Ischemic Preconditioning and the β-Adrenergic Signal Transduction Pathway
Background—Previous studies from our laboratory showed cyclic increases in tissue cAMP during a multiple-cycle preconditioning (PC) protocol, followed by attenuated cAMP accumulation during sustained ischemia. The aim of this study was to determine whether ischemia-induced activation of the β-adrenergic signaling pathway could act as a trigger in eliciting protection.
Methods and Results—Isolated perfused rat hearts were preconditioned by 3×5 minutes of global ischemia, interspersed by 5 minutes of reperfusion. β-Adrenergic responsivity was assessed by measurement of tissue cAMP generation after β-adrenergic agonist administration at the end of the PC protocol. Tissue cAMP, adenylyl cyclase, and protein kinase A (PKA) activities and β-adrenergic receptor characteristics were assessed at different times. The role of cAMP generation in eliciting PC was studied by investigation of functional recovery during reperfusion after 25 minutes of global ischemia after (1) cAMP increases in the trigger period were prevented with the β-adrenergic blocker alprenolol 7.5×10−5 mol/L and (2) increases in cAMP were elicited by administration of forskolin 10−7 and 10−6 mol/L or isoproterenol 10−8, 10−7, and 10−6 mol/L. Intermittent ischemia resulted in reduced β-adrenergic responsivity at the end of the protocol, although Bmax and Kd values of the β-adrenergic receptor population and adenylyl cyclase and PKA activities were increased. Abolishment of cyclic increases in cAMP before sustained ischemia attenuated myocardial protection against ischemia, whereas agonists elicited protection. No clear correlation between protection and β-adrenergic desensitization was observed.
Conclusions—Ischemia-induced activation of the β-adrenergic signaling pathway during preconditioning should also be considered a trigger in eliciting preconditioning.
Despite considerable uncertainty regarding the exact mechanism whereby preconditioning elicits protection against ischemia, it is generally accepted that it is a receptor-mediated process with protein kinase C (PKC) activation as the common final pathway.1 2 3 However, the role of PKC could not be corroborated by all workers.1 3 As knowledge of preconditioning accumulated, it became apparent that several signal transduction pathways may participate.
Involvement of the β-adrenergic signal transduction pathway in preconditioning has recently been suggested: we4 and others5 6 have shown that preconditioning attenuates cAMP generation during sustained ischemia in rat and rabbit myocardium. Sandhu et al5 proposed reduced stimulation of the β-adrenergic receptor secondary to reduced release of norepinephrine7 rather than desensitization as an explanation. However, changes in the different components of the β-adrenergic signal transduction pathway could also explain the observation. We have recently shown that a multiple-episode preconditioning protocol is characterized by cyclic increases in cAMP, coinciding with opposite changes in phosphodiesterase activity, indicating that this pathway is changed during both preconditioning and sustained ischemia.8
We hypothesized that the repeated elevations of cAMP during short episodes of ischemia were caused by the release of endogenous catecholamines, resulting in downregulation of the β-adrenergic signal transduction pathway, contributing to attenuation of cAMP generation during sustained ischemia and functional improvement during reperfusion.
The aims of this study were to use a well-characterized model of multiple-episode–induced ischemic preconditioning to evaluate whether (1) treatment with reserpine would abolish the cyclic increases in cAMP; (2) downregulation of the β-adrenergic signal transduction pathway occurred by evaluation of the β-adrenergic receptor characteristics, PKA activity, and β-adrenergic responsiveness of preconditioned myocardial tissue; and (3) abolishment of or agonist-induced cyclic increases in cAMP would prevent or elicit myocardial protection against ischemia, respectively, and how the latter related to β-adrenergic receptor desensitization and cAMP generation during sustained ischemia.
Male Wistar rats weighing 200 to 250 g were used. Before anesthesia (30 mg pentobarbital IP), rats were allowed free access to food and water. This investigation conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication No. 85-23, revised 1985).
The perfusion technique of the isolated perfused working heart (not electrically stimulated, preload 15 cm H2O, afterload 100 cm H2O) and measurement of mechanical activity were as described previously.9 Krebs-Henseleit bicarbonate solution was used as buffer, containing, in mmol/L: NaCl 119, NaHCO3 24.9, KCl 4.74, KH2PO4 1.19, MgSO4 0.6, Na2SO4 0.59, CaCl2 1.25, and glucose 10. Oxygenation was done with 95% O2/5% CO2. Drugs were administered via a side arm into the aortic cannula. Normothermic zero-flow global ischemia was used for preconditioning. Mechanical activity was determined by measurement of aortic flow, coronary flow, heart rate, peak systolic pressure, and total work performance.
Adenylyl Cyclase and β-Adrenergic Receptor Assays
Sarcolemmal membranes were prepared by a modification of the method described by Strasser et al.10 Adenylyl cyclase activity was determined as described by Salomon et al,11 and radiolabeled cAMP was eluted from alumina columns.12 β-Receptor characteristics were determined by use of the β-antagonist [125I]iodocyanopindolol as specific ligand as described by Strasser et al.10 The Kd and Bmax values of the β-adrenergic receptors were calculated by use of the Enzfitter computer program (Robin J. Leatherbarrow, published by Elsevier-Biosoft).
β-Adrenergic Receptor Responsiveness: Biochemical Analyses
Hearts were freeze-clamped at various times during the experimental protocol with precooled Wollenberger tongs and immediately plunged into and stored in liquid nitrogen. cAMP analyses were done as described before4 with a commercially available [3H]cAMP assay system (Amersham).
Tissue cAMP-dependent PKA activity was determined with an assay from Gibco BRL. Frozen tissue was homogenized in an extraction buffer (5 mmol/L EDTA, 50 mmol/L Tris, pH 7.5) and centrifuged for 10 minutes at 600g, and the supernatant was diluted to ≈5 μg protein/10 μL extract. Four assay conditions were used for each sample (with and without inhibitor and with and without cAMP). Results were expressed as pmol activated PKA · min−1 · mg protein−1.
Role of Release of Endogenous Catecholamines in Causing Cyclic Increases in cAMP
Endogenous catecholamines were depleted by administration of reserpine 7 mg/kg IP 24 hours before experimentation (dose according to Reference 1515 ). Hearts from reserpinized and untreated animals were first perfused retrogradely (15 minutes), then for 15 minutes in the working mode (mimicking the exact technique used previously).9 Nonpreconditioned (Non-PC) hearts were subsequently perfused retrogradely for 30 minutes, whereas preconditioned (PC) hearts were subjected to 3 episodes of 5 minutes of global ischemia, interspersed by 5 minutes of retrograde reperfusion. Hearts were freeze-clamped after 30 minutes total perfusion time (controls) as well as at the end of each 5-minute period of global ischemia (PC1−, PC2−, and PC3−).
Evaluation of β-Adrenergic Responsiveness
Non-PC and PC hearts were perfused as described above. At 58 minutes of perfusion time, forskolin 10−6 mol/L or isoproterenol 10−8, 10−7, or 10−6 mol/L was added to the perfusate, and the hearts were freeze-clamped after 2 minutes for cAMP determination. Control hearts were freeze-clamped after 60 minutes of perfusion without administration of drugs (see Figure 2A⇓ for protocol).
Similar experiments were performed on rats pretreated with pertussis toxin in saline (25 μg/kg)16 or vehicle only 24 hours before experimentation with isoproterenol 10−7 mol/L.
Characterization of Changes in β-Adrenergic Signaling Pathway During Preconditioning
Non-PC hearts were perfused as described above and freeze-clamped after 30 and 60 minutes total perfusion time. PC hearts were freeze-clamped after 30 minutes total perfusion time (before onset of preconditioning), at PC1−, PC2−, and PC3−, and at the end of PC1+, PC2+, and PC3+ for determination of β-adrenergic receptor characteristics and adenylyl cyclase and PKA activities.
Role of β-Adrenergic Receptor Activation in Eliciting Protection
To assess β-adrenergic receptor blockade during preconditioning, pilot experiments showed that the nonselective β1/β2-adrenergic blocker alprenolol 7.5×10−5 mol/L abolished the increase in cAMP induced by intermittent ischemia and that 10 minutes of washout was sufficient to remove the drug totally (see Figure 3A⇓ for protocol).
Control hearts were perfused for 115 minutes. In a separate series, alprenolol was administered at 20 minutes total perfusion time for 3 episodes of 5 minutes interspersed by 5-minute periods of perfusion with buffer, followed by perfusion in the retrograde (75 minutes) and working (10 minutes) modes.
Non-PC hearts were perfused as above, in the absence or presence of alprenolol. After the last administration of the drug, the hearts were perfused with buffer for 10 minutes, followed by 25 minutes of global ischemia and 20 minutes of reperfusion (10 minutes retrograde, 10 minutes working heart).
PC hearts were preconditioned with 3×5 minutes of ischemia in the absence or presence of alprenolol. Alprenolol was administered 5 minutes before the onset of PC1− and during PC1+ and PC2+. After PC3−, the drug was washed out for 10 minutes before the onset of ischemia. Evaluation of the recovery potential was done by addition of epinephrine 10−6 mol/L at the end of each protocol and monitoring of function for 10 minutes.
To assess agonist-induced increases in cAMP and functional recovery after 25 minutes of global ischemia, hearts were perfused retrogradely for 15 minutes, followed by 15 minutes of working heart. In separate groups, tissue cAMP was then elevated experimentally by (1) 3×5 minutes of global ischemia, (2) forskolin 10−7 or 10−6 mol/L for 1×5 or 3×5 minutes, and (3) isoproterenol 10−8 or 10−7 mol/L for 1×5 or 3×5 minutes.
Non-PC hearts were perfused retrogradely for 30 minutes after the initial stabilization period. All hearts were then subjected to 25 minutes of global ischemia and 20 minutes of reperfusion for evaluation of functional recovery (for protocol, see Figure 4A⇓).
β-Adrenergic receptor responsiveness of the above groups was assessed by measurements of tissue cAMP (1) before onset of sustained ischemia, with isoproterenol 10−7 mol/L; and (2) at the end of 25 minutes of global ischemia (for protocol, see Figures 5A⇓ and 6A⇓).
The number of samples in each group studied is listed in the table and figures. All data are given as the mean and SEM. Multiple comparisons were analyzed by 1-way ANOVA, and the Bonferroni correction was applied.
cAMP During Preconditioning: Role of Endogenous Catecholamines
Hearts of nonreserpinized rats subjected to preconditioning exhibited increases in tissue cAMP similar to those previously described,4 whereas depletion of catecholamines by prior reserpination abolished these increases at every time point studied (Figure 1⇓).
Responsiveness of the β-Adrenergic Signal Transduction Pathway
Isoproterenol 10−8, 10−7, or 10−6 mol/L caused a significant increase in cAMP content of Non-PC hearts compared with untreated control hearts, whereas cAMP of all PC groups remained unchanged (Figure 2⇓). Forskolin elicited a similar 2-fold increase in tissue cAMP in both Non-PC and PC hearts, indicating that a change had been induced at the β-adrenergic receptor level. The diminished response to β-adrenergic stimulation was not due to increased Gi protein activity, because cAMP generation in response to isoproterenol 10−7 mol/L of hearts from pertussis toxin–treated rats did not differ from hearts of untreated animals.
Changes in β-Adrenergic Signaling Pathway During Preconditioning
Preconditioning caused a gradual increase in receptor density (Bmax) and Kd (ie, decrease in affinity), which became significant at PC3− (Table 1⇓). At the end of preconditioning, ie, immediately preceding sustained ischemia, Bmax was increased by 39% and affinity decreased by 35.5%.
Although adenylyl cyclase activity increased significantly from PC1− to PC3−, it was reduced at PC3+. cAMP-dependent PKA activity followed a pattern similar to that of adenylyl cyclase activity.
β-Adrenergic Receptor Blockade and Preconditioning-Induced Protection
Transient administration of alprenolol 7.5×10−5 mmol during the preconditioning protocol prevented the characteristic increase in cAMP (preconditioning without alprenolol: controls 302±26, PC1 to 398±8*, and PC3 to 392±26*, *P<0.05 versus controls; preconditioning with alprenolol: controls 264±16, PC1 to 310±20, and PC3 to 297±17 pmol/g wet wt; n=5 per series). Alprenolol had no effect on mechanical recovery during reperfusion of non-PC hearts: both treated and untreated hearts failed to produce aorta output (Figure 3⇓). The functional recovery of hearts preconditioned in the presence of alprenolol was significantly less than that of untreated PC hearts but better than that of Non-PC hearts. Although the cardiac output of both PC groups could be increased by administration of epinephrine, it remained lower in the alprenolol-treated hearts. This was not a drug effect, because the response to epinephrine of control hearts treated with alprenolol but not subjected to ischemia was similar to that of untreated hearts (results not shown).
Protection Due to Cyclic Increases in cAMP
Pilot studies showed that forskolin had a dose-dependent effect on cAMP: compared with control values of 330±48, forskolin 10−8, 10−7, and 10−6 mmol/L administered for 5 minutes caused an elevation of cAMP to 363±12, 539±24, and 1231±16 pmol/g wet wt, respectively (n=3 per series) (Figure 4⇓). When administered repeatedly, forskolin increased cAMP to the same extent every time. Forskolin 10−8 (results not shown) and 10−7 mol/L (Figure 4⇓) administered either once or 3 times failed to improve functional recovery compared with Non-PC hearts. However, similar treatment with forskolin 10−6 mmol/L significantly improved functional recovery during reperfusion, although the protection was less than that of ischemic preconditioning.
Administration of isoproterenol 10−8 and 10−7 mol/L for 5 minutes increased tissue cAMP to 428±26 and 1036±55 pmol/g wet wt, respectively, causing significant improvement in functional recovery (compared with Non-PC hearts) and protection similar to that observed with ischemic preconditioning. However, administration of isoproterenol for 3×5 minutes resulted in no increases in cAMP after the first administration, causing complete mechanical failure upon reperfusion.
β-Adrenergic Receptor Responsiveness Before the Onset of Sustained Ischemia and cAMP Generation During Sustained Ischemia
Ischemic PC, forskolin 10−6 mol/L, 3×5 minutes, and isoproterenol 10−7 mol/L, 1×5 minutes resulted in significant desensitization of the β-adrenergic response at the onset of sustained ischemia and a significant reduction in cAMP accumulation at the end of 25 minutes of ischemia (Figures 5⇓ and 6⇓). Forskolin 10−7 mol/L (3×5 or 1×5 minutes), forskolin 10−6 mol/L (1×5 minutes), and isoproterenol 10−8 mol/L (1×5 minutes) did not cause β-adrenergic receptor desensitization and reduced cAMP accumulation during 25 minutes of sustained ischemia.
The results suggest that the cyclic changes in cAMP levels and PKA activation during the preconditioning protocol may indeed contribute to desensitization of the β-adrenergic receptor and subsequent protection against ischemic damage.
What Causes the Cyclic Increases in cAMP During Preconditioning?
Our data suggest that release of endogenous catecholamines is mainly responsible for the cyclic increases in tissue cAMP and PKA activity during preconditioning: prior reserpination and β-adrenergic blockade with alprenolol both abolished cAMP generation at PC1−, PC2−, and PC3−. The findings that release of norepinephrine occurred within 2 minutes of ischemia in the rat heart17 and that a substantial reduction in synaptic norepinephrine content occurred during brief ischemia, which was terminated upon reperfusion,18 support our findings.
Other mechanisms that may be involved are (1) adenylyl cyclase activation, probably due to PKC activation,10 and (2) upregulation of the β-adrenergic receptor due to loss of ATP.10 However, cAMP generation in response to β-adrenergic receptor stimulation by norepinephrine will become less as affinity of the receptor for its ligand is reduced. Another possibility is a marked reduction in the activities of both cAMP and cGMP phosphodiesterases during PC1−, PC2−, and PC3−.8
Attenuation of the β-Adrenergic Response to Sustained Ischemia: β-Adrenergic Desensitization
Although Sandhu et al5 could not demonstrate reduced β-adrenergic receptor sensitivity in a rabbit model, our data clearly show that downregulation of the β-adrenergic signaling pathway can be attributed to desensitization of the β-adrenergic receptor: isoproterenol elicited a marked increase in cAMP in non-PC hearts but not in PC hearts, whereas forskolin caused similar increases in cAMP in both groups. Pertussis toxin pretreatment failed to increase cAMP in PC hearts, excluding a role for Gi protein.
The following situation thus prevails: a significant (39%) increase in Bmax and a reduction (35%) in the affinity of the β-adrenergic receptor for its ligand concomitant with normalization of adenylyl cyclase and phosphodiesterase activities. Clearly, the reduced affinity of the receptor for its ligand overrides the effect of the increase in Bmax. Desensitization can be mediated by changes in the functional state of the receptor induced by phosphorylation19 20 and by the number of receptors present on the cell surface.19 The significant elevation in PKA activity could be particularly important in the desensitization process.19 21 Also, phosphorylation of the receptor by the β-adrenergic receptor kinases, particularly β-ARK-1 (or GRK2), during ischemia22 may be involved. The role of the latter kinases in preconditioning is not known.
Other factors that may also contribute to the reduction in cAMP during ischemia are (1) protection of myocardial autonomic nerve terminals by preconditioning,23 (2) reduced release of endogenous catecholamines during sustained ischemia,7 and (3) the significant increase in cAMP and cGMP phosphodiesterase activities during sustained ischemia in PC hearts.8 Reduced accumulation of cAMP may conceivably be protective because of the reduction in energy demand and calcium influx.
Mechanism of Protection: cAMP Related?
The question we addressed was whether the cyclic increases in cAMP (and thus PKA activation) occurring during preconditioning are involved in eliciting protection against ischemia. Two possibilities should be considered: cAMP/PKA may play a role in β-adrenergic desensitization and/or may phosphorylate an unknown target downstream. We investigated the former possibility.
Our results underscore the significance of β-adrenergic stimulation during the PC protocol: abolishment of the increases in cAMP during PC attenuated functional recovery during reperfusion (Figure 3⇑). The reverse approach was to establish whether repeated generation of cAMP before the onset of sustained ischemia could elicit protection. Indeed, such a relationship could be established, depending on the mode of administration and concentration of the agonists (Figure 4⇑).
No clear correlation between elevation in tissue cAMP levels before sustained ischemia and subsequent protection was observed. Although isoproterenol 10−8 mol/L was capable of eliciting protection at cAMP levels comparable to those observed during a PC protocol (428±26 versus 398±8 pmol/g wet wt, respectively), further elevation to 1036±55 pmol/g wet wt by isoproterenol 10−7 mol/L did not further enhance recovery. Furthermore, only at a concentration of 10−6 mol/L and cAMP levels of >1000 pmol/g wet wt did forskolin elicit protection. However, this may be because forskolin increases cAMP in a compartmentalized manner.24 The complete mechanical failure observed after repeated stimulation with isoproterenol may be a result of the well-established harmful effects of excess cAMP.25
The mechanism of protection elicited by β-adrenergic agonists is complex. With preconditioning or isoproterenol 10−7 mol/L, 1×5 minutes, or forskolin 10−6 mol/L, 3×5 minutes, the following pattern emerged: significant desensitization of the β-adrenergic receptor (Figure 5⇑) and a reduced cAMP content after 25 minutes of sustained ischemia, whereas the converse was true for forskolin 10−7 mol/L, 1×5 minutes. However, with isoproterenol 10−8 mol/L or forskolin 10−6 mol/L administered for 1×5 minutes, functional protection was not associated with either of the above. In view of these discrepancies, we suggest the following possible mechanisms: (1) the activation of PKA observed in all agonist-treated hearts (results not shown) may act via desensitization of the β-adrenergic receptor and/or phosphorylation of a protective protein and (2) agonist-induced increases in heart rate: rapid pacing can protect against ischemia via NO production.26 27 It is possible that both factors play a role in the protection observed.
PKA Versus PKC
The results obtained provide proof that the protection elicited by PC may be partially dependent on activation of the β-adrenergic signaling pathway, which implies a role for PKA. This possibility is supported by previous findings that repeated stimulation with norepinephrine or isoproterenol mimics ischemic preconditioning.28 Also, repeated β-adrenergic stress induced a long-term cardiac adaptation manifested by a reduction of harmful ischemic changes due to cardiac stress 24 and 48 hours after preconditioning.29
Although PKC activation has long been advocated as the main signal transduction pathway in PC, we30 and others (reviewed in References 1 and 31 3 ) could find no evidence for this. The finding that β-adrenergic blockade effectively reduced a 3-cycle PC-induced protection (Figure 4⇑), as opposed to the failure of α1-adrenergic or PKC blockade to abolish protection,30 is indicative of a role for this pathway in our model.
Finally, in view of the results obtained in the present study, we propose that, in addition to stimulation of G protein–coupled receptors such as the muscarinic, angiotensin II, or opioid receptors,1 2 or PKC activation, ischemia-induced activation of the β-adrenergic signaling pathway should also be considered as a trigger or contributory factor in eliciting PC. The exact role of PKA still needs to be resolved.
This study was supported by the MRC, South Africa, Harry Crossley Trust.
- Received December 11, 1998.
- Revision received April 19, 1999.
- Accepted April 27, 1999.
- Copyright © 1999 by American Heart Association
Brooks G, Hearse DJ. Role of protein kinase C in ischemic preconditioning: player or spectator? Circ Res. 1996;79:627–630.
Zimkhovich BZ, Przyklenk K, Kloner RA. Role of protein kinase C as a cellular mediator of ischemic preconditioning: a critical review. Cardiovasc Res. 1998;40:9–22.
Sandhu R, Thomas U, Diaz RJ, Wilson GJ. Effect of ischemic preconditioning of the myocardium on cAMP. Circ Res. 1996;78:137–147.
Sandhu R, Diaz RJ, Mao GD, Wilson GJ. Ischemic preconditioning: differences in protection and susceptibility to blockade with single-cycle versus multicycle transient ischaemia. Circulation. 1997;96:984–995.
Seyfarth M, Richardt G, Mizsnyak A, Kurz T, Schömig A. Transient ischemia reduces norepinephrine release during sustained ischemia: neural preconditioning in the isolated rat heart. Circ Res. 1996;78:573–580.
Moolman JA, Genade S, Winterbach R, Harper IS, Williams K, Lochner A. Preconditioning with a single short episode of global ischemia in the isolated working rat heart: effect on structure, mechanical function and energy metabolism for various durations of sustained global ischemia. Cardiovasc Drugs Ther. 1995;9:103–115.
Strasser RH, Braun-Dullaeus R, Walendzik H, Marquetant R. α1-Receptor–independent activation of protein kinase C in acute myocardial ischemia: mechanisms for sensitization of the adenylyl cyclase system. Circ Res. 1992;70:1304–1312.
Lowry AD, Rosenbrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem. 1951;193:265–275.
Toombs CG, Wiltse AL, Shebuski RJ. Ischaemic preconditioning fails to limit infarct size in reserpinized rabbit myocardium. Circulation. 1993;88:2351–2358.
Lawson CS, Coltart DJ, Hearse DJ. The antiarrhythmic action of ischaemic preconditioning in rat hearts does not involve functional Gi proteins. Cardiovasc Res. 1993;27:681–687.
Banerjee A, Locke-Winter C, Rogers KB, Mitchell MB, Brew EC, Cairns C, Bensard D, Harken AH. Preconditioning against myocardial dysfunction after ischemia and reperfusion by an α1-adrenergic mechanism. Circ Res. 1993;73:656–670.
Podzuweit T, Winkelmann A, Müller A, Vogt A. Protective mechanisms, I: IP modulates catecholamine release. J Mol Cell Cardiol. 1995;27:A161. Abstract.
January B, Seibold A, Whaley B, Hipkin RW, Lin D, Schonbrunn A, Barber R, Clark RB. β2-Adrenergic receptor desensitization, internalization, and phosphorylation in response to full and partial agonists. J Biol Chem. 1997;272:23871–23879.
Clark RB, Friedman J, Dixon RA, Strader CD. Identification of a specific site required for rapid heterologous desensitization of the β-adrenergic receptor by cAMP-dependent protein kinase. Mol Pharmacol. 1989;36:343–348.
Ungerer M, Kessebohm K, Kronsbein K, Lohse MJ, Richardt G. Activation of β-adrenergic receptor kinase during myocardial ischemia. Circ Res. 1996;79:455–460.
Miyazaki T, Zipes DP. Protection against autonomic denervation following acute myocardial infarction by preconditioning ischemia. Circ Res. 1989;64:437–448.
Opie LH. The Heart: Physiology and Metabolism. 2nd ed. New York, NY: Raven Press; 1991:454.
Ferdinandy P, Csont T, Csonka C, Török M, Dux M, Nemeth J, Horvath LI, Dux L, Szilvassy Z, Jancso G. Capsaicin-sensitive local sensory innervation is involved in pacing-induced preconditioning in rat hearts: role of nitric oxide and cGRP? Naunyn Schmiedebergs Arch Pharmacol. 1997;356:356–363.
Asimakis GK, Inners-McBride K, Conti VR, Yang C. Transient β-adrenergic stimulation can precondition the rat heart against postischaemic contractile dysfunction. Cardiovasc Res. 1994;28:1726–1734.