- Email: [email protected]
IN VITRO INTERACTION BETWEEN SPIRAMYCIN AND POLYMORPHONUCLEAR NEUTROPHILS OXIDATIVE METABOLISM
Pharmacological Research, Vol. 37, No. 3, 1998
IN VITRO INTERACTION BETWEEN SPIRAMYCIN AND POLYMORPHONUCLEAR NEUTROPHILS OXIDATIVE METABOLISM I. MOUTARD, B. GRESSIERU a , C. BRUNET, T. DINE, M. LUYCKX, F. TEMPLIER, M. CAZIN and J.C. CAZIN Laboratoire de Pharmacologie, Pharmacocinetique et Pharmacie Clinique, Faculte´ des ´ Sciences Pharmaceutiques et Biologiques, rue du Professeur Laguesse, B.P. 83, 59006 Lille cedex, France and a Laboratoire de Biologie du Centre Hospitalier d’Armentieres, rue Sadi ` Carnot, BP 189, 59421 Armentieres ` cedex, France Accepted 8 January 1998
PMNs are a major component of body defense against microbial invasion, involving reactive oxygen species in great quantity, which could benefit from antibiotic therapy. Recently, possible antibiotic effects on phagocyte functions Žimpairment or stimulation of reactive oxygen species production. were studied. In our study, an in ¨ itro evaluation was made on macrolide activity on phagocyte respiratory burst functions, using assay of superoxide anion ŽO .2y . in response to four stimuli systems: N-formyl Met-Leu-Phe ŽfMLP., an analogue of bacterial chemotactic factors; 4b-phorbol 12-myristate 13-acetate ŽPMA., a direct activator of protein kinase C ŽPKC.; calcium ionophore ŽA23187., which acts directly on calcium influx; and a bacterial strain, Staphylococcus aureus. We have shown that spiramycin, at therapeutic plasma concentrations, increased O .2y generation by bacteria and fMLP-stimulated PMNs, with rate of 26% for 1 m g mly1 and 34% for 5 m g mly1 , respectively. This pro-oxidant effect, however weaker, was observed when PMNs were stimulated by PMA. A weak anti-oxidant effect was observed with A23187. For higher concentrations, spiramycin decreased strongly O .2y production, with IC 50 values of 74 m g mly1 , 154 m g mly1 , 296 m g mly1 and 400 m g mly1 when PMNs were stimulated with bacteria, A23187, fMLP and PMA, respectively. The effect of spiramycin seemed to result from an intracellular mechanism by intervention of PMN oxidative metabolism ŽNADPH-oxidase activation., rather than a simple chemical interaction, because no effect has been observed in acellular models. For higher spiramycin concentrations, the decrease of O .2y production observed could not be taken into consideration because this concentration was not used in therapy. The enhanced of O .2y production observed could be used in therapy, so as to increase PMNs bactericidal activity. Q 1998 The Italian Pharmacological Society KEY
WORDS:
spiramycin, PMNs, respiratory burst, PMNs bactericidal activity.
INTRODUCTION Polymorphonuclear neutrophils ŽPMNs. represent an important mechanism in host defense against a large variety of micro-organisms w1x. The aptitude of the host to recover from serious bacterial infections is dependent not only on antibiotic antimicrobial activity, but of correct PMNs functions. Destruction of intruding micro-organisms occurs as a result of a U
Corresponding author: B. Gressier, Laboratoire de Pharmacologie, Faculte ´ des Sciences Pharmaceutiques et Biologiques, B.P. 83, 59006 Lille cedex, France. 1043]6618r98r030197]05r$25.00r0rfr980292
complex event sequence initiated by ingestion and sequestration of microbes in the phagosome w2x. Concurrent with these processes, the surrounding medium oxygen is reduced to the superoxide anion ŽO .2y . by NADPH-oxidase, which leads subsequently to the formation of other toxic metabolites, such as hydroxyl radical ŽOH ? ., hydrogen peroxide ŽH 2 O 2 ., hypochlorous acid ŽHOCl. and N-chloramines ŽRNCl., each of these compounds have powerful microbicidal properties w1]3x. Recent reports on various antibiotics effects, on PMNs biological functions and especially their oxidative metabolism, have been published w4, 5x. The possibility that antimicrobial agents can induce supQ1998 The Italian Pharmacological Society
198
pression or really improve the PMNs functions is particularly important since these drugs are frequently administered to patients who, in addition to infections, modified their mechanisms of host defense w6x. Recently, in ¨ itro studies have shown that antibiotics, especially macrolides, allocated phagocyte functions w7, 8, 9x. Macrolides are characterised by their aptitude to penetrate and to concentrate into phagocytes, to reach intracellular concentrations considerably higher than those of other antibiotics. The aptitude of these substances to penetrate phagocyte cells could be an important factor concerning the therapy of infections caused by organisms which survive intracellularly after ingestion w10x. However, the question to know what antibiotics cause what functions Žsuch as respiratory burst, phagocytosis, chemotaxis. is again a matter of debate, essentially due to methodological differences between studies. There is also a considerable confusion concerning the possible clinic relevance of observed effects. In our study, we evaluated the effect of an antibiotic macrolide, spiramycin, on phagocyte respiratory burst, using an assay of O .2y, with an incubation of 1 h and four stimuli systems with different transductional mechanisms: N-formyl-Met-Leu-Phe ŽfMLP., an analogous oligopeptide of bacterial chemotactic factors, mimicking the activation of PMNs by bacteria w11x; 4b-phorbol 12-myristate 13-acetate ŽPMA., a phorbol ester, direct activator of proteine kinase C ŽPKC. w12x; calcium ionophore ŽA23187., which acts directly on the calcium influx; and a bacterial strain of Staphylococcus aureus.
MATERIALS AND METHODS
Drug and chemical products Buffer compounds, chemical products, reagents and spiramycin were purchased from Sigma Chemical Company ŽSt Louis, MO, USA.. Spiramycin was dissolved in phosphate buffered saline ŽPBS. 0.1 M, pH 7.4 to reach final concentrations ranging from 1 to 500 m g mly1 .
Bacterial strain Staphylococcus aureus was isolated from a clinical sample obtained from the Centre Hospitalier d’ArŽFrance.. Bacterial suspension was admentieres ` justed to an optical density of 0.75 at 550 nm, which corresponds to a bacterial concentration of 9 = 10 8 mly1 , as determined by the McFarland Ordinary Equipment ŽApi System, Biomerieux, Marcy l’Etoile, ´ France.. Bacteria were inoculated at a final concentration of 9 = 10 7 mly1 .
Isolation of human polymorphonuclear neutrophils (PMNs ) PMNs were purified from fresh heparinised venous
Pharmacological Research, Vol. 37, No. 3, 1998
blood of healthy human subjects. Briefly, PMNs were isolated by a density gradient technique followed by isotonic ammonium chloride haemolysis w13x. Cells isolated by this technique were always ) 95% viable as determined by Trypan blue exclusion, and 95% of cells were PMNs. Cells were suspended in Hank’s Hepes ŽHH. buffer at pH 7.4.
Cellular ¨ iability after action of agents To control the cellular viability after exposure to reagents, the cytosolic enzyme lactate dehydrogenase ŽLDH., was determined after each cellular experience w14x. PMNs were incubated at 378C in the presence of drugs or solvents plus stimulants and LDH liberation was determined by lysis of the cells at the end of the experiment. Results were expressed in percentage of LDH activity.
Cellular superoxide anion assay
O .2y generation by stimulated PMNs was measured by determining the superoxide dismutase-inhibitable reduction of horse cytochrome C w15x. Firstly, 1.5 = 10 6 PMNs were incubated for 60 min at 378C with or without various concentrations of spiramycin. After addition of cytochrome C Ž20 m M., PMNs were stimulated by fMLP Ž1 m M. and cytochalasin B Ž10 m M. or PMA Ž160 nM., or A23187 Ž100 m M., or bacterial suspension Ž9 = 10 7 ., for 60 min at 378C. Incubation time of stimulation and stimuli concentrations were chosen in order to obtain the same concentration of O .2y for every stimulated references. Others experiments have been made without any stimuli in order to inspect the effect of the drug on the basal O .2y release. After centrifugation at q48C, the supernatants were measured at 550 nm with a Kontron Uvikon 860 spectrophotometer against a reference containing a similar volume of PMNs suspension, identical concentrations of stimuli and drugs, and bovine erythrocyte superoxide dismutase ŽSOD, 250 UI mly1 .. O .2y concentration was calculated as a function of absorption change rate according to the Beer-Lambert law with a coefficient of extinction of E 550 s 2.1 = 10y2 mMy1 cmy1 . The results were expressed as a percentage of O .2y liberated, as previously described w16x.
Xanthine oxidase assay
To ensure that change in cellular O .2y assay was due to an effect of drug on phagocytes and was not simply due to a scavenger effect, we tested with an acellular model by using a xanthine oxidase assay. O 2.y was generated by the hypoxanthine-xanthine oxidase system w17x, in the same amount as in the cellular model. Reaction mixtures contained EDTA Ž1 mM., hypoxanthine Ž0.1 mM., cytochrome C Ž20 m M. and various concentrations of spiramycin, in a final volume of 1.5 ml buffered in KH 2 PO4 ]KOH Ž50 mM. pH 7.4. The reaction was started by adding
Pharmacological Research, Vol. 37, No. 3, 1998
xanthine oxidase Ž0.07 U mly1 . and the rate of reduced cytochrome C was measured at 550 nm. Amount of O .2y generated was calculated as previously described w18x. When spiramycin was added to the cell-free system, the initial rate of O .2y was not significantly modified.
Statistical analysis Data obtained was subjected to statistical analysis using a non-parametric test ŽWilcoxon. as previously reported w19x, where P-values F 0.05 were considered significant.
199
m g mly1 with fMLP and 26% for 1 m g mly1 with bacteria. When PMA was used as stimulus, the observed pro-oxidant effect was weaker, 12% for 5 m g mly1 . When A23187 was used as stimulus, no pro-oxidant effect was observed. At the highest non-therapeutic concentration, spiramycin strongly decreased PMNs oxidative response, with 50% inhibitory concentration ŽIC 50 . values of 74 m g mly1 , 154 m g mly1 , 296 m g mly1 and 400 m g mly1 when PMNs were stimulated by bacteria, A23187, fMLP and PMA respectively. The effect of the drug at different concentrations on the basal O .2y release was undertaken by using unstimulated PMN. As shown in Table I, the results do not show any activity of spiramycin. This was shown by the absence of O .2y release.
RESULTS
Cellular ¨ iability The potential cytotoxicity of spiramycin was tested at concentrations from 0.1 m g mly1 to 500 m g mly1 . As shown in Fig. 1, PMNs viability, evaluated by LDH liberation, was not substantially modified after 1 h of incubation in presence of spiramycin up to a concentration of 400 m g mly1 .
Superoxide anion (O2.y ) production When spiramycin was added to the cell-free system, the initial rate of O 2 was not significantly modified. As shown in Fig. 2, spiramycin increased significantly O .2y production by PMNs induced by fMLP or bacteria, near to therapeutic plasma concentrations Ž3 m g mly1 .. This pro-oxidant effect was 34% for 5
DISCUSSION
PMNs are a major component of host defense against microbial invasion, and this microbicidal mechanism partly depends on reactive oxygen species generation Žsuperoxide anion, hydrogen peroxide, hydroxyl radical, chlorines and chloramines.. Besides their antimicrobial activities, a lot of antibiotics show immunomodulator effects on host defense systems w4, 5x. The properties of these antibiotics could serve as a new strategy for treatment of infections. In this context, recent years have seen a renewal of interest in the achievement of synergic interaction between therapeutic agents and host defense systems: various
Fig. 1. Effect of spiramycin on LDH activity. Values are the means " SEM of 6 separate experiments. different from spiramycin free control Ž P- 0.05..
U
Significantly
Pharmacological Research, Vol. 37, No. 3, 1998
200
Fig. 2. Effect of spiramycin on O .2y production by stimulated PMNs in four different systems: Ž`. fMLP; Ž^. PMA; Žq. A23187 and Žv. bacteria. Values are the means " SEM of 6 separate experiments. U Significantly different from spiramycin free control Ž P- 0.05.. Ž`., fMLP; Ž^., PMA; Žq., A23187; Žv., Staph. Aureus.
Table I Effect of spiramycin on basal O .2y release Spiramycin concentrations (m g ml y 1) } 0.1 1 10 100 200 300 400 PMA Ž160 nM.UU
O2.y concentrations (m M) 0.43 0.47 0.38 0.52 0.52 0.23 0.47 0.24 11.47U
Values are the means " SEM of 6 separate experiments. Significantly different from spiramycin free control Ž P - 0.05.. UU Assays without spiramycin. U
investigations were undertaken concerning the effect of anti-infectious agents on PMNs functions, especially on macrolides and oxidative metabolism of these phagocyte cells. In our study, we studied spiramycin activity on PMNs respiratory burst response, especially on O .2y production. At therapeutic and supratherapeutic concentrations, our results demonstrated that such interaction with PMNs was not a result of spiramycin cytotoxicity. On the other hand, at usual therapeutic concentrations, we observed a significant pro-oxidant effect of this macrolide in two systems: fMLP and bacteria, with activation of 34% for 5 m g mly1 and 26% for 1 m g mly1 , respectively. Since no effect was observed on O .2y production in the acellular model, we suggested that spiramycin did not directly interact with extracellular and intracellular reactive oxygen species
by a radical scavenger mechanism, but that this macrolide acted on PMNs oxidative metabolism. Further, the probable mechanism of the activation of O .2y production could be a result of phospholipase C ŽPLC. stimulation, and so activation of diacylglycerol ŽDAG. generation and therefore of NADPH-oxidase w12x. Various investigators have already demonstrated this pro-oxidant activity concerning other macrolides, such as josamycin w7, 20x, erythromycin and miocamycin w21x and dirithromycin w22x. These observations, as well as antimicrobicidal activity of these drugs, could lead to a more efficient therapy for infections due to intracellular pathogens. However, our results were in disagreement with these of Labro et al. w23x and Dumas et al. w24x who have not observed any effect of spiramycin on O .2y generation. This could be explained by different techniques used in the laboratories. Indeed, Labro et al. w23x have studied spiramycin activity, for concentrations ranging from 1 to 100 m g mly1 , on O .2y production by PMA-stimulated PMNs: macrolide was dissolved with dimethylsulfoxide ŽDMSO. 10 mg mly1 , which could interfere with PMNs metabolism. Moreover, they have used an incubation time of 30 min between macrolide and PMNs, whereas it was demonstrated that incubation time allowing the best cellularrextracellular concentration ratio was obtained at 60 min. On the other hand, at supratherapeutic concentrations and at concentrations not obtained in PMNs on account of the low cellular concentrationrextracellular concentration ratio ŽCrE. of 15 at the end of 1 h w23x, that were unthinkable in therapeutic situations, spiramycin seemed to damage considerably the PMNs respiratory burst response in the four
Pharmacological Research, Vol. 37, No. 3, 1998
stimuli systems. Moreover, IC 50 was weaker in the A23187 system than in the fMLP and PMA systems, so we could suggest that the probable mechanism of in ¨ itro inhibition of O .2y generation was due to an inhibition of calcium influx, as Arbo and Santos w25x and Hand et al. w5x have already shown for clindamycin, and Nelson et al. w26x for erythromycin. However, we could not exclude a mechanism of PKC inhibition, as showed in the cellular PMA system. It would be interesting to evaluate ex ¨ i¨ o the effect of spiramycin treatment on neutrophil oxidative burst because, in ¨ i¨ o, cooperation of other physiological factors may influence the intraphagocytic activity of spiramycin, especially this interaction on O .2y metabolism. Moreover, t would permit us to study this interaction for longer incubation times. Nevertheless, this study has shown a pro-oxidant effect at therapeutic concentrations and by this increase of the O .2y production by PMNs, at therapeutic concentrations, in the two stimuli systems ŽfMLP and bacteria ., spiramycin indirectly favoured the bactericidal activity of PMNs by an increase in their oxidative metabolism, which represented a synergistic interaction between spiramycin and the host defense system.
ACKNOWLEDGEMENTS The authors wish to thank S. Battez-Lebegue for technical assistance.
REFERENCES 1. Babior BM. Oxidants from phagocytes: agents of defense and destruction. Blood 1984; 64: 959]66. 2. Roos D. The respiratory burst of phagocytic leukocytes. Drug In¨ est 1991; 3: 48]53. 3. Strieter RM, Kasahara K, Allen RM, Standiford TJ, Rolfe MW, Becker FS, Chensue SW, Kunkel ST. Cytokine-induced neutrophil-derived interleukine 8. Am J Pathol 1992; 141: 397]407. 4. Venezio FR, Di Vincenzo CA. Effects of aminoglycoside antibiotics on polymorphonuclear leukocyte function in vivo. Antimicrob Agents Chemother 1985; 27: 712]14. 5. Hand LW, Hand LD, King-Thompson NL. Antibiotic inhibition of the respiratory burst response in human polymorphonuclear leukocytes. Antimicrob Agents Chemother 1990; 34: 863]70. 6. Labro MT, El Benna J. Anti-infectious agents on polymorphonuclear neutrophils. Eur J Clin Microbiol Infect D 10 w2x: 124]31. 7. Labro MT, Babin-Chevaye C. Synergistic interaction of josamycin with human neutrophils bactericidal function in vitro. J Antimicrob Chemother 1989; 24: 731]40. 8. Labro MT, El Benna J, Abdelghaffar H. Modulation of human polymorphonuclear neutrophil function by m a crolide s: pre lim in a ry da ta con ce rn in g dirithromycin . J Antimicrob Chemother 1993; 31: 51]64. 9. Umeki S. Anti-inflammatory action of erythromycin. Its inhibitory effect on neutrophil NADPH-oxidase activity. Chest 1993; 104: 1191]3.
201
10. Prokesch RC, Hand WL. Antibiotic entry into human polymorphonuclear leukocytes. Antimicrob Agents Chemother 1982; 21: 373]80. 11. Bauldry SA, Wylke RL, Bass DA. Phospholipase A2 activation in human neutrophils differencial actions of diacylglycerol and alkylacylglycerols in priming cells for stimulation by N-formyl-Met-Leu-Phe. J. Biol Chem 1988; 263: 16787]95. 12. Rossi F. The O .2y-forming NADPH-oxidase of the phagocytes: nature, mechanisms of activation and function. Biochim Biophys Acta 1986; 853: 65]89. 13. Cabanis A, Gressier B, Lebegue S, Brunet C, Dine T, Luyckx M, Cazin M, Cazin JC. A rapid density gradient technique for separating polymorphonuclear granulocytes. APMIS 1994; 102: 119]21. 14. Srinivas VK, Habibullah CM, Ayesha Q, Hassan SI, Khaleel UR., Rehman K, Mohsins S. Lactate dehydrogenase: a marker of cellular integrity. Meth Find Exp Clin Pharmacol 1993; 15: 709]13. 15. Cohen HJ, Chovaniec ME. Superoxide generation by digitonin-stimulated guinea pig granulocytes. A basis for continious assay for monitoring superoxide production and for the study of the activation of the generation system. J Clin In¨ est 1978; 61: 1081]7. 16. Khalfi F, Gressier B, Brunet C, Dine T, Luyckx M, Cazin M, Cazin JC. Involvement of the extracellular calcium in the release of elastase and the human neutrophils oxidative burst. Cell Mol Biol 1996; 42: 1211]18. 17. Aruoma OL, Halliwell B, Hoey BM, Butler J. The antioxidant action of N-acetyl cysteine: its reaction with hydrogen peroxyde, hydroxyl radical, superoxyde anion and hypochlorous acid. Free Radical Biol Med 1989; 6: 593]7. 18. Cabanis A, Gressier B, Brunet C, Luyckx M, Dine T, Cazin M, Cazin JC. Effect of the protein kinase C inhibitor GF 109 203X on the elastase release and respiratory burst of human neutrophils. Gen Pharmacol 1996; 27: 1409]14. 19. Khalfi F, Gressir B, Brunet C, Dine T, Luyckx M, Cazin M, Cazin JC. Effects of calcium antagonist diltiazem on leukocyte elastase and on reactive oxygen species production in human neutrophils. Pharmacol Res 1996; 33: 117]21. 20. Labro MT, El Benna J. Synergistic bactericidal interaction of josamycin with human neutrophils in vitro. J Antimicrob Chemother 1990; 26: 515]24. 21. Bacci P, Grossi A, Vannuchi AM, Rafanelli D, Casini A, De Luca M. Effects of miocamycin and erythromycin on polymorphonuclear cell function. J Chemother 1992;4: 268]70. 22. Moutard I, Gressier B, Brunet C, Dine T, Luyckx M, Cazin M, Cazin JC. In vitro interaction between dirithromycin or its metabolite, erythromycylamine, and oxidative polymorphonuclear metabolism. J Antibiot 1997; 50: 53]7. 23. Labro MT, El Benna J, Babin-Chevaye C. Comparison of the in-vitro effect of several macrolides on the oxidative burst of human neutrophils. J Antimicrob Chemother 1989; 24: 561]72. 24. Dumas R, Brouland JP, Tedone R, Descote J. Influence of macrolide antibiotics on the chemiluminescence of zymosan-activated human neutrophils. Chemother 1990; 36: 381]4. 25. Arbo A, Santos JI. The in vitro effects of clindamycin on polymorphonuclear leukocyte function. Drug In¨ est 1990; 2: 235]41. 26. Nelson S, Summer WR, Terry PB, Warr GA, Jakab GJ. Erythromycin-induced suppression of pulmonary antibacterial defenses. A potential mechanism of surinfection in the lung. Am Re¨ Respir Dis 1987; 136: 1207]12.
IN VITRO INTERACTION BETWEEN SPIRAMYCIN AND POLYMORPHONUCLEAR NEUTROPHILS OXIDATIVE METABOLISM I. MOUTARD, B. GRESSIERU a , C. BRUNET, T. DINE, M. LUYCKX, F. TEMPLIER, M. CAZIN and J.C. CAZIN Laboratoire de Pharmacologie, Pharmacocinetique et Pharmacie Clinique, Faculte´ des ´ Sciences Pharmaceutiques et Biologiques, rue du Professeur Laguesse, B.P. 83, 59006 Lille cedex, France and a Laboratoire de Biologie du Centre Hospitalier d’Armentieres, rue Sadi ` Carnot, BP 189, 59421 Armentieres ` cedex, France Accepted 8 January 1998
PMNs are a major component of body defense against microbial invasion, involving reactive oxygen species in great quantity, which could benefit from antibiotic therapy. Recently, possible antibiotic effects on phagocyte functions Žimpairment or stimulation of reactive oxygen species production. were studied. In our study, an in ¨ itro evaluation was made on macrolide activity on phagocyte respiratory burst functions, using assay of superoxide anion ŽO .2y . in response to four stimuli systems: N-formyl Met-Leu-Phe ŽfMLP., an analogue of bacterial chemotactic factors; 4b-phorbol 12-myristate 13-acetate ŽPMA., a direct activator of protein kinase C ŽPKC.; calcium ionophore ŽA23187., which acts directly on calcium influx; and a bacterial strain, Staphylococcus aureus. We have shown that spiramycin, at therapeutic plasma concentrations, increased O .2y generation by bacteria and fMLP-stimulated PMNs, with rate of 26% for 1 m g mly1 and 34% for 5 m g mly1 , respectively. This pro-oxidant effect, however weaker, was observed when PMNs were stimulated by PMA. A weak anti-oxidant effect was observed with A23187. For higher concentrations, spiramycin decreased strongly O .2y production, with IC 50 values of 74 m g mly1 , 154 m g mly1 , 296 m g mly1 and 400 m g mly1 when PMNs were stimulated with bacteria, A23187, fMLP and PMA, respectively. The effect of spiramycin seemed to result from an intracellular mechanism by intervention of PMN oxidative metabolism ŽNADPH-oxidase activation., rather than a simple chemical interaction, because no effect has been observed in acellular models. For higher spiramycin concentrations, the decrease of O .2y production observed could not be taken into consideration because this concentration was not used in therapy. The enhanced of O .2y production observed could be used in therapy, so as to increase PMNs bactericidal activity. Q 1998 The Italian Pharmacological Society KEY
WORDS:
spiramycin, PMNs, respiratory burst, PMNs bactericidal activity.
INTRODUCTION Polymorphonuclear neutrophils ŽPMNs. represent an important mechanism in host defense against a large variety of micro-organisms w1x. The aptitude of the host to recover from serious bacterial infections is dependent not only on antibiotic antimicrobial activity, but of correct PMNs functions. Destruction of intruding micro-organisms occurs as a result of a U
Corresponding author: B. Gressier, Laboratoire de Pharmacologie, Faculte ´ des Sciences Pharmaceutiques et Biologiques, B.P. 83, 59006 Lille cedex, France. 1043]6618r98r030197]05r$25.00r0rfr980292
complex event sequence initiated by ingestion and sequestration of microbes in the phagosome w2x. Concurrent with these processes, the surrounding medium oxygen is reduced to the superoxide anion ŽO .2y . by NADPH-oxidase, which leads subsequently to the formation of other toxic metabolites, such as hydroxyl radical ŽOH ? ., hydrogen peroxide ŽH 2 O 2 ., hypochlorous acid ŽHOCl. and N-chloramines ŽRNCl., each of these compounds have powerful microbicidal properties w1]3x. Recent reports on various antibiotics effects, on PMNs biological functions and especially their oxidative metabolism, have been published w4, 5x. The possibility that antimicrobial agents can induce supQ1998 The Italian Pharmacological Society
198
pression or really improve the PMNs functions is particularly important since these drugs are frequently administered to patients who, in addition to infections, modified their mechanisms of host defense w6x. Recently, in ¨ itro studies have shown that antibiotics, especially macrolides, allocated phagocyte functions w7, 8, 9x. Macrolides are characterised by their aptitude to penetrate and to concentrate into phagocytes, to reach intracellular concentrations considerably higher than those of other antibiotics. The aptitude of these substances to penetrate phagocyte cells could be an important factor concerning the therapy of infections caused by organisms which survive intracellularly after ingestion w10x. However, the question to know what antibiotics cause what functions Žsuch as respiratory burst, phagocytosis, chemotaxis. is again a matter of debate, essentially due to methodological differences between studies. There is also a considerable confusion concerning the possible clinic relevance of observed effects. In our study, we evaluated the effect of an antibiotic macrolide, spiramycin, on phagocyte respiratory burst, using an assay of O .2y, with an incubation of 1 h and four stimuli systems with different transductional mechanisms: N-formyl-Met-Leu-Phe ŽfMLP., an analogous oligopeptide of bacterial chemotactic factors, mimicking the activation of PMNs by bacteria w11x; 4b-phorbol 12-myristate 13-acetate ŽPMA., a phorbol ester, direct activator of proteine kinase C ŽPKC. w12x; calcium ionophore ŽA23187., which acts directly on the calcium influx; and a bacterial strain of Staphylococcus aureus.
MATERIALS AND METHODS
Drug and chemical products Buffer compounds, chemical products, reagents and spiramycin were purchased from Sigma Chemical Company ŽSt Louis, MO, USA.. Spiramycin was dissolved in phosphate buffered saline ŽPBS. 0.1 M, pH 7.4 to reach final concentrations ranging from 1 to 500 m g mly1 .
Bacterial strain Staphylococcus aureus was isolated from a clinical sample obtained from the Centre Hospitalier d’ArŽFrance.. Bacterial suspension was admentieres ` justed to an optical density of 0.75 at 550 nm, which corresponds to a bacterial concentration of 9 = 10 8 mly1 , as determined by the McFarland Ordinary Equipment ŽApi System, Biomerieux, Marcy l’Etoile, ´ France.. Bacteria were inoculated at a final concentration of 9 = 10 7 mly1 .
Isolation of human polymorphonuclear neutrophils (PMNs ) PMNs were purified from fresh heparinised venous
Pharmacological Research, Vol. 37, No. 3, 1998
blood of healthy human subjects. Briefly, PMNs were isolated by a density gradient technique followed by isotonic ammonium chloride haemolysis w13x. Cells isolated by this technique were always ) 95% viable as determined by Trypan blue exclusion, and 95% of cells were PMNs. Cells were suspended in Hank’s Hepes ŽHH. buffer at pH 7.4.
Cellular ¨ iability after action of agents To control the cellular viability after exposure to reagents, the cytosolic enzyme lactate dehydrogenase ŽLDH., was determined after each cellular experience w14x. PMNs were incubated at 378C in the presence of drugs or solvents plus stimulants and LDH liberation was determined by lysis of the cells at the end of the experiment. Results were expressed in percentage of LDH activity.
Cellular superoxide anion assay
O .2y generation by stimulated PMNs was measured by determining the superoxide dismutase-inhibitable reduction of horse cytochrome C w15x. Firstly, 1.5 = 10 6 PMNs were incubated for 60 min at 378C with or without various concentrations of spiramycin. After addition of cytochrome C Ž20 m M., PMNs were stimulated by fMLP Ž1 m M. and cytochalasin B Ž10 m M. or PMA Ž160 nM., or A23187 Ž100 m M., or bacterial suspension Ž9 = 10 7 ., for 60 min at 378C. Incubation time of stimulation and stimuli concentrations were chosen in order to obtain the same concentration of O .2y for every stimulated references. Others experiments have been made without any stimuli in order to inspect the effect of the drug on the basal O .2y release. After centrifugation at q48C, the supernatants were measured at 550 nm with a Kontron Uvikon 860 spectrophotometer against a reference containing a similar volume of PMNs suspension, identical concentrations of stimuli and drugs, and bovine erythrocyte superoxide dismutase ŽSOD, 250 UI mly1 .. O .2y concentration was calculated as a function of absorption change rate according to the Beer-Lambert law with a coefficient of extinction of E 550 s 2.1 = 10y2 mMy1 cmy1 . The results were expressed as a percentage of O .2y liberated, as previously described w16x.
Xanthine oxidase assay
To ensure that change in cellular O .2y assay was due to an effect of drug on phagocytes and was not simply due to a scavenger effect, we tested with an acellular model by using a xanthine oxidase assay. O 2.y was generated by the hypoxanthine-xanthine oxidase system w17x, in the same amount as in the cellular model. Reaction mixtures contained EDTA Ž1 mM., hypoxanthine Ž0.1 mM., cytochrome C Ž20 m M. and various concentrations of spiramycin, in a final volume of 1.5 ml buffered in KH 2 PO4 ]KOH Ž50 mM. pH 7.4. The reaction was started by adding
Pharmacological Research, Vol. 37, No. 3, 1998
xanthine oxidase Ž0.07 U mly1 . and the rate of reduced cytochrome C was measured at 550 nm. Amount of O .2y generated was calculated as previously described w18x. When spiramycin was added to the cell-free system, the initial rate of O .2y was not significantly modified.
Statistical analysis Data obtained was subjected to statistical analysis using a non-parametric test ŽWilcoxon. as previously reported w19x, where P-values F 0.05 were considered significant.
199
m g mly1 with fMLP and 26% for 1 m g mly1 with bacteria. When PMA was used as stimulus, the observed pro-oxidant effect was weaker, 12% for 5 m g mly1 . When A23187 was used as stimulus, no pro-oxidant effect was observed. At the highest non-therapeutic concentration, spiramycin strongly decreased PMNs oxidative response, with 50% inhibitory concentration ŽIC 50 . values of 74 m g mly1 , 154 m g mly1 , 296 m g mly1 and 400 m g mly1 when PMNs were stimulated by bacteria, A23187, fMLP and PMA respectively. The effect of the drug at different concentrations on the basal O .2y release was undertaken by using unstimulated PMN. As shown in Table I, the results do not show any activity of spiramycin. This was shown by the absence of O .2y release.
RESULTS
Cellular ¨ iability The potential cytotoxicity of spiramycin was tested at concentrations from 0.1 m g mly1 to 500 m g mly1 . As shown in Fig. 1, PMNs viability, evaluated by LDH liberation, was not substantially modified after 1 h of incubation in presence of spiramycin up to a concentration of 400 m g mly1 .
Superoxide anion (O2.y ) production When spiramycin was added to the cell-free system, the initial rate of O 2 was not significantly modified. As shown in Fig. 2, spiramycin increased significantly O .2y production by PMNs induced by fMLP or bacteria, near to therapeutic plasma concentrations Ž3 m g mly1 .. This pro-oxidant effect was 34% for 5
DISCUSSION
PMNs are a major component of host defense against microbial invasion, and this microbicidal mechanism partly depends on reactive oxygen species generation Žsuperoxide anion, hydrogen peroxide, hydroxyl radical, chlorines and chloramines.. Besides their antimicrobial activities, a lot of antibiotics show immunomodulator effects on host defense systems w4, 5x. The properties of these antibiotics could serve as a new strategy for treatment of infections. In this context, recent years have seen a renewal of interest in the achievement of synergic interaction between therapeutic agents and host defense systems: various
Fig. 1. Effect of spiramycin on LDH activity. Values are the means " SEM of 6 separate experiments. different from spiramycin free control Ž P- 0.05..
U
Significantly
Pharmacological Research, Vol. 37, No. 3, 1998
200
Fig. 2. Effect of spiramycin on O .2y production by stimulated PMNs in four different systems: Ž`. fMLP; Ž^. PMA; Žq. A23187 and Žv. bacteria. Values are the means " SEM of 6 separate experiments. U Significantly different from spiramycin free control Ž P- 0.05.. Ž`., fMLP; Ž^., PMA; Žq., A23187; Žv., Staph. Aureus.
Table I Effect of spiramycin on basal O .2y release Spiramycin concentrations (m g ml y 1) } 0.1 1 10 100 200 300 400 PMA Ž160 nM.UU
O2.y concentrations (m M) 0.43 0.47 0.38 0.52 0.52 0.23 0.47 0.24 11.47U
Values are the means " SEM of 6 separate experiments. Significantly different from spiramycin free control Ž P - 0.05.. UU Assays without spiramycin. U
investigations were undertaken concerning the effect of anti-infectious agents on PMNs functions, especially on macrolides and oxidative metabolism of these phagocyte cells. In our study, we studied spiramycin activity on PMNs respiratory burst response, especially on O .2y production. At therapeutic and supratherapeutic concentrations, our results demonstrated that such interaction with PMNs was not a result of spiramycin cytotoxicity. On the other hand, at usual therapeutic concentrations, we observed a significant pro-oxidant effect of this macrolide in two systems: fMLP and bacteria, with activation of 34% for 5 m g mly1 and 26% for 1 m g mly1 , respectively. Since no effect was observed on O .2y production in the acellular model, we suggested that spiramycin did not directly interact with extracellular and intracellular reactive oxygen species
by a radical scavenger mechanism, but that this macrolide acted on PMNs oxidative metabolism. Further, the probable mechanism of the activation of O .2y production could be a result of phospholipase C ŽPLC. stimulation, and so activation of diacylglycerol ŽDAG. generation and therefore of NADPH-oxidase w12x. Various investigators have already demonstrated this pro-oxidant activity concerning other macrolides, such as josamycin w7, 20x, erythromycin and miocamycin w21x and dirithromycin w22x. These observations, as well as antimicrobicidal activity of these drugs, could lead to a more efficient therapy for infections due to intracellular pathogens. However, our results were in disagreement with these of Labro et al. w23x and Dumas et al. w24x who have not observed any effect of spiramycin on O .2y generation. This could be explained by different techniques used in the laboratories. Indeed, Labro et al. w23x have studied spiramycin activity, for concentrations ranging from 1 to 100 m g mly1 , on O .2y production by PMA-stimulated PMNs: macrolide was dissolved with dimethylsulfoxide ŽDMSO. 10 mg mly1 , which could interfere with PMNs metabolism. Moreover, they have used an incubation time of 30 min between macrolide and PMNs, whereas it was demonstrated that incubation time allowing the best cellularrextracellular concentration ratio was obtained at 60 min. On the other hand, at supratherapeutic concentrations and at concentrations not obtained in PMNs on account of the low cellular concentrationrextracellular concentration ratio ŽCrE. of 15 at the end of 1 h w23x, that were unthinkable in therapeutic situations, spiramycin seemed to damage considerably the PMNs respiratory burst response in the four
Pharmacological Research, Vol. 37, No. 3, 1998
stimuli systems. Moreover, IC 50 was weaker in the A23187 system than in the fMLP and PMA systems, so we could suggest that the probable mechanism of in ¨ itro inhibition of O .2y generation was due to an inhibition of calcium influx, as Arbo and Santos w25x and Hand et al. w5x have already shown for clindamycin, and Nelson et al. w26x for erythromycin. However, we could not exclude a mechanism of PKC inhibition, as showed in the cellular PMA system. It would be interesting to evaluate ex ¨ i¨ o the effect of spiramycin treatment on neutrophil oxidative burst because, in ¨ i¨ o, cooperation of other physiological factors may influence the intraphagocytic activity of spiramycin, especially this interaction on O .2y metabolism. Moreover, t would permit us to study this interaction for longer incubation times. Nevertheless, this study has shown a pro-oxidant effect at therapeutic concentrations and by this increase of the O .2y production by PMNs, at therapeutic concentrations, in the two stimuli systems ŽfMLP and bacteria ., spiramycin indirectly favoured the bactericidal activity of PMNs by an increase in their oxidative metabolism, which represented a synergistic interaction between spiramycin and the host defense system.
ACKNOWLEDGEMENTS The authors wish to thank S. Battez-Lebegue for technical assistance.
REFERENCES 1. Babior BM. Oxidants from phagocytes: agents of defense and destruction. Blood 1984; 64: 959]66. 2. Roos D. The respiratory burst of phagocytic leukocytes. Drug In¨ est 1991; 3: 48]53. 3. Strieter RM, Kasahara K, Allen RM, Standiford TJ, Rolfe MW, Becker FS, Chensue SW, Kunkel ST. Cytokine-induced neutrophil-derived interleukine 8. Am J Pathol 1992; 141: 397]407. 4. Venezio FR, Di Vincenzo CA. Effects of aminoglycoside antibiotics on polymorphonuclear leukocyte function in vivo. Antimicrob Agents Chemother 1985; 27: 712]14. 5. Hand LW, Hand LD, King-Thompson NL. Antibiotic inhibition of the respiratory burst response in human polymorphonuclear leukocytes. Antimicrob Agents Chemother 1990; 34: 863]70. 6. Labro MT, El Benna J. Anti-infectious agents on polymorphonuclear neutrophils. Eur J Clin Microbiol Infect D 10 w2x: 124]31. 7. Labro MT, Babin-Chevaye C. Synergistic interaction of josamycin with human neutrophils bactericidal function in vitro. J Antimicrob Chemother 1989; 24: 731]40. 8. Labro MT, El Benna J, Abdelghaffar H. Modulation of human polymorphonuclear neutrophil function by m a crolide s: pre lim in a ry da ta con ce rn in g dirithromycin . J Antimicrob Chemother 1993; 31: 51]64. 9. Umeki S. Anti-inflammatory action of erythromycin. Its inhibitory effect on neutrophil NADPH-oxidase activity. Chest 1993; 104: 1191]3.
201
10. Prokesch RC, Hand WL. Antibiotic entry into human polymorphonuclear leukocytes. Antimicrob Agents Chemother 1982; 21: 373]80. 11. Bauldry SA, Wylke RL, Bass DA. Phospholipase A2 activation in human neutrophils differencial actions of diacylglycerol and alkylacylglycerols in priming cells for stimulation by N-formyl-Met-Leu-Phe. J. Biol Chem 1988; 263: 16787]95. 12. Rossi F. The O .2y-forming NADPH-oxidase of the phagocytes: nature, mechanisms of activation and function. Biochim Biophys Acta 1986; 853: 65]89. 13. Cabanis A, Gressier B, Lebegue S, Brunet C, Dine T, Luyckx M, Cazin M, Cazin JC. A rapid density gradient technique for separating polymorphonuclear granulocytes. APMIS 1994; 102: 119]21. 14. Srinivas VK, Habibullah CM, Ayesha Q, Hassan SI, Khaleel UR., Rehman K, Mohsins S. Lactate dehydrogenase: a marker of cellular integrity. Meth Find Exp Clin Pharmacol 1993; 15: 709]13. 15. Cohen HJ, Chovaniec ME. Superoxide generation by digitonin-stimulated guinea pig granulocytes. A basis for continious assay for monitoring superoxide production and for the study of the activation of the generation system. J Clin In¨ est 1978; 61: 1081]7. 16. Khalfi F, Gressier B, Brunet C, Dine T, Luyckx M, Cazin M, Cazin JC. Involvement of the extracellular calcium in the release of elastase and the human neutrophils oxidative burst. Cell Mol Biol 1996; 42: 1211]18. 17. Aruoma OL, Halliwell B, Hoey BM, Butler J. The antioxidant action of N-acetyl cysteine: its reaction with hydrogen peroxyde, hydroxyl radical, superoxyde anion and hypochlorous acid. Free Radical Biol Med 1989; 6: 593]7. 18. Cabanis A, Gressier B, Brunet C, Luyckx M, Dine T, Cazin M, Cazin JC. Effect of the protein kinase C inhibitor GF 109 203X on the elastase release and respiratory burst of human neutrophils. Gen Pharmacol 1996; 27: 1409]14. 19. Khalfi F, Gressir B, Brunet C, Dine T, Luyckx M, Cazin M, Cazin JC. Effects of calcium antagonist diltiazem on leukocyte elastase and on reactive oxygen species production in human neutrophils. Pharmacol Res 1996; 33: 117]21. 20. Labro MT, El Benna J. Synergistic bactericidal interaction of josamycin with human neutrophils in vitro. J Antimicrob Chemother 1990; 26: 515]24. 21. Bacci P, Grossi A, Vannuchi AM, Rafanelli D, Casini A, De Luca M. Effects of miocamycin and erythromycin on polymorphonuclear cell function. J Chemother 1992;4: 268]70. 22. Moutard I, Gressier B, Brunet C, Dine T, Luyckx M, Cazin M, Cazin JC. In vitro interaction between dirithromycin or its metabolite, erythromycylamine, and oxidative polymorphonuclear metabolism. J Antibiot 1997; 50: 53]7. 23. Labro MT, El Benna J, Babin-Chevaye C. Comparison of the in-vitro effect of several macrolides on the oxidative burst of human neutrophils. J Antimicrob Chemother 1989; 24: 561]72. 24. Dumas R, Brouland JP, Tedone R, Descote J. Influence of macrolide antibiotics on the chemiluminescence of zymosan-activated human neutrophils. Chemother 1990; 36: 381]4. 25. Arbo A, Santos JI. The in vitro effects of clindamycin on polymorphonuclear leukocyte function. Drug In¨ est 1990; 2: 235]41. 26. Nelson S, Summer WR, Terry PB, Warr GA, Jakab GJ. Erythromycin-induced suppression of pulmonary antibacterial defenses. A potential mechanism of surinfection in the lung. Am Re¨ Respir Dis 1987; 136: 1207]12.