Preconditioning enhanced glucose uptake is mediated by p38 MAP kinase not by phosphatidylinositol 3-kinase.

Ischemia is reported to stimulate glucose uptake, but the signaling pathways involved are poorly understood. Modulation of glucose transport could be important for the cardioprotective effects of brief intermittent periods of ischemia and reperfusion, termed ischemic preconditioning. Previous work indicates that preconditioning reduces production of acid and lactate during subsequent sustained ischemia, consistent with decreased glucose utilization. However, there are also data that preconditioning enhances glucose uptake. The present study examines whether preconditioning alters glucose transport and whether this is mediated by either phosphatidylinositol 3-kinase (PI3K) or p38 MAP kinase. Langendorff-perfused rat hearts were preconditioned with 4 cycles of 5 min of ischemia and 5 min of reperfusion, with glucose as substrate. During the last reflow, glucose was replaced with 5 mM acetate and 5 mM 2-deoxyglucose (2DG), and hexose transport was measured from the rate of production of 2-deoxyglucose 6-phosphate (2DG6P), using (31)P nuclear magnetic resonance. Preconditioning stimulated 2DG uptake; after 15 min of perfusion with 2DG, 2DG6P levels were 165% of initial ATP in preconditioned hearts compared with 96% in control hearts (p < 0.05). Wortmannin, an inhibitor of PI3K, did not block the preconditioning induced stimulation of 2DG6P production, but perfusion with SB202190, an inhibitor of p38 MAP kinase, did attenuate 2DG6P accumulation (111% of initial ATP, p < 0. 05 compared with preconditioned hearts). SB202190 had no effect on 2DG6P accumulation in nonpreconditioned hearts. Preconditioning stimulation of translocation of GLUT4 to the plasma membrane was not inhibited by wortmannin. The data demonstrate that ischemic preconditioning increases hexose transport and that this is mediated by p38 MAP kinase and is PI3K-independent.

Glucose transport into cells is a tightly regulated process that responds to hormonal and other stimuli to provide substrate for energy generation and for glycogen synthesis. A well studied regulator of glucose metabolism is insulin, which increases glucose transport by stimulating insertion of the glucose transporter (GLUT4) into the plasma membrane from an intracellular storage compartment (reviewed in Refs. 1 and 2). Activation of phosphatidylinositol 3-kinase (PI3K) 1 appears to be necessary for insulin-induced stimulation of glucose transport (3)(4)(5), and wortmannin, an inhibitor of PI3K, has been reported to block the increase in glucose uptake in response to insulin stimulation (3). Insulin also increases glycogen synthase activity. Overexpression of a membrane-targeted PI3K mimics the effect of insulin on glucose transport but not on glycogen synthase activity (6). Thus the effects of insulin on glucose metabolism are primarily mediated through PI3K but other signaling pathways also appear to be involved.
Glucose transport can also be enhanced by hypoxia by pathways independent of insulin and PI3K; however, the signaling pathways responsible are poorly understood (7)(8)(9)(10)(11)(12). In skeletal muscle, hypoxia and contraction have been shown to increase glucose transport and translocation of glucose transporters (13) by a mechanism that is not inhibited by wortmannin (14,15). Furthermore, contraction does not activate PKB (16) or PI3K (17). Similarly in heart, ischemia causes an increase in glucose uptake that is not inhibited by wortmannin (7). Phorbol esters also increase glucose transport by a mechanism that is not inhibited by wortmannin (8,18). Tyrosine kinase inhibitors also stimulate GLUT4 translocation by a wortmannin insensitive pathway (19); however, unlike contraction (16,17), tyrosine kinase inhibition does appear to activate PI3K, although this activation is not required for GLUT4 translocation (19). ␤-Adrenergic agonists are also reported to increase glucose transport by a wortmannin-insensitive pathway (11,12,20,21). In adipocytes, Shimizu et al. (20) have reported that norepinephrine causes an increase in glucose transport without an increase in translocation of glucose transporter but rather by increasing the activity of glucose transporters in the plasma membrane. The PI3K-independent pathway(s) responsible for glucose transporter translocation have not been identified, but nitric oxide (10) and protein kinase C (22) have been suggested to be involved. Thus multiple signaling pathways can result in glucose transporter translocation.
In addition to PI3K-independent pathways leading to increased glucose transporter translocation, there are PI3K-independent mechanisms that enhance activity of glucose transporters present in the plasma membrane (23,24). A recent study suggests that insulin-stimulated glucose transport is regulated not only by translocation of transporters, which in the case of insulin is mediated by PI3K, but also by activation of transporters after they are inserted into the plasma membrane, which requires p38 MAP kinase (23). Anisomycin, an activator of c-Jun NH 2 -terminal kinase and p38 MAP kinase, stimulates glucose transport by a mechanism that does not involve enhanced translocation of glucose transporters to the plasma membrane but rather stimulates transporters already in the plasma membrane (24).
Brief intermittent periods of ischemia and reflow, termed ischemic preconditioning, have been shown to protect the myocardium against injury produced by a subsequent sustained period of ischemia (25)(26)(27). Preconditioning has been shown to significantly reduce infarct size, arrhythmias, and postischemic contractile dysfunction (25)(26)(27)(28)(29). Intense investigation of the mechanisms responsible for the protective effects of preconditioning has revealed numerous potential mediators and downstream effectors of preconditioning, but cause and effect relationships have not been fully delineated. Preconditioning has been shown to reduce the production of lactate (25) and to attenuate the fall in intracellular pH during the sustained period of ischemia (27). These observations would suggest that glucose utilization is decreased during ischemia in preconditioned hearts. However, there also are data to suggest that preconditioning enhances glucose uptake (30,31). In the present study, we investigated whether preconditioning enhanced glucose transport during subsequent reflow periods and whether the effects of preconditioning involved either PI3K or p38 MAP kinase.
We report that preconditioning stimulates glucose uptake, but this enhanced glucose uptake is not blocked by the inhibition of PI3K with wortmannin. However, the preconditioning induced increase in glucose uptake is blocked by the addition of SB202190, an inhibitor of p38 MAP kinase, suggesting a role for p38 MAP kinase in the stress-induced stimulation of glucose transport. Preconditioning stimulates translocation of GLUT4 to the plasma membrane, but this translocation is not blocked by wortmannin.

EXPERIMENTAL PROCEDURES
Isolated Rat Heart Preparation-In this study, all rats received humane care in accordance with the Guide for the Care and Use of Laboratory Animals (53). Male Harlan Sprague-Dawley rats (170 -250 g) were anesthetized with intraperitoneal pentobarbitone (ϳ25 mg). The animals were heparinized (200 units intravenously), the heart was rapidly excised, and the aorta was cannulated. Retrograde perfusion was begun under constant pressure (90 cm of H 2 O). The nonrecirculating perfusate was a Krebs-Henseleit buffer (KH) containing: 120 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO 4 , 1.2 mM KH 2 PO 4 , 1.25 mM CaCl 2 , 25 mM NaHCO 3 , and 11 mM glucose. The buffer was maintained at pH 7.4 by bubbling with a mixture of 95% O 2 /5% CO 2 and at a temperature of 37°C.
Measurement of Glucose Transport by 2-Deoxyglucose 6-Phosphate Production by 31 P NMR-2-Deoxyglucose (2DG), an analog of glucose, is transported into the cell by the glucose transporter and is phosphorylated by hexokinase, resulting in the formation of 2-deoxyglucose 6-phosphate (2DG6P). The latter cannot leave the cell and is metabolized very slowly; hence its accumulation provides a measure of glucose uptake and phosphorylation (32). Rat hearts were perfused with phosphate-containing Krebs-Henseleit buffer using acetate (5 mM) as the sole exogenous substrate for ATP production. 2DG (5 mM) was added to the perfusate, and 2DG6P production was monitored by 31 P NMR expressed as a percentage of the initial ATP content prior to any intervention, drug treatment, or the addition of 2DG. Rat hearts were placed in a 20-mm NMR tube and bathed in perfusate for optimal magnetic susceptibility matching. 31 P NMR spectra were obtained at 161.9 MHz using a Varian Unity Plus 400 MHz wide bore NMR spectrometer. The temperature of the heart was maintained at 37 Ϯ 0.5°C by the variable temperature unit of the Varian spectrometer. We shimmed on the proton signal from the heart and routinely obtained a nonspinning line width at one-half height of ϳ0.2 ppm. Spectra were acquired every 2 min using a 2-s interval between scans with a pulse angle of 70°(33 s). The spectral width was Ϯ 3603 Hz, and 4000 data points were collected. The free induction decay was multiplied by an exponential function corresponding to a 20-Hz line broadening before Fourier transformation.
To determine whether preconditioning stimulates glucose transport and whether this stimulation is affected by inhibition of p38 MAP kinase or PI3K, nine groups of hearts were studied (Fig. 1). To assess hexose transport, 2DG6P production was measured during 20 min of perfusion with 2DG ϩ acetate. In group I (control), hearts were perfused aerobically for 20 min with glucose and then were switched to 2DG ϩ acetate buffer. Groups II (SB202190 (SB)) and III (wortmannin (WM)) were identical to group I, except 10 M SB or 100 nM WM was present prior to and at the start of perfusion with 2DG ϩ acetate. These concentrations were chosen based on those used previously (23,33). Hearts were perfused for 15 min with glucose, and then SB202190 or wortmannin was added 10 -15 min prior to perfusion with 2DG ϩ acetate and remained for an additional 5 min during perfusion with 2DG ϩ acetate. Hearts were then perfused for an additional 15 min with 2DG ϩ acetate. In group IV (preconditioning), hearts were perfused with glucose containing buffer through 4 cycles of 5 min of ischemia and 5 min of reperfusion, and during the last reflow, the perfusate was shifted to 2DG ϩ acetate. This reflow period was extended to 20 min to allow measurement of 2DG6P accumulation. Group V (PCϩSB) was identical to group IV except that 10 M SB202190 was added during the third reflow period of the preconditioning protocol (in the presence of glucose), and SB202190 was also present for the first 5 min of the final 20-min reflow period after the fourth cycle of preconditioning ischemia. Hearts were perfused with 2DG ϩ acetate for the 20-min reflow period as in group IV. Group VI (PCϩWM) was identical to group IV except 100 nM WM was added during the third reflow period of the preconditioning protocol (in the presence of glucose), and WM was also present for the first 5 min of the final 20-min reflow period after the fourth cycle of preconditioning ischemia. Group VII (PCϩWM throughout) was identical to group IV except that 100 nM wortmannin was added 5 min prior to the start of the preconditioning protocol and was present throughout 4 cycles of preconditioning ischemia and reperfusion, including the first 5 min of the final 20-min reflow period. In group VIII (insulin), hearts were perfused for 10 min with KH, and then insulin was added for 10 min followed by perfusion with 2DG ϩ acetate without insulin. Group IX (insulin ϩ WM) was identical to group VIII except that 100 nM WM was added 5 min prior to and continued during the addition of insulin.
Phospho-PKB Western Blotting-Five groups of rat hearts were snap frozen for phospho-PKB analysis at the end of the treatment period. Control hearts were frozen after 60 min of control perfusion with KH. PC hearts were frozen at the end of the fourth 5-min reflow of the preconditioning protocol. PCϩWM hearts were treated with WM throughout PC and were frozen at the end of the fourth reflow of PC. WM-treated hearts were frozen after 20 min of perfusion with KH followed by 25 min of WM perfusion. Insulin-treated hearts were frozen after 30 min of perfusion with KH followed by 15 min of perfusion with insulin (5 IU/liter).
Frozen hearts were powdered in a prechilled mortar and pestle with liquid nitrogen. Ice-cold lysis buffer (75 mM NaCl, 20 mM HEPES, 2.5 mM MgCl 2 , 0.1 mM EDTA, 0.1% Triton X-100, 10 g/ml aprotinin, 20 mM glycerophosphate, 0.5 mM dithiothreitol, 0.1 mM sodium orthovanadate, 200 g/ml phenylmethylsulfonyl fluoride, 1 mM sodium fluoride, pH 7.7) was added to the powdered tissue. The homogenate was centrifuged at 45,000 ϫ g at 4°C for 10 min. Aliquots of supernatant containing equal amounts of protein (Pierce protein assay) were boiled in sample loading buffer for 5 min before loading on 10% SDS-polyacrylamide gel and semi-dry transferred electrophoretically to nitrocellulose membranes. The nonspecific binding sites on the membrane were blocked with 5% nonfat milk in Tris-buffered saline plus Tween-20 (TBST; 20 mM Tris-HCl, 150 mM NaCl, 0.1% Tween-20, pH 7.4). The membranes were incubated in 5% bovine serum albumin in TBST containing phospho-PKB polyclonal antibodies (1:1000 dilution) for 1 h and then incubated in horseradish peroxidase-conjugated anti-rabbit IgG antibody in TBST for 1 h. Finally, the immunoreactive bands were visualized by a chemiluminescence reagent and quantified by densitometry.
p38 MAP Kinase Assay-The activity of p38 MAP kinase was measured in tissue homogenates using a commercial kit (Upstate Biotechnology). Hearts were snap frozen and homogenized as described by Bogoyevitch et al. (34). The activity of p38 MAP kinase in the homogenates was determined by measuring the p38 MAP kinase activation of MAPKAP kinase-2, which can transfer the ␥ phosphate of [␥-32 P]ATP to a specific peptide substrate. The phosphorylated substrate is then separated from the residual [␥-32 P]ATP by differential binding to P81 phosphocellulose paper. After extensive washing of the phosphocellulose paper with phosphoric acid, the bound radioactivity is determined by liquid scintillation counting. Activity was normalized to protein. Negative controls were performed without substrate peptide and with unactivated MAPKAP kinase-2. Four groups of hearts were studied. A group of control hearts was frozen after 30 min of aerobic perfusion. The PC hearts were frozen after preconditioning with four cycles of 5 min of ischemia and 5 min of reflow. The PCϩSB hearts were frozen after preconditioning with four cycles of ischemia/reperfusion with 10 M SB202190 added to the perfusate during the third reflow period and present in the perfusate throughout the 4th reflow period of the preconditioning protocol, and the SB hearts were frozen after 20 min of perfusion with KH followed by 10 min of treatment with 10 M SB202190.
GLUT4 Translocation Measurements-At the end of the treatment period, prior to ischemia, rat hearts were snap frozen and stored in liquid nitrogen. Four groups of rat hearts were included in the subcellular membrane fractionation study. Control hearts were frozen after 60 min of control perfusion. PC hearts were frozen at the end of the fourth 5-min reflow. PCϩWM hearts were treated with WM throughout PC and were frozen at the end of the fourth reflow of PC. Insulintreated hearts were frozen after 30 min of perfusion with KH followed by 15 min of perfusion with insulin (5 IU/liter).
Membrane fractionation was performed using a modification of the method of Rett et al. (35). The hearts were powdered with a prechilled mortar and pestle and dissolved in ice-cold buffer (0.2 M Tris-HCl, 10 mM EDTA, 255 mM sucrose, 0.2 mg/ml benzamidine, 10 g/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride, 100 mM sodium fluoride, 10 mM sodium pyrophosphate, 1 mM sodium orthovanadate, pH 7.4). The homogenates were centrifuged at 8700 ϫ g for 20 min. The supernatant, containing intracellular membranes, was centrifuged at 48,000 ϫ g for 15 min, and the resulting supernatant was centrifuged again for 60 min at 293,000 ϫ g.
The pellet obtained with the first centrifugation was resuspended in buffer, homogenized again in a dounce homogenizer, and centrifuged twice at 750 ϫ g and 270 ϫ g for 10 min each. The resulting supernatants from the two centrifugations were combined and centrifuged for 30 min at 10,000 ϫ g, and the resulting supernatant was centrifuged again at 48,000 ϫ g for 60 min. The pellet, containing plasma membranes, was resuspended in buffer and layered on a 38.3% sucrose gradient and was centrifuged at 80,000 ϫ g for 50 min. The interphase was collected and centrifuged again at 183,000 ϫ g for 40 min. GLUT4, in both the intracellular and plasma membranes, was measured by Western blotting. Na,K-ATPase-␣ 1 was also measured to confirm the purity and lack of contamination of the fractions. The plasma membrane fractions from the four groups of hearts showed similar levels of Na,K-ATPase.
GLUT1, GLUT4, and Na,K-ATPase Western Blotting-Protein samples were loaded on a 10% SDS-polyacrylamide gel and transferred electrophoretically to nitrocellulose membranes. The nonspecific binding sites on the membrane were blocked with 5% nonfat milk in TBST, and the membrane subsequently was incubated with antibodies against GLUT1 and GLUT4 at a dilution of 1:500 in 1% bovine serum albumin in TBST for 1 h and with Na,K-ATPase-␣ 1 antibody with a dilution of 1:5000 in 1% bovine serum albumin in TBST. The membrane was then incubated in horseradish peroxidase-conjugated anti-rabbit IgG antibody for GLUT1 and GLUT4 and in horseradish peroxidase-conjugated anit-mouse antibody for Na,K-ATPase-␣ 1 in 1% bovine serum albumin in TBST. Finally, the immunoreactive bands were visualized by a chemiluminescence reagent and quantified by densitometry.
Materials-SB202190 (Calbiochem) was dissolved in dimethyl sulfoxide (Me 2 SO) and diluted to the final concentration with perfusate immediately before use. Wortmannin (Calbiochem) was made up as a 1 mM stock in Me 2 SO and diluted in buffer immediately before use. The phospho-PKB antibody was obtained from New England Biolabs Inc., Beverly, MA, and the GLUT1 and GLUT4 antibodies were obtained from Alpha Diagnostic International, San Antonio, TX. Na,K-ATPase-␣ 1 antibody was purchased from Upstate Biotechnology, Inc., Lake Placid, NY. All other chemicals and reagents were obtained from either Sigma or Calbiochem.
Statistics-Values are expressed as means Ϯ S.E. Statistical analyses were performed by a Systat 5 program using fully factorial (M)ANOVA. When the analysis of variance demonstrated that significant differences existed, a post hoc test (Fisher) was performed. A value of p Ͻ 0.05 was considered significant.

RESULTS
To investigate directly whether brief intermittent periods of ischemia and reperfusion stimulate glucose uptake, we used 31 P NMR to monitor the uptake and phosphorylation of the glucose analog, 2DG, which is phosphorylated and accumulated as 2DG6P when perfusate glucose is replaced with 2DG and acetate. Hearts were perfused (see Fig. 1) with 2DG and acetate (the latter to provide substrate for mitochondrial respiration) under control conditions and following insulin addition or preconditioning; in the latter protocol, the glucose was replaced with 2DG and acetate during the fourth reflow of the preconditioning protocol. Fig. 2, A and B, shows typical 31 P NMR spectra demonstrating that preconditioning enhanced 2DG6P production. Fig. 2, A and B, shows resonance peaks for 2DG6P, the extracellular and intracellular inorganic phosphate (ext. P i and int. P i , respectively), phosphocreatine, and the ␥, ␣, and ␤ phosphate resonances of ATP. 2DG6P is normally not present in the heart, but on the addition of 2DG to the perfusate, the 2DG is transported into the myocytes and phosphorylated, and the 2DG6P is trapped intracellularly. The resulting accumulation of 2DG6P allows us to quantify the rate of this process as a function of time, which is manifest by an increase in the 2DG6P resonance peak height and area. We normalize the 2DG6P peak area to the area under the initial ␤-ATP resonance peak prior to any treatment and prior to addition of 2DG, and these data are presented in Fig. 2C, which summarizes data from multiple hearts. After 15 min of perfusion with 2DG and acetate, 2DG6P levels were 96 Ϯ 15% of the initial ATP in control hearts compared with 165 Ϯ 13% of the initial ATP in preconditioned hearts (p Ͻ 0.05, Fig. 2C).
We were interested in investigating the signaling pathway(s) involved in this increase in glucose transport observed in preconditioned hearts. Insulin stimulation of glucose transport is mediated by a PI3K-sensitive pathway (1,(3)(4)(5). To investigate whether PI3K is involved in the preconditioning induced stimulation of glucose transport, we added wortmannin, an inhibitor of PI3K, prior to and during preconditioning and the sub- sequent 2DG perfusion. The addition of 100 nM wortmannin, either throughout the preconditioning period (157 Ϯ 21% initial ATP) or added during the 3rd reflow period (179 Ϯ 31% initial ATP), did not block the preconditioning induced stimulation of glucose uptake. Because wortmannin had similar effects regardless of when it was added, for simplicity of presentation these data are combined in Fig. 3. To establish that we had an effective concentration of wortmannin, we tested whether 100 nM wortmannin inhibited insulin-stimulated glucose uptake. As shown in Fig. 3, insulin resulted in a large stimulation of glucose uptake; 2DG6P levels were 273 Ϯ 21% of control ATP in insulin-treated hearts. However in the presence of 100 nM wortmannin, insulin did not significantly increase 2DG6P levels (92 Ϯ 9% versus 96 Ϯ 15% for control; values are a percentage of the initial ATP). Thus, wortmannin, at a concentration that eliminated the insulin stimulated glucose uptake, had no effect on preconditioning stimulated glucose uptake.
We next investigated whether the stress-activated MAP kinase pathway might be involved in stimulating glucose uptake during preconditioning. p38 MAP kinase is reported to be stimulated by ischemia (34,36), and we confirmed that p38 MAP kinase is activated by a standard preconditioning protocol consisting of 4 cycles of 5 min of ischemia and 5 min of reperfusion (Fig. 4). Hearts were frozen after the 4th reflow, and p38 kinase activity was assessed. There is approximately a 2-fold increase in p38 MAP kinase activity in preconditioned hearts as compared with aerobic hearts frozen after a 30-min perfusion with Krebs-Henseleit buffer. We also tested whether the presence of 10 M SB202190, an inhibitor of p38 MAP kinase, would prevent p38 MAP kinase activation. As shown in Fig. 4, the addition of SB202190 completely eliminated the preconditioning time course of accumulation of 2DG6P in a control heart, and B presents data in a preconditioned heart (PC). Hearts were perfused with 2DGϩacetate as illustrated in Fig. 1. The labeled peaks are 2DG6P, extracellular inorganic phosphate (ext P i ), intracellular inorganic phosphate (int P i ), phosphocreatine (PCr), and the 3 peaks of ATP (␥, ␣, and ␤). C shows the time course of 2DG6P production averaged from four experiments for each group. Values are means Ϯ S.E. At all times from 1 min on, the PC values are significantly different than control (p Ͻ 0.05). induced increase in p38 MAP kinase activity. Treatment of aerobic hearts with SB202190 had no significant effect on p38 activity. Because p38 MAP kinase is activated by preconditioning, it is plausible that it might mediate the increase in glucose transport observed in preconditioned hearts. We therefore examined whether SB202190 would affect uptake of 2DG. Fig. 3 shows that the increase in 2DG6P accumulation in preconditioned hearts was attenuated markedly in the presence of SB202190 (111 Ϯ 20% of the initial ATP after 15 min, p Ͻ 0.05 compared with preconditioned hearts). SB202190 alone had no significant effect on 2DG6P accumulation in hearts that had not been subjected to transient ischemia. Taken together, these data suggest that brief periods of ischemia and reperfusion result in an increase in glucose transport, which persists during the reflow periods, and this increase in glucose transport can be attenuated by the p38 MAP kinase inhibitor, SB202190.
We tested whether preconditioning altered phosphorylation of PKB, a kinase directly downstream of PI3K. As shown in Fig.  5, the addition of insulin caused a 2.9-fold stimulation of phospho-PKB compared with control. Interestingly, preconditioning caused a 1.5-fold increase in PKB phosphorylation, and this increase was blocked by wortmannin. Indeed the addition of wortmannin to either control or preconditioned hearts reduced PKB phosphorylation below baseline levels (35% of control), suggesting that there is some baseline activation of PKB in control hearts.
The data show that preconditioning stimulates phosphorylation of PKB, and this phosphorylation is inhibited by wortmannin. However, the preconditioning induced stimulation of glucose transport is not inhibited by wortmannin. We therefore investigated whether preconditioning stimulated translocation of GLUT4 and whether this was inhibited by wortmannin. As shown in Fig. 6, preconditioning caused a decrease in GLUT4 in the intracellular membranes and a concomitant increase in GLUT4 at the plasma membrane; however, this translocation of GLUT4 was not inhibited by the addition of 100 nM wortmannin, a dose of wortmannin that inhibited the preconditioning induced increase in phosphorylation of PKB. The lack of inhibition of GLUT4 translocation by wortmannin is consistent with the lack of effect of wortmannin on preconditioning stimulated glucose transport. For comparison we also show the effects of insulin, which caused a decrease in GLUT4 in the intracellular membranes and a concomitant increase in GLUT4 at the plasma membrane. As shown in Fig. 7, preconditioning did not alter the cellular content of either GLUT1 or GLUT4.

DISCUSSION
As reviewed recently (1,2), there are many hormones or stimuli reported to regulate glucose transport. Insulin stimulation of glucose transport has received considerable study and the signaling pathways involved are understood in more detail than other regulators of glucose transport. Insulin binding to its receptor results in activation of the receptor tyrosine kinase, which through insulin receptor substrate phosphorylation, recruits and activates PI3K, which in turn leads to regulated insertion of the GLUT4 transporter in the plasma membrane. This mechanism is based on studies showing that inhibitors of PI3K block insulin stimulation of glucose transport (3,4) as well as studies showing that expression of a dominant negative PI3K blocks the increase in glucose transport (5). Studies also report that expression of a constitutively active PI3K results in at least a partial stimulation of glucose transport (6,37,38). In contrast, Jiang et al. (33) showed that the addition of a cellpermeable form of phosphatidylinositol 3,4,5-triphosphate did not stimulate glucose transport; however, the addition of the cell-permeable phosphatidylinositol 3,4,5-triphosphate did stimulate glucose transport when insulin was added with wortmannin. Thus, although the data show that PI3K is necessary for insulin stimulation of glucose transport, it is unclear whether activation of PI3K is sufficient. A recent report (23) suggests that insulin stimulation of glucose transport is a twostep process involving translocation of transporters to the plasma membrane followed by activation of the transporters inserted into the membrane. The signals downstream of PI3K are also under investigation. There are reports that protein kinase C isoforms (, , ⑀) are involved (39,40). There are data suggesting a role for PKB/Akt in signaling the increase in glucose transport (41)(42)(43)(44)(45), and it is also suggested that ADP  ribosylation factor-related proteins that bind PI3K might be involved in the regulation of glucose transport (46).
In contrast to our understanding of the signal transduction mechanisms involved in insulin stimulation of glucose transport, little is known about the signals involved in enhanced glucose transport by other hormones and mediators. Nitric oxide (47)(48)(49), hypoxia, ischemia (7), phorbol esters (8), adenosine (50), and ␣ (12) and ␤ (11) adrenergic agonists all stimulate glucose transport by mechanisms that are not blocked by inhibition of PI3K. In skeletal muscle, hypoxia causes an increase in translocation of glucose transporters to the plasma membrane by a mechanism that is distinct from the PI3K pathway used by insulin; hypoxia-induced translocation of GLUT4 is not blocked by wortmannin (9). There is also a recent report suggesting that wortmannin does not block ischemiainduced translocation of GLUT4 (51). However, there are no data on the heart regarding the mechanism responsible for the increase in glucose transport. We investigated the signaling pathways involved in the stimulation of glucose transport, which occurs with ischemic preconditioning. As shown in Fig. 3, SB202190, but not wortmannin, blocks the preconditioning induced increase in glucose uptake. These data suggest that p38 MAP kinase, but not PI3K, is involved in signaling the preconditioning induced increase in glucose uptake. These data are consistent with reports that p38 MAP kinase can be a mediator of increased glucose uptake (23,24,52). Thus there appear to be at least two separate mechanisms for increasing glucose transport.
We also find that preconditioning increases phosphorylation of PKB, and this increase in phosphorylation of PKB is blocked by wortmannin. However, the increased phosphorylation of PKB is not necessary for the preconditioning induced increase in glucose transport or the increase in GLUT4 translocation, as the wortmannin addition blocks the increase in PKB phosphorylation but has no effect on glucose transport or GLUT4 translocation. Taken together these data suggest that preconditioning results in an increase in PI3K, which activates PKB, but it also results in activation of p38 MAP kinase, and the activation of p38 MAP kinase is responsible for stimulation of glucose transport. Wortmannin inhibits the phosphorylation of PKB, but does not block glucose transport or GLUT4 translocation. These data are similar to a recent study (19) showing that tyrosine phosphatase inhibitors stimulate PI3K, increase phosphorylation of PKB, and also stimulate glucose uptake; however, wortmannin blocks the phosphorylation of PKB but not the stimulation of glucose transport (19).
In summary, preconditioning stimulated glucose uptake, and this stimulation was inhibited by SB202190 but not by wortmannin. Preconditioning also stimulated phosphorylation of PKB, which was blocked by wortmannin, but this inhibition of PKB phosphorylation did not block the preconditioning induced stimulation of either GLUT4 translocation or glucose uptake. The data demonstrate that preconditioning stimulates glucose uptake and that this involves p38 MAP kinase and is independent of PI3K.