Molecular Determinants of Monosodium Urate Crystal-Induced Murine Peritonitis: A Role for Endogenous Mast Cells and a Distinct Requirement for Endothelial-Derived Selectins1
- Stephen J. Getting1,
- Roderick J. Flower1,
- Luca Parente2,
- Rinaldo de Médicis3,
- André Lussier3,
- Barry A. Woliztky4,
- Marco A. Martins1 and
- Mauro Perretti1
- 1Department of Biochemical Pharmacology, The William Harvey Research Institute, London, United Kingdom (S.J.G., R.J.F., M.A.M., M.P.);2Istituto di Farmacologia e Farmacognosia, Facolt[aa]a di Farmacia, Palermo, Italy (L.P.); 3Rheumatic Disease Unit, Faculty of Medicine, Université de Sherbrooke, Sherbrooke, Canada (R.d.M., A.L.);4Department of Inflammation/Autoimmune Diseases, Hoffman-La Roche Inc., Nutley, New Jersey (B.A.W.)
Abstract
Injection of monosodium urate (MSU) crystals, the etiological cause of gouty arthritis, into murine peritoneal cavities produced an intense recruitment of polymorphonuclear leukocytes (PMN). After 3 mg MSU crystal injection, cell influx was maximal (∼ 10 × 106 cells per mouse) at 6 hr postinjection and sustained up to the 24 hr time-point. In mice depleted of mast cells by administration of compound 48/80 72 hr before challenge with MSU crystals a lower PMN influx was measured (58% reduction). The occurrence of endogenous mast cell activation, in the MSU response, was validated by the observation that MSU challenge reduced by more than 90% the number of intact mast cells recovered in the peritoneal washes. Pretreatment of mice with a histamine H1 antagonist (tripolidine; 0.5 mg/kg) or a platelet-activating factor receptor antagonist (WEB2086; 10 mg/kg) significantly reduced by 50 to 60% the number of PMN recovered from the peritoneal cavities. The molecular determinants of this process of leukocyte recruitment were also investigated. Treatment of mice with an anti-CD62P or anti-CD62E monoclonal antibody (mAb; 100 μg i.v.) produced a distinct inhibition of PMN recruitment measured at 6 hr, whereas only a combined administration of both monoclonal antibodies was effective in reducing by 60% the influx of PMN caused by the MSU crystals within 24 hr. In conclusion, these data highlight a role for endogenous mast cells and for endothelial-derived selectins in MSU crystal-induced PMN recruitment into the peritoneal cavity, and may be useful to dissect molecular mechanism(s) which may be operating in gouty arthritis.
Deposition of MSU and CPPD crystals in the joint articular space is the etiological cause of acute inflammatory conditions such as gout and pseudogout, respectively (McCarthy et al., 1962; Dieppeet al., 1979). Clinically, these inflammatory diseases are associated with edema and erythema of the joints with consequent severe pain. A strong infiltration of leukocytes in the intraarticular and periarticular space is also characteristic of these pathologies. In particular, PMN are the predominant cell type recovered from these inflammatory joints (Terkeltaub, 1992; Dieppe et al., 1979).
The clear relationship between MSU crystals and gouty arthritis has, in the past, prompted the characterization of experimental models of crystal-induced inflammation. For instance, injection of crystals into a preformed rat air-pouch (Brooks et al., 1987) or within the rat pleural cavity (Sedgwick et al., 1985) produced an intense PMN accumulation associated with generation of chemoattractants such as leukotriene B4 (Brooks et al., 1987). More recently, the ability of MSU crystals to activate human neutrophils in vitro has been reported. Incubation of these cells with tumor necrosis factor produces release of both interleukin-1 and interleukin-1 receptor antagonist, and coaddition of MSU crystals potentiate the release of the former without affecting the latter (Roberge et al., 1994). Furthermore, in similar conditions the synthesis of a CXC chemokine selective for PMN, such as interleukin-8, has also been observed, but not that of a CC chemokine selective for monocytes, such as monocyte chemoattractant protein-1 (Hachica et al., 1995). In addition, MSU crystal activation of mononuclear phagocytes, which are normally found in the joint space, also induces secretion of interleukin-8 (Terkeltaub et al., 1991). All these studies provide further evidence that gout is mainly a PMN driven pathology.
The events that regulate PMN tropism towards the inflammatory sites have been the subject of extensive investigation in recent years (Springer, 1994). In the initial phase, the rolling of leukocytes on the endothelium is a prerequisite for subsequent adhesion (Von Andrianet al., 1992; Lawrence and Springer, 1991) and it is mainly mediated by the selectins (Lasky, 1992; Abbassi et al., 1993); interaction between the β2-integrins on the leukocyte membrane and their counterpart (members of the immunoglobulin superfamily) on the endothelium sustains the firm adhesion (Hynes, 1992; Carlos and Harlan, 1990). Less is known about the actual emigration process, as characterized by pseudopodia emission and opening of the endothelial gaps, although it is clear that it requires G-protein-mediated cell activation (Springer, 1994).
Among the selectins, L-selectin (CD62L) is constitutively expressed on PMN membrane and is required for the rolling of cells, mediated by carbohydrate ligands, on the endothelial surface. P-selectin (CD62P) and E-selectin (CD62E) are expressed on the endothelium under different molecular mechanisms: P-selectin expression is both constitutive and inducible upon treatment with several agents, e.g.,histamine (Asako et al., 1994); E-selectin is expressed, at least in vitro, only after stimulation of the endothelium with pro-inflammatory cytokines (Springer, 1994).
We have recently characterized the existence of an endogenous pathway that controls the PMN extravasation process and it is centered on the action of leukocyte-derived lipocortin 1. Activation of this pathway by exogenous administration of lipocortin 1 mimetics, such as its N-terminus peptide acetyl 2–26 produces a significant inhibition of PMN accumulation in distinct models of acute inflammation (Perrettiet al., 1993; Getting et al., 1997).
We describe a novel experimental mouse peritonitis model, in which a strong PMN accumulation is observed after challenge with MSU crystals. This was used to characterize the mechanisms of MSU crystal-induced PMN recruitment, mainly in the light of recent understandings of the mechanisms responsible for the initiation of the leukocyte extravasation process, i.e., role of endogenous mast cells and of endothelial-derived selectins. The effect of the lipocortin 1-derived peptide Ac2–26 was also assessed to investigate the potential therapeutic application of agents that activate the lipocortin 1 pathway in gout.
Materials and Methods
Animals.
Male Swiss Albino mice (20–22 g body weight) were purchased from Banton and Kingsman (T.O. strain; Hull, Humberside, UK), and maintained on a standard food pellet diet with tap water ad libitum using a 12:00 hr light/dark cycle. Animals were used 3 to 4 days after the arrival.
MSU-induced PMN recruitment.
The peritonitis was induced by injection of 1 to 5 mg MSU crystals, in 0.5 ml PBS (0.1 M, pH 7.4). At different time-points, animals were euthanized by CO2 exposure, peritoneal cavities washed with 3 ml of PBS containing 3 mM EDTA and 25 U/ml of heparin. Aliquots of the lavage fluids were then stained with Turk’s solution (0.01% crystal violet in 3% acetic acid) and differential counting performed using a Neubauer hemacytometer and a light microscope (Olympus B061). Mononuclear cells and PMN could be easily identified. The large predominance of neutrophils in the PMN population in 6-hr lavage fluids was confirmed in cytospin preparations stained with May-Grunwald and Giemsa, confirming that >95% of PMN were neutrophils (not shown). Data are reported as 106 PMN/mouse.
Lavage fluids were then centrifuged at 400 × g 10 min and supernatants stored at -20°C before evaluation of the β-glucuronidase activity (see below). Protein contents of cell-free lavage fluids was also determined after staining in Blue Coomassie. Readings were made at 570 nm wavelength in a plate-reader, and absorbance values were converted to mg/ml after comparison with a standard curve constructed with bovine serum albumin (0–5 mg/ml).
Other model of peritonitis.
CPPD crystal- or zymosan-induced PMN recruitment were assessed either at 6 or 24 hr postchallenge. The dose of 3 mg was chosen for each of these insoluble phlogogens since comparison with an identical dose of MSU crystals was made. The number of PMN recruited into the murine peritoneal cavities was quantified as described above.
Endogenous mast cell.
The role played by resident peritoneal mast cells on PMN accumulation elicited by MSU crystal injection was validated by depleting animals of their endogenous mast cells, a procedure already employed successfully to delineate the role of the chemokine interleukin-8 (Perretti et al., 1994). As adapted from Diaz et al. (1996), compound 48/80 was given i.p. (10 μg in 0.5 ml) 72 hr before MSU crystals, a time chosen to assure that any acute inflammatory response elicited by the mast cell activator would have subsided. Control mice received sterile PBS. PMN recruitment was assessed 6 hr after MSU crystal (3 mg) injection as described above.
As a way to monitor mast cell activation in vivo, the number of intact mast cells in different inflammatory conditions was determined. Cell pellets were stained in 0.1% toluidine blue (in 50% ethanol, 7.6% paraformaldehyde, 1% acetic acid). This staining allows the clear identification of intact mast cells since it interacts with the heparinic component of the cytoplasmic granules. Data are reported as 103 intact mast cells per cavity.
The ability of MSU crystals to cause mast cell degranulation in vitro was also assessed. For this purpose, peritoneal cells were lavaged as described above and washed twice in RPMI-1640 medium supplemented with 2% fetal calf serum. Pellets were then resuspended at 5 × 106 cells/ml and incubated at 37°C in a shaking water bath with medium, MSU crystals or CDDP crystals at a final concentration of 1 mg/ml, selected from a previous study (Hachicaet al., 1995). Aliquots (0.5 ml) of cell suspension were taken at 30, 60 and 120 min, centrifuged at 1000 r.p.m. × 10 min in a minifuge: supernatants were discarded and the resulting pellets were stained overnight with 100 μl Toluidine blue solution. The number of intact mast cells was then determined using a Neubauer hemacytometer and light microscopy.
Drug treatment.
The H1-antagonists, tripolidine (0.5 mg/kg) and diphenhydramine (9 mg/kg), or the PAF receptor antagonist WEB2086 (10 mg/kg) were given i.p. simultaneously with the local injection of MSU crystals (3 mg) (Harris et al., 1996). Peptide Ac2–26 was given s.c. and the dose (200 μg per mouse) was chosen from a recent study in which the peptide effectively inhibited PMN accumulation into the mouse peritoneal cavity in response to challenge with zymosan (Getting et al., 1997). The mAbs were administered i.v. according to published protocols (Rosen and Gordon, 1987; Watson et al., 1991). Mice were treated i.v. with 100 μg rat anti-mouse CD11b mAb, rat anti-mouse P-selectin or rat anti-mouse E-selectin mAb 1 hr before challenge with MSU crystals (3 mg). Control animals received an equal dose of nonimmune rat IgG. The sulfated polysaccharide, fucoidin, was given i.v. 2 hr before MSU crystals at the dose of 0.3 mg per mouse (Harriset al., 1995). In most cases PMN accumulation was evaluated at the 6 hr time-point. However, the different role of selectins was investigated also at 24 hr post-MSU crystal challenge: in this case, mAbs were given i.v. 6 hr after the crystals, and PMN accumulation was quantified at the 24 time-point as described above.
β-Glucuronidase activity.
β-Glucuronidase activity in the supernatants was measured according to a published protocol (Iwamura et al., 1993). A total of 250 μl of cell-free lavage fluids was incubated with the substrate phenolphthalein-β-glucuronic acid (1 mM) in 0.5 ml total volume and kept in a water bath at 37°C with gentle shaking for 18 hr. Reactions were stopped by addition of 1 ml of ice cold glycine buffer (200 mM) in 200 mM NaCl (pH 10.4). Absorbance values, measured at 550 nm using a 96-well plate multi-reader, were transformed into U/ml of lavage fluid, using a standard curve constructed with 0 to 2000 U β-glucuronidase. Data are reported as U per × 106 cells per mouse.
Materials.
MSU and CPPD crystals were prepared by a previously described method (Roberge et al., 1994; Denko and Whitehouse, 1976). A boiling solution of 0.03 M MSU, pH 7.5, was prepared by dissolving equimolar quantities of uric acid and sodium hydroxide and filtering with an Acropor membrane filter (AN-3000, 3 μM; Gelman, Ann Arbor, MI). Sodium chloride (0.1 M final concentration) was added to accelerate and improve the uniformity of the crystallization. CPPD was obtained by mixing a calcium nitrate solution (0.1 M final concentration) with an acidic solution of sodium pyrophosphate (0.025 final concentration in Na2P2O7and 0.03 M HNO3). The milky white precipitate formed CPPD crystals after 1 day at 50 to 60°C. The crystals were characterized by x-ray diffraction (Rigaku Geirflex D/max), by examination under phase and polarizing microscopy and by scanning electron microscopy. The MSU and CPPD crystals showed triclinic morphological characteristics. Their dimensions, determined by scanning microscopy, were 10 × 1 × 1 μm to 25 × 1.5 × 1.5 μm, and 12 × 1.4 × 1.4 to 25 × 1.7 × 1.7 μm for MSU and CPPD, respectively. Both preparations were free of endotoxin as determined by the Limulus Assay (Whittaker, Walkersville, MS). The in vitro efficacy of these crystals on human PMN has been recently validated (Hachica et al., 1995).
The lipocortin 1-derived peptide Ac2–26 (acetyl-AMVSEFLKQAWFIENEEQEYVVQTVK) was prepared by The Advance Biotechnology Centre (The Charing Cross and Westminster Medical School, London, UK) using solid phase step-wise synthesis. Purity was more than 90% as assessed by high pressure liquid chromatography and capillary electrophoresis (data furnished by the manufacturer).
Zymosan type A, PBS, EDTA sodium salt, fucoidin, phenolphthalein-β-glucuronic acid, glycine buffer, β-glucuronidase and all other chemicals were obtained from Sigma Chemical Co. (Poole, U.K.). mAbs, rat anti-murine P-selectin (clone 5H1) and E-selectin (clone 9A9) were produced at Hoffman-La Roche as described (Norton et al., 1993; Labow et al., 1994), and contained less than 1 endotoxin U/mg. An isotype-matched control rat IgG was also prepared at Hoffman-La Roche. The rat anti-murine CD11b mAb (clone C5.6) was purchased from Serotec (Oxford, UK). Tripolidine-HCl and diphenhydramine-HCl were purchased from Research Biomedical International (Natick, MA), whereas WEB2086 was a generous gift of Boheringer Ingleheim (Ingleheim am Rheim, Germany).
Statistics.
Statistical differences were calculated on original values using analysis of variance followed by Bonferroni test for intergroup comparisons (Berry and Lindgren, 1990), or by unpaired Student’s t test (two-tailed) when only two groups were compared. A threshold value of P < .05 was taken as significant.
Results
Characterization of MSU crystal-induced peritonitis.
Intraperitoneal injection of MSU crystals produced an intense PMN accumulation at the 6 hr time-point (fig.1A). The dose-response curve was relatively steep, with a maximal effect (ranging between 8 and 12 × 106 PMN per mouse, n = 32) at the 3-mg dose and a lower effect seen at the highest dose tested of 5 mg (probably due to a toxic action of the crystals). The dose of 3 mg was selected for the subsequent experiments.
Characterization of MSU crystal-induced PMN accumulation into murine peritoneal cavities. A, Dose-response. Mice received 0.5 ml i.p. of sterile PBS alone (dose 0 group) or supplemented with different doses of MSU crystals at time 0. PMN influx was quantified 6 hr later. Values are mean ± S.E.M. ofn = 6 mice per group. All doses of MSU produced a significant (P < .05) PMN accumulation when compared with the PBS group. B, Time-course. Mice received 0.5 ml i.p. of sterile PBS alone or supplemented with 3 mg MSU crystals at time 0. PMN accumulation into the peritoneal cavities was quantified at the reported time-points. Values are mean ± S.E.M. ofn = 5 to 9 mice per group. MSU crystals produced a significant (P < .05) PMN accumulation when compared with the PBS-treated group at all time-points.
MSU crystal-induced PMN accumulation into the murine peritoneal cavity was a time-dependent process, fully activated between 2 and 6 hr postchallenge (with an approximately influx rate of 1.6 × 106 PMN per hour in this time interval) (fig.1B). PMN recruitment plateaued between 4 and 24 hr, remaining very high at the latter time-point (∼ 10 × 106 PMN per cavity, n = 32). This process essentially subsided 48 hr after crystal injection (fig. 1B). The 6-hr time-point was selected for most of the subsequent experiments, although in some cases PMN numbers at 24 hr postinjection were also assessed.
Table 1 reports the time-dependency of plasma protein extravasation and β-glucuronidase release after MSU crystal injection. In the case of both parameters, maximal values were obtained at 2 hr post-MSU challenge with subsequent lower amounts being detected at the other time-points.
Protein content and β-glucuronidase activity in peritoneal lavage fluids after MSU crystal injection
Table 2 shows a comparison between the ability of MSU crystals, CPPD crystals and zymosan particles (all insoluble phlogogens) to elicit PMN accumulation into the peritoneal cavity. In contrast to MSU crystals, the response to CPPD crystals appeared to be delayed, such that a higher number of PMN was recovered at the 24-hr rather than the 6-hr time-point. Zymosan particles provoked the highest PMN influx at 6 hr, but the cell influx was greatly diminished at 24 hr postchallenge.
Comparison of the PMN recruitment into murine peritoneal cavities in response to distinct insoluble inflammatory stimuli
Role of endogenous mast cells in MSU crystal-induced inflammation.
A number ranging from 10 to 13 × 103 of intact mast cells were routinely recovered from untreated mice (n = 10). Local administration of compound 48/80 (10 μg) produced a marked depletion in intact mast cells as assessed 72 hr postinjection: only 0.8 × 103 cells could be counted (n = 6; P < .05). The effect of this depletion treatment on MSU crystal-induced leukocyte accumulation was then evaluated. A remarkable reduction (58%) in the 6-hr PMN influx was observed when MSU crystals were injected into mice pretreated with compound 48/80 compared to PBS pretreated animals (fig. 2A). This indicates that resident mast cells are playing an important role in the cellular response activated by these crystals. To substantiate this proposition, the effect of MSU crystals on the number of intact mast cells recovered from the peritoneal cavities was quantified. Figure 2B shows that this cell type is relatively sensitive to manipulation such that a reduction in number was also seen after i.p. injection of 250 μl PBS. However, 3 mg of MSU crystals substantially diminished (>95% reduction) the number of intact mast cells found at 2 hr postinjection. As expected, values remained low for the entire period under observation (up to 24 hr) (fig. 2B).
A role for endogenous mast cells in MSU crystal-induced PMN recruitment in vivo. A, Depletion experiments. Mice received 10 μg i.p. of compound 48/80 in 0.5 ml PBS, or PBS alone, 72 hr before i.p. challenge with PBS (0.5 ml) or MSU crystals (3 mg). PMN accumulation was quantified 6 hr later. Values are mean ± S.E.M. of n = 6 mice per group.*P < .05 vs. MSU group in PBS-pretreated animals. B, Effect on mast cell numbers in vivo. Mice received 0.5 ml i.p. of sterile PBS alone or supplemented with 3 mg MSU crystals at time 0. The number of intact mast cells recovered in the lavage fluids at different times post-challenge is shown. Values are mean ± S.E.M. ofn = 4 mice per group. MSU crystals produced a significant (P < .05) reduction in intact mast cell numbers when compared with the PBS-treated group at all time-points.
The pronounced effect of MSU crystal was not the result of a nonspecific action. For instance, zymosan administration reduced by ∼ 60% the number of intact mast cells at the 4-hr time-point (4.6 × 103 mast cells per cavity,n = 6).
The possibility that the MSU crystals could directly activated peritoneal mast cells was investigated in a set of in vitroexperiments. Coincubation of peritoneal cells with 1 mg/ml MSU crystals produced a marked reduction in the number of intact mast cells already at 30 min postchallenge (from 11.30 ± 0.72 × 103 to 3.50 ± 0.25 × 103 cells per aliquot; 69% reduction,n = 4 experiments). Similar lower counts were seen at 60 and 120 min: 3.60 ± 0.51 × 103 and 3.88 ± 0.27 × 103 cells per aliquot, respectively. Co-incubation of cells with CPPD crystals (1 mg/ml) produced a lesser degree of mast cell degranulation, such that 9.3 ± 1.1 × 103 intact mast cells were counted at the 30-min time-point (18% reduction, n = 3 experiments; not significant). A slightly more potent effect was seen at 60 and 120 min, with 7.7 ± 1.0 and 7.5 ± 0.6 × 103 intact mast cells, respectively (n = 3). Addition of the medium alone did not modify mast cell counts to any extent within the 120 min under observation, thus indicating also that the culture conditions did not provoke a nonspecific damage to this cell type.
Role of endogenous histamine and PAF in MSU crystal-induced inflammation.
Treatment of mice with the PAF antagonist, WEB2086, reduced significantly the 6 hr PMN accumulation observed in response to MSU crystal challenge by more than 50% (fig.3). A similar effect was achieved by pretreating mice with the H1 antagonist, tripolidine (47% reduction, n = 17, P < .05). Interestingly, a combined treatment with both antagonists did not produce an additive effect, such that a 67% of inhibition of PMN influx was then measured (n = 6, not significantvs. either single treatment) (fig. 3). The involvement of H1 receptors in the MSU response was validated further using diphenhydramine: 10.3 ± 0.20 × 106 PMN were recovered 6 hr after MSU crystal challenge in control mice, and this figure was reduced to 4.70 ± 0.31 × 106 cells in animals pretreated with diphenhydramine (9 mg/kg i.p.) (54% reduction, n = 6; P < .05).
Endogenous mediators involved in MSU crystal-induced PMN recruitment in vivo. Animals were treated with the PAF antagonist, WEB2086 (10 mg/kg i.p.), or with the H1 antagonist, tripolidine (0.5 mg/kg i.p.), or with a combination of both drugs (given i.p.) simultaneously with the challenge with MSU crystals (3 mg in 0.5 ml PBS i.p.). PMN accumulation into the peritoneal cavities was assessed 6 hr later. Values are mean ± S.E.M. of (n) mice per group. Dashed line indicates the migration seen in the absence of MSU crystal challenge.*P < .05 vs. control group.
Role of endothelial selectins in MSU crystal-induced inflammation.
Administration of fucoidin (0.3 mg per mouse i.v., corresponding to ∼10 mg/kg, -2 hr) significantly reduced MSU crystal-induced PMN accumulation (mean ± S.E.M.): 8.6 ± 0.2 × 106 PMN per mouse in saline-pretreated mice (n = 6) and 3.3 ± 0.3 × 106 PMN in fucoidin-pretreated animals (62% reduction; n = 10; P < .05).
The role of endothelial selectins was then investigated. Intravenous injection of anti-CD62P mAb alone significantly attenuated by 35% the 6-hr cell influx (n = 12, P < .05) (fig.4A). A similar degree of inhibition was also seen in the group of mice treated with the anti-CD62E mAb (45% reduction, n = 10; P < .05), whereas almost a 70% reduction in PMN accumulation was measured after a combined treatment with both mAbs (n = 13; P < .05vs. control IgG, and P < .05 vs. anti-CD62P mAb alone) (fig. 4A).
Role of endothelial selectins in MSU crystal-induced PMN recruitment in vivo. Control rat IgG (100 μg in 100 μl), rat anti-mouse CD62P mAb (CD62P; 100 μg), rat anti-mouse CD62E mAb (CD62E; 100 μg) or a combination of anti-CD62P and CD62E mAbs (CD62P/E; 100 μg of each) was given i.v. 1 hr before i.p. administration of MSU crystals (3 mg in 0.5 ml sterile PBS). PMN migration was evaluated either 6 hr (A) or 24 hr (B) after administration of the crystals. Values are mean ± S.E.M. of (n) mice per group. *P < .05vs. control rat IgG group.
A different scenario was obtained when PMN influx was investigated at a later time-point. If the anti-CD62P mAb, or the anti-CD62E mAb, were administered alone 6 hr after challenge with the MSU crystal, no inhibition in the number of cells recovered at the 24-hr time-point was seen (fig. 4B). In contrast, a combined treatment with both mAbs was still effective (60% reduction, n = 4; P < .05) (fig. 4B).
Effect of anti-CD11b mAb and peptide Ac2–26 on MSU crystal-induced inflammation.
A potent inhibition of PMN accumulation (-60%,n = 6; P < .05) was seen after s.c. administration of the lipocortin 1 pharmacophore, peptide Ac2–26 (fig.5). A similar degree of inhibition was also measured after treatment of mice with the anti-CD11b mAb (-68%,n = 10; P < .05) .
Modulation of MSU crystal-induced PMN recruitmentin vivo by anti-CD11b mAb or peptide Ac2–26. Mice received 200 μg of peptide Ac2–26 s.c., or 100 μl PBS (n = 6 in both cases), 30 min before i.p. challenge with 3 mg MSU crystals (in 0.5 ml sterile PBS). In a separate set of experiments, animals were treated with control rat IgG (n = 17) or rat anti-murine CD11b mAb (250 μg i.v.; n = 10) 1 hr before i.p. challenge with 3 mg MSU crystals (in 0.5 ml sterile PBS). In all cases PMN elicitation was quantified at the 6 hr time-point. Values are mean ± S.E.M.*P < .05 vs. appropriate control group.
Discussion
Gouty arthritis is a neutrophil driven disease triggered by deposition of MSU crystals in the joint space (McCarthy et al., 1962; Brandt and Schumaker, 1995; Terkeltaub, 1992). The ability to reproduce experimentally the marked cellular influx in response to MSU crystals represents a convenient way to investigate the molecular mechanisms underlying this specific pathology. Previous studies have reported that the rat pleural cavity (Sedgwick et al., 1985) or a preformed rat air-pouch (Brooks et al., 1987) provide convenient tissue sites to observe PMN influx in response to challenge with insoluble crystals. Here, we report that an i.p. injection of MSU crystals causes a remarkable peritoneal accumulation of murine PMN. Importantly, the cellular infiltration into the cavity was sustained up to 24 hr, and this is different from other models of acute inflammation in which insoluble particles are injected (i.e., zymosan peritonitis). Another validation of this experimental system came from the inability of CPPD crystals to produce a rapid PMN influx similar to that obtained after challenge with MSU. All these characteristics prompted us to begin an investigation into the molecular mechanisms responsible for the intense and sustained accumulation of PMN elicited by MSU crystals in our murine model.
PMN accumulation in response to MSU crystal injection was clearly dissociated from plasma protein extravasation and β-glucuronidase release, both peaking at 2 hr post-challenge. It therefore means that the enzymatic activity is mainly derived from the resident macrophage (Adams and Hamilton, 1992), rather than from the infiltrating PMN (Iwamura et al., 1993).
It is well known that resident mast cells are in close proximity to the site of leukocyte extravasation (i.e., the postcapillary venule) (Granger and Kubes, 1994), and recent studies have shown that this cell type plays a central role in the initiation of the PMN recruitment process (reviewed in Kubes and Granger, 1996). Mast cells may release an array of mediators (histamine, PAF, pluripotent cytokines, etc.) which are likely to activate the endothelium to attract the circulating leukocyte, cause arrest and direct it to the site of inflammation. We showed that a drastic reduction of resident intact mast cell numbers in the mouse peritoneal cavity, achieved by administration of compound 48/80 according to a well-established protocol (Diaz et al., 1996), produced a marked suppression of PMN accumulation in response to MSU crystal challenge. The ability of these crystals to activate mouse mast cells was confirmed by the dramatic suppression (>90%) of intact mast cell counts followingin vivo administration and, mainly, by the information obtained with in vitro experiments. With the latter protocol we could detect that MSU crystals produced a maximal mast cell degranulation within 30 min of addition.
We then sought to identify, at least in part, which mast cell mediators could contribute to the PMN influx generated by crystal injection. Mast cell-derived histamine has been shown to induce leukocyte rolling on the endothelium of postcapillary venules via up-regulation of P-selectin on the endothelial cell surface (Asako et al., 1994). The same study showed that this was effected by activation of H1 receptors. In our study, when mice were pretreated with the potent and specific H1antagonist tripolidine, a remarkable attenuation of MSU-induced PMN accumulation into the peritoneal cavity was observed. Similar data have also been obtained with a chemically unrelated H1antagonist, diphenhydramine. These observations clearly point to mast cell-derived histamine as to a mediator involved in the cellular response to MSU crystals.
A more recent study has described mast cell-derived PAF as a mediator able to induce leukocyte adhesion to the post-capillary endothelium (Gaboury et al., 1996). PAF action is mediated by β2-integrins, because an anti-CD18 mAb suppresses PAF-induced cell adhesion (Kubes et al., 1990). The effect of PAF antagonist, WEB2086 (Casals-Stenzel et al., 1987), was tested in this study, finding again a significant inhibition of PMN recruitment activated by MSU crystals. Importantly, a combined treatment with tripolidine and WEB2086 did not produce an additive effect. Because the rolling phenomenon is a prerequisite to subsequent firm adhesion (Lawrence and Springer, 1991), these data confirm the suggestion that, in the inflammatory response activated by MSU crystals, mast cell-derived histamine and PAF are mediating leukocyte rolling and adhesion, respectively, in a sequential manner.
The ability of fucoidin to inhibit PMN accumulation caused by MSU crystals indicated that endogenous selectins were involved in the process of cell recruitment (Lindbom et al., 1992). In murine systems a redundancy of endothelial-derived selectins has been found, such that both adhesion molecules need to be blocked, with specific mAbs, to inhibit PMN extravasation in experimental inflammation (reviewed in Ley and Tedder, 1995). P-selectin (CD62P), but not E-selectin (CD62E), blockade alone is sufficient per se only in the initial phases of the inflammatory response activated by thioglycollate (4 hr time-point) (Labow et al., 1994) or by the chemokine KC (2 hr time-point) (Harris et al., 1996), when PMN recruitment is really modest. It was therefore somehow surprising to observe that the intense PMN accumulation seen 6 hr post-MSU crystal challenge could be inhibited either with the anti-CD62P mAb or with the anti-CD62E mAb. This indicates not only that both selectins are expressed after challenge with these crystals, but also that they are acting through different pathways (or by interacting with distinct ligands). Indeed, a combined treatment with both mAbs produced essentially an additive inhibitory effect. A different scenario was seen when PMN recruitment at times later than 6 hr was investigated. In this case the phenomenon of redundancy was observed and only a combined treatment with both mAbs was able to suppress PMN influx into the murine peritoneal cavities (probably they are now interacting with an identical ligand). It is of interest that the degree of inhibition attained (60%) was similar to that observed with the combined treatment at the 6 hr time-point (∼70%). The fact that high PMN counts are found at 24 hr post-MSU crystal challenge, and that this is inhibited with antiselectin mAbs, suggests that active cell recruitment is occurring in the 6- to 24-hr time period. These data are in agreement with a recent study conducted in the pig skin, in which CD62E induction within the 4- to 24-hr time interval after intradermal administration of MSU crystals was demonstrated (Chapman et al., 1996).
We have previously compared the ability of an anti-CD11b mAb to inhibit PMN recruitment in murine models of experimental inflammation with that of the lipocortin 1 pharmacophore, the N-terminus derived peptide Ac2–26 (Perretti et al., 1993). Both agents were also tested in our study, observing again a similar degree of inhibition (60–70%). The result of the anti-CD11b mAb experiment adds MSU crystals to the list of inflammatory stimuli that cause mouse PMN recruitment via the β2-integrin CD11b/CD18. This list includes zymosan, C5a, interleukin-1 and CXC chemokines specific for PMN (Perretti et al., 1993; Harriset al., 1995). Similarly, the efficacy of the lipocortin 1 mimetic peptide Ac2–26 may suggest, at least theoretically, a potential therapeutic application for drugs arising from this line of research. Gout is a disease characterized by massive neutrophil accumulation into the joint space, and the data generated in this study may suggest that this pathology could be susceptible to systemic administration with lipocortin 1 mimetics. Modulation of endogenous lipocortin 1 levels in circulating leukocytes by antiinflammatory glucocorticoid hormones (Perretti and Flower, 1996) may account, at least in part, for their therapeutic application in acute gout (Brandt and Schumaker, 1995).
In conclusion, we describe a novel murine model of MSU crystal-induced inflammation. Taking advantage of 1) the simplicity of the model, 2) the intense and sustained PMN accumulation observed and 3) the availability of specific tools, we have attempted to shed light on mechanism(s) which may be operating in the pathology of gout.
Footnotes
-
Send reprint requests to: Dr. Mauro Perretti, Department of Biochemical Pharmacology, The William Harvey Research Institute, Charterhouse Square, London EC1 M 6BQ, United Kingdom.
-
↵1 This work was supported by an endowment to the William Harvey Research Institute by the Ono Pharmaceutical Co. (Osaka, Japan). R.J.F. is a Principal Research Fellow of the Wellcome Trust. L.P. is supported by grants from Consiglio Nazionale delle Ricerche nos. 95.02383.CT04 and 96.03339.CT04. M.A.M. is recipient of a CNPq fellowship (Brasil).
- Abbreviations:
- CPPD
- calcium pyrophosphate dihydrate
- EDTA
- ethylenediaminetetraacetic sodium salt
- mAb
- monoclonal antibody
- MSU
- monosodium urate
- peptide Ac2–26
- lipocortin 1-derived N-terminus peptide
- PBS
- phosphate-buffered saline
- PAF
- platelet-activating factor
- PMN
- polymorphonuclear leukocyte
-
- Received February 24, 1997.
- Accepted June 2, 1997.
- The American Society for Pharmacology and Experimental Therapeutics
