INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

SINCE FURCHGOTT AND ZAWADZKI (7) first discovered what is now known as nitric oxide (NO), there has been considerable interest in characterizing the physiological and pathophysiological role of this substance. Its involvement in modulation of vascular tone has since been well established (7), and several other physiological effects have been demonstrated, such as inhibition of platelet aggregation (28) and modulation of leukocyte adhesion (16, 19). In conditions such as myocardial ischemia-reperfusion (MI/R) injury, changes in NO metabolism as well as production of oxygen-derived free radicals (1) are thought to play a critical role in the extension of tissue injury during the reperfusion period.

One purported mediator of tissue injury during reperfusion is peroxynitrite (ONOO; see Ref. 32). This reaction product of NO and superoxide is formed almost exclusively at equimolar concentrations of these two reactants (23). At high micromolar and low millimolar concentrations, ONOO promotes DNA strand breakage (12) and nitration of phenolic-containing substances (10), resulting in cellular dysfunction.

Other studies, however, have demonstrated that ONOO at high nanomolar to low micromolar concentrations exerts beneficial effects similar to those of NO, including vasodilation of human (15) coronary arteries and inhibition of leukocyte-endothelium interactions (20, 25). A potential molecular mechanism that may account for the beneficial effects of ONOO is the formation of S-nitrosothiols (34). S-nitrosothiols have previously been shown to relax vascular smooth muscle by activation of guanylyl cyclase (11) and protect in the setting of ischemia-reperfusion (4). S-nitrosothiols have also been shown to occur endogenously in low micromolar concentrations in human plasma (30), as well as in human airways (8). S-nitrosothiols may act as an endogenous reservoir of NO as well as an NO donor (i.e., release of NO; see Refs. 3 and 30).

It is necessary to define the maximally achievable concentrations of ONOO that may be formed in vivo. It has previously been shown that ONOO is the reaction product of NO and superoxide at equimolar concentrations of these reactants. Moreover, significant imbalances in the concentrations of these two reactants markedly curtail the formation and oxidative potential of ONOO (23). Thus the maximally achievable concentration of ONOO is limited by the highest amount of either NO or superoxide produced in vivo. Physiological concentrations of NO occur in the range of 1-20 nM (13) and are thought to increase to maximal levels in the low micromolar range (i.e., 2-5 µM) in disease states when inducible NO synthase is activated. Therefore, it follows that maximally achievable concentrations of ONOO that may be formed in vivo would also be in the low micromolar range (i.e., 2-5 µM), and concentrations above these levels would probably not be formed in vivo.

Two important recent findings lend support to the contention that ONOO may actually be beneficial rather than cytotoxic. First, physiologically achievable concentrations of ONOO significantly attenuated neutrophil-endothelium interactions in an in vivo rat mesentery preparation (20). Second, at low micromolar concentrations, ONOO decreased the extension of necrotic tissue in an in vivo model of MI/R injury (25). These were the first studies to demonstrate that exogenously administered ONOO may be protective in the in vivo setting, thus confirming the results of the in vitro reports. However, in neither of these two reports was the dose-response relationship of ONOO examined, enabling assessment of physiologically relevant or maximally achievable concentrations of ONOO.

Thus the main purposes of this study were 1) to determine the dose-response relationship of ONOO in a well-characterized model of feline MI/R injury and 2) to provide data on the cellular mechanism(s) responsible for the observed cardioprotective effects of ONOO including data clarifying whether ONOO can act as a direct NO donor or a NO carrier via S-nitrosothiol formation.

	METHODS

Top Abstract Introduction Methods Results Discussion References

MI/R in vivo. All procedures employed in the present study are in accordance with the guidelines of the American Physiological Society for the care and use of experimental animals. Adult male cats (2.4-3.2 kg) were anesthetized with pentobarbital sodium (30 mg/kg, iv). An intratracheal cannula was inserted through a midline incision, and cats were placed on mechanical ventilation (small animal respirator; Harvard, Dover, MA). The right femoral vein was cannulated for administration of additional pentobarbital sodium as needed to maintain a surgical plane of anesthesia. The right femoral artery was cannulated for continuous measurement of mean arterial blood pressure (MABP) and for withdrawal of blood samples. For administration of ONOO, a catheter was inserted directly into the lumen of the left ventricle through the apex with care taken not to interfere with the mitral valve, chordae tendinae, or the papillary muscles. After a midsternal thoracotomy, the anterior pericardium was incised and a 3-0 silk ligature was placed around the left anterior descending (LAD) coronary artery 8-10 mm from its origin. Standard lead II of the scalar electrocardiogram (ECG) was used to determine heart rate (HR) and S-T segment elevation. The ECG and MABP were continuously monitored on a model 78304A unit oscilloscope (Hewlett Packard, Palo Alto, CA) and recorded on an oscillographic recorder every 20 min (model 2107-4490-00; Gould, Cleveland, OH). The pressure-rate index (PRI), employed as an index of myocardial oxygen demand, was calculated as the product of MABP and HR/1,000. Two milliliters of blood were drawn every 2 h to count circulating white blood cells according to established procedures (24).

Experimental protocol. After completion of all surgical procedures, the cats were allowed to stabilize for 30 min before baseline readings of ECG and MABP were recorded. In cats subjected to MI/R, ischemia was induced by tightening the previously placed reversible silk ligature around the LAD so that the vessel was completely occluded. This was designated as time 0. Eighty minutes after coronary occlusion (i.e., 10 min before reperfusion), intraventricular infusion of ONOO in pH 8.4 saline or pH 8.4 saline alone was initiated and maintained throughout the 4.5-h reperfusion period. Ten minutes later, the LAD ligature was untied, and the ischemic myocardium was allowed to reperfuse for 270 min.

Cats were randomly divided into five groups: 1) six sham MI/R cats receiving 0.9% NaCl at pH 8.4, 2) six MI/R cats receiving 0.9% NaCl, pH 8.4, as a vehicle, 3) six MI/R cats receiving ONOO (0.2 µM) in 0.9% NaCl, pH 8.4, 4) seven MI/R cats receiving ONOO (2 µM) in 0.9% NaCl, pH 8.4, and 5) six MI/R cats receiving ONOO (20 µM) in 0.9% NaCl, pH 8.4. Sham MI/R cats were subjected to the same surgical procedures and observed for the same duration of time as MI/R cats except that the LAD coronary artery was not occluded. Concentrations of ONOO >20 µM could not be employed due to severe changes in osmolarity and pH of the infusion solution.

Quantification of myocardial area at risk and necrotic area. At the end of the 270-min reperfusion, the ligature around the LAD was tightened again. Twenty milliliters of 0.5% Evans blue (Sigma Chemical, St. Louis, MO) were rapidly injected into the left ventricle to stain the area of myocardium that was perfused by the patent, nonoccluded coronary arteries [i.e., left circumflex (LCX) and right coronary arteries] according to previously described methods (24). With the use of this method, the atria and right ventricle were separated from the rest of the heart and discarded. The left ventricle was then sliced parallel to the atrioventricular groove into 3-mm-thick sections. The ischemic, unstained portion of the heart [i.e., area at risk (AAR)] was separated from the stained, nonischemic portion of the left ventricle (i.e., area not at risk). The ischemic portion was then sliced into 1-mm-thick sections and incubated in 0.1% nitroblue tetrazolium dye (NBT; Sigma) for 7 min at 37°C. In the presence of viable coenzymes and dehydrogenases, NBT is converted to a blue azo dye, thus indicating ischemic but nonnecrotic tissue. The irreversibly injured or necrotic portion of the AAR that did not stain with the NBT was separated from the stained, ischemic nonnecrotic portion of the myocardium. The three portions of the myocardium (i.e., nonischemic, ischemic nonnecrotic, and ischemic necrotic) were subsequently weighed. Results were expressed as the AAR indexed to the total left ventricular mass (total LV) and the area of necrotic tissue indexed to either the AAR or the total LV.

Autologous cat PMN isolation and labeling. Peripheral blood (20 ml) was collected from the cannulated femoral artery just before thoracotomy, and polymorphonuclear leukocytes (PMNs) were isolated by the method of Lafrado and Olson (17). PMN preparations obtained by this method were in general >95% pure and >95% viable.

Isolated autologous cat PMNs were subsequently labeled with a fluorescent dye (PKH2 Green Fluorescent Cell Linker Kit; Sigma Immunochemicals) according to the method of Yuan and Fleming (35). One milliliter of diluent was added to a cell pellet containing ~10 × 10⁶ cells. One milliliter of PKH2-GL dye solution (4 µM) was added to the cell suspension and incubated for 7 min at room temperature. Two milliliters of phosphate-buffered saline (PBS) with 10% platelet-poor plasma were then added to stop the labeling reaction. Cells were collected and centrifuged at 400 g for 10 min and were resuspended in PBS. This labeling procedure does not affect either the normal morphology or function of PMNs (33).

PMN adherence to the cat coronary endothelium. PMNs were isolated and fluorescently labeled as described above. Both LAD and nonischemic LCX coronary segments were isolated from ischemic-reperfused cat hearts and placed into warmed Krebs-Henseleit (KH) buffer. These ex vivo artery segments were cut 2-3 mm in length. The segments were placed endothelial-surface-up into a cell culture dish filled with 3 ml of oxygenated KH buffer and incubated in culture dishes with autologous labeled PMNs for 20 min at 37°C. After incubation, the coronary segments were rinsed lightly to remove nonadherent PMNs and placed on microscope slides. Adherent PMNs were counted using epifluorescence microscopy (Nikon, Tokyo, Japan) on five separate fields from each vessel segment and expressed as PMNs per square millimeter of coronary vascular endothelial surface area as previously described (24).

Additionally, LAD and LCX segments were isolated from cats that had not undergone ischemia-reperfusion (i.e., sham MI/R cats). These segments underwent the same procedure as above but were stimulated for 10 min with 2 U/ml thrombin, with or without ONOO (2 µM), sodium nitrite (NaNO₂) acidified to pH 2.0 at 2 µM, and hemoglobin (Hb; 20 µg/ml). The Hb was added 1 min before the ONOO or NaNO₂ addition, and the ONOO or NaNO₂ was added 3 min before thrombin stimulation. Adherence of autologous PMNs was then evaluated as above. In pilot experiments, we determined that ONOO had no direct effect on the staining protocol of the PMNs. This was done by coincubation of labeled rat PMNs with rat superior mesenteric artery endothelial segments in the presence of 1 µM histamine with and without 20 µM ONOO, as outlined for cat coronary artery segments above. There was no significant difference in PMN adherence between histamine-treated segments without ONOO (103 ± 6 PMNs/mm²) and histamine-treated segments given 20 µM ONOO (100 ± 16 PMNs/mm²). This higher concentration of ONOO is one that does not inhibit PMN adherence, and so it clearly shows that ONOO does not influence the fluorescent dye used to label the PMNs.

Isolated cat coronary artery ring vasoactivity. Both LAD and LCX coronary arteries were isolated from ischemic-reperfused cats and placed into warmed KH buffer as described above. Arteries were cut into rings of 2- to 3-mm length. The rings were then mounted on stainless steel hooks, transferred to tissue baths, and connected to FT-03 force transducers (model 7; Grass Instrument, Quincy, MA) as described previously (21). Relaxation of isolated cat LAD and LCX coronary artery rings to the endothelium-dependent dilators [100 nM acetylcholine (ACh) and 1 µM calcium ionophore A-23187] and to the endothelium-independent dilator acidified nitrite (NaNO₂, 100 µM) was calculated as the percent decrease from the peak U-46619 (100 nM)-induced precontraction value as previously described (21).

In several additional studies, cat coronary artery rings from nonischemic-reperfused cats were isolated and set up as described above. However, the rings were de-endothelialized by gentle application of a cotton swab. The removal of the endothelium was done to eliminate any confounding influences of endogenous NO production on relaxation of the coronary artery rings. These de-endothelialized rings relaxed fully to 100 µM acidified NaNO₂ but only <5% to 100 nM ACh. Both concentrations were selected to yield maximal vasorelaxation. Both ONOO (40 µM at pH 8.4) and acidified NaNO₂ at 100 µM, with or without Hb (100 µg/ml), were added to the coronary artery ring baths to determine whether ONOO was a direct NO donor. KH buffer alone at either pH 2.0 or 8.4 exerted no effect on vasocativity of cat coronary artery rings.

Immunohistochemical localization of P-selectin in the cat myocardium. To determine the effect of ONOO on endothelial surface expression of P-selectin after ischemia and reperfusion, two additional cats in each of the four MI/R groups were exposed to 90 min of ischemia and 30 min of reperfusion and received either vehicle or the appropriate dose of ONOO (i.e., 0.2, 2, or 20 µM) 10 min before reperfusion. Two additional cats served as sham-operated controls and were not subject to ischemia. After 30 min of reperfusion, the hearts were removed, and the aorta was cannulated and perfused with KH buffer for 3 min, followed by perfusion with ice-cold 4% paraformaldehyde in PBS for 3 min. Slices of cardiac tissue were dehydrated using graded acetone washes at 4°C. Tissue sections were embedded in plastic (Immunobed; Polysciences, Warrington, PA), and 4-µm-thick sections were cut and transferred to Vectabond-coated slides (Vector Laboratories, Burlingame, CA). Immunohistochemical localization of P-selectin was accomplished using the avidin-biotin immunoperoxidase technique (Vectastain ABC reagent; Vector Laboratories) and the monoclonal antibody PB1.3 (Cytel, San Diego, CA), which is a neutralizing antibody against P-selectin that only recognizes cell surface-expressed P-selectin (33). Positive staining was defined as a venule displaying brown reaction product on >50% of the circumference of its endothelium, as previously described (33). Ten sections from each heart and 50 venules per tissue section were examined, and the percentage of positive staining venules was then calculated.

S-nitrosoglutathione detection via HPLC. Reduced and oxidized glutathione (GSH and GSSG, respectively; Sigma Chemical) were prepared immediately before each experiment as standard stock 1 M solutions in normal KH buffer, from which serial dilutions were made. S-nitrosoglutathione (GSNO) was also prepared immediately before each experiment, according to the method of Park (26). One milliliter of a 1 mM GSH solution was incubated for 5 min at 37°C with increasing concentrations of ONOO (1-1,000 µM) in the presence of the copper chelator bathocuproinedisulfonic acid (Sigma Chemical). An aliquot of the GSH-ONOO mixture was then treated with iodoacetic acid to generate the S-carboxymethyl (CM) derivatives of compounds presenting free sulfhydryl groups and then treated overnight with ethanolic 1-fluoro-2,4-dinitrobenzene to form the N-dinitrophenyl (DNP) derivative (29). GSNO is stable during this reaction, and N-DNP-GSNO is obtained. GSH, GSSG, and GSNO yields were based on similarly derivatized standard solutions of known concentrations.

The CM-DNP derivatives were separated by chromatography using an anion exchange HPLC method similar to the procedure developed by Reed et al. (29), utilizing the acetate buffer described as the aqueous elution solvent. Compound resolution was obtained with a Whatman Partisil 10 SAX 25 cm HPLC cartridge column (Whatman, Clifton, NJ) equilibrated with 5% acetate buffer in methanol and developed with a 26-min linear gradient to 60% acetate buffer starting at 3 min after sample injection (20 µl), followed by 5 min of isocratic 60% acetate buffer in methanol, and finally 6 min of re-equilibration with 5% acetate buffer in methanol. The flow rate was 1.25 ml/min, and elution bands were detected by absorbance at 365 nm. The retention time for N-DNP-GSNO was 26.0 min, 33.5 min for N-DNP-S-CM-glutathione, and 34.9 min for DNP-derivatized GSSG.

ONOO. ONOO was obtained from Dr. Harry Ischiropoulos (University of Pennsylvania, Philadelphia, PA) and from Alexis (San Diego, CA); the ONOO was freshly prepared, delivered via express carrier in dry ice overnight, and stored at 80°C. ONOO from both sources was synthesized from acidified NaNO₂ and hydrogen peroxide according to the method of Beckman et al. (2). The concentration of ONOO was monitored before use in each experiment by measuring the extinction coefficient at 302 nm after the addition of 5 µl of ONOO in 3 ml of 1 N sodium hydroxide at pH 12. Only ONOO aliquots that exhibited a concentration >95% of the stipulated concentrations based on their extinction coefficient were used. At the end of each experiment, the concentration of the ONOO that was used was determined again. All of the ONOO utilized during the experiment maintained >90% of the original value at the end of reperfusion, thus confirming that native ONOO itself was delivered to each cat. Aliquots of ONOO were diluted in an appropriate volume of freshly prepared, ice-cold, pH 8.4 saline. The pH of this saline solution was adjusted by addition of an appropriate volume of 0.1 N NaOH directly to the normal saline. The ONOO, pH 8.4 saline solution was then infused at a rate of 1 ml/h to achieve the desired concentration in the coronary circulation (i.e., 0.2, 2, or 20 µM). This was determined by the infusion rates of ONOO employed, coupled with estimating the stroke volume of a 3-kg cat to be ~3 ml and the transit time of blood from the left ventricular chamber to the coronary arteries to be 4-6 s. A totally decomposed form of ONOO prepared from the same stock of ONOO was found in preliminary tests to be inactive in all aspects of this study, thus eliminating the effects of nitrite and hydrogen peroxide from consideration.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Effect of ONOO on the extent of myocardial necrosis. To ascertain the effects of infusion of different concentrations of ONOO on the degree of myocardial salvage of necrotic tissue after reperfusion, we assessed the amount of necrotic tissue and expressed it both as a percentage of the AAR and as a percentage of the total LV (Fig. 1). The AAR indexed to the total LV was not significantly different among any of the MI/R groups studied, indicating that the severity of the ischemic insult was comparable in all four MI/R groups. However, the mass of necrotic myocardial tissue expressed as a percentage of the AAR or total LV was different among the myocardial ischemia groups. Only infusion of the intermediate concentration of ONOO (2 µM) directly into the left ventricle significantly protected the myocardium from developing a substantial amount of necrosis (14.4 ± 0.6 vs. 30.3 ± 3.2% in the vehicle group, P < 0.01). Infusion of the low (0.2 µM) and high (20 µM) concentrations of ONOO exerted no protective effects in limiting the progression of myocardial necrosis after reperfusion. Neither the effects of the high nor of the low concentration of ONOO (i.e., 0.2 or 20 µM) were statistically significant when compared with the vehicle MI/R group. Thus maximally achievable concentrations of ONOO in the low micromolar range (2 µM) exerted marked cardioprotective effects, but lower (0.2 µM) or higher concentrations of ONOO (20 µM) were without significant effect.

View larger version (51K): [in this window] [in a new window]   Fig. 1.   Tissue wet weights of area at risk (AAR) as a percentage of total left ventricular (Total LV) wet weight (left), necrotic tissue as a percentage of the AAR (middle), and necrotic tissue as a percentage of Total LV wet weight (right) for the 4 groups of cats that underwent myocardial ischemia and reperfusion (MI/R). Heights of bars are means and brackets SE. Only cats receiving peroxynitrite (ONOO, 2 µM) exhibited significantly attenuated cardiac necrosis. Nos. at the bottom of the bars indicate no. of cats studied. NS, not significant.

Effects of ONOO on endothelial dysfunction. Because endothelial dysfunction is an early and critical event in neutrophil-mediated myocardial reperfusion injury, we tested endothelial dysfunction by comparing vasoactivity of isolated coronary vascular rings with the endothelium-dependent vasodilators ACh and the calcium ionophore A-23187 and with the endothelium-independent vasodilator acidified NaNO₂. Figure 2 summarizes the responses of the isolated coronary rings to the receptor-mediated endothelium-dependent dilator ACh. In the ischemic-reperfused LAD, only the rings isolated from MI/R cats given the intermediate concentration of ONOO (2 µM) exhibited a significantly preserved endothelial function. ACh responses in MI/R cats receiving the low (0.2 µM) and high (20 µM) concentrations of ONOO demonstrated a similar degree of endothelial dysfunction when compared with responses in cats given the vehicle. In the control LCX rings, all of the groups showed a similar high degree of vasorelaxation to ACh, except for the 20 µM ONOO group, which exhibited an attenuated vasorelaxation. As shown in Fig. 3, ischemic-reperfused LAD coronary rings demonstrated a similar pattern of relaxation as that for ACh, with 2 µM ONOO conferring significant preservation of endothelial function and 0.2 and 20 µM ONOO having no protective effect. However, challenge of A-23187 in the control LCX rings showed a similar high degree of vasorelaxation in all MI/R groups, including the high 20 µM ONOO concentration. This suggests that 20 µM ONOO contributes to endothelial dysfunction only at the level of the muscarinic receptor. Figure 4 shows that both LAD and LCX coronary artery rings fully relaxed to the endothelium-independent dilator acidified NaNO₂, demonstrating that the coronary vascular smooth muscle functioned normally in all conditions. Thus only the intermediate concentration of ONOO (2 µM) exerted a significant vasculoprotective effect, consistent with the findings on myocardial necrosis.

View larger version (44K): [in this window] [in a new window]   Fig. 2.   Summary of vasorelaxant responses of coronary artery rings isolated from sham-operated control cats and cats subjected to MI/R. Left anterior descending (LAD) coronary artery rings (left) and nonischemic left circumflex (LCX) coronary artery rings (right) were challenged with 100 nM ACh. Bar heights are means, brackets are SE, and nos. at the bottom of the bars indicate no. of rings studied.

View larger version (44K): [in this window] [in a new window]   Fig. 3.   Summary of vasorelaxant responses of coronary artery rings isolated from sham-operated control cats and cats subjected to MI/R. LAD coronary artery rings (left) and LCX coronary artery rings (right) were challenged with 1 µM A-23187, a calcium ionophore. Bar heights are means, brackets are SE, and nos. at the bottom of the bars indicate no. of rings studied.

View larger version (48K): [in this window] [in a new window]   Fig. 4.   Summary of vasorelaxant responses of coronary artery rings isolated from sham-operated control cats and cats subjected to MI/R. LAD coronary artery rings (left) and LCX coronary artery rings (right) were challenged with 100 µM acidified sodium nitrite (NaNO2), an endothelium-independent dilator. Bar heights are means, brackets are SE, and nos. at the bottom of the bars indicate no. of rings studied.

Effect of in vivo ONOO on neutrophil adherence to ex vivo coronary endothelium. The initial steps of neutrophil-mediated reperfusion injury involve rolling along and firm adherence of neutrophils to the vascular endothelium. Therefore, we assessed the extent of neutrophil adherence to the LCX and the LAD of ex vivo vascular segments obtained at the conclusion of each protocol (Fig. 5). Few neutrophils adhered to control LCX arterial segments in any of the MI/R groups, with the exception of the highest ONOO concentration (20 µM). Also, few neutrophils adhered to the endothelium of LAD coronary artery segments isolated from control nonischemic cats. In contrast, LAD segments isolated from ischemic-reperfused cats receiving only vehicle showed a marked increase (i.e., 3- to 4-fold) in PMN adherence compared with nonischemic controls (P < 0.001). However, LAD segments isolated from ischemic-reperfused cats receiving 2 µM ONOO exhibited a significantly lower PMN adherence than that of ischemic-reperfused cats receiving only vehicle (P < 0.01). Neither the low nor the high ONOO concentrations inhibited PMN adherence to the endothelium. These data suggest that only at 2 µM does ONOO significantly modulate PMN-endothelial interactions in an in vivo setting of MI/R injury.

View larger version (40K): [in this window] [in a new window]   Fig. 5.   Effects of ONOO or pH 8.4 saline on adherence of unstimulated neutrophils (PMNs) to coronary endothelium. Sham-operated controls (Sham) as well as rings from cats subjected to MI/R were studied. In the 4 MI/R groups, ischemic-reperfused LAD and LCX coronary endothelium were studied. Data are expressed as no. of neutrophils adhering per square millimeter of coronary artery endothelial surface area. Bar heights are means, brackets are SE, and nos. at the bottom of the bars indicate no. of coronary artery segments studied.

Effect of ONOO on immunohistochemical localization of P-selectin. P-selectin is a key mediator of the early steps of neutrophil-mediated reperfusion injury (33). Therefore, a possible mechanism for the observed cardioprotective effects of ONOO may involve its ability to modulate P-selectin expression on the vascular endothelium. The percentage of coronary venules staining positively for P-selectin in sham MI/R cats as well as in the area not at risk in MI/R groups was similarly low and in the range of 10% (Fig. 6). Ninety minutes of ischemia followed by 30 min of reperfusion resulted in a significant increase in the percentage of venules staining positively for P-selectin in untreated ischemic-reperfused cats, as well as in the low and high ONOO-treated groups. This represents a five- to sixfold increase in the surface expression of P-selectin under these conditions. This increased expression of P-selectin on the coronary microvasculature was significantly attenuated by infusion of ONOO only at 2 µM (P < 0.01). Therefore, it appears that only 2 µM ONOO was able to modulate surface expression of P-selectin on vascular endothelium, resulting in a potential beneficial effect.

View larger version (30K): [in this window] [in a new window]   Fig. 6.   Percentage of coronary venules staining positive for P-selectin in 5 experimental groups of cats. Sham-operated controls (Sham) as well as cats subjected to MI/R were studied. Cats received either ONOO (0.2, 2, or 20 µM) or vehicle (saline at pH 8.4). Bar heights represent mean values, and brackets indicate SE. Two cats were studied in each group, 10 sections were studied in each cat heart, and 50 venules were counted in each section. ANAR, area not at risk.

Effect of ONOO on de-endothelialized cat coronary artery rings. We also investigated the ability of ONOO to act as a direct NO donor in solution as a potential protective mechanism of action of ONOO. Coronary vascular rings from nonischemic control cats were isolated as described above. To avoid any confounding effects of endogenous NO production, the endothelium of the rings was removed by gentle rubbing with a cotton swab. As shown in Fig. 7, the rings relaxed <10% when challenged with the endothelium-dependent dilator ACh (100 nM), confirming the removal of the endothelium. Rings were then challenged with ONOO (10-40 µM), in the presence or absence of 100 µg/ml Hb (a known NO scavenger). These concentrations of ONOO were selected since they significantly relaxed cat coronary arteries to maximize its potential NO donor properties. ONOO vasorelaxed the rings essentially the same with or without Hb. As a positive control, the direct NO donor acidified NaNO₂ (100 µM) was utilized. Incubation of the vascular rings with Hb markedly blunted the NaNO₂-induced vasorelaxation. These data suggest that ONOO does not directly donate NO in solution under the conditions of these studies.

View larger version (24K): [in this window] [in a new window]   Fig. 7.   Relaxation of de-endothelialized cat coronary vascular rings to ONOO (40 µM) and acidified NaNO2 (100 µM), with or without hemoglobin (Hb, 100 µg/ml). When appropriate, rings were preincubated with Hb for 5 min before challenge with either ONOO or NaNO2. Bar heights are means, brackets are SE, and nos. at the bottom of the bars indicate no. of rings studied.

Effect of exogenously administered ONOO on PMN adherence to nonischemic coronary artery segments in vitro. In an effort to further determine whether ONOO is a direct NO donor in solution, we examined the ability of ONOO to modulate PMN-endothelium interaction in the presence or absence of Hb (a known NO scavenger) in coronary artery segments isolated from control, nonischemic-reperfused cats. Stimulation of the isolated cat coronary endothelium with 2 U/ml thrombin resulted in a fourfold increase in PMN adherence to the endothelium (Fig. 8). ONOO (2 µM), both in the absence and presence of Hb, resulted in a significant reduction of PMN adherence to the thrombin-stimulated cat coronary endothelium (P < 0.01 vs. thrombin alone). Incubation of the coronary artery segments with acidified NaNO₂ (2 µM) also decreased PMN adherence, but this effect was markedly blunted by preincubation with 20 µg/ml Hb. Because Hb failed to reduce cat PMN adherence to thrombin-stimulated nonischemic cat coronary artery segments in the presence of ONOO, these data reinforce the notion that ONOO does not appear to be a direct NO donor in solution.

View larger version (27K): [in this window] [in a new window]   Fig. 8.   In vitro adherence of autologous cat PMNs to thrombin-stimulated nonischemic cat coronary segments. ONOO at 2 µM significantly attenuated PMN adherence compared with thrombin alone (P < 0.01). Preincubation with Hb (20 µg/ml) had no effect on the ability of ONOO to inhibit PMN adherence to the vascular endothelium. All values are means ± SE. Nos. at the bottom of bars indicate the no. of segments studied. KH, Krebs-Henseleit buffer control.

Formation of GSNO from ONOO and glutathione. Because the mechanism of ONOO-induced cardioprotection does not appear to be due to the direct NO-donating properties of ONOO, we examined the possibility that ONOO leads to S-nitrosothiol formation. Different concentrations of ONOO (1-1,000 µM) were incubated for 5 min with the reduced form of GSH (1 mM) in the presence of the copper chelator bathocuproinedisulfonic acid (Sigma Chemical), and the products of this reaction were then analyzed via HPLC. Figure 9 demonstrates that, as the concentration of ONOO increases, the GSH concentration decreases as it is converted to GSSG and GSNO. The maximal yield of GSNO is 4.5 µM, which occurs at 1 mM ONOO, and the highest GSSG yield of ~300 µM occurs at 0.1 mM ONOO. In the maximally achievable range of 2-5 µM ONOO, the yield of GSNO is 1 µM, which is ~0.1-1% of the initial 1 mM GSH concentration in this study. Furthermore, because the physiological levels of thiols in plasma have been measured at 500 µM (30), the reaction of ONOO with this concentration of thiols would yield plasma levels of S-nitrosothiols of ~5 µM, which is in the physiologically relevant range (30). Therefore, these data suggest that a potential mechanism of ONOO-induced cardioprotection may involve the formation of S-nitrosothiols in vivo.

View larger version (18K): [in this window] [in a new window]   Fig. 9.   HPLC analysis of ONOO-induced S-nitrosoglutathione (GSNO) formation. One milliliter of a 1 mM reduced glutathione (GSH) solution was incubated for 5 min at 37°C with increasing concentrations of ONOO (1-1,000 µM). GSNO formation was then measured via HPLC as described in METHODS. A concentration-response relationship is observed with increasing concentrations of ONOO inducing formation of increasing concentrations of GSNO. GSSG, oxidized glutathione.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

To our knowledge, this study is the first investigation of the dose-response relationships of ONOO in an in vivo model of feline MI/R injury. We have observed that only maximally achievable concentrations of ONOO (i.e., 2 µM) afforded significant cardioprotective effects and decreased the extension of myocardial necrosis upon reperfusion of the previously ischemic myocardium. This cardioprotection was followed by a significant preservation of the vasodilator capacity of isolated coronary vascular rings, a marked attenuation of PMN adherence to the vascular endothelium, as well as a marked inhibition of P-selectin expression on the vascular endothelium.

The cardioprotective effect of ONOO was not due to any overt differences in systemic hemodynamic, electrophysiological, or hematological variables between the four groups of MI/R cats. In all groups, there were no significant differences in any of these variables observed before coronary occlusion. However, a few minutes after LAD occlusion, the S-T segment of the ECG was markedly elevated in all MI/R groups. After reperfusion, the S-T segment decreased to nearly control values, indicating that successful reperfusion had occurred in all MI/R groups. Peak S-T segment elevation (at 20-40 min postocclusion) was also similar in all MI/R groups, indicating that the ischemic insult was comparable in all groups. Furthermore, no significant differences in the PRI or circulating white blood cell counts were observed, suggesting that myocardial oxygen demand was comparable in all groups and that ONOO had no direct effect on the number of circulating leukocytes. Thus direct effects of ONOO on hematological, electrocardiographic, and hemodynamic actions could not have accounted for the observed cardioprotective or vasculoprotective effects of ONOO.

As previously mentioned, maximal levels of ONOO formed in vivo would probably not exceed the 2-5 µM range. Also, the extremely short half-life of ONOO at physiological pH (i.e., ~1 s) would prevent accumulation of ONOO to higher concentrations. Furthermore, van der Vliet et al. (31) have demonstrated that the chemical fate of ONOO is directly dependent on the constituents that are present in its surrounding environment. Thus it is of primary importance to ensure that normal physiological buffers and ONOO detoxification mechanisms are present when undertaking studies on the biological actions of ONOO. In addition, nitrotyrosine formation, which was once thought to be the "footprint" of in vivo formation of ONOO, is now clearly shown to be primarily due to neutrophil myeloperoxidase-driven formation of nitryl chloride and nitrogen dioxide (5, 6) in diseases where neutrophils are present at sites of inflammation. Therefore, the presence of nitrotyrosine can no longer be considered a reliable index of ONOO production in vivo.

In this study, 2 µM ONOO resulted in a 52% attenuation of cardiac necrosis and a significant preservation of coronary artery vasorelaxation to endothelium-dependent dilators, whereas concentrations of ONOO 10 times higher were without significant beneficial effect. Also, 0.2 µM ONOO would probably be in the lower limit of physiologically achievable concentrations and could be below the threshold for any beneficial effects to be observed. This is consistent with the findings that 0.2 µM ONOO had no significant effect when compared with vehicle in all aspects of this study.

We demonstrated that the cardioprotective effects of the intermediate concentration of ONOO (2 µM) may be due to its ability to modulate P-selectin expression on the vascular endothelium. This cardioprotective mechanism of ONOO is shared with NO (4). It has previously been shown that the primary early dysfunction in ischemia-reperfusion injury is the loss of endothelium-derived NO (21), which subsequently leads to an early upregulation of adhesion molecules (i.e., P-selectin) on the vascular endothelium, promoting leukocyte-endothelial interaction (19). This allows leukocytes, primarily PMNs, to release their cytotoxic mediators, such as oxygen-derived free radicals, proteases, and cytokines, in close proximity to the vascular endothelium and cardiomyocytes, provoking further cardiomyocyte injury. Thus it is no surprise that agents that decrease the interaction of neutrophils with P-selectin have been a major focus of ischemia-reperfusion research (4, 9, 19). For instance, NO replacement therapies have been shown to decrease P-selectin expression on the endothelium (4, 9). Therefore, a proposed mechanism for the observed cardioprotection exerted by ONOO may involve the ability to modulate P-selectin expression on the vascular endothelium, as shown in Fig. 6. The ability of ONOO to modulate adhesion molecule expression on the surface of endothelial cells is supported by the observation that ONOO (2 µM) significantly decreased PMN adhesion to the coronary vascular endothelium. Thus a significant component of the cardioprotective mechanism of ONOO may be due to reduced expression of cell adhesion molecules on the vascular endothelium, which subsequently inhibits coronary PMN-endothelium interactions.

Another potential mechanism for the observed cardioprotection of ONOO is the ability of ONOO to be a direct NO donor in solution, which could account for its NO-like effects. As demonstrated in Figs. 7 and 8, the ability of ONOO to induce vasorelaxation and to inhibit PMN adhesion to vascular endothelial segments in vitro was not modified by preincubation of the arterial segments with Hb, which is known to scavenge free NO in solution. These data suggest that ONOO is not a direct NO donor in solution and therefore may be acting via a different process, perhaps through an endogenous NO carrier mechanism.

One such process may be the formation of S-nitrosothiols. As shown in Fig. 9, a near maximally achievable concentration of ONOO (1 µM) was able to react with glutathione to form GSNO. This route of S-nitrosothiol formation by ONOO is in agreement with other studies (22, 34). In human plasma, albumin is the most abundant thiol species (~500 µM; see Ref. 30). Therefore, the reaction of physiologically achievable concentrations of ONOO with this concentration of thiol compound (i.e., 500 µM) would yield ~5 µM S-nitrosothiol. This concentration of S-nitrosothiol is in line with the demonstration by Stamler et al. (30) of the presence of ~7 µM S-nitrosothiols in human plasma under normal conditions. Moreover, these low concentrations of S-nitrosothiols have been shown to exert physiologically important effects such as inhibition of platelet aggregation (27). In addition, S-nitrosothiols have also been shown to be present in human airways (8) and are used therapeutically to inhibit platelet aggregation during human coronary angioplasty (18). Recently, Davidson et al. (3) have demonstrated that endogenous ONOO production was responsible for prolongation of bovine pulmonary artery smooth muscle relaxation when these arteries were challenged with low concentrations of NO. These authors proposed that a probable mechanism for the observed effect of ONOO occurred through formation of S-nitrosothiols. These S-nitrosothiols may be involved in ONOO-mediated stimulation of guanylyl cyclase and accumulation of cGMP (22), particularly since direct release of NO cannot account for the short-term effects of S-nitrosothiols in terms of inducing vasorelaxation, stimulating guanylyl cyclase, and inhibiting platelet aggregation (14). Therefore, in light of these reports, ONOO may induce vasorelaxation and modulate PMN-endothelial interactions through the formation of low micromolar concentrations of S-nitrosothiols, which may undergo NO exchange reactions for transfer of the NO moiety to low-molecular-weight thiol groups in tissues. As a result, the NO moiety would not be initially released in solution, and Hb would have no inhibitory effect, as was the case in our study. Further studies using isolated tissues are needed to verify this potential mechanism of ONOO-induced protection.

In summary, we have demonstrated that 2 µM ONOO exerted significant cardioprotective and vasculoprotective effects and inhibited P-selectin expression on the vascular endothelium after ischemiareperfusion. Also, this ONOO concentration decreased PMN adhesion to coronary artery segments both ex vivo and in vitro. A potential mechanism for the observed effects of ONOO may involve the formation of S-nitrosothiols, which have similar effects as NO. This mechanism is supported by the fact that, in vitro, ONOO was able to nitrosate glutathione and form GSNO as shown in Fig. 9. However, higher concentrations of ONOO (i.e., 20 µM), which are probably above the maximally achievable range of ONOO formation in vivo, did not exert any beneficial effects in reducing reperfusion injury. In this connection, a decrease in vasodilator reserve was observed in nonischemic reperfused coronary arteries when challenged to endothelium and receptor-dependent dilator ACh, suggesting that the high concentration of ONOO may have had a detrimental effect at the level of muscarinic receptors that are present on the vascular endothelium. Thus maximally achievable concentrations of ONOO in the 2-5 µM range may actually be cardioprotective, whereas concentrations of ONOO above these levels may have detrimental effects in the setting of MI/R injury and would probably not be encountered in vivo.