Mechanism of action of novel NO-releasing furoxan derivatives of aspirin in human platelets

Abstract

Incorporation of a nitric oxide (NO)-releasing moiety in aspirin can overcome its gastric side effects. We investigated the NO-release patterns and antiplatelet effects of novel furoxan derivatives of aspirin (B8 and B7) in comparison to existing antiplatelet agents. Cyclooxygenase (COX) activity was investigated in purified enzyme using an electron paramagnetic resonance-based technique. Concentration-response curves for antiplatelet agents +/- the soluble guanylate cyclase inhibitor, ODQ (50 microM) were generated in platelet-rich plasma (PRP) and washed platelets (WP) activated with collagen using turbidometric aggregometry. NO was detected using an isolated NO electrode. The furoxan derivatives of aspirin (B8, B7) and their NO-free furazan equivalents (B16, B15; all 100 microM) significantly inhibited COX activity (P < 0.01; n = 6) in vitro and caused aspirin-independent, cGMP-dependent inhibition of collagen-induced platelet aggregation in WP. B8 was more potent than B7 (PRP IC(50) = 0.62 +/- 0.1 microM for B8; 400 +/- 89 microM for B7; P < 0.0001. WP IC(50)s = 0.6 +/- 0.1 and 62 +/- 10 microM, respectively). The NO-free furazan counterparts were less potent antiplatelet agents (WP IC(50)s = 54 +/- 3 microM and 62 +/- 10 microM, respectively; P < 0.0001, B8 vs B16). Of the hybrids investigated, only B8 retained antiplatelet activity in PRP.NO release from furoxan-aspirin hybrids was undetectable in buffer alone, but was accelerated in the presence of either plasma or plasma components, albumin (4%), glutathione (GSH; 3 microM) and ascorbate (50 microM), the effects of which were additive for B7 but not B8. NO generation from furoxans was greatly enhanced by platelet extract, an effect that could largely be explained by the synergistic effect of intracellular concentrations of GSH (3 mM) and ascorbate (1 mM). We conclude that the decomposition of furoxan-aspirin hybrids to generate biologically active NO is catalysed by endogenous agents which may instil a potential for primarily intracellular delivery of NO. The blunting of the aspirin effects of furoxan hybrids is likely to be due to loss of the acetyl moiety in plasma; the observed antiplatelet effects are thereby primarily mediated via NO release. Compounds of this class might represent a novel means of inhibiting platelet aggregation by a combination of NO generation and COX inhibition.

Figure 1

Structural formulae of furoxan hybrids (B8 and B7), furazan hybrids (B16 and B15) and the nitrooxy-ester NCX4016.

Figure 2

EPR-based COX-activity assay. (a) Typical 3-peak EPR spectra obtained in the absence (control; COX+AA) and presence of aspirin or hybrid compound (all 100 μM) after correction for background autoxidation. EPR settings: B0-field, 3356 G; sweep width; 50 G, microwave frequency, 9.3 GHz; sweep time, 30 s; modulation amplitude, 1500 mG; microwave power, 20 mW. (b) Effect of aspirin, salicylic acid, furoxans, furazans and NCX4016 (all 100 μM) on EPR signals generated from COX-1 after treatment with substrate (AA). In each case, drug incubations were for 10 min prior to the baseline EPR reading. Readings shown were taken 1.5 min after the addition of AA. ^*P<0.05, ^**P<0.01; one-way ANOVA with Dunnet's post hoc test vs control: n=6–10. Values are mean±s.e.m.

Figure 3

Collagen (2.5 μg ml⁻¹)-induced platelet aggregation in PRP and WP. (a) Effect of NCX4016, salicylic acid and aspirin on collagen-induced platelet aggregation in PRP. ^***P<0.0001, n=6–7. (b) The effect of the guanylate cyclase inhibitor ODQ (50 μM; 15 min preincubation) on responses to NCX4016 in WP. ^**P=0.002, n=6. Values are mean±s.e.m. Statistical analysis by two-way ANOVA.

Figure 4

Collagen (2.5 μg ml⁻¹)-induced platelet aggregation in PRP and WP. (a) Effect of B8 (±ODQ; 50 μM, 15 min preincubation) and its NO-free equivalent, B16, on collagen-induced platelet aggregation in PRP. P<0.0001 (B8+ODQ vs B8 alone, n=6–9). (b) The effect of ODQ (50 μM) on responses to B8 in WP. P<0.0001 (+ODQ vs B8 alone), n=6–9. (c) Effect of B7 (±ODQ; 50 μM) and its NO-free equivalent B15 on collagen-induced platelet aggregation in PRP. ^***P<0.0001, n=6–7. (d) The effect of ODQ (50 μM) on the B7 response in WP. ^***P<0.0001, n=6–7. Values are mean±s.e.m. Statistical analysis by two-way ANOVA.

Figure 5

NO release recorded over 10 min from B8 (100 μM) or B7 (500 μM) in various media. (a) NO release from B8 in media related to plasma conditions: Tyrode's buffer was reconstituted with approximate plasma concentrations of the plasma constituents, albumin (4%), GSH (3 μM) and ascorbate (50 μM). Inset shows typical 10 min traces of NO release recorded via the NO electrode in Tyrode's buffer with or without ascorbate (1 mM) and GSH (3 mM). (b) Shows typical NO release from B8 in media related to platelet conditions: Tyrode's buffer 3 and 1 mM are approximate intracellular concentrations of glutathione and ascorbate, respectively; n=6–7. GSH=Glutathione. Ascorb=ascorbate. Values are mean±s.e.m. (c) Shows NO release from B7 in media related to plasma conditions. Inset shows typical 10 min traces of NO release from B7 recorded via the NO electrode in Tyrode's buffer with or without ascorbate (1 mM) and GSH (3 mM) (d) shows NO release from B7 in media related to platelet conditions.

Figure 6

A sample of WP was suspended in 1 ml 0.5% Triton X and then homogenized to release platelet extract before centrifugation to remove membrane fraction. The dilution factor on x-axis was calculated by volume Triton X (1 ml)/(number of platelets × average platelet volume). (a) NO release from 16 platelet extracts treated with 100 μM B8. The triangles show samples treated in the same way but with a 10 min preincubation with 500 μM DTNB before addition of B8. (b) NO release from 13 platelet extracts treated with 500 μM B7. The triangles show samples treated in the same way but with a 10 min preincubation with 500 μM DTNB before addition of B7.