Cardiovascular Drugs and Therapy 17 199–208 2003

C© 2003 Kluwer Academic Publishers. Manufactured in The Netherlands

BASIC PHARMACOLOGY

In Vivo Cardioprotection by N-Acetylcysteineand Isosorbide 5-Mononitrate in a Rat Modelof Ischemia-Reperfusion

Laura Calvillo1, Serge Masson1, Monica Salio1,Loredana Pollicino1, Noeleen De Angelis1,Fabio Fiordaliso1, Antonio Bai1, Pietro Ghezzi2,Francesco Santangelo3, and Roberto Latini1

1Departments of Cardiovascular Research and 2Biochemistryand Molecular Pharmacology, Istituto di RicercheFarmacologiche “Mario Negri”, Milano; 3Zambon Group SpA,Bresso, Italy

Summary. Aims: We evaluated the effect of N-acetylcys-

teine (NAC, infused i.v.), isosorbide 5-mononitrate (IS5MN,

by gavage), or their combination on cardiac injury in an in

vivo rat model of 30-min ischemia followed by 24 hours or

7 days of reperfusion.

Results: When administered immediately prior to reper-

fusion with continuous infusion for 24 h, the combination

of NAC+IS5MN reduced infarct size (29 ± 6 vs. 59 ± 4%

area-at-risk, p<0.01) and the infiltration of polymorphonu-

clear leukocytes (226 ± 15 vs. 315 ± 18 cells mm−2 of area-

at-risk, p=0.002) and monocytes/macrophages (118 ± 8 vs.

194 ± 22 cells mm−2, p=0.012), compared to vehicle. NAC

or IS5MN alone did not reduce infarct size at 24 hours of

reperfusion. The same dose regimen of NAC and IS5MN

did not reduce infarct size with permanent ischemia for

24 hours not followed by reperfusion. After 7 days of reper-

fusion (3 days of treatment with NAC+ IS5MN or vehicle

and 4 days of wash-out), infarct size was similar in the vehi-

cle and NAC+IS5MN groups, but LV end-diastolic pressure

and diastolic LV chamber wall stress were significantly lower

in the animals treated with NAC+IS5MN (5 ± 1 mmHg and

62 ± 7 dyne mm−2, respectively) compared to vehicle (9 ± 1

mmHg and 123 ± 18 dyne mm−2, p<0.05).

Conclusion: We demonstrate in a rat model of cardiac

ischemia-reperfusion treated with NAC and IS5MN, ac-

cording to a regimen that mimicked a clinical situation

(drugs started at time of reperfusion), that the short-

term benefit seen after 24 h of reperfusion (51% reduc-

tion of infarct size) is maintained after one week, possibly

through modulation of the inflammatory response to cardiac

injury.

Key Words. heart, ischemia/reperfusion injury, infarct size,

N-acetylcysteine, isosorbide 5-mononitrate, antioxidant,

rat

Introduction

The pharmacological and surgical procedures that re-store coronary flow in the ischemic myocardium arebeneficial. Nevertheless, the overall benefit might beincreased by reducing untoward effects of reperfusion

such as arrhythmias, microvascular damage, stunningand cardiomyocyte death [1]. Indeed, it was firstdemonstrated 30 years ago that in vitro reperfusion ofhearts previously rendered anoxic caused macroscopictissue damage [2]. Since this pioneer observation, it hasbeen shown that Ca2+ overload and unopposed gen-eration of oxygen-derived free radicals are involvedin ischemia-reperfusion injury [3,4]. Under oxidativestress, the unbalance between the generation of reac-tive oxygen and nitrogen species (•O−

2 , H2O2, •OH, •NOand peroxynitrite) and their removal by antioxidantsystems results in peroxidation of cardiac membranelipids and oxidation of thiol groups, that may ultimatelycompromise cardiac function.

Reduced glutathione (GSH) is a major antioxidantin cardiomyocytes, as a direct free-radical scavengeror as a substrate for GSH peroxidase [5]. The levelsof myocardial non-protein sulfhydryl containing agents(mainly GSH) are severely depressed during reperfu-sion after prolonged ischemia in vivo [6]. Exogenousglutathione can restore compromised antioxidant de-fences during cardiac ischemia-reperfusion injury, butsuffers a limited bioavailability [7]. N-acetylcysteine(NAC) is an orally active derivative of cysteine whichrepletes glutathione levels and acts as a free radicalscavenger for species such as •OH, H2O2, HOCl−•. Therelative contribution of the 2 actions in various set-tings is still under discussion [8,9]. Beside its use asan antidote for paracetamol poisoning, NAC has beenevaluated in the setting of heart diseases where sev-eral small clinical studies have demonstrated that itmay be an effective therapeutic agent. In particular,NAC has been shown to potentiate the clinical efficacyof nitrate derivatives in acute myocardial infarction, un-stable angina and chronic heart failure [10–13]. There

Address for correspondence: Dr Roberto Latini, Department ofCardiovascular Research, Istituto di Ricerche Farmacologiche“Mario Negri”, 20157 Milano, Italy. E-mail: latini@marionegri.it

199

200 Calvillo et al.

are at least two hypotheses to explain this synergis-tic benefit. Firstly, NAC may enhance the bioavailabil-ity of NO by forming different biologically vasoactiveintermediates, including S-nitrosocysteine, that playsa role in the molecular mechanism of vascular smoothmuscle relaxation [14–18]. Secondly, NAC may reducethe development of nitrate tolerance during continu-ous nitrate therapy since the depletion of sulfhydrylgroups has been postulated as a mechanism contribut-ing to this phenomenon [19]. This hypothesis is how-ever controversial since NAC failed to reverse toler-ance to the anti-anginal effect of isosorbide dinitrate[20].

Studies on the effects of NAC in animal models ofcardiac ischemia-reperfusion injury yielded contrast-ing results as some showed no protective effects [21–23] while others a significant reduction of cardiac dam-age [8,24]. These contradictions may be due to profounddifferences in the experimental protocols for relevantvariables such as the species considered, the duration ofischemia and/or of reperfusion, and the timing of NACadministration. All published studies were concludedwithin 24 hours of reperfusion, so that no data are avail-able for longer periods of observation, when functionaland structural consequences of ischemia/reperfusion in-jury become evident. Although some studies reportedthat glyceryl trinitrate confers protective effects in ex-perimental cardiac ischemia not followed by reperfu-sion [25,26], no published data exist on the effect of an-other NO donor, isosorbide 5-mononitrate (IS5MN), inischemia and reperfusion.

In the present study, we examined the effects ofNAC on cardiac injury in a rat model of 30-min is-chemia followed by 24 hours or 7 days of reperfu-sion. NAC was administered at the beginning of reper-fusion, as i.v. bolus and then followed by continuousi.v. infusion. Since clinical benefits in acute coronarysyndromes were observed when NAC was adminis-tered in combination with a nitrate derivative, we as-sessed the effect of this combination and of either agentalone.

Materials and Methods

Animal surgery

Male Sprague-Dawley rats (range 200–285 g) wereanaesthetised with 50 mg/kg i.p. of sodium pentobar-bital and ventilated (65 breaths min−1, tidal volume1.0 ml · 100 g BW−1) through an endotracheal cannula.Silastic tubing was introduced via the right externaljugular vein, into the right atrium. The other end wasguided under the skin layer toward the interscapu-lar area where it was connected, 5 minutes prior toreperfusion, to an osmotic minipump implanted subcu-taneously.

Left coronary artery (LAD) was ligated with 4-0 silksuture (Eticon 682H) after exteriorisation of the heartthrough a 15 mm opening of the 4th intercostal space. A

plain knot was tied over two pieces of suture which werepulled after 30 min to start reperfusion. After start ofreperfusion, the chest was closed under negative pres-sure, and the rat weaned from mechanical ventilationunder ECG monitoring. Ischemia was confirmed in allrats included in the present analysis by the appearanceof ventricular ectopies on the ECG and discolorationof the heart surface, while reperfusion was verified byreddening of the previously discoloured area. Sham-operated rats underwent the same surgical procedures,without ligation of LAD. All rats received a sc injectionof buprenorphine (0.30 mg kg−1) upon awakening to re-duce acute post-surgery pain.

Procedures involving animals and their care wereconducted in conformity with the institutional guide-lines that are in compliance with national (D.L. n. 116,G.U., suppl. 40, 18 Febbraio 1992, Circolare n. 8, G.U., 14Luglio 1994) and international laws and policies (EECCouncil Directive 86/609, OJ L 358,1, Dec 12, 1987;Guide for the Care and Use of Laboratory Animals,U.S. National Research Council, 1996).

Experimental protocol

The study included two sets of experiments with reper-fusion times of either 24 h (short term) or 7 days (longterm) after 30 min of ischemia, and a group of animalswith permanent ischemia for 24 hours (no reperfusion).

In the short term protocol, 78 rats were operated and29 died within 24 hours of surgery. The animals wererandomly assigned to one of the 4 groups:

(a) 30 min ischemia, 24 h reperfusion receiving vehicle(n = 21);

(b) 30 min ischemia, 24 h reperfusion receiving NAC(n = 8);

(c) 30 min ischemia, 24 h reperfusion receiving IS5MN(n = 9);

(d) 30 min ischemia, 24 h reperfusion receivingNAC + IS5MN (n = 11).

NAC (given as a sterile injectable solution at 20% w/v,Hidonac) was first given as a single bolus in the jugularvein 5 min prior to reperfusion (2 mmol/kg) and then im-mediately after as a continuous infusion (1.0–1.2 mmolkg−1 24 h−1, Alzet model 2ML1). IS5MN was given byoral gavage at a dose of 200 mg kg−1 bid in water. In thecombination therapy, the drugs were administered asdescribed for the single treatments. Animals assignedto the vehicle group were infused with saline via os-motic pumps and subjected to water gavage twice aday.

For the experiment with permanent ischemia for 24hours (no reperfusion), a total of 20 rats were operatedand 10 died within 24 hours of sugery. The animals wereassigned to the following groups:

(e) 24 h ischemia receiving vehicle (n = 4)(f ) 24 h ischemia receiving NAC + IS5MN (n = 6)

NAC and IS5MN in Ischemia-Reperfusion 201

Administration of drugs (or vehicle) and processing ofthe hearts to determine the necrotic area were thesame as that described for the 24-hour reperfusionexperiment.

In the long term protocol, 46 rats were operated and18 died. The animals were assigned to groups as follows:

(g) sham-operation (n = 11);(h) 30 min ischemia, 7 day reperfusion receiving vehicle

(n = 8);(i) 30 min ischemia, 7 days reperfusion receiving

NAC + IS5MN (n = 9);

NAC and IS5MN were administered for 3 days post-reperfusion according to the routes and dosages de-scribed above, and the animals were sacrificed onday 7.

There were no evidences of extremely large infarctsin the animals found dead prior to sacrifice, as checkedwith tetrazolium chloride (TTC) staining. No signs ofacute heart failure, such as pulmonary edema, were ob-served in animals whose death was witnessed. Animalsfound dead prior to sacrifice were not included in finalanalyses of hemodynamic and histological data.

Determination of necrotic area, area at risk,

polymorphonuclear (PMN) leukocyte and

monocyte/macrophage infiltration

After 24 h of reperfusion the chest was reopened un-der anaesthesia and ventilation, the ligature around thecoronary artery was re-tightened and 2 ml of 5% EvansBlue in water were injected into the right ventricu-lar chamber to distinguish the area-at-risk (AAR) fromthe perfused area of the left ventricle. The heart wasblotted, excised, quickly frozen at −20◦C in OCT, andchopped transversally into ten to fourteen 1-mm serialsections with a McIlwain tissue chopper.

The tissue not stained with Evans blue was dissectedfrom the stained tissue in all sections, except for a me-dian one that was fixed in formaldehyde, embedded inparaffin and then processed for the quantification of in-filtrating inflammatory cells (see below). The unstainedtissue was then incubated for 20 min at 20◦C in TTC,then transferred to 4% buffered formaldehyde. The ar-eas of perfused tissue (blue), viable ischemic tissue (red)and dead ischemic tissue (white) were measured by animage analyser (AIS, Analytical Imaging Station, ver-sion 3.0) and used to determine the volume of perfusedLV, AAR and necrotic area (Fig. 1).

Polymorphonuclear (PMN) infiltration in the AARwas determined in 5 µm sections, cut from the paraffinembedded median section of the heart, with a specificstain (Naphtol AS-D chloroacetate esterase, Sigma).PMN were counted in the AAR at a magnification of400×, and expressed as cells/mm2 of AAR.

Quantification of infiltrating monocytes/macro-phages was performed using the monoclonal antibodyED-1 (Serotec) and the avidin-biotin-enzyme complex

peroxidase method for detection. Nuclei were counter-stained with haematoxylin. The number of positivelystained ED-1 cells were counted in at least 30 gridfields (area 0.09 mm2) at a magnification of 400× andexpressed by mm2 of AAR.

Hemodynamics

Rats reperfused for 7 days underwent a differentprotocol which included hemodynamic evaluation andfixation-perfusion of the heart for histomorphometricevaluation.

Under pentobarbital anaesthesia (50 mg kg−1

i.p.), a microtip pressure transducer (Millar SPC-320,Houston, TX, USA) connected to an amplifier andrecorder (Windowgraph, Gould Electronics, ValleyView, OH, USA) was inserted into the right carotidartery to measure systolic (SBP) and diastolic (DBP)blood pressure, and heart rate (HR). The pressuretransducer was then advanced into the LV to measureLV systolic (LVSP) and end-diastolic (LVEDP) pres-sure and the first derivatives (positive and negative) ofLV pressure over time (±dP dt−1).

Fixation-perfusion of the heart

Immediately after the hemodynamic measurements, aPE-60 cannula was inserted into the abdominal aortaand the heart was arrested in diastole by injecting 4 mlof KCl 0.5 M into the aorta. The heart was perfusedretrogradely for 3 min with a Ca2+- and Mg2+-free PBScontaining 10 U ml−1 of heparin followed by phosphate-buffered 4% formaldehyde for 30 min at 100 mmHg.Finally the LV was filled with the same solution at apressure equivalent to the LVEDP measured in vivoand the hearts were kept in this fixative for at least24 h, before the evaluation of the LV morphology.

Histomorphometric evaluation

The atria were trimmed, then the RV free wall andLV inclusive of septum were separately weighed. TheLV internal length was measured using a calibratedprobe. The heart was cut into 3 slices (basal, medianand apical) of similar thickness along its long axis andthe LV morphology of the median slice was evaluatedwith an image analyser (AIS, version 3.0) before em-bedding in paraffin. The section was placed on a transil-luminator, an image acquired with a fixed photographiclens and digitised. The LV wall thickness (LVWT) wasmeasured at 8 different points in the septum and freewall. The LV chamber area was measured by tracingthe chamber perimeter manually, and the LV diameterwas determined from this area. LV chamber volumewas calculated according to the formula: chamber vol-ume (mm3) = r2 × LV inner length × 4π /3, where r isthe chamber radius.

Left ventricular diastolic average wall stress (WS,dyne mm−2) was calculated according to the formula:

202 Calvillo et al.

Fig. 1. Identification of non-viable ischemic left ventricular tissue and infarcted area after ischemia/reperfusion. Rats were subjectedto 30 min of ischemia followed by 24 hours or 7 days of reperfusion. After 24 hours, the ventricular tissue not stained with Evans bluewas dissected from the stained tissue and incubated in tetrazolium chloride (TTC), staining in red the viable ischemic tissue and whitethe dead ischemic tissue. After 7 days, paraffin-embedded 5 µm histological sections were stained with hematoxylin-eosin (H&E) andinfarct area measured as described in the Methods section. The figure compares the extent of acute and long-term tissue damage afterlarge (left panels) and small infarcts (right panels).

WS = (LVEDP × Rm × 13.33)/(2 × LVWT), whereLVEDP is the LV end-diastolic pressure (mmHg),LVWT = [(mean septum thickness) + (3 × mean freewall thickness)]/4, and Rm = chamber radius + (0.5 ×LVWT).

Each slice was then embedded in paraffin and 5 µmsections stained with hematoxylin-eosin (H&E, Fig. 1).The infarct area was measured on 2 equally spaced sec-tions cut between the ligature and the cardiac apex at amagnification of 40×, by counting the number of pointsof an ocular grid with 121 points (Olympus) overlyinginfarcted or spared myocardium of the left ventricle andexpressed as a percentage of the total number of pointsoverlying the LV.

The left ventricular free wall myocyte nuclei werecounted and the myocyte volume per nucleus was cal-culated on the 2 histological sections used for scar sizemeasurement, as previously described [27]. Quantifi-cation of infiltrating monocytes/macrophages was per-formed as described for the 24-hour reperfusion exper-iment.

Statistical analysis

All data are reported as mean ± standard error of themean for the number of animals indicated. Compar-isons between experimental groups were performedby ANOVA followed by Dunnett’s multiple comparisontest. A two-tailed p value less than 0.05 was consideredsignificant. All tests were performed with the softwareGraphPad Prism version 3.00 for Windows.

Results

After 24 hours of reperfusion, mortality rates of thefour experimental groups were not statistically differ-ent and averaged 37% of the animals operated. TheAAR was also similar in all groups, being 37 ± 2,42 ± 3, 45 ± 3 and 39 ± 5% of LV area for vehicle,NAC, IS5MN and NAC +IS5MN groups, respectively(ANOVA, p = 0.50). On the other hand, the infarctarea was significantly reduced in the animals treatedwith the combination NAC + IS5MN (29 ± 6% AAR)

NAC and IS5MN in Ischemia-Reperfusion 203

Fig. 2. Infarct area of rats with ischemia/reperfusion injury.Rats were subjected to 30 min of ischemia followed by 24 hoursof reperfusion. NAC (1.2 mmol kg−1 ev), IS5MN (200 mg kg−1

bid po), their combination or vehicle were started at reperfusion.Myocardial perfusion was delineated on serial transversal1-mm thick sections stained with Evans Blue, leaving the AARunstained. The necrotic area of the non perfused tissue wasfurther stained with tetrazolium chloride to quantify infarctarea with an automatic analyzer. ∗p < 0.05 vs. vehicle byDunnett’s multiple comparison test.

when compared to the other groups (vehicle: 59 ± 4%,NAC: 49 ± 8%, IS5MN: 45 ± 5%, ANOVA: p = 0.0033,NAC + IS5MN vs. vehicle: p < 0.01) (Fig. 2).

Infarct size after 24 h of permanent ischemia (noreperfusion) was 88±7% of AAR in vehicle and 83±7%in NAC +IS5MN group (p = N.S.)

Figure 3a shows leukocyte infiltration in the AARafter 24 hours of reperfusion. The combination ther-apy significantly reduced leukocyte density from 315 ±18 cells mm−2 of AAR in the vehicle group to 226 ±15 cells mm−2 of AAR (p = 0.002). There were 11 ± 1leukocytes mm−2 in the interventricular septum ofthese animals. Similarly, monocyte/macrophage infil-tration in the AAR of the group of animals treated

Fig. 3. Leukocyte and monocyte/macrophage infiltration in the myocardium of rats with ischemia/reperfusion injury. Rats weresubjected to 30 min of ischemia followed by 24 hours of reperfusion. NAC (1.2 mmol kg−1 ev) + IS5MN (200 mg kg−1 bid po), orcorresponding vehicle were started at reperfusion. After sacrifice, a median section of the heart was fixed in formaldehyde to determineleukocyte (Fig. 3a) and monocyte/macrophage (Fig. 3b) infiltration in AAR. Cell density expressed as number of cells mm−2 .

with NAC + IS5MN (118 ± 8 cells mm−2) was signif-icantly lower than in the vehicle group (194 ± 22 cellsmm−2, p = 0.012, Fig. 3b). There were 6 ± 1 mono-cytes/macrophages mm−2 in the perfused myocardiumof these animals. Given the benefits afforded by thecombination therapy after 24 hours of reperfusion, fur-ther investigations were restricted to this group oftreatment.

Long-term (7 days of reperfusion) mortalitywas not significantly different between vehicle andNAC + IS5MN groups (χ2 = 0.93, p = 0.33) and av-eraged 43% of the animals operated. None of the sham-operated rats died. Infarct area was similar in the ve-hicle (18 ± 1% of LV area) and NAC +IS5MN groups(19 ± 3%). There were 24.7 ± 2.5 106 myocyte nucleiin the left ventricular free wall of the sham-operatedanimals. Myocyte loss amounted to 30% (residual num-ber 17.4 ± 2.1 106 myocyte nuclei) and 32% (16.8 ± 1.7106 myocyte nuclei) in the vehicle and NAC +IS5MNgroups, respectively (p< 0.05 vs. sham for each group).The volume of viable myocytes was significantly higherin the vehicle (20,610 ± 639 µm3) than in the shamgroup (15,584 ± 539 µm3, p< 0.0001). The combinedtreatment did not affect this hypertrophic processsince myocyte volume was 19,963 ± 718 µm3 in theNAC +IS5MN group (p < 0.001 vs. sham, not dif-ferent from vehicle). The number of infiltrating mono-cytes/macrophages was not statistically different inthe AAR of the vehicle (3385 ± 966 cells mm−2) andNAC + IS5MN groups (3439 ± 939 cells mm−2).

Hemodynamic variables were measured 4 days aftertermination of drug therapy that was given for 3 days(Table 1). There was no significant difference betweenthe experimental groups for systemic blood pressure,but heart rate was higher (ca. 50 beats min−1) in the twogroups of operated animals compared to the sham group(ANOVA p < 0.0001). LV end-diastolic pressure was

204 Calvillo et al.

Table 1. Hemodynamics of rats with ischemia/reperfusion

Sham (n = 10) Vehicle (n = 16) NAC +IS5MN (n = 9) ANOVA (p value)

SBP (mmHg) 123 ± 6 116 ± 2 114 ± 4 0.28DBP (mmHg) 94 ± 4 92 ± 2 89 ± 4 0.59HR (beat min−1) 393 ± 15∗ 451 ± 11 441 ± 9 <0.0001LVEDP (mmHg) 5 ± 1∗ 9 ± 1 5 ± 1∗ 0.0086LVSP (mmHg) 118 ± 6 118 ± 3 109 ± 5 0.31LV + dP/dt (mmHg s−1) 8650 ± 409 8884 ± 364 9111 ± 570 0.80

Rats were subjected to 30 min of ischemia followed by 7 days of reperfusion. Drug therapy (NAC 1.2 mmol kg−1 ev + IS5MN 200 mg/kg bid po, or correspondingvehicle) was started at reperfusion and lasted 3 days. A group of sham-operated was included. Systemic and LV hemodynamics were measured on day 7 beforesacrifice. Data are expressed as mean ± SEM. SBP, DBP: systolic and diastolic blood pressure, HR: heart rate, LVEDP: LV end-diastolic pressure, LVSP:LV systolic pressure, LV +dP dt−1: positive time derivative of LV pressure. ∗: p < 0.05 vs. vehicle (Dunnett’s multiple comparison test).

Table 2. Body weight, cardiac weight, and histomorphometric variables of hearts from rats with ischemia/reperfusion injury

Sham (n = 10) Vehicle (n = 16) NAC +IS5MN (n = 9) ANOVA (p value)

Body weight (g) 256 ± 9 262 ± 5 266 ± 9 0.67Heart weight (mg) 1051 ± 41 1086 ± 38 1090 ± 68 0.83LV weight (mg) 808 ± 27 835 ± 31 827 ± 48 0.86RV weight (mg) 222 ± 13 219 ± 9 229 ± 19 0.86LV chamber diameter (mm) 6.1 ± 0.3 6.5 ± 0.3 6.1 ± 0.3 0.55LV chamber length (mm) 12.6 ± 0.3 12.0 ± 0.2 12.0 ± 0.4 0.26LV chamber volume (mm3) 255 ± 27 274 ± 26 239 ± 27 0.66Diastolic wall stress (dyne mm−2) 80 ± 9 123 ± 18 62 ± 7∗ 0.022

Rats were subjected to 30 min of ischemia followed by 7 days of reperfusion. Drug therapy (NAC 1.2 mmol kg−1 ev + IS5MN 200 mg kg−1 bid po, orcorresponding vehicle) was started at reperfusion and lasted 3 days. A group of animal was sham-operated. After 7 days of reperfusion, the heart wasarrested in diastole and fixed under controlled hydrostatic pressure. ∗: p < 0.05 vs. vehicle (Dunnett’s multiple comparison test).

significantly elevated in the vehicle group (9 ± 1 mmHg)compared to the sham-operated animals (5 ± 1 mmHg).Three days treatment with the combination of NAC andIS5MN completely normalized LVEDP (5 ± 1 mmHg,p < 0.05 vs. vehicle).

There were no significant differences in body weight,heart and LV weights, and LV chamber diameter andvolume between sham-operated and vehicle-treatedanimals with I/R injury on day 7 (Table 2). The combi-nation therapy did not affect these morphometric vari-ables. The average diastolic LV chamber wall stresswas significantly lower in the animals treated withNAC +IS5MN (62 ± 7 dyne mm−2) compared to thevehicle group (123 ± 18 dyne mm−2, p < 0.05), due tothe reduction of LV end-diastolic pressure.

Discussion

In the present study, we showed that treatment of ratssubjected to 30 min of cardiac ischemia followed by24 hour of reperfusion, with the combination of a thiolantioxidant (NAC) and a nitrate derivative (IS5MN),but not with either drug alone, significantly reduced in-farct size. This protection was maintained over a longerperiod of time since, at 7 days post-reperfusion (withonly 3 days of treatment with the combination therapy),

LV end-diastolic pressure and diastolic wall stress, animportant determinant of LV remodelling, were signif-icantly lower than in untreated animals.

The dose of NAC used in the present experiment(333 mg kg−1 as an i.v. bolus before reperfusion plus200 mg kg−1 day−1 for 1 or 3 consecutive days, i.e. a cu-mulative dose ranging from 0.5 to 1.0 g kg−1 BW i.v.) isamong the highest used in previously published reportson the protective effects of NAC in different experi-mental models of organ damage [8,23,28–31]. The doseused has been shown to be safe in the rat during longterm (28 weeks) administration (data on file, ZambonGroup). Given the scarcity of reports on the effect ofIS5MN in rat models of myocardial infarction, the doseof IS5MN (200 mg kg−1 BW b.i.d. by oral gavage) wasselected on the basis of a previous study showing a re-duction of systolic blood pressure by 10–15 mmHg overthe first 4 hours after CAO [32] and the fact that hy-potension in the first hours after MI has been shown tobe detrimental [25]. It is therefore unlikely that higherdoses of NAC or IS5MN alone would have led to a sub-stantial benefit.

Infarct size reduction after 24 hours of reperfusionaveraged 51% in the group of animals treated withthe combination therapy. In the first hours of reper-fusion of a previously ischemic area, PMN migrateinto the cardiac tissue and release compounds toxic

NAC and IS5MN in Ischemia-Reperfusion 205

to the cardiac myocytes (ROS, proteolytic enzymes,leukotrienes, thromboxane A2 and the complement).This acute inflammatory response contributes to car-diac damage [33], therefore, various pharmacologicalapproaches have been proposed to protect the my-ocardium during this phase [34]. Permanent ischemiais associated to a weaker inflammatory response. Here,we show that combination therapy with NAC + IS5MNgiven at reperfusion, significantly reduced PMN in-filtration in the area-at-risk by 28% and mono-cyte/macrophage infiltration by 39% after 24 hours.Interpretation of this observation is not straight-forward since reduced infiltration of polymorphonu-clear leukocytes and macrophages in the infarctedarea could be a consequence of reduced tissue dam-age according to mechanisms that are not related tothe anti-inflammatory property of NAC and/or dueto a true anti-inflammatory mechanism. However, re-duction of monocytes or leukocytes adhesion to en-dothelial cells by NAC has already been documented[35,36] and could explain the reduced infiltration shownhere. The observation that infarct size was unaffectedby NAC + IS5MN when the heart was not reper-fused, further supports the hypothesis that an anti-inflammatory/anti-ROS action contributed to the pro-tective effect of the combined treatment.

Our data do not provide a mechanistic explanationfor the protective effect of NAC. However, several hy-potheses can be put forward:

(1) It is possible that NAC inhibits the inflammatoryresponse driven by the ischemic event. In fact,anoxia has long been recognized as an inducer forproduction of inflammatory cytokines directly onmononuclear cells [37], and this is thought to be atthe basis of the inflammatory response associatedwith ischemic diseases. From this point of view,NAC would likely act by inhibiting the synthesisof inflammatory cytokines, including TNF and IL-1[38,39] and chemokines [40], possibly through inhi-bition of the transcription factor NF-κB [41,42] aswell as by decreasing the upregulation of adhesionmolecules [43]. Therefore, NAC might be protec-tive by inhibiting PMN accumulation or productionof cytokines such as TNF, that contribute to the is-chemic damage [44].

(2) NAC is as a potent free radical scavenger [9].(3) Thiol-containing compounds like the ACE in-

hibitor zofenopril and NAC may be cardiopro-tective through their interaction with the ATP-sensitive potassium channel [45]. The critical roleof the sulfhydryl group has been also underlinedin a rat model of heart failure after myocardial in-farction where thiol donors (zofenopril and NAC)restored NO-dependent vasodilation induced byacetylcholine in isolated aortic ring preparations[46].

(4) NAC improves coronary and peripheralendothelium-dependent vasodilation [47].

(5) It is also possible, however, that in vivo the inflam-matory response will not only result from a directeffect of anoxia on cytokine-producing cells but alsoas a consequence of tissue injury. This has been sug-gested in other models of ischemia/reperfusion in-jury, in the kidney and the brain, where inhibitionof apoptosis reduced the inflammatory response[48,49]. Therefore, NAC might reduce inflammationin our model by preventing apoptosis, as previouslyobserved in isolated cardiac myocytes subjected tooxidative stress [50,51].

On the other hand, an apparent early reduction in in-farct size might be due to a delay in injury onset ratherthan to a real protective effect. Moreover, interferingwith the process of inflammation might be detrimen-tal over long term since it could prevent the formationof scar tissue that later contributes to the mechanicalcohesion of the damaged ventricle [52]. Most experi-mental studies showing reduction of infarct size aftercardiac ischemia-reperfusion injury by drugs are con-cluded within 24 hours of reperfusion, when infarct ex-pansion is still very active, so that it is not possible toinfer whether the benefits observed acutely will trans-late in a better function and/or morphology of the in-jured ventricle over a longer period of time. We there-fore planned experiments with a reperfusion time of1 week, when the acute phase of reparative processesis considered to be concluded. After 7 days, animalsof the vehicle group showed a mild LV dysfunction, asrevealed by a slight increase in LV end-diastolic pres-sure (9 ± 1 mmHg) when compared to sham (5 ± 1mmHg, p < 0.05). The combined therapy with NACand IS5MN fully normalized this hemodynamic vari-able, even though it was given only for the first 3 days ofreperfusion. In this respect, acute hypotensive effectsof the combined therapy leading to beneficial decreasein afterload cannot be excluded since hemodynamicmeasurements were performed on day 7, after 4 daysof wash-out. Early hypotension with NAC +IS5MN, ifit occurred, was not severe enough to increase earlymortality.

A benefit of NAC on tolerance to nitrate cannotbe excluded, although it cannot be proven. NO donorsalone have been shown to reduce infarct size in onestudy in the rat [26] and in the pig [53]. In our case, wefound a non significant reduction (25%) in MI size withIS5MN alone; the effect, if any, might have been limitedby early development of tolerance.

The experimental model, as expected, did not lead toprofound alterations of LV morphology after 1 week ofreperfusion: cardiac weights and LV chamber dimen-sions were not significantly different between sham-operated and vehicle-treated animals with ischemia-reperfusion injury. After 7 days of reperfusion, infarctarea was similar in the vehicle (18 ± 1% of LV area)and NAC +IS5MN (19±3%) groups. This result is con-firmed by the observation that the number of resid-ual myocytes was identical in the LV free wall of the

206 Calvillo et al.

vehicle and NAC +IS5MN groups. The infarct area af-ter 24 hours of reperfusion, when normalized by LVarea rather than by AAR, was not significantly dif-ferent than that measured after 7 days of reperfusion,even if different histological methods were used at eachtime point: infarct area amounted to 19 ± 3, 21 ± 4,20 ± 2 and 11 ± 2% of LV area after 24 hours of reper-fusion in the vehicle, NAC, IS5MN and NAC +IS5MNgroups, respectively. However, the combination ther-apy succeeded in normalizing the average diastolic LVwall stress, an important contributor to infarct expan-sion and LV remodelling, mainly due to the decrease inLVEDP. The better LV function observed at 7 days withNAC +IS5MN cannot be easily explained by merelyextrapolating the 51% decrease in MI size at 24 h sincescar size was not different.

Limitations

In the present study, we assumed that reperfusion fol-lowed ischemia, at least to a certain extent of the areaat risk, based on 3 kinds of evidences, each one notfully satisfactory by itself, but quite convincing whenconsistent:

1. epicardial blushing,2. ST normalization starting with reperfusion,3. morphology of infarction, never transmural, as com-

monly observed for permanent ligation.

Moreover, in a separate series of rats, we injectedin vivo Evans Blue without re-occlusion of coronaryartery: the whole heart, including area at risk, stainedblue at serial sectioning, thus suggesting completereperfusion.

In conclusion, we demonstrate in a rat model of car-diac ischemia and reperfusion treated with the com-bination of NAC and IS5MN, according to a protocolthat mimicked clinical situation (drugs given at timeof reperfusion, intravenously for NAC), that the short-term benefits seen after 24 h of reperfusion with a 51%reduction of infarct area is maintained after one weekinasmuch as LV diastolic wall stress was fully normal-ized by a 3-day treatment. Reduction of leukocyte in-filtration could be responsible, at least in part, for thisbenefit.

Acknowledgments

This study was supported by a grant from Zambon Group, Bresso,Italy.

References

1. Verma S, Fedak PWM, Weisel RD, et al. Fundamentals ofreperfusion injury for the clinical cardiologist. Circulation2002;105:2332–2336.

2. Hearse DJ, Humphrey SM, Chain EB. Abrupt reoxygena-tion of the anoxic potassium-arrested perfused rat heart: Astudy of myocardial enzyme release. J Molec Cell Cardiol1973;5:395–407.

3. Kaplan P, Lehotsky J, Racay P. Role of sarcoplasmic retic-ulum in the contractile dysfunction during myocardial is-chemia and reperfusion. Physiol Res 1997;46:333–339.

4. Bolli R, Jeroudi MO, Patel BS, et al. Direct evidence thatoxygen-derived free radicals contribute to postischemic my-ocardial dysfunction in the intact dog. Proc Natl Acad Sci(USA) 1989;86:4695–4699.

5. Dhalla NS, Elmoselhi AB, Hata T, Makino N. Status of my-ocardial antioxidants in ischemia-reperfusion injury. Cardio-vasc Res 2000;47:446–456.

6. Lesnefsky EJ, Dauber IM, Horwitz LD. Myocardialsulfhydryl pool alterations during reperfusion after briefand prolonged myocardial ischemia in vivo. Circ Res 1991;68:605–613.

7. Ramires PR, Ji LL. Glutathione supplementation and train-ing increases myocardial resistance to ischemia-reperfusionin vivo. Am J Physiol 2001;281:H679–H688.

8. Sochman J, Kolc J, Vrana M, Fabian J. Cardioprotective ef-fects of N-acetylcysteine: The reduction in the extent of in-farction and occurrence of reperfusion arrhythmias in thedog. Int J Cardiol 1990;28:191–196.

9. Aruoma OI, Halliwell B, Hoey BM, Butler J. The antioxidantaction of N-acetylcysteine: Its reaction with hydrogen per-oxide, hydroxyl radical, superoxide, and hypochlorous acid.Free Radic Biol Med 1989;6:593–597.

10. Horowitz JD, Antman EM, Lorell BH, Barry WH, SmithTW. Potentiation of the cardiovascular effects of nitro-glycerin by N-acetylcysteine. Circulation 1983;68:1247–1253.

11. Mehara A, Shotan A, Ostrzega E, Hsueh W, Vasquez-Johnson J, Elkayam U. Potentiation of isosorbide dinitrateeffects with N-acetylcysteine in patients with chronic heartfailure. Circulation 1994;89:2595–2600.

12. Arstall MA, Yang J, Stafford I, Betts WH, Horowitz JD. N-acetylcysteine in combination with nitroglycerin and strep-tokinase for the treatment of evolving acute myocardial in-farction. Circulation 1995;92:2855–2862.

13. Ardissino D, Merlini PA, Savonitto S, et al. Effect of trans-dermal nitroglycerin or N-acetylcysteine, or both, in thelong-term treatment of unstable angina pectoris. J Am CollCardiol 1997;29:941–947.

14. Ignarro LJ, Lippton H, Edwards JC, et al. Mechanism of vas-cular smooth muscle relaxation by organic nitrates, nitrites,nitroprusside and nitric oxide: Evidence for the involvementof S-nitrosothiols as active intermediates. J Pharmacol ExpTher 1981;218:739–749.

15. Loscalzo J. N-acetylcysteine potentiates inhibition ofplatelet aggregation by nitroglycerin. J Clin Invest 1985;76:703–708.

16. Stamler J, Mendelsohn ME, Amarante P, et al. N-acetylcysteine potentiates platelet inhibition by endothe-lium-derived relaxing factor. Circ Res 1989;65:789–795.

17. Cooke JP, Stamler J, Andon N, Davies PF, McKinley G,Loscalzo J. Flow stimulates endothelial cells to release anitrovasodilator that is potentiated by reduced thiol. Am JPhysiol 1990;259:H804–H812.

18. Scharfstein JS, Keany JF, Jr, Slivka A, et al. In vivo transferof nitric oxide between a plasma protein-bound reservoirand low molecular weight thiols. J Clin Invest 1994;94:1432–1439.

NAC and IS5MN in Ischemia-Reperfusion 207

19. Elkayam U. Tolerance to organic nitrates: Evidences, mech-anisms, clinical relevance and strategies for prevention. AnnIntern Med 1991;14:667–677.

20. Parker JO, Farrel B, Lahey KA, Rose BF. Nitrate toler-ance: The lack of effect of N-acetylcysteine. Circulation1987;76:572–576.

21. Forman MB, Puett DW, Cates CU, et al. Glutathione re-dox pathway and reperfusion injury. Effect of N-acetylcys-teine on infarct size and ventricular function. Circulation1988;78:202–213.

22. Tripathi Y, Hedge BM. Effect of N-acetylcysteine on myocar-dial infarct size and reperfusion in dogs. Indian J PhysiolPharmacol 1998;42:50–56.

23. Kingma JG, Jr, Rouleau JL. Effect of N-acetylcysteine ontissue necrosis during acute myocardial infarction in rabbits.Can J Cardiol 1989;5:321–326.

24. Alberola A, Such L, Gil F, Zaragoza R, Morcillo EJ. Pro-tective effect of N-acetylcysteine on ischaemia-induced my-ocardial damage in canine heart. Naunyn-Schmiederberg’sArch Pharmacol 1991;343:505–510.

25. Jugdutt B. Myocardial salvage by intravenous nitroglyc-erin in conscious dogs: Loss of beneficial effect with markednitroglycerin-induced hypotension. Circulation 1983;68:673–684.

26. Chiariello M, Brevetti G, Ambrosio G, Cataffo A, Pollice U,Condorelli M. Long-term protection of ischemic myocardiumby nitroglycerin ointment. Cardiovasc Res 1984;8:321–325.

27. Anversa P, Beghi C, Kikkawa Y, Olivetti G. Myocardialresponse to infarction in the rat. Morphometric measure-ment of infarct size and myocyte cellular hypertrophy. AmJ Pathol 1985;18:484–492.

28. Davreux CJ, Soric I, Nathens AB, et al. N-acetyl cysteineattenuates acute lung injury in the rat. Shock 1997;8:432–438.

29. Schmidt W, Walther A, Gebhard MM, Martin E, SchmidtH. Influence of N-acetylcysteine treatment on endotoxin-induced microcirculatory disturbances. Intensive Care Med1998;24:967–972.

30. Victor VM, Guayerbas N, Garrotte M, Del Rio M, De LaFuente M. Modulation of murine macrophage function byN-acetylcysteine in a model of endotoxic shock. BioFactors1999;10:347–357.

31. Cuzzocrea S, Mazzon E, Costantino G, Serraino I, De SarroA, Caputi A. Effects of N-acetylcysteine in a rat model of is-chemia and reperfusion injury. Cardiovasc Res 2000;47:537–548.

32. Riva E, Kurosaki M, Porzio S, Latini R, Lagrasta C,Olivetti G. Effects of an early treatment with lisinopril andisosorbide-5-mononitrate on hemodynamics and late ven-tricular remodelling in rats with 9-week myocardial infarc-tion. Cardioscience 1995;6:139–146.

33. Hansen PR. Role of neutrophils in myocardial ischemia andreperfusion. Circulation 1995;91:1872–1885.

34. Latini R, Masson S, Bertini R, Maggioni AP, Ghezzi P,Calvillo L. Cardiac protection by pharmacological modu-lation of inflammation. Expert Opin Investig Drugs 2001;10:1913–1924.

35. Weber C, Erl W, Pietsch A, Strobel M, Ziegler-HeitbrockHW, Weber PC. Antioxidants inhibit monocyte adhesionby suppressing nuclear factor-kappa B mobilization and in-duction of vascular cell adhesion molecule-1 in endothelialcells stimulated to generate radicals. Arterioscler Thromb1994;4:1665–1673.

36. Schmidt H, Schmidt W, Muller T, Bohrer H, GebhardMM, Martin E. N-acetylcysteine attenuates endotoxin-induced leukocyte-endothelial cell adhesion and macro-molecular leakage in vivo. Crit Care Med 1997;25:858–863.

37. Ghezzi P, Dinarello CA, Bianchi M, Rosandich ME, RepineJE, White CW. Hypoxia increases production of interleukin-1 and tumor necrosis factor by human mononuclear cells.Cytokine 1991; 3:189–194.

38. Peristeris P, Clark BD, Gatti S, et al. N-acetylcysteine andglutathione as inhibitors of tumor necrosis factor production.Cell Immunol 1992;40:390–399.

39. Leff JA, Wilke CP, Hybertson BM, Shanley PF, Beehler CJ,Repine JE. Postinsult treatment with N-acetyl-L-cysteinedecreases IL-1-induced neutrophil influx and lung leak inrats. Am J Physiol 1993; 265:L501–L506.

40. Sato M, Miyazaki T, Nagaya T, et al. Antioxidants in-hibit tumor necrosis factor-alpha mediated stimulation ofinterleukin-8, monocyte chemoattractant protein-1, and col-lagenase expression in cultured human synovial cells. JRheumatol 1996; 23:432–438.

41. Schreck R, Rieber P, Baeuerle PA. Reactive oxy-gen intermediates as apparently widely used mes-sengers in the activation of the NF-kappa B tran-scription factor and HIV-1. EMBO J 1991;10:2247–2258.

42. Cargnoni A, Ceconio C, Gaia G, Agnoletti L, Ferrari R. Cel-lular thiols redox status: A switch for NF-kB activation dur-ing myocardial post-ischaemic reperfusion. J Molec Cell Car-diol 2002;34:997–1005.

43. Marui N, Offermann MK, Swerlick R, et al. Vascular cell ad-hesion molecule-1 (VCAM-1) gene transcription and expres-sion are regulated through an antioxidant-sensitive mech-anism in human vascular endothelial cells. J Clin Invest1993,92:1866–1874.

44. Maekawa N, Wada H, Kanda T, et al. Improved my-ocardial ischemia/reperfusion injury in mice lacking tu-mor necrosis factor-alpha. J Am Coll Cardiol 2002;39:1229–1235.

45. Sargent CA, Sleph PG, Dzwnczyk S, et al. Cardioprotec-tion in ischemic rat hearts with SH-containing angiotensin-converting enzyme inhibitor zofenopril: Possible involve-ment of the ATP-sensitive potassium channel. J PharmacolExp Ther 1993;265:609–618.

46. Buikema H, Monnink SHJ, Tio RA, Crijns HJGM, De ZeeuwD, Van Gilst WH. Comparison of zofenopril and lisino-pril to study the role of the sulfhydryl-group in improve-ment of endothelial dysfunction with ACE-inhibitors in ex-perimental heart failure. Br J Pharmacol 2000;30:1999–2007.

47. Andrews NP, Prassad A, Quyyumi AA. N-acetylcysteine im-proves coronary and peripheral vascular function. J Am CollCardiol 2001;37:117–123.

48. Daemen MA, van’t Veer C, Denecker G, et al. Inhibition ofapoptosis induced by ischemia-reperfusion prevents inflam-mation. J Clin Invest 1999;104:541–549.

49. Rabuffetti M, Sciorati C, Tarozzo G, Clementi E, ManfrediAA, Beltramo M. Inhibition of caspase-1-like activity by Ac-Tyr-Val-Ala-Asp-chloromethyl ketone induces long-lastingneuroprotection in cerebral ischemia through apoptosis re-duction and decrease of proinflammatory cytokines. J Neu-rosci 2000;20:4398–4404.

50. Inserte J, Taimor G, Hofstaetter B, Garcia-Dorado D, PiperHM. Influence of simulated ischemia on apoptosis induction

208 Calvillo et al.

by oxidative stress in adult cardiomyocytes of rats. Am JPhysiol 2000;278:H94–H99.

51. Yamamoto S, Seta K, Morisco C, Vatner SF, Sadoshima J.Chelerythrine rapidly induces apoptosis through generationof reactive oxygen species in cardiac myocytes. J Molec CellCardiol 2001;33:1829–1848.

52. Mannisi JA, Weisman HF, Bush DE, Dudeck P, Healy B.Steroid administration after myocardial infarction promotesearly infarct expansion. J Clin Invest 1987;79:1431–1439.

53. Galie N, Guarnieri C, Ussia GP, et al. Limitation of myocar-dial infarct size by nicorandil after sustained ischemia in pigs.Br J Pharmacol 1995;26:477–484.