Neutrophil accumulation induced by bacterial lipopolysaccharide: effects of dexamethasone and annexin 1

Abstract

Annexin 1 (ANX-1) can reduce leucocyte migration in response to cytokines and chemokines in some rodent models of inflammation. However, its effectiveness against an inflammatory stimulus as strong as bacterial lipopolysaccharide (LPS) is unknown. Thus, we have examined whether ANX-1 can modulate LPS-induced neutrophil accumulation in the rat, as assessed by intravital microscopy and by myeloperoxidase (MPO) assay. The anti-inflammatory glucocorticoid, dexamethasone (DEX) was also studied for comparison. LPS superfusion induced adhesion of leucocytes to the endothelium and a subsequent increase in emigration from rat post-capillary venules over 2 h as assessed by intravital microscopy. Either ANX-1 or DEX was able to attenuate this adhesion and emigration of leucocytes. MPO activity in the lung, kidney and ileum was elevated after a 6-h exposure to LPS (intraperitoneal), indicating accumulation of neutrophils in these tissues. DEX attenuated the LPS-induced increase in MPO in the ileum but had no effect on MPO in the lungs or kidneys. This would suggest that the underlying mechanism by which neutrophils accumulate in the ileum, and more generally in the gastrointestinal compartment, is different from other vascular beds. ANX-1 had no effect on the LPS-induced increase in MPO activity in any of the tissues studied. Thus, from these data, ANX-1 appears to reduce leucocyte adhesion and emigration induced by a short-term (2 h), but not a longer (6 h) exposure to LPS.

Introduction

In response to an inflammatory stimulus, leucocytes may infiltrate tissues as part of the host immune response. This process occurs in several stages. First, upon application of an inflammatory stimulus neutrophils roll on the endothelium of post-capillary venules, a process mediated by the selectins, CD62P, CD62E, CD62L [1]. The leucocytes may then become activated and adhere firmly via β₂-integrins [2]. There are many compounds that can activate leucocytes, including chemoattractants, such as complement factor C5a [3], and chemokines, such as IL-8 [4]. Following adhesion, leucocytes undergo cytoskeletal reorganization before emigrating through junctions between endothelial cells into the extravascular space, a process called diapedesis.

Bacterial lipopolysaccharide (LPS), a potent stimulator of inflammation, can produce endotoxaemia and, in extremis, multiple organ failure. LPS releases an array of proinflammatory mediators, such as complement factor C5a [5], cytokines like tumour necrosis factor-alpha (TNF-α) [6], and chemokines like IL-8 [6] and CINC [7,8]. Furthermore, LPS can cause changes in adhesion molecules, such as shedding of l-selectin and expression of CD11/CD18 integrin [9,10]. These factors may combine to bring about neutrophil accumulation and infiltration into tissues that, in turn, may lead to tissue injury through the production of superoxide radicals.

Since neutrophils contribute to the tissue injury observed in endotoxaemia, it may be beneficial to prevent their accumulation. Glucocorticoids have long been used as anti-inflammatory agents and they have been shown to reduce leucocyte accumulation in many models [11,12], including LPS-induced leucocyte trafficking across rat mesenteric post-capillary venules [7]. However, glucocorticoids possess a wide range of side-effects. Thus, of increasing interest is the research into anti-inflammatory compounds that prevent neutrophil infiltration yet possessing less adverse effects than glucocorticoids.

Annexin 1 (ANX-1), also known as lipocortin 1, is an endogenous protein that is widely distributed among cells involved in the inflammatory response [13]. Since anti-ANX-1 antibodies can block some of the anti-inflammatory effects of dexamethasone [14] it has been suggested that ANX-1 mediates some, though by no means all [15], of the anti-inflammatory actions of glucocorticoids. Indeed, exogenous ANX-1 reduces neutrophil infiltration in several animal models and against various inflammatory stimuli including IL-1β and zymosan (for review see [16]). However, the effect of exogenous ANX-1 has yet to be investigated in a rat vascular bed or in a model of inflammation as severe as that produced by LPS.

Thus, we have investigated the anti-inflammatory effects of dexamethasone (DEX) and ANX-1 by their ability to modulate (i) LPS-induced leucocyte adhesion and infiltration, assessed by intravital microscopy of rodent mesenteric post-capillary venules, and (ii) neutrophil accumulation in various tissues, assessed by the myeloperoxidase (MPO) assay, following a longer exposure to LPS.

Materials and methods

Intravital microscopy

The rat mesenteric preparation was performed as described previously [12,17]. Male Sprague–Dawley rats (170–220 g supplied by Tuck, UK) were fasted 24 h prior to use. Animals were anaesthetized (Inactin; 120 mg/kg, i.p.) and the trachea cannulated to facilitate spontaneous breathing. The jugular vein was also cannulated for administration of drugs. The abdominal area was shaved, a midline laparotomy performed and the mesenteric vascular bed exteriorized and placed on a Plexiglas viewing stage. This was mounted on a light microscope (Zeiss Axioskop FS, eyepiece magnification × 10) with a water immersion objective lens (× 40 magnification) and then transluminated with a 100-W halogen light source. Images were acquired by a colour camera (Hitachi CCD, model KPC571; Tokyo, Japan), displayed on a colour video monitor (Sony Trinitron, model PVM 1440QM) and then recorded on a VHS video recorder (Sony, model SVO-9500 MDP) for subsequent off-line analysis. A video time-date generator (FOR.A video timer, model VTG-33; Tokyo, Japan) projected the time, date and stopwatch function on to the monitor. Erythrocyte velocity was measured using an optical Doppler velocimeter (Microcirculation Research Institute, Texas A & M University, College Station, TX). The mesentery was superfused with warmed (37°C), gassed (5% CO₂/95% N₂) bicarbonate-buffered solution (g/l: NaCl 7·71; KCl 0·35; MgSO₄·7H₂O 0·145; NaHCO₃ 1·51; CaCl₂ 0·22; pH 7·4) at a rate of 2 ml/min. Rat body temperature was maintained at 37°C via the use of a heating lamp.

Protocol for intravital microscopy

A ‘real time’ protocol was used. Briefly, after a suitable mesenteric venule (25–40 μm diameter and less than eight adherent cells) was found the preparation was superfused with buffer and allowed to stabilize for 20 min before a basal recording of 10 min was made. The mesentery was then superfused (2 ml/min) with LPS (serotype 0127:B8, 1 μg/ml) in the bicarbonate buffer or without LPS in control animals. To investigate the effects of ANX-1 on leucocyte adhesion and emigration induced by LPS, animals were treated with ANX-1 (100 μg/kg, i.v.) or vehicle (PBS) 20 min after the LPS superfusion was started. Alternatively, to investigate the effect of DEX on adhesion and emigration induced by LPS, DEX (1 mg/kg s.c.) was given 1 h prior to the LPS superfusion. DEX was given as a pre-treatment in order that its genomic effects would have time to work, while ANX-1 was given therapeutically, i.e. after the inflammatory stimulus, as it produces its effects rapidly [18]. The doses chosen have been validated in previous studies [12,18]. The mesenteric post-capillary venule was then recorded onto video cassette for 2 h and the 10-min periods at 10, 30, 60 and 120 min after the LPS were analysed. The following parameters were monitored: (i) rolling velocity (Vwbc), calculated from 20 randomly selected rolling leucocytes; (ii) cell adhesion (the number of cells that adhere to the endothelium for > 30 s; (iii) cell emigration (the number of cells outside the vessel but within the field of view (50 μm); and (iv) blood velocity. From the mean blood velocity (Vmean = centreline blood velocity/1·6) and the vessel diameter the wall shear could be calculated assuming a cylindrical geometry. Wall shear rate was calculated by the Newtonian definition: SR = 8000 × (Vmean/diameter) [19].

MPO study

In initial experiments, male Sprague–Dawley rats (170–220 g) received LPS (1 mg/kg i.p., serotype 0127:B8), or sterile saline (i.p.) for controls, and 6 h or 18 h later the animals were exsanguinated and 50 ml of sterile saline perfused through the animal via the left ventricle to flush blood from the organs. Tissues (kidney, ileum, lung and liver) were then harvested, snap-frozen in liquid nitrogen and kept at −80°C until ready for the MPO assay.

In subsequent experiments, DEX (0·1–1 mg/kg, s.c.) or sterile PBS was given 1 h prior to i.p. administration of LPS (1 mg/kg) or sterile saline. Alternatively, rats received ANX-1 (100 μg/kg, s.c.) concomitantly with the LPS. Tissues were collected 6 h later as described earlier.

For the MPO assay 50 mg of tissue were homogenized on ice in 1 ml of a 0·5% hexa-decyl-trimethyl ammonium bromide (HTAB) in MOPS (10 mm, pH 7) buffer. After homogenization, samples were centrifuged (4000 g) for 20 min and then the supernatant was removed and aliquoted for the assay. The MPO assay was performed as previously described [20,21]. Briefly, the sample supernatant was allowed to react with a solution of tetra-methyl benzidine (TMB; 1·6 mm) and 0·1 mm H₂O₂. The change in absorbance was measured with a spectrophotometer at 650 nm and compared with a standard curve of MPO (0·03125–1 U/ml).

Statistical analysis

Data points were compared with an unpaired t-test or a Mann–Witney U-test as appropriate.

Materials

LPS (serotype 0127:B8), NaCl, KCl, MOPS, HTAB, TMB, H₂O₂ and MPO for the standard curve were all obtained from Sigma Chemical Co. (Poole, UK). NaHCO₃ and CaCl₂ were obtained from BDH (Lutterworth, UK). MgSO₄·7H₂O was obtained from FSA Lab Supplies (Loughborough, UK). Dexamethasone was obtained from David Bull Labs (distributed by Central Laboratories Ltd, Dublin, Eire). Inactin was obtained from Research Biochemicals Int. (Natick, MA). Highly purified human recombinant ANX-1 was supplied by Dr E. Solito (INSERM U332, Institute Cochin de Génétique Moleculaire, Paris, France) and endotoxin contamination was < 20 pg/ml as measured by the Limulus amoebocyte chromogenic assay [18]. ANX-1 was kept at a stock solution of 0·4 mg/ml (in 20 mm 2-N-morpholinoethanesulphonic acid, pH 6). Subsequent dilutions were in PBS pH 7·4 and were made on the day of the experiment.

Results

Intravital microscopy

Intravital microscopy of the rat mesentery was utilized initially. LPS (1 μg/ml superfusion) induced adhesion of leucocytes to the endothelium that had reached a maximum within 30 min (Fig. 1a). This increase in adhesion was followed by a subsequent increase in emigration that was still increasing by 120 min (Fig. 1b). LPS superfusion had no significant effect on leucocyte rolling velocity (Fig. 2a), blood velocity (Fig. 2b) or shear rate (Table 1) in mesenteric post-capillary venules. ANX-1 (100 μg/kg, i.v.) was able to attenuate the LPS-induced adhesion and emigration of leucocytes (Fig. 1a,b) without altering rolling velocity (Fig. 2a), blood velocity (Fig. 2b) or shear rate (Table 1). DEX (1 mg/kg s.c.) given 1 h prior to an LPS superfusion blocked the increase in leucocyte adhesion and emigration. Leucocyte rolling velocity was significantly greater in DEX-treated animals compared with control (at 60 min) or LPS-treated (at 30 and 60 min) animals. DEX pre-treatment had no significant effect on blood flow or shear rate throughout the experiment.

Fig. 1.

Effect of annexin 1 (ANX-1; 100 μg/kg, i.v., hatched columns) or dexamethasone (DEX; 1 mg/kg s.c., cross-hatched columns) on the changes in leucocyte adhesion (a) and emigration (b) induced by a lipopolysaccharide (LPS) superfusion (1 μg/ml, ▪) in mesenteric post-capillary venules as assessed by intravital microscopy. □, Time-matched controls superfused with normal bicarbonate buffer. The open arrow indicates the onset of the LPS superfusion while the closed arrow indicates the time of the ANX-1 injection. Data are the mean and s.e.m. mean of n = 5–6 rats. *P < 0·05 versus controls; **P < 0·05 versus LPS-treated rats.

Fig. 2.

Effect of annexin 1 (ANX-1; 100 μg/kg, i.v., hatched columns) or dexamethasone (DEX; 1 mg/kg s.c., cross-hatched columns) on the changes in leucocyte rolling velocity (a) and blood flow (b) induced by a lipopolysaccharide (LPS) superfusion (1 μg/ml, ▪) in mesenteric post-capillary venules as assessed by intravital microscopy. Time-matched controls (□) were superfused with normal bicarbonate buffer. The open arrow indicates the onset of the LPS superfusion, while the closed arrow indicates the time of the ANX-1 injection. Data are the mean and s.e.m. mean of n = 5–6 rats. *P < 0·05 versus controls; **P < 0·05 versus LPS-treated rats.

Table 1.

Effect of dexamethasone (DEX; 1 mg/kg, s.c.), and annexin 1 (ANX-1; 100 μg/kg, i.v.) on changes in shear stress (s⁻¹) induced by a lipopolysaccharide (LPS) superfusion (1 μg/ml) on rat mesenteric post-capillary venules

	0 min	30 min	60 min	120 min
Control	451·5 ± 54·5	427·7 ± 57·4	416·3 ± 51·0	403·8 ± 55·5
LPS (1 μg/ml)	435·8 ± 48·3	450·8 ± 61·3	436·7 ± 47·7	367·1 ± 62·9
LPS + ANX-1	343·4 ± 34·2	347·4 ± 45·5	354·2 ± 47·2	328·3 ± 37·5
LPS + DEX	384·5 ± 35·5	418·9 ± 56·0	370·1 ± 52·8	241·8 ± 59·6

Data are the mean ± s.e.m. mean of n = 5–6 experiments.

Systemic endotoxaemia

When rats were treated with LPS (1 mg/kg, i.p.) for 6 h, the MPO activity in the lung, kidney and ileum (Fig. 3), but not the liver (control 8·29 ± 2·36 versus LPS 9·63 ± 2·73 U/mg tissue, n = 5) was elevated over levels found in control rats. Since the increase in MPO activity induced by LPS had waned by 18 h in all tissues studied except the kidney (Fig. 3), the 6 h time point was used to investigate the effects of DEX or ANX-1. ANX-1 (100 μg/kg s.c.) had no effect on the increased MPO activity in any of the tissues induced by LPS for 6 h (Fig. 4). DEX (1 mg/kg, s.c.) pre-treatment attenuated the increase in MPO in the ileum at 6 h but had no affect on MPO in the lungs or kidneys (Fig. 4). The lower dose of DEX administered (0·1 mg/kg s.c.) did not significantly affect the LPS-induced increase in MPO activity in any of the tissues collected.

Fig. 3.

Effect of lipopolysaccharide (LPS; 1 mg/kg i.p., ▪) or sterile saline (control, □) on myeloperoxidase (MPO) activity in the lung (a), ileum (b) and kidney (c) at 6 h and 18 h post-injection. Data are the mean and s.e.m. mean of n = 5–12 samples. *P < 0·05 compared with controls.

Fig. 4.

Effect of dexamethasone (DEX; 0·1 mg/kg, s.c., horizontally hatched; 1 mg/kg, s.c., diagonally hatched columns), or annexin 1 (ANX-1; 100 μg/kg, s.c., cross-hatched columns) on the lipopolysaccharide (LPS; 1 mg/kg i.p., ▪)-induced increase in myeloperoxidase (MPO) activity in the lung (a), ileum (b) and kidney (c) at 6 h. □, Control injection of sterile saline. Data are the mean and s.e.m. mean of n = 6–12 samples. *P < 0·05 compared with controls; **P < 0·05 compared with LPS-treated rats.

Discussion

Bacterial LPS is a potent stimulus of neutrophil adhesion and infiltration. LPS induced an increase in leucocyte adhesion and emigration in our rat intravital microscopy model, in keeping with previous studies [7,22]. When injected intraperitoneally, LPS increases in the circulation, until it peaks after 30 min [23]. This then increases the systemic levels of proinflammatory cytokines, such as IL-1 and TNF-α [6,24] and chemokines such as CINC [7,8], as well as activating the complement cascade [5]. Furthermore, LPS can affect the expression of adhesion molecules, with L- and P-selectin [22,25] and the integrins CD11/CD18 [9] all being important in LPS-induced leucocyte adhesion to post-capillary venules.

Leucocyte migration in many models of inflammation, including those caused by IL-1 [26] and zymosan [27], can be reduced by ANX-1. Indeed, it is not just exogenous ANX-1 that reduces neutrophil accumulation, as experiments with anti-ANX-1 antibodies show that endogenous ANX-1 contributes to the anti-migratory effects of glucocorticoids [14]. In static conditions, endogenous ANX-1 does not appear to modulate adhesion of leucocytes to endothelial cell monolayers [28]. However, in murine post-capillary venules in vivo, exogenous ANX-1 can reduce adhesion to a significant extent [18], an observation that is in agreement with the findings presented here. Reduced leucocyte adhesion does not appear to be the main mechanism by which ANX-1 attenuates leucocyte trafficking. Instead, ANX-1 seems to promote leucocyte detachment and to prolong the process of diapedesis [14,18]. The exact mechanisms by which ANX-1 achieves this are unknown, although it does first require that endogenous leucocytic ANX-1 be externalized [28].

Although ANX-1 was able to inhibit emigration of leucocytes caused by LPS over 2 h in mesenteric post-capillary venules, no effect was observed on neutrophil accumulation in the ileum after 6 h. Why there is a difference between 2 h and 6 h is unknown. It may be that the difference observed is methodological in origin, since the route of administration differed and subcutaneous ANX-1 may not be absorbed and distributed into the circulation. This route of administration was chosen for the MPO work since it was a more convenient method of dosing conscious animals and it has been shown to be effective for ANX-1-derived peptides [27]. An alternative explanation is that ANX-1 may have been degraded over the 6-h period of the experiment so that it was no longer present at a pharmacologically effective concentration in the plasma. A more likely suggestion may be that there are differences in the regional blood flow at 6 h. For instance, following LPS, the blood pressure falls and nitric oxide synthase is induced [29], an occurrence that typically takes place within 4–6 h. Peripheral vasoconstriction and a fall in blood pressure are hallmarks of LPS-induced shock and may lead to poor organ perfusion. Thus, at 6 h, blood flow and shear rates in the tissues may fall so that leucocytes become adherent regardless of ANX-1 treatment.

For comparison against ANX-1, DEX was also utilized as an anti-inflammatory agent in our experiments. DEX greatly reduced leucocyte adhesion and emigration induced by LPS in mesenteric post-capillary venules. These findings were somewhat unsurprising, since DEX has been previously shown to reduce leucocyte rolling and adhesion to IL-1β [12] and LPS [7] and glucocorticoids can down-regulate some cytokines and chemokines (for review see [30]). DEX was also able to raise significantly leucocyte rolling velocity compared with both LPS and control treatments. This is probably associated with an interference with the selectins, since both L- and P-selectin mediate leucocyte rolling induced by LPS [22]. Furthermore, DEX has been shown to alter LPS-induced changes in L-selectin expression in the rat [7], and has been shown to attenuate the increase in P- and L-selectin expression induced by LPS in the mouse [31]. That DEX can increase leucocyte rolling velocity compared with control rats may indicate that there is a small amount of activation in our intravital model that seems to affect leucocyte rolling velocity, but not adhesion or emigration. Indeed, previous studies have shown that activation and increased leucocyte rolling is a consequence of disturbing the microcirculation [32]. We found that DEX was also able to reduce MPO levels in the ileum of LPS-treated rats, which is in keeping with our data from intravital microscopy experiments in the mesentery. Furthermore, this may add credence to the theory that blood flow changes are responsible for the lack of effect of ANX-1 at 6 h. For instance, DEX can produce genomic effects, such as reducing nitric oxide production by decreasing nitric oxide synthase (NOS) induction [33,34], that may result in improved organ perfusion at these later times.

DEX was unable to affect the increase in neutrophil accumulation induced by LPS in the lung and kidney at 6 h. Similar findings where DEX was ineffective at preventing an increase in MPO in the lungs of mice induced by i.p. LPS have been reported [35]. Histological sections of rat [23] or mouse [35] lung following i.p. treatment with LPS show an increase in neutrophils in the pulmonary vessels but that these neutrophils had not extravasated. Thus, neutrophil accumulation in the lung does not seem to be due to emigration and may be due to trapping in small vessels. This indicates a clear difference in how neutrophils accumulate in lungs and the ileum/mesentery in response to i.p. LPS, and may explain why DEX is not effective attenuating neutrophil gathering in the lung. It may be that a similar mechanism is in operation to cause renal neutrophil accumulation. An additional factor which may account for the tissue differences in neutrophil accumulation is that, whilst LPS increases in the circulation following i.p. injection, it may still be present locally in the ileum, thus triggering the release of local chemotactic agents.

In conclusion, here we show for the first time that ANX-1 itself reduces neutrophil trafficking induced by LPS. Furthermore, this is the first study that reports the ability of ANX-1 to interfere directly with neutrophil extravasation in the rat, since previous studies had been performed with ANX-1-derived N-terminal peptides [21]. Our observations of rat mesenteric post-capillary venules point to a cellular mechanism for the anti-inflammatory action exerted by ANX-1 in several rat models including carrageenin paw oedema [36] and zymosan pleurisy [37]. ANX-1 however, was ineffective at preventing neutrophil accumulation in tissues after a more prolonged exposure to LPS. Finally, our data also suggest that DEX may modulate neutrophil accumulation by distinct mechanisms in different tissues.