density marker beads (Amersham Biosciences). MCs were collected at band density of 1.088 g·mL⁻¹, and their purity was >95% positive when tested by flow cytometry using CD117–FITC mAb (BD Biosciences, Oxford, England).

MC stimulation

Purified peritoneal MC at 0.5 × 10⁵ cells mL⁻¹ was stimulated in 24‐well plates (VWR, Leicestershire, England) with saline or LPS (1 μg·mL⁻¹) for 4 h or LPS (1 μg·mL⁻¹) and CMP 48/80 (10 μg·mL⁻¹) for 20 min at 37°C with 5% CO₂ atmosphere, with or without cromolyn sodium. The cell‐culture supernatants were either removed, spun and analysed by elisa for the presence of histamine (SPI bio, Strasbourg, France) or AnxA1 (MyBiosource Inc, San Diego, CA, USA), or were added to chemotaxis plates.

Neutrophil isolation

Mouse bone marrow cells were harvested by flushing marrow from femurs and tibias with RPMI‐1640 medium. Neutrophils were isolated from bone marrow cells by density centrifugation with Histopaque‐1077 and Histopaque‐1119. The purity of isolated neutrophils was routinely >95% as assessed by light microscopic analysis of the cells stained with Diff‐Quick (Wako Pure Chemical Industries, Osaka, Japan) and >98% viable as assessed by a trypan blue exclusion test.

Neutrophil labelling

Isolated neutrophils were stained with Calcein AM (Sigma, UK; 1 μL of Calcein per 4 million cells) for 30 min at 37°C with 5% CO₂ atmosphere. The cells were then washed and used in the chemotaxis assay.

Neutrophil chemotaxis assay

MCs were degranulated as described above, and the supernatant was collected. Neutrophils (4 × 10⁶ cells mL⁻¹) were added to the upper chamber of a Neuroprobe 96‐well disposable chemotaxis plate (Neuro Probe Inc. MD, USA, 5 μm pore size) in RPMI containing 10% FBS, 2 mmol·L⁻¹ L⁻¹‐glutamine, 100 U·mL⁻¹ penicillin and 100 mg·mL⁻¹ streptomycin. MC supernatant was added to the lower chamber. Plates were incubated for 1.5 h (37.5°C, 5% CO₂). The number of migrated cells was assessed using a spectrofluorometer (excitation 485 nm and emission 528 nm). In some cases, calcein‐AM‐labelled neutrophils were pre‐incubated (10 min) with Boc2 (20 μM) or WRW4 (10 μM) before being incubated with the MC supernatant described above. This neutrophil‐MC supernatant cocktail was added to the upper chamber of a Neuroprobe chemotaxis plate and vehicle or the chemoattractant fMLP (10 nM) added to the lower wells. Plates were incubated for 1.5 h (37.5°C, 5% CO₂) and read as above.

Myeloperoxidase (MPO) activity

Mesenteric tissue samples were collected from mice at 0, 20, 60 or 120 min post‐LPS or saline vehicle treatment. MPO was measured as a marker for mesenteric neutrophil infiltration. Mesenteric tissue homogenates and MPO standards (Sigma, UK) were placed onto a 96‐well plate, and 200 μL of o‐dianisidine (Sigma, UK) solution and 10 μL of 0.1% H₂O₂ (Sigma, UK) were added. The absorbance was read after 5 min at 405 nm and expressed as units mg⁻¹ of wet tissue (Gavins et al., 2007).

Histamine release

Histamine release was measured using an elisa (SPI Bio). The assay was conducted following the manufacturer's protocols. A standard curve ranging from 0.39 to 50 nM histamine was prepared using the reagent provided, and the optical density was then read within 60 min in a microplate reader (Titertek™, Vienna, Austria) at 405 nm.

AnxA1 elisa

AnxA1 release was measured using an elisa (MyBiosource). The assay was conducted following the manufacturer's protocols. A standard curve ranging from 0.625 to 40 ng·mL⁻¹ was prepared using the reagent provided, and the optical density was then read within 60 min in a microplate reader (Titertek™) at 450 nm.

Data analysis and statistical procedures

All data were analysed using GraphPad Prism 6. Data are expressed as mean ± SEM with n values given in the respective figure legends. When determining statistical significance between two groups, an unpaired t‐test was carried out and, where appropriate, corrected for multiple comparisons using the Holm–Šídák method. Multiple groups were analysed by one‐way ANOVA or non‐parametric Kruskal–Wallis test, and post hoc comparisons were performed by Bonferroni or Dunn's multiple comparison test respectively. Differences were considered statistically significant if P < 0.05. The figures have been graphically presented on a range‐specific axis. The data and statistical analysis comply with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2015).

Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Southan et al., 2016), and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 (Alexander et al., 2015a,b).

Results

LPS induces early leukocyte‐endothelial cell interactions in the murine mesentery

LPS induced heightened cellular trafficking within the mesenteric microcirculation as early as 20 min when determined by IVM. These findings are displayed in the representative image of the mesenteric vascular bed taken 20 min post‐LPS challenge (Figure 1A). Whilst no increase in leukocyte adhesion at that early time point was observed (Figure 1C), the number of emigrated leukocytes had already increased significantly (saline = 2.4 ± 0.3 cells vs. LPS = 5.8 ± 0.6 cells; Figure 1D). Both parameters (Figure 1C, D), as well as leukocyte rolling velocity (Figure 1B) and wall shear rate (Table 1), were significantly (P < 0.05) higher in LPS‐treated mice at 60 and 120 min.

Figure 1.

Effects of LPS on the mesenteric microcirculation. Mice were treated with LPS (10 μg per mouse, i.p.) or saline vehicle. At 0, 20, 60 or 120 min post‐LPS, the mesentery was exteriorised for visualization of post‐capillary venules by IVM. (A) Image of mesenteric microcirculation taken 20 min post‐LPS treatment. Arrows with short tails = rolling leukocytes; arrowheads = adherent leukocytes; and arrows with long tails = emigrated leukocytes. Scale bar = 20 μm. Leukocyte–endothelial cell interactions were quantified in terms of (B) leukocyte rolling velocity [expressed as white blood cell velocity (V_WBC)], (C) number of adherent (stationary for ≥30 s) leukocytes 100 μm⁻¹ length and (D) number of emigrated leukocytes 100 × 50 μm². FITC‐conjugated albumin was injected i.v. (0.25 mg·g⁻¹ body weight) and allowed to circulate for 5–10 min to quantifiy albumin leakage as a measure of PPE. (E) Image of a representative mesenteric vessel containing FITC‐conjugated albumin. Average fluorescence intensity in three areas of equal size was measured: Fl_in, Fl_out and Bk in an area with no obvious leakage. Scale bar = 20 μm. (F) Albumin leakage at 0, 20, 60 or 120 min post‐LPS was determined as follows: [(Fl_out × bk)/(Flⁱⁿ × bk)] × 100%. Data are expressed as mean ± SEM. n = 6 mice per group. ^* P < 0.05 versus saline vehicle‐treated counterpart.

Table 1.

Haemodynamic parameters in the mesenteric microcirculation of WT mice

Genotype	Treatment	Time post‐LPS (min)	Vessel diameter (μm)	Wall shear rate (s−1)
	Saline	0	28.4 ± 2.2	546 ± 12.6
		20	25.2 ± 3.1	529 ± 11.4
		60	28.8 ± 2.1	556.8 ± 9.5
		120	30.2 ± 1.9	558 ± 9.5
	LPS	0	26.8 ± 2.9	531 ± 16.3
		20	30.1 ± 2.1	532 ± 13.6
		60	26.8 ± 2.1	450 ± 31.4*
		120	28.2 ± 1.4	334 ± 34.0*

LPS dose administered = 10 μg per mouse. Data are mean ± SEM. n = 5 mice per group.

^* P < 0.05 versus saline at corresponding time point.

LPS induces early mesenteric but not systemic inflammatory responses

Another cardinal sign of an inflammatory response, plasma protein extravasation (PPE) measured by FITC‐albumin leakage using IVM, was found to follow the same pattern as leukocyte emigration: LPS induced an acute increase at 20 min post treatment, which remained significantly elevated at 60 and 120 min (Figure 1E, F. P < 0.05). Despite the initial acute local inflammatory responses in the mesentery, no early systemic effects were recorded when the pro‐inflammatory cytokines TNF‐α, IL‐1β and IL‐6 were measured in serum of saline‐ or LPS‐treated mice (Table 2). All three cytokines were found to be significantly (P < 0.05) raised at 60 and 120 min following LPS injection.

Table 2.

Altered levels of pivotal pro‐inflammatory cytokines in serum post endotoxin challenge

Time post‐treatment (min)	TNF‐α (pg·mL−1)		IL‐1β (pg·mL−1)		IL‐6 (pg·mL−1)
	Saline	LPS	Saline	LPS	Saline	LPS
0	0	0	0	0	0	0
20	0	0	0	0	0	0
60	43.71 ± 4.4	304.6 ± 29.2*	91.9 ± 9.3	321.7 ± 29.6*	60.7 ± 11.3	157.1 ± 14.8*
120	88.7 ± 4.8	663.3 ± 35.4*	84.3 ± 4.8	702.7 ± 33.9*	55.9 ± 6.6	513.8 ± 49.7*

Data are mean ± SEM. n = 5 mice per group.

^* P < 0.05 versus saline at corresponding time point. A value of 0 indicates levels below the detectable limit of the kit (31.5 pg·mL⁻¹).

The early acute inflammatory response is not specific to LPS

Having established that an early cellular recruitment (i.e. emigration of leukocytes) occurs in the murine mesenteric microvasculature in response to LPS, we next tested whether this phenomenon was pathogen‐specific. We used two additional inflammagens: zymosan, a glucan derived from yeast acting as a ligand for TLR2 (Volman et al., 2005), and thioglycollate, a non‐specific leukocyte chemotactic agent (Vemula et al., 2010). When comparing rolling velocity, we found both zymosan and thioglycollate to induce a significant increase when compared with saline control as well as to LPS (Figure 2A) whilst only thioglycollate was found to increase leukocyte adherence at 20 min (Figure 2B). In regard to leukocyte emigration, zymosan followed the established trend for LPS and caused increased extravasation, whilst thioglycollate did not differ from saline control (Figure 2C). Finally, both inflammagens mirrored the effects observed with LPS on PPE (Figure 2D). Although the experiments showed that the acute inflammatory mesenteric response was not specific to LPS, we continued to use LPS throughout the rest of the study based on the superior inflammatory profile in regard to endothelial‐leukocyte interactions.

Figure 2.

Rapid effects of different inflammagens on mesenteric leukocyte–endothelial cell interactions. Mice were treated with LPS (10 μg per mouse, i.p.), zymosan (1 mg in 0.5 mL PBS i.p.), ammonium thioglycollate (250 μL of a 3% solution, i.p.) or saline vehicle. After 20 min, the mesentery was exteriorized under anaesthesia for visualization of post‐capillary venules by IVM. Leukocyte–endothelial cell interactions were quantified in terms of (A) leukocyte rolling velocity [expressed as white blood cell velocity (V_WBC)], (B) number of adherent (stationary for ≥30 s) leukocytes 100 μm⁻¹ length and (C) number of emigrated leukocytes 100 × 50 μm⁻². (D) FITC‐conjugated albumin was injected i.v. (0.25 mg·g⁻¹ body weight) and allowed to circulate for 5–10 min to quantifiy albumin leakage {[(Fl_out × bk)/(Flⁱⁿ × bk)] × 100%)} as a measure of PPE. Data are mean ± SEM. n = 5 mice per group. ^* P < 0.05 versus saline vehicle‐treated counterpart and ^# P < 0.05 versus LPS‐treated counterpart.

Neutrophils are the predominant cell type in the inflamed murine mesentery

Since De Filippo et al. (2008) previously described the inflammatory response to LPS in the murine mesentery to be mainly mediated by neutrophils, we aimed to confirm these results in our model. Analysis of histological slides showed the above‐mentioned leukocyte population to consist predominantly of neutrophils (Figure 3A). Having this established, we used a second neutrophil assay, MPO, as a marker of neutrophil infiltration and found a significant increase within the mesentery as early as 20 min after LPS treatment (Figure 3B). These results first confirm the observations made by IVM of an early LPS‐induced leukocyte emigration in the murine mesentery and, second, identifies these cells to be predominantly neutrophils.

Figure 3.

Neutrophils are the predominant cell type in the inflamed murine mesentery. Mice were treated with LPS (10 μg per mouse, i.p.) or saline vehicle for 0, 20, 60 or 120 min. (A) Haematoxylin and eosin staining of the mesentery shows adherent neutrophils (aN) and emigrated neutrophils (eN) following LPS stimulation. (B) MPO activity was measured in the mesenteric tissue samples. Data are mean ± SEM. n = 5 mice per vehicle group and n = 6 mice per LPS group. ^* P < 0.05 versus saline vehicle‐treated counterpart.

LPS causes an early MC activation in the mesentery

As MCs lie in close proximity to vessel walls, we hypothesized that the early LPS‐induced inflammatory response in the mesenteric microvasculature could be at least in part driven by LPS‐induced MC activation. Therefore, we tested the influence of LPS on early MC activation in the murine mesentery, and as such, MC activation was evaluated by measuring the uptake of ruthenium red dye (expressed at each time point as % of the maximum uptake after complete stimulation with CMP 48/80), which is selectively and quantitatively taken up by activated MCs. We found a significantly increased uptake of dye in the LPS group, observable within 5 min of treatment (Figure 4A). To further determine the degree of early MC activation following LPS stimulation, we used the established MC activator CMP 48/80 to test if this would result in an additional MC activation above LPS‐induced levels. Importantly, in saline‐treated mice, MCs were rapidly activated upon addition of CMP 48/80, whilst those from LPS‐treated mice did not further respond to that stimulus suggesting a near‐complete early MC activation following LPS treatment (Figure 4A).

Figure 4.

Mesenteric MC activation and effects on leukocyte–endothelial cell interactions. (A) Mice were treated with LPS (10 μg per mouse, i.p.) or saline vehicle. After 20 min, the mesentery was exteriorized and superfused with bicarbonate‐buffered saline containing 0.001% ruthenium red, which is selectively taken up by activated MCs. Images of the mesentery were recorded under a ×10 objective for 15 min, after which time the MC destabilizing agent CMP 48/80 (1 μg·mL⁻¹) was added to the superfusion buffer. Dye uptake expressed at each time point as % of the maximum intensity reached after complete stimulation with CMP 48/80, 25 min post exteriorization or after 10 min superfusion with CMP 48/80. (B, C, D) Mice were treated with vehicle (saline i.p.), sodium cromolyn (CRO, 10 mg·kg⁻¹ i.p.) or Ac2‐26 (100 μg per mouse i.v.), 15 min prior to injection of LPS (10 μg per mouse i.p.). After 20 min, the mesentery was exteriorized for visualization of post‐capillary venules by IVM. Leukocyte–endothelial cell interactions were quantified in terms of (B) number of adherent (stationary for ≥30 s) leukocytes 100 μm⁻¹ length, (C) number of emigrated leukocytes 100 × 50 μm⁻² and (D) FITC‐conjugated albumin was injected i.v. (0.25 mg·g⁻¹ body weight) and allowed to circulate for 5–10 min to quantify albumin leakage {[(Fl_out × bk)/(Flⁱⁿ × bk)] × 100%)} as a measure of PPE. Data are expressed as mean ± SEM. n = 6 mice per vehicle group and n = 5 mice per LPS group. ^* P < 0.05 versus corresponding saline vehicle‐treated counterpart.

MC activation is involved in the early inflammatory response in the mesenteric microvasculature

Having established MC to be activated early by LPS, we next tested the hypothesis that MC activation was dynamically involved in the acute inflammatory responses (neutrophil adherence and emigration as well as PPE) being observed in the murine mesenteric microvasculature. Thus, we first administered the non‐specific MC stabilizer cromolyn sodium, to prevent MC activation towards a pro‐inflammatory phenotype. In addition, as AnxA1 has previously been described to be a key regulator of MC reactivity, acting as an endogenous MC stabilizer and to further promote the release of anti‐inflammatory mediators, we further aimed to test the influence of the AnxA1 mimetic peptide Ac2‐26. When analysing leukocyte‐endothelial cell interactions and PPE 20 min after LPS stimulation, we found that neither pretreatment with cromolyn sodium nor Ac2‐26 had any significant influence on the previously described slight differences in cell adherence (Figure 4B). However, both reagents significantly reduced the numbers of emigrated neutrophils (P < 0.05 respectively, Figure 4C) as well as PPE (P < 0.05 respectively, Figure 4D). These results demonstrate that both preventing the pro‐inflammatory activation (cromolyn sodium) and enhancing an anti‐inflammatory phenotype (Ac2‐26) in MCs were able to regulate the early in vivo inflammatory response in the mesenteric microvasculature. We further confirmed these anti‐inflammatory findings with FPR2/ALX specific ligands, WKYMVm and ATL, and in addition, we also showed that an exacerbation of inflammation was possible by targeting FPR2/ALX with the specific pro‐inflammatory agonist SAA (Figure S1). Furthermore, our results show that Ac2‐26 caused an increase on leukocyte rolling at 60 min and 120 min (Figure S2A) and decreased leukocyte adhesion at 120 min and emigration at both time points (Figure S2B+C).

MCs are a source of AnxA1 and mediate early neutrophil recruitment via AnxA1‐FPR2/ALX

Having established that MC activation is involved in early neutrophil migration and PPE and that Ac2‐26 (enhancing the anti‐inflammatory phenotype in MCs) is able to block this process like cromoyln sodium in vivo, we aimed to further investigate the relationship of neutrophils and MCs via the AnxA1‐FPR2/ALX pathway in the sequence of neutrophil recruitment in vitro. Using an in vitro transmigration assay, equiped with freshly isolated MC from the peritoneal cavity (De Filippo et al., 2013) and bone marrow‐derived neutrophils, we found that supernatant collected from the LPS‐stimulated MC had chemotactic activity and was able to promote neutrophil migration in vitro (Figure 5A). Furthermore, the LPS‐stimulated MC supernatant induced an increase in AnxA1 release from neutrophils (as measured by elisa, Figure 5B).

Figure 5.

MCs mediate early neutrophil recruitment in inflammation via FPR2/ALX. Supernatants from purified MC stimulated with saline or LPS (1 μg·mL⁻¹) for 4 h at 37°C with 5% CO₂ atmosphere were added to (A) a chemotaxis chamber to ascertain the ability of the supernatant to act as a neutrophil chemoattractant or directly to (B) neutrophils to measure whether the supernatant could induce AnxA1 release from neutrophils. MC was also treated for 20 min with saline, LPS (1 μg·mL⁻¹) or LPS + CMP 48/80 (10 μg·mL⁻¹) with or without the addition of cromolyn sodium (CRO), and supernatants were assayed using elisas for (C) histamine and (D) AnxA1 release or were added to (E) isolated murine neutrophils, which had been labelled with calcein AM and pre‐incubated with saline, the FPR pan antagonist Boc2 (20 μM) or the FPR2‐specific antagonist WRW4 (10 μM) for 10 min. The neutrophil‐MC supernatant cocktail was incubated together for 10 min prior to adding to a chemotaxis plate for 20 min with fMLP in the lower compartment. Data are expressed as mean ± SEM of three experiments with n = 6 mice per group or three experiments (chemotaxis). ^* P < 0.05 versus saline vehicle‐treated counterpart. ^# P < 0.05 versus vehicle + LPS + CMP 48/80. ^§ P < 0.05 versus neutrophil + LPS + CMP 48/80. ^& P < 0.05 versus neutrophil + vehicle + LPS + CMP 48/80 + CRO.

Next, we took purified MC and pretreated them with cromolyn or saline followed by a further (20 min) stimulation with either saline or LPS + CMP48/80. Using the obtained supernatant, we then tested it for presence of the key MC‐derived pro‐inflammatory molecule, histamine. LPS + CMP48/80 treatment resulted in a significant increase in histamine release (Figure 5C), which was blocked by the pretreatment with cromolyn. Since previous studies have shown that cromones are able to provoke the release of AnxA1 from cord‐derived human MCs (Yazid et al., 2010a,b), we next quantified the release of AnxA1 from cells in the above described set of experiments. We found that LPS + CMP48/80 resulted in a significant release of MC‐elicited AnxA1, which was exacerbated when cells were pretreated with the MC stabilizer cromolyn, enhancing the MCs' anti‐inflammatory phenotype.

Finally, having demonstrated that MC not only promotes the release of AnxA1 from neutrophils but also, when stabilized, presents an anti‐inflammatory phenotype (demonstrated by decreased production of histamine and increase in AnxA1 release), we wanted to address whether, when MC and neutrophils are coupled together, the production of AnxA1 from MC acts on neutrophils to help resolve inflammation. Figure 5E shows that the combination of neutrophils coupled with the supernatant from stimulated (LPS + CMP48/80) MC results in a robust in vitro neutrophil migration towards the chemoattractant fMLP. This response was significantly reduced when MC were pretreated with cromolyn. Based on our results in Figure 4D, this suggests a role for AnxA1 in limiting neutrophil recruitment, and therefore, we tested this hypothesis by pre‐incubating neutrophils with either the FPR pan‐antagonist Boc2 or the FPR2 selective antagonist WRW4 before exposure to MC supernatant. Both Boc2 and WRW4 caused a significant increase in neutrophil migration versus vehicle, suggesting the importance of neutrophil FPR2 in this process of MC‐mediated neutrophil migration via AnxA1 (Figure 6).

Figure 6.

Schematic overview of the important anti‐inflammatory role that MC plays in early neutrophil recruitment. LPS activates MCs causing them to release histamine and pro‐inflammatory cytokines, including IL‐6, which in turn leads to a rapid recruitment of neutrophils, within 20 min of LPS administration. MCs are ideally positioned in close proximity to the vasculature to initiate this early phase of neutrophil recruitment. These recruited neutrophils release factors such as MPO that also leads to the recruitment of more neutrophils. Mast stabilizing drugs, such as cromolyn sodium, destabilize MCs and cause the down‐regulation of histamine and the release of AnxA1. MC AnxA1 acts upon the FPR2/ALX on neutrophils that have arrived within the stimulated tissue, to help promote the anti‐inflammatory effect. This demonstrates a positive role for the MC in tissue inflammation via the FPR2/ALX pathway.

Discussion

We have used the technique of IVM, coupled with pharmacological approaches, to assess haemodynamic parameters within the mesenteric microcirculation, and the state of MC activation surrounding the vascular bed under study. We have demonstrated that upon LPS stimulation, MCs contribute to the early stages of neutrophil recruitment and that this process could be blocked by preventing the pro‐inflammatory activation (using cromolyn sodium) or enhancing an anti‐inflammatory phenotype (using Ac2‐26) in MCs. We further demonstrated in vitro that stimulated MCs induce neutrophil migration and that MC stabilization enhances the release of AnxA1, which acts upon neutrophil FPR2/ALX to limit further neutrophil recruitment. These findings demonstrate thus a positive role for MCs in the resolution of inflammation.

MCs are multifunctional and highly effective tissue‐dwelling cells found in most tissues of the body, especially in sites that are in close contact to the external environment, for example, skin, airways and intestines. Within seconds of stimulation, MC becomes activated, degranulate and release a variety of soluble factors including: histamine, proteases, for example, tryptase and chymase, and pre‐formed TNF‐α, followed by the production of lipid‐derived eicosanoids, for example, prostaglandin D₂ and leukotriene C₄ (Urb and Sheppard, 2012). MCs are also capable of selectively releasing pro‐inflammatory mediators, such as IL‐6, without degranulation (Theoharides et al., 2012). In the present study, exposure of the mouse mesentery to CMP48/80 (which bypasses the high‐affinity IgE receptor and acts directly on G‐proteins to produce MC degranulation) caused a rapid selective uptake of ruthenium red uptake by MCs, demonstrating that MCs are activated very rapidly after stimulation (Urb and Sheppard, 2012). Furthermore, it suggests that MCs were playing a role in the early neutrophil recruitment and oedema formation.

The dose of LPS used in our study was relatively low, chosen deliberately as a refinement tool and to activate the vasculature without causing either mortality or a drastic decrease in microvascular perfusion, as is seen at higher doses and which would complicate the inflammatory response and cause an increase in animal numbers (Zanetti et al., 1992; Damazo et al., 2005; Hughes et al., 2013).

Whilst MCs clearly participate in the induction and/or propagation of certain inflammatory diseases (Theoharides et al., 2012) and have been best characterized by their role in allergic reactions, there is growing interest in understanding how MCs may play a role in immunoregulation and in leukocyte recruitment (Abraham and St John, 2010; De Filippo et al., 2013). Depletion of MCs causes heightened acute neutrophil emigration and oedema, suggesting that MCs play a protective role in terms of neutrophil emigration and albumin leakage in response to LPS. Further evidence for a protective role of MCs has emerged from studies involving MC deficient mice. For example, Echtenacher et al. (1996) showed that Kit^W/Kit^W‐v (MC‐deficient) mice display a significantly increased mortality rate versus their wild‐type counterparts following acute septic peritonitis, which could be corrected by restoration of the deficiency. Thus, MCs are required for host defence in some murine models of bacterial infection, although the mechanisms utilized have until now been largely undefined.

The injection of a wide range of substances, including LPS, zymosan and thioglycollate (Rao et al., 1994; Henderson et al., 2003; Volman et al., 2005; Chen et al., 2008; Vemula et al., 2010), into the peritoneal cavity has been shown to induce the hallmarks of inflammation including pain, swelling and leukocyte infiltration (Doherty et al., 1995; Kubes and Gaboury, 1996; Jirillo et al., 2002). Here, although we found that both zymosan and thioglycollate induced fairly similar responses to those observed with LPS, they did not induce all inflammatory readouts that were measured, that is, rolling velocity, leukocyte adherence at 20 min, leukocyte emigration at 20 min and PPE (adherence being unchanged in zymosan treatment, yet increased in thioglycollate and LPS, and, conversely, emigration being unchanged in thioglycollate treatment, yet increased in zymosan and LPS). These differences are likely due to the inflammatory mediators and profiles instigated by each of the injected reagents and their involvement in ‘adapted homeostasis’ (Newson et al., 2014). Innate immune‐mediated responses to low‐dose LPS or low dose zymosan (Rao et al., 1994) resolve within days, although more recently, a previously overlooked second wave of leukocyte influx into tissues has been shown to occur and to persist for weeks (Newson et al., 2014). Studies reported by others on the use of thioglycollate‐induced peritonitis have also noted specific inflammatory changes such as a population of monocyte‐derived macrophages that persist in the peritoneum for at least 2 months post‐resolution (Newson et al., 2014). As described above, we chose to use LPS throughout the rest of the study based on the superior inflammatory profile in regard to endothelial–leukocyte interactions.

Previous studies have shown that MCs do not degranulate in vitro in response to LPS (Theoharides et al., 2012); however, acting via TLR4, LPS is able to cause the secretion of a number of different factors from murine MCs including IL‐6, IL‐13 and TNFα (McCurdy et al., 2001; Supajatura et al., 2001). Our data concur with findings of other groups (De Filippo et al., 2013) that LPS‐stimulated MC supernatant has chemotactic activity. Furthermore, there was an increase in AnxA1 release from MCs into the supernatant. AnxA1 is present in a number of myeloid cells including neutrophils and MCs (Oliani et al., 2000) and, once secreted from existing intracellular pools of activated cells, acts in a paracrine/autocrine fashion, utilizing FPRs to produce its biological effects (Perretti and D'Acquisto, 2009; Yazid et al., 2010a,b).

AnxA1 is an endogenous regulator of MC function, and administration of this protein has been shown to exert a tonic inhibitory influence on MC reactivity, for example, administration of Ac2‐26 to rats inhibits antigen‐induced histamine release in the pleural cavity (Bandeira‐Melo et al., 2005). AnxA1 is cleaved in activated MCs, and this cleavage is required for the phosphorylation of cytosolic phospholipase A₂ and subsequent inflammatory reactions, such as the production of eicosanoids.

Cromoglycate‐like drugs, for example, cromoyln sodium, have been described as ‘MC stabilizers’, due to their ability to prevent the release of histamine when stimulated by different agonists (Yazid et al., 2010a,b). It has been confirmed that PKC activation is responsible for the phosphorylation of AnxA1 in both cord‐derived MCs (Yazid et al., 2013) and in U937 cells (Yazid et al., 2009), occurring within 5 min when stimulated by cromones. Yazid et al. (2010a,b) also demonstrated that AnxA1 phosphorylation acts either wholly or partially through the FPR system to regulate the limit and extent of mediator release. It is also important to note that the AnxA1‐FPR system can be activated in the absence of MC degranulation by drugs such as glucocorticoids or cromones (Sinniah et al., 2016). Other groups have also suggested that the GPCR GP35 (which is present on MCs, eosinophils and basophils) also may play a role in the effects of cromones, although this may be more relevant to asthma and allergy rather than infection (Jenkins et al., 2010; Yang et al., 2010).

The addition of cromoyln sodium in this study resulted in a decrease in neutrophil extravasation and oedema. These effects were also mirrored by the administration of Ac2‐26. Both these compounds have been shown to inhibit leukocyte emigration, for example, cromoglycate exhibits a protective role in neutrophil‐dependent pathologies including intestinal (Kanwar and Kubes, 1994) and pulmonary (Vural et al., 2000) ischaemia–reperfusion models, and AnxA1 and its mimetic peptide have been shown to cause the detachment of leukocytes in the brain (Gavins et al., 2007) and periphery (Gavins et al., 2012). Da Silva et al. (2011) have also shown the importance of MCs in a model of endotoxin‐induced uveitis and suggest that AnxA1 may be an innovative form of therapy for uveitis by preventing MC activation and/or leukocyte infiltration.

Adhesion molecules on leukocytes and endothelial cells are up‐regulated in both brain and mesentery in response to LPS (Eppihimer et al., 1996; Henninger et al., 1997), supporting leukocyte–endothelial cell interactions. In our 2 h study presented here, we found that Ac2‐26 was able to promote anti‐inflammatory effects by decreasing mesenteric leukocyte adhesion and emigration in LPS‐treated mice. These findings support our previous studies in the brain (Gavins et al., 2012). The ability of neutrophils to prevent further leukocyte extravasation is clearly complex. Tissue resident macrophages have been previously studied and shown to interact with MCs. De Filippo et al. (2013) have shown that in response to LPS, MCs are able to quickly recruit neutrophils via release of preformed CXC chemokines (CXCL1/CXCL2), and this is followed by slower macrophage‐mediated recruitment of leukocytes deeper into the extravascular tissue (Abtin et al., 2014). Furthermore, Wolf et al. (2008) demonstrated that extravasating neutrophils phagocytosed by perivascular macrophages are able to negatively regulate further leukocyte extravasation by secreting a variety of resolution mediators such as resolvins, lipoxins, TGF‐β and IL‐10. Therefore, it is clear from our study and others that MCs play important modulatory roles in inflammation. Further, investigations are required to fully elucidate the functions of these versatile sentinel cells in leukocyte trafficking (Nourshargh and Alon, 2014). Additionally, it must be mentioned that MCs are also able to produce extracellular traps [mast cell extracellular traps (MCETs), which are composed of DNA, histones, tryptase and the antimicrobial peptide LL‐37] that encompass and kill organisms (von Kockritz‐Blickwede et al., 2008). However, the importance of these MCETs in clearing all type of inflammation has yet to be established, and it may be that more indirect effects of MCs in coordinating host innate and adaptive responses may be more important in the balance of host defence (von Kockritz‐Blickwede et al., 2008).

We have demonstrated in this study that MCs play a protective role in the regulation of extravasated neutrophils by the release of endogenous AnxA1, which then acts on neutrophil FPR2/ALX. MCs also cause the release of AnxA1 from the neutrophil itself, which can act in autocrine and paracrine way to resolve the inflammation. We suggest, based on other data in the literature, that the feedback of AnxA1 on the MC is likely to be in‐part dependent upon Fpr1 (Sinniah et al., 2016), and not solely on FPR2/ALX, as we found with the neutrophil. Sinniah et al. (2016) showed that whilst blocking FPR2/ALX in cord‐derived MCs either pharmacologically (using an antagonist) or genetically (by using cord‐derived MCs obtained from Fpr2/3‐null mice) prevented the action of nedocromil (MC stabilizing drug) on PGD₂ generation and release, it failed to completely block inhibitory action on histamine release, suggesting the involvement of another FPR. The fact that the effects of AnxA1 are mediated through different members of the FPR family has been reported before. Intact AnxA1 is a 37 kDa protein that exerts its anti‐inflammatory and pro‐resolving actions mainly through the dimeric FPR2/ALX, although it has also been shown to be able to elicit these actions via FPR1 in the epithelium (Leoni et al., 2013). Agonists of the classic GPCR, FPR2/ALX are associated with both pro‐inflammatory (e.g. SAA and cathelicidin) and pro‐resolving (e.g. AnxA1 and LXA₄) signalling pathways (Dahlgren et al., 2000; Perretti and D'Acquisto, 2009), which probably depend upon different homo‐ or hetero‐dimers of these receptors (Cooray et al., 2013). Although FPR2/ALX is also termed the ALX receptor, there remains some uncertainty as to whether lipoxin A₄ binds to FPR2/ALX, as evidence has been presented for (Chiang et al., 2006; Perretti and D'Acquisto, 2009) and against (Forsman and Dahlgren, 2009; Hanson et al., 2013) this phenomenon. Some of these discrepancies may lie in the fact that lipoxins are rapidly modified or degraded and thereby lose their biological activity (Forsman and Dahlgren, 2009). Nonetheless, it is now known that distinct FPR2/ALX domains are required for signalling by different agonists (Damazo et al., 2006) and the versatility of FPR2/ALX receptors appears to rely on the activation of receptor dimmers in a biased fashion (Perretti and D'Acquisto, 2009).

In the case of AnxA1, Ehrchen et al. (2007) showed that it can activate FPR2/ALX homodimerization but not the pro‐inflammatory SAA, and the peptide Ac2‐26 activates FPR2/ALX heterodimerization.

Upon cell activation, AnxA1 can be cleaved at the N‐terminal region by proteases such as elastase and proteinase‐3 (Vago et al., 2016). Differences in signalling mechanisms observed here could be due the result of a combination of signalling effects produced by AnxA1 or clipped forms, for example, Ac2‐26. Although peptides derived from the AnxA1 N‐terminal region have been shown to activate all three receptors of the FPR family (Perretti and D'Acquisto, 2009), it is important to mention here that some studies have suggested that the effects of certain peptides, for example, Ac9–25, transduce certain inhibitory signals in neutrophils through a receptor distinct from the members of the FPR family (Karlsson et al., 2005). These findings suggest that short AnxA1 mimetic peptides might fulfil other functions at variance to those reported for the parental protein (Perretti and D'Acquisto, 2009).

In summary, it is important to understand how neutrophils are recruited to control infection or injury. This study provides evidence for a role of MCs as sentinel cells in neutrophil extravasation and resolution of inflammation. This observation bears clinical relevance, since MCs are more and more being recognized as important gatekeepers in regulating immune responses beyond their known role in allergic responses. The described possibility to enhance the anti‐inflammatory phenotype of MCs to balance immune responses might allow more targeted therapy for inflammatory and allergic diseases. Further work is still needed to clarify whether MCs act as first alarm signals to recruit neutrophils upon detection of pathogenic challenge or whether they are activated by upstream signalling events (De Filippo et al., 2013). However, this study shows a role for MCs in the resolution of inflammation via the AnxA1‐FPR2/ALX pathway.

Author contributions

E.L.H. designed and performed experiments, interpreted results and helped write manuscript. F.B. provided scientific input, interpreted results and helped write manuscript. J.B. designed experiments, provided scientific input and helped write manuscript. R.J.F. provided scientific input and helped write manuscript. F.N.E.G. designed and performed experiments, provided scientific input and interpreted results and helped write manuscript.

Conflict of interest

The authors declare no conflicts of interest.

Declaration of transparency and scientific rigour

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research recommended by funding agencies, publishers and other organisations engaged with supporting research.

Supporting information

Figure S1 Effects of FPR2/ALX agonists on LPS‐induced inflammation in the mesenteric microcirculation. Mice were treated with Ac2‐26 (100 μg per mouse i.v.), ATL (4.0 μg per mouse i.v.), WKYMVm (100 μg per mouse i.v.), or SAA (0.2 nmol i.v.), 15 min prior to injection of LPS (10 μg per mouse i.p.). After 20 min, the mesentery was exteriorized for visualization of post‐capillary venules by IVM. Leukocyte‐endothelial cell interactions were quantified in terms of A) number of adherent (stationary for ≥30 s) leukocytes per 100 μm length, B) number of emigrated leukocytes per 100 x 50 μm² and C) FITC‐conjugated albumin was injected i.v. (0.25 mg·g⁻¹ body weight) and allowed to circulate for 5–10 min to quantify albumin leakage ([(Fl_out x bk)/(Flⁱⁿ x bk)] x 100%)) as a measure of plasma protein extravasation. Data are expressed as mean ± SEM. n = 5 mice per group. ^* P < 0.05 versus antiinflammatory ligands Ac2‐26, ATL and WKYMVm.

Figure S2 Effects of Ac2‐26 on LPS‐induced inflammation in the mesenteric microcirculation. Mice were treated with LPS (10 μg per mouse, i.p.) or saline vehicle. Ac2‐26 (100 μg per mouse i.v.) was administered 15 min prior to injection of LPS. At 0, 20, 60 or 120 min post‐LPS, the mesentery was exteriorized for visualization of post‐capillary venules by IVM. Leukocyte‐endothelial cell interactions were quantified in terms of A) leukocyte rolling velocity (expressed as V_WBC), B) number of adherent (stationary for ≥30 s) leukocytes per 100 μm length, C) number of emigrated leukocytes per 100 x 50 μm². Data are expressed as mean ± SEM. n = 6 mice per group. ^* P < 0.05 versus saline vehicle‐treated counterpart.

Supporting info item

Supporting info item

Hughes, E. L. , Becker, F. , Flower, R. J. , Buckingham, J. C. , and Gavins, F. N. E. (2017) Mast cells mediate early neutrophil recruitment and exhibit anti‐inflammatory properties via the formyl peptide receptor 2/lipoxin A₄ receptor. British Journal of Pharmacology, 174: 2393–2408. doi: 10.1111/bph.13847.