Guang S. Liu

Opening of Mitochondrial KATP Channels Triggers the Preconditioned State by Generating Free Radicals

The critical time for opening mitochondrial (mito) KATP channels, putative end effectors of ischemic preconditioning (PC), was examined. In isolated rabbit hearts 29±3% of risk zone infarcted after 30 minutes of regional ischemia. Ischemic PC or 5-minute exposure to 10 &mgr;mol/L diazoxide, a mito KATP channel opener, reduced infarction to 3±1% and 8±1%, respectively. The mito KATP channel closer 5-hydroxydecanoate (200 &mgr;mol/L), bracketing either 5-minute PC ischemia or diazoxide infusion, blocked protection (24±3 and 28±6% infarction, respectively). However, 5-hydroxydecanoate starting 5 minutes before long ischemia did not affect protection. Glibenclamide (5 &mgr;mol/L), another KATP channel closer, blocked the protection by PC only when administered early. These data suggest that KATP channel opening triggers protection but is not the final step. Five minutes of diazoxide followed by a 30-minute washout still reduced infarct size (8±3%), implying memory as seen with other PC triggers. The protection by diazoxide was not blocked by 5 &mgr;mol/L chelerythrine, a protein kinase C antagonist, given either to bracket diazoxide infusion or just before the index ischemia. Bracketing preischemic exposure to diazoxide with 50 &mgr;mol/L genistein, a tyrosine kinase antagonist, did not affect infarction, but genistein blocked the protection by diazoxide when administered shortly before the index ischemia. Thus, although it is not protein kinase C-dependent, the protection by diazoxide involves tyrosine kinase. Bracketing diazoxide perfusion with N-(2-mercaptopropionyl) glycine (300 &mgr;mol/L) or Mn(III)tetrakis(4-benzoic acid) porphyrin chloride (7 &mgr;mol/L), each of which is a free radical scavenger, blocked protection, indicating that diazoxide triggers protection through free radicals. Therefore, mito KATP channels are not the end effectors of protection, but rather their opening before ischemia generates free radicals that trigger entrance into a preconditioned state and activation of kinases.

Ischemic preconditioning depends on interaction between mitochondrial KATP channels and actin cytoskeleton

Both mitochondrial ATP-sensitive K+ (KATP) channels and the actin cytoskeleton have been proposed to be end-effectors in ischemic preconditioning (PC). For evaluation of the participation of these proposed end effectors, rabbits underwent 30 min of regional ischemia and 3 h of reperfusion. PC by 5-min ischemia + 10-min reperfusion reduced infarct size by 60%. Diazoxide, a mitochondrial KATP-channel opener, administered before ischemia was protective. Protection was lost when diazoxide was given after onset of ischemia. Anisomycin, a p38/JNK activator, reduced infarct size, but protection from both diazoxide and anisomycin was abolished by 5-hydroxydecanoate (5-HD), an inhibitor of mitochondrial KATP channels. Isolated adult rabbit cardiomyocytes were subjected to simulated ischemia by centrifuging the cells into an oxygen-free pellet for 3 h. PC was induced by prior pelleting for 10 min followed by resuspension for 15 min. Osmotic fragility was assessed by adding cells to hypotonic (85 mosmol) Trypan blue. PC delayed the progressive increase in fragility seen in non-PC cells. Incubation with diazoxide or pinacidil was as protective as PC. Anisomycin reduced osmotic fragility, and this was reversed by 5-HD. Interestingly, protection by PC, diazoxide, and pinacidil could be abolished by disruption of the cytoskeleton by cytochalasin D. These data support a role for both mitochondrial KATP channels and cytoskeletal actin in protection by PC.Both mitochondrial ATP-sensitive K+(KATP) channels and the actin cytoskeleton have been proposed to be end-effectors in ischemic preconditioning (PC). For evaluation of the participation of these proposed end effectors, rabbits underwent 30 min of regional ischemia and 3 h of reperfusion. PC by 5-min ischemia + 10-min reperfusion reduced infarct size by 60%. Diazoxide, a mitochondrial KATP-channel opener, administered before ischemia was protective. Protection was lost when diazoxide was given after onset of ischemia. Anisomycin, a p38/JNK activator, reduced infarct size, but protection from both diazoxide and anisomycin was abolished by 5-hydroxydecanoate (5-HD), an inhibitor of mitochondrial KATP channels. Isolated adult rabbit cardiomyocytes were subjected to simulated ischemia by centrifuging the cells into an oxygen-free pellet for 3 h. PC was induced by prior pelleting for 10 min followed by resuspension for 15 min. Osmotic fragility was assessed by adding cells to hypotonic (85 mosmol) Trypan blue. PC delayed the progressive increase in fragility seen in non-PC cells. Incubation with diazoxide or pinacidil was as protective as PC. Anisomycin reduced osmotic fragility, and this was reversed by 5-HD. Interestingly, protection by PC, diazoxide, and pinacidil could be abolished by disruption of the cytoskeleton by cytochalasin D. These data support a role for both mitochondrial KATP channels and cytoskeletal actin in protection by PC.

Potassium channels and preconditioning of isolated rabbit cardiomyocytes: effects of glyburide and pinacidil.

Calcium tolerant rabbit cardiomyocytes, isolated by collagenase perfusion, were preincubated for varying periods of time followed by resuspension in fresh media and centrifugation into an ischaemic pellet with restricted extracellular fluid. Pellets were incubated for 240 min under oil at 37 degrees C to mimic severe ischaemia. Time to onset of ischaemic contracture (rod to square transformation) and trypan blue permeability following resuspension in 85 mOSM media were monitored at sequential times. The protocol of Series 1 was a 5-10 min pre-incubation, immediately followed by ischaemic pelleting. Preincubation with pinacidil (50 microM) protected cells from ischaemic insult, but pinacidil added only into the ischaemic pellet did not protect. Protection was abolished by the protein kinase (PKC) inhibitors chelerythrine (10 microM) added with pinacidil and calphostin C (200nM) added only into the ischaemic pellet. Neither PKC inhibitor had an effect on injury of untreated ischaemic myocytes (data not shown). Series 2-5 were preconditioning protocols with a 10 min intervention period, followed by a 30 min oxygenated drug-free period, prior to ischaemic pelleting. In series 2 pinacidil protected cells from ischaemic insult and this protection was abolished when glyburide (10 microM) was present during preincubation, or during post-incubation and ischaemia. Glyburide only partially inhibited the protection when glyburide was added only into the ischaemic pellet. In Series 3, 8-sulfophenyltheophyline (SPT)(100 microM) or adenosine deaminase during preincubation, or SPT only added into the ischaemic pellet abolished pinacidils protection. In Series 4, cardiomyocytes were ischaemically preconditioned by pelleting for 10 min followed by 30 min reoxygenation. Glyburide during initial ischaemic blocked protection, but when added during post incubation and into the final pellet protection was not reduced. In Series 5 8-cyclopentyl-1,3,dipropylxanthine (DPCPX) (10 microM) added into the final pellet abolished protection by pinacidil, but not protection following ischaemic preconditioning. In contrast to pinacidil, ischaemically preconditioned cells maintain protection in the presence of glyburide, indicating that: (1) pinacidil does not exactly mimic preconditioning and (2) ischaemically preconditioned cells do not require opened K+ATP channels for protection, although they appear to be important during initiation of the preconditioned state. It is hypothesized that pinacidil opening of K+ channels may facilitate induction of preconditioning.

Cyclosporine A limits myocardial infarct size even when administered after onset of ischemia.

OBJECTIVE The role of the immunosuppressant cyclosporine A as a preconditioning-mimetic in the rabbit heart was examined. METHODS Cyclosporine A, a potent protein 2B or calcium/calmodulin-dependent phosphatase (PP) inhibitor, was administered isolated rabbit hearts starting either 15 min prior to or 10 or 20 min after the onset of a 30 min period of regional ischemia and continuing until the onset of reperfusion. The effect of pretreatment with a second PP2B antagonist, FK-506, was also examined. In an additional protocol L-NAME was perfused for 50 min starting 5 min before the 45-min infusion of cyclosporine A. After 2 h of reperfusion infarct size was measured with triphenyltetrazolium chloride. In a second study left ventricular biopsies of isolated rabbit hearts were obtained to measure the effect of cyclosporine A on dephosphorylation of [32P] phosphorylase kinase by calcium/calmodulin-dependent phosphatases. RESULTS Pretreatment with cyclosporine A resulted in only 10.0%, infarction of the risk zone, significantly less than that in untreated control hearts (28.7%, p < 0.001) but comparable to the extent of infarction in ischemically preconditioned hearts (10.0% p < 0.001 vs. control). Equivalent protection was also observed in hearts with treatment delayed for 10 min following the onset of ischemia (10.4% infarction, p < 0.001 vs. control). However, protection waned when cyclosporine A was administered only during the last 10 min of the 30-min ischemic period (25.5% infarction, p = n.s. vs. control). Pretreatment with FK-506 also resulted in myocardial salvage (10.4% infarction, p < 0.001 vs. control). When hearts were exposed to a co-infusion of L-NAME and cyclosporine A, protection was still evident (18.1% infarction, p < 0.05 vs. L-NAME), although not as robust as that seen with the PP2B blocker alone. In hearts pretreated with cyclosporine A dephosphorylation of [32P] phosphorylase kinase by calcium/calmodulin-dependent phosphatases was inhibited by 67%. CONCLUSIONS Cyclosporine A and FK-506, potent PP2B inhibitors, can protect the ischemic rabbit heart, and at least cyclosporine A continues to be effective when infusion is delayed until after the onset of ischemia. The mechanism of this protection may be related to inhibition of phosphatases and prolongation of the phosphorylation state of ischemic cells.

CGX-1051, a peptide from conus snail venom, attenuates infarction in rabbit hearts when administered at reperfusion

CGX-1051, isolated from the venom of the marine snail Conus purpurasens, was previously noted to interact with potassium channels. Since potassium channels play an important role in cardiac physiology, we assessed the effect of CGX-1051 on infarct size in a rabbit heart model of ischemia/reperfusion. A coronary branch was occluded for 30 minutes followed by 3 hours of reperfusion in in situ and 2 hours in in vitro preparations. Infarct size was measured with triphenyltetrazolium chloride staining and expressed as a percent of the risk zone. In in situ studies, a bolus intravenous injection of CGX-1051, either 10 or 100 &mgr;g/kg, administered 5 minutes before reperfusion, reduced infarct size from 40.4 ± 2.8% of the risk zone in untreated animals to 19.8 ± 3.8% and 15.0 ± 1.9%, respectively. One &mgr;g/kg CGX-1051 was not protective. To see if the salvage was sustained, two groups of rabbits underwent 72 hours of reperfusion. The dose of 10 &mgr;g/kg infused 5 minutes before reperfusion reduced infarct size from 37.0 ± 1.6% in untreated rabbits to 15.5 ± 2.0%. When administered 10 minutes after reperfusion had begun, 100 &mgr;g/kg CGX-1051 had no effect. CGX-1051 also reduced infarct size in crystalloid-perfused, isolated rabbit hearts suggesting that protection did not depend on circulating leukocytes. The mitochondrial KATP inhibitors glibenclamide and 5-hydroxydecanoate and the MEK½, ERK and hence, inhibitor PD 98059 aborted protection from CGX-1051. These data indicate that functionally active ERK and mitochondrial KATP channels are necessary for protection. CGX-1051 caused no hemodynamic alterations at any dose tested. We conclude that CGX-1051 has a powerful anti-infarct effect when given just before reperfusion.

The mechanism of protection from 5 (N-ethyl-N-isopropyl)amiloride differs from that of ischemic preconditioning in rabbit heart

We investigated the effects of 5-(N-ethyl-N-isopropyl)amiloride (EIPA) on infarction in isolated rabbit hearts and cardiomyocytes. Thirty min of regional ischemia caused 29.6±2.8% of the risk zone to infarct in untreated Krebs buffer-perfused hearts. Treatment with EIPA (1 μM) for 20 min starting either 15 min before ischemia or 15 min after the onset of ischemia significantly reduced infarction to 5.4±2.0% and 7.0±1.0%, respectively (p<0.01 versus untreated hearts). In both cases salvage was very similar to that seen with ischemic preconditioning (PC) (7.1±1.5% infarction). Unlike the case with ischemic preconditioning, however, protection from EIPA was not blocked by 50μM polymyxin B, a PKC inhibitor, or 1μM glibenclamide, a KATP channel blocker. Forty-five min of regional ischemia caused 51.0±2.9% infarction in untreated hearts. Ischemic preconditioning reduced infarction to 23.4±3.1% (p<0.001 versus untreated hearts). In these hearts with longer periods of ischemia pretreatment with EIPA reduced infarction similarly to 28.8±2.1% (p<0.01 versus untreated hearts). However, when EIPA was combined with ischemic PC, no further reduction in infarction was seen (23.8±3.5% infarction). To further elucidate the mechanism of EIPAs cardioprotective effect, this agent was also examined in isolated rabbit cardiomyocytes. Preconditioning caused a delay of about 30 min in the progressive increase in osmotic fragility that occurs during simulated ischemia. In contrast, EIPA had no effect on the time course of ischemia-induced osmotic fragility. Furthermore, EIPA treatment did not alter the salutary effect of ischemic preconditioning when the two were combined in this model. We conclude that Na+/H+ exchange inhibition limits myocardial infarction in the isolated rabbit heart by a mechanism which is quite different from that of ischemic preconditioning. Despite the apparently divergent mechanisms, EIPAs cardioprotective effect could not be added to that of ischemic or metabolic preconditioning in these models.