And Ischemic Preconditioning–Mediated Renal Protection

J Am Soc Nephrol 12: 233–240, 2001

Protein Kinase C and Gi/o Proteins Are Involved in Adenosine- and Ischemic Preconditioning–Mediated Renal Protection

H. THOMAS LEE and CHARLES W. EMALA Department of Anesthesiology, College of Physicians and Surgeons of Columbia University, New York, New York.

ϩ Abstract. Renal ischemic reperfusion (IR) injury is a signifi- dil (K ATP channel opener), bradykinin, methacholine, or mor- cant clinical problem in anesthesia and surgery. Recently, it phine before renal ischemia. Twenty-four h later, plasma cre- was demonstrated that both renal ischemic preconditioning atinine was measured. Separate groups of rats received (IPC) and systemic adenosine pretreatment protect against pertussis toxin intraperitoneally 48 h before being subjected to renal IR injury. In cardiac IPC, pertussis toxin-sensitive G- the above protective protocols. IPC and adenosine pretreatment proteins (i.e.,Gi/o), protein kinase C (PKC), and ATP-sensitive protected against renal IR injury. Pretreatment with pertussis ϩ potassium (K ATP) channels are implicated in this protective toxin and chelerythrine abolished the protective effects of both signaling pathway. The aim of this study was to elucidate the renal IPC and adenosine. However, glibenclamide pretreatment signaling pathways that are responsible for renal protection had no effect on either renal IPC or adenosine-induced renal ϩ mediated by both IPC and adenosine pretreatment. In addition, protection, indicating no apparent role for K ATP channels. because A1 adenosine receptor antagonist failed to block renal Moreover, pinacidil, bradykinin, methacholine, and morphine IPC, whether activation of bradykinin, muscarinic, or opioid failed to protect renal function. Therefore, the conclusion is receptors can mimic renal IPC was tested because these recep- that cellular signal transduction pathways of renal IPC and tors have been implicated in cardiac IPC. Rats were acutely adenosine pretreatment in vivo involve Gi/o proteins and PKC ϩ pretreated with chelerythrine or glibenclamide, selective block- but not K ATP channels. Unlike cardiac IPC, bradykinin, mus- ϩ ers of PKC and K ATP channels, respectively, before IPC or carinic, and opioid receptors do not mediate renal IPC. adenosine pretreatment. Some rats were pretreated with pinaci-

Renal dysfunction secondary to ischemic reperfusion (IR) in- demonstrated that IPC protects renal function and morphology jury contributes to frequent and grave clinical morbidity in in rats after 45 min of ischemia and 24 h of reperfusion (5). aortovascular surgery and anesthesia (1,2). Postoperative acute Extensive studies of cardiac IPC have implicated pre-isch- renal failure implies a poor clinical prognosis and is frequently emic activation of adenosine receptors (AR), specifically, A1 associated with many other life-threatening complications, in- AR, as a predominate mechanism that mediates protection cluding sepsis and multiorgan failure (1–3). In high-risk pa- (6,7). In addition, it is hypothesized in cardiac models of IPC tients undergoing major vascular surgery and anesthesia, the that stimulation of A1 AR results in activation and transloca- mortality and morbidity rate from perioperative acute renal tion of protein kinase C (PKC) from the cytosolic to the failure has changed little during the past 30 yr; the incidence of membrane compartment, resulting in phosphorylations of un- renal dysfunction after aortovascular surgery is reported to be known cytoprotective proteins (8,9). In addition to PKC, many as high as 50% (3). studies of cardiac protection with IPC and A1 AR activation Murry (4), in 1986, first reported the protective effects of suggest that activation and opening of ATP-sensitive potas- “ischemic preconditioning (IPC)” against IR injury in cardiac ϩ sium (K ATP) channels are involved (6,10). In models of muscle by showing that multiple brief ischemic periods before ϩ cardiac protection, PKC and K ATP channel activation are a prolonged ischemic period lessened myocardial dysfunction thought to be coupled to cell surface A1 AR via pertussis and infarction size after the reperfusion period. We recently toxin-sensitive G-proteins (i.e.,Gi/o, (11,12)). Therefore, we ϩ hypothesized that Gi/o, PKC, and K ATP channels are intermediate signaling proteins involved in adenosine- and IPC-medi- Received March 30, 2000. Accepted June 21, 2000. ated renal protection from IR injury as has been demonstrated Correspondence to Dr. H. Thomas Lee, Department of Anesthesiology, Co- in the heart. lumbia Presbyterian Medical Center, P&S Box 46, 630 West 168th Street, New We recently demonstrated that systemic adenosine pretreat- York, NY 10032. Phone: 212-305-0586; Fax: 212-305-8980; E-mail: [email protected] ment protects renal function via A1 AR activation and mimics renal IPC (5). However, unlike most models of cardiac pre- 1046-6673/1202-0233 Journal of the American Society of Nephrology conditioning, renal IPC was not blocked by an A1 AR antag- Copyright © 2001 by the American Society of Nephrology onist. This suggests either that IPC- and adenosine-mediated 234 Journal of the American Society of Nephrology J Am Soc Nephrol 12: 233–240, 2001 protection follow completely different cellular signaling path- 45 min of left renal ischemia followed by reperfusion. Twenty-four h ways with a common physiologic end point or that multiple later, animals were killed and serum Cr levels were measured. endogenous agonists with common intermediate signaling ϩ pathways are involved in renal IPC. Evidence from cardiac Role of K ATP Channels in IPC- and Adenosine- preconditioning suggests that endogenously released agonists Mediated Renal Protection ϩ other than adenosine, e.g., bradykinin, acetylcholine, and opi- To determine the potential role of K ATP channels in renal IPC- or adenosine-induced renal protection, we subjected the rats to the fol- oid agonists, can also mimic IPC (13–17). ϩ The distal portion (S segment) of the proximal tubule lowing protocols after right nephrectomy: (1) high dose K ATP chan- 3 ϩ located in the outer medulla of the kidney is the primary site of nel antagonist (glibenclamide) controls (H-Glib Sham), glibenclamide (6 mg/kg intravenously) 30 min before sham operations; (2) injury in renal ischemia and reperfusion (18,19) because of its high dose glibenclamide and ischemia-reperfusion (H-GlibϩIR), glib- marginal oxygenation under normal physiologic conditions enclamide (6 mg/kg intravenously) 30 min before 45 min of left renal coupled with high basal metabolic demand (19–21). These ischemia followed by reperfusion; (3) low dose glibenclamide before renal tubular cells express the A1 AR (22) as well as the adenosine (L-GlibϩADO), glibenclamide (1 mg/kg intravenously) 30 bradykinin (23,24), muscarinic (25,26), and opioid (27,28) min before 10 min of adenosine pretreatment followed by 45 min of receptors. Therefore, the second hypothesis of the current study left renal ischemia followed by reperfusion; (4) high dose gliben- was that bradykinin, muscarinic, or opioid receptors may clamide before adenosine (H-GlibϩADO), glibenclamide (6 mg/kg intravenously) 30 min before 10 min of adenosine pretreatment fol- mimic the protection induced by A1 AR activation. lowed by 45 min of left renal ischemia followed by reperfusion; (5) high dose glibenclamide before IPC (H-GlibϩIPC), glibenclamide (6 Materials and Methods mg/kg intravenously) 30 min before IPC treatment followed by 45 General Surgical Preparation min of left renal ischemia followed by reperfusion; and (6) pinacidil All protocols were approved by the Institutional Animal Care and ϩ ␮ pretreatment group, pinacidil, a K ATP channel opener, (100 g/kg Use Committee of Columbia University. Surgical procedures have per min ϫ 10 min intravenously) 2 min before 45 min of left renal been previously described (5). Briefly, adult male Wistar rats were ischemia followed by reperfusion. Twenty-four h later, animals were anesthetized with pentobarbital intraperitoneally (45 mg/kg body wt killed and serum Cr levels were measured. or to effect) and allowed to breathe room air spontaneously. After 500 U of heparin (intraperitoneally) and maintaining the body temperature Potential Roles of Muscarinic, Bradykinin, and Opioid at 37°C, the right femoral artery and vein were cannulated with heparinized (10 U/ml) polyethylene tubing (PE-50). After a 10-min Receptors in Renal Protection stabilization period, a midline laparotomy was performed. A right To determine whether other agonists with similar signaling path- nephrectomy was performed, and the left renal artery and vein were ways in renal cells mimic renal protection afforded by adenosine or isolated. IPC, methacholine (a muscarinic agonist), bradykinin, or morphine was given to rats before ischemia and reperfusion. Rats were divided to the following groups after right nephrectomy: (1) methacholine IPC and Adenosine Pretreatment pretreatment group (MCh), methacholine (2 mg/kg, intraperitoneally) For IPC and adenosine pretreatment protocols, rats were subjected 15 min before 45 min of left renal ischemia followed by reperfusion; to the following protocols after right nephrectomy as described pre- (2) bradykinin pretreatment group (BK), bradykinin (500 ␮g/kg per viously (5): (1) control group (SHAM), isolation of left renal artery min ϫ 10 min intravenously) terminating 2 min before 45 min of left and vein only; (2) ischemia-reperfusion group (IR), 45 min of left renal ischemia followed by reperfusion; and (3) morphine pretreat- renal ischemia followed by reperfusion; (3) ischemic preconditioning ment group (Morph), morphine (5 mg/kg intraperitoneally) 15 min group (IPC), four cycles of 8 min of left renal ischemia separated by before 45 min of left renal ischemia followed by reperfusion. Twenty- 5 min of reperfusion periods before 45 min of left renal ischemia four h later, animals were killed and serum Cr levels were measured. followed by reperfusion; and (4) adenosine pretreatment group The doses of chelerythrine, glibenclamide, methacholine, bradykinin, (ADO), adenosine (1.75 mg/kg per min ϫ 10 min, intravenously) 2 and morphine were selected based on previous in vivo studies min before 45 min of left renal ischemia followed by reperfusion. (15,28–31). Twenty-four h later, animals were killed and serum creatinine (Cr) levels were measured. Role of Pertussis Toxin–Sensitive G-Proteins in Renal Protection Role of PKC in IPC- and Adenosine-Mediated To determine the potential role of pertussis toxin–sensitive G-proteins Renal Protection in renal IPC- and adenosine-induced renal protection, we initially pre- To determine the potential role of PKC in renal IPC- or adenosine- treated rats for 48 h with pertussis toxin 25 ␮g/kg intraperitoneally. Pilot induced renal protection, we subjected the rats to the following studies demonstrated that all of the rats that were pretreated intraperito- protocols after right nephrectomy: (1) PKC antagonist (chelerythrine) neally with 25 ␮g/kg pertussis toxin and subjected to 45 min of renal controls (CheϩSham), chelerythrine (5 mg/kg, intraperitoneally) 15 ischemia died within 8 h during the reperfusion period with evidence of min before sham operations; (2) chelerythrine and ischemia-reperfu- elevated body temperature and distended fluid-filled small bowel. There- sion (CheϩIR), chelerythrine 15 min before being subjected to 45 min fore, three separate and independent changes in the protocol were made of left renal ischemia followed by reperfusion; (3) chelerythrine in an attempt to enhance survival after pertussis toxin treatment and before adenosine (CheϩADO), chelerythrine 15 min before 10 min of ischemia-reperfusion: (1) The dose of pertussis toxin was reduced to 10 adenosine pretreatment followed by 45 min of left renal ischemia ␮g/kg intraperitoneally, (2) the reperfusion period was shortened to6hin followed by reperfusion; and (4) chelerythrine before IPC some rats, and (3) the ischemic period was shortened from 45 to 30 min. (CheϩIPC), chelerythrine 15 min before IPC treatment followed by Despite lowering the pertussis toxin dose to 10 ␮g/kg, none of the rats J Am Soc Nephrol 12: 233–240, 2001 Gi/o and PKC in Renal Ischemic Preconditioning 235

survived more than 10 h after 45 min of renal ischemia. The Cr measured Results at6hofreperfusion indicated that pertussis toxin pretreatment blocked Protective Effects of Renal IPC and Systemic IPC- and adenosine-mediated renal protection (see the Results section). Adenosine Pretreatment ␮ In an attempt to prolong the survival after pertussis toxin (10 g/kg Forty-five min of renal ischemia and 24 h of reperfusion (IR) intraperitoneally) and renal ischemia, the ischemic time interval was resulted in significant rises in Cr (4.6 Ϯ 0.3 mg/dl [n ϭ 9]) reduced to 30 min, resulting in a greater that 60% survival during 24 h of ϭ Ϯ reperfusion. Subsequently, the following protocols were performed: (1) compared with the sham-operated group (Cr 0.8 0.1 mg/dl ϭ Ͻ ischemia-reperfusion group with modified ischemia protocol (IR30), 30 [n 12]; P 0.01; Figure 1). IPC significantly improved min of left renal ischemia and reperfusion; (2) IPC group with modified renal function (Cr ϭ 2.5 Ϯ 0.4 mg/dl [n ϭ 9]; P Ͻ 0.05) after ischemia protocol (IPC30), four cycles of 8 min of left renal ischemia 45 min of renal ischemia and 24 h of reperfusion compared separated by 5 min of reperfusion periods before 30 min of left renal with animals that were subjected to IR injury alone (Figure 1). ischemia followed by reperfusion; (3) adenosine pretreatment group with Systemic adenosine (ADO) pretreatment also resulted in sig- modified ischemia protocol (ADO30), adenosine (1.75 mg/kg per min ϫ nificant improvements in renal function (Cr ϭ 1.7 Ϯ 0.4 mg/dl 10 min intravenously) until 2 min before 30 min of left renal ischemia [n ϭ 12]; P Ͻ 0.01; Figure 1) compared with animals that were ϩ followed by reperfusion; (4) pertussis controls (PTX SHAM), pertussis subjected to IR injury alone. toxin (10 ␮g/kg intraperitoneally) 48 h before the sham operation; (5) ϩ pertussis toxin and ischemia-reperfusion group (PTX IR30), pertussis Roles of PKC and Kϩ Channels in Renal IPC- and toxin (10 ␮g/kg intraperitoneally) 48 h before 30 min of left renal ATP ischemia followed by reperfusion; (6) pertussis toxin before IPC group Adenosine-Induced Renal Protection (PTXϩIPC30), pertussis toxin (10 ␮g/kg intraperitoneally) 48 h before Chelerythrine (Che, 5 mg/kg, intraperitoneally), a PKC an- being subjected to four cycles of 8 min of left renal ischemia separated by tagonist, given 15 min before IPC or adenosine infusion abol- 5 min of reperfusion periods before 30 min of left renal ischemia ished the renal protection induced by IPC (Cr ϭ 4.1 Ϯ 0.4 followed by reperfusion; (7) pertussis toxin before adenosine group mg/dl [n ϭ 9]) or adenosine (Cr ϭ 4.4 Ϯ 0.5 mg/dl [n ϭ 10]; (PTXϩADO30), pertussis toxin (10 ␮g/kg intraperitoneally) 48 h before Figure 1). Chelerythrine itself had no effect on renal function receiving an systemic intravenous infusion of adenosine (1.75 mg/kg per of sham-operated rats (Cr ϭ 0.7 Ϯ 0.1 mg/dl [n ϭ 4]) or of rats min ϫ 10 min) until 2 min before 30 min of left renal ischemia followed that underwent 45 min of renal ischemia and 24 h of reperfu- by reperfusion. sion (Cr ϭ 4.4 Ϯ 0.8 mg/dl [n ϭ 3]). In contrast to the effects of chelerythrine, both the low ϩ (L-Glib, 1 mg/kg) and high (H-Glib, 6 mg/kg) doses of glib- Effectiveness of K ATP Channel Blockade by ϩ Intravenous Glibenclamide In Vivo enclamide, a selective antagonist for K ATP channels, given 30 To test the effectiveness of glibenclamide in blocking Kϩ min before adenosine failed to block the renal protection by ATP ϭ Ϯ channels in vivo, we measured hemodynamic and metabolic parame- systemic adenosine pretreatment (low dose: Cr 1.8 0.5 ϩ mg/dl [n ϭ 4]; high dose: Cr ϭ 1.7 Ϯ 0.4 mg/dl [n ϭ 6]; Figure ters. Rats received 0.3 mg/kg pinacidil (K ATP channel opener) intravenously while a second group was pretreated with 6 mg/kg of 2). High-dose glibenclamide also failed to block the renal glibenclamide intravenously 30 min before receiving pinacidil (0.3 mg/kg intravenously). Maximal changes in mean arterial BP were recorded for each animal. In addition, blood glucose was measured colorimetrically by Antec Diagnostics (Farmingdale, NY) 60 min after receiving glibenclamide (6 mg/kg intravenously).

Measurement of Cr Plasma Cr levels were measured spectrophotometrically using a commercially available quantitative colorimetric assay (Sigma, St. Louis, MO).

Materials Adenosine, pertussis toxin, and methacholine were dissolved in sterile, isotonic saline. All other drugs were dissolved in 50% DMSO. Figure 1. Pretreatment with a protein kinase C (PKC) antagonist Solutions were made daily. Pentobarbital was purchased from Henry blocks adenosine- and ischemic precondition (IPC)-mediated renal Schein Veterinary Co. (Indianapolis, IN). All other drugs were ob- protection from ischemic reperfusion (IR) injury. Comparison of tained from Sigma Chemical Company. mean creatinine (Cr) measured from sham-operated (Sham, n ϭ 12), chelerythrine (5 mg/kg intraperitoneally) treatment before sham operation (CheϩSham, n ϭ 4), ischemic-reperfusion (IR, n ϭ 9), Statistical Analyses ischemic-preconditioned (IPC, n ϭ 9), adenosine-pretreated (ADO, n A one-way ANOVA was used to compare mean values across ϭ 9), chelerythrine treatment before ischemia and reperfusion multiple treatment groups with a Dunnett post hoc multiple compar- (CheϩIR, n ϭ 3), chelerythrine treatment before IPC (CheϩIPC, n ϭ ison test, e.g., SHAM versus IPC. In all cases, a probability statistic 9), and chelerythrine treatment before adenosine (CheϩADO, n ϭ 10) less than 0.05 was taken to indicate significance. All data are ex- animals. *, P Ͻ 0.05 versus sham; #, P Ͻ 0.05 versus IR. Error bars pressed throughout the text as mean Ϯ SEM. represent 1 SEM. 236 Journal of the American Society of Nephrology J Am Soc Nephrol 12: 233–240, 2001 protection by IPC (Cr ϭ 1.9 Ϯ 0.3 mg/dl [n ϭ 6]; Figure 2). Glibenclamide (6 mg/kg intravenously) given alone had no effect on renal function of sham-operated rats (Cr ϭ 1.1 Ϯ 0.1 mg/dl [n ϭ 3]) or of rats that were subjected to ischemia and reperfusion (Cr ϭ 4.0 Ϯ 0.2 mg/dl [n ϭ 4]). Moreover, pretreatment with pinacidil, a Kϩ channel opener, failed to protect renal function (Cr ϭ 4.0 Ϯ 0.3 mg/dl [n ϭ 3]).

ϩ Effectiveness of K ATP Channel Antagonism by Glibenclamide In Vivo ϩ Activation of vascular K ATP channels leads to hypotension in vivo. Figure 3 shows the hypotensive response (maximum Figure 3. Effectiveness of glibenclamide in antagonizing vascular drop in mean arterial BP ϭ 78 Ϯ 16 mmHg [n ϭ 3]) to ϩ ϩ K ATP channels as demonstrated by attenuated pinacidil-induced pinacidil (0.3 mg/kg intravenously), a K ATP channel opener. reductions in mean arterial BP (mmHg). Glibenclamide pretreatment Pretreatment with glibenclamide (6 mg/kg intravenously) 30 (H-Glib, 6 mg/kg intravenously [n ϭ 3]) significantly attenuated the min before pinacidil bolus significantly attenuated the hypo- hypotension induced by pinacidil (0.3 mg/kg intravenously [n ϭ 3]). tension (maximum drop in mean arterial BP ϭ 13 Ϯ 6 mmHg *, P Ͻ 0.05 versus pinacidil-only group. [n ϭ 3]; P Ͻ 0.05), indicating an effective in vivo blockade of ϩ vascular K ATP channels. We also measured blood glucose levels before and 60 min Pertussis Toxin Abolished Protective Effects of Renal after glibenclamide (6 mg/kg intravenously). Blockade of pan- IPC and Adenosine Pretreatment ϩ creatic K ATP channels results in an increased release of in- The bradycardic responses to adenosine and R-phenyliso- sulin and subsequent fall in blood glucose. Glibenclamide propyladenosine (R-PIA) are known to be mediated via acti- Ϯ significantly decreased the blood glucose level from 224 40 vation of A1 AR that couple to Gi/o (32,33). Forty-eight h of mg/dl (n ϭ 3) to 31 Ϯ 11 mg/dl 60 min after injection (P Ͻ either 10 or 25 ␮g/kg pertussis toxin treatment abolished the 0.01). Control rats that were not treated with glibenclamide had bradycardic effects of R-PIA and adenosine (data not shown), Ϯ ϭ significantly higher blood glucose levels (152 15 mg/dl [n indicating effective in vivo blockade of Gi/o by our pertussis 3]) 60 min after initiation of the surgical procedure (preoper- toxin pretreatment regimen. ative blood glucose ϭ 208 Ϯ 12 mg/dl [n ϭ 3]). Taken Pertussis toxin pretreatment alone had no effect on animals’ together, these physiologic effects of glibenclamide suggest an hemodynamic profile, apparent well-being, or appearance. More- ϩ effective blockade of K ATP channels in the present study. over, pertussis toxin–treated rats that underwent sham operations had similar renal function (Cr ϭ 1.1 Ϯ 0.1 mg/dl [n ϭ 2]) compared with the sham-operated controls (Cr ϭ 0.8 Ϯ 0.1 mg/dl [n ϭ 12]). However, none of the pertussis toxin–treated animals (either 10 or 25 ␮g/kg) survived more than 8 h after being subjected to 45 min of renal ischemia. Therefore, in the initial group of experiments, the reperfusion period was shortened to 6 h in both the pertussis toxin–treated and control groups. Forty-five min of renal ischemia and6hofreperfusion resulted in significant rises in Cr (2.7 Ϯ 0.2 mg/dl [n ϭ 6]), although these rises as expected were much less than those of rats that were subjected to 45 min of renal ischemia and 24 h of reperfusion (Cr ϭ 4.6 Ϯ 0.3 mg/dl [n ϭ 9]). IPC (Cr ϭ 1.4 Ϯ 0.1 mg/dl [n ϭ 6]) and R-PIA ϭ Ϯ ϭ (A1 AR agonist) pretreatment (Cr 1.9 0.1 mg/dl [n 6]) also protected renal function after 45 min of renal ischemia and the ϩ Figure 2. K ATP channels are not involved in adenosine- or IPC- shorter reperfusion period of 6 h. However, pretreatment with mediated renal protection from IR injury. Comparison of mean Cr pertussis toxin (25 ␮g/kg intraperitoneally) 48 h before renal IPC measured from sham-operated (Sham, n ϭ 12), high-dose gliben- ϭ Ϯ ϭ ϭ Ϯ (Cr 3.0 0.4 mg/dl [n 6]) and A1 AR agonist (Cr 2.7 clamide (6 mg/kg intravenously) treatment before sham operation 0.4 mg/dl, [n ϭ 6]) abolished renal protection by either A AR ϩ ϭ ϭ 1 (H-Glib Sham, n 4), ischemic-reperfusion (IR, n 9), ischemic- activation or IPC. ϭ ϭ preconditioned (IPC, n 9), adenosine-pretreated (ADO, n 9), In an attempt to enhance survival beyond 24 h, both the dose high-dose glibenclamide treatment before ischemia and reperfusion of pertussis toxin and the duration of renal ischemia were (H-GlibϩIR, n ϭ 4), high-dose glibenclamide treatment before IPC (H-GlibϩIPC, n ϭ 6), low-dose glibenclamide (1 mg/kg intrave- reduced in the next series of experiments, which resulted in nously) treatment before adenosine (L-GlibϩADO, n ϭ 6), high-dose more than 60% of the animals surviving 24 h of reperfusion. glibenclamide treatment before adenosine (H-GlibϩADO, n ϭ 6), Thirty min of renal ischemia and 24 h of reperfusion resulted and pinacidil treatment before ischemia and reperfusion (PINϩIR, n in significant rises in Cr (4.1 Ϯ 0.2 mg/dl [n ϭ 3]; Figure 4). ϭ 3) animals. *, P Ͻ 0.05 versus sham; #, P Ͻ 0.05 versus IR. IPC (Cr ϭ 1.5 Ϯ 0.2 mg/dl [n ϭ 3]) and adenosine pretreat- J Am Soc Nephrol 12: 233–240, 2001 Gi/o and PKC in Renal Ischemic Preconditioning 237

Figure 4. Pretreatment with pertussis toxin blocks adenosine- and Figure 5. Bradykinin, methacholine, or morphine did not mimic IPC-mediated renal protection from IR injury. Comparison of mean adenosine-mediated renal protection from IR injury. Comparison of ϭ Cr measured from sham-operated (Sham, n 4), 48-h pertussis toxin mean Cr measured from sham-operated (Sham, n ϭ 12), IR (IR, n ϭ ϩ ϭ treatment before sham operation (PTX Sham, n 2), ischemic- 9), adenosine-pretreated (ADO, n ϭ 9), bradykinin-pretreated (BK, ϭ ϭ reperfusion (IR30, n 3), ischemic-preconditioned (IPC30, n 3), n ϭ 6), methacholine-pretreated (MCh, n ϭ 6), and morphine- ϭ adenosine-pretreated (ADO30, n 3), 48-h pertussis toxin treatment pretreated (Morph, n ϭ 4) animals before ischemia and reperfusion. ϩ ϭ before ischemia and reperfusion (PTX IR30, n 4), 48-h pertussis *, P Ͻ 0.05 versus sham; #, P Ͻ 0.05 versus IR. toxin treatment before IPC (PTXϩIPC30, n ϭ 3), and 48 h pertussis toxin treatment before adenosine (PTXϩADO30, n ϭ 3) animals. For this series of study, all groups underwent 30 min of global renal PKC. In contrast to the findings in most models of cardiac and ischemia and 24 h of reperfusion. *, P Ͻ 0.05 versus sham; #, P Ͻ ϩ skeletal muscle preconditioning (6,10,34), K channels are 0.05 versus IR. ATP not involved in renal protection. In addition, unlike adenosine (6,7) and unlike some models of cardiac preconditioning (13– 15,35), pretreatments with bradykinin, methacholine, or mor- ment (Cr ϭ 1.3 Ϯ 0.1 mg/dl [n ϭ 3]) protected renal function phine, agonists that are known to mimic IPC in cardiac muscle, in these rats (Figure 4). The rats that were pretreated for 48 h failed to protect renal function after ischemia and reperfusion. with 10 ␮g/kg pertussis toxin intraperitoneally and then sub- The current study is the first to identify signaling intermediates jected to 30 min of renal ischemia and 24 h of reperfusion also that are responsible for renal IPC- and adenosine-induced renal exhibited significant impairments of renal function (Cr ϭ 4.4 protection. Ϯ 0.5 mg/dl [n ϭ 4]). However, pretreatment with 10 ␮g/kg The results of our studies have similarities and differences to pertussis toxin 48 h before renal IPC (Cr ϭ 5.0 Ϯ 0.3 mg/dl [n IPC studies in the heart. In our previous study (5), we were able ϭ ϭ Ϯ ϭ 3]) or systemic adenosine (Cr 5.3 0.5 [n 3]) to protect renal function with pre-ischemic A1 AR activation abolished their renal protective effects (Figure 4). These data and mimic renal IPC. However, we were unable to block the suggest that pertussis toxin–sensitive G-proteins are signaling protective effects of renal IPC with an A1 AR antagonist. We intermediates in both adenosine- and IPC-mediated protection hypothesized either that IPC- and adenosine-induced protec- of renal IR injury. tion in the kidney follows completely different cellular signaling pathways or that multiple endogenous agonists are in- Methacholine, Bradykinin, and Morphine Failed to volved in renal IPC. This second hypothesis led us to test Protect Renal Function whether activation of other endogenous receptors such as mus- Intravenous bradykinin at 500 ␮g/kg per min caused a carinic, bradykinin, or opioid receptors protected renal function similar degree of hypotension (systolic BP, approximately 60 against IR injury. In some studies of cardiac IR injury, other to 70 mmHg) as observed with intravenous adenosine (5). receptors that demonstrate common intracellular signaling in- Systemic methacholine caused transient reduction in BP but termediates with adenosine, such as bradykinin (14,15), ace- was associated with markedly increased salivary and lacrimal tylcholine (13), phenylephrine (36), and opioids (35), mimic secretions. Pretreatments with systemic methacholine (Cr ϭ IPC. That multiple agonists (bradykinin, morphine, acetylcho- 3.8 Ϯ 0.1 mg/dl [n ϭ 6]), bradykinin (Cr ϭ 4.6 Ϯ 0.1 mg/dl line, and phenylephrine) are able to induce cardiac precondi- ϭ ϭ Ϯ ϭ [n 5]), or morphine (Cr 4.3 0.2 mg/dl [n 4]) failed tioning suggested to us that A1 AR activation per se may not be to protect renal function (Figure 5). This is in contrast to the only agonist inducing protection but that any agonist that protections obtained with either 10 min of systemic adenosine stimulates common second messenger signaling intermediates pretreatment or with renal IPC (5). (e.g.,Gi/o and PKC), may also protect the heart against IR injury. However, unlike studies in the heart, we did not find Discussion that activation of muscarinic, bradykinin, or opioid receptors

The current in vivo study in the rat kidney demonstrates that mimicked renal IPC, suggesting that activation of Gi/o alone renal IPC- and adenosine-mediated renal protection from IR was not sufficient to mediate renal protection. Alternatively, it injury involve both pertussis toxin–sensitive G-proteins and is possible that the rat renal cells that are protected by IPC and 238 Journal of the American Society of Nephrology J Am Soc Nephrol 12: 233–240, 2001

ϩ adenosine pretreatment may not express muscarinic, bradyki- K ATP channels are present in various cell types in the nin, or opioid receptors that couple to Gi/o. kidney, including the proximal tubule, the thick ascending limb The current study demonstrated that PKC plays a role in of Henle’s loop, and the cortical collecting duct (43). Under renal IPC- and adenosine-induced renal protection in vivo.We physiologic conditions, these channels have a high open prob- used chelerythrine, a highly selective and specific PKC antag- ability and function to regulate renal blood flow, reabsorption ϭ ␮ ϩ onist that inhibits the PKC catalytic domain (Ki 0.7 M of electrolytes and solutes, renin release, K secretion, and (37)), to block the physiologic effects of PKC to determine diuresis (43–45). In the heart and skeletal muscle, gliben- ϩ whether PKC activation is required for renal IPC- and ade- clamide, a selective K ATP channel blocker, at doses ranging nosine-induced renal protection. It is accepted that PKC plays from 0.3 mg/kg to 3 mg/kg blocked the protective effects of ϩ a critical role in mediating IPC- and A1 AR-mediated cardiac IPC and adenosine (31,34,46). Moreover, the K ATP channels protection (8,30). PKC modulates both short- and long-term play a role to protect the brain against IR injury (47). These cellular responses after IR injury, such as ion channel regula- data suggest that IPC in these tissues may be mediated by the ϩ tion, new protein synthesis, and cellular proliferation. Coupling activation of K ATP channels coupled to A1 AR and that these of A1 AR to PKC via Gi/o in renal tubules has been confirmed channels may serve an endogenous protective role. However, previously (38–40). Currently, it is not conclusively known the current study does not support the hypothesis that renal which subtypes of PKC are involved in mediating cardiac protection by IPC or adenosine pretreatment involves the ac- ␦ ⑀ ϩ preconditioning, although evidence exists that the and/or tivation of K ATP channels. Neither the high (6 mg/kg) nor the isoforms may be involved in the heart (8). low (1 mg/kg) doses of glibenclamide abolished the renal

A1 AR, including those present in the kidney, couple to protection afforded by adenosine or IPC. Moreover, pinacidil, ϩ intracellular effectors via Gi/o (pertussis toxin–sensitive G- aK ATP channel activator, failed to protect renal function. Our proteins (22,32)). In cardiac IPC from a number of species hemodynamic and blood glucose data show that 6 mg/kg ϩ including the rat, Gi/o are intermediates in modulating ade- glibenclamide intravenously effectively blocked the K ATP nosine’s protective effect (11). Moreover, cardiac precondi- channels in vivo. This finding makes the kidney unique com- ϩ tioning in vivo is abolished after the pertussis toxin treatment pared with the brain, heart, and skeletal muscle, where K ATP (12,41). Therefore, we hypothesized that by blocking Gi/o with channels are known to play an important role in tissue protec- ϩ pertussis toxin, we may prevent the renal protective effects of tion. The role of K ATP channels in excitable tissues such as IPC and adenosine. The doses of pertussis toxin used in this brain, heart, and skeletal muscle seems to be different from the ␮ ϩ study (25 and 10 g/kg) were based on previous studies of role of K ATP channels in cells of epithelial origin, e.g., renal cardiac IPC in rats (11,12,41). Pertussis toxin treatment for tubules, in terms of cytoprotection. 48 h abolished the protective effects of renal IPC and adeno- In summary, we extended our in vivo studies in the rat to

sine pretreatment, supporting a role for Gi/o proteins as inter- show that both Gi/o and PKC are signaling intermediates in- mediates in renal IPC- and adenosine-induced renal protection volved in renal protection induced by either IPC or adenosine from IR injury. Although we did not measure the degree of pretreatment. Unlike the findings in cardiac models of protec- ϩ adenosine diphosphate ribosylation of Gi/o in this study, that A1 tion from IR injury, K ATP channels are not signaling inter- AR-mediated bradycardic responses to R-PIA and adenosine mediates in renal protection. Moreover, bradykinin, methacho-

were abolished in pertussis toxin–treated rats suggests that A1 line, or opioid receptor activation does not induce renal AR-Gi/o coupling was blocked with pertussis toxin treatment. protection from IR injury. These studies offer a mechanistic Endoh et al. (42) also demonstrated that pertussis toxin at insight into the signaling intermediates that are responsible for doses ranging from 1.25 to 10 ␮g/kg dose-dependently atten- adenosine-induced and IPC-induced renal protection from IR uated the negative inotropic and chronotropic effects of atrial injury. muscarinic receptor activity; the pertussis toxin dose of 10 ␮ g/kg maximally inhibited the receptor-Gi/o interactions. Acknowledgments That pertussis toxin treatment associated with in vivo renal This work was funded in part by intramural grant support from the ischemia led to significant mortality and morbidity is a limi- Department of Anesthesiology, College of Physicians and Surgeons of tation in our study. None of the rats survived for 24 h after 45 the Columbia University. min of global renal ischemia with either 10 or 25 ␮g/kg of pertussis toxin. We were able to improve the survival rate by reducing the dose of pertussis toxin to 10 ␮g/kg and the References ischemic time to 30 min. With this reduced dose of pertussis 1. Maher ER, Robinson KN, Scoble JE, Farrimond JG, Browne DR, toxin and reduced ischemic interval, we enhanced the survival Sweeny P, Morehead JF: Prognosis of critically-ill patients with at 24 h of reperfusion and again showed that pertussis toxin acute renal failure: APACHE II score and other predictive fac- tors. Q J Med 72: 857–866, 1989 pretreatment blocked both adenosine- and IPC-mediated pro- 2. McCarthy JT: Prognosis of patients with acute renal failure in the tection from renal IR injury. It is unclear why the combination intensive care unit: A tail of two eras. Mayo Clin Proc 71: of pertussis toxin and renal ischemia in vivo is detrimental to 117–126, 1996 rat survival. Sham-operated rats that received 10 to 25 ␮g/kg of 3. Aronson S, Blumenthal R: Perioperative renal dysfunction and pertussis toxin seemed normal before and after the surgical cardiovascular anesthesia: Concerns and controversies. J Cardio- procedures. thorac Vasc Anesth 17: 117–130, 1998 J Am Soc Nephrol 12: 233–240, 2001 Gi/o and PKC in Renal Ischemic Preconditioning 239

4. Murry CE, Jennings RB, Reimer KA: Preconditioning with isch- proliferation in mesangial cells. Immunopharmacology 45: 57– emia. A delay of lethal cellular injury in ischemic myocardium. 62, 1999 Circulation 74: 1124–1136, 1986 25. Paulmichl M, Woll E, Lang F: Pertussis toxin-dependent and 5. Lee HT, Emala CW: Protective effects of renal ischemic precon- -independent hormonal effects on cultured renal epithelioid cells. ditioning and adenosine pretreatment: Role of A(1) and A(3) FEBS Lett 234: 263–266, 1988 receptors. Am J Physiol Renal Physiol 278: F380–F387, 2000 26. Mohuczy-Dominiak D, Garg LC: Muscarinic receptors in 6. Auchampach JA, Gross GJ: Adenosine A1 receptors, KATP chan- MDCK cells are coupled to multiple messenger systems. Am J nels, and ischemic preconditioning in dogs. Am J Physiol 264: Physiol 263: C1289–C1294, 1992 H1321–H1336, 1993 27. Wittert G, Hope P, Pyle D: Tissue distribution of opioid receptor 7. Thornton JD, Liu GS, Olsson RA, Downey JM: Intravenous gene expression in the rat. Biochem Biophys Res Commun 218: pretreatment with A1-selective adenosine analogues protects the 877–881, 1996 heart against infarction. Circ Res 85: 659–665, 1992 28. el Awady ES, Walker LA: Effects of morphine on the renal 8. Kawamura S, Yoshida KI, Miura T, Mizukami Y, Matsuzaki M: handling of sodium and lithium in conscious rats. J Pharmacol Ischemic preconditioning translocates PKC-␦ and -⑀ which me- Exp Ther 254: 957–961, 1990 diate functional protection in isolated heart. Am J Physiol 275: 29. Ping P, Zhang J, Cao X, Li RC, Kong D, Tang XL, Qiu Y, H2266–H2271, 1998 Manchikalapudi S, Auchampach JA, Black RG, Bolli R: PKC- 9. Downey JM, Liu GS, Thornton JD: Adenosine and the anti- dependent activation of p44/p42 MAPKs during myocardial infarct action of preconditioning. Cardiovasc Res 27: 3–8, 1993 ischemia-reperfusion in conscious rabbits. Am J Physiol 276: 10. Van Winkle DM, Chien GL, Wolff RA, Soifer BE, Kuzume K, Davis RF: Cardioprotection provided by adenosine receptor ac- H1468–H1481, 1999 30. Fryer RM, Schultz JE, Hsu AK, Gross GJ: Importance of PKC tivation is abolished by blockade of the KATP channel. Am J Physiol 266: H829–H839, 1994 and tyrosine kinase in single or multiple cycles of preconditioning in rat hearts. Am J Physiol 276: H1229–H1235, 1999 11. Lasley RD, Mentzer RM: Pertussis toxin blocks adenosine A1 receptor mediated protection of the ischemic rat heart. J Mol Cell 31. Schultz JE, Yao Z, Cavero I, Gross GJ: Glibenclamide-induced Cardiol 25: 815–821, 1993 blockade of ischemic preconditioning is time dependent in intact 12. Schultz JE, Hsu AK, Barbieri JT, Li PL, Gross GJ: Pertussis rat heart. Am J Physiol 272: H2607–H2615, 1997 toxin abolishes the cardioprotective effect of ischemic precondi- 32. Stiles GL: Adenosine receptors. J Biol Chem 267: 6451–6454, tioning in intact rat heart. Am J Physiol 275: H495–H500, 1998 1992 13. Yao Z, Gross GJ: Acetylcholine mimics ischemic precondition- 33. Ungerer M, Stocker M, Richardt G: A1 adenosine receptors and ing via a glibenclamide-sensitive mechanism in dogs. Am J muscarinic cholinoceptors in myocardial ischemia. Naunyn Physiol 264: H2221–H2225, 1993 Schmiedebergs Arch Pharmacol 354: 44–52, 1996 14. Wall TM, Sheehy R, Hartman JC: Role of bradykinin in myocardial 34. Pang CY, Neligan P, Xu H, He W, Zhong A, Hopper R, Forrest preconditioning. J Pharmacol Exp Ther 270: 681–689, 1994 CR: Role of ATP-sensitive Kϩ channels in ischemic precondi- 15. Goto M, Liu Y, Yang XM, Ardell JL, Cohen MV, Downey JM: tioning of skeletal muscle against infarction. Am J Physiol 273: Role of bradykinin in protection of ischemic preconditioning in H44–H51, 1997 rabbit hearts. Circ Res 77: 611–621, 1995 35. Schultz JE, Hsu AK, Gross GJ: Morphine mimics the cardiopro- 16. Schultz Je, Hsu AK, Nagase H, Gross GJ: TAN-67, a delta 1-opioid tective effect of ischemic preconditioning via glibenclamide- receptor agonist, reduces infarct size via activation of Gi/o proteins sensitive mechanism in the rat heart. Circ Res 78: 1100–1104, ϩ and K ATP channels. Am J Physiol 274: H909–H914, 1998 1996 17. Schultz JE, Hsu AK, Gross GJ: Ischemic preconditioning in the 36. Tsuchida A, Liu Y, Liu GS, Cohen MV, Downey JM: alpha intact rat heart is mediated by delta1- but not mu- or kappa- 1-adrenergic agonists precondition rabbit ischemic myocardium opioid receptors. Circulation 97: 1282–1289, 1998 independent of adenosine by direct activation of protein kinase 18. Venkatachalam MA, Bernard DB, Donohoe JF, Levinsky NG: Isch- C. Circ Res 75: 576–585, 1994 emic damage and repair in the rat proximal tubule: Differences among 37. Lasley RD, Noble MA, Mentzer RM: Effects of protein kinase C the S1, S2, and S3 segments. Kidney Int 14: 31–49, 1978 inhibitors in in situ and isolated ischemic rabbit myocardium. 19. Lieberthal W, Nigam SK: Acute renal failure. I. Relative impor- J Mol Cell Cardiol 29: 3345–3356, 1997 tance of proximal vs. distal tubular injury. Am J Physiol 275: 38. Coulson R, Proch PS, Olsson RA, Chalfant CE, Cooper DR: F623–F631, 1998 Upregulated renal adenosine A1 receptors augment PKC and 20. Dinour D, Agmon Y, Brezis M: Adenosine: An emerging role in the control of renal medullary oxygenation? Exp Nephrol 1: glucose transport but inhibit proliferation. Am J Physiol 270: 152–157, 1993 F263–F274, 1996 21. Agmon Y, Dinour D, Brezis M: Disparate effects of adenosine 39. Schwiebert EM, Karlson KH, Friedman PA, Dietl P, Spielman A1- and A2-receptor agonists on intrarenal blood flow. Am J WS, Stanton BA: Adenosine regulates a chloride channel via Physiol 265: F802–F806, 1993 protein kinase C and a G protein in a rabbit cortical collecting 22. Spielman WS, Arend LJ: Adenosine receptors and signalling in duct cell line. J Clin Invest 89: 834–841, 1992 the kidney. Hypertension 17: 117–130, 1991 40. Burnatowska-Hledin MA, Spielman WS: Effects of adenosine on 23. Schanstra JP, Alric C, Marin-Castano ME, Girolami J, Bascands cAMP production and cytosolic Ca2ϩ in cultured rabbit medul- J: Renal bradykinin receptors: Localisation, transduction path- lary thick limb cells. Am J Physiol 260: C143–C150, 1991 ways and molecular basis for a possible pathological role [Re- 41. Thornton JD, Liu GS, Downey JM: Pretreatment with pertussis view]. Int J Mol Med 3: 185–191, 1999 toxin blocks the protective effects of preconditioning: Evidence 24. Alric C, Pecher C, Bascands JL, Girolami JP: Effect of brady- for a G-protein mechanism. J Mol Cell Cardiol 25: 311–320, kinin on tyrosine kinase and phosphatase activities and cell 1993 240 Journal of the American Society of Nephrology J Am Soc Nephrol 12: 233–240, 2001

42. Endoh M, Maruyama M, Iijima T: Attenuation of muscarinic Glyburide on renal tubule transport and potassium-channel ac- cholinergic inhibition by islet-activating protein in the heart. tivity. Renal Physiol Biochem 18: 169–182, 1995 Am J Physiol 249: H309–H320, 1985 46. Toombs CF, Moore TL, Shebuski RJ: Limitation of infarct size 43. Quast U: ATP-sensitive Kϩ channels in the kidney. Naunyn in the rabbit by ischaemic preconditioning is reversible with Schmiedebergs Arch Pharmacol 354: 213–225, 1996 glibenclamide. Cardiovasc Res 27: 617–622, 1993 44. Bailey MA, Walter SJ: Renal effects of glibenclamide: A mi- 47. Perez-Pinzon MA, Born JG: Rapid preconditioning neuroprotection cropuncture study. J Pharmacol Exp Ther 285: 464–467, 1998 following anoxia in hippocampal slices: Role of the Kϩ ATP 45. Wang T, Wang WH, Robbenhaar GK, Giebisch G: Effects of channel and protein kinase C. Neuroscience 89: 453–459, 1999