Hypoxia preconditions rabbit myocardium via adenosine and catecholamine release

J Mol Cell Cardiol 27, 1527-1534 (1995)

Hypoxia Preconditions Rabbit Myocardium via Adenosine and Catecholamine Release Michael V. Cohen 1, Robert S. Walsh 2, Mahiko Goto I and James M. Downey 1 1Departments of Medicine, Physiology, and Surgery, University of South Alabama, Mobile, Alabama, 2Present address: Department of Surgery, East Tennessee State University, James H. Quillen College of Medicine, Johnson City, Tennessee, USA Received 2 June 1994, acceptedin revisedform 8 November 1994) M. V. COHEN,R. S. WALSH,M. GOTOANDJ. M. DOWNEY.Hypoxia Preconditions Rabbit Myocardium via Adenosine and Catecholamine Release. Journal of Molecular and Cellular Cardiology (1995) 27, 1527-1534. It has been proposed that brief hypoxia can substitute for ischemia in the preconditioning of cardiac tissue and salvage of ischemic myocardium. To elucidate a possible mechanism isolated rabbit hearts were subjected to a 30-rain period of regional ischemia by occluding a previously snared coronary artery. Following 2 h of reperfusion infarct size was measured by staining left ventricular slices with triphenyltetrazolium chloride. In control hearts infarction averaged 28.7+ 1.9% of the risk zone. If the hearts were preconditioned with 5 rain global ischemia/10 rain reperfusion prior to the regional ischemia, then infarction was significantly reduced to 7.2 +2.0% (P<0.01). When global hypoxia (p02 of perfusate 42.0 + 2.1 mmHg) for ten rain substituted for the five rain period of global ischemia, protection was comparable to that observed after ischemic preconditioning (10.2 + 1.5% infarction, P<0.01 v control). During hypoxic perfusion adenosine release increased 16-fold over baseline levels. This protection could not be blocked by adding either the adenosine antagonist 8-(p-sulfophenyl)theophylline or the ~1-adrenergic blocker phenoxybenzamine to the hypoxic perfusate. However, co-administration of both agents to the hypoxic perfusate successfully aborted protection (22.6 + 2.9% infarction, P N.S. v control). Therefore, 10 rain of hypoxia releases both norepinephrine and adenosine in sufficient quantities such that either can completely precondition the heart. © 1995 AcademicPress Limited KEY WORDS:Adenosine receptor; ~l-adrenergic receptor; Ischemic preconditioning; Myocardial infarction; Myocardial salvage.

Introduction

hypothesis is that any of the m a n y receptors which couples to PKC should also be able to trigger the protected or preconditioned state. This has proven to be the case as M2-muscarinic (Thornton et al., 1993a; Hendrikx et al., 1993; Yao and Gross, 1993; Przyklenk and Kloner, 1994), ~z-adrenergic (Tsuchida et al., 1994), and angiotensin II (Lin et al., 1995) receptor agonists all seem to be equipotent with adenosine for triggering the preconditioned state. Typically a period of 5 rain of regional or global ischemia is used to initiate preconditioning

It is now well accepted that brief exposure of the heart to ischemia protects the myocardium against infarction from a subsequent ischemic insult. The mechanism of this protection is still not fully understood but recent studies from this laboratory suggest that adenosine released during the brief or preconditioning ischemia triggers the protection by stimulating protein kinase C (PKC) (Ytrehus et al., 1994a). One of the testable corollaries of the PKC

Please address all correspondence to: Michael V. Cohen, M.D., Department of Physiology. MSB 3050, University of South Alabama. College of Medicine, Mobile, AL 36688, USA.

0022-28281951081527+08 $12.00/0

1527

© 1995 Academic Preas Limited

1528

M.V. Cohen et al.

in the rabbit, and apparently in that species adenosine is the predominant receptor agonist released during such a short ischemia since blockade of only adenosine receptors completely prevents the protection (IAu et al., 1991). Several studies have reported that a protection similar to that observed after ischemic preconditioning can be induced with a period of hypoxic perfusion in lieu of a brief period of myocardial ischemia in isolated rat hearts (Neely and Grotyohann, 1984; Tani et al., 1991; Noto et al., 1992; Zhai et al., 1993; Lasley et aI., 1993) and in an in vivo canine model (Shizukuda et al., 1992). Indeed we were able to protect rabbit hearts against infarction when 10 min of hypoxia immediately preceded a 30-min period of regional ischemia, and we were able to block that protection with an adenosine receptor antagonist (Walsh et al., 1995). The protection in the latter protocol, however, was much less than that seen with ischemic preconditioning. One obvious explanation for the reduced protection is the possible injurious effect of 40 rain of continuous metabolic deprivation of the myocardium since the 10-rain preconditioning hypoxic period immediately preceded the 30-rain of regional ischemia without any intervening period ofreflow. Therefore, in the present study we wanted to test whether adding 10m in of reoxygenation after the 10 rain of hypoxia would eliminate the additional component of injury and result in protection equipotent to that of ischemic preconditioning. We also wanted to test Whether adenosine !would continue to be the sole mediator of this protection.

snare. The heart was rapidly excised and the heart mounted on a Langendorff apparatus by the aortic root. Coronary perfusion was resumed within 1 min.

Isolated perfused heart preparation The hearts were perfused at 100 mmHg constant pressure with modified Krebs-Henseleit buffer [NaC1 = 118.5 mM, KC1=4.7 raM, NaHC03 = 24.8 mu, KH2PO4= 1.2 raM, MgS04= 1.2 mM, CaC12=2.5 mM, and dextrose= 10 mM (Sigma Chemical Co., St. Louis, MO, USA)] aerated with 95% 02/5% COx at 37°C (pH 7.4). The perfusate was filtered prior to use through a 0.45 #m porosity filter to remove any particulate matter. A small latex balloon on the tip of a polyethylene catheter was inserted across the mitral valve into the left ventricle. The catheter was attached to a pressure transducer (Cobe Industrial, Lakewood, CO, USA), and left ventricular pressure and heart rate were contin,uously monitored on a Beckman R411 Dynagraph recorder. Volume was added to the balloon to adjust the left ventricular end-diastolic pressure to 5 mmHg. Two stainless steel hook electrodes were placed on the right atrium. A Grass (Natick, MA~ USA) square wave stimulator was used to pace the hearts at 200 beats per min with pulses of 4 ms and 5 V. Coronary flow was measured by collection of right heart effluent.

Experimental protocol

Materials and Methods Surgical preparation All animals were treated in accordance with "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences and published by the US Nation~ Institutes of Health (NIH publication No. 85-23, revised 1985). Adult New Zealand White rabbits (1.7-3.1 kg) were anesthetized with intravenous sodium pentobarbital (30mg/kg). Tracheotomy was performed for mechanical ventilation (MD Industries, MoNle, AL, USA) with 100% oxygen. Left thoracotomy was performed in the fourth intercostal space and the pericardium opened. A prominent branch of the left coronary artery was identified and a silk ligature was passed around this branch for later use as a

All hearts first had a 30 min equilibration period with oxygenated perfusate, and subsequently a 30min period of regional ischemia produced by coronary artery occlusion followed by 2 h reperfusion. The appearance of bulging dyskinetic myocardium distal to the ligature attested to the adequacy of the snare. Figure 1 depicts the experimental protocol used. Control hearts (group 1) received only the above described ischemia and reperfusion. In group 2 ischemic preconditioning was produced with a 5rain period of global ischen~i'a followed by 10 rain reperfusion prior to the regional ischemia.,Hypoxia hearts :.(group 3) were perfused with KrebsHenseleit buffer of identical composition as described above which was bubbled-with9 5% N2/5% C02 (pH = 7.4) to make the h e a r ( ~ o x i c for 10 min and then reperfused for 10rain with oxygenated buffer prior to the 30 min regional ischemia. In

Hypoxia and Preconditioning

Control

Hypoxia PBZ ~pT j Ischaemic

~]Normoxic perfusion [ ]

Hypoxic ~ perfusion

Regional ~-~ Global ischaemia ischaemia

Figure l Experimental protocols. All rabbits had an initial stabilization period and then 30 rain of regional ischemia foilowing coronary occlusion and subsequently 120 rain of reperfusion. [n some hearts 10 min of perfusion with hypoxic buffer preceded the period of regional ischemia by 10 min and outcomes were contrasted with those observed in hearts undergoing ischemic preconditioning (PC) with 5 rain of global ischemia and 10 min of reflow. Pretreatment of hypoxic hearts with the a-adrenergic antagonist phenoxybenzamine (PBZ) and/or the adenosine blocker 8-(p-sulfophenyl) theophylline (SPT) was performed to evaluate receptor mechanisms.

group 4, the non-selective adenosine receptor antagonist, 8-(p-sulfophenyl)theophylline (SPT) (Research Biochemicals Incorporated, Natick, MA, USA) was added to the perfusate at lOOpM concentration and was infused for five min prior to, during and five min after the hypoxia. In group 5, the ~radrenergic receptor blocker phenoxybenzamine (PBZ) was infused into the perfusate to produce a concentration of 10 JIM for 10 min prior to the hypoxic perfusion period. In group 6 the SPT and PBZ treatment of groups 4 and 5, respectively, were combined. Both oxygenated and hypoxic buffers were monitored throughout the protocols for pH, Po2, and Pco2 with a blood gas analyser (ABL 30, Radiometer Copenhagen, Denmark).

Infarct size analysis At the completion of the reperfusion period, the coronary artery snare was again tightened, and yellow fluorescing zinc cadmium sulfide particles ( l - 1 0 p m ) (Duke Scientific, Palo Alto, CA, USA) suspended in Krebs-Henseleit buffer were injected into the coronary arteries via the aortic root to identify the area at risk supplied by the selected coronary artery. The heart was weighed, frozen at - 8 ° C for 24 h, and sliced from apex to base into

1529

3 mm transverse sections• Sections were incubated at 38°C in 1% buffered 2,3,5-triphenyltetrazolium chloride (Sigma Chemical Co.. St• Louis, MO, USA) ( p H = 7.4) for 10 rain which allowed demarcation of the necrotic myocardium. Slices were placed in 10% formalin. After compression to a standard 2 m m thickness between glass plates, the sections were illuminated with long-wave ultraviolet light• The perfused myocardium appeared with yellow fluorescence. The area devoid of fluorescence represented the area at risk (AR). Triphenyltetrazolium chloride is a vital stain that reacts with NADH in the presence of dehydrogenase enzymes and cofactors to form a red pigment (Fishbein et al., 1981). Necrotic tissue loses these enzymes and cofactors during reperfusion and therefore does not stain. Infarcted myocardium was visible as pale areas adjacent to the red-stained viable myocardium. These boundaries were traced upon plastic overlays and the areas of the infarcted region (IS) and the AR were determined for each slice with a digitizing tablet and summed. Tissue volumes were determined by multiplying measured areas by slice thickness. The volume of tissue infarcted was expressed as a percentage of the myocardium at risk of infarction for each heart. Adenosine measurements To assess the amount of adenosine released by the hypoxic myocardium four additional hearts were prepared as above• During perfusion of the hea~s with oxygenated buffer coronary effluent dripp~g from the right heart was collected during two successive l-min periods. Then after initiation of perfusion with hypoxic Krebs buffer five 2-min samples were collected. Samples were frozen at - 70°C until analysed for adenosine by HPLC. Adenosine was separated from other purmes by using a reverse phase column (Supelco LC-18S,~upelco Inc. Bellefonte, PA, USA) and a 1% (pH 5.3) to 25% (pH 5.58) methanol in 100 mM KH2PO4 gradient (Model 600E system controller, 717 Wisp, 486 detector, Maxima 820 chromatography software, Waters, Division of Mifiipore, Milford, MA, USA). All peaks were detected by absorbance changes at 254 nm. The peaks of interest in the sample were identified by comparing retention times to those of known standards, and peak areas were quantified. •

4

,

lit"

Statistics

~,

Data are expressed as mean_+s.E.M. Comparisons among and between groups were determined by

1530

M.V. Cohen et al.

Table 1 Bodyand heart weights and risk zone and infarction volumes in isolated rabbit hearts Control Animal number Body weight ( k g ) Heart weight (g) Risk zone volume (cm3) Infarct volume (cm3) Infarct/risk zone ratio (%)

Ischemic preconditioning Hypoxia

5 2.26_+0.21 8.16_+0.45 0.73___0.04 0.20_+0.02 28.7_+1.9

5 2.05_+0.15 7.50_+0.28 0.81_+0.10 0.05_+0.01" 7.2_+2.0*

6 2.09_+0.08 8.25_+0.41 0.97_+0.07 0.09___0.021" 10.2_+1.5"

SPT 10 1.97_+0.04 8.08_+0.28 0.93_+0.08 0.12_+0.02 12.7_+1.6"

PBZ 4 1.88_+0.41 7.95+0.41 1.09_+0.07 0.07_+0.0216.7_+1.9"

SPT + PBZ 9 2.00_+0.07 7.28_+0.24 0.80_+0.07 0.19_+0.03 23.3_+2.1

M e a n + S.E,M.

Significanceof paired comparisonwith control:"I'P
Results Thirty-five rabbits were used in the six groups of this study. Numbers of animals in each group can be found in Table 1. There were no differences in either body or heart weight or risk zone size between any of the groups (Table 1). Baseline hemodynamics were comparable in all groups (Table 2). The Po2 of the control buffer perfusate averaged 745.5 + 16.0 mmHg. Changing from bubbled oxygen to nitrogen successfully lowered the perfusate pO2 to 42.0 +_2.1 mmHg. As noted in Table 2 global hypoxia markedly depressed left ventricular developed pressure in all groups, whereas depression with regional ischemia was more modest. Changes in coronary flow were generally small in most groups except for the absence of flow during global iscl~emia in the rabbits with ischemic preconditioning (group 2). Neither SPT nor PBZ produced s'ignificant alterations in either developed pressure or coronary flow. Developed pressure and coronary flow gradually declined in all groups during the final 2-h reperfusion period. Infarct size data are presented in Table 1 and Figure 2. In control hearts infarction averaged 28.7% of the risk zone. As anticipated, ischemic preconditioning resulted in significant myocardial salvage with only 7.2% infarction of the jeopardized tissue (P<0.01) (not shown in Fig. 2). In hearts exposed to global hypoxia in lieu of brief global ischemia protection was comparable. This hypoxiamediated protection was not affected by either the non-specific adenosine antagonist SPT or the irreversible • l-adrenergic blocker PBZ. Doses of drugs used in the present experiments have previously

been shown to effectively block the protection initiated by ischemic preconditioning (Liu et al., 1994) and phenylephrine infusion (Tsuchida et al., 1994), respectively. Only when SPT and PBZ were simultaneously infused prior to hypoxia could protection be blocked. Under these conditions infarction was no different than that seen in control hearts (22.6+2.9%, P N.S.). Although average risk zone volumes were not different among the six rabbit groups (Table 1), it is recognized that in the rabbit heart the percentage infarction is dependent on risk zone volume (Ytrehus et al., 1994b). To ensure that changes in risk zone size were not causing shifts in infarct volumes that might be misinterpreted, a plot of infarct v risk zone sizes is presented in Figure 3. There is little overlap between all data points from the two unprotected groups (control and SPT+PBZ) (open circles) and those from the protected hypoxic (group 3) and ischemically preconditioned (group 2) rabbits and animals treated with SPT (group 4) and PBZ (group 5) (filled-in circles). For the same risk zone volumes infarcts were smaller in the latter four groups. These results again indicate true differences between protected and unprotected groups. Adenosine levels were quantitated in the coronary effluent of four additional hearts before and during perfusion with hypoxic buffer. Baseline concentration averaged 0.32___0.17 #M. Adenosine levels rose continuously during the 10 rain of global hypoxia. Peak adenosine levels of 5.12 + 1.90/aM were observed in the last sample collected between 8-10 rain of hypoxia. Hence hypoxia resulted in a 16-fold increase in released adenosine.

Discussion Recently brief periods of hypoxia prior to longer periods of ischemia in the isolated rat heart have

Hypoxia and Preconditioning Table 2

1531

Hemodynamic data in isolated rabbit hearts LV dias P mmHg

LVDP mmHg

cor flow ml/min

107.4_+4.7 75.5 _+9.0 94.3 _+10.5 82.9 _+11.9 49.9 + 13.7"

54.8 _+3.6 38.8 _+4.0* 50.2 _+3.5 34.0_+ 1.8" 28.8 _+ 1.3"

Baseline 5 min regional ischemia 10 rain reperfusion 60 rain reperfusion 120 min reperfusion

Control 8.3 -+ 1.5 8.0 _+1.3 13.4 + 2.8* 10.2 + 2.2 9.7-+ 1.9

Baseline 5 min global ischemia 10 min reperfusion 5 min regional ischemia 10 min reperfusion 60 min reperfusion 120 rain reperfusion

7.0 _+2.7 7.3 _+ 1.8 8.1 _+2.5 8.8 _+1.9 12.2 _+2.3 11.3 +_1.8 13.3 -+ 1.7

103.1 _+9.1 2.1 _+0.5f 86.6 _+5.1 78.4 _+7.3 83.3 -+ 6.9 65.7 -+ 5.2 47.7 _+14.5

55.4 +_6.4 0f 56.4_+ 5.0 43.6 -+ 5.2 50.8 4- 6.4 38.8 _+6.9* 36.0 _+6.3f

Baseline 10 min hypoxia 10 rain reoxygenation 5 vain regional ischemia 10 min reperfusion 60 rain reperfusion 120 min reperfusion

Hypoxia 6.4 _+2.1 24.5 _+6.1 17.7 _+4.4 14.4 +_3.9 15.0 -+ 5.2 15.8 _+5.4 13.6 _+4.6

113.3 _+14.2 32.7 _+6.7f 86.1 -+ 9.7* 78.1 _+13.0f 80.4 _+12.8f 70.8 +_12.Of 63.3 -+ 10.8t

68.7 _+2.8 66.0 _+3.4 66.5 _+2.8 57.5 _+4.4 57.3 _+3.8 46.0 -+ 5.6 37.0 _+2.8

Baseline SPT l O m i n hypoxia + SPT 10 min reoxygenation 5 min regional ischemia 10 min reperfusion 60 rain reperfusion 120 min reperfusion

3.3 _+0.9 1.2 -+0.3 11.7_+ 3.2f 4.9 + 0 . 7 5.1 _+1.1 5.3 _+0.9 4.3 _+1.0 4.6_+ 1.5

90_+ 7.0 98.8 _+3.1 33.6 + 3 . 8 f 77.9 + 6.9 53.1 _+4.4f 68.6 _+3.9* 55.8 _+3.2t 52.1 _+3.3f

50.5 + 2 . 0 44.3 _+2.2 51.9+4.6 53.9 + 3.5 38.0_+ 3.0 49.5 _+3.7 36.4 _+2.7 29.0_+ 2.7

Baseline 10 min hypoxia 10 min reoxygenation 5 rain regional ischernia 10 rain reperfusion 60 rain reperfusion 120 rain reperfusion

7.5 -+ 2.4 9.0_+ 2.4 10.5 -+ 2.6 8.5 + 0.6 8.8 _+ 1.3 13.5 -+ 2.5 13.0 + 2.4 11.5 _+3.2

84.0-+ 11.2 69.5 _+10.8 14.7-+ 6.6t 74.5 _+8.3 43.5 + 4 . 1 t 67.0 +_8.5 57.5 + 3.2 55.3 -+ 1.9

59.5 _+6.1 53.0 _+5.5 57.5 + 5.9 66.3 + 9.3 44.0_+ 6.4 61.5 _+6.4 46.0 -+ 6.0 35.0_+ 7.0

Baseline PBZ PBZ + SPT 10 rain hypoxia + SPT 10 rain reoxygenation 5 rain regional ischemia 10 rain reperfusion 60 rain reperfusion 120 rain reperfusion

SPT + PBZ 5.2 _+0.5 6.4_+ 1.1 7.5 -+ 1.4 26.9 ± 4 . 4 t 15.5 _+3.7 15.8 + 3.4 16.6 -+ 3.6 16.0 _+3.8 12.9 _+3.4

113.6 _+8.6 98.5 + 7.3 90.2 _+9.8 23.9 _+1.9t 84.0 _+7.4* 55.4 + 6. i t 72.1 _+7.2t 58.4 -+ 6.8t 47.4-+ 7.0t

55.1 _+5.6 46.4 + 5.6 37.2 _+4.7 39.6 + 4 . 4 43.0-+ 3.8 25.8 -+ 2.7t 40.6 +_3.5 29.0 _+2.8t 21.4_+ 2.9t

PC

SPT

PBZ PBZ

Mean + S.I~.M. Significanceof paired comparison with baseline:*P<0.05 tP
1532

M.V. Cohen et al.

50

o~

40

~.

30

N

10' Hypo~da+ 10' Reoxygenation (isolated heart)

* P < 0.01

o

8

o o

~ ~.0

0

o

~

o

0

0

*

O±

•

8 Control

Hypoxia

Hypoxia

Hypoxia

s+~ P~z

Hypoxia +

~ + ~

Figure 2 Infarction as a percentage of myocardium at risk for control rabbit hearts with only 30 rain of regional ischemia and subsequent reperfusion as well as hearts perfused with hypoxic buffer for 10 min before regional ischemia. The latter hearts had significantly less infarction, and this protection was not affected by pretreatment with either phenoxybenzamine (PBZ) or 8-(p-sulfophenyl)theophylline (SPT). However, co-administration of both of these antagonists successfully blocked the beneficial effect of preconditioning with hypoxic perfusate.

0.3

o Unprotectedgroups

o

• Protected groups

o O o

O

0.2

~0.1

O

P

I

0.25

0.50

o

Oe

O

go

o

0

• o

o

•

• •

••

~

0.75

I'

1.00

•

I

1.25

1.50

Risk zone (cm a)

Figure 3 Plot of risk volume v infarct size for unprotected (groups 1 and 6) (open circles) and protected (groups 2-5) (filled-in circles) hearts. For comparable risk zones infarcts were much larger in the unprotected groups. The minimal overlap between protected and unprotected groups documents that differences in the size of risk zones did not account for observ&l variability in infarct size.

been demonstrated to reduce the time to ischemic contracture (Lasley et d., 1993) and enhance the recovery ofpostischemic ventricular function (Neely and Grotyohann, 1984; Tani e t a / . , 1991; Noto et d., 1992; Zhai et al., 1993; Lasley et a/., 1993). However, because post-ischemic functional abnormalities might be the result of both myocardial stunning and infarction, it was unclear whether an anti-infarct effect had been realized in this model, although Zhai (1993) had documented less CPK

release during the initial 15 rain of reperfusion in the previously hypoxic hearts compared to control animals. Shizukuda et al. (1992) observed that perfusion of a coronary territory for 5 min with hypoxemic blood prior to a 60-min cessation of coronary flow in dogs resulted in significantly smaller infarcts than in animals with only the coronary occlusion. This observation led them to propose that the hearts had been protected by the ischemic preconditioning phenomenon. Although changes in cardiac metabolism were suspected as a cause of this beneficial effect of hypoxia, only Lasley and colleagues (1993) attempted to identify a cellular mechanism. They were unable to block the protective effect of hypoxia with the specific A~adenosine antagonist 8-cyclopentyl-l,3-dipropylxanthine (DPCPX). Walsh et al. (1995) also studied the effect of hypoxia in an isolated rabbit heart model, and evaluated the importance of the reoxygenation period. They perfused the heart with buffer with an average Po2 of 33 mmHg for 10 min immediately before 30 min of regional ischemia. Therefore, there was no reoxygenation separating the hypoxic and ischemic periods. These hearts exposed to hypoxic perfusate had significantly smaller infarcts ( 2 1 . 0 + 4 . 2 % of risk zone) than control hearts (38.2___2.8% of risk zone, P<0.05) and this protection could be completely blocked when the adenosine antagonist SPT was included in the buffer perfusate. However, the protective effect of hypoxia was not nearly as comprete as that of ischemic preconditioning, and this difference was attributed to the accumulating injury from the continuous 40-min period of oxygen deprivation in the hypoxia-ischemia protocol. One of the aims of the present study was to determine whether hypoxia could indeed substitute for the brief ischemic period of a standard ischemic preconditioning protocol in the rabbit heart. Therefore, perfusion of the heart with hypoxic buffer was followed by a period of reoxygenation. As detailed above~ when the heart was exposed to this sequence of hypoxia-reoxygenation-ischemia, the observed prot~cti0n was no different from that experienced by h e ~ s subjected to ischemic preconditioning. This evidence confirms the striking protective effect of hyp0xia, and further supports the contention of Walsh et al. (1995) that the lesser protection observed in their hypoxic hearts was most likely a result of the specific protocol in which global hypoxia was immediately followed by regional ischemia. The present data also support a receptor-mediated mechanism for this protection. Unlike the data of

Hypoxia and Preconditioning Walsh et al. (1995), however, SPT-blockade alone was unsuccessful in aborting the protective effect of hypoxia. SPT is a hydrophilic compound that remains in the extracellular space. Therefore, in isolated rabbit hearts it can be administered for a specified period of time and then washed out during a drug-free perfusion interval. As noted previously in the isolated rabbit heart (unpublished observation) a 5-min wash out is sufficient to restore the ability of exogenous adenosine to decrease heart rate and dilate coronary arteries. Liu et al. (1994) demonstrated that SPT infusion beginning 5 min before and ending 5 min after a 5-min period of preconditioning ischemia successfully blocked the protection of ischemic preconditioning. But in the present study only combined adenosine and ~ladrenergic receptor blockade could effectively block protection. The reasons for the difference between the present results and those of Walsh et al. (1995) are unclear, but presumably are related to the reoxygenation period between hypoxia and ischemia. The dual receptor activation in these hypoxic animals was unexpected. The salvage of ischemic myocardium by ischemic preconditioning and consequent production of endogenous adenosine is effectively blocked by adenosine receptor antagonists alone (Liu et al., 1991; Lin et al., 1994). Exogenous substances, e.g. adenosine (Liu et al., 1991) and ~l-adrenergic (Tsuchida et al., 1994) agonists, administered in lieu of preconditioning ischemia are also capable of triggering protection, presumably because of their ability to activate PKC. ~-Adrenergic receptor antagonists can block the protective effect of the cq-agonist phenylephrine (Tsuchida et al., 1994) or tyramine-induced norepinephrine release (Thornton et aI., 1993b). However, neither adrenergic antagonist [phenoxybenzamine (Tsuchida et aI., 1994) or BE 2254 (Thornton et al., 1993b)] could block the protective effect of ischemic preconditioning. Ischemic tissue releases both adenosine (Dorheim et al., 1990) and catecholamines (SchOmig, 1990). But presumably only tissue adenosine levels reach the necessary threshold for activation of protein kinase C to produce protection in the ischemically preconditioned heart. The existence of such a threshold is suggested by the observation that two successive 2-min coronary occlusions are unsuccessful at triggering the preconditioning phenomenon, whereas a single 5 rain occlusion can (Van Winkle et aI., 1991). As demonstrated by the adenosine measurements, adenosine released after lOmin of hypoxia results in a 16-fold increase over baseline levels. In a prior in-

1533

vestigation using the microdialysis technique to measure interstitial adenosine, peak adenosine levels during 10 rain of occlusion were elevated to a comparable degree (Cohen et al., 1995). It is hypothesized that hypoxia additionally results in sufficient release of endogenous catecholamines to reach threshold for protection independent of adenosine release. Only simultaneous dual blockade prevented protection. This redundant signalling pathway seems to be an important adaptation of the myocardinm against the deleterious effects of ischemia and hypoxia. There are now at least four receptors with either known or putative coupling to protein kinase C which can initiate protection of ischemic myocardium. Exogenous ~l-adrenergic (Tsuchida et aI., 1994), angiotensin (Liu et al., 1995), adenosine (Liu et a/., 1991; Liu et al., 1994), and muscarinic (Thornton et al., 1993a; Hendrikx et al., 1993; Yao and Gross, 1993; Przyldenk and Kloner, 1994) agonists can each successfully precondition the heart. In the present study it was seen that both adrenergic and adenosine receptors participated in the protection of ischemic myocardium during hypoxic preconditioning. It is possible that under differing hemodynamic conditions other combinations of these receptors may participate in myocardial protection. Therefore, before receptor mechanisms can ever be effectively excluded as a mechanism of protection, blockade of several in tandem may be required.:

Acknowledgements The authors acknowledge the technical assistance of Jack J. F. Daly. Portions of this work were presented at the Scientific Sessions of the American College of Cardiology at Atlanta, GA, March 13-17, 1994. This work was supported by an American Heart Association, Alabama Affiliate, grant-in-aid (R.S.W.) and grants from the National Heart, Lung and Blood Institute HL-50688 (M.V.C.) and HL-20648 (J.M.D.).

References COHEN MV, SNELL KS, TSUCHmA A, VAN WYLEN DGL, DOWNEY ]M, 1995. Effects of anesthesia and K+m~ channel blockade on interstitial adenosine accumulation in ischemic rabbit myocardium. Basic Res Cardiol (in press).

DoRitmM TA, WANe T, M~TZ~ RM JR, VAN WYU~ DGL, 1990. Interstitial purine metabolites during

1534

M.V. Cohen et al.

regional myocardial ischemia. J Surg Res 48: 4 9 1 497. FISHBEINMC, MEERBAUMS, RIT J, LANDOU, KANMATSUSE K, MERCIER JC, CORDAY E, GANZ W, 1981. Early phase acute myocardial infarct size quantification: validation of the triphenyl tetrazolium chloride tissue enzyme staining technique. Am Heart I 101: 593-600. HENDRIKX M, TOSmMA Y, MUBAGWA K, FLAMEN(] W, 1993. Muscarinic receptor stimulation by carbachol improves functional recovery in isolated, blood perfused rabbit heart. Cardiovasc Res 27: 980-989. LASLEY RD, ANDERSON GM, M ~ r z ~ RM JR, 1993. Ischemic and hypoxic preconditioning enhance postischemic recovery of function in the rat heart. Cardiovasc Res 27: 565-570. LIU GS, THORNTON J, VAN WINKLE DM, STANLEYA~gVH, OLSSON RA, DOWNEY JM, 1991. Protection against hlfarction afforded by preconditioning is mediated by A1 adenosine receptors in rabbit heart. Circulation 84: 350-356. LnJ GS, RICaARDS SC, OtSSON RA, MULr.ANE K, WALSH RS, DOWNEYJM, 1994. Evidence that the adenosine A3 receptor may mediate the protection afforded by preconditioning in the isolated rabbit heart. Cardiovasc Res 28: 1057-1061. LIU Y, rI~UCHIDA A, COHEN MY, DOWNEY JM, 1995. Pretreatment with angiotensin II activates protein kinase C and limits myocardial infarction in isolated rabbit hearts, l Mol Ceil Cardiol 27: 883-892. NKSLY JR, GROTYOHANNLW, 1984. Role of glycolytic products in damage to ischemic myocardium: dissociation of adenosine triphosphate levels and recovery of function of reperfused ischemic hearts. Circ Res 55: 816-824. Noro T, Mxtrgn T, UgASE K, E~DOH A, TsocnmA A, ItMURA O, 1992. Effect of anoxic perfusion on ischemic myocardial injury in isolated rat hearts, lap Heart J 33: 679-691. Pe,zYra,ENK K, KT.ONERRA, 1994. Acetylcholine acts as a "preconditioning-mimetic" in the canine model. (Abstract). J Am Coil Cardiol 23: 396A. ScnOr,flG A, 1990. Catecholzmines in myocardial ischemia: systemic and cardiac release. Circulation 82 (Suppl if): II-13-II-22. SmzortroA Y, MAI,I,gr RT, I.~E S-C, Dowl~mY I-IF, 1992.

Hypoxic preconditioning of ischemic canine myocardium. Cardiovasc Res 26: 534-542. TANI M, ASAKUKAY, EBIHAKAY, SHINMURAK, NAKAMURA Y, 1991. Effect of preconditioning (PC) with anoxic perfusion (HP) on sacroplasmic reticular Ca 2+ uptake (CasR) in reperfused rat hearts. (Abstract). Circulation 84 (Suppl II): II--433. THORNTON JD, LIU GS, DOWNEY]M, 1993a. Pretreatment with pertussis toxin blocks the protective effects of preconditioning: evidence for a G-protein mechanism. J Mol Ceil Cardiol 25: 311-320. THORNTON JD, DALYIF, COHEN MV, Y^Na X-M, DOWNEY IM, 1993b. Catecholamines can induce adenosine receptor-mediated protection of the myocardinm but do not participate in ischemic preconditioning in the rabbit. Circ Res 73: 649-655. TSUCHmA A, LIU Y, LIU GS, COHEN MV, DOWNEY JM, 1994. ~l-Adrenergic agonists precondition rabbit ischemic myocardium independent of adenosine by direct activation of protein kinase C. Circ Res 75: 576-585. WALSHRS, BORGESM, THORNTON~D, COHENMY, DOWNEY JM, 1995. Hypoxia preconditions rabbit myocardium by an adenosine receptor-mediated mechanism. Can ] Cardiol 11: 141-146. VAN WINKLE DM, THORNTON JD, DOWNEY JM, DOWNEY JM, 1991. The natural history of preconditioning: cardioprotection depends on duration of transient ischemia and time to subsequent ischemia. Coron Artery Dis 2: 613-619. YAO Z, GROSSG]', 1993. Role of nitric oxide, muscarinic receptors, and the ATP-sensitive K + channel in mediating the effects of acetylcholine to mimic preconditioning in dogs. Circ Res 73: 1193-1201. YTRrams K, LIu Y, DOWNEYJM, 1994a. Preconditioning protects ischemic rabbit heart by protein kinase C activation. Am J Physiol 266: H1145-H1152. YTRmIUS K, LIu Y, TSUCHIDAA, MIURA T, LIu GS, YANG X-M, HERBERT D, COHEN MY, DOWNEY J]~, 1994b. Rat and rabbit heart infarction: effects of anesthesia, perfusate, risk zone, and method of infarct sizing. Am J Physiol 267: H2383-H2390. 7a-In1 X, LAWSON CS, CAW AC, HEARSE DJ, 1993. Preconditioning and post-ischemic contractile dysfunction: the role of impaired oxygen delivery vs extracellular metabolite accumulation. I Mol Cell Cardiol 25: 847-857.