All animal procedures were conducted according to the ethical guidelines of the University of Glasgow under Schedule 1 of the Animals (Scientific Procedures) Act 1986. Male Wistar rats (150?180?g) were killed by concussion followed by exsanguination. The superior mesenteric artery was cannulated and the mesenteric arterial vasculature dissected from the intestines and suspended in a heated organ bath as previously described (^{McNeish et al., 2002; Stirrat et al., 2008}). Mesenteries were then perfused at 37??C using a peristaltic pump (Minipuls 3, Gilson) at a flow rate of 15?ml/min with Krebs solution containing (mM): NaCl, 118; KCl, 4.7; CaCl₂, 2.5; KH₂PO₄, 1.2; MgSO₄, 1.2; NaHCO₃, 25; glucose, 11.5; and gassed with 95% O₂/5% CO₂. Tissues were allowed to equilibrate for at least 30?min before the beginning of each experiment. Perfusion pressure was measured using Gould Statham P32 ID transducers via a side arm located immediately proximal to the inflow cannula and displayed on a PowerLab data acquisition system (AD Instruments, Hastings, U.K.).

In order to observe EDHF-mediated dilator responses to acetylcholine, the perfusion pressure was raised to ~?120?mm Hg using the thromboxane-mimetic, U46619 (10?100?nM). The effects of nitric oxide and cyclooxygenase products were routinely blocked using N^G-nitro-L-arginine methyl ester (L-NAME, 100??M) and indomethacin (3??M), respectively. Vasodilator responses (~?50% max) to acetylcholine (10?nmol) were elicited at 15?min intervals during a 3?h study period by injecting into the perfusion fluid 10??l volumes of stock solution using a Hamilton micro-syringe, and expressed as a percentage of the U46619-induced perfusion pressure.

As will be seen in the Results, the ability of ascorbate (50??M) to induce a rapid constriction and a slowly developing (max at 3?h) blockade of acetylcholine-induced, EDHF-mediated dilatation (^{Stirrat et al., 2008}) was abolished by the Cu²⁺ ion chelator, cuprizone (30??M), and enhanced by CuSO₄ (0.5??M). As a consequence, except when otherwise stated, all further experiments with ascorbate (50??M) were conducted on tissues perfused with Krebs containing CuSO₄ (0.5??M). A number of agents were investigated for their ability to inhibit the constrictor and EDHF-blocking actions of ascorbate in the presence of CuSO₄. These were: the superoxide scavengers, superoxide dismutase (SOD, 100?u/ml), tiron (300??M) and Tempol (1?mM); the NADPH oxidase inhibitor, diphenyleneiodonium (3??M); the xanthine oxidase inhibitor, allopurinol (100??M); the mitochondrial inhibitors, rotenone (1??M) and myxothiazole (0.3??M); the hydrogen peroxide scavenger, catalase (1200?u/ml); and the hydroxyl radical scavenger, mannitol (10?mM).

Time-matched control experiments were conducted in the absence of ascorbate and CuSO₄ to determine the reproducibility of the dilator action of acetylcholine. In addition, the selectivity of the blockade of acetylcholine-induced, EDHF-mediated vasodilatation by ascorbate in the presence of CuSO₄ was assessed by examining the effects on dilatation to the endothelium-independent agent, levcromakalim (5?nmol).

Experiments were conducted to determine if FeCl₃ (0.5 and 100??M) could mimic the ability of CuSO₄ to facilitate ascorbate (50??M)-induced constriction and blockade of acetylcholine-induced, EDHF-mediated dilatation. Before these experiments were conducted, Cu²⁺-free conditions were produced by purging the perfusion system with Krebs containing cuprizone (30??M). Cuprizone was not, however, present during these experiments to guard against any possibility of chelation of Fe³⁺ ions.

Other experiments were conducted to determine if hydrogen peroxide (30 and 50??M) could mimic the ability of ascorbate to induce constriction and blockade of acetylcholine-induced, EDHF-mediated dilatation. These experiments were conducted in the presence of cuprizone (30??M) to ensure any effects seen were due to hydrogen peroxide per se and not to other reactive oxygen species formed following its interaction with trace Cu²⁺ ions.

The ability of Cu²⁺ and Fe³⁺ ions to catalyse the formation of hydrogen peroxide from ascorbate was assessed using the FOX2 (ferrous oxidation in xylenol orange, version 2) method (^{Nourooz-Zadeh et al., 1994}). The FOX2 reagent was prepared by adding one volume of solution 1 (1?mM xylenol orange and 2.5?mM ammonium ferrous sulphate in 250?mM sulphuric acid) to 9 volumes of solution 2 (4.4?mM butylated hydroxytoluene in HPLC grade methanol). Experiments were carried out at 37??C in organ baths filled with the same Krebs solution used for the biological experiments and gassed with 95% O₂/5% CO₂?CuSO₄ (0.5??M) or FeCl₃ (100??M) was added to the Krebs solution and 100??l samples (time zero) taken for assay. Ascorbate (50??M) was then added to the Krebs solution and 100??l samples taken for assay at 2, 30, 60, 120 and 180?min. Once taken, samples were immediately added to 1?ml of FOX2 reagent and incubated at room temperature for 30?min. Absorbance at 560?nm was read using a Pye Unicam SP6-550 spectrophotometer, with hydrogen peroxide levels assessed using standards of known concentration (0?100??M). Fresh standards and blanks were prepared for each time point because their absorbance did not remain stable during the 3?h duration of experiments.

Acetylcholine chloride, ammonium ferrous sulphate, allopurinol, ascorbic acid, butylated hydroxytoluene, cuprizone, deferoxamine, diphenyleneiodonium, indomethacin, myxothiazole, L-NAME (N^G-nitro-L-arginine methyl ester), mannitol, rotenone, Tempol, tiron and U46619 (9,11-dideoxy-11?,9?-epoxy-methanoprostaglandin F_2?) were all obtained from Sigma. Catalase (bovine liver) was obtained from Calbiochem. Hydrogen peroxide was obtained from BDH. Levcromakalim was a gift from GlaxoSmithKline (Harlow, UK). Xylenol orange disodium salt was obtained from Fluka. All drugs were dissolved and diluted in 0.9% saline except indomethacin (1?mM stock), which was dissolved in Na₂CO₃ (0.4?mg/ml), rotenone (10?mM stock in ethanol), levcromakalim (0.1?M stock in 70% ethanol), cuprizone and U46619 (10?mM and 1?mM stocks, respectively, in 50% ethanol), allopurinol, diphenyleneiodonium and myxothiazole (0.1?M, 30?mM and 3?mM stocks, respectively, in DMSO).

Results are expressed as the mean???S.E.M. of n separate observations, each from a separate tissue. Graphs were drawn and statistical comparisons made using one-way analysis of variance and Bonferroni's post-test with the aid of a computer program, Prism (GraphPad, San Diego, USA). A probability (P) less than or equal to 0.05 was considered significant.

In the rat mesenteric vascular bed, perfused at 15?ml/min in the presence of L-NAME (100??M) and indomethacin (3??M), perfusion pressure was raised to ~?120?mm Hg using U46619 (10?100 nM). Under these conditions, acetylcholine (10?nmol) induced powerful EDHF-mediated dilatation (49.8???5.4%, n?=?8; Fig. 2). As previously reported (^{Stirrat et al., 2008}), treatment with ascorbate (50??M) induced a rapid rise in perfusion pressure (53.8???11.8?mm Hg, n?=?9; Figs. 1 and 2) and a slowly developing blockade of acetylcholine-induced, EDHF-mediated dilatation (reduced to 25.2???5.6% at 3?h, n?=?7, Fig. 2). The Cu²⁺ chelator, cuprizone (30??M), abolished the ability of ascorbate to induce constriction and block acetylcholine-induced dilatation (Figs. 1 and 2). Moreover, treatment with CuSO₄ (0.5??M) enhanced the ability of ascorbate to induce constriction and block acetylcholine-induced dilatation. In contrast, following prior purging of the perfusion system with cuprizone (30??M) to create Cu²⁺-free conditions, FeCl₃ (0.5 and 100??M) lacked the ability of Cu²⁺ to support ascorbate-induced constriction or blockade of acetylcholine-induced dilatation. In addition, the Fe³⁺ chelator, deferoxamine (100??M), had no statistically significant influence on the poor ability of ascorbate to induce constriction or blockade of acetylcholine-induced dilatation under these conditions.

Our previous findings with ascorbate on the rat mesentery showed that it produced selective blockade of acetylcholine-induced, EDHF-mediated dilatation, with no effects on the endothelium-independent dilatation induced by levcromakalim (^{Stirrat et al., 2008}). Fig. 3 shows, however, that in the additional presence of CuSO₄ (0.5??M), which enhances the time-dependent blockade of acetylcholine (10?nmol)-induced dilatation, vasodilator responses to levcromakalim (5?nmol) were also inhibited.

The ability of ascorbate (50??M) in the presence of CuSO₄ (0.5??M) to induce both constriction and blockade of acetylcholine (10?nmol)-induced, EDHF-mediated dilatation was unaffected by the superoxide scavengers, SOD (100?u/ml) and Tempol (1?mM), but was abolished by tiron (300??M; Fig. 4). The NADPH oxidase inhibitor, diphenyleneiodonium (3??M), and the mitochondrial inhibitors, rotenone (1??M) and myxothiazole (0.3??M), each induced substantial falls in U46619-induced perfusion pressure (of 81.2???11.5, 32.7???12.0 and 24.8???9.0?mm Hg, respectively, n?=?5?6). They each also depressed the ability of ascorbate to induce constriction, but had no effect on its blockade of acetylcholine-induced dilatation (Fig. 4). Neither the xanthine oxidase inhibitor, allopurinol (100??M), nor the hydroxyl radical scavenger, mannitol (10?mM), had any effect on ascorbate-induced constriction or the blockade of acetylcholine-induced dilatation. The hydrogen peroxide scavenger, catalase (1200?u/ml), had no effect on ascorbate-induced constriction, but abolished the blockade of acetylcholine-induced dilatation (Figs. 1 and 4).

The ability of catalase to inhibit the blockade of acetylcholine-induced dilatation by ascorbate in the presence of CuSO₄ prompted us to examine the generation of hydrogen peroxide. In the absence of tissues, the addition of ascorbate (50??M) to Krebs solution containing CuSO₄ (0.5??M) at 37??C led to the rapid generation of hydrogen peroxide (Fig. 5): 6.6???1.8??M (n?=?6) was detected at 2?min and a maximum of 40.2???3.6??M was seen at 1?h, with only a slight decline during the remainder of the 3?h incubation. The accumulation of hydrogen peroxide under these conditions was almost abolished by catalase (1200?u/ml). In contrast to CuSO₄, FeCl₃ (100??M) failed to generate appreciable levels of hydrogen peroxide from ascorbate.

The generation of hydrogen peroxide during the oxidation of ascorbate by Cu²⁺, coupled with the ability of catalase to inhibit the associated blockade of acetylcholine-induced dilatation but not the constriction, led us to examine which of these vascular actions could be mimicked by exogenous hydrogen peroxide. We found that hydrogen peroxide failed to produce constriction of the rat mesentery at any concentration tested (1?100??M). It did, however, induce dilatations of 13.9???2.8%, 23.0???6.5% and 76.7???21.2% (n?=?5), at concentrations of 10, 30 and 100??M respectively. Hydrogen peroxide (30??M) produced a small but significant blockade of acetylcholine (10?nmol)-induced dilatation at 3?h (control 49.8???5.5%; treated 28.2???4.0%, n?=?6, P?<?0.05), but had no significant effect on dilatation to levcromakalim (5?nmol; control 48.7???7.1%; treated 38.7???4.1%, n?=?9). Raising the hydrogen peroxide concentration to 50??M led to a time-dependent blockade of acetylcholine-induced dilatation and blocked levcromakalim-induced dilatation at 2 and 3?h (Fig. 6).