Ivana Vojnovic

Cellular mechanisms of acetaminophen: role of cyclo-oxygenase

Acetaminophen is one of the most commonly used drugs for the safe and effective treatment of pain and fever. Acetaminophen works by lowering cyclo‐oxygenase products preferentially in the central nervous system, where oxidant stress is strictly limited. However, the precise mechanism of action for acetaminophen on cyclo‐oxygenase activity is debated. Two theories prevail. First, it is suggested that acetaminophen selectively inhibits a distinct form of cyclo‐oxygenase, cyclo‐oxygenase‐3. Second, it is suggested that acetaminophen has no affinity for the active site of cyclo‐oxygenase but instead blocks activity by reducing the active oxidized form of cyclo‐oxygenase to an inactive form. Here, we have used an in vitro model of cyclo‐oxygenase‐2 activity (A549 cells stimulated with IL‐1β) to show that acetaminophen is an effective inhibitor of cyclo‐oxygenase activity in intact cells. However, acetaminophen, unlike nonsteroidal anti‐inflammatory drugs (NSAIDs), cannot inhibit activity in broken cell preparations. The inhibitory effects of acetaminophen were abolished by increasing intracellular oxidation conditions with the cell‐permeable hydroperoxide t‐butylOOH. Similarly the inhibitory effects of the cyclo‐oxygenase‐2 selective inhibitor rofecoxib or the mixed cyclo‐oxygenase‐1/cyclo‐oxygenase‐2 inhibitors ibuprofen and naproxen were significant reduced by t‐butylOOH. By contrast, the inhibitory effects of indomethacin or diclofenac, which also inhibit both cyclo‐oxygenase‐1 and cyclo‐oxygenase‐2, were unaffected by t‐butylOOH. These observations dispel the notion that cyclo‐oxygenase‐3 is involved in the actions of acetaminophen and provide evidence that supports the theory that acetaminophen interferes with the oxidation state of cyclo‐oxygease. Moreover, they suggest for the first time that the inhibitory effects of some NSAIDs, including the newly introduced cyclo‐oxygenase‐2 selective inhibitor rofecoxib, owe part of their inhibitory actions to effects on oxidation state of cyclo‐oxygenase. Our data with t‐butylOOH and NSAIDs illustrates an, as yet, undeveloped therapeutic window for the “cyclo‐oxygenase inhibitor”. Specifically, combining active site selectively with actions on enzyme oxidation state would allow for a broader range of tissue selective drugs.

A771726, the active metabolite of leflunomide, directly inhibits the activity of cyclo-oxygenase-2 in vitro and in vivo in a substrate-sensitive manner

The immunosuppressive and anti‐inflammatory drug leflunomide has several sites of action, although its precise mode of action is unknown. Here we show in vitro and in vivo that leflunomide and/or its active metabolite A771726, inhibit the activity of cyclo‐oxygenase (COX) at doses below those that affect protein expression. In J774.2 macrophages treated with endotoxin for 24 h to induce COX‐2 and iNOS, leflunomide and A771726 inhibited more potently the accumulation of PGE2 (A771726, IC50 3.5 μg ml−1) than of NO2 (A771726, IC50 380 μg ml−1). At high concentrations (>300 μg ml−1) A771726 also exhibited the expression of COX‐2 and iNOS proteins. In A549 cells treated for 24 h with interleukin‐1β, to induce COX‐2, A771726 potently inhibited PGE2 synthesis (IC50 0.13 μg ml−1). In the same cells, A771726 was notably less active (IC50, 52 μg ml−1) at inhibiting the formation of PGE2 stimulated by exposure to 30 μM arachidonic acid. In a human whole blood assay, measuring the accumulation of TxB2 in response to calcium ionophore as a measure of COX‐1 activity and in response to incubation with bacterial endotoxin as a measure of COX‐2 activity, leflunomide inhibited COX‐1 and COX‐2 with IC50 values of 31 and 185 μg ml−1; for A771726 the corresponding values were 40 and 69 μg ml−1. Pre‐treatment of rats with leflunomide or A771726 (10 mg kg−1, i.p.) inhibited the plasma accumulation of 6‐keto‐PGF1α but not NO2/NO3 following infusion of endotoxin. Injection of a bolus of arachidonic acid following 6 h infusion of endotoxin caused a marked acute rise in plasma 6‐keto‐PGF1α which was inhibited only by higher doses of A771726 (50 mg kg−1, i.p.). In conclusion, leflunomide via A771726 can directly inhibit the activity of COX, an effect that appears blunted both by increases in substrate supply and possibly by plasma binding. Only at much higher drug levels does leflunomide and/or A771726 inhibit the induction of COX‐2 or iNOS proteins.

Stronger inhibition by nonsteroid anti-inflammatory drugs of cyclooxygenase-1 in endothelial cells than platelets offers an explanation for increased risk of thrombotic events

Recent data have suggested that regular consumption of nonsteroid anti‐inflammatory drugs (NSAIDs), particularly selective inhibitors of cyclo‐oxygenase‐2 (COX‐2), is associated with an increased risk of thrombotic events. It has been suggested that this is due to NSAIDs reducing the release from the endothelium of the antithrombotic mediator prosta‐glandin I2 as a result of inhibition of endothelial COX‐2. Here, however, we show that despite normal human vessels and endothelial cells containing cyclo‐oxygenase‐1 (COX‐1) without any detectable COX‐2, COX‐1 in vessels or endothelial cells is more readily inhibited by NSAIDs and COX‐2‐selective drugs than COX‐1 in platelets (e.g., log IC50±SEM values for endothelial cells vs. platelets: naproxen −5.59±0.07 vs. −4.81±0.04; rofecoxib–4.93±0.04 vs. −3.75±0.03; n=7). In broken cell preparations, the selectivities of the tested drugs toward endothelial cell over platelet COX‐1 were lost. These observations suggest that variations in cellular conditions, such as endogenous peroxide tone and substrate supply, and not the isoform of cyclo‐oxygenase present, dictate the effects of NSAIDs on endothelial cells vs. platelets. This may well be because the platelet is not a good representative of COX‐1 activity within the body as it produces prostanoids in an explosive burst that does not reflect tonic release from other cells. The results reported here can offer an explanation for the apparent ability of NSAIDs and COX‐2‐selective inhibitors to increase the risk of myocardial infarction and stroke.—Mitchell, J. A., Lucas, R., Vojnovic, I., Hasan, K., Pepper, J. R., Warner, T. D. Stronger inhibition by nonsteroid anti‐inflammatory drugs of cyclooxygenase‐1 in endothelial cells than platelets offers an explanation for increased risk of thrombotic events. FASEB J. 20, 2468–2475 (2006)

Aspirin and the in vitro linear relationship between thromboxane A2-mediated platelet aggregation and platelet production of thromboxane A2

Summary.  Background: Currently, ‘aspirin resistance’, the anti‐platelet effects of non‐steroid anti‐inflammatory drugs (NSAIDs) and NSAID‐aspirin interactions are hot topics of debate. It is often held in this debate that the relationship between platelet activation and thromboxane (TX) A2 formation is non‐linear and TXA2 generation must be inhibited by at least 95% to inhibit TXA2‐dependent aggregation. This relationship, however, has never been rigorously tested. Objectives: To characterize, in vitro and ex vivo, the concentration‐dependent relationships between TXA2 generation and platelet activity. Method: Platelet aggregation, thrombi adhesion and TXA2 production in response to arachidonic acid (0.03–1 mmol L−1), collagen (0.1–30 μg mL−1), epinephrine (0.001–100 μmol L−1), ADP, TRAP‐6 amide and U46619 (all 0.1‐30 μmol L−1), in the presence of aspirin or vehicle, were determined in 96‐well plates using blood taken from naïve individuals or those that had taken aspirin (75 mg, o.d.) for 7 days. Results: Platelet aggregation, adhesion and TXA2 production induced by either arachidonic acid or collagen were inhibited in concentration‐dependent manners by aspirin, with logIC50 values that did not differ. A linear relationship existed between aggregation and TXA2 production for all combinations of arachidonic acid or collagen and aspirin (P < 0.01; R2 0.92; n = 224). The same relationships were seen in combinations of aspirin‐treated and naïve platelets, and in blood from individuals taking an anti‐thrombotic dose of aspirin. Conculsions: These studies demonstrate a linear relationship between inhibition of platelet TXA2 generation and TXA2‐mediated aggregation. This finding is important for our understanding of the anti‐platelet effects of aspirin and NSAIDs, NSAID–aspirin interactions and ‘aspirin resistance’.

Effects of non-steroidal anti-inflammatory drugs on cyclo-oxygenase and lipoxygenase activity in whole blood from aspirin-sensitive asthmatics vs healthy donors

Cyclo‐oxygenase (COX) and lipoxygenase (LO) share a common substrate, arachidonic acid. Aspirin and related drugs inhibit COX activity. In a subset of patients with asthma aspirin induces clinical symptoms associated with increased levels of certain LO products, a phenomenon known as aspirin‐sensitive asthma. The pharmacological pathways regulating such responses are not known. Here COX‐1 and LO activity were measured respectively by the formation of thromboxane B2 (TXB2) or leukotrienes (LT) C4, D4 and E4 in whole blood stimulated with A23187. COX‐2 activity was measured by the formation of prostaglandin E2 (PGE2) in blood stimulated with lipopolysaccharide (LPS) for 18 h. No differences in the levels of COX‐1, COX‐2 or LO products or the potency of drugs were found in blood from aspirin sensitive vs aspirin tolerant patients. Aspirin, indomethacin and nimesulide inhibited COX‐1 activity, without altering LO activity. Indomethacin, nimesulide and the COX‐2 selective inhibitor DFP [5,5‐dimethyl‐3‐(2‐isopropoxy)‐4‐(4‐methanesulfonylphenyl)‐2(5H)‐furanone] inhibited COX‐2 activity. NO‐aspirin, like aspirin inhibited COX‐1 activity in blood from both groups. However, NO‐aspirin also reduced LO activity in the blood from both patient groups. Sodium salicylate was an ineffective inhibitor of COX‐1, COX‐2 or LO activity in blood from both aspirin‐sensitive and tolerant patients. Thus, when COX activity in the blood of aspirin‐sensitive asthmatics is blocked there is no associated increase in LO products. Moreover, NO‐aspirin, unlike other NSAIDs tested, inhibited LO activity in the blood from both aspirin sensitive and aspirin tolerant individuals. This suggests that NO‐aspirin may be better tolerated than aspirin by aspirin‐sensitive asthmatics.

Influence of plasma protein on the potencies of inhibitors of cyclooxygenase-1 and -2

It is widely believed that the potencies of nonsteroid anti‐inflammatory drugs (NSAIDs) as inhibitors of cyclooxygenase (COX) are influenced by protein binding in the extracellular fluid, since NSAIDs are bound to circulating albumin by well over 95%. This is an important point because the protein concentrations in synovial fluid and the central nervous system, which are sites of NSAID action, are markedly different from those in plasma. Here we have used a modified whole‐blood assay to compare the potencies of aspirin, celecoxib, diclofenac, indomethacin, lumiracoxib, meloxicam, naproxen, rofecoxib, sodium salicylate, and SC560 as inhibitors of COX‐1 and COX‐2 in the presence of differing concentrations of protein. The potencies of diclofenac, naproxen, rofecoxib, and salicylate, but not aspirin, celecoxib, indomethacin, lumiracoxib, meloxicam, or SC560, against COX‐1 (human platelets) increased as protein concentrations were reduced. Varying protein concentrations did not affect the potencies of any of the drugs against COX‐2, with the exception of sodium salicylate (A549 cells). Clearly, our findings show that the selectivity of inhibitors for COX‐1 and COX‐2, which are taken to be linked to their efficacy and side effects, may change in different extracellular fluid conditions. In particular, selectivity in one body compartment does not demonstrate selectivity in another. Thus, whole‐body safety or toxicity cannot be linked to one definitive measure of COX selectivity.

Blockade of the purinergic P2Y12 receptor greatly increases the platelet inhibitory actions of nitric oxide

Circulating platelets are constantly exposed to nitric oxide (NO) released from the vascular endothelium. This NO acts to reduce platelet reactivity, and in so doing blunts platelet aggregation and thrombus formation. For successful hemostasis, platelet activation and aggregation must occur at sites of vascular injury despite the constant presence of NO. As platelets aggregate, they release secondary mediators that drive further aggregation. Particularly significant among these secondary mediators is ADP, which, acting through platelet P2Y12 receptors, strongly amplifies aggregation. Platelet P2Y12 receptors are the targets of very widely used antithrombotic drugs such as clopidogrel, prasugrel, and ticagrelor. Here we show that blockade of platelet P2Y12 receptors dramatically enhances the antiplatelet potency of NO, causing a 1,000- to 100,000-fold increase in inhibitory activity against platelet aggregation and release reactions in response to activation of receptors for either thrombin or collagen. This powerful synergism is explained by blockade of a P2Y12 receptor-dependent, NO/cGMP-insensitive phosphatidylinositol 3-kinase pathway of platelet activation. These studies demonstrate that activation of the platelet ADP receptor, P2Y12, severely blunts the inhibitory effects of NO. The powerful antithrombotic effects of P2Y12 receptor blockers may, in part, be mediated by profound potentiation of the effects of endogenous NO.

Endothelin in human inflammatory bowel disease: comparison to rat trinitrobenzenesulphonic acid-induced colitis

There have been suggestions that endothelins (ET-1, ET-2, ET-3) are involved in the pathogenesis of human inflammatory bowel disease (IBD). Furthermore, the non-selective endothelin receptor antagonist, bosentan, ameliorates colonic inflammation in TNBS colitis in rats. However, no studies have measured the tissue expression and release of endothelins in human IBD in direct comparison to experimental TNBS colitis. Mucosal biopsies were obtained from 114 patients (42 Crohns colitis, 35 ulcerative colitis and 37 normal) and compared to whole colonic segments from rats with TNBS colitis. ET-1/2 levels were reduced in human IBD but greatly increased in experimental TNBS colitis. RT-PCR indicated ET-2 was the predominant endothelin isoform in human IBD whereas ET-1 prevailed in the TNBS model. No associations were found between human IBD and tissue expression, content or release of ET-1/2. Our study shows, therefore, that unlike TNBS colitis in rats, in which ET-1/2 levels are greatly elevated and ET receptor antagonists are efficacious, there is no significant link between endothelins and human IBD.

Heparin but not citrate anticoagulation of blood preserves platelet function for prolonged periods

Summary.  Background: Current guidelines state that platelet aggregation studies should be conducted within 4 h of venepuncture because of the decline in sensitivity to platelet agonists. This constrains studies of platelet activity in clinical situations where samples need to be transported or there are unavoidable delays prior to assessment. Objectives: The aim of the present study was to compare systematically the responses of platelets stored in the presence of either citrate or heparin, the two most widely used anti‐coagulants, using a range of standard techniques. Methods: Blood was taken from healthy volunteers and either assessed immediately or stored at ambient temperature (18–25 °C) for 24 h. Platelet reactivity to a range of agonists was determined by a combination of 96‐well plate techniques; light transmission aggregometry, thrombi adhesion, ATP and ADP release, and TxA2 release; by whole blood aggregometry; and by PFA‐100. Results and conclusions: Testing using 96‐well plate techniques allowed for the simultaneous measurement of responses to multiple concentrations of multiple agonists. The responses of platelets from blood anti‐coagulated with heparin were predominantly preserved in all assays after 24 h storage, whereas, responses of platelets stored in blood anti‐coagulated with citrate were greatly diminished. Consequently, anti‐coagulation with heparin, but not citrate, preserves platelet responses for up to 24 h as determined by a range of techniques.

The effects and metabolic fate of nitroflurbiprofen in healthy volunteers

Nitric oxide–donating nonsteroidal anti‐inflammatory drugs (NO‐NSAIDs) are a new class of cyclooxygenase (COX) inhibitors. To investigate whether these drugs actually release nitric oxide (NO), we labeled the nitroxy group of nitroflurbiprofen with nitrogen 15 to determine the metabolic fate of this compound in humans.