367
Pharmacological Research, Vol. 23, No. 4, 1991
EFFECTS OF TENOXICAM ON SUPEROXIDE ANION FORMATION, ß-GLUCURONIDASE RELEASE AND fMLP BINDING IN HUMAN NEUTROPHILS : COMPARISON WITH OTHER NSAIDs S . COLLI, S . COLOMBO, E . TREMOLI, E. STRAGLIOTTO and S . NICOSIA Institute of Pharmacological Sciences and E . Grossi Paoletti Center, University of Milan, via Balzaretti 9, 20133 Milan, Italy Received in final form 25 September .1990
SUMMARY Non-steroidal anti-inflammatory drugs (NSAIDs) are considered to exert their activity by interfering with the generation of arachidonate metabolites in various cells, mainly in neutrophils and monocytes . The inhibition of cellular cyclooxygenase enzyme, however, does not always correlate with the in vivo activity of these drugs . Recent evidence indicates that several NSAIDs may interfere with the stimulus-response coupling of inflammatory cells . In this study, the effects of tenoxicam, an oxicam derivative with a thienothiazine structure, on neutrophil activation were evaluated by the assessment of the following parameters : (1) superoxide anion generation by neutrophils and whole blood stimulated with N-formyl-methionyl-leucyl-phenylalanine (fMLP), the calcium ionophore A23187 and serum treated zymosan (STZ) ; (2) ß-glucuronidase release from neutrophils stimulated with fMLP, A23187 and STZ ; (3) binding of [ 3H]fMLP to intact neutrophils . The results were compared to those obtained using piroxicam and diclofenac . Tenoxicam, added in vitro to whole blood, at concentrations ranging between 10 -5 and 3 X 10 -4 M, significantly inhibited the generation of superoxide anion induced by fMLP, A23187 and STZ . The activity of tenoxicam on whole blood was similar to that of piroxicam, whereas diclofenac had only minimal effects on this experimental system. In isolated cells tenoxicam inhibited the generation of superoxide anion induced by A23187 and STZ . In addition, at the 3 x 10 -4 M concentration, tenoxicam and diclofenac similarly inhibited O Z - generation by neutrophils stimulated with fMLP, whereas piroxicam only minimally affected this parameter . Tenoxicam also slightly, but not significantly, inhibited ß-glucuronidase release by isolated neutrophils induced by all the agonists used . Specific binding of [3H]fMLP to neutrophils was inhibited by the three NSAIDs tested in a dose-dependent fashion and tenoxicam was the most potent . The Correspondence to : S. Colli .
1043-6618/91/040367-13/$03 .00/0
0 1991 The Italian Pharmacological Society
368
Pharmacological Research, Vol. 23, No. 4, 1991
affinities (Kd) of tenoxicam, piroxicam and diclofenac were 1 .11, 1 .80 and 2.70 X 10 -5 M, respectively. The mechanism of inhibition of [ 3 H]fMLP binding by tenoxicam was non-competitive . It is concluded that tenoxicam, at concentrations achievable in plasma at steady state, effectively inhibits some of the processes involved in neutrophil activation, which bear some relevance in the inflammatory disease . KEY WORDS : NSAIDs, neutrophils, fMLP, superoxide anion. INTRODUCTION
Neutrophil activation is known to play a major role in the inflammatory reaction, as a consequence of the release of cytotoxic species [1, 2], which, in turn, may injure surrounding tissues . Anti-inflammatory agents, and in particular non-steroidal antiinflammatory drugs (NSAIDs) have been considered to exert their activity through inhibition of arachidonate metabolism in inflammatory cells [3] . Their antiinflammatory effect, however, does not always correlate with their potency in interfering with the synthesis of eicosanoids [4]. On the other hand, it has been reported that several NSAIDs affect various responses of neutrophils, including the release of lysosomal enzymes and the generation of oxygenated species [5, 6] . Chemotactic agents initiate various cellular responses in neutrophils, including shape change, chemotaxis, enzyme secretion and synthesis of oxygenated species . In particular, chemoattractants, such as N-formyl-methionyl-leucyl-phenylalanine (fMLP), a synthetic oligopeptide resembling the chemotactic factors produced by bacteria, have been shown to recognize specific binding sites on neutrophils [7] . Interestingly, the affinity of such binding sites for a series of formyl peptides parallels their potency as chemotactic agents [8]. In turn, the chemotactic activity strictly correlates with the ability to activate one of the neutrophil functions involved in the inflammatory process, that is lysosomal enzyme release [9] . On the basis of these considerations, it appears of interest to evaluate the effects of NSAIDs on various aspects of neutrophil activation, including the stimulusresponse coupling. In particular, in this study we investigated the effects of tenoxicam, a newly developed oxicam derivative with a thienothiazine structure [10-12] on : (1) OZ - generation by human polymorphonuclear leucocytes (PMN) and whole blood; (2) ß-glucuronidase release by polymorphonuclear leucocytes ; (3) [3H]fMLP binding to intact polymorphonuclear leucocytes . The effects of tenoxicam have been compared with those of the related compound piroxicam and with the structurally unrelated NSAID, diclofenac . The latter NSAIDs have been used as standard comparative drugs because their capacity in interfering with various aspects of neutrophil activation in vitro has already been described [13-15] .
MATERIALS AND METHODS Materials
Tenoxicam was kindly supplied by Prodotti Roche (Milano, Italy) . Piroxicam, diclofenac, fMLP, superoxide dismutase (SOD), cytochalasin B, zymosan,
Pharmacological Research, Vol. 23, No . 4, 1991
369
cytochrome c (type III), bovine serum albumin (BSA) were from Sigma Chemical Co. (St. Louis, MO, USA); calcium ionophore A23187 was from Calbiochem, Behring Corp . (San Diego, CA, USA) . fMLP was prepared as stock solution in dimethylsulphoxide and stored at - 20°C . Cytochalasin B and calcium ionophore were dissolved in ethanol and kept at - 20°C . The radiolabelled peptide, N-formylmethionyl-leucyl-[ 3H]phenylalanine ([ 3H]fvlLP), with a specific activity of 56 .9 Ci/ mmol was purchased from New England Nuclear (NEN Dupont Co., Firenze). Dextran T500 and Ficoll-Paque were obtained from Pharmacia (Uppsala, Sweden) ; phosphate buffered saline (PBS) from Flow Labs (UK); phenolphthalein #-nglucuronide from Serva (Heidelberg, FRG) . Preparation of serum treated zymosan (STZ )
Zymosan (10 mg/ml) was boiled in saline for 30 min, centrifuged and washed twice in saline . Opsonization was carried out by adding washed zymosan to fresh human serum (10 mg/ml) . The suspension was then incubated for 30 min at 37°C with stirring (300 r.p .m.) . After centrifugation, the supernatant was discarded, the pellet resuspended in saline, divided into aliquots and kept at - 80°C until use . Polymorphonuclear leucocyte (PMN) isolation
Venous blood was collected from healthy volunteers and anticoagulated with 3 .8% sodium citrate (9 :1 v/v) for cell isolation, or with heparin (2 U/ml of blood) for studies with whole blood . Purified preparations of PMN were obtained according to Boyum [ 16] . Briefly, whole blood was centrifuged at 100 g for 18 min to obtain platelet rich plasma which was discarded . The residue was processed for PMN isolation using Dextran sedimentation of red cells followed by centrifugation of the PMN enriched suspension of Ficoll-Paque . The obtained pellet, containing erythrocytes and PMN was subjected to hypotonic lysis and purified cells were resuspended, at the desired concentration, in PBS without calcium and magnesium containing BSA (0 .25%) and glucose (0 .05%). Cell suspension contained 97% PMN . Viability, determined by the exclusion of trypan blue, was 98% . Cell counts were performed by phase contrast microscopy . Superoxide anion (O2 - ) generation
This parameter was monitored by the SOD-inhibitable reduction of ferricytochrome c [17] in the presence (fMLP) and in the absence (calcium ionophore A23187, STZ) of cytochalasin B (5,ug/ml) . Tenoxicam, diclofenac and piroxicam were dissolved in DMSO at 3 X 10- ' M. Appropriate dilutions were performed in DMSO . Samples, including blank samples, contained less than 0 .1% DMSO. Samples (1 ml total volume) containing PMN (3 X 105) or whole blood (50 µl), 1 mg/ml ferricytochrome c, cytochalasin B and different drugs were preincubated for 10 min at 37°C before the addition of the stimuli . To samples containing cell suspensions, a solution of divalent cations (1 mm Ca 2+/Mg2 +) was added. Following the addition of the various stimuli, the incubation was carried out for 20 min and then stopped by centrifugation (8000 g) at 4°C for 3 min . Supernatants were collected, absorption determined spectrophotometrically at
3 70
PharmacologicalResearch, Vol. 23, No. 4, 1991
550 nm and quantification performed using an extinction coefficient of 24 500 m - '- cm - ' . Appropriate controls containing SOD (30 µg/ml), prepared for each sample, verified that 02- formation was responsible for cytochrome c reduction . Results were expressed as nmol 0 2 - / 106 PMN or leucocytes released in 20 min . Release of ß-glucuronidase
Samples (1 ml) containing 1 X 10 6 PMN were incubated in the presence or in the absence of increasing concentrations of tenoxicam, diclofenac or piroxicam at 37°C for 15 min with 10 -6 M fMLP in the presence of 5 µg/ml cytochalasin B or 1 mg/ ml STZ . At the end of the incubation period after centrifugation at 8000 g for 3 min the supernatants were collected . To determine the release of ß-glucuronidase, a modification of the method of Brittinger et al. [18] was used . Briefly, the incubation mixture containing 0 .2 ml of 0 .01 M phenolphthalein mono-ßglucuronide in 10% ethanol, 0 .1 ml sodium acetate buffer 0 .2 M pH 4 .5, 1 .5 ml glycine 0.2 M pH 10 .5 and 0 .2 ml of supernatant was incubated at 37°C for 18 h . Absorbance was read at 550 nm and compared with the standard curve for phenolphthalein absorption. Results were expressed in ,ug of phenolphthalein released in 15 min x 106 PMN . [3H]fMLP binding assay
All binding studies were carried out, in polystyrene tubes (final volume of 250,ul), at 37°C in PBS supplemented with 1 mm Cal +/Mg 2 +, 0.1% BSA, 0 .02% glucose. Tenoxicam, piroxicam and diclofenac were dissolved in DMSO at 200 mm and the serial dilutions were performed in PBS containing 3% DMSO ; therefore, all the samples contained DMSO at the final concentration of 0 .3% . The concentration of radioligand was 20 rim . Non-specific binding was defined as amount of [ 3H]fMLP bound in the presence of 10 µM unlabelled fMLP, and was approximately 20-25% of total binding . The incubation was started by adding 100 µl of PMN (2 .5-3 X 10' cells/ml) . After 20 min, the incubation was stopped by rapid filtration through Whatman GF/C filters, followed by two 3-ml washes with ice-cold saline buffered with Tris-HCl 1 mm pH 7 .4 . Each determination was done in triplicate . The radioactivity was extracted from filters with 10 ml Filtercount (Packard, Milan, Italy) and then counted in a liquid scintillation counter . Computer analysis of binding data
The data were fitted by an iterative program for non-linear regression analysis (LIGAND) [19] in order to obtain Kd (dissociation constant for the radiolabelled ligand) and K, (dissociation constant for unlabelled ligands). The data were fitted both to a one- and a two-site model . The two-site model was discarded when the improvement in goodness-of-fit was not statistically significant (P> 0 .05) according to the F test used on the `extra sum of squares' principle . In order to evaluate the statistical significance of the difference of the I,s, the curves for different ligands were analysed simultaneously, both without any constraint on the parameter values and imposing the identity of the K ;s . The I,s were considered not significantly different when the constrained fitting was not significantly worse (P> 0 .05) .
371
Pharmacological Research, Vol. 23, No. 4, 1991
Statistical analysis
The significance of differences between control and treated samples was assessed by Student t-test (paired) and by ANOVA one way followed by Dunnett test .
Table I Effect of tenoxicam on superoxide anion (nmol/10 6 cells) generation by PMN Tenoxicam (M)
0 10 -, 10 -7 10 -6 10 -5 10 -4 3 x 10 -4 *P< 0 .05 ; **P< 0 .02
fMLP (10- ' M)
A23187 (5X10 -6 M)
STZ (1 mg/ml)
51 .8±7 .2 53 .4±8 .1 53 .9±8 .8 53 .9±7 .2 52 .6±6 .2 53 .4 ± 6 .9 26 .9 ± 6 .9*
30 .1±7 ..4 28 .7±7 .7 29 .1±7 .6 27 .8±7 .9 28 .0±8 .8 21 .8 ± 6 .0** 16.3±4 .7**
52 .7±2 .8 56 .3±5 .3 53 .3±3 .3 55 .5±5 .5 55 .4±6 .0 46 .9±3 .4* 35 .5 ± 6 .7*
versus 0 ; values are the mean
± SEM of 5
separate experiments, each
performed in triplicate .
RESULTS Tenoxicam significantly inhibited, at the highest concentrations used, the generation of superoxide anion by PMN stimulated with fMLP, A23187 and opsonized zymosan (Table I) . Figure 1 shows the comparison between the effects of tenoxicam and diclofenac on this parameter. Tenoxicam was more active than diclofenac in inhibiting 02- generation by PMN stimulated with STZ (Fig . 1, panel c), whereas the two compounds displayed a similar effect on 02- generation stimulated with fMLP or A23187 (Fig . 1, panels a and b) . When the effects of tenoxicam (3 x 10 -4 M) were compared to those of an equimolar concentration of piroxicam, it appeared that tenoxicam was significantly more active than piroxicam in inhibiting 02- generation by PMN stimulated with fMLP, while the difference was not statistically significant in the case of PMN stimulated with STZ (Fig . 2) . To assess the effects of the three NSAIDs on leucocytes in a more physiological milieu, studies on superoxide anion generation by whole blood were also carried out. As shown in Table II, tenoxicam inhibited O 2 - formation in whole blood in a fashion similar to that observed with isolated cells . It should be noted, however, that the drug appeared to be more potent in whole blood with respect to isolated cells as far as fMLP is concerned . Indeed, significant reduction in 02- synthesis in whole blood was achieved at 10 - s M tenoxicam . Inhibition of 02- generation by whole blood was also recorded in the presence of diclofenac (Fig . 3) . Tenoxicam, however, was more potent than diclofenac in inhibiting 02- generation by leucocytes in whole blood, using all the three stimuli . At variance with the results obtained in isolated cells, tenoxicam and piroxicam exerted a similar effect on superoxide anion generation by whole blood (Fig . 4) .
372
Pharmacological Research, Vol. 23, No . 4, 1991
(a) 80z CL
Io 0 r . N
0
o
E c
(b)
fMLP (10-7M)
A23187 (5 x 10" e M
50 -
)
70 60 50 40 30 20 010
100
300
100
(c) STZ (1 mg/ml) 70-
-13 -A---
60-
Tenoxicam Diclofenac
504030200 , 10
100
300
X
10 -6
M
Fig . 1 . Comparison between the effects of tenoxicam and diclofenac on 0 2 - generation by PMN stimulated with fMLP, A23187 and STZ . Data represent the mean ± SEM of 5 separate experiments each performed in triplicate . °°P< 0 .01 versus diclofenac .
The effects of tenoxicam on the release of /3-glucuronidase, following incubation of neutrophils with either STZ or fMLP, were also investigated . As shown in Table III, tenoxicam, at 10 - 'm, only slightly affected the release of this lysosomal enzyme. Table IV compares the effects of the three NSAIDs at the 3 X 10 -4 M concentration on this parameter . Tenoxicam, as well as piroxicam and diclofenac, concentration dependently inhibited the specific binding of [ 3H]fMLP to intact PMN . Figure 5 shows the inhibition curves for the three agents . Although tenoxicam displayed a slightly higher affinity than piroxicam and diclofenac for the [3 H]fMLP binding sites, such difference did not reach statistical significance . In fact, computer assisted analysis of the curves with the LIGAND program did not show a statistically significant worsening of the fitting when the identity of the Ids of the two ligands was imposed .
373
Pharmacological Research, Vol. 23, No. 4, 1991
60-
C N U
0
Fig. 2 . Effects of tenoxicam (Ten) and piroxicam (Pir), 3 X 10 - a M, on 0 Z - generation by PMN stimulated with fMLP (10 -7 M) and STZ (1 mg/ml) . Data represent the mean ±sEM of 5 separate experiments, each performed in triplicate . *P< 0 .01 versus Pir.
Table II Effect of tenoxicam on superoxide anion (nmol/10 6 leucocytes) generation by whole blood Tenoxicam (M)
0 10 - " 10 -7 10 -6 5 10_ 10-4 3 x 1()-4
fMLP (10-6 M)
A23187 (2X10 -5 M)
STZ (1 mg/ml)
16 .8±1 .9 17 .0±1 .7 14.7±0 .6 15 .0±0 .8 15 .0±2 .0* 8 .8 ± 1 .4* 8 .8 ± 1 .1**
16 .7±0 .7 15 .5±0 .7 16 .0±0 .9 14 .6±1 .1 14 .7±1 .2 8 .8 ± 1 .4*** 8.0 ± 1 .0***
15 .4±1 .3 11 .7±1 .7 13 .1 ± 1 .4 11 .2±2 .3 10 .5±1 .8 9 .9 ± 1 .9** 9 .4 ± 1 .9**
*P
Pharmacological Research, Vol. 23, No. 4, 1991
EFFECTS OF TENOXICAM ON SUPEROXIDE ANION FORMATION, ß-GLUCURONIDASE RELEASE AND fMLP BINDING IN HUMAN NEUTROPHILS : COMPARISON WITH OTHER NSAIDs S . COLLI, S . COLOMBO, E . TREMOLI, E. STRAGLIOTTO and S . NICOSIA Institute of Pharmacological Sciences and E . Grossi Paoletti Center, University of Milan, via Balzaretti 9, 20133 Milan, Italy Received in final form 25 September .1990
SUMMARY Non-steroidal anti-inflammatory drugs (NSAIDs) are considered to exert their activity by interfering with the generation of arachidonate metabolites in various cells, mainly in neutrophils and monocytes . The inhibition of cellular cyclooxygenase enzyme, however, does not always correlate with the in vivo activity of these drugs . Recent evidence indicates that several NSAIDs may interfere with the stimulus-response coupling of inflammatory cells . In this study, the effects of tenoxicam, an oxicam derivative with a thienothiazine structure, on neutrophil activation were evaluated by the assessment of the following parameters : (1) superoxide anion generation by neutrophils and whole blood stimulated with N-formyl-methionyl-leucyl-phenylalanine (fMLP), the calcium ionophore A23187 and serum treated zymosan (STZ) ; (2) ß-glucuronidase release from neutrophils stimulated with fMLP, A23187 and STZ ; (3) binding of [ 3H]fMLP to intact neutrophils . The results were compared to those obtained using piroxicam and diclofenac . Tenoxicam, added in vitro to whole blood, at concentrations ranging between 10 -5 and 3 X 10 -4 M, significantly inhibited the generation of superoxide anion induced by fMLP, A23187 and STZ . The activity of tenoxicam on whole blood was similar to that of piroxicam, whereas diclofenac had only minimal effects on this experimental system. In isolated cells tenoxicam inhibited the generation of superoxide anion induced by A23187 and STZ . In addition, at the 3 x 10 -4 M concentration, tenoxicam and diclofenac similarly inhibited O Z - generation by neutrophils stimulated with fMLP, whereas piroxicam only minimally affected this parameter . Tenoxicam also slightly, but not significantly, inhibited ß-glucuronidase release by isolated neutrophils induced by all the agonists used . Specific binding of [3H]fMLP to neutrophils was inhibited by the three NSAIDs tested in a dose-dependent fashion and tenoxicam was the most potent . The Correspondence to : S. Colli .
1043-6618/91/040367-13/$03 .00/0
0 1991 The Italian Pharmacological Society
368
Pharmacological Research, Vol. 23, No. 4, 1991
affinities (Kd) of tenoxicam, piroxicam and diclofenac were 1 .11, 1 .80 and 2.70 X 10 -5 M, respectively. The mechanism of inhibition of [ 3 H]fMLP binding by tenoxicam was non-competitive . It is concluded that tenoxicam, at concentrations achievable in plasma at steady state, effectively inhibits some of the processes involved in neutrophil activation, which bear some relevance in the inflammatory disease . KEY WORDS : NSAIDs, neutrophils, fMLP, superoxide anion. INTRODUCTION
Neutrophil activation is known to play a major role in the inflammatory reaction, as a consequence of the release of cytotoxic species [1, 2], which, in turn, may injure surrounding tissues . Anti-inflammatory agents, and in particular non-steroidal antiinflammatory drugs (NSAIDs) have been considered to exert their activity through inhibition of arachidonate metabolism in inflammatory cells [3] . Their antiinflammatory effect, however, does not always correlate with their potency in interfering with the synthesis of eicosanoids [4]. On the other hand, it has been reported that several NSAIDs affect various responses of neutrophils, including the release of lysosomal enzymes and the generation of oxygenated species [5, 6] . Chemotactic agents initiate various cellular responses in neutrophils, including shape change, chemotaxis, enzyme secretion and synthesis of oxygenated species . In particular, chemoattractants, such as N-formyl-methionyl-leucyl-phenylalanine (fMLP), a synthetic oligopeptide resembling the chemotactic factors produced by bacteria, have been shown to recognize specific binding sites on neutrophils [7] . Interestingly, the affinity of such binding sites for a series of formyl peptides parallels their potency as chemotactic agents [8]. In turn, the chemotactic activity strictly correlates with the ability to activate one of the neutrophil functions involved in the inflammatory process, that is lysosomal enzyme release [9] . On the basis of these considerations, it appears of interest to evaluate the effects of NSAIDs on various aspects of neutrophil activation, including the stimulusresponse coupling. In particular, in this study we investigated the effects of tenoxicam, a newly developed oxicam derivative with a thienothiazine structure [10-12] on : (1) OZ - generation by human polymorphonuclear leucocytes (PMN) and whole blood; (2) ß-glucuronidase release by polymorphonuclear leucocytes ; (3) [3H]fMLP binding to intact polymorphonuclear leucocytes . The effects of tenoxicam have been compared with those of the related compound piroxicam and with the structurally unrelated NSAID, diclofenac . The latter NSAIDs have been used as standard comparative drugs because their capacity in interfering with various aspects of neutrophil activation in vitro has already been described [13-15] .
MATERIALS AND METHODS Materials
Tenoxicam was kindly supplied by Prodotti Roche (Milano, Italy) . Piroxicam, diclofenac, fMLP, superoxide dismutase (SOD), cytochalasin B, zymosan,
Pharmacological Research, Vol. 23, No . 4, 1991
369
cytochrome c (type III), bovine serum albumin (BSA) were from Sigma Chemical Co. (St. Louis, MO, USA); calcium ionophore A23187 was from Calbiochem, Behring Corp . (San Diego, CA, USA) . fMLP was prepared as stock solution in dimethylsulphoxide and stored at - 20°C . Cytochalasin B and calcium ionophore were dissolved in ethanol and kept at - 20°C . The radiolabelled peptide, N-formylmethionyl-leucyl-[ 3H]phenylalanine ([ 3H]fvlLP), with a specific activity of 56 .9 Ci/ mmol was purchased from New England Nuclear (NEN Dupont Co., Firenze). Dextran T500 and Ficoll-Paque were obtained from Pharmacia (Uppsala, Sweden) ; phosphate buffered saline (PBS) from Flow Labs (UK); phenolphthalein #-nglucuronide from Serva (Heidelberg, FRG) . Preparation of serum treated zymosan (STZ )
Zymosan (10 mg/ml) was boiled in saline for 30 min, centrifuged and washed twice in saline . Opsonization was carried out by adding washed zymosan to fresh human serum (10 mg/ml) . The suspension was then incubated for 30 min at 37°C with stirring (300 r.p .m.) . After centrifugation, the supernatant was discarded, the pellet resuspended in saline, divided into aliquots and kept at - 80°C until use . Polymorphonuclear leucocyte (PMN) isolation
Venous blood was collected from healthy volunteers and anticoagulated with 3 .8% sodium citrate (9 :1 v/v) for cell isolation, or with heparin (2 U/ml of blood) for studies with whole blood . Purified preparations of PMN were obtained according to Boyum [ 16] . Briefly, whole blood was centrifuged at 100 g for 18 min to obtain platelet rich plasma which was discarded . The residue was processed for PMN isolation using Dextran sedimentation of red cells followed by centrifugation of the PMN enriched suspension of Ficoll-Paque . The obtained pellet, containing erythrocytes and PMN was subjected to hypotonic lysis and purified cells were resuspended, at the desired concentration, in PBS without calcium and magnesium containing BSA (0 .25%) and glucose (0 .05%). Cell suspension contained 97% PMN . Viability, determined by the exclusion of trypan blue, was 98% . Cell counts were performed by phase contrast microscopy . Superoxide anion (O2 - ) generation
This parameter was monitored by the SOD-inhibitable reduction of ferricytochrome c [17] in the presence (fMLP) and in the absence (calcium ionophore A23187, STZ) of cytochalasin B (5,ug/ml) . Tenoxicam, diclofenac and piroxicam were dissolved in DMSO at 3 X 10- ' M. Appropriate dilutions were performed in DMSO . Samples, including blank samples, contained less than 0 .1% DMSO. Samples (1 ml total volume) containing PMN (3 X 105) or whole blood (50 µl), 1 mg/ml ferricytochrome c, cytochalasin B and different drugs were preincubated for 10 min at 37°C before the addition of the stimuli . To samples containing cell suspensions, a solution of divalent cations (1 mm Ca 2+/Mg2 +) was added. Following the addition of the various stimuli, the incubation was carried out for 20 min and then stopped by centrifugation (8000 g) at 4°C for 3 min . Supernatants were collected, absorption determined spectrophotometrically at
3 70
PharmacologicalResearch, Vol. 23, No. 4, 1991
550 nm and quantification performed using an extinction coefficient of 24 500 m - '- cm - ' . Appropriate controls containing SOD (30 µg/ml), prepared for each sample, verified that 02- formation was responsible for cytochrome c reduction . Results were expressed as nmol 0 2 - / 106 PMN or leucocytes released in 20 min . Release of ß-glucuronidase
Samples (1 ml) containing 1 X 10 6 PMN were incubated in the presence or in the absence of increasing concentrations of tenoxicam, diclofenac or piroxicam at 37°C for 15 min with 10 -6 M fMLP in the presence of 5 µg/ml cytochalasin B or 1 mg/ ml STZ . At the end of the incubation period after centrifugation at 8000 g for 3 min the supernatants were collected . To determine the release of ß-glucuronidase, a modification of the method of Brittinger et al. [18] was used . Briefly, the incubation mixture containing 0 .2 ml of 0 .01 M phenolphthalein mono-ßglucuronide in 10% ethanol, 0 .1 ml sodium acetate buffer 0 .2 M pH 4 .5, 1 .5 ml glycine 0.2 M pH 10 .5 and 0 .2 ml of supernatant was incubated at 37°C for 18 h . Absorbance was read at 550 nm and compared with the standard curve for phenolphthalein absorption. Results were expressed in ,ug of phenolphthalein released in 15 min x 106 PMN . [3H]fMLP binding assay
All binding studies were carried out, in polystyrene tubes (final volume of 250,ul), at 37°C in PBS supplemented with 1 mm Cal +/Mg 2 +, 0.1% BSA, 0 .02% glucose. Tenoxicam, piroxicam and diclofenac were dissolved in DMSO at 200 mm and the serial dilutions were performed in PBS containing 3% DMSO ; therefore, all the samples contained DMSO at the final concentration of 0 .3% . The concentration of radioligand was 20 rim . Non-specific binding was defined as amount of [ 3H]fMLP bound in the presence of 10 µM unlabelled fMLP, and was approximately 20-25% of total binding . The incubation was started by adding 100 µl of PMN (2 .5-3 X 10' cells/ml) . After 20 min, the incubation was stopped by rapid filtration through Whatman GF/C filters, followed by two 3-ml washes with ice-cold saline buffered with Tris-HCl 1 mm pH 7 .4 . Each determination was done in triplicate . The radioactivity was extracted from filters with 10 ml Filtercount (Packard, Milan, Italy) and then counted in a liquid scintillation counter . Computer analysis of binding data
The data were fitted by an iterative program for non-linear regression analysis (LIGAND) [19] in order to obtain Kd (dissociation constant for the radiolabelled ligand) and K, (dissociation constant for unlabelled ligands). The data were fitted both to a one- and a two-site model . The two-site model was discarded when the improvement in goodness-of-fit was not statistically significant (P> 0 .05) according to the F test used on the `extra sum of squares' principle . In order to evaluate the statistical significance of the difference of the I,s, the curves for different ligands were analysed simultaneously, both without any constraint on the parameter values and imposing the identity of the K ;s . The I,s were considered not significantly different when the constrained fitting was not significantly worse (P> 0 .05) .
371
Pharmacological Research, Vol. 23, No. 4, 1991
Statistical analysis
The significance of differences between control and treated samples was assessed by Student t-test (paired) and by ANOVA one way followed by Dunnett test .
Table I Effect of tenoxicam on superoxide anion (nmol/10 6 cells) generation by PMN Tenoxicam (M)
0 10 -, 10 -7 10 -6 10 -5 10 -4 3 x 10 -4 *P< 0 .05 ; **P< 0 .02
fMLP (10- ' M)
A23187 (5X10 -6 M)
STZ (1 mg/ml)
51 .8±7 .2 53 .4±8 .1 53 .9±8 .8 53 .9±7 .2 52 .6±6 .2 53 .4 ± 6 .9 26 .9 ± 6 .9*
30 .1±7 ..4 28 .7±7 .7 29 .1±7 .6 27 .8±7 .9 28 .0±8 .8 21 .8 ± 6 .0** 16.3±4 .7**
52 .7±2 .8 56 .3±5 .3 53 .3±3 .3 55 .5±5 .5 55 .4±6 .0 46 .9±3 .4* 35 .5 ± 6 .7*
versus 0 ; values are the mean
± SEM of 5
separate experiments, each
performed in triplicate .
RESULTS Tenoxicam significantly inhibited, at the highest concentrations used, the generation of superoxide anion by PMN stimulated with fMLP, A23187 and opsonized zymosan (Table I) . Figure 1 shows the comparison between the effects of tenoxicam and diclofenac on this parameter. Tenoxicam was more active than diclofenac in inhibiting 02- generation by PMN stimulated with STZ (Fig . 1, panel c), whereas the two compounds displayed a similar effect on 02- generation stimulated with fMLP or A23187 (Fig . 1, panels a and b) . When the effects of tenoxicam (3 x 10 -4 M) were compared to those of an equimolar concentration of piroxicam, it appeared that tenoxicam was significantly more active than piroxicam in inhibiting 02- generation by PMN stimulated with fMLP, while the difference was not statistically significant in the case of PMN stimulated with STZ (Fig . 2) . To assess the effects of the three NSAIDs on leucocytes in a more physiological milieu, studies on superoxide anion generation by whole blood were also carried out. As shown in Table II, tenoxicam inhibited O 2 - formation in whole blood in a fashion similar to that observed with isolated cells . It should be noted, however, that the drug appeared to be more potent in whole blood with respect to isolated cells as far as fMLP is concerned . Indeed, significant reduction in 02- synthesis in whole blood was achieved at 10 - s M tenoxicam . Inhibition of 02- generation by whole blood was also recorded in the presence of diclofenac (Fig . 3) . Tenoxicam, however, was more potent than diclofenac in inhibiting 02- generation by leucocytes in whole blood, using all the three stimuli . At variance with the results obtained in isolated cells, tenoxicam and piroxicam exerted a similar effect on superoxide anion generation by whole blood (Fig . 4) .
372
Pharmacological Research, Vol. 23, No . 4, 1991
(a) 80z CL
Io 0 r . N
0
o
E c
(b)
fMLP (10-7M)
A23187 (5 x 10" e M
50 -
)
70 60 50 40 30 20 010
100
300
100
(c) STZ (1 mg/ml) 70-
-13 -A---
60-
Tenoxicam Diclofenac
504030200 , 10
100
300
X
10 -6
M
Fig . 1 . Comparison between the effects of tenoxicam and diclofenac on 0 2 - generation by PMN stimulated with fMLP, A23187 and STZ . Data represent the mean ± SEM of 5 separate experiments each performed in triplicate . °°P< 0 .01 versus diclofenac .
The effects of tenoxicam on the release of /3-glucuronidase, following incubation of neutrophils with either STZ or fMLP, were also investigated . As shown in Table III, tenoxicam, at 10 - 'm, only slightly affected the release of this lysosomal enzyme. Table IV compares the effects of the three NSAIDs at the 3 X 10 -4 M concentration on this parameter . Tenoxicam, as well as piroxicam and diclofenac, concentration dependently inhibited the specific binding of [ 3H]fMLP to intact PMN . Figure 5 shows the inhibition curves for the three agents . Although tenoxicam displayed a slightly higher affinity than piroxicam and diclofenac for the [3 H]fMLP binding sites, such difference did not reach statistical significance . In fact, computer assisted analysis of the curves with the LIGAND program did not show a statistically significant worsening of the fitting when the identity of the Ids of the two ligands was imposed .
373
Pharmacological Research, Vol. 23, No. 4, 1991
60-
C N U
0
Fig. 2 . Effects of tenoxicam (Ten) and piroxicam (Pir), 3 X 10 - a M, on 0 Z - generation by PMN stimulated with fMLP (10 -7 M) and STZ (1 mg/ml) . Data represent the mean ±sEM of 5 separate experiments, each performed in triplicate . *P< 0 .01 versus Pir.
Table II Effect of tenoxicam on superoxide anion (nmol/10 6 leucocytes) generation by whole blood Tenoxicam (M)
0 10 - " 10 -7 10 -6 5 10_ 10-4 3 x 1()-4
fMLP (10-6 M)
A23187 (2X10 -5 M)
STZ (1 mg/ml)
16 .8±1 .9 17 .0±1 .7 14.7±0 .6 15 .0±0 .8 15 .0±2 .0* 8 .8 ± 1 .4* 8 .8 ± 1 .1**
16 .7±0 .7 15 .5±0 .7 16 .0±0 .9 14 .6±1 .1 14 .7±1 .2 8 .8 ± 1 .4*** 8.0 ± 1 .0***
15 .4±1 .3 11 .7±1 .7 13 .1 ± 1 .4 11 .2±2 .3 10 .5±1 .8 9 .9 ± 1 .9** 9 .4 ± 1 .9**
*P
