1264

Cardiac Stress Protein Elevation 24 Hours AfterBrief Ischemia or Heat Stress Is AssociatedWith Resistance to Myocardial Infarction

Michael S. Marber, MRCP; David S. Latchman, PhD;J. Malcolm Walker, MD, FRCP; Derek M. Yellon, PhD, FACC

Background. To test the hypothesis that the heat shock response is associated with myocardial salvage,the heat stress protein (HSP) content of cardiac tissue was increased by either ischemic or thermal stress.Methods and Results. Rabbits were divided into four groups. Ischemic pretreatment (n= 15) comprised

four 5-minute episodes of coronary ligation separated by 10 minutes of reperfusion. The correspondingcontrol group (n=21) underwent surgical preparation without coronary ligation. Thermal pretreatment(n=16) involved whole-body temperature elevation to 420C for 15 minutes; corresponding controls (n= 15)were treated with anesthetic alone. Twenty-four hours later, hearts were removed for HSP estimation orinfarct size assessment after a 30-minute coronary ligation. Myocardial HSP72 content assessed byWestern blotting was elevated by both ischemic and thermal pretreatments (2.5±0.2 units, n=4, and2.8±0.3 units, n=4, mean±SEM; P=NS, respectively) compared with the corresponding control groups(1.0±0.3, n=4, P<.01 and 03±0.1, n=4, Ps.01, respectively). HSP60 was preferentially elevated byischemic pretreatment. After a 30-minute coronary occlusion and 120 minutes of reperfusion, ischemicand thermal pretreatments limited infarct size as a percentage of the volume at risk by 28.8±5.2% vs52.0±5.2%, P<.01 and 32.8±3.8% vs 56.9±6.5%, Ps.01, respectively.

Conclusions. Myocardial stress protein induced by either sublethal thermal or ischemic injury isassociated with myocardial salvage. Our findings suggest that stress protein elevation, rather than thenonspecific effects of thermal or ischemic stress, may be responsible for the myocardial protection seen inthis model. Our observations may have important implications regarding myocardial adaptation to briefperiods of ischemia. (Circulation. 1993;88:1264-1272.)KEY WORDs * proteins * ischemia * myocardium * reperfusion

W hen any living cell is exposed to a sublethalelevation of environmental temperature, aseries of adaptive changes occur that serve

to protect that cell from subsequent increases in tem-perature.' A group of proteins known as the heat shockor heat stress proteins (HSPs) are the major proteinssynthesized during such stress and play a pivotal role inproviding this protection.2'3 A similar increase in stressprotein synthesis is seen after a variety of nonthermalstresses; for example, in myocardial tissue, stress pro-tein synthesis is increased by ischemic and mechanicalinjury.4-6 It seems likely, therefore, that stress proteinsare involved in recovery from nonthermal as well asthermal injury. This conclusion is supported by the factthat stress proteins raised by one form of sublethalinjury seem capable of protecting against a subsequentbut different injury, a phenomenon known as cross-

Received October 8, 1992; revision accepted April 30, 1993.From the Hatter Institute for Cardiovascular Studies (M.S.M.,

J.M.W., D.M.Y.), Division of Cardiology, University CollegeLondon Medical School, and the Division of Molecular Pathology(D.S.L.), University College London Medical School.Correspondence to Professor Derek Yellon, Director and Head

of Division, the Hatter Institute for Cardiovascular Studies, Divi-sion of Cardiology, University College Hospital, Gower Street,London WC1E 6AU, England.

tolerance.7 In keeping with this phenomenon, a numberof investigators have demonstrated that whole-bodyhyperthermia increases myocardial stress protein con-tent and renders the isolated blood- or buffer-perfusedheart resistant to subsequent ischemia and reperfu-sion.8-10 However, our attempts to use whole-body heatstress to limit infarction after a 45-minute coronaryligation were successful in the isolated blood-perfusedheart' but unsuccessful in vivo,12 leading to the specu-lation that whole-body heat stress may have deleteriousaspects that negate any beneficial effects secondary tomyocardial stress protein elevation." In contrast, in thein situ rat heart, Donnelly and coworkers13 have dem-onstrated that whole-body heat stress can reduce infarctsize after a 35-minute but not a 45-minute coronaryocclusion.The purpose of this study was to examine in more

detail the relation between myocardial stress proteincontent and the subsequent resistance of the in siturabbit heart to ischemia. An established protocol4 ofrepetitive short coronary artery occlusions was used toselectively elevate myocardial stress protein content,thereby avoiding whole-body heat stress and its possibledeleterious consequences. More importantly, brief cor-onary occlusions, unlike thermal stress, are a physiolog-ically more relevant stimulus for stress protein induc-

by guest on Novem

ber 13, 2017http://circ.ahajournals.org/

Dow

nloaded from

Marber et al Stress Protein and Myocardial Infarction 1265

Groupl IschemiaI e*a10-10*10* 24 mllHours _ 120 minutes

Group2 Sham Ischemia150 rnins chest ope 14l ors _~ 120 minutes 1

Group3 Heat Stress| Heat Stres 42 for 15'24 Hourse 120 minutes l

Group4 Sham Heat Stress fAnaesthetic oni 24 Hours 120 minutes l

* = Coronary Ligation

FIG 1. Summary of experimental protocols. Group 1, isch-emic pretreatment, was subjected to four 5-minute coronary

ligations separated by 10-minute periods of reperfusion. Thecorresponding sham ischemia group (group 2) underwentidentical surgicalpreparation but without coronary ligation. Inthe heat stress group (group 3), after anesthesia, rectaltemperature was raised to 420C for 15 minutes by wrappingrabbits in an electric warming blanket. Rabbits then recoveredat room temperature. The corresponding sham heat stressgroup (group 4) received anesthetic agent without warming.

All groups were allowed to recover for 24 hours before eithermyocardial stress protein estimation (n=4 for each group) or

a 30-minute coronary ligation followed by reperfusion andinfarct size assessment (n=10 for each group).

tion, having been likened to the clinical situation duringunstable angina.4

MethodsExperimental Protocols and Exclusions

For the purposes of this study, 67 New Zealand Whiterabbits (2.0 to 3.0 kg) were divided into four experimen-tal groups. Within each group, the end point was eitherstress protein estimation or infarct size assessment (seeprotocol, Fig 1). Experiments were performed sequen-tially between the groups, except when an experimentalexclusion necessitated repetition. Hemodynamic, tem-perature, and blood gas data from experiments thatwere excluded were not used for subsequent analysis.Group 1, ischemic pretreatment, consisted of 15

rabbits. There was 1 exclusion because of respiratoryobstruction and death 2 hours after repetitive ischemia.Four rabbits were used for stress protein estimation and10 rabbits for infarct size assessment.Group 2, sham ischemic pretreatment, consisted of 21

rabbits with 7 exclusions. In 3 rabbits, coronary ligationwas unsuccessful (no risk area evident after micro-sphere injection), 2 rabbits failed to reperfuse ade-quately (as judged by a large area of infarction withoutintramyocardial hemorrhage), and 2 animals died dur-ing the 30-minute coronary ligation secondary to hypo-tension and progressive acidosis. Four rabbits were usedfor stress protein estimation and 10 for infarct sizeassessment.

Group 3, heat stress, consisted of 16 rabbits, with 2exclusions. One rabbit died during the 30-minute coro-nary ligation, and 1 experiment had to be excludedbecause of inadequate staining of viable myocardiumwith triphenyltetrazolium. Ten rabbits were used forinfarct size assessment and 4 rabbits for stress proteinestimation.Group 4, sham heat stress, consisted of 15 rabbits.

There was 1 exclusion because of inadequate tissuestaining. Ten rabbits were used for infarct size assess-ment and 4 rabbits for stress protein estimation.

Ischemic and Sham Ischemic PretreatmentsRabbits were anesthetized with intramuscular fenta-

nyl 100 ug/kg and fluanisone 3 mg/kg (Janssen Phar-maceuticals, Oxford, UK), followed by intraperitonealdiazepam 2 mg/kg. Anesthesia was maintained by fen-tanyl and fluanisone administered every 30 minutes.Once the rabbits were adequately anesthetized, theywere orally intubated, and limb lead ECG electrodes(Medicotest, Rugmarken, Denmark) were attached.The rabbits were mechanically ventilated with 100%oxygen at a tidal volume of 5 mL/kg delivered at a rateof 1 Hz. A marginal ear vein was cannulated to admin-ister fluids and drugs, and 30 mg of intramuscularamoxycillin was given prophylactically.The heart was exposed through a median sternotomy,

and a coronary artery (usually an anterolateral branchof the circumflex) was underrun with 3/0 silk suture ona tapered needle. The free ends of the suture werepassed through a soft vinyl tube so as to form a snare toocclude the artery. Rabbits then received 1000 units ofheparin before the first 5-minute coronary artery liga-tion. Successful ligation was confirmed by myocardialcyanosis with bulging and changes in the amplified ECGsignal (ECG amplifier and series 3000 recorder, GouldInc, Cleveland, Ohio). Reperfusion was confirmed by amyocardial blush and resolution of ECG changes. Afterfour 5-minute coronary ligations separated by 10 min-utes of reperfusion, the vinyl occluder was removed, andthe coronary suture was left in situ while the sternotomywas closed by suturing muscle and then skin layers.Animals were given 10 mL/kg of 0.9% saline intrave-nously, allowed to breathe spontaneously, and eventu-ally extubated.The surgical preparation for the sham ischemic pre-

treatment group was identical to that described above.However, although the pericardium was opened, acoronary artery was neither underrun nor occluded,thus avoiding mechanical manipulation of the myocar-dium, which may in itself act as a trigger for stressprotein induction.6

Heat Stress and Sham Heat Stress PretreatmentsRabbits were anesthetized with pentobarbitone 40

mg/kg delivered via a marginal ear vein. In the heat-stressed group, rectal temperature was raised to at least420C for 15 minutes by wrapping the anesthetized rabbitin an electric warming blanket. Animals were thenallowed to recover at room temperature.Sham heat stress rabbits were identically anesthetized

and wrapped for similar periods with the blanket notturned on.

IL

-10cj

0,011

1

by guest on Novem

ber 13, 2017http://circ.ahajournals.org/

Dow

nloaded from

1266 Circulation Vol 88, No 3 September 1993

Stress Protein EstimationTwenty-four hours after the pretreatment interven-

tion, rabbits were reanesthetized with intravenous pen-tobarbitone (60 mg/kg), and hearts were removed forstress protein determination (see Fig 1). Excised heartswere washed and briefly retrogradely perfused with icedsaline to remove blood and albumin. In the ischemicpretreatment group, the coronary tie was retightenedand Coomassie brilliant blue R250 dye (BDH Chemi-cals, Poole, England) was introduced into the coronaryperfusate. Hearts were then removed from the perfu-sion rig, and atria, fat, and right ventricular free wallwere removed. In the ischemic pretreatment group, thearea without dye (risk zone) was separated from stained(perfused) myocardium. Left ventricular specimenswere then rapidly frozen in liquid nitrogen.At a later date, myocardial specimens were crushed

and homogenized in 2x concentrated sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer (20% glycerol and 6% SDS in0.12 mol/L Tris at pH 6.8), and protein concentrationswere estimated by use of the Pierce BCA reagent(Pierce, Rockford, Ill) and equalized by further additionof sample buffer where necessary; 2-mercaptoethanolto a final concentration of 10% was added beforeboiling. Samples were then centrifuged and stored at-700C. Subsequently, samples were thawed, recentri-fuged, and further diluted in sample buffer to allowloading of approximately 30 ttg of total protein per laneof slab gel.

Proteins were separated by SDS-PAGE on 0.8-mm-thick, 12.5% acrylamide gels according to Laemmli14Equivalence of loading and adequacy of sample prepa-ration were determined by visualization of proteins withCoomassie blue stain. When loading conditions wereoptimal, three identical gels were prepared, each loadedwith four heart samples derived from each of the fourgroups. The proteins on two gels were transferred ontoa nitrocellulose membrane (Hybond C, Amersham,Bucks, UK) by Western blotting. The other identical gelwas stained with Coomassie brilliant blue to visualizeprotein for subsequent densitometry.One nitrocellulose membrane was washed in phos-

phate-buffered saline with 0.1% dried skimmed milkpowder to block nonspecific binding sites. After washing,the membrane was incubated at 370C for 1 hour withmouse monoclonal IgG cross-reactive to the inducible72-kD heat shock protein (Stressgen, Sidney, Canada) at1:1000 dilution. After repeated washing, the membranewas incubated with horseradish peroxidase-conjugatedrabbit anti-mouse IgG (DAKO, Denmark) at 1:1000dilution at room temperature for 1 hour. The filter wasthen washed and developed by use of enhanced chemilu-minescence detection (Amersham, Bucks, UK).The other identical nitrocellulose membrane was

blocked with 3% skimmed milk powder and after wash-ing at room temperature for 3 hours was exposed to amouse monoclonal IgG raised against human heatshock protein 60 at 1:2000 dilution.15 Subsequent meth-ods were as described above with the second antibody at1:2000 dilution. The monoclonal used recognizes only

does not cross-react with the cytosolic homologue(TCP1).15The relative levels of heat shock proteins 72 and 60

were determined using densitometry (model 620 video-densitometer with Biorad analyst 2 version 3.1 software,Biorad, Hemel Hempstead, UK), normalizing to theactin band on the Coomassie-stained gel. This proce-dure adjusts for slight variations in protein loadingbetween samples.16

Myocardial Infarction and Infarct Size AssessmentApproximately 24 hours after the above pretreat-

ments, animals were reanesthetized with pentobarbi-tone (40 mg/kg) via the marginal ear vein. ECG elec-trodes were attached, and the trachea was opened andintubated via a midline cervical incision. Mechanicalventilation was as previously described; however, thetidal volume was reduced to 4 mL/kg in keeping with thereduction in ventilatory dead space. The right commoncarotid artery was cannulated with a short rigid poly-thene cannula attached to a pressure transducer(P23XL, Gould) for continuous recording of arterialpressure and intermittent arterial blood gas estimations(ABL2, Radiometer, Copenhagen, Denmark). Through-out the procedure, rectal temperature was monitoredand maintained between 38.5°C and 39.0°C by an elec-tric warming pad. The chest was opened (or reopened)via a midline sternotomy, and a coronary artery wasidentified. Rabbits were given 1000 units of heparinintravenously, and the coronary artery was ligated asdescribed for the ischemic pretreatment group. In thecase of the ischemic pretreatment group, the existingcoronary tie was reused to ensure that the same coro-nary bed was rendered ischemic as on the previous day.In the sham ischemia group, identification of a suitablecoronary vessel was often difficult, since the surface ofthe heart was encased in a thin fibrinous exudate.

After a 30-minute coronary occlusion, reperfusionwas confirmed by the appearance of a myocardial blush.After 120 minutes of reperfusion, a further 1000 units ofheparin was given, and the heart was removed andretrogradely perfused with 0.9% saline. After blood hadbeen washed out of the coronary vasculature, the cor-onary snare was retightened and fluorescent micro-spheres were infused via the aortic cannula. Demarca-tion of the myocardial surface at risk (area withoutspheres) was confirmed under UV light. The heart wasthen frozen at - 18C overnight. The next day, while stillfrozen, the heart was sliced at 2-mm intervals at rightangles to its long axis. The slices were then incubated at37°C in triphenyltetrazolium chloride 10 mg/mL ofphosphate buffer at pH 7.4. When the viable myocar-dium had stained, slices were placed in 10% formalde-hyde solution. Approximately 24 hours later, heart sliceswere placed caudal surface upward and compressedbetween two glass plates separated by 2-mm spacers.Risk areas and areas of infarction were traced, photo-graphically enlarged, and planimetered (Summa Graph-ics, Summa Sketch II, Seymour, Conn). The area ofinfarction (no tetrazolium staining) was expressed as apercentage of the area at risk of infarction (no fluores-cent microspheres). The volume of myocardial tissue at

the mitochondrial form of the 60-kD stress protein and risk and the volume of infarction were calculated by

by guest on Novem

ber 13, 2017http://circ.ahajournals.org/

Dow

nloaded from

Marber et al Stress Protein and Myocardial Infarction 1267

MW Heat Stress Ischemic Sham Ischemic Sham Heat Stress(kDa) MFflFTFT MT1 MM1 FTM12 MT M [MM MF

J5%

II I II I

*1p=NS I!

p<1O.1 0 0|p<.01 ' I p4.05

FIG 2. Densitometric assessment of Westernblot loaded with four specimensfrom each ofthefour intervention groups andprobed against heatstress protein (HSP) 72. The graded induction ofHSP72 between groups seen after immunoblot-ting (upper portion of figure) becomes moreobvious after densitometric assessment (lowerportion offigure). On the basis ofoptical densityratios, heat stress and ischemic pretreatmentresulted in a similar level of HSP72 induction(2.8±0.3 units vs 2.5±0.2 units, respectively).However, sham ischemia caused a greater induc-tion than sham heat stress (1.0±0.3 units vs0.3±0.1 units, respectively). All comparisonswere by ANOVA.

multiplying the corresponding areas by the depth (2mm) of the tissue slices.

StatisticsAll values are expressed as mean±SEM. All compar-

isons between groups were assessed for significance witha one-way ANOVA, Fisher's protected least significantdifference method being used for comparisons withinthe ANOVA table. Changes in hemodynamic parame-ters within groups over the course of the experimentwere compared by two-way ANOVA with a post hocpaired t test for comparison of variables at baseline andend of reperfusion. An unpaired t test was used tocompare baseline temperatures between heat stress andsham heat stress groups. Association between rate-pressure product (RPP) and infarct size was tested bythe Spearman rank correlation method. Statistical sig-nificance was defined as P<.05.

ResultsTemperature Changes During Heat StressBefore warming, there was no significant difference

between basal rectal temperature within sham or heatstress groups (39.1+0.1°C, n=14, and 39.0±0.10C,n= 14, P=NS, respectively). After the heat stress rabbitswere wrapped in an electric warming blanket, an aver-

age of 42.6±2.9 minutes elapsed before the rectaltemperature reached 420C. The peak temperature re-

corded was 42.3±0.10C, and the time that the rectaltemperature remained above 420C was 17.6±1.4minutes.

Myocardial 72-kD Stress Protein ContentAn initial protocol with only 6 hours of recovery after

ischemic pretreatment caused some increase in HSP72,but this was not significantly different from basal expres-sion and was significantly less than the elevation seen 24hours after heat stress. The recovery time after ischemicpretreatment was therefore increased to 24 hours.

Fig 2 demonstrates that significant induction of the72-kD stress protein occurred 24 hours after heat stressand ischemic pretreatments. This protein was detected inall myocardial samples examined (including those from

the nonischemic area, data not shown). When the blotwas examined densitometrically, the graded induction ofstress protein with differing interventions became moreapparent with thermal pretreatment, resulting in margin-ally greater induction than ischemic pretreatment. How-ever, the mean stress protein content in each of thepretreatment groups was greater than in correspondingcontrols. Between the control groups, there was a signif-icant induction in the sham ischemic pretreatment group.The reason for this may lie in the greater surgical traumaoccurring with sham ischemic pretreatment, althoughother differences exist between these two control groups,including the method of ventilation and anesthetic regi-mens. Compared with the sham heat stress group, therewas an approximately eightfold HSP72 induction bythermal, sevenfold by ischemic, and threefold by shamischemic pretreatments.

Myocardial 60-kD Stress Protein ContentFig 3 demonstrates that the marked variation seen

between experimental groups for the 72-kD stress pro-tein was not apparent when an identically loaded gelwas blotted and probed for the 60-kD stress protein. Inthis model, cardiac HSP60 was not significantly elevatedby whole-body heat stress. The only intervention asso-ciated with any change in HSP60 content was ischemicpretreatment, which resulted in an approximately 1.5-to 2-fold induction compared with other interventiongroups. Therefore, differences exist in stress proteininducibility, with HSP72 being more inducible thanHSP60, whereas HSP60 is preferentially, though onlymodestly, induced by ischemia.

Infarct SizeRisk volume, infarct volume, and percentage of risk

area infarcted are shown in Fig 4. The volumes ofmyocardial tissue at risk after coronary ligation werenot significantly different between intervention groups.However, thermal or ischemic pretreatment resulted inboth a significant reduction in absolute infarct volumeand a reduction in infarct volume expressed as a per-centage of the volume of tissue at risk. The volume ofinfarction as a percentage of the volume at risk for

46>

4

cj20.

I

0]

by guest on Novem

ber 13, 2017http://circ.ahajournals.org/

Dow

nloaded from

1268 Circulation Vol 88, No 3 September 1993

MW Heat Stress Ischemic Sham Ischemic Sham Heat Stress(kDa) mrrr mmrflr mlrm mmTM

69>-

5 -1wc

o 4 -

LO 3 -

0

2-

0

* I* .sf

.1

FIG 3. Densitometric assessment of Westernblot loaded with four specimensfrom each ofthefour intervention groups andprobed against heatstress protein (HSP)60. The lane arrangementgel loading conditions and figure layout areidentical to Fig 2. HSP60 levels appeared quitesimilar between groups. The HSP60 band did,however, appear denser in the lanes containingsamples from the ischemically pretreated hearts.Densitometry confirmed that ischemic pretreat-ment was associated with a selective induction ofHSP60. Optical density ratios, ischemic pretreat-ment vs sham ischemia, heat stress, and shamheat stress were 3.8±0.4 vs 24±0.3, 2.6±0.3,and 2.1±0.2, respectively. All comparisons wereby ANOVA.

ischemic pretreatment compared with control was28.8+5.2% vs 52.0±5.2%, Ps.01, respectively, and forthermal pretreatment compared with control,32.8+3.8% vs 56.9±6.5%, P..01, respectively. There-fore, both ischemic and thermal pretreatment resultedin myocardial protection.

In all treatment groups, the infarcted areas exhibitedintramyocardial hemorrhage. The absence of any hem-orrhage was rare and usually indicative of a failure toreperfuse (see exclusions).

Hemodynamic and Metabolic DataFig 5 describes the changes in RPP that occur before,

during, and after the 30-minute coronary occlusion foreach group. There were no differences in RPP betweengroups before coronary occlusion or during ischemia,implying that myocardial work and therefore oxygendemand was likely to be similar between groups.During reperfusion, the RPP was consistently higher

in the heat stress group, and at various time points, itwas significantly different from all other groups. Thispreservation of RPP is likely to be independent ofinfarct size, since a similar reduction in infarct size wasseen with ischemic pretreatment but was not associated

1.6.

-^1.2.E0

E 0.8.

>0.4

o J

Ischemia I Heat Stress

Sham Ischemia 0 Sham Heat Stress )

p=NS

with any significant advantage in RPP. In an attempt tofurther elucidate the reason for this apparent advan-tage, we correlated RPP at the end of reperfusion withinfarct volume as a ratio of heart mass (volume to heartweight ratio). As seen in Fig 6, there was no correlationbetween infarct size and RPP for any group. Thisfinding suggests, surprisingly, that in this experimentalmodel, infarct size does not appear to be the majordeterminant of cardiac work.The Table includes the hemodynamic as well as

acid/base balance data for each experimental group(the RPP is displayed in graphic form in Fig 5). Therewere no significant differences between groups in termsof acid/base balance or heart rate.

DiscussionThis study demonstrates that both ischemic and heat

stress pretreatments elevate myocardial HSP72 to asimilar extent and are associated with a similar reduc-tion of infarct size. This reduction occurs despite similarRPPs before and during prolonged coronary arteryocclusion and therefore suggests that the protection is adirect result of enhanced myocardial resistance to in-farction. This enhanced resistance to infarction is at

-80

FIG 4. Bar graphs of risk and infarctzone volumes. Risk zones were demarcated

60 by the absence offluorescent microspheresE and infarct zones by absence of staining

.4 with triphenyltetrazolium. Both ischemic40 and heat stress pretreatments were associ-

2 ated with a significant reduction in infarctL volume as well as infarct zone as a per-

20 centage ofrisk zone. All comparisons were

by ANOVA.

L0

by guest on Novem

ber 13, 2017http://circ.ahajournals.org/

Dow

nloaded from

Marber et al Stress Protein and Myocardial Infarction 1269

*-* Sham lchemia261-

24 -

o-V Heat Stress*-* Sham Heat Stress

-20 0 20 40 60 80 100 120 140 160Minutes

FiG 5. Graph ofrate-pressureproduct before, during, and after30-minute coronary occlusion. The shaded area represents theperiod ofcoronary occlusion. Although no significant differencesexisted before or during coronary occlusion, upon reperfusion,rate-pressure product was consistently higher in the heat stressgroup. Heat stress was the only intervention without a significantdrop in rate-pressure productfor the duration ofthe experiment.The minus-15-minute hemodynamic data were collected whilethe chest was closed, and the minus-S-minute data were collectedwhen the chest and pericardium were reopened. *Ps.05 heatstress vs ischemia and sham ischemia; tP5.05 heat stress vssham ischemia and sham heat stress; #P5.05, ##P..01 foreach group at 150 minutes vs baseline byANOVA.

least associated with HSP72, and it is tempting topostulate that it occurs as a direct consequence ofmyocardial stress protein induction.

Basis for the Reduction in Myocardial Infarct SizeThe design of this study is similar to and the findings

are consistent with those of Donnelly et al.13 Thoseinvestigators were able to demonstrate significant in-farct size reduction after whole-body heat stress; how-ever, ischemic pretreatment with a single 20-minutecoronary occlusion failed to reduce infarct size but alsofailed to increase myocardial HSP72 to the level seenafter heat stress. Those investigators, therefore, con-cluded that the absolute levels of HSP72 may be impor-tant in conferring protection from ischemic injury. Thisconclusion is supported and extended by the findings ofour study, which demonstrate that greater ischemic ele-vation of myocardial HSP72 is indeed associated withprotection. Moreover, a modest elevation of HSP72, asoccurred with sham ischemic pretreatment, is not asso-ciated with protection. In addition, our study suggeststhat HSP60 induction appears not to be a prerequisite formyocardial salvage, whereas an appropriate level ofHSP72 may be. However, the relation between whole-body heat stress and subsequent myocardial salvage iscomplicated by the fact that in the isolated rat heart,protection seems to involve an increase in myocardialcatalase activity rather than an induction of HSP72, sinceprotection can be abolished by inhibition of myocardialcatalase with 3-amino-triazole.17 In a preliminary report,endogenous catalase activity was also increased by re-peated ischemia and reperfusion18; therefore, a common

R2--0.08p=0.6

0

o Heat Stress

* Sham Heat Stress

0 Ischeona

* Sham Ischemia

0.02 0.04 0.06 0.08Infarct voVHeart Weight (cffl Ig)

R2-0.01c P-0.8

01

01

0.02 0.04 0.06 0.08Infarct vol/Heart Weight (cel Ig)

01

01

R2=0.08E p=0.7

|20. 8 8

!10.0 .

A

0.0 0.04 0.08 0.12 0.16 0.2Infarct vol/Heart Weight (cl Ig)

,30. * R2=0.03

E pM0.8

120 -

10

0.0 0.04 0.08 0.12 0.16 0.2 0.24Infarct vol/Heart Weight (cdl Ig)

FIG 6. Scatterplots showing relation between infarct volume normalized to heart weight and rate-pressure product at 120 minutesof reperfusion. R2 represents the coefficient of determination. There is no correlation between normalized infarct volume andrate-pressure product in any of the intervention groups. This finding implies that infarct volume does not determine rate-pressureproduct in our experimental model.

X= C° 22 -

X.I2 E 20-

E 18-

am 16-

14 -

12 -

°40

d 30

&20

;§10-

i _

0.0

a, 30.mEE20 _

10.

j nn

_

UMv

by guest on Novem

ber 13, 2017http://circ.ahajournals.org/

Dow

nloaded from

1270 Circulation Vol 88, No 3 September 1993

Hemodynamic, Arterial pH, and Oxygen Tension Changes During Myocardial InfadtionPeriod of coronary artery Period of myocardial

occlusion reperfusion

Baseline Preocclusion 5 min 15 min 30 min 45 min 60 min 90 min 120 min 150 minHeart rate (bpm)

Ischemia 297±6 277±7 272±9 278±6 270±5 257±6 258±7 255±5 269±13 271±1111Sham ischemia 293±11 277±9 271±10 271±11 284±15 267±10 271±8 268±10 270±11 260±1411Heat stress 273±9 273±10 269±9 269±8 265±9 256±10 259±10 266±12 262±11 264±11Sham heat stress 281±7 277±10 270±13 265±14 272±12 263±11 266±11 264±9 261±10 260±91

Arterial systolic pressure (mm Hg)Ischemia 92±7 90±7 76±10 77±6 74±6 70±5 70±6 67±6 73±6 74±811Sham ischemia 91±6 76±4 67±4 72±3 75±6 68±6 68±5 67±4 65±5 71±6Heat stress 99±6 86±6 83±6 87±6 84±7 85±6.4* 89±6 94±6t 88±6§ 87±6tSham heat stress 92±6 86±5 78±6 82±7 81±5 75±5 73±5 78±5 61±6 63±7¶

Arterial diastolic pressure (mm Hg)Ischemia 73±7 71±8 59±9 60±6 56±6 52±5 52±5 49±5 55±6 55±8¶Sham ischemia 73±6 59±4 53±4 58±3 59±5 53±6 52±5 50±4 50±5 52±611Heat stress 81±7 73±7 70±6 75±6 67±8 69±7 74±6* 76±5t 72±7 71±6Sham heat stress 78±7 70±5 63±7 68±8 65±7 62±8 57±7 58±6 43±6 49±6¶

Mean arterial pressure (mm Hg)Ischemia 81±7 80±8 66±9 68±6 63±6 60±5 60±6 56±5 62±7 63±8¶Sham ischemia 79±5 66±4 59±4 64±4 65±5 59±6 60±5 56±4 57±5 58±611Heat stress 88±6 80±7 77±6 80±6 76+7 77±7 82±6* 84±6t 79±7t 78±6Sham heat stress 84±7 76±4 74±8 75±7 76±6 68±6 68±6 71±6 56±6 60±71

Arterial blood pHIschemia 7.49±0.02 7.50±0.03 7.46±0.02 7.45±0.02 7.47±0.02 7.44±0.02 7.44±0.02Sham ischemia 7.53±0.02 7.48±0.02 7.43±0.04 7.44±0.02 7.43±0.01 7.40±0.02 7.39±0.0111Heat stress 7.51±0.02 7.53±0.04 7.46±0.02 7.44±0.02 7.45±0.01 7.45±0.01 7.45±0.01Sham heat stress 7.46±0.02 7.45±0.04 7.41±0.03 7.42±0.03 7.42±0.02 7.38±0.04 7.38±0.0411

Arterial blood oxygen (kPa)Ischemia 47.9±6.7 51.3±5.8 48.9±5.6 47.2±6.1 47.8±4.7 50.7±5.3 54.2±4.5Sham ischemia 60.3±3.1 59.1±3.1 55.1±2.0 51.4±4.7 48.4±4.6 50.1±3.5 54.9±3.1Heat stress 61.5±4.2 58.9±2.3 56.3±3.3 59.4±1.6 55.9±4.0 59.6±2.3 59.5±1.7Sham heat stress 52.3±5.8 54.7±2.0 50.6±2.6 53.1±2.0 50.5±3.4 50.5±5.0 55.0±6.1

n= 10 for each group. bpm, Beats per minute. Baseline data were recorded before the chest was opened. Mean weights (kg): ischemia,2.57±0.14; sham ischemia, 2.40±0.12; heat stress, 2.58±0.22; sham heat stress; 2.45+0.16.

*P<.05, tP5.01 vs ischemia and sham ischemia; tP5.05, §P..01 vs sham ischemia and sham heat stress using ANOVA; IIP..05, ¶P..01for each individual group at 150 minutes vs baseline using ANOVA.

mechanism independent of stress proteins may underliethe protection seen with both ischemic and heat stresspretreatments. Somewhat against this explanation is theobservation that heat stress is protective in a number ofdifferent biological systems against injuries not obviouslyinvolving oxidant stress.219 More importantly, a recentreport suggests that heart cells transfected so as tooverexpress HSP72 become resistant to anoxic injury.20 Apossible explanation for these apparently contradictoryfindings is the suggestion that stress proteins may in someway modify catalase activity.21 This is supported by theobservation that the increase in myocardial catalase afterheat stress is not transcriptionally regulated.22The mechanisms whereby stress proteins may directly

limit infarct size are necessarily speculative, since at themolecular level, the key sites of injury during ischemiaare poorly understood. The most attractive possibilitiesinvolve the stabilization of the cytoskeleton duringischemia by a13-crystallin (an HSP homologue).23 Alter-

natively, the chaperoning function (of the HSP90 andHSP70 families) of stress proteins may allow the correctrefolding of proteins sublethally damaged by ischemia/reperfusion or may prevent inappropriate molecularinteractions between these damaged proteins and oth-erwise viable proteins."7 Another possibility may simplybe that the general inhibition of protein synthesisaccompanying the stress protein response24 helps pre-serve the activity of more essential processes duringischemia, a hypothesis in keeping with the finding thatcell turnover is lower in myocyte cultures that overex-press HSP72 and are resistant to simulated ischemia.20

In the ischemic pretreatment group, the protectionobserved could be an ischemia-triggered increase incollateral flow or stunning. Both these explanations areunlikely, since it is unclear whether the oxygen require-ments and susceptibility to infarction are increased,decreased, or normal in stunned myocardium.25 In aprevious study, a pattern of ischemic pretreatment

by guest on Novem

ber 13, 2017http://circ.ahajournals.org/

Dow

nloaded from

Marber et al Stress Protein and Myocardial Infarction 1271

identical to that used in our experiments produced onlya very mild degree of stunning.4 In addition, the rabbitis a species deficient in collateral vessels26; therefore,any increase in collateral blood flow occurring as aresult of ischemic pretreatment would have to be on thebasis of new (mitotic) vessel growth, and this is unlikelyto occur over a 24-hour period.

Relation Between Ischemic Pretreatment andIschemic PreconditioningThe protocol used in this study and protocols known

to trigger "ischemic preconditioning" are identical.27Ischemic preconditioning, however, is a short-lived phe-nomenon with protective benefits waning within 2 hoursof reperfusion.27 This study therefore raises the possi-bility of a delayed, but perhaps longer-lasting, secondwindow of protection (SWOP) (a biphasic protectiveeffect). Evidence in support of such a phenomenonexists in neuronal tissue, since the stress protein contentof neuronal tissue can be elevated by 2-minute, dailyepisodes of ischemia, a treatment that also seems toincrease the neuronal resistance to more prolongedischemia.",29 Preliminary findings from other laborato-ries confirm a similarly delayed protection at 24 hoursafter "classic" preconditioning of the dog heart (withrepeated coronary occlusions)30 and also 24 hours afterrapid pacing of the rabbit heart.31 These experimentalfindings may explain the apparent benefit of a 27-dayhistory of angina before myocardial infarction,32 al-though differences in treatment and collateral vesselformation may complicate the issue.The relation between the early phase of protection

after brief episodes of ischemia (as seen in ischemicpreconditioning) and the later phase of protection (asseen in this study) is speculative. An obvious inconsis-tency is the fact that multiple coronary occlusions are aseffective as a single occlusion in triggering precondition-ing33 but differ in their ability to induce myocardialHSP72.4 This inconsistency may explain why protectionagainst ischemic arrhythmia was not seen 24 hours or 72hours after a single 5-minute coronary occlusion in therat.34

ConclusionsThe results of this study indicate that substantial

myocardial HSP72 induction is possible with eithersublethal thermal or ischemic pretreatment, whereasHSP60 is preferentially elevated by ischemic pretreat-ment. This observation is the first indication that myo-cardial HSP72 may be more important than other stressproteins, since myocardial salvage was seen after a30-minute coronary occlusion with both ischemic andthermal pretreatments. In addition to limiting infarctsize, whole-body heat stress has other nonspecific effectsthat act to maintain RPP.Our results provide evidence for what may prove to

be a clinically relevant myocardial adaptive responseoccurring 24 hours after short episodes of cardiacischemia that serves to increase myocardial resistanceto infarction.

AcknowledgmentsThe anti-60-kD heat shock protein antibody was a gift from

Dr G.A. Rook, Department of Medical Microbiology, Univer-

sity College London Medical School. M.S.M. is the recipient ofan intermediate fellowship from the British Heart Foundation.We thank the Hatter Foundation for continued support.

References1. Hightower LE. Heat shock, stress proteins, chaperones, and pro-

teotoxicity. Cell. 1991;66:191-197.2. Li CG, Werb Z. Correlation between synthesis of heat shock

proteins and development of thermotolerance in chinese hamsterfibroblasts. Proc NatlAcad Sci USA. 1982;79:3218-3222.

3. Sanchez Y, Linquist SL. HSP104 required for induced thermotol-erance. Science. 1990;248:1112-1115.

4. Knowlton AA, Brecher P, Apstein CS. Rapid expression of heatshock protein in the rabbit after brief cardiac ischemia. J ClinInvest. 1991;87:139-147.

5. Dillmann WH, Mehta H, Barrieux A, Guth BD, Neeley WE, RossJ. Ischemia of the dog heart induces the appearance of a cardiacmRNA coding for a protein with migration characteristics similarto heat-shock/stress protein 71. Circ Res. 1986;59:110-114.

6. Knowlton AA, Eberli FR, Brecher P, Romo GM, Owen A,Apstein CS. A single myocardial stretch or decreased systolic fibershortening stimulates the expression of heat shock protein 70 in theisolated, erythrocyte-perfused rabbit heart. J Clin Invest. 1991;88:2018-2025.

7. Yellon DM, Latchman DS, Marber MS. Stress proteins: anendogenous route to cardiac protection: fact or fiction? CardiovascRes. 1993;27:158-161.

8. Currie RW, Ross BM, Davis TA. Heat-shock response is asso-ciated with enhanced post ischemic ventricular recovery. Circ Res.1989;63:543-549.

9. Yellon DM, Pasini E, Cargnoni A, Marber MS, Latchman DS,Ferrari R. The protective role of heat stress in the ischaemic andreperfused rabbit myocardium. J Mol Cell Cardiol. 1992;24:895-907.

10. Liu X, Engelman RM, Moraru II, Lu D, Nath D, Rousou JA, FlackJE, Deaton DW, Maulik N, Das DK. Heat shock: a new approachfor myocardial preservation in cardiac surgery. Circulation. 1992;86(suppl II):II-358-II-363.

11. Walker DM, Pasini E, Kucukoglu S, Marber MS, Iliodromitis E,Ferrari R, Yellon DM. Heat stress limits infarct size in the isolatedperfused rabbit heart. Cardiovasc Res. 1993;27:962-967.

12. Yellon DM, Iliodromitis E, Latchman DS, Van Winkle DM,Downey JM, Williams FM, Williams TJ. Whole body heat shockfails to limit infarct size in the reperfused rabbit heart. CardiovascRes. 1992;26:342-346.

13. Donnelly TJ, Sievers RE, Vissern FU, Welch WJ, Wolfe CL. Heatshock protein induction in rat hearts: a role for improved myo-cardial salvage after ischemia and reperfusion? Circulation. 1992;85:769-778.

14. Laemmli UK. Cleavage of structural proteins during the assemblyof the head of bacteriophage T4. Nature. 1970;227:680-685.

15. Sharif M, Worrall JG, Singh B, Gupta RS, Lydyard PM, LambertC, McCulloch J, Rook G. The development of monoclonal anti-bodies to the human mitochondrial 60 kDa heat shock protein andtheir use in studying the expression of the protein in rheumatoidarthritis. Arthritis Rheum. 1992;35:1427-1433.

16. Norton PM, Latchman DS. Levels of 9OkDa heat shock proteinand resistance to glucocorticoid-mediated cell killing on a range ofhuman and murine lymphocyte cell lines. J Steroid Biochem. 1987;33:149-154.

17. Karmazyn M, Mailer K, Currie RW. Acquisition and decay ofheat-shock-enhanced postischemic ventricular recovery. Am JPhysiol. 1990;259:H424-H431.

18. Flack J, Engelman R, Rousou J, Iyergar J, Kimura Y, Prasad R,Das DK. Preconditioning the heart by repeated stunning improvesmyocardial salvage. Circulation. 1990;82(suppl III):III-575.Abstract.

19. Barbe MF, Tytell M, Gower DJ, Welch WJ. Hyperthermia protectsagainst light damage in the retina. Science. 1988;241:1817-1820.

20. Dillmann WH, Chi S-H, Sayen MR, Mestril R. Decreased ischemia(Ism) induced injury in rat heart derived H9c2 cells expressinghuman inducible heat shock protein 70 (hHSP70i). J Mol CellCardioL 1992;24(suppl III):S30. Abstract.

21. Kukreja RC, Hess ML. The oxygen free radical system: fromequations through membrane protein interactions to cardio-vascular injury and protection. Cardiovasc Res. 1992;26:641-655.

by guest on Novem

ber 13, 2017http://circ.ahajournals.org/

Dow

nloaded from

1272 Circulation Vol 88, No 3 September 1993

22. Currie RW, Tanguay RM. Analysis of RNA for transcripts forcatalase and SF71 in rat hearts after in vivo hyperthermia. BiochemCell Biol. 1991;69:375-382.

23. Bennardini F, Wrzosek A, Chiesi M. a3-Crystallin in cardiactissue: association with actin and desmin filaments. Circ Res. 1992;71:288-294.

24. Havre PA, Hammond GL. Isolation of a translation-inhibitingpeptide from myocardium. Am J Physiol. 1988;255:H1024-H1031.

25. Ohgoshi Y, Goto Y, Futaki S, Yaku H, Kawaguchi 0, Suga H.Increased oxygen cost of contractility in stunned myocardium ofdog. Circ Res. 1991;69:975-988.

26. Maxwell M, Hearse DJ, Yellon DM. Species variation in thecoronary collateral circulation during regional myocardialischaemia: a critical determinant of the rate of evolution and extentof myocardial infarction. Cardiovasc Res. 1987;21:737-746.

27. Walker DM, Yellon DM. Ischaemic preconditioning: from mech-anisms to exploitation. Cardiovasc Res. 1992;26:734-739.

28. Kitagawa K, Matsumoto M, Tagaya M, Hata R, Ueda H, NiinobeM, Handa N, Fukunaga R, Kimura K, Mikoshiba K, Kamada T.Ischemic tolerance phenomenon found in the brain. Brain Res.1990;528:21-24.

29. Nowak TS. Protein synthesis and heat shock/stress response afterischemia. Cerebrovasc Brain Metab Rev. 1990;2:345-366.

30. Yamashita N, Kuzuya T, Hoshido S, Fuji H, Oe H, Kitabatake A,Tada M, Kamada T. Relationship between time interval frompreconditioning to sustained ischemia and its effects on limitinginfarct size. J Mol Cell Cardiol. 1992;24(suppl I):S150. Abstract.

31. Szekeres L, Szilvassy Z, Udvary E, Vegh A. Moderate stress bycardiac pacing may induce both short-term and long-term cardio-protection based on different mechanisms. J Mol Cell Cardiol.1992;24(suppl I):S120. Abstract.

32. Muller DW, Topol EJ, Califf RM, Sigmon KN, Gorman L, GeorgeBS, Kereiakes DJ, Lee KL, Ellis SG. Relationship between ante-cedent angina pectoris and short-term prognosis after thrombolytictherapy for acute myocardial infarction: Thrombolysis and Angio-plasty in Myocardial Infarction (TAMI) study group. Am Heart J.1990;119:224-231.

33. Li GC, Vasquez BS, Gallagher KP, Lucchesi BR. Myocardialprotection with preconditioning. Circulation. 1990;82:609-619.

34. Shiki K, Hearse DJ. Preconditioning of ischemic myocardium:reperfusion-induced arrhythmias. Am J Physiol. 1987;22:H1470-H1476.

by guest on Novem

ber 13, 2017http://circ.ahajournals.org/

Dow

nloaded from

M S Marber, D S Latchman, J M Walker and D M Yellonresistance to myocardial infarction.

Cardiac stress protein elevation 24 hours after brief ischemia or heat stress is associated with

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1993 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.88.3.1264

1993;88:1264-1272Circulation. 

http://circ.ahajournals.org/content/88/3/1264World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer available in the

Permissions in the middle column of the Web page under Services. Further information about this process isOnce the online version of the published article for which permission is being requested is located, click Request

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Circulation Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on Novem

ber 13, 2017http://circ.ahajournals.org/

Dow

nloaded from