Objectives: Nitric oxide (NO) has complex effects on myocardial function particularly following ischaemia–reperfusion. The goal of this study was to examine the result of repetitive myocardial stunning on myocardial NO release and expression of inducible (iNOS) and constitutive (eNOS) NO synthases. Methods and results: Propofol anaesthetised pigs underwent ten, 2-min episodes of circumflex artery occlusion (n=6) or acted as sham operated controls (n=4). Measurements of segment shortening demonstrated a fall in function in the ischaemic territory to 52.5±7.3% (mean±S.E.M.) of baseline shortening 30 min after the stunning stimulus, recovering to 92±8.7% 5.5 h later. Function remained stable in sham controls. The change in venous–arterial [NO] between baseline and 6 h reperfusion was found to be significantly different between the two groups (0.2±0.7 in stunned vs. −4.3±1.6 μM in shams; P<0.02). Western blotting and band optical density used to compare tissue from stunned territory (S), non-stunned territory (IC) and sham control animals (SC) demonstrated this was associated with an increase in the expression of both iNOS (S: 93±13.4, IC: 37±2.4 and SC: 25±4 [arbitrary units], P<0.01 and P=0.031) and eNOS (S: 104±7.4, IC; 62.5±7.4 and SC; 75.7±0.6, P<0.03 and P<0.01) in stunned myocardium. Immunocytochemistry localised iNOS reactivity to vascular smooth muscle cells and cardiomyocytes in stunned tissue and eNOS reactivity to endothelial cells. Conclusion: Recovery from repetitive myocardial stunning is associated with the increased expression of both iNOS and eNOS and would be compatible with a protective role for both these enzymes. This finding has possible relevance for both the late window of ischaemic preconditioning and myocardial hibernation.
Time for primary review 50 days.
Nitric oxide (NO) is synthesised from the amino acid l-arginine by a family of nitric oxide synthases (NOS). Three isoforms of NO synthase have been identified: endothelial type or constitutive NOS (eNOS), macrophage type or inducible NOS (iNOS) and neuronal type . Of these isoforms eNOS has been found to be constitutively expressed in cardiac endothelial  and endocardial cells  in a wide variety of species, including the pig, and in cardiomyocytes in both rodents and humans . By contrast, iNOS is a cytokine-inducible, pro-inflammatory enzyme that has not been widely studied in the pig heart but has been demonstrated under pathological conditions in cardiomyocytes, vascular smooth muscle cells, microvascular endothelium, endocardium and monocyte/macrophages in an assortment of other species .
Investigation into the actions of NO has revealed a spectrum of actions that may modulate the heart’s response to ischaemia–reperfusion injury. However, despite much study and subsequent controversy over the effects of NO manipulation in ischaemia–reperfusion injury following prolonged (>30 min) [6,7] and briefer [8–10] coronary occlusion, little is known about the expression of eNOS and iNOS following brief, repeated episodes of ischaemia–reperfusion leading to stunning.
This has recently become of greater interest as NO has been implicated as the mediator of the late window of ischaemic preconditioning following single or repeated brief episodes of ischaemia [11,12]. Such preconditioning protects against both infarction and stunning and has been hypothesised to be mediated through the expression of iNOS. This hypothesis is supported by evidence showing that aminoguanidine, a relatively selective inhibitor of iNOS, inhibits late preconditioning , however, demonstration of iNOS protein expression in response to such preconditioning stimuli is as yet lacking.
The results of investigation into the late effects of ischaemia–reperfusion on vascular function are also of interest with regard to expression of eNOS and iNOS. Although it is widely recognised that reperfusion injury may lead to an acute reduction in NO mediated vascular dilatation , myocardial stunning is associated with a delayed, enhanced effect on coronary endothelial function. A single 5-min episode of coronary occlusion in the conscious dog has been shown to lead to an increase in vasodilator response to acetylcholine and bradykinin, beginning 6 h after reperfusion . This enhanced effect was associated with the increased production of NO and could be blocked by the inhibition of NO by intracoronary NG-nitro-l-arginine. The NOS isoform responsible for the increased NO production however was not investigated in this study.
Lastly, the phenomenon of repetitive myocardial stunning has recently been proposed as the underlying mechanism of myocardial hibernation , following the demonstration of normal resting myocardial blood flow in most hibernating segments . In this hypothesis repetitive episodes of stress-induced myocardial ischaemia lead to myocardial stunning from which there is insufficient time for recovery to occur prior to the next episode of ischaemia. Therefore the effect of repetitive stunning on NOS expression may also have relevance in this phenomenon.
In view of the potential relationship between NO production and both preconditioning and hibernation we designed the present study to determine the effect of repetitive myocardial stunning on myocardial NO production and on the expression of both the constitutive and inducible forms of NOS. The results show for the first time the increased expression of both isoforms of NOS in response to repetitive myocardial stunning.
2.1 Experimental preparation
This investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). Large white pigs of either sex (34–39 kg) were briefly sedated with halothane 2–4%. Anaesthesia was induced with propofol (Zeneca) (0.83–1.66 mg/kg i.v. bolus) following which the animals were intubated and ventilation adjusted to maintain PO2, PCO2 and pH within normal ranges . Body temperature was monitored by a continuously reading rectal thermometer and maintained (37–39.6°C) by heating pad. Anaesthesia was maintained by propofol infusion (8–15 mg/kg/h) and supplemented immediately prior to the operative procedure by sodium thiopentone (0.2–0.5 g total dose). Lead II of the surface ECG was continuously monitored. The right femoral artery was cannulated and a 6F catheter (Sims Portex) advanced into the thoracic aorta. A left thoracotomy was performed at the level of the 4th intercostal space and the heart supported in a pericardial cradle. The circumflex coronary artery and accompanying vein were dissected free and a hydraulic occluder (2–3 mm) (In Vivo Metric, Healdsburg, CA, USA) was implanted around the artery taking care not to damage or occlude side branches. A loose elastic snare was placed around the vein to allow temporary arrest of flow from the anterior interventricular vein. The vein was cannulated distal to the snare with a saline-flushed polyurethane intravenous cannula (I.D. 0.8 mm) and positioned to allow sampling predominantly from the circumflex territory. Left ventricular (LV) pressure was measured by a solid state pressure transducer (model P7, Konigsberg Instruments, Pasadena, CA, USA) placed in the LV via an apical incision. Two pairs of subendocardial piezoelectric sonomicrometric crystals (Triton Technologies) were placed parallel to muscle fibre orientation within the territories supplied by the circumflex and left anterior descending coronary arteries to measure regional segment shortening. Finally, two epicardial pacing wires (Ethicon, model FEP 13) were sutured to the left atrial appendage and the heart paced on demand at 120 beats/min (Coats Pacesetter, model 446) to ensure an equal minimal heart rate between animals.
2.2 Experimental measurements
All hemodynamic measurements were processed, continuously displayed and recorded via a system 6 mainframe (Triton Technologies) linked to a Ca recorder (DISS, USA). Aortic pressure was measured with a strain-gauge transducer (Ohmeda, model 5299-704). Both LV and aortic pressure gauges were calibrated with a mercury manometer. Heart rate measurement was derived from the LV pressure pulse. LV dP/dt was also derived from the LV pressure signal and used to time the measurement of segment shortening (% segment shortening=end diastolic segment length−end systolic segment length/end systolic segment length×100). Final values were averaged from ten consecutive heartbeats with the exclusion of ectopic beats.
2.3 Experimental protocol
After instrumentation all animals were allowed to stabilise for 30 min following which regional ischaemia was induced by 10, 2-min inflations of the hydraulic occluder, each interspersed by 2 min of reperfusion (n=6). Reperfusion was ascertained by recovery in regional function and ST segment change. Recordings of aortic and LV pressure, dP/dt and segment shortening were made at baseline, throughout the period of ischaemia reperfusion, at 5, 15 and 30 min of reperfusion and at half-hourly intervals thereafter up to 6 h. Paired aortic and local coronary venous blood samples were taken over 30 s into lithium heparin at baseline and at the end of the experiment. Serum samples were also collected at baseline and at the end of the experimental period from the aorta. Sham operated control animals underwent the same procedure but without circumflex occlusion (n=4).
2.4 Nitric oxide analysis
Paired arterial and local coronary vein samples for NO analysis were placed on ice and then cold centrifuged at 811 g for 10 min. Plasma aliquots were stored at −40°C. All plasma samples subsequently underwent cold ethanol deproteinization prior to NO analysis, thus leading to a 1 in 3 dilution. Plasma [NO] was measured in duplicate by acidified vanadium chloride reduction of its oxidation products, nitrite and nitrate (NOx) and ozone chemiluminesence (NO analyser 280, Sievers, USA). Pigs had been fasted for ≅12 h prior to the experiment to reduce the effect of dietary nitrate on the assay. Standard curves were constructed for each experiment using NaNO3 (3.125–100 μM). Correlation coefficients for all curves were r>0.98. Regional nitric oxide production fraction was calculated from: Local coronary vein [NO]−aortic [NO]/aortic [NO]×100. This calculation was performed to reduce the effect of any baseline drift of the analytical equipment used to measure [NOx].
2.5 Creatine kinase MB isoenzyme analysis
Creatine kinase MB fraction was blindly assayed in serum samples stored at −70°C, using a commercially available immunoconcentration assay (Hybritec, Liège, Belgium) previously validated for use in pigs (company information).
2.6 Western blotting
Myocardial samples from the circumflex territory distal to the occluder, as determined by the distribution of obtuse marginal branches, were taken from both stunned and sham operated animals (sham control) at the termination of the experiment (6 h post reperfusion). Samples were also obtained from the left anterior descending territory in stunned animals (internal control). Care was taken to ensure that tissue directly affected by implantation of the sonomicrometric crystals was not included. Samples were then snap frozen in liquid nitrogen and stored at −80°C. Protein was extracted by homogenisation of tissue (Ultra-Tarrax, Janke and Kunkel, Staufen, Germany) in lysis buffer containing: chymostatin 10 μg/ml, leupeptin 1 μg/ml, bestatin 40 μg/ml, pepstatin 1 μg/ml, TLCK 50 μg/ml, 0.1 mM DTT (dl-dithiothreitol), 50 mM Tris, pH 7.4, 1 mM EDTA and 20 mM CHAPS (3-[(3-cholamidopropyl)-dimethylammonio]-propanesulfonate). The homogenate was centrifuged at 4°C for 30 min at 26 891 g and the supernatant stored at −70°C. All protein extracts were analysed simultaneously for protein concentration (Bio-Rad DC protein assay, Ca) to ensure subsequent equal protein loading between samples. Protein extracts were separated on 7.6% SDS polyacrylamide gels (40 μg of protein per lane) and transferred onto a nitrocellulose membrane by wet electroblotting for 2 h. Blots were blocked for 1 h at room temperature with 5% non-fat dry milk in Tris-buffered saline and 0.1% Tween-20 (TBS-T) and then washed in TBS-T. Incubation with the primary antibody was performed overnight at 4°C. iNOS detection was separately performed with both a polyclonal and a monoclonal antibody. The polyclonal antiserum (1 in 1000) (code SA-200) was the kind gift of Dr. J. Pollock and was raised to a 15-amino-acid peptide based on the human form of iNOS (Biomol Research Laboratories, USA). The monoclonal antibody (1 in 5000) was a mouse anti-macrophage IgG1 (Transduction Labs. N39120). Further monoclonal antibodies were employed to identify eNOS (1 in 2000) (Transduction Labs. N30020) and CD31, an endothelial cell marker, (1 in 3000) (Dako). After washing, blots were incubated at room temperature for 1 h with secondary antibody (biotinylated goat antiserum to rabbit IgG or horse antiserum to mouse IgG, both 1 in 1000) (Vector Laboratories, Peterborough, UK) washed and incubated for a further 1 h with streptavidin conjugated horseradish peroxidase (1 in 5000) and again washed. Specific proteins were detected by enhanced chemiluminesence (Amersham) according to the manufacturers instructions. Prestained protein markers (Bio-Rad Broadrange) were used for molecular mass determinations. Cell lysate from mouse macrophages stimulated for 12 h with IFNχ (10 ng/ml and lipopolysaccharide (1 μg/ml) was used as a positive control for iNOS (Transduction Labs.). Cultured pig aortic endothelial cells were homogenised and used as a positive control for eNOS.
After scanning blots onto computer (Hewlett-Packard ScanJet 4C) individual bands were analysed for optical density using scion image (based on NIH image for the Macintosh) running on a IBM compatible computer. The area analysed for each band was kept constant for each blot analysed. Background density on the blot was subtracted from the densitometric data of each band.
Tissue blocks were taken as described for western blotting, placed overnight in 1% paraformaldehyde before washing in PBS–sucrose and snap freezing in OCT (Tissue Tek). The avidin–biotin–peroxidase complex method was used to stain sections. Endogenous peroxidase was blocked with 0.03% (v/v) hydrogen peroxide in 70% methanol for 15 min followed by washing in 10 mM phosphate-buffered saline, pH 7.1–7.4 (PBS) (three washes of 5 min each). Non-specific binding was blocked by incubation with 3% (v/v) normal goat serum for the polyclonal antibody and 3% (v/v) normal horse serum for the monoclonal antibody. Sections were blotted to remove excess serum and incubated overnight at 4°C with optimally diluted antisera to iNOS (SA-200 1 in 800) or eNOS (1 in 3000). Negative controls were performed by omission of the primary antibody for iNOS and incubation with isotype matched irrelevant antibody for eNOS. Sections were washed in PBS and then incubated for 45 min at room temperature with biotinylated goat antiserum to rabbit IgG or with horse antiserum to mouse IgG (Vector Laboratories), diluted 1:100. Following further washes in PBS, freshly prepared avidin–biotin–peroxidase complex (Vectastain, Vector Laboratories) was applied for 45 min. Peroxidase activity was revealed with diaminobenzidine and hydrogen peroxide and haematoxylin used as a counterstain. Sections were dehydrated, cleared and mounted in DPX (Merck, UK).
Antibody specificity in the case of iNOS (SA-200) was confirmed by immunostaining sections after overnight incubation with 12 nmol/ml of the peptide used to raise the antibody. Lastly, haematoxylin and eosin staining was performed on all sections for histological assessment of tissue necrosis.
2.8 Statistical analysis
Data are reported as mean±S.E.M. Hemodynamic data were analysed using two-way repeated measures ANOVA (time and group). If significant differences between groups were detected then parameter estimates were obtained from the model to compare groups at each time point. NO production, protein band optical density and total deficit of segment shortening are compared using the Mann–Whitney test for non-parametric unpaired data to compare stunned vs. sham controls and with Wilcoxon’s signed rank test to compare stunned vs. internal controls. Changes in [NOx] within groups were compared with a paired t test. A P value of <0.05 is taken as significant. Statistical analysis was performed using spss for Windows 95.
Of the thirteen pigs instrumented for this study, six were assigned to the stunned group and four to the control group. Three animals were excluded, two had nondominant circumflex vessels and failed to show substantial reductions in segment shortening during vessel occlusion and one developed ventricular fibrillation during the ischaemia–reperfusion sequence.
3.2 Systemic hemodynamics and regional
3.2.1 Myocardial function
Systemic hemodynamic effects in both stunned and control groups are summarised in Table 1. No significant group differences were seen for heart rate, mean arterial pressure, left ventricular end diastolic pressure, rate pressure product or left ventricular dP/dt max.
|Baseline||Occl 1||Rep 1||Occl 10||Rep 10||30 min||60 min||120 min||180 min||240 min||300 min||360 min|
|SS (% EDL)|
|Baseline||Occl 1||Rep 1||Occl 10||Rep 10||30 min||60 min||120 min||180 min||240 min||300 min||360 min|
|SS (% EDL)|
Data are mean±S.E.M. Data at reperfusion 1 and 10 were recorded 2 min after the end of the previous occlusion. HR, heart rate; MAP, mean arterial pressure; LVEDP, left ventricular end diastolic pressure; RPP, rate pressure product (systolic blood pressure×heart rate/1000). Stunned n=6, control n=4.
*P<0.05 vs. baseline within same group.
#P<0.05 vs. corresponding value control group.
The effects of repetitive episodes of myocardial ischaemia on regional myocardial function are summarised in Table 1 and Fig. 1. Repeated measures comparison of segment shortening between stunned (S), internal control (IC) and sham control (SC) groups showed a significant difference (P<0.031 and P<0.008, respectively). The results of individual time point comparisons are presented in Table 1. During coronary occlusion segment shortening in the ischaemic territory became negative, indicating segment lengthening due to systolic bulging of the myocardium. Immediately upon reperfusion function recovered rapidly, initially showing a small overshoot (21.5±7.1% increase during the first reperfusion falling to −12.6±9.1% after the tenth) in function in the first 1 min of each reperfusion period before falling below pre-occlusion levels immediately prior to the next occlusion. Overall segment shortening fell in response to ischaemia–reperfusion reaching a nadir 30 min after the final reperfusion. At this time function had fallen to 52.5±7.3% of baseline. It steadily recovered over the next 5.5 h reaching 92±7.9% of baseline at the termination of the experiment (Fig. 1). Internal control segment shortening (Table 1) showed a small increase in function as a compensatory response to ischaemia though otherwise function in both the internal and sham control segments remained stable throughout. The total deficit in segment shortening (an integrated measure of the magnitude and duration of postischaemic dysfunction) differed significantly from sham controls (247.3±18.4 vs. 354.8±9.4 [arbitrary units], P=0.0095).
3.3 Nitric oxide production
At baseline arterial [NOx] was 85±21 in S and 120±38 μM in SC (P=ns) and venous [NOx] was 101±24 in S and 144±44 μM in SC (P=ns). The venous–arterial difference was 15.5±4.2 and 24.2±6.7 μM in S and SC (P=ns) (Table 2) resulting in an extraction fraction of 21±5% and 22±4% respectively. Six hours into reperfusion arterial [NOx] was 56±15 in S and 77±25 μM in SC (P=ns) and venous [NOx] was 73±16 in S and 89±28 μM in SC (P=ns). The venous–arterial difference was 16.0±3.4 (P=ns cf baseline) and 11.5±3.5 μM (P<0.07 cf baseline) in S and SC resulting in an extraction fraction of 36±9 and 16±1, respectively. The change in venous–arterial [NOx] between baseline and 6 h reperfusion was found to be significantly different between the two groups (0.2±0.7 in S vs. −4.3±1.6 μM in SC; P<0.02).
3.4 Creatine kinase MB fraction
Serum assayed for CKMB at both baseline and the termination of the experiment showed no significant rise in either sham control or stunned animals (CKMB <2 U/l in all samples). Thus indicating no evidence of myocardial necrosis in either group.
3.5 Western blotting
3.5.1 Inducible nitric oxide synthase
Stunned myocardial tissue and positive control homogenate repeatedly demonstrated a discrete protein band at ≈130 000, the molecular weight of iNOS (Fig. 2, top panel) with both antibodies tested. Fainter bands were seen at the same level for both internal and sham control tissues. Optical density analysis of the protein bands revealed a 3.7-fold diference in iNOS concentration in stunned vs. sham controls (92.5±13.4 vs. 24.8±4, P=0.031) and a 2.5-fold increase compared with internal controls (92.5±13.4 vs. 36.7±2.4, P<0.01) (Fig. 2, bottom panel). A second band ≈10 000 above that corresponding to iNOS was also detected in all samples. This band appeared to vary in intensity in parallel with the bands seen at the same molecular weight as the positive control. The possibility that these bands represent an unequal fragmentation of an iNOS homodimer could not be confirmed and they were thus not included in the densitometric analysis.
3.5.2 Endothelial nitric oxide synthase and CD31
As eNOS is largely expressed in endothelial cells, homogenates used in Western blotting for eNOS were also incubated with a primary antibody to the endothelial marker CD31 confirming that similar amounts of endothelial cells had been sampled in each group (Fig. 3, bottom right panel). Incubation with eNOS (Fig. 3, top panel) in contrast showed a significant increase in expression in stunned compared with internal and sham controls (Fig. 3, bottom left panel) (103.6±7.4 vs. 62.5±7.4, P<0.03 and 75.7±0.6, P<0.01 respectively).
No evidence of cell necrosis in either stunned or control tissues was seen on haematoxylin or eosin stained sections.
Immunoreactivity to iNOS was seen to localise to arterial and venous smooth muscle cells and cardiomyocytes in stunned myocardial tissue (Fig. 4). No reactivity was seen in sham controls. Faint reactivity was seen in venous smooth muscle in two out of six of the internal controls. Specificity of the antibody staining was confirmed by successful preabsortion of the antibody (SA-200) with iNOS peptide (data not shown). No reactivity was seen in sections processed without the addition of the primary antibody. eNOS reactivity was seen in endothelial cells of both large and small arteries/arterioles in stunned and internal and sham control tissue (Fig. 5). However, no reactivity could be identified in myocytes in any group (Fig. 5). No group difference could be detected by assessment of staining intensity or distribution on light microscopy.
Our results show that repetitive brief episodes of ischaemia–reperfusion in this open chested swine model led to a temporary impairment of myocardial contractile function with no evidence of myocardial necrosis and was thus consistent with the presence of myocardial stunning . Recovery from this state of reduced contractile function largely occurred over a period of 6 h and was associated with maintained release of NO by the myocardium compared with a tendency for a fall in NO release by sham controls. This maintenance of NO production in turn appeared to be mediated by the increased expression of both eNOS, within endothelial cells, and iNOS within myocytes and vascular smooth muscle cells.
The finding that the anaesthetised pig heart has a positive transmyocardial NOx gradient at baseline is consistent with previous investigations indicating a significant production of NO by the coronary circulation of dogs . Furthermore, NOx analysis in humans has demonstrated that plasma [NOx] may show considerable subject variability  in the same way as was noted in our animals. Plasma NOx is predominantly determined by the concentration of nitrate which is in turn largely dependent on previous dietary intake and renal excretion  and it therefore seems likely that variations in these factors may explain the large interanimal variation in arterial [NOx]. The fall in arterial NOx over the period of the study is consistent with the continued excretion of nitrate and lack of intake, though the average 35% reduction in arterial NOx over this time period would suggest a slightly longer halflife than that proposed for dogs . At the end of the experiment all subjects continued to show a net production of NOx though there was a trend for this to be lower in the controls compared with baseline, a finding consistent with the proposed inhibitory effect of propofol on NOS . In contrast recovery from myocardial stunning was associated with a preservation in NOx production and comparison of the change in NOx production from baseline between the groups indicated a relative increase in the stunned group.
The relative augmentation of NO production following recovery from repetitive myocardial stunning and the upregulation of eNOS expression in endothelial cells is consistent with the findings of Kim et al. . They have previously demonstrated a delayed, enhanced response of coronary vasodilatation to the NO dependent vasodilators acetylcholine and bradykinin in conscious dogs following a single 10-min coronary occlusion. Furthermore, this occurred at a time when coronary sinus NO release was increased. Indeed, the onset of increased endothelium dependent vasodilatation occurred over a remarkably similar time frame to the increase in eNOS expression seen in our model .
Similarly, recent investigations into the mechanism of the late window of ischaemic preconditioning, triggered by repeated brief episodes of ischaemia reperfusion, have implicated an increase in NO production as the protective mediator against both myocardial infarction  and stunning  in rabbits. It seems likely from these studies that this occurs via the induction of iNOS, though an increase in iNOS expression was not shown to have occurred. Our study now supports these findings and provides evidence that repetitive stunning can lead to upregulation of iNOS protein within myocytes.
The upregulation of iNOS at a time when myocardial function is recovering initially might seem contradictory given that iNOS is generally perceived as a damaging enzyme . Previous studies have described the expression of iNOS in a variety of disorders associated with myocardial dysfunction including septic shock, dilated cardiomyopathy, myocarditis and transplant rejection . There is also substantial evidence from in vitro studies that exogenous and cytokine induced NO can be negatively inotropic . However, it seems that a dichotomy of action for iNOS/NO exists since NO manipulation following episodes of ischaemia–reperfusion has been shown to have both detrimental and beneficial effects . Furthermore, recent studies have shown that induction of iNOS, by the prior administration of monophosphoryl lipid A, mediated myocardial protection . Indeed, iNOS has now been suggested to have beneficial actions in a variety of clinical conditions including protection from transplant atherosclerosis , obliterative bronchiolitis  and accommodation of transplanted cells , possibly by reducing apoptosis . This double edged effect of iNOS/NO is likely to be due to variations in the amount of NO produced and the milieu into which it is released, factors that are determined by the individual model under investigation. For example, NO is known to react with both the superoxide anion to produce the potentially damaging oxidant, peroxynitrite  and to interact with other pro-inflammatory, inducible enzyme systems  that may be present following some forms of injury.
We hypothesise that the expression of NOS in our model may be a protective adaptation against further episodes of ischaemia and may be mediated via the known beneficial actions of NO which can help to protect against reperfusion injury in both the long and short term. These include enhanced coronary vasodilation , reduced calcium influx in to myocytes , reduced myocardial oxygen consumption , opening of KATP channels , inhibition of white cell  and platelet adhesion  to endothelial cells and the interaction with the heat shock proteins HSP70  and HSP90 .
The mechanisms of NOS induction in this study remain open to speculation. iNOS is known to have multiple consensus sites for transcription factors including nuclear factor-κB (NF-κB) response elements in its promoter/enhancer region . NF-κB itself has been demonstrated to be activated by reactive oxygen intermediates  that are produced in myocardial stunning . In addition NF-κB has been shown to be induced in rat myocardium following a 15-min coronary occlusion and that this in turn lead to the expression of iNOS . Cytokines also upregulate iNOS , including interleukin-6 which has been shown to be produced in a cardiac bypass model of stunning . The transcriptional and post-transcriptional control of eNOS expression is less well described. The human eNOS promotor is known to bind a variety of transcription factors including AP-1 and 2, retinoblastoma control element, shear stress response element, NF-1, steroid regulatory element, Sp-1 and cAMP response element . eNOS mRNA abundance can also be regulated in response to oxygen concentration in cell culture experiments .
The strong reactivity to eNOS in endothelial cells from control subjects made the unregulation of this protein difficult to detect by immunocytochemistry. Therefore, this technique was reserved as a qualitative assay for localising expression while quantification was assesed by Western blotting. The lack of eNOS expression in cardiomyocytes in this study is contrary to previous studies in both humans  and rodents . This may be either a true species difference or a result of differences in antibody detection. However, this result is not completely inconsistent with reported data. Kim et al.  have argued that the increase in NO production in response to brief ischaemia–reperfusion in dogs is likely to originate from coronary vessels rather than cardiomyocytes as myocardial contractile responses to acetylcholine and bradykinin remained intact.
In conclusion, repetitive myocardial stunning is associated with the preservation of NO production and the upregulated expression of iNOS and eNOS. This finding has implications for coronary vascular autoregulation following stunning and the late window of ischaemic preconditioning. It also raises the question of the role of NOS in myocardial hibernation, quite possibly resulting from repetitive episodes of stunning. We speculate that in contrast to the usually accepted damaging role of iNOS it may be acting in this model in a protective fashion.
The authors thank Miss M. Nohadani, Miss R. Strong and Mrs. J. Beckett for their invaluable technical support throughout the completion of this project. Dr. C.S.R. Baker is supported by a British Heart Foundation Junior Fellowship grant.
Morishima N. The effect of modulation of the l-arginine nitric oxide pathway on myocardial stunning following repetitive coronary occlusion in dogs. Hiroshima J Med Sci 1995;119–127.
- cardiac myocyte
- nitric oxide
- myocardial stunning
- hibernating myocardium
- ischemic preconditioning
- western blotting
- endothelial cells
- reperfusion therapy
- physiologic reperfusion
- muscle, smooth, vascular
- nitric oxide synthase
- family suidae