Flurbinitroxybutylester: A novel anti-inflammatory drug has enhanced antithrombotic activity

llmmbosis

Pergamon

Research, Vol. 79, No. 1, pp. 7Ml. 1995 Coovrieht 0 1995 Elsevier Science Ltd F&k-d-in the USA. All rights reserved 0049-3848195 $9.50 + .w

0049-3848(95)00092-5

FLURBINITROXYBUTYLESTER: DRUG HAS ENHANCED

A NOVEL ANTI-INFLAMMATORY ANTITHROMBOTIC ACTIVITY

Giuseppe Cirino, Carla Cicala, Franca Mancuso, Anwar R. Baydoun’

and JohnL.

Wallace2

Department of Experimental Pharmacology, via Domenico Montesano 49, 80131 Naples, Italy. ‘Biology Research Centre, Biomedical Sciences Division, King’s College, Campden Hill Road, London W8 7AH, U.K. *Department of Pharmacology and Therapeutics, University of Calgary, Calgary, Alberta T2N 4Nl; Canada

(Received 13 December 1994 by Editor G. De Gaetano; revised/accepted

Abstract

24 April 1995)

We have recently shown that the introduction of a nitroxybutylester moiety into flurbiprofen, to form Flurbi-NO, results in a compound with markedly reduced undesired effects in the gastrointestinal tract. This effect has been shown to be linked to nitric oxide release from the Flurbi-NO. Here we have investigated whether this is associated with a reduction in platelet aggregability in vivo, as assessed in a mouse model of thromboembolism and a rat model of platelet aggregation, and found in both models that Flurbi-NO is more potent than flurbiprofen at inhibiting collagen-induced platelet aggregation. Further in vitro studies using human washed platelets and cells in culture suggest that this is due to the release of NO from Flurbi-NO following the action of (possibly plasma) esterases. Together with our earlier data, these results strongly suggest that FlurbiNO and other members of this class of drugs, have particular potential as antithrombotic agents devoid of gastrointestinal side effects,

Key words: NSAID, Nitric Oxide, Cyclooxygenase, thrombosis Corresponding author: Dr. Giuseppe Cirino, Department of Experimental Domenico Montesano 49, 80 13 1 Naples Italy

73

Pharmacology,

via

74

ANTITHROMBOTIC

DRUG

Vol. 79, No. 1

Platelet aggregation and adhesion within the blood vessel is promoted by the platelet production of thromboxane A,. At the same time it is limited by the endothelial and platelet formation of nitric oxide (NO), and by the release of prostacyclin from the endothelium. Changes in the balance of this mediator production explain the actions of drugs that influence platelet adhesion and aggregation. Thus, non steroidal anti-inflammatory drugs (NSAID) potently inhibit aggregation, via their ability to inhibit cyclooxygenase and so depress the production of thromboxane AZ. Similarly, platelet aggregation and adhesion is reduced by agents such as sodium nitroprusside, that stimulate platelet guanylate cyclase, or compounds that boost the activity of the NO system, such as L-arginine. In concert with this, L-NMMA, or other inhibitors of the synthesis of NO, tend to enhance platelet aggregation (l-3). Theoretically, therefore, a compound that inhibits the production of thromboxane A, while at the same time stimulating guanylate cyclase, should have marked potential as an anti-platelet agent. We have previously shown that introduction of the nitroxybutylester moiety onto the carboxyl group of the NSAID flurbiprofen, to form Flurbi-NO, produces a compound with markedly reduced adverse gastrointestinal effects but the same beneficial biological activity (4). As FlurbiNO, unlike its progenitor, has the potential to influence platelet aggregation through both the thromboxane A, and NO pathways, we have tested the hypothesis that Flurbi-NO may have an enhanced anti-thrombotic effect compared to flurbiprofen.

MATERIALS

AND METHODS

Reagents Flurbiprofen, porcine esterase, collagen (type 6) and epinephrine hydrochloride, were purchased from Sigma, St. Louis. Collagen (Kollagenreagens Horm) for use in the mouse anti-thrombotic assay was obtained from Hormon Chemie (Munchen, Germany). Flurbinitroxybutylester (FlurbiNO) was a generous gift of Dr. Piero de1 Soldato of Nicox. (Dublin). Flurbiprofen and FlurbiNO (figure 1) were dissolved in a minimum quantity of either dimethylsulfoxide or methanol (20100 ul) to obtain a stock solution which was diluted to a suspension with the appropriate buffer.

CH - COO(CH,) 4- ONO,

FIG. 1. Structure of flurbinitroxybutylester

Vol. 79, No. 1

ANTITHROMBOTIC

DRUG

75

Statistical analysis Data were analyzed for statistical significance by using ANOVA for multiple comparisons followed by Bonferroni correction. Mortality data were analyzed by using Fisher’s exact test. A value of p < 0.05 was taken as significant. Platelet aggregation in vivo The ability of flurbiprofen or Flurbi-NO to inhibit platelet aggregation was assessed using an in vivo model, as described by Pinon (5). Croups of 5 male Wistar rats (200-250 g), permitted water ad libitum, were deprived of food 18-20 hours after which time they were given, p.o., flurbiprofen (2.5, 15 or 20 m&g), Flurbi-NO (5, 20 or 30 mgikg ), or vehicle. One hour later, the rats were anaesthetized with 10% urethane (1 g/kg, i.p.) and the left jugular vein and right carotid artery were cannulated. Collagen (type 6, Sigma) was then administered intravenously at a dose of 2 mgkg. Three minutes later, 2 blood samples (A and B) were collected from the carotid artery into 2.5 ml disposable syringes, as follows: sample A, 0.4 ml blood into 1.6 ml EDTA/formalin buffer (EDTA tetrasodium salt 24 mM, KH2P0, 1.3 mM, NaH,PO, 5.4 mM, formalin [40% sol] 2.5% v/v); sample B, 0.4 ml blood into 1.6 ml EDTA buffer (EDTA tetrasodium salt 24 mM, KI$PO, 3.3 mM, N%HPO, 13.4 mM). Samples were then transferred to 5 ml polystyrene tubes and allowed to stand for 15 min at room temperature. After this time the platelet aggregates in sample A had become fixed by the formalin, whereas those in sample B were dissociated and the platelets made unresponsive by EDTA. Samples were than centrifuged for 10 minutes at 200 g and the platelet count in the supernatant of each sample was then assessed using light microscopy by an observer who was unaware of the treatment the rats had received. The count from sample B represented total platelet number, whereas that from sample A represented only the non-aggregated platelets. Results were then expressed as percent of aggregation, calculated as follows: ([l - (Platelet count in sample A) / (Platelet count in sample B) ] x 100) (5).

Mouse anti-thrombotic assay Mice (male Swiss, 15-20 g, Charles River) were fasted overnight and then treated, p.o., with flurbiprofen, Flurbi-NO (5, 10 or 20 mgkg, for both) or vehicle. l-3 hours later the mice (each group, n=l O-1 5) were injected into the tail vein with 0.1 ml of a mixture of collagen (150 ug/ml) plus epinephrine hydrochloride (100 uM) diluted in 0.154 M sodium chloride solution. The injection of this mixture causes death within 3 minutes in > 90 % of control animals (6). Platelet aggregation in vitro Platelets were obtained from healthy human volunteers and isolated as described previously (7). Platelet aggregation was recorded using a Payton platelet aggregometer (model 300 BD-5). Suspensions of platelets (0.4 ml, containing lO*platelets) were preincubated at 37 “C for 10 min in the presence of flurbiprofen, Flurbi-NO (0.001 to 500 PM, for both) or vehicle (methanol, 20 ~1). The platelets were then stimulated by addition of collagen (10 ug/ml) and aggregation monitored for 15 min thereafter. In an additional series of experiments, platelets were preincubated for 10 min with flurbiprofen, flurbi-NO (0.1 uM, for both) or vehicle in the presence or absence of esterase (porcine liver, 10 U; Sigma). Aggregation in response to

ANTITHROMBOTIC

76

DRUG

Vol. 79, No. 1

collagen was then examined as in the experiments above. Data were expressed as percent of platelet aggregation relative to the vehicle treated group. Evuluation of the spontaneous release of NO To measure the release of NO from Flurbi-NO and its subsequent conversion to nitrite, Flurbi-NO (100 pg/ml) was incubated (37 “C) in the absence or presence of cultured cells (5774.2 macrophages or bovine aortic endothelial cells) in a final volume of 10 ul of DMSO and 100 ul of culture medium. The medium was removed at 0, 1, 10, 20, 30, 60, 120 and 180 min, and sometimes 24 h, and the concentration of nitrite measured using the Griess reaction (8).

RESULTS Effect on platelet aggregation in vivo Flurbiprofen (15 and 20 mg/kg) and Flurbi-NO (5, 20 and 30 mgkg) significantly inhibited collagen induced platelet aggregation in viva, in dose-dependent manners (calculated IC,, values, flurbiprofen -40 mgkg, Flurbi-NO -10 mgkg; figure 2). Flurbiprofen (2.5 mg/kg) was without significant effect. Interestingly, at equimolar doses Flurbi-NO produced significantly greater antiaggregatory effects than flurbiprofen. Thus, 20 and 30 mg/kg Flurbi-NO were more antiaggregatory than 15 and 20 mg/kg flurbiprofen, respectively (p < 0.01, for both).

501 E

40

‘E 2

30

f! g u 0 e

I

Control

20 10 0L Cont

2.5 5

1520

2030

Dose (mglkg)

FIG.

Effect of flurbiprofen

2.

and Flurbi-NO on collagen induced human platelet aggregation in vivo

ANTITHROMBOTIC

Vol. 79, No. 1

0 2

DRUG

77

Flurbiprofen

50

E z $

25

5 IO Dose (mglkg)

FIG. 3. Effect of flurbiprofen and Flurbi-NO on mice survival. Drugs were administered 3 hours before the chaleenge with collagen plus epinephrine

Effect in mouse anti-thrombotic assay Administration of the mixture of collagen and epinephrine to vehicle treated mice resulted in a high rate of mortality, with only 13% surviving (n=60; figure 3). Retreatment with flurbiprofen 3 hours prior to the challenge increased survival only at the highest dose tested (20 mgkg), whereas Flurbi-NO significantly increased survival at doses of 10 and 20 mg/‘kg (figure 2). These differences were reflected in the calculated IC, values (flurbiprofen -22 mg/kg, Flurbi-NO -13 mg/kg). When flurbiprofen and Flurbi-NO were given 1 hour prior to the challenge with collagen and epinephrine, both drugs elicited significant dose-dependent increases in survival at all doses tested. Thus, Flurbi-NO at doses of 5, 10 and 20 mgkg increased the survival rate to 55% (n=9; p < O.OOOl), 66% (n=15; p
lZfiect on platelet aggregation in vitro Flurbiprofen was more potent than Flurbi-NO at inhibiting the in vitro aggregation of human washed platelets in response to collagen (IC,, values, -4 x 1Om9M and -1 x lo-* M, respectively; figure 4). For instance, 0.01 uM flurbiprofen inhibited platelet aggregation by -55 %, whereas 0.7 uM Flurbi-NO was required to give the same inhibition. Addition of esterase (10 U/ml) enhanced the activity of Flurbi-NO such that at a concentration of 0.1 PM, for example, it was equiactive with flurbiprofen (figure 5). Esterase was without direct effect on platelet aggregation.

78

ANTITHROMBOTIC

DRUG

Vol.

79, No. 1

100

75

0 no estefase Srme&erase

p
-

50

25

0

Flurbipmfen

methanol

FlurbiiNO

FIG. 4 Inhibition

Spontaneous

release

of collagen induced human platelet aggregation by flurbiprofen and Flurbi-NO of‘ NO

No release of nitrite (f 1 QA) from Flurbi-NO was detected when the compound was incubated either in the absence (up to 24 h) or presence (up to 3 h) of celis (n=6 for each). .-2 loo-

a- Fiurbiprofen

0.001

0.01

0.1

1

IO

Dose

pM [log]

100

FIG. 5. Effect of esterase (10 units/ml) on the inhibitory effect of Flurbi-NO collagen induced human platelet aggregation

and flurbiprofen on

Vol. 79, No. 1

ANTITHROMBOTIC

DRUG

DISCUSSION Here we show, using two different assays, that Flurbi-NO has an enhanced anti-thrombotic activity in vivo in comparison to its parent compound, flurbiprofen. Indeed, taking into account the differences in molecular weight (flurbiprofen 264 g, Flurbi-NO 361 g) it is clear that on a molar basis the nitroxybutylester is a far better anti-platelet agent. The difference in potencies between flurbiprofen and Flurbi-NO are not explained by differences in their abilities to inhibit cyclooxygenase, for we have previously shown that Flurbi-NO and flurbiprofen inhibit prostaglandin synthesis similarly in the rat gastric mucosa (4). This difference is also not explained by different selectivities for the two isoforms of cyclooxygenase, as both flurbiprofen and Flurbi-NO inhibit the activities of cyclooxygenase I and II, in intact and broken cell assays (9). Thus, it most probable that the increased anti-thxombotic effect of FlurbiNO is due to the linkage of the nitroxybutylester moiety onto flurbiprofen. Interestingly, when comparing the potencies of flurbiprofen and Flurbi-NO as inhibitors of platelet aggregation in vitro we found that the parent compound was more active than the nitroxybutylester derivative. However, this difference was overcome when Flurbi-NO was tested in the presence of esterase. This suggests that in washed human platelets, where plasma and plasma enzymes have been removed, the metabolic pathways cleaving Flurbi-NO to flurbiprofen are no longer present. This would agree with other in vivo experiments in which we have found that one of the major breakdown products of the closely related diclofenac nitroxybutylester is the alcohol obtained by hydrolysis of the ester bond (10). Thus, it would appear that the full activity of Flurbi-NO is best determined in the presence of these catabolic processes. Indeed, we found that in solution, either in the presence of absence of homogenous cell populations FlurbiNO did not release detectable amounts of nitrite. Clearly, therefore, Flurbi-NO is a relatively stable compound requiring enzymatic processing for its full biological activity to be revealed. It is tempting to speculate that the additional anti-thrombotic effect of Flurbi-NO is explained by its ability to interfere with some other(s) of the complex mechanisms involved in thrombus formation, such as leucocyte adhesion to the endothelium or the interaction between platelets and leucocytes. This suggestion is given strong support by our previous observation that oral administration of flurbiprofen to rats causes an increased adherence of leucocytes in the postcapillary venules whereas Flurbi-NO does not produce this effect (4). This is also entirely in agreement with the suggestion that NO is an important endogenous inhibitor of leucocyte adherence in these vessels (1 l-12). Thus, as it is known that an early event in the pathogenesis of NSAID-induced gastic mucosal injury is the adherence of leukocytes to the endothelium of post capillary venules (1 I), and as leucocytes (polymorphonuclear and monocytes) participate in thrombus generation and in the triggering and amplification of the platelet response (13-14) it is feasible that Flurbi-NO depresses platelet aggregation via its ability to reduce the interactions between leucocytes and platelets. In conclusion, it is clear that the enhanced anti-thrombotic activity of Flurbi-NO together with its lack of gastrointestinal side effects suggests this

80

ANTITHROMBOTIC

DRUG

Vol. 79, No. 1

compound, and other members of this class of drugs, have particular potential as anti-thrombotic agents.

REFERENCES 1. RADOMSKI, M.W., PALMER, R.M.J. and MONCADA S. The anti-aggregating properties of vascular endothelium: interactions between prostacyclin and nitric oxide. Br. J. Pharmacol. 92,639~646 1987. 2. RADOMSKI, M.W., PALMER, R.M.J. and MONCADA S. Comparative pharmacology of endothelium-derived relaxing factor, nitric oxide and prostacyclin in platelets. Br. J. Pharmacol. 92, 181-187, 1987. 3. RADOMSKI, M.W., PALMER, R.M.J. and MONCADA S. An L-arginine/nitric oxide pathway present in human platelets regulates aggregation. Proc. Natl. Acad. Sci. U.S.A.-~, 5193-5197, 1990. 4. WALLACE, J.L., REUTER, B., CICALA, C., McKNIGHT, W., GRISHAM, M. and G. CIRINO. Novel non steroidal anti-inflammatory drug derivatives with markedly reduced ulcerogenic properties in the rat. Gastroenterology 107, 173-179, 1994. 5. PINON, J.F. In vivo study of platelet aggregation in rats. J. Pharm. Methods u,79-84, 1984. 6. DI MINNO, G. and SILVER, M.J. Mouse antithrombotic assay: a simple method for the evaluation of antithrombotic agents in vivo. Potentiation of antithrombotic activity by ethyl alcohol. J. Pharm. Exp. Ther. 225, 57-60 ,1983. 7. RADOMSKI, M. and MONCADA, S. An improved method for washing of human platelets with prostacyclin. Thromb. Res. 3, 383-386, 1983. 8. GROSS, S.S. & LEVI, R. (1992). Tetrahydrobiopterin synthesis: An absolute requirement for cytokine-induced nitric oxide generation by vascular smooth muscle. J. Biol. Chem. m,2572225729. 9. MITCHELL, J.A., CIRINO, G., AKARASEREENONT, P., WALLACE, J.L., FLOWER, R.J. and VANE, J.R. Flurbinitroxybutylester: a novel anti-inflammatory drug devoid of ulcerogenic activity, inhibits cycle-oxygenase-I and cycle-oxygenase-2. Can. J. Physiol. Pharmacol. 22 suppl. 1, 270,1994 10. BENONI, G., ADAMI, A., TERZI, M., GRIGOLINI, L., CUZZOLIN, L., DEL SOLDATO, P. Plasma concentrations and pharmacokinetic parameters of nitrofenac using a simple and sensitive HPLC method. Can. J. Physiol. Pharmacol. 72 suppl. 1, 273, 1994. 11. ASAKO, H., KUBES, P., WALLACE, J.L., GAGINELLA, T., WOLF, R.E., GRANGER, D.N. Indomethacin-induced leukocyte adhesion in mesenteric venules: role of lipoxygenase products. Am. J. Physiol. 262, 4903-4908, 1990. 12. ASAKO, H., KUBES, P., WALLACE, J.L., WOLF, R.E., GRANGER, D.N. Modulation of leukocyte adhesion in rat mesenteric venules by aspirin and salicylate. Gastroenterology @&146152, 1992. 13. DEL MASCHIO, A., EVANGELISTA, V., RAJTAR, G., CHEN, Z.M., CERLETTI, C. and DE GAETANO, G. Platelet activation by polymorphonuclear leukocytes exposed to chemotactic

Vol. 79, No. 1

ANTITHROMBOTIC

DRUG

81

agents. Am. J. Physiol. 2%, H870-H879, 1990. 14. MAUGERI, N., EVANGELISTA, V., PICCARDONI, P., DELL’ELBA, G., CELARDO, A., DE GAETANO, G. and CERLETTI, C. TransceHuIar metaboiism of arachidonic acid: increased platelet thromboxane generation in the presence of activated polymorphonuclear leukocytes. Blood SO, 447-451, 1992.