reperfusion injuryis a complex set of events that paradoxically increases tissue injury when oxygenated blood is supplied to ischemic tissues (1,2, 6, 26). Myocardial infarction (MI) develops largely as a result of damage to ischemic and reperfused cardiac myocytes and endothelial cells. Current therapy for MI can achieve rapid and efficient reperfusion of the occluded or infarct-related coronary artery using thrombolytic treatment or direct coronary artery angioplasty. However, there is no generally accepted treatment or therapeutic intervention for the attendant reperfusion injury, which, if minimized or halted, would likely result in improved ventricular function, the major determinant of long-term survival in humans (6, 26).

Several lines of evidence suggest that a major pathway in reperfusion injury is mediated by activation of nuclear factor-κB (NF-κB). NF-κB is a dimer of members of the Rel family of proteins. The classic form of NF-κB is composed of p50 (NF-κB1) and p65 (RelA) (4, 16). In most cells, NF-κB is inactive and sequestered in the cytoplasm due to binding by an inhibitor protein (IκB). During activation of NF-κB, IκB is phosphorylated by the IκB kinase complex (37), targeting it for ubiquitination and degradation by the proteasome. The unbound NF-κB is transported to the nucleus where it binds to specific promotor sequences and regulates gene expression, including those involved in the inflammation and likely reperfusion injury (3, 16, 32). Importantly, proteasome inhibitors have been shown to inhibit IκB degradation and block NF-κB nuclear translocation (2, 15).

In this study, we addressed the hypothesis that inhibiting proteasome activity would reduce reperfusion injury in the in situ heart by a mechanism that involved inhibition of activation of inducible NF-κB. The proteasome inhibitor PS-519 (2) was shown to significantly reduce proteasome activity in circulating leukocytes. Also, with the use of a 1-h ischemia protocol followed by a 3-h reperfusion protocol, PS-519 was shown to block activation of myocardial NF-κB by electromobility shift assay (EMSA); to reduce myocardial serum creatine kinase (CK), CK muscle-brain fraction (CK-MB), and troponin I (TnI) release; to preserve regional myocardial function measured by segmental shortening; to reduce the area of MI; and to exhibit no acute toxicity. Our data support the hypothesis that proteasome inhibition reduces reperfusion injury in an experimental model of MI by a mechanism that involves inhibition of activation of NF-κB. These data provide proof of the concept for considering the therapeutic potential of proteasome inhibitors in humans following reperfusion of an infarct-related coronary artery.

MATERIALS AND METHODS

Experimental Animals

Male and female pigs (Sus scrofa) between 6 and 8 mo of age were used. All pigs were bred and raised at the Francis Owen Blood Research Laboratory at the University of North Carolina, Chapel Hill. All animals were treated according to the standards set in theGuide for the Care and Use of Laboratory Animals [National Institutes of Health (NIH) Publication No. 85-23]. All procedures were in accordance with institutional guidelines under approved protocols at the University of North Carolina.

Myocardial Ischemia, Infarction, Reperfusion, and PS-519 Treatment Protocols in Pigs

The pigs were anesthetized with ketamine (10 mg/kg im), intubated, and maintained with isofluorane (2–4% in air supplemented with 1 l/min oxygen). A median sternotomy was performed, and the pigs were placed on positive pressure ventilation. Arterial blood gases were monitored, and the ventilator was adjusted to maintain pH 7.35–7.45, Pco₂ 35–45 mmHg, and Po₂ >100 mmHg.

The left anterior descending coronary artery (LAD) was dissected free of surrounding tissues at two sites, and an encircling suture was placed proximally and a Doppler flow velocity crystal distally. Measurements of coronary blood flow (CBF) velocity were performed in vivo using a custom-made Doppler system (Dr. Craig J. Hartley, Baylor College of Medicine, Houston, TX). CBF was also confirmed by injecting contrast media in the left main coronary artery under fluoroscopic guidance. A Simmons catheter (SIMI/II, Cook; Bloomington, IN) was placed in the coronary sinus for cardiac venous sampling (12). An eight-channel strip-chart oscillograph (Linearcorder WR3310, Western Graphtec; Irvine, CA) was used to monitor and record CBF velocity, blood pressure via a Millar catheter, and heart rate by electrocardiogram. Body temperature was monitored by thermometers via both rectal and myocardial temperature probes and maintained at temperatures of 36 to 38°C with warming blankets (TP-500, Gaymar; Orchard Park, NY).

PS-519 (1 mg/kg, Millennium Pharmaceuticals; Cambridge, MA) was dissolved in an excipient consisting of an equal volume of propylene glycol (1,2-propanediol, P-1009, Sigma Chemical; St. Louis, MO) and sterile saline (0.9% NaCl, 2F7124, Baxter Healthcare; Deerfield, IL). PS-519 dissolved in excipient (PS-519-treated group) or excipient alone (control group) was administered over ∼1–3 min via a coronary angiography catheter in a supraaortic valvular region immediately after baseline samples were taken and just before LAD occlusion.

The LAD was then occluded for 1 h to produce an anterior MI and subsequently reopened and allowed to reperfuse for 3 h. Occlusion and reperfusion were verified by changes in Doppler flow velocity signals and by angiographic visualization. Myocardial biopsies of ischemic and nonischemic myocardium were obtained by using a myocardial bioptome (T-Rex, Boston Scientific SciMed; Maple Grove, MN). Blood samples for serum CK, CK-MB, TnI, and white blood cell (WBC) 20S proteasome activity analysis were taken from the coronary sinus (see Leukocyte Isolation and 20S Proteasome Activity Analysis and Measurement of Serum K, CK-MB, and TnI Levels). Biopsy and blood samples were taken at time points indicated on Table 1. The experiment was concluded with administration of an overdose of pentobarbital sodium (6 grains/4.5 kg iv) The hearts were then removed and processed for quantification of the area of MI as detailed below.

Myocardial Segment Shortening Analysis

Myocardial segment shortening was analyzed in ischemic and nonischemic myocardium in four control and four PS-519-treated pigs. Subepicardial, multidirectional ultrasonic 0.7-mm dimension crystals (Sonometrics; Ontario, Canada) were implanted 10 mm apart in the minor axis plane to subtend the myocardial region supplied by the LAD and to measure longitudinal segment length shortening with high-frequency sampling (>100 Hz) (8). Segmental shortening in the ischemic and nonischemic regions was compared, at the indicated time points (Table 1), using Sonosoft sonomicrometer analysis software (version 3.1.0, Sonometrics).

Leukocyte Isolation and 20S Proteasome Activity Analysis

Leukocyte isolation.

WBCs were isolated from heparinized venous blood using Nycoprep 1.077 (Nycomed; Oslo, Norway) according to the manufacturer's recommendations with modifications: the leukocytes were isolated by centrifugation (800 g for 30 min at room temperature), washed in 1× phosphate-buffered saline, recentrifuged (400g for 10 min), and stored at −80°C.

Fluorometric 20S proteasome activity analysis.

The prepared WBCs were lysed with EDTA (5 mmol/l, pH 8.0) by incubation for 1 h and then centrifuged (6,600 g, 10 min, 4°C). Lysed cells were then resuspended in buffer (1 mol/l dithiothreitol and 1 mol/l HEPES), centrifuged (2,200 g, 15 min, 4°C), and then mixed with substrate buffer (20 mmol/l HEPES, 0.5 mmol/l EDTA, and 0.35% SDS) and 60 μmol/l Ys substrate (succinyl-Leu-Leu-Val-Tyr-amido-4-methyl-coumarin, Bachem; Bubendorf, Switzerland) at pH 8.0 and 37°C. The rate of substrate cleavage per 20S proteasome activity was determined (expressed as pmol AMC · s⁻¹ · mg protein⁻¹, where AMC is 7-amino-4-methylcoumarin).

Measurement of Serum CK, CK-MB, and TnI Levels

Serum for CK, CK-MB, and TnI analysis was prepared from blood by centrifugation separation in SST Gel and Clot Activator Vacutainers (Becton Dickinson; Franklin Lakes, NJ). Total CK was measured by an enzymatic method using a Vitros 950 chemical analyzer (Johnson & Johnson Clinical Diagnostics; Rochester, NY) according to the manufacturer's recommendations. CK-MB and TnI were measured by a microparticle enzyme immunoassay method using an Axsym chemical analyzer (Abbot; Abbot Park, IL) according to the manufacturer's recommendations.

Nuclear Extractions and EMSA for NF-κB

Nuclear proteins were extracted, and a 5-μg aliquot was incubated with either a γ-³²P-labeled double-stranded DNA probe (∼2.5 ng) containing a κB site from the class I major histocompatibility complex promoter or a constitutive nuclear protein (Oct-1, sc-232, Santa Cruz Biotechnology; Santa Cruz, CA) to control for protein loading and separated by EMSA as described (10,18)

Quantification of Area at Risk and Myocardial Infarction

Area at risk.

After the 3 h of reperfusion, the LAD was reoccluded, and 20 ml of a freshly made solution of 20% thioflavin-S (T-1892, Sigma-Aldrich; St. Louis, MO) in sterile saline (0.9% NaCl, 2F7124, Baxter Healthcare) were injected into the aortic valve sinuses adjacent to the ostia of the coronary arteries with fluoroscopic guidance immediately after administration of the pentobarbitol overdose and just before cessation of the heartbeat. A cardiectomy was performed, and the heart was cut from apex to base into seven serial sections of ∼10 mm thickness. The myocardial sections were then placed on a light box mounted with a digital camera (Q-mini, Konica Photo Imaging; Englewood Cliffs, NJ), fitted with a polarizer, and illuminated by an ultraviolet lamp (UVP systems; Upland, CA). Myocardial sections were photographed in plane alongside a metric ruler for subsequent calibration of the computer-assisted measurements of the surface area. Thioflavin S is fluorescent under ultraviolet light in perfused tissue. Myocardium that did fluoresce was defined as thioflavin positive and therefore as nonischemic. Myocardium that did not fluoresce was defined as thioflavin negative and therefore as ischemic (21,22). The area of the ischemic tissue was defined as the ischemic area at risk (AAR).

Myocardial infarction.

After fluorescent thioflavin imaging, the sections were incubated at room temperature in 500 ml of a freshly made solution of 1% triphenyltetrazolium chloride (TTC, T-8877, Sigma-Aldrich) for 15 min. The sections were then photographed with the same camera system but under incandescent illumination. TTC forms a red precipitate in viable tissue, and myocardium that stained red was defined as TTC positive and therefore as viable. Myocardium that did not stain red appeared pale yellow under incandescent light and was defined as TTC negative and therefore as infarcted myocardium (35). The area of infarcted tissue was defined as the area of MI. On completion of imaging, the sections were fixed in 10% buffered formalin.

Quantification of MI and AAR With Computer-Assisted Planimetry

Images were downloaded from the digital camera into Adobe PhotoShop 5.5 (Adobe Systems; Seattle, WA), and the area of MI and AAR were determined using NIH Image 1.61 (Bethesda, MD). Each digital image was individually calibrated for area before measurements were taken. The following three surface areas were measured: TTC-positive (viable) myocardium, thioflavin S-positive (perfused) myocardium, and total myocardium. The following formulas were then applied to the measurements: 1) area of MI = total myocardial section area − TTC-positive myocardium; 2) percent area of MI = area of MI ÷ total myocardial section area × 100;3) AAR = total myocardial section area − thioflavin-positive myocardium; 4) percent AAR = AAR ÷ total myocardial section area × 100; 5) nonischemic myocardium = total myocardial section area − AAR; and 6) percent nonischemic myocardium = nonischemic myocardium ÷ total myocardial section area × 100.

Detection of Apoptosis in Reperfused Sections of Entire Myocardium

Cross sections of the entire heart from control and PS-519-treated pigs were prepared by immersion fixation for 30 min in 10% buffered formalin followed by a change to fresh fixative and immersion for an additional 24 h. The fixed tissues were paraffin embedded, and 5-μm sections were mounted on ProbeOn Plus slides (15-188-52, Fisher Scientific; Pittsburgh, PA). Myocardial cells exhibiting apoptosis were detected by the terminal deoxynucleotidyl transferase-mediated nick-end labeling (TUNEL) immunohistochemistry method using a Trevigen Apoptotic Cell system (TACS) Blue Label reaction (CardioTACS apoptosis kit, 4827-30-K, Trevigen; Gaithersburg, MD). This system was used according to the manufacturer's recommendations, with modifications made to the quenching and incubation times that optimized detection of positive cells using controls. Positive controls were prepared by digestion of myocardial sections with TACS-Nuclease in appropriate buffer for 60 min at 37°C. Negative controls were by incubation with buffer only.

Statistical Methods

The data for proteasome activity, coronary sinus CK, area of MI, AAR, and area of nonischemic myocardium are described as means ± SD. The exact Wilcoxon test was used for comparisons between PS-519-treated and control groups. P values <0.05 were considered statistically significant.

RESULTS

Myocardial Reperfusion Injury in Pigs

Twenty-eight pigs were entered into these experiments. Control animals (n = 13) were given excipient alone, whereas PS-519-treated animals (n = 15) were given excipient with a single dose of PS-519 as described in materials and methods. Six pigs, two control and four PS-519 treated, did not complete the protocol because of irreversible ventricular fibrillation during the ischemia and reperfusion protocol and were excluded from analysis. The total number of pigs is too small to determine whether these arrthymias were due to drug toxicity or the large percentage of the left ventricular being ischemic (see Table5). Two PS-519-treated pigs did not complete the protocol because of failure to reperfuse completely and were also excluded. In summary, 11 control and 9 PS-519-treated pigs completed the protocol. There was no significant difference in heart rate and blood pressure between the two groups (data not shown).

PS-519 Inhibits 20S Proteasome Activity

PS-519 significantly reduced the proteasome activity in circulating WBCs from 16.67 to 1.11 pmol AMC · s⁻¹ · mg protein⁻¹ (Table 2). The control group had no significant change in WBC proteasome activity with the administration of excipient (Table 2). Pre-PS-519-treated proteasome activity observed in all animals was comparable (Table 2). These results demonstrate that PS-519 is very effective at inhibiting proteasome activity in pigs as previously reported in rats (9). Inhibition of 20S proteasome activity in WBCs closely correlated with the inhibition found in other tissues including the myocardium (data not shown).

An additional pig was given PS-519 at 1.0 mg/kg intravenously and then monitored for several weeks without MI. Proteasome activity decreased 93% (11.5 to 0.8 pmol AMC · s⁻¹ · mg protein⁻¹) within 30 min of administration and returned to 50% of baseline (5.7 pmol AMC · s⁻¹ · mg protein⁻¹) at 24 h. Importantly, no clinical toxicity was detected during the recovery period.

PS-519 Preserves Regional Myocardial Function as Measured by Segmental Shortening

Ultrasonic sonomicrometer measurements of longitudinal myocardial segmental shortening were taken in four control and four PS-519-treated pigs to determine whether PS-519 preserved regional myocardial function. Baseline-adjusted measurements were calculated by taking the absolute distance of segmental shortening at baseline, defining that as 100%, and then reporting the relative change in the distance of segmental shortening at later time points. In the ischemic region, baseline-adjusted segmental shortening was significantly greater in the PS-519-treated pigs compared with the control pigs at both 1 h of ischemia (102 ± 22% vs. 47 ± 17%) and at 3 h of reperfusion (63 ± 23% vs. 32 ± 13%, P < 0.05) (Table3). In the nonischemic region, baseline-adjusted segmental shortening was not significantly different between control and PS-519-treated pigs during the experiment (data not shown).

NF-κB Is Activated by Reperfusion and Is Inhibited by PS-519

Nuclear extracts from the myocardium were prepared from five control and five PS-519-treated animals and were analyzed for NF-κB binding activity by EMSA to determine whether PS-519 was effective at inhibiting NF-κB activation and to determine whether ischemia and/or reperfusion induced activation of NF-κB in vivo. Activated NF-κB was not detected during ischemia but was consistently detected in the myocardium of the control animals following 1 h of reperfusion (Fig. 1, A vs.B). Importantly, activated NF-κB was not detected after ischemia and reperfusion in the myocardium of the PS-519-treated animals. Differences between control and PS-519-treated groups cannot be accounted for by differences in loading of nuclear extracts as evidenced by EMSA performed using a constitutive octamer probe to detect Oct-1 DNA binding activity (Fig. 1, C and D), although it is noted that there is some variability of binding of Oct-1 in these experiments.

PS-519 Blocks Increase of Serum CK, CK-MB, and TnI Levels

CK and TnI release from the damaged myocardium are clinically accepted measures of myocardial cell death (11). Baseline-adjusted CK levels observed in the PS-519-treated animals were 17% less relative to the control animals after 1 h of ischemia. Baseline-adjusted CK, CK-MB, and TnI levels were ∼47%, 67%, and 76% less, respectively, at 3 h of reperfusion (Table 4). Baseline-adjusted values were calculated by subtracting each pig's initial serum value from all subsequent values to express net change in levels that occurred during the experiment. The difference in the two groups is not significant because of the wide variation in values, which is typical of these experiments. However, the marked reduction in mean CK, CK-MB, and TnI release in PS-519-treated animals strongly suggests that the difference was a direct effect of administration of this proteasome inhibitor.

PS-519 Reduces Area of MI and AAR

Consistent with the ability of PS-519 to block NF-κB activation and to reduce serum CK levels, there were significant reductions in the area of MI and the area of MI relative to the AAR for the PS-519-treated animals (89.7% and 90.4% reductions,P < 0.05 and 0.03, respectively, Table5 and Fig.2). The AAR for the heart sections distal to the LAD occlusion were not significantly different between the PS-519-treated and control animals (55.9% and 44.8%,P = not significant, respectively, Table 5 and Fig. 2). The area of MI and AAR were measured in 9 of 9 PS-519-treated and 9 of 11 control animals because two sets of control group images were lost during processing. Importantly, all 9 control pigs had a TTC-positive area consistent with an MI, whereas only 1 of 9 treated pigs had a detectable TTC-positive area. These data indicate that a single dose of the proteasome inhibitor PS-519 can reduce myocardial injury during 3 h of reperfusion.

Apoptosis Was Not Detected in Control or PS-519-Treated Pig Hearts

The number of cells undergoing apoptosis as judged by TUNEL staining was <0.01% in both the infarct and noninfarct zones, and the difference between PS-519-treated and control tissues was not significant. Positive controls exhibited >95% of cells having positive TUNEL staining, and negative controls exhibited <0.01% (data not shown).

DISCUSSION

In this MI and reperfusion model, treatment with PS-519 significantly reduced 20S proteasome activity in circulating leukocytes, markedly reduced release of CK, CK-MB, and TnI from the myocardium, inhibited activation of inducible NF-κB in ischemic and reperfused myocardium, preserved regional myocardial function measured by segmental shortening, and reduced the area of MI. Taken together, these data demonstrate that proteasome inhibition by treatment with PS-519 significantly reduces myocardial reperfusion injury in an open-chest porcine model by a mechanism that apparently includes inhibition of activation of NF-κB.

Prior studies have shown that inhibition of inducible NF-κB activation affects myocardial reperfusion injury. In one study, double-stranded DNA oligonucleotides were transfected into rat myocardium in vivo using the hemagglutinating virus of Japan (HVJ) liposome (27). Both a reduction in activation of inducible NF-κB and reduced myocardial reperfusion injury were seen. Compared with this method, our ability to easily solubilize and administer PS-519 and monitor its effect on proteasome activity in circulating WBC make it a potentially more attractive pharmacological agent than one that requires transfection reagents. Another study used PS-519 in an ex vivo Langendorff rat heart preparation perfused with plasma and WBCs (9). Myocardial reperfusion injury mediated by WBCs was markedly reduced. In a mouse model of acute pancreatitis and myocardial infarction, proteasome inhibition resulted in a smaller infarct compared with controls (15). Our study utilized a different and larger species in which the heart was perfused with whole blood under systemic pressure and showed inhibition of NF-κB in the reperfused myocardium by EMSA; reduction of myocardial CK, CK-MB, and TnI release; preservation of regional myocardial function measured by segmental shortening; and reduction in the area of MI. Other prior studies using pig models have investigated MI following 45–90 min of ischemia and 2–24 h of reperfusion. Our study used 1 h of ischemia followed by 3 h of reperfusion, and the AAR and AAR infarcted observed in our control pigs fall within the range of these previously published results (14, 19, 20, 23, 29,30, 33). Our model is relevant to human disease where myocardial damage is sustained within minutes to hours following heart attack. Considered together, our data then confirm and extend those previous studies with a unique therapeutic agent that can be readily produced in sufficient quantities to administer to humans (31).

The mechanisms of cell death that occur within the initial hours of an MI and that may depend on NF-κB are not completely understood (6, 26). Activation of NF-κB has been shown to regulate programmed cell death or apoptosis and under many experimental conditions NF-κB functions to block apoptosis (34). The degree of apoptosis detected by TUNEL staining in our study was not significantly different between untreated and treated hearts but was quite low (<0.01%). TUNEL staining alone did not allow us to determine the type of cell undergoing apoptosis, myocardial, or other. Apoptosis is felt to occur over several hours to days, and the degree to which it contributes to myocardial damage is controversial (1). Studies of longer duration may be required to detect a greater degree of apoptosis in control animals (5, 13, 38). Within the limits of our study, it appears that apoptosis is not a major cardioprotective target of the proteasome inhibitor PS-519. Moreover, proteasome inhibitors are not specific for inhibiting NF-κB alone, and our data leave open the possibility introduced in previously published studies that other mediators besides NF-κB may be involved in the reduction of reperfusion injury observed in our study (17,24, 25). Other possible molecular networks have been recently reviewed (28).

Another role for NF-κB in promoting reperfusion injury is its capacity to upregulate transcription of genes encoding proteins involved in the recruitment of inflammatory cells. NF-κB is known to regulate genes such as interleukin-8 and ICAM-1, which have been demonstrated to be involved in reperfusion injury (7, 36). It is not known whether infiltration of inflammatory cells occurred in the 3-h reperfusion period of our experiment and whether PS-519 was effective at blocking this mechanism. Clearly, additional studies are needed to address these broader issues and others, including determining the effect of PS-519 treatment beginning after the onset of ischemia and assessing the impact of prior MI and atherosclerotic coronary disease.

Taken together, our data and those of others support the hypothesis that proteasome inhibition reduces reperfusion injury in multiple organs, including the myocardium and the brain by a mechanism that includes reduction of inducible NF-κB activation. Acute treatment with PS-519 is well tolerated in our large animal model, and its efficacy at inhibiting proteasome function is easy to monitor in circulating leukocytes. In a recent phase I trial in normal human volunteers (31), PS-519 was well tolerated in 1- or 3-day dosing regimens, inhibited 20S proteasome activity in circulating WBCs in a dose-dependant fashion, and was shown to have an acceptable pharmacological toxicity profile. The use of either this or other proteasome inhibitors or NF-κB inhibitors may be beneficial for treating patients at risk for reperfusion injury during myocardial infarction.

FOOTNOTES

• This work was supported by Grants 1-P20-DE-123474 and 1-P-60-DE-13079.

• Address for reprint requests and other correspondence: T. C. Nichols, Depts. of Medicine and Pathology, Francis Owen Blood Research Laboratory, 350 South Old Fayetteville Rd., Univ. of North Carolina, Chapel Hill, NC 27516 (E-mail:[email protected]unc.edu).

• The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

• First published November 7, 2002;10.1152/ajpheart.00851.2002