Clinical and Experimental Pharmacology and Physiology

(2002)

29

,

956–962

HEAT STRESS INCREASES ENDOTHELIUM-DEPENDENT RELAXATIONS AND PREVENTS REPERFUSION-INDUCED

ENDOTHELIAL DYSFUNCTION

Vincent Richard, Karine Laude, Cecile Artigues, Nathalie Kaeffer, Jean-Paul Henry and Christian Thuillez

INSERM E9920, IFRMP 23, Department of Pharmacology, Rouen University Medical School, Rouen, France

SUMMARY

1. Heat stress has been widely used to stimulate theexpression of stress proteins and is associated with variouscardiovascular changes, including anti-ischaemic effects.However,

the

effect

of

heat

stress

on

endothelial

function

isless clear.

2. Heat stress was induced in anaesthetized rats by increas-ing body temperature to 42

�

C for 15 min. Twenty-four hourslater, segments of rat aorta and mesenteric and coronaryarteries were mounted in organ chambers.

3. Heat stress markedly increased relaxation to acetylcho-line (ACh) in all three blood vessels studied, without affectingthe response to the nitric oxide (NO) donor sydnonimine-1.

4. Heat stress also increased aortic relaxation to histamineand the calcium ionophore A23187.

5. In the aorta, an inhibitor of NO synthesis abolished theresponse to ACh in both control and heat stressed-rings,whereas a cyclo-oxygenase inhibitor had no effect.

6. Heat stress also prevented completely the impairedresponse to ACh in coronary arteries isolated from ratssubjected to myocardial ischaemia and reperfusion.

7. Thus, heat stress increases the stimulated release of NOthe rat aorta and mesenteric and coronary arteries andprevents reperfusion-induced injury at the level of thecoronary endothelium.

Key words: coronary circulation, endothelial function,ischaemia, nitric oxide, reperfusion.

INTRODUCTION

Various stress conditions promote the expression and synthesis ofstress proteins, such as heat shock proteins (HSP) or anti-oxidantenzymes. Induction of these proteins is considered to be a defencemechanism of cells against metabolic and environmental insults.Although a number of different stimuli trigger the release of stressproteins, including myocardial hypoxia or ischaemia,

1–4

expression

of stress proteins is often induced experimentally by transienthyperthermia or heat stress (HS).

2,5,6

Indeed, numerous biologicaleffects have been associated with HS, including protection againstarrhythmias,

7

calcium paradox and, in particular, ischaemia/reperfusion injury.

2,5,8–10

Although the cardiac effects of HS have been widelyinvestigated, the potential vascular changes induced by thisintervention

are

less

clear.

Induction

of

HSP

or

anti-oxidantscan be detected in the vascular wall or in cultured endothelialcells.

11

However, whether such expression is indeed associatedwith vascular functional changes, especially at the level of theendothelium, is largely unknown. Furthermore, whether thepreviously described protective effects of HS against ischaemia/reperfusion injury extend to endothelial cells is also unknown.

Thus, the present study was designed to assess: (i) whether priorHS affects endothelium-dependent relaxation in various rat arterypreparations (aorta, mesenteric and coronary arteries); and (ii)whether HS also prevents coronary endothelial dysfunctioninduced by myocardial ischaemia/reperfusion in rats.

METHODS

Induction of HS

The present investigation conforms with the Guide for the Care and use ofLaboratory Animals published by the US National Institute of Health (NIHpublication no. 85–23, revised 1996).

12

Experiments were performed inmale Wistar rats (Charles River, Saint Aubin les Elbeuf, France), weighingbetween 300 and 400 g. Rats were anaesthetized with 40 mg/kg, i.p.,sodium pentobarbital and placed in a temperature-controlled incubator.Body temperature was then allowed to increase to 42

°

C and was monitoredusing a rectal thermometer. The raised temperature was maintained for15 min, after which time the animals were allowed to return to their normalbody temperature. The animals were then left to recover for 24 h before theexperiments. Control animals were treated identically except that bodytemperature was not raised.

In vitro

vascular studies (normal arteries)

Aorta

Twenty-four hours after the initial rise in temperature, rats were re-anaesthetized with sodium pentobarbital (40 mg/kg, i.p.). A thoracotomywas performed and the thoracic aorta was removed and cut into rings. Therings were then mounted in organ chambers for the recording of isometrictension. Relaxation responses to increasing concentrations of acetylcholine(ACh), the nitric oxide (NO) donor sydnonimine-1 (SIN-1), histamine orthe calcium ionophore A23187 were obtained in arteries precontracted withphenylephrine (10

−

6

mol/L). In some experiments, arteries were incubated

Correspondence: Vincent Richard, INSERM E9920, Faculté deMédecine, 22 Bd Gambetta, 76183 Rouen Cedex, France. Email: Vincent.Richard@University-rouen.fr

Received 7 February 2002; revision 21 March 2002; accepted 9 April2002.

Endothelial effects of heat stress

957

with

N

G

-nitro-

L

-arginine (

L

-NNA; 10

−

5

mol/L), an inhibitor of NO syn-thesis, or diclofenac (10

−

5

mol/L), an inhibitor of cyclo-oxygenase. Inexperiments involving

L

-NNA, the concentration of phenylephrine wasadjusted in order to obtain levels of precontraction similar to those reachedin untreated rings.

Mesenteric and coronary arteries

Twenty-four hours after HS, the heart and intestine were removed andimmediately placed in cold oxygenated Krebs’ buffer. Small segments ofthe left coronary and mesenteric arteries were mounted in a small vesselmyograph for isometric tension recording.

13–15

Relaxation responses wereassessed

after

precontraction

with

phenylephrine

(mesenteric

arteries)

or5-hydroxytryptamine (5-HT; coronary arteries).

Evaluation of reperfusion-induced coronary endothelial injury

Twenty-four hours after HS, rats were anaesthetized with sodium pento-barbital administered intraperitoneally, intubated and ventilated. A leftthoracotomy was performed and the heart exposed. Animals were subjectedto 20 min ischaemia (left coronary artery occlusion) followed by 60 minreperfusion. This protocol is similar to that used in previous studies

14–16

andhas been shown to induce severe structural injury to the endothelium,

17

together with a markedly decreased response to ACh in the presence ofmaintained response to NO donors.

14

Sham animals were treated identicallyexcept

that

the

artery

was

not

occluded.

At

the

end

of

reperfusion

orsham

surgery,

the

heart

was

removed,

the

left

coronary

artery

wasdissected

out

and

a

small

segment

was

taken

and

placed

in

a

myograph,

Fig. 1

Relaxation responses induced by increasing concentrations of acetylcholine (ACh) in the (a) aorta and (b) mesenteric and (c) coronary arteriesisolated from sham (

�

) or heat-stressed (

�

) rats. Relaxations are expressed as a percentage of contractions induced by phenylephrine (aorta and mesentericarteries) or 5-hydroxytryptamine (coronary arteries). Data are the mean

±

SEM of eight animals in each group. *

P

< 0.05,

†

P

< 0.01 compared with sham.

Fig. 2

Relaxation responses induced by increasing concentrations of the nitric oxide donor sydnonimine-1 (SIN-1) in the (a) aorta and (b) mesenteric and(c) coronary arteries isolated from sham (

�

) or heat-stressed (

�

) rats. Relaxations are expressed as a percentage of contractions induced by phenylephrine(aorta and mesenteric arteries) or 5-hydroxytryptamine (coronary arteries). Data are the mean

±

SEM of eight animals in each group.

958

V Richard

et al.

as

described

above,

for

the

measurement

of

contractions

induced

by5-HT

or

relaxations

induced

by

ACh

or

the

NO

donor

SIN-1

(in5-HT-precontracted arteries).

Data analysis

All results are expressed as the mean±SEM and results were comparedusing Student’s t-test or one-way ANOVA followed, when the ANOVA wassignificant, by a Tukey test for multiple comparisons. P ≤ 0.05 was con-sidered statistically significant.

RESULTS

Effect of HS on the response to ACh and SIN-1

The effect of HS on the response of isolated aorta and mesentericand coronary arteries to ACh is shown in Fig. 1 (n = 8 per group).In isolated aorta, HS markedly increased the response to ACh.Indeed, the maximal response to ACh was increased from 75 ± 4 to91 ± 3%. This increase was present at concentrations as low as10−8 mol/L (9 ± 2 vs 24 ± 9% for sham and HS, respectively;P < 0.05) and was also reflected by the fivefold decrease in EC50

(from 0.26 ± 0.06 to 0.05 ± 0.01 �mol/L; P < 0.01). Heat stressalso increased the responses to ACh in mesenteric arteries (Fig. 1;maximal relaxations from 75 ± 4 to 91 ± 3%; P < 0.01) andcoronary arteries (Fig. 1; maximal relaxations from 75 ± 4 to91 ± 3%; P < 0.01).

In contrast with responses to ACh, HS did not affect the endo-thelium-independent responses to the NO donor SIN-1 in the threevessels tested (Fig. 2). Similar results were obtained with the NOdonor sodium nitroprusside (data not shown).

Effect of HS on contractile responses and basal NO release in the aorta

Heat stress did not affect the response to KCl 100 mmol/L inrings with endothelium (3.1 ± 0.3 vs 3.4 ± 0.4 g for sham and HS,respectively) or in rings without endothelium (3.5 ± 0.3 vs3.7 ± 0.9 g for sham and HS, respectively). The responses tophenylephrine in aorta taken from sham and HS rats (normalizedto KCl 100 mmol/L) are shown in Fig. 3. Heat stress did not affectresponses to phenylephrine in rings without endothelium or in ringswith endothelium either untreated or treated with the NO synthase(NOS) inhibitor L-NNA. Both endothelium removal and incubationof endothelium-intact rings with L-NNA markedly potentiated thecontractile responses to phenylephrine. This reflects removal of theinhibitory effect of basal NO release on contraction and, thus, canbe considered as an index of the basal release of NO. However, thepotentiation of the contractile response by L-NNA or by endo-thelium removal was not affected by HS, suggesting that HS didnot affect the basal release of NO.

Effect of NOS and cyclo-oxygenase inhibitors on responses to ACh in the aorta

The effect of L-NNA and diclofenac on responses to ACh in aortastaken from sham or HS rats is shown in Fig. 4. NG-Nitro-L-arginineabolished the response to ACh in both control and HS rings,suggesting that the increased response after HS is due to NO.Diclofenac induced slight, non-significant decreases in therelaxation induced by ACh in both sham and HS arteries.However, diclofenac did not affect the increased responseinduced by HS.

Fig. 3 Contractile responses induced by increasing concentrations of phenylephrine in aortas isolated from sham (open symbols) or heat-stressed (closedsymbols) rats, in rings (a) with or (b) without endothelium, in the absence (�, �) or presence (�, �) of the nitric oxide synthase inhibitor NG-nitro-L-arginine(10−5 mol/L). Contractions are expressed as a percentage of contractions induced by 100 mmol/L KCl. Data are the mean±SEM of eight animals in eachgroup.

Endothelial effects of heat stress 959

Fig. 4 Relaxation responses induced by increasing concentrations of acetylcholine (ACh) in aortas isolated from sham (open symbols) or heat-stressed(closed symbols) rats, in the absence (�, �) or presence (�, �) of the (a) nitric oxide synthase inhibitor NG-nitro-L-arginine or (b) the cyclo-oxygenaseinhibitor diclofenac (both 10−5 mol/L). Relaxations are expressed as a percentage of contractions induced by phenylephrine. Data are the mean±SEM of eightanimals in each group. *P < 0.05, †P < 0.01 compared with sham.

Fig. 5 Relaxation responses induced by increasing concentrations of (a) histamine or (b) A23187 in aortas isolated from sham (open symbols) or heat-stressed (closed symbols) rats, in the absence (�, �) or presence (�, �) of the nitric oxide synthase inhibitor NG-nitro-L-arginine (10−5 mol/L). Relaxationsare expressed as a percentage of contractions induced by phenylephrine. Data are the mean±SEM of eight animals in each group. *P < 0.05, †P < 0.01compared with sham.

960 V Richard et al.

Responses to histamine and A23187 in the aorta

Responses to histamine in aortas taken from sham or HS rats areshown in Fig. 5. In arteries taken from sham rats, histamineinduced modest relaxation responses (maximal relaxation24 ± 3%), which were markedly increased by HS (60 ± 9%). Theresponses to histamine were abolished by L-NNA in both controland HS rings.

Responses to the calcium ionophore A23187 are shown in Fig. 5.In this preparation, A23187 induced relaxations only at twoconcentrations (3 � 10−8 and 10−7 mol/L) and induced contractionsat higher concentrations. Heat stress significantly increasedresponses to A23187 (maximal relaxation from 56 ± 7 to75 ± 10%; P < 0.05). NG-Nitro-L-arginine abolished the responsesto A23187 in both control and HS rings.

Effect of HS on reperfused coronary arteries

Compared with non-ischaemic arteries, ischaemia/reperfusion withor without prior HS had no effect on the contractile responsesto 5-HT or the relaxation to the NO donor SIN-1 (data not shown).Ischaemia/reperfusion markedly reduced the response to ACh(maximal responses 61 ± 4 and 30 ± 8% for non-ischaemic (n = 6)and ischaemia/reperfusion (n = 9), respectively; P < 0.01; Fig. 6).The impaired response to ACh after ischaemia/reperfusion wascompletely prevented by HS (response to ACh 55 ± 6%, n = 8;P < 0.01 vs ischaemic/reperfused; NS vs non-ischaemic; Fig. 6).

DISCUSSION

The major finding of the present study was that prior HS increasedrelaxation to ACh and other endothelium-dependent vasodilators in

rat arteries. Thus, HS is one of the very few interventions capableof increasing endothelium-dependent responses in normal arteries(i.e. in the absence of any endothelial dysfunction). Furthermore,we demonstrated that HS was also protective against endothelialinjury induced by ischaemia and reperfusion.

The effect of HS on the response to ACh was observed in threedifferent arteries (aorta and mesenteric and coronary arteries),suggesting that this is a generalized effect, involving variousvascular territories and different sized vessels. Indeed, the markedincrease in the response to ACh after HS appeared similar at thelevel of a large conduit artery (the aorta) to that of a smallperipheral artery, as well as to that of an epicardial coronary artery.Moreover, the marked increase in the response to ACh was notassociated with changes in the response to the NO donor SIN-1.This suggests that it is not due to changes in smooth muscle cellreactivity, but is, indeed, due to an increased endothelial response.

At the level of the aorta, we found that the relaxation responsesto ACh in tissues from both sham and HS rats were abolishedby L-NNA, an inhibitor of NO synthesis. In contrast, the effect ofHS on responses to ACh was not affected by the cyclo-oxygenaseinhibitor diclofenac. Thus, the increased response does not appearto be due to changes in the production of vasoactive prostanoidsbut, rather, appears to be mediated entirely by increased productionof NO.

The effect of HS was not limited to the response to ACh, becausewe could also demonstrate a marked increase in the relaxationresponses to histamine. Again, these relaxations appeared entirelymediated by NO, because they were abolished by L-NNA. Further-more, we also found that HS increased the receptor-independantrelaxations induced by the calcium ionophore A23187. Thissuggests that the effect of HS cannot be ascribed to changes in thetransduction pathways coupling the ACh or histamine receptors toNOS but, rather, to a true increase in the calcium-mediated releaseof biologically active NO from the endothelium. However, whetherHS also increased calcium-independent activation of endothelial(e) NOS (such as occurs in response to changes in shear stress)cannot be answered from the present study and would requireexperiments in cannulated arteries in order to evaluate the effect ofHS on flow-mediated vasodilatation.

We found that HS did not affect the response to phenylephrinein aortic rings without or with endothelium, in the absence or thepresence of the NOS inhibitor L-NNA. In addition, although bothendothelium removal and incubation of endothelium-intact ringswith L-NNA markedly increased the response to phenylephrine,these increases were not affected by HS. Such an increasedcontractile response is usually considered to reflect removal of theinhibitory effect of the basal release of NO on contraction and, thus,can be considered as a functional index of basal NO release.18,19 Inthis context, our experiments suggest that, in contrast with thestimulated release, HS does not appear to affect the basal release ofNO from the endothelium.

The mechanisms by which HS increases NO-mediated relax-ations in our experiments are not clear. Obviously, one possibilityis that this effect is mediated by the expression of HSP, which, inturn, somehow facilitate the release of NO. However, although it ishighly likely, based on the available literature, that HS, as per-formed in our experiments, does lead to the induction of HSP, thelink between such expression and any type of biological responseis difficult to demonstrate directly.

Fig. 6 Relaxation responses induced by increasing concentrations ofacetylcholine (ACh) in coronary arteries isolated from rats subjected tosham surgery (�; n = 6) or to ischaemia followed by reperfusion, in theabsence (�; n = 9) or presence (�; n = 8) of prior heat stress. Relaxationsare expressed as a percentage of contractions induced by 5-hydroxy-tryptamine. Data are the mean±SEM of eight animals in each group.*P < 0.05, †P < 0.01 compared with ischaemia/reperfusion.

Endothelial effects of heat stress 961

One possible mechanism by which HS increases endothelium-dependent relaxations is through an upregulation of eNOS (typeIII). However, this hypothesis is unlikely because HS did not affectthe basal release of NO, as would be expected from an increasedexpression of eNOS. Another possibility is that the increasedendothelium-dependent relaxation is the consequence of a HS-induced expression of anti-oxidant enzymes (such as superoxidedismutase (SOD), catalase or glutathione peroxidase), leading to adecreased production of reactive oxygen species. Indeed, HS limitsthe production of free radicals in rat isolated hearts20 and increasesthe expression of SOD.21,22 Moreover, some of the biologicaleffects of HS can be reduced by inhibitors of catalase23 or antisenseagaint SOD.22 Because free radicals are potent inactivators of NO,a decreased production of free radicals secondary to increasedactivity of anti-oxidant enzymes could lead to increased NO-dependent relaxations.

Another finding of the present experiments is that HS completelyprevented the impaired response to ACh in coronary arteriesisolated from rats subjected to ischaemia and reperfusion. Thisdemonstrates that, in addition to increasing endothelium-dependentrelaxations in normal arteries, HS also protects the endotheliumagainst reperfusion-induced dysfunction and, thus, that the myo-cardial protective effects of HS already described in situations ofischaemia/reperfusion also extend to endothelial cells. A protectiveeffect of HS of the coronary microcirculation during ischaemia/reperfusion has already been described in rat isolated hearts24 but,to our knowledge, the present study is the first to assess suchprotection at the level of large arteries and after ischaemia/reperfusion in vivo.

As demonstrated previously by electron microscopy of coronaryarteries,17 the decreased response to ACh reflects marked structuralinjury to the endothelium and endothelial cell death. Thus, theincreased response after heat stress most likely reflects the pro-tective effects of this intervention against reperfusion-inducedendothelial injury and death, as demonstrated with preconditiong.17

This suggests that the mechanisms of the HS-induced restoredresponse to ACh after reperfusion differ from those of the increasedresponse in normal arteries.

As in the case of normal arteries, the endothelial protective effectof HS after ischaemia/reperfusion could be mediated by theexpression of HSP9,25,26 or by the increased expression of the anti-oxidant enzymes it induces. Indeed, in our model reperfusion,injury to the endothelium is mediated by free radicals because it isprevented by free radical scavengers given during reperfusion.15 Inaddition to directly inactivating NO, oxygen-derived free radicalsmay trigger the expression of neutrophil adhesion molecules, suchas intercellular adhesion molecule-1, and, thus, allow neutrophil-mediated endothelial injury to the endothelium. Thus, an increasedactivity of anti-oxidant enzymes after HS would, in turn, lead to alesser production of free radicals, which could explain the endo-thelial protection.

In conclusion, our experiments demonstrate that HS increasesendothelium-dependent relaxations in normal arteries andprevents impaired relaxations in a model of endothelialdysfunction (i.e. reperfusion injury). Identification of themolecular mechanisms responsible for these changes inendothelial function could lead to the discovery of new means toincrease the physiological release of NO both in physiologicaland pathological situations.

REFERENCES

1. Dillmann WH, Mehta HB, Barrieux A, Guth BD, Neeley WE, Ross JJ.Ischemia of the dog heart induces the appearance of a cardiac mRNAcoding for a protein with migration characteristics similar to heat-shock/stress protein 71. Circ. Res. 1986; 59: 110–14.

2. Marber MS, Latchman DS, Walker JM, Yellon DM. Cardiac stressprotein elevation 24 hours after brief ischemia or heat stress isassociated with resistance to myocardial infarction. Circulation 1993;88: 1264–72.

3. Das DK, Engelman RM, Kimura Y. Molecular adaptation of cellulardefences following preconditioning of the heart by repeated ischaemia.Cardiovasc. Res. 1993; 27: 578–84.

4. Benjamin IJ, Kroger B, Williams RS. Activation of the heat shocktranscription factor by hypoxia in mammalian cells. Proc. Natl Acad.Sci. USA 1990; 87: 6263–7.

5. Currie RW, Karmazyn M, Kloc M, Mailer K. Heat-shock response isassociated with enhanced postischemic ventricular recovery. Circ.Res. 1988; 63: 543–9.

6. Marber MS, Walker JM, Latchman DS, Yellon DM. Myocardialprotection after whole body heat stress in the rabbit is dependent onmetabolic substrate and is related to the amount of inducible 70 kDheat stress protein. J. Clin. Invest. 1994; 93: 1087–94.

7. Steare SE, Yellon DM. The protective effect of heat stress againstreperfusion arrhythmias in the rat. J. Mol. Cell. Cardiol. 1993; 25:1471–81.

8. Karmazyn M, Mailer K, Currie RW. Acquisition and decay of heat-shock-enhanced postischemic ventricular recovery. Am. J. Physiol.1990; 259: H424–31.

9. Yellon DM, Pasinin 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. Donnelly TJ, Sievers RE, Vissern FL, Welch WJ, Wolfe CL. Heatshock protein induction in rat hearts. A role for improvedmyocardial salvage after ischemia and reperfusion? Circulation1992; 85: 769–78.

11. Lu D, Maulik N, Moraru II, Kreutzer DL, Das DK. Molecularadaptation of vascular endothelial cells to oxidative stress. Am. J.Physiol. 1993; 264: C715–22.

12. Institute for Laboratory Animal Research, National Research Council.Guide for the Care and Use of Laboratory Animals. National AcademyPress, Washington DC. 1996.

13. Mulvany MJ, Halpern W. Contractile properties of small arterialresistance vessels in spontaneously hypertensive and normotensiverats. Circ. Res. 1977; 41: 19–26.

14. Richard V, Kaeffer N, Tron C, Thuillez C. Ischemic preconditioningprotects against coronary endothelial dysfunction induced by ischemiaand reperfusion. Circulation 1994; 89: 1254–61.

15. Kaeffer N, Richard V, Thuillez C. Delayed coronary endothelialprotection 24 hours after preconditioning. Role of free radicals.Circulation 1997; 96: 2311–16.

16. Richard V, Danielou E, Kaeffer N, Thuillez C. Delayed endothelialprotective effects of monophosphoryl lipid A after myocardialischemia and reperfusion in rats. J. Mol. Cell. Cardiol. 1999; 31:1117–23.

17. Kaeffer N, Richard V, Francois A, Lallemand F, Henry JP, ThuillezC. Preconditioning prevents chronic reperfusion-induced coronaryendothelial dysfunction in rats. Am. J. Physiol. 1996; 271: H842–9.

18. Clozel M, Kuhn H, Hefti F. Effects of angiotensin converting enzymeinhibitors and of hydralazine on endothelial function in hypertensiverats. Hypertension 1990; 16: 532–40.

19. Gray G, Clozel M, Baumgartner H. Effects of calcium channelblockade on the aortic intima in spontaneously hypertensive rats.Hypertension 1993; 22: 569–76.

20. Mocanu MH, Steare SE, Evans MCW, Nugent JH, Yellon DM. Heatstress attenuated free radical release in the isolated perfused rat heart.Free Radic. Biol. Med. 1993; 15: 459–63.

962 V Richard et al.

21. Hass MA, Massaro D. Regulation of the synthesis of superoxidedismutase in rat lungs during oxidant and hyperthermic stresses.J. Biol. Chem. 1988; 263: 776–81.

22. Yamashita N, Hoshida S, Otsu K, Taniguchi N, Kuzuya T, Hori M.Involvement of cytokines in the mechanism of whole-body hyper-thermia-induced cardioprotection. Circulation 2000; 102: 452–8.

23. Kingma JG, Simard D, Rouleau JR, Tanguay M, Currie RW. Effect of3-aminotriazole on hyperthermia-mediated cardioprotection in rabbits.Am. J. Physiol. 1996; 270: H1165–71.

24. Joyeux M, Bouchard JF, Lamontagne D, Godin-Ribuot D, Ribuot C.Heat stress-induced protection of endothelial function against

ischaemic injury is abolished by ATP-sensitive potassium channelblockade in the isolated rat heart. Br. J. Pharmacol. 2000; 130:345–50.

25. Marber MS, Mestril R, Chi S, Sayen R, Yellon DM, Dillmann WH.Overexpression of the rat inducible 70 kD heat stress protein in atransgenic mouse increases the resistance of the heart to ischemicinjury. J. Clin. Invest. 1995; 95: 1446–56.

26. Plumier J-CL, Ross BM, Currie RW et al. Transgenic mice expressingthe human heat shock protein 70 have improved post-ischemicmyocardial recovery. J. Clin. Invest. 1995; 95: 1854–60.