- Email: [email protected]
Indomethacin prevents the long-lasting inhibitory effect of aspirin on human platelet cyclo-oxygenase activity
PROSTAGLANDINS
INDOMETHACIN PREVENTS THE LONG-LASTING INHIBITORY EFFECT OF ASPIRIN ON HUMAN PLATELET CYCLO-OXYGENASE ACTIVITY
M. Livio, A. Del Maschio, C. Cerletti and G. de Gaetano Laboratory of Cardiovascular Clinical Pharmacology Istituto di Ricerche Farmacologiche "Mario Negri", Milan, Italy
ABSTRACT Aspirin and indomethacin do interact with the same site on This suggestion is based on in vitro studies on cycle-oxygenase. ram seminal vesicles and in vivo drug interaction studies on rat platelets. The purpose of the present study was to ascertain whether the same interaction also occurred after administration of both drugs to human volunteers. Platelet aggregation induced by sodium arachidonate or by collagen, and formation of platelet MDA and TxB2 were measured before, two and 48 hours after ingestion of either indomethacin (50 mg) or aspirin (500 mg) or of both drugs (30 minutes apart). While the inhibitory effect of indomethacin on these parameters was short lasting, that of aspirin persisted for at least 48 hours. However, when both drugs were given concurrently, the long-lasting effect of aspirin was no longer detectable. Since competition at levels other than platelets was unlikely, this study indicates that indomethacin and aspirin inhibit human platelet cycle-oxygenase by the same basic mechanism. Acetylation of the enzyme appears to be a secondary mechanism which makes the inhibitory effect of aspirin persistent. INTRODUCTION Aspirin and other non-steroidal antiinflammatory drugs (NSAIDS) inhibit platelet cycle-oxygenase activity. Aspirin inhibits this enzyme by irreversible acetylation of its active site (l), whereas the mechanism whereby indomethacin and other NSAIDs inhibit it is not known. Experiments in vitro with washed rabbit platelets failed to demonstrate any interaction between indomethacin and aspirin, suggesting that the two drugs differ as regards their site of inhibition on cycle-oxygenase (2). Using radio-labelled compounds, Stanford et al. (3) have shown that indomethacin, unlike aspirin, does not covalently modify cyclooxygenase from ram seminal vesicles. Nevertheless, indomethacin can prevent enzyme acetylation by aspirin. This suggests that the two drugs may bind to the same site on the enzyme. Using microsomes from ram seminal vesicles, Humes et al. (4) have recently shown that NSAIDs react in vitro with two sites on the cycle-oxygenase,the catalytic site and a putative supplementary
JUNE
1982 VOL. 23 NO. 6
787
PROSTAGLANDINS
one. Interaction with the latter is essential for the efficacy of NSAIDs as cycle-oxygenase inhibitors. The existence of a putative supplementary binding site on platelet cycle-oxygenase has been confirmed by in vivo drug interaction studies in the rat (5). It has been shownnndomethacin pretreatment of animals prevents the inhibitory effect of aspirin on platelet malondialdehyde (MDA) and thromboxane (TX) generation. Although indomethacin and aspirin are often administered concurrently in clinical practice (for instance in rheumatic patients), interaction studies of these two drugs on platelet function and prostaglandin formation in man have apparently not been carried out. Evidence is presented here that administration of indomethacin to human volunteers prevents the long-lasting effect of a subsequent single dose of aspirin on plateletcyclo-oxygenase activity and platelet aggregation. This suggests that indomethacin and aspirin interfere with the same biochemical mechanism leading to prostaglandin generation and platelet aggregation. MATERIALS AND METHODS In vitro studies. The following aggregating agents were used: -T--~-----_~~--arachidonic acid (AA) (Sigma, St. Louis, Missouri,> 99% pure) and acid-soluble collagen (Hormon Chemie, Munchen, WG). A solution of 50 mM sodium arachidonate was prepared by dissolving arachidonic acid in 100 mM sodium carbonate, as described by Silver et al. (6). Blood was drawn from the anticubital vein of apparently healthy subjects who declared they had taken no drugs for at least two weeks before the experiment. Nine volumes of blood were mixed with one volume of trisodium citrate (3.8%). Platelet-rich plasma (PRP) was prepared by centrifuging the titrated blood at 200 x g for 15 min at room temperature. PRP (250 ~1) was stirred in the cuvette of an aggregometer (Elvi 840, Elvi Logos, Milan, Italy) at 37°C with microliter amounts first of sodium salicylate (1 mM final cont., Farmitalia Carlo Erba, Milan, Italy), indomethacin (0.5-3.0 PM final cont., Merck, Sharp and Dohme, Rome, Italy) or redistilled water for 1 min, then with aspirin (50-250 PM, lysine salt, Flectadol, Maggioni, Milan, Italy) or redistilled water for one more minute. Then 0.4-1.0 mM sodium arachidonate was added to the cuvette and platelet aggregation was followed for 3 min. Ex vivo studies. Three healthy volunteers (2 men and one woman), __--------~~_-31-38 years old, participated in the study. Each subject ingested: firstly, 50 mg inhomethacin (Indocid, Merck, Sharp and Dohme); then, three days later, when platelet function was completely restored, 500 mg aspirin (AspirinR, Bayer, Italy) and two weeks later both indomethacin and aspirin 30 min apart for the interaction study. Blood was sampled before and at different times after drug ingestion as detailed in the Results section (Table I). 0.6 ml of PRP, prepared as described for in vitro studies, was stirred for 1 min at 37°C in the aggregometer cuvette and challenged with various final concentrations of the aggregating agents to determine
JUNE
1982 VOL. 23 NO. 6
PROSTAGLANDINS
the threshold aggregating concentrations (TAC). The TAC was defined as the smallest concentration of each aggregating agent which gave irreversible pldtelet aggregation starting within 3 min of addition of the aggregating agent to PBP. At that time a portion of each sample was pipetted off, diluted in an equal volume 0: cold ethanol, spun and the supernatant used for thromboxane B2 (TxB2) radioimmunoassay as described by Ingermanh'ojensket al. (7). Highly specific rabbit antibodies against TxB2 were provided by J.B. Smith (Cardeza Foundation, Thomas Jefferson University, Philadelphia, Pa., USA). Cross reactivity of this antibodies with arachidonic acid was less than 0.01% and with other prostanoid compounds less than 0.02%. The remainder of the sample was processed for MDA spectrophotometric assay as previously described (8). RESULTS In vitro studies. Indomethacin (0.5-5 p) induced concentratim--__-----_-----dependent inhibition of platelet aggregation stimulated by 0.4-0.6 mM AA. At none of the concentrations tested was indomethacin able to prevent the inhibitory effect of aspirin, the latter used at concentrations which induced variable degrees of inhibition (40-100 PM). Fig. 1 reports a representative experiment showing lack of prevention by indomethacin (at the maximal ineffective concentration, 0.6 PM) of the inhibitory effect of aspirin (at the minimal concentration causing complete inhibition of aggregation,75 PM). Additivity rather than prevention of inhibition was found when indomethacin and aspirin were combined at concentrations inducing partial inhibition of platelet aggregation (data not shown). In contrast preincubation of platelet with 1 mM sodium salicylate (which did not modify AA-induced aggregation) resulted in complete prevention of the inhibitory effect of either 5 M indomethacin or 75 ?M tr aspirin. Ex vivo studies. Fig. 2 reports platelet aggregation tracings T__--_----_---obtained in one volunteer given either 50 mg indomethacin or 500 mg aspirin or both drugs (30 min apart). Two hours after administra-tion of either the single drugs or the combined schedule, platelet aggregation induced by 0.5 mM sodium arachidonate was completely blocked. Platelet aggregation still appeared completely inhibited 48 h after aspirin but had returned to normal after indomethacin. Prolongation of the latent period but almost normal platelet aggregation was observed 48 h after ingestion of both drugs together. A similar pattern of platelet aggregation was seen when collagen was used as aggregating agentingead of sodium arachidonate. Table I summarizes the results with the two stimuli in the three volunteers studied. The threshold aggregating concentration (as defined above) of sodium arachidonate and collagen was determined for each of them before starting the drug study. In all instances no aggregation
JUNE
1982 VOL. 23 NO. 6
PROSTAGLANDINS
I
ASA INDO+ASA
C or
INDO
INDO
SAL + INDO C or SAL
1 min
Figure 1. Representative tracings of platelet aggregation induced by sodium arachidonate (0.4 mM, final cont.). PRP was preincubated with 75 PM aspirin (ASA), 0.6 p (upper panel) and 5 PM (middle panel) indomethacin (INDO), 1 mM sodium salicylate (SAL) or a combination of the drugs before addition of sodium arachidonate as indicated. C = Controls. For further details see Materials and Methods.
790
JUNE
1982 VOL. 23 NO. 6
PROSTAGLANDINS
_0
40
2 HOURS
AFTER
INGESTION
Figure 2. Representative tracings of platelet aggregation induced by sodium arachidonate (0.5 mM, final cone.) in PRP from a healthy volunteer. Blood was collected before and at various intervals after ingestion of indomethacin (50 mg), aspirin (500 mg) or both drugs (30 min apart) as indicated. For further details see Materials and Methods.
could be obtained by raising the concentration of sodium arachidonate to 1.6 mM and that of collagen to 4 pg/ml in the samples collected two hours after single or combined drug administration. After 48 h the TACs reverted toward basal values after indomethacin, but the inhibitory effect of aspirin persisted apparently unmodified. Administration of indomethacin before aspirin resulted 48 h later in almost complete recovery of the platelet aggregation response to either stimulus. Platelet cycle-oxygenase activity was measured both as MDA and TxB2 generated on incubation of PRP samples with sodium arachidonate. The results, summarized in Fig. 3, show that both products of arachidonic acid metabolism were undetectable 2 h after administration of single or combined drugs. They remained undetectable 48 h
JUNE
1982 VOL. 23 NO.
b
PROSTAGLANDINS
Table I. Threshold aggregating concentrations of sodium arachidonate and collagen determined before and at intervals after oral administration of indomethacin (50 mg), aspirin (500 mg) or both drugs (30 min apart).Individual values for the three subjects studied. Sodium arachidonate (ml.0
Drug
Collagen @g/ml)
0.7 - 0.5 - 0.9
1.0 - 1.5 - 2.0
Indomethacin
2h 48 h
>1.6 ->1.6 -)1.6 0.9 - 0.6 - 0.9
> 4.0 -)4.0 ->4.0 1.0 - 1.3 - 1.5
Aspirin
2h 48 h
p1.6 ->1.6 -91.6 F1.6 ->1.6 ->1.6
J4.0 -)4.0 -)4.0 > 4.0 ->4.0 ->4.0
Indomethacin + 2 h Aspirin 48 h
)1.6 ->1.6 ->1.6 1.0 - 1.0 - 1.0
p4.0 ->4.0 ->4.0 1.0 - 1.5 - 1.5
after aspirin but averaged 60-80% of basal values 48 h after indomethacin alone or indomethacin plus aspirin. DISCUSSION This study shows that indomethacin prevents the in vivo inhibitory effect of aspirin on human platelet cycle-oxygenase activity. This was measured by platelet aggregation induced by sodium arachidonate or collagen and by the formation of MDA and TxB2, two metabolites of arachidonic acid. Although interaction of indomethacin with aspirin has already been reported in ram seminal vesicles in vitro (4) and in rat platelets in vivo (5), this is the first demonstration of this kind of interxin human volunteers in vivo. Previous in vitro studies onhuman (9) or washed rabbit platelets (2) failed to demonstrate any interaction between indomethacin and aspirin. In the present study we were also unable to show any prevention by indomethacin of aspirin inhibition of arachidonic acid-induced aggregation of human platelets in vitro. Additivity rather than competition between the two drugs was observed. In the same experimental system we found that sodium salicylate, although apparently inactive by itself, did prevent the inhibitory effect of either indomethacin or aspirin. This finding could indicate that all three drugs interact with the same site on plateIt has recently been suggested that NSAIDs let cycle-oxygenase. interact with cycle-oxygenase at a site related to but distinct from the active substrate site (4,5). The various drugs may have different affinities for each site. Salicylate in particular may have a greater affinity for the binding than the active site, whereas indomethacin shows similar affinity for both sites.
792
JUNE
1982 VOL. 23 NO. 6
PROSTAGLANDINS
El
TxE12
100
INDOMETHACIN
100
. LA.
ASPIRIN
J-:
0
INDOMETHACIN + ASPIRIN
l
2
40
HOURS
AFTER
INGESTION
Figure 3. Cycle-oxygenase activity (measured as MDA and TxB2 generation induced by threshold concentrations of sodium arachidonate) Individual values (0) and in PW from three healthy volunteers. means (-) are reported as percentages of basal values. Blood was collected before and at various intervals after ingestion of indomethacin (50 mg), aspirin (500 mg) or both drugs (30 min apart) as indicated. For further details see Materials and Methods.
If one accepts the above enzyme model, indomethacin appears only to interact with the binding site at concentrations inhibiting the enzyme's activity, while salicylate binds to the enzyme at concentrations not affecting its function. This could explain the results of our in vitro studies in which enzyme activity was measured: at effective concentrations of indomethacin, additivity with aspirin was found. At ineffective concentrations, indomethacin did not prevent aspirin's effect. In contrast, the relatively high concentrations of salicylate used, leaving the eneyme activity unaltered but interacting with its binding site, stopped aspirin gaining
JUNE
1982 VOL. 23 NO. 6
793
PROSTAGLANDINS
access to the latter, and consequently prevented the inhibitory effect of this drug. Our interpretation is supported by the fact that the only report of interaction between indomethacin and aspirin in vitro was based on measurements of enzyme acetylation by radiolabelled aspirin rather than enzyme activity (3). The concentration of indomethacin used in the latter study inhibited cycle-oxygenase activity by more than 90%, so that no interaction with'aspirin would have been observed if enzyme activity rather than enzyme acetylation had been measured. We therefore devised an in vivo experiment which took advantage of the short-lasting effect ofmethacin and the persistent inhibition of plateyet cycle-oxygenase activity by aspirin (10). Indomethacin was administered 30 min before aspirin and blood was collected 2 and 48 h later. After 2 h platelet aggregation induced by sodium arachidonate or collagen and arachidonic acid metabolism was completely blocked, and the inhibitory effect of indomethacin alone had almost completely disappeared after 48 h. At 48 h platelet cycle-oxygenase from volunteers given aspirin alone was still completely inhibited. As shown in Figs. 2 and 3 and in Table I, indomethacin pretreatment almost entirely prevented the aspirin effect 48 h later. These results suggest that the interaction between indomethacin and aspirin may have been masked in our in vitro studies and in previous reports (2,9) by the inhibitory activity of indomethacin itself. The indomethacin-aspirin interaction observed in vivo and the interaction of either drug with sodium salicylate in support the concept that the three drugsdnteract with the same site on cycle-oxygenase. Since salicylate is inactive by itself on this platelet enzyme, it may well interact with aspirin or indomethacin at the putative supplementary binding site, rather than directly on the substrate active site. Consequently it has been proposed that indomethacin and aspirin interact with the same site, a phenomenon essential for their action as cycle-oxygenase inhibitors (4,5). The mechanism of action of both drugs would therefore be the same. Acetylation by aspirin of an aminoacid critical to the action of cycle-oxygenase (11) would be a secondary effect, which renders this drug's inhibitory effect irreversible rather than actually inducing it (5,10,12,13). The possibility that the interaction observed in the present study occurred at a level different from cycle-oxygenase is unlikely since, as already mentioned, the interaction between indomethacin and aspirin was shown using a purified cycle-oxygenase enzyme preparation (3). Moreover in a study on rheumatic patients there was no difference in serum levels of salicylate after oral administration of aspirin alone and after a concurrent oral dose of indomethacin (14). Thus the possibility of altered aspirin absorption by indomethacin in the gastrointestinal tract,although cannot be ruled out with certainty,does not appear to explain the reduced effect of aspirin observed in our study. Finally, prevention by indomethacin of aspirin binding to plasma proteins would have resulted in increased availability of free
JUNE
1982 VOL. 23 NO. 6
PROSTAGLANDINS
aspirin for the platelet enzyme and thus in an increased inhibitory effect. ACKNOWLEDGEMENTS This work was supported by the Italian National Research Council (Program "Farmacologia Clinica e Malattie Rare"). Judith Baggott, Anna Mancini, Antonella Bottazzi and Vincenzo and Felice de Ceglie helped prepare the manuscript. M.L. is a Fellow of the Italian National Research Council. REFERENCES Roth, G.J., N. Stanford, and P.W. Majerus. Acetylation of prostaglandin synthase by aspirin. Proc. Natl. Acad. Sci. USA -72: 3073, 1975.
2)
Vargaftig, B.B. The inhibition of cycle-oxygenase of rabbit platelets by aspirin is prevented by salicylic acid and by phenanthrolines. 'Eur. J. Pharmacol. -50: 231, 1978.
3)
Stanford, N., G.J. Roth, T.Y. Shen, and P.W. Majerus. Lack of covalent modification of prostaglandin synthetase (cyclooxygenase) by indomethacin. Prostaglandins -13: 669, 1977.
4)
Humes, J.L., C.A. Winter, S.J. Sadowski, and F.A. Kuehl Jr. Multiple sites on prostaglandin cyclooxygenase are determinants in the action of nonsteroidal antiinflammatory agents. Proc. Natl. Acad. Sci. USA -78: 2053, 1981.
5)
Cerletti, C., M. Livio , and G. de Gaetano. Non-steroidal antiinflammatory drugs react with two sites on platelet cyclooxygenase. Evidence from -in vivo drug interaction studies in rats. Biochim. Biophys. Acta 714: 122 (1982).
6)
Silver, M.J., W. Hoch, J.J. Kocsis, C.M. Ingerman, and J.B. Smith. Arachidonic acid causes sudden death in rabbits. Science 183: 1085, 1974.
7)
Ingerman-Wojenski, C., M.J. Silver, J.B. Smith, and E. Macarak. Bovine endothelial cells in culture produce thromboxane as well as prostacyclin. J. Clin. Invest. -67: 1292, 1981.
8)
Rajtar, G., C. Cerletti, M. Livio, and G. de Gaetano. Sulphinpyrazone preventsin -- vivo the inhibitory effect of aspirin on rat platelet cycle-oxygenase activity. Biochem. P'narmacol. -30: 2773, 1981.
9)
Zucker, M.B., and J. Peterson. Effect of acetylsalicylic acid, other nonsteroidal anti-inflammatory agents, and dipyridamole on human blood platelets. J. Lab. Clin. Med. -76: 66, 1970.
JUNE
1982 VOL. 23 NO. 6
795
PROSTAGLANDINS
10)
O'Brien, J.R. Effect of anti-inflammatory Lancet -1: 894, 1968.
11)
Roth, G.J., and C.J. Siok. Acetylation of the NH2-terminal serine of prostaglandin synthetase by aspirin. J. Biol. Chem. 253: 3782, 1978.
12)
Merino, J., M. Livio, G. Rajtar, and G. de Gaetano. Salicylate reverses in vitro aspirin inhibition of rat platelet and vascular prostagla;;;?iGeration. Biochem. Pharmacol. -29: 1093, 1980.
13)
Livio, M., G. Rajtar, J. Merino, and G. de Gaetano. Malondialdehyde formation in rat platelet-rich plasma. II. Modification of the reaction kinetics by aspirin and indomethacin. Thromb. Haemost. -44: 52, 1980.
14)
Kaldestad, E., T. Hansen, and H.K. Brath. Interaction of indomethacin and acetylsalicylic acid as shown by the serum concentrations of indomethacin and salicylate. Eur. J. Clin. Pharmacol. -9: 199, 1975.
agents on platelets.
Robert R. Gorman Editor: Received: l-27-82 4-15-82 Accepted:
796
JUNE
1982 VOL. 23 NO. 6
INDOMETHACIN PREVENTS THE LONG-LASTING INHIBITORY EFFECT OF ASPIRIN ON HUMAN PLATELET CYCLO-OXYGENASE ACTIVITY
M. Livio, A. Del Maschio, C. Cerletti and G. de Gaetano Laboratory of Cardiovascular Clinical Pharmacology Istituto di Ricerche Farmacologiche "Mario Negri", Milan, Italy
ABSTRACT Aspirin and indomethacin do interact with the same site on This suggestion is based on in vitro studies on cycle-oxygenase. ram seminal vesicles and in vivo drug interaction studies on rat platelets. The purpose of the present study was to ascertain whether the same interaction also occurred after administration of both drugs to human volunteers. Platelet aggregation induced by sodium arachidonate or by collagen, and formation of platelet MDA and TxB2 were measured before, two and 48 hours after ingestion of either indomethacin (50 mg) or aspirin (500 mg) or of both drugs (30 minutes apart). While the inhibitory effect of indomethacin on these parameters was short lasting, that of aspirin persisted for at least 48 hours. However, when both drugs were given concurrently, the long-lasting effect of aspirin was no longer detectable. Since competition at levels other than platelets was unlikely, this study indicates that indomethacin and aspirin inhibit human platelet cycle-oxygenase by the same basic mechanism. Acetylation of the enzyme appears to be a secondary mechanism which makes the inhibitory effect of aspirin persistent. INTRODUCTION Aspirin and other non-steroidal antiinflammatory drugs (NSAIDS) inhibit platelet cycle-oxygenase activity. Aspirin inhibits this enzyme by irreversible acetylation of its active site (l), whereas the mechanism whereby indomethacin and other NSAIDs inhibit it is not known. Experiments in vitro with washed rabbit platelets failed to demonstrate any interaction between indomethacin and aspirin, suggesting that the two drugs differ as regards their site of inhibition on cycle-oxygenase (2). Using radio-labelled compounds, Stanford et al. (3) have shown that indomethacin, unlike aspirin, does not covalently modify cyclooxygenase from ram seminal vesicles. Nevertheless, indomethacin can prevent enzyme acetylation by aspirin. This suggests that the two drugs may bind to the same site on the enzyme. Using microsomes from ram seminal vesicles, Humes et al. (4) have recently shown that NSAIDs react in vitro with two sites on the cycle-oxygenase,the catalytic site and a putative supplementary
JUNE
1982 VOL. 23 NO. 6
787
PROSTAGLANDINS
one. Interaction with the latter is essential for the efficacy of NSAIDs as cycle-oxygenase inhibitors. The existence of a putative supplementary binding site on platelet cycle-oxygenase has been confirmed by in vivo drug interaction studies in the rat (5). It has been shownnndomethacin pretreatment of animals prevents the inhibitory effect of aspirin on platelet malondialdehyde (MDA) and thromboxane (TX) generation. Although indomethacin and aspirin are often administered concurrently in clinical practice (for instance in rheumatic patients), interaction studies of these two drugs on platelet function and prostaglandin formation in man have apparently not been carried out. Evidence is presented here that administration of indomethacin to human volunteers prevents the long-lasting effect of a subsequent single dose of aspirin on plateletcyclo-oxygenase activity and platelet aggregation. This suggests that indomethacin and aspirin interfere with the same biochemical mechanism leading to prostaglandin generation and platelet aggregation. MATERIALS AND METHODS In vitro studies. The following aggregating agents were used: -T--~-----_~~--arachidonic acid (AA) (Sigma, St. Louis, Missouri,> 99% pure) and acid-soluble collagen (Hormon Chemie, Munchen, WG). A solution of 50 mM sodium arachidonate was prepared by dissolving arachidonic acid in 100 mM sodium carbonate, as described by Silver et al. (6). Blood was drawn from the anticubital vein of apparently healthy subjects who declared they had taken no drugs for at least two weeks before the experiment. Nine volumes of blood were mixed with one volume of trisodium citrate (3.8%). Platelet-rich plasma (PRP) was prepared by centrifuging the titrated blood at 200 x g for 15 min at room temperature. PRP (250 ~1) was stirred in the cuvette of an aggregometer (Elvi 840, Elvi Logos, Milan, Italy) at 37°C with microliter amounts first of sodium salicylate (1 mM final cont., Farmitalia Carlo Erba, Milan, Italy), indomethacin (0.5-3.0 PM final cont., Merck, Sharp and Dohme, Rome, Italy) or redistilled water for 1 min, then with aspirin (50-250 PM, lysine salt, Flectadol, Maggioni, Milan, Italy) or redistilled water for one more minute. Then 0.4-1.0 mM sodium arachidonate was added to the cuvette and platelet aggregation was followed for 3 min. Ex vivo studies. Three healthy volunteers (2 men and one woman), __--------~~_-31-38 years old, participated in the study. Each subject ingested: firstly, 50 mg inhomethacin (Indocid, Merck, Sharp and Dohme); then, three days later, when platelet function was completely restored, 500 mg aspirin (AspirinR, Bayer, Italy) and two weeks later both indomethacin and aspirin 30 min apart for the interaction study. Blood was sampled before and at different times after drug ingestion as detailed in the Results section (Table I). 0.6 ml of PRP, prepared as described for in vitro studies, was stirred for 1 min at 37°C in the aggregometer cuvette and challenged with various final concentrations of the aggregating agents to determine
JUNE
1982 VOL. 23 NO. 6
PROSTAGLANDINS
the threshold aggregating concentrations (TAC). The TAC was defined as the smallest concentration of each aggregating agent which gave irreversible pldtelet aggregation starting within 3 min of addition of the aggregating agent to PBP. At that time a portion of each sample was pipetted off, diluted in an equal volume 0: cold ethanol, spun and the supernatant used for thromboxane B2 (TxB2) radioimmunoassay as described by Ingermanh'ojensket al. (7). Highly specific rabbit antibodies against TxB2 were provided by J.B. Smith (Cardeza Foundation, Thomas Jefferson University, Philadelphia, Pa., USA). Cross reactivity of this antibodies with arachidonic acid was less than 0.01% and with other prostanoid compounds less than 0.02%. The remainder of the sample was processed for MDA spectrophotometric assay as previously described (8). RESULTS In vitro studies. Indomethacin (0.5-5 p) induced concentratim--__-----_-----dependent inhibition of platelet aggregation stimulated by 0.4-0.6 mM AA. At none of the concentrations tested was indomethacin able to prevent the inhibitory effect of aspirin, the latter used at concentrations which induced variable degrees of inhibition (40-100 PM). Fig. 1 reports a representative experiment showing lack of prevention by indomethacin (at the maximal ineffective concentration, 0.6 PM) of the inhibitory effect of aspirin (at the minimal concentration causing complete inhibition of aggregation,75 PM). Additivity rather than prevention of inhibition was found when indomethacin and aspirin were combined at concentrations inducing partial inhibition of platelet aggregation (data not shown). In contrast preincubation of platelet with 1 mM sodium salicylate (which did not modify AA-induced aggregation) resulted in complete prevention of the inhibitory effect of either 5 M indomethacin or 75 ?M tr aspirin. Ex vivo studies. Fig. 2 reports platelet aggregation tracings T__--_----_---obtained in one volunteer given either 50 mg indomethacin or 500 mg aspirin or both drugs (30 min apart). Two hours after administra-tion of either the single drugs or the combined schedule, platelet aggregation induced by 0.5 mM sodium arachidonate was completely blocked. Platelet aggregation still appeared completely inhibited 48 h after aspirin but had returned to normal after indomethacin. Prolongation of the latent period but almost normal platelet aggregation was observed 48 h after ingestion of both drugs together. A similar pattern of platelet aggregation was seen when collagen was used as aggregating agentingead of sodium arachidonate. Table I summarizes the results with the two stimuli in the three volunteers studied. The threshold aggregating concentration (as defined above) of sodium arachidonate and collagen was determined for each of them before starting the drug study. In all instances no aggregation
JUNE
1982 VOL. 23 NO. 6
PROSTAGLANDINS
I
ASA INDO+ASA
C or
INDO
INDO
SAL + INDO C or SAL
1 min
Figure 1. Representative tracings of platelet aggregation induced by sodium arachidonate (0.4 mM, final cont.). PRP was preincubated with 75 PM aspirin (ASA), 0.6 p (upper panel) and 5 PM (middle panel) indomethacin (INDO), 1 mM sodium salicylate (SAL) or a combination of the drugs before addition of sodium arachidonate as indicated. C = Controls. For further details see Materials and Methods.
790
JUNE
1982 VOL. 23 NO. 6
PROSTAGLANDINS
_0
40
2 HOURS
AFTER
INGESTION
Figure 2. Representative tracings of platelet aggregation induced by sodium arachidonate (0.5 mM, final cone.) in PRP from a healthy volunteer. Blood was collected before and at various intervals after ingestion of indomethacin (50 mg), aspirin (500 mg) or both drugs (30 min apart) as indicated. For further details see Materials and Methods.
could be obtained by raising the concentration of sodium arachidonate to 1.6 mM and that of collagen to 4 pg/ml in the samples collected two hours after single or combined drug administration. After 48 h the TACs reverted toward basal values after indomethacin, but the inhibitory effect of aspirin persisted apparently unmodified. Administration of indomethacin before aspirin resulted 48 h later in almost complete recovery of the platelet aggregation response to either stimulus. Platelet cycle-oxygenase activity was measured both as MDA and TxB2 generated on incubation of PRP samples with sodium arachidonate. The results, summarized in Fig. 3, show that both products of arachidonic acid metabolism were undetectable 2 h after administration of single or combined drugs. They remained undetectable 48 h
JUNE
1982 VOL. 23 NO.
b
PROSTAGLANDINS
Table I. Threshold aggregating concentrations of sodium arachidonate and collagen determined before and at intervals after oral administration of indomethacin (50 mg), aspirin (500 mg) or both drugs (30 min apart).Individual values for the three subjects studied. Sodium arachidonate (ml.0
Drug
Collagen @g/ml)
0.7 - 0.5 - 0.9
1.0 - 1.5 - 2.0
Indomethacin
2h 48 h
>1.6 ->1.6 -)1.6 0.9 - 0.6 - 0.9
> 4.0 -)4.0 ->4.0 1.0 - 1.3 - 1.5
Aspirin
2h 48 h
p1.6 ->1.6 -91.6 F1.6 ->1.6 ->1.6
J4.0 -)4.0 -)4.0 > 4.0 ->4.0 ->4.0
Indomethacin + 2 h Aspirin 48 h
)1.6 ->1.6 ->1.6 1.0 - 1.0 - 1.0
p4.0 ->4.0 ->4.0 1.0 - 1.5 - 1.5
after aspirin but averaged 60-80% of basal values 48 h after indomethacin alone or indomethacin plus aspirin. DISCUSSION This study shows that indomethacin prevents the in vivo inhibitory effect of aspirin on human platelet cycle-oxygenase activity. This was measured by platelet aggregation induced by sodium arachidonate or collagen and by the formation of MDA and TxB2, two metabolites of arachidonic acid. Although interaction of indomethacin with aspirin has already been reported in ram seminal vesicles in vitro (4) and in rat platelets in vivo (5), this is the first demonstration of this kind of interxin human volunteers in vivo. Previous in vitro studies onhuman (9) or washed rabbit platelets (2) failed to demonstrate any interaction between indomethacin and aspirin. In the present study we were also unable to show any prevention by indomethacin of aspirin inhibition of arachidonic acid-induced aggregation of human platelets in vitro. Additivity rather than competition between the two drugs was observed. In the same experimental system we found that sodium salicylate, although apparently inactive by itself, did prevent the inhibitory effect of either indomethacin or aspirin. This finding could indicate that all three drugs interact with the same site on plateIt has recently been suggested that NSAIDs let cycle-oxygenase. interact with cycle-oxygenase at a site related to but distinct from the active substrate site (4,5). The various drugs may have different affinities for each site. Salicylate in particular may have a greater affinity for the binding than the active site, whereas indomethacin shows similar affinity for both sites.
792
JUNE
1982 VOL. 23 NO. 6
PROSTAGLANDINS
El
TxE12
100
INDOMETHACIN
100
. LA.
ASPIRIN
J-:
0
INDOMETHACIN + ASPIRIN
l
2
40
HOURS
AFTER
INGESTION
Figure 3. Cycle-oxygenase activity (measured as MDA and TxB2 generation induced by threshold concentrations of sodium arachidonate) Individual values (0) and in PW from three healthy volunteers. means (-) are reported as percentages of basal values. Blood was collected before and at various intervals after ingestion of indomethacin (50 mg), aspirin (500 mg) or both drugs (30 min apart) as indicated. For further details see Materials and Methods.
If one accepts the above enzyme model, indomethacin appears only to interact with the binding site at concentrations inhibiting the enzyme's activity, while salicylate binds to the enzyme at concentrations not affecting its function. This could explain the results of our in vitro studies in which enzyme activity was measured: at effective concentrations of indomethacin, additivity with aspirin was found. At ineffective concentrations, indomethacin did not prevent aspirin's effect. In contrast, the relatively high concentrations of salicylate used, leaving the eneyme activity unaltered but interacting with its binding site, stopped aspirin gaining
JUNE
1982 VOL. 23 NO. 6
793
PROSTAGLANDINS
access to the latter, and consequently prevented the inhibitory effect of this drug. Our interpretation is supported by the fact that the only report of interaction between indomethacin and aspirin in vitro was based on measurements of enzyme acetylation by radiolabelled aspirin rather than enzyme activity (3). The concentration of indomethacin used in the latter study inhibited cycle-oxygenase activity by more than 90%, so that no interaction with'aspirin would have been observed if enzyme activity rather than enzyme acetylation had been measured. We therefore devised an in vivo experiment which took advantage of the short-lasting effect ofmethacin and the persistent inhibition of plateyet cycle-oxygenase activity by aspirin (10). Indomethacin was administered 30 min before aspirin and blood was collected 2 and 48 h later. After 2 h platelet aggregation induced by sodium arachidonate or collagen and arachidonic acid metabolism was completely blocked, and the inhibitory effect of indomethacin alone had almost completely disappeared after 48 h. At 48 h platelet cycle-oxygenase from volunteers given aspirin alone was still completely inhibited. As shown in Figs. 2 and 3 and in Table I, indomethacin pretreatment almost entirely prevented the aspirin effect 48 h later. These results suggest that the interaction between indomethacin and aspirin may have been masked in our in vitro studies and in previous reports (2,9) by the inhibitory activity of indomethacin itself. The indomethacin-aspirin interaction observed in vivo and the interaction of either drug with sodium salicylate in support the concept that the three drugsdnteract with the same site on cycle-oxygenase. Since salicylate is inactive by itself on this platelet enzyme, it may well interact with aspirin or indomethacin at the putative supplementary binding site, rather than directly on the substrate active site. Consequently it has been proposed that indomethacin and aspirin interact with the same site, a phenomenon essential for their action as cycle-oxygenase inhibitors (4,5). The mechanism of action of both drugs would therefore be the same. Acetylation by aspirin of an aminoacid critical to the action of cycle-oxygenase (11) would be a secondary effect, which renders this drug's inhibitory effect irreversible rather than actually inducing it (5,10,12,13). The possibility that the interaction observed in the present study occurred at a level different from cycle-oxygenase is unlikely since, as already mentioned, the interaction between indomethacin and aspirin was shown using a purified cycle-oxygenase enzyme preparation (3). Moreover in a study on rheumatic patients there was no difference in serum levels of salicylate after oral administration of aspirin alone and after a concurrent oral dose of indomethacin (14). Thus the possibility of altered aspirin absorption by indomethacin in the gastrointestinal tract,although cannot be ruled out with certainty,does not appear to explain the reduced effect of aspirin observed in our study. Finally, prevention by indomethacin of aspirin binding to plasma proteins would have resulted in increased availability of free
JUNE
1982 VOL. 23 NO. 6
PROSTAGLANDINS
aspirin for the platelet enzyme and thus in an increased inhibitory effect. ACKNOWLEDGEMENTS This work was supported by the Italian National Research Council (Program "Farmacologia Clinica e Malattie Rare"). Judith Baggott, Anna Mancini, Antonella Bottazzi and Vincenzo and Felice de Ceglie helped prepare the manuscript. M.L. is a Fellow of the Italian National Research Council. REFERENCES Roth, G.J., N. Stanford, and P.W. Majerus. Acetylation of prostaglandin synthase by aspirin. Proc. Natl. Acad. Sci. USA -72: 3073, 1975.
2)
Vargaftig, B.B. The inhibition of cycle-oxygenase of rabbit platelets by aspirin is prevented by salicylic acid and by phenanthrolines. 'Eur. J. Pharmacol. -50: 231, 1978.
3)
Stanford, N., G.J. Roth, T.Y. Shen, and P.W. Majerus. Lack of covalent modification of prostaglandin synthetase (cyclooxygenase) by indomethacin. Prostaglandins -13: 669, 1977.
4)
Humes, J.L., C.A. Winter, S.J. Sadowski, and F.A. Kuehl Jr. Multiple sites on prostaglandin cyclooxygenase are determinants in the action of nonsteroidal antiinflammatory agents. Proc. Natl. Acad. Sci. USA -78: 2053, 1981.
5)
Cerletti, C., M. Livio , and G. de Gaetano. Non-steroidal antiinflammatory drugs react with two sites on platelet cyclooxygenase. Evidence from -in vivo drug interaction studies in rats. Biochim. Biophys. Acta 714: 122 (1982).
6)
Silver, M.J., W. Hoch, J.J. Kocsis, C.M. Ingerman, and J.B. Smith. Arachidonic acid causes sudden death in rabbits. Science 183: 1085, 1974.
7)
Ingerman-Wojenski, C., M.J. Silver, J.B. Smith, and E. Macarak. Bovine endothelial cells in culture produce thromboxane as well as prostacyclin. J. Clin. Invest. -67: 1292, 1981.
8)
Rajtar, G., C. Cerletti, M. Livio, and G. de Gaetano. Sulphinpyrazone preventsin -- vivo the inhibitory effect of aspirin on rat platelet cycle-oxygenase activity. Biochem. P'narmacol. -30: 2773, 1981.
9)
Zucker, M.B., and J. Peterson. Effect of acetylsalicylic acid, other nonsteroidal anti-inflammatory agents, and dipyridamole on human blood platelets. J. Lab. Clin. Med. -76: 66, 1970.
JUNE
1982 VOL. 23 NO. 6
795
PROSTAGLANDINS
10)
O'Brien, J.R. Effect of anti-inflammatory Lancet -1: 894, 1968.
11)
Roth, G.J., and C.J. Siok. Acetylation of the NH2-terminal serine of prostaglandin synthetase by aspirin. J. Biol. Chem. 253: 3782, 1978.
12)
Merino, J., M. Livio, G. Rajtar, and G. de Gaetano. Salicylate reverses in vitro aspirin inhibition of rat platelet and vascular prostagla;;;?iGeration. Biochem. Pharmacol. -29: 1093, 1980.
13)
Livio, M., G. Rajtar, J. Merino, and G. de Gaetano. Malondialdehyde formation in rat platelet-rich plasma. II. Modification of the reaction kinetics by aspirin and indomethacin. Thromb. Haemost. -44: 52, 1980.
14)
Kaldestad, E., T. Hansen, and H.K. Brath. Interaction of indomethacin and acetylsalicylic acid as shown by the serum concentrations of indomethacin and salicylate. Eur. J. Clin. Pharmacol. -9: 199, 1975.
agents on platelets.
Robert R. Gorman Editor: Received: l-27-82 4-15-82 Accepted:
796
JUNE
1982 VOL. 23 NO. 6