Insulin increases glycolysis without further

Clinical Science (2004) 107, 213–220 (Printed in Great Britain)

Insulin increases glycolysis without further vasodilation in porcine coronary arteries exposed to hypoxia Ole FRØBERT∗ †, Jens P. BAGGER‡, Ulf SIMONSEN†, Sten LUND§ and Claus H. GRAVHOLT§ ∗

Department of Cardiology S, Aalborg University Hospital, DK-900 Aalborg, Denmark, †Department of Pharmacology, University of Aarhus, DK-8000 Aarhus C, Denmark, ‡Imperial College School of Medicine, Department of Cardiology, Hammersmith Hospital, London W12 0HS, U.K., and §Department of Endocrinology M, University Hospital Aarhus, DK-8000 Aarhus C, Denmark

A

B

S

T

R

A

C

T

In acute ischaemia, glucose–insulin–potassium administration reduces mortality and β-adrenoceptor antagonists have favourable effects on the outcome of ischaemic heart disease. The present study was designed to investigate whether insulin (1.4 × 10−7 M) and the β-adrenoceptor antagonist, propranolol (10−5 M), increase hypoxic vasodilation in correspondence with changes in glycolysis. Porcine coronary arteries, precontracted with 10−5 M prostaglandin F2α , were mounted in a pressure myograph and a microdialysis catheter was inserted in the tunica media. Hypoxic vasodilation, interstitial lactate/pyruvate ratio and interstitial glucose were measured at low (2 mM) and high (20 mM) glucose concentrations. Hypoxia (60 min) caused vasodilation and doubled the lactate/pyruvate ratio. Treatment with insulin quadrupled the lactate/pyruvate ratio during hypoxia, but did not change hypoxic vasodilation. Propranolol blocked isoprenaline-evoked vasodilation, but hypoxic increases in lactate/pyruvate ratio and vasodilation did not change. The combination of insulin and propranolol did not cause further changes compared with each drug added alone, although the combination increased vasoconstriction during reoxygenation. Interstitial glucose fell during hypoxia at an organ bath glucose concentration of 2 mM, and rose at a glucose concentration of 20 mM. Addition of insulin and propranolol alone or in combination had no effect on interstitial glucose concentration. Accordingly, arteries were found to contain only minute amounts of the glucose transporter isoform GLUT4. Our findings suggest that insulin increases arterial glycolysis, but treatment with insulin, propranolol, or both, is not associated with enhanced coronary vasodilation during hypoxia.

INTRODUCTION The administration of GIK (glucose–insulin–potassium) in myocardial infarction has been reported to reduce mortality [1,2], and the favourable effects of acute and longterm administration of β-adrenoceptor antagonists on

outcome are well established [3,4]. Hypoxia, insulin and β-adrenoceptor antagonists induce a shift from lipid to carbohydrate utilization in the myocardium [5]. Hypoxia stimulates translocation of GLUT4 to the cellular membrane in rat hearts [6] and forms NADH2 that inhibits pyruvate dehydrogenase [7]. In the anoxic

Key words: coronary artery, glucose, hypoxia, insulin, vasodilation. Abbreviations: GIK, glucose–insulin–potassium; LAD, left anterior descending coronary artery; MLC, myosin light chain; PGF2α , prostaglandin F2α ; PCr, phosphocreatine; Po2 , partial pressure of O2 ; PSS, physiological saline solution; RR, exchange fraction. Correspondence: Dr Ole Frøbert (email [email protected]).

C

2004 The Biochemical Society

213

214

O. Frøbert and others

myocardium, the rate of conversion of glycogen into glucose and lactate helps the survival of the anoxic region, and the administration of glucose plus insulin could be expected to promote this process [5]. In addition to the improvement of myocardial function, treatment with insulin and β-adrenoceptor antagonists could have beneficial effects on the coronary circulation. It has been suggested [8,9] that at least two functionally independent glycolytic pathways exist in porcine coronary smooth muscle, aerobic glycolysis related to Na+ /K+ -ATPase and oxidative metabolism in response to increases in smooth muscle tension. The latter may be influenced by hypoxia and insulin, which both have been described to evoke coronary arterial vasodilation [10,11]. Hypoxia-induced vasodilation is endotheliumindependent in systemic conduit arteries and several causal mechanisms have been suggested (for a comprehensive review, see [11]): release of adenosine [12,13], opening of ATP-sensitive K+ channels [13,14], increased intracellular pH [15], decreased Ca2+ sensitivity [16] and decreased energy state [17]. Insulin is an endotheliumdependent vasodilator in canine conduit coronary arteries [18] and a constrictor in endothelium-denuded rat arterioles [19]. Insulin causes endothelium-independent vasorelaxation in porcine epicardial arteries contracted with endothelin-1 [20]. β-Adrenoceptor antagonists inhibit vasodilatation in vitro [21] and forearm vascular conductance response to hypoxia is reduced after addition of propranolol [22]. According to the literature available, addition of insulin, propranolol or both during hypoxia could either: (i) reduce systemic vascular dilation through ‘improved’ metabolism securing a higher level of ATP or PCr (phosphocreatine) [17]; or (ii) increase vascular dilation (in line with findings of reduced ischaemic myocardial contracture following perfusion with high glucose/insulin [23] and improved glycaemic status [5] that increases hypoxic vasodilation as found previously in rabbit aorta [24]). We hypothesized that insulin and β-adrenoceptor antagonism increase coronary artery smooth muscle glycolysis and enhance vasodilation further during hypoxia. In order to test the hypothesis, we measured coronary artery diameters while sampling interstitial metabolites by means of microdialysis in the arterial media layer. Information on the concentration of both lactate and its redox partner pyruvate is likely to provide more information on the metabolic state of the coronary artery than are measurements of lactate alone. Therefore we measured interstitial lactate/pyruvate ratios and glucose concentration in relation to changes in artery diameter during oxygenated and hypoxic conditions. The response to insulin was validated by measurement of the glucose transporter isoform GLUT4, and the response to βadrenoceptor antagonism was validated by addition of isoprenaline. C


Figure 1 Schematic representation of the experimental set-up

A cylindrical segment of porcine LAD was mounted in a vessel chamber. The internal pressure was controlled by adjustment of two fluid-filled reservoirs mounted on a pressure column and the external diameter was automatically determined by video imaging. A microdialysis catheter was placed in the smooth muscle interstitium of the coronary artery. Glucose, lactate and pyruvate concentrations in the dialysate were measured by an automated spectrophotometric kinetic enzymic analyser.

MATERIALS AND METHODS Tissue Hearts from 70–90 kg pigs were obtained shortly after slaughter [the investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996)]. The aorta was cannulated and the coronary circulation was perfused with 100 ml of PSS [physiological saline solution; 119 mM NaCl, 25 mM NaHCO3 , 4.7 mM KCl, 1.2 mM MgSO4 , 1.5 mM CaCl2 and 20 mM Hepes (pH 7.4) and 2 or 20 mM glucose depending on protocol] bubbled with 5 % CO2 /95 % O2 . The hearts were bathed in PSS at 5 ◦ C for approx. 2 h before the start of the experiment. The LAD (left anterior descending coronary artery) was carefully dissected and the proximal 3–4 cm of the artery left intact. The arterial segments usually had one to three branches in the proximal part. The branches were ligated with 5–0 silk sutures during dissection in cold PSS. The anterior surface of the LAD was marked in both ends with India Ink and the in situ length of the segment was recorded.

Pressure myograph Cylindrical arterial segments (1.5–2 cm) were mounted at both ends on stainless steel cannulae and fastened with sutures in a 20 ml vessel chamber (Figure 1). The arterial segments were bathed in PSS continuously bubbled with 5 % CO2 in O2 [Po2 (partial pressure of O2 ) > 650 mmHg; ISO2-D oxygen analyser; WPI, Sarasota, FL, U.S.A.] and the temperature was raised

Insulin increases glycolysis

gradually to 37 ◦ C. The internal pressure was controlled by adjustment of two reservoirs containing PSS mounted on a pressure column and connected to the cannulae. Pressure transducers close to the ‘arterial’ end of each cannula continuously measured the internal pressure. The segments were stretched to the in situ length by operating a micrometer device on one of the cannulae and a pressure of 40 mmHg was applied to the vessel during a stabilizing period of 1 h. We used this intravascular pressure because our previous experiments [25] demonstrated this to be optimal in terms of coronary arterial response to an agonist in a pressure myograph. The external diameter of the arterial segment was automatically determined by video imaging at a frequency of 20 Hz. Internal arterial pressure, the outer diameter and a video image of the arterial segment were monitored continuously and stored on computer (Vessel View software version 1.2; Danish Myo Technology, Aarhus, Denmark). Before the start of the experiments, the video dimension analyser was calibrated with a 3000 × 3000 µm phantom image in the horizontal and vertical directions. At the end of the experiments, the calibration was verified by repeated measurements of the phantom.

Microdialysis and calculations Microdialysis catheters (CMA/7 microdialysis probe; CMA, Stockholm, Sweden) were placed in the smooth muscle interstitium of the coronary artery mounted in the pressure myograph. The microdialysis catheters have a 6 kDa molecular-mass cut-off, an outer diameter of 0.24 mm and a membrane length of 2 mm. Perfusion of the catheter was started immediately at a rate of 0.3 µl/min with a physiological perfusate (147 mM Na+ , 4 mM K+ , 2.3 mM Ca2+ and 156 mM Cl− ; Perfusion fluid T1; CMA). The catheters were perfused for 60 min prior to the experiments. The effluent from the microdialysis catheters (dialysate) was collected in consecutive 30 min fractions (9 µl) starting before induction of hypoxia. There was a delay of 12 min between the passage of the perfusate through the microdialysis catheter and collection in vials. This time delay was considered in the calculations and all data are presented in real time. Glucose, lactate and pyruvate concentrations in the dialysate were measured by an automated spectrophotometric kinetic enzymic analyser (CMA 600). To determine the exchange fraction (RR) of lactate over the microdialysis membrane, a small amount of [14 C]lactate (0.95 µCi/ml; 3.51 pmol/l) was included in the perfusate as an internal reference marker. RR was measured in 11 animals and calculated as (perfusated.p.m. − dialysated.p.m. )/perfusated.p.m. , where perfusated.p.m. and dialysated.p.m. are the radioactivity in the perfusate and dialysate respectively, measured in d.p.m., and interstitial lactate concentration was calculated as [lactate]dialysate /RR, where [lactate]dialysate is the concen-

tration of lactate in the dialysate. The mean RR increased throughout the study period [0.29 + − 0.04 (before hypoxia), 0.37 + 0.04 (0–30 min hypoxia), 0.37 + − − 0.05 (30–60 min hypoxia) and 0.48 + 0.05 (reoxygenation − period)]. The values of RR for these time periods differed significantly from each other (P < 0.01, as determined by repeated measures ANOVA). The mean RR values for the different time periods were used for further calculations of lactate concentration.

Procedures During a stabilizing period of 1–2 h, contraction of the coronary arteries was induced by addition of KPSS (PSS in which NaCl was exchanged for KCl on a equimolar basis to a final KCl concentration of 125 mM) until a reproducible response was obtained (at least twice). After an additional stabilizing period with PSS in the organ bath, arterial contraction was induced with PGF2α (prostaglandin F2α ; 10−5 M; Dinoprost, Upjohn, Germany). On a random basis, the glucose concentration was changed to either 2 or 20 mM and, depending on the protocol, insulin (1.4 × 10−7 M; Novo Nordisk, Bagsværd, Denmark) and propranolol (10−5 M; Sigma, St Louis, MO, U.S.A.) alone, a combination of insulin and propranolol or none of these substances (i.e. control situation) were added to the organ bath. In order to examine the segments under definite β-adrenoceptor activation, the β-agonist isoprenaline (10−6 M; Sigma) alone or isoprenaline and propranolol were added to the organ bath in some experiments. All organ bath substance changes were also applied to the coronary artery lumen. When a stable contraction had been established, + hypoxia was induced by 5 % CO2 /95 % N2 (Po2 = 35 − 5 mmHg). Simultaneously, PSS in the lumen was exchanged with deoxygenated PSS. Oxygenation was reestablished after 60 min of hypoxia by washing with PSS aerated with 5 % CO2 /95 % O2 and the arterial response during the following 45 min was recorded (reoxygenation period). pH in the organ bath and the dialysate was monitored with a small pH electrode (MI410; Microelectrodes. Bedford, NH, U.S.A.) connected to a pH meter (PHM210; Radiometer, Copenhagen, Denmark) and remained stable at pH 7.4 throughout the study period.

GLUT4 measurement In some experiments, coronary artery segments were snap-frozen in liquid nitrogen directly after removal and were used to determine total crude membranes GLUT4 protein content (microsomes and plasma membranes) as described previously [26]. For comparison, samples of myocardium, striated muscle (pectoralis major) and fat from the chest were collected from the same pigs. The tissue was homogenized carefully in ice-cold HES buffer [20 mM Hepes, 5 mM EDTA and 255 mM sucrose (pH 7.2)] containing proteinase inhibitors {1 µg/ml each C


215

216


of antipain, aprotinin, pepstatin and leupeptin and 100 µmol/l AEBSF [4-(2-aminoethyl)benzenesulphonyl fluoride]} and centrifuged at 320 000 g for 60 min at 4 ◦ C to obtain a total membrane fraction. The pellet was resuspended and solubilized in PBS containing 3 % (w/v) Thesit and proteinase inhibitors (as mentioned above). The homogenate was rotated for 1 h at 4 ◦ C and then centrifuged at 30 000 g for 30 min at 4 ◦ C. Aliquots of the solubilized crude membranes (7.5 µg) were subjected to SDS/PAGE, followed by immunoblotting analysis using a polyclonal antibody raised against the C-terminal peptide of GLUT4. Proteins were finally visualized by enhanced chemiluminescence (BioWest, Miami, FL, U.S.A.) and quantified by a UVP BioImaging System (UVP, Upland, CA, U.S.A.).

Statistics

Values are presented as means + − S.E.M. and number of vessels (one per pig). Because the coronary artery diameter varied between pigs, the steady-state diameter induced by PGF2α was used as an internal standard (0 %). To test for differences at different time points (baseline, 30 min hypoxia, 60 min hypoxia and reoxygenation), statistical comparisons were performed using ANOVA with a Student–Newman–Keuls post hoc test. To test for changes in relative diameter, lactate/pyruvate ratio and glucose with time and intervention, data were tested compared with baseline by using Student’s paired t test. Differences were considered statistically significant when P < 0.05.

Table 1 Peak transient vasodilation within 3 min following insulin and propranolol alone or in combination during normoxic conditions ∗ Values are means + − S.E.M., n = 4. P < 0.05 compared with control. The concentrations of insulin and propranolol were 1.4 × 10−7 M and 10−5 M respectively.

Organ bath glucose 2 mM Control Propranolol Insulin Propranolol + insulin 20 mM Control Propranolol Insulin Propranolol + insulin

Diameter increase (%) 3.69 + − 0.94 2.23 + − 0.46 1.81 + − 0.87 ∗ 11.73 + − 4.91 2.27 + − 0.42 4.38 + − 0.93 2.67 + − 1.13 ∗ 9.90 + − 1.37

RESULTS Coronary diameters Data were obtained from 50 arteries (one per pig). There was a transient increase in coronary artery diameter within the first 3 min after changing the PSS in the organ bath during normoxic conditions. When the PSS contained both insulin and propranolol, the transient diameter increase was significantly greater than control (Table 1). Following hypoxia, there was a consistent increase in diameter in all groups. There was no significant difference between the increase in diameter after 30 min of hypoxia with the 2 mM (11.0 + − 2.1 %, n = 20) and 20 mM (12.6 + − 2.1 %, n = 21; P = 0.61) organ bath glucose concentrations. Similarly, diameters did not differ after 60 min of hypoxia (16.1 + − 1.9 % for 2 mM glucose compared with 13.2 + − 1.9 % for 20 mM glucose; P = 0.3; Figure 2). In the 2 mM glucose group, addition of insulin tended to increase diameters compared with control (25.4 + − 2.2 % compared with 15.6 + − 3 % respectively; P = 0.08). Following 45 min of reoxygenation with both 2 mM and 20 mM organ bath glucose, the presence of insulin and propranolol resulted in a significant coronary C


Figure 2 Effect of hypoxia and reoxygenation on coronary artery diameter (top panel), interstitial lactate/pyruvate ratio (middle panel) and interstitial glucose (bottom panel)

Results depict 2 and 20 mM organ bath glucose. Arteries were examined without drugs (control), with propranolol alone, with insulin alone or with a combination of both propranolol and insulin. Norm., normoxia; Hyp.30, following 30 min of hypoxia; Hyp.60, following 60 min of hypoxia; Rep. 45, following 45 min of reoxygenation. ∗ P < 0.05 compared with control (ANOVA).


artery constriction compared with control (2 mM glucose, − 12.3 + − 1.3 % compared with 4.4 + − 1.5 % respectively; P < 0.05; 20 mM glucose, − 8.5 + 1.8 % compared − with 0.5 + 0.8 % respectively; P < 0.05). − Isoprenaline (10−6 M) evoked pronounced vasodilation, which was not increased further by hypoxia (n = 4; results not shown). Propranolol abolished isoprenalineevoked vasodilation and, in this case, hypoxia increased vessel diameter by 16.3 + − 1.6 % (n = 5), which did not differ from control (13.2 + − 1.9 %; P = 0.9).

Lactate/pyruvate ratio Microdialysis sampling was carried out in 53 arteries during hypoxia and reoxygenation and in 24 timematched control arteries. Propranolol and insulin alone or in combination had no effect on the concentration of lactate or pyruvate in the dialysate during normoxic conditions. Isoprenaline, in the absence or presence of propranolol, had no effect on lactate/pyruvate ratio. Baseline lactate (corrected for RR) demonstrated a considerable range in absolute values from artery to artery (range, 0.21–1.98 mM; mean + − S.E.M., 0.73 + − 0.05 mM). The results are therefore presented as lactate/pyruvate ratios. The lactate/pyruvate ratio increased with hypoxia (Figure 2). Following 60 min of hypoxia, the lactate/ pyruvate ratio increased significantly more with insulin compared with control in both glucose concentration groups (Figure 2). There were no significant correlations between diameter changes and lactate concentration (r = − 0.40) or between diameter and lactate/pyruvate ratios (r = 0.19).

Interstitial glucose Propranolol and insulin alone or in combination had no effect on the concentration of glucose in the dialysate during normoxic conditions. Isoprenaline, in the absence or presence of propranolol, increased dialysate glucose throughout by 3–5 mM (results not shown). Baseline dialysate crude glucose values demonstrated a large range between arteries (2 mM organ bath glucose, range 0.32–1.64 mM; mean + − S.E.M., 0.77 + − 0.05 mM; 20 mM, range 3.50–9.05 mM; mean + − S.E.M., 6.49 + − 0.23 mM), but the change during hypoxia was uniform within each glucose group and there were no withingroup differences. In the 2 mM glucose group, there was a consistent fall in interstitial glucose (− 43.8 + − 3.9 % after 60 min hypoxia; P < 0.01), whereas, in the 20 mM group, hypoxia induced a significant increase (16.5 + − 3.3 % after 60 min hypoxia; P < 0.01; Figure 2). There was no significant correlation between diameter changes and interstitial glucose concentration (r = − 0.43).

Figure 3 Comparison of total crude membrane GLUT4 content in coronary arteries, myocardium, striated muscle and fat

GLUT4 content was determined using 7.5 µg of total membrane protein from each. Data are means + − S.E.M. (n = 4–5). The GLUT4 content is expressed as a percentage of the internal control.

of GLUT4 was observed in the myocardium, GLUT4 was found consistently in all examined arteries (n = 5) at < 6 % of the content in myocardium (Figure 3).

DISCUSSION Hypoxia-related changes in smooth muscle contractility and glycolysis interact at several levels (Figure 4), and we expected to find a relationship between stimulated and unstimulated glycolytic activity on the one hand and hypoxic porcine coronary artery smooth muscle contractile function on the other. Our main findings were as follows. (i) Interstitial lactate/pyruvate ratio approximately doubled during hypoxia. With insulin, the hypoxic lactate/pyruvate ratio quadrupled. (ii) Interstitial glucose increased during hypoxia when the organ bath glucose concentration was 20 mM and decreased when it was 2 mM. Interstitial glucose concentrations were unchanged by insulin and propranolol, consistent with our finding of only small amounts of GLUT4 in the coronary arteries. (iii) The coronary artery diameter increase in response to hypoxia was independent of organ bath glucose concentration and to the addition of propranolol (with or without isoprenaline) and insulin alone or in combination. Despite significant changes in the concentrations of glycolytic substances during hypoxia, the only impact on contractile activity was the finding of augmented posthypoxic constriction following the administration of a combination of insulin and propranolol.

GLUT4 Assessment of total crude membrane GLUT4 protein content after equal loading of total membrane protein (7.5 µg) is shown in Figure 3. Although, the majority

Glycolysis We have reported previously [27] increases in interstitial lactate concentration during hypoxia when using an organ C


217

218


Figure 4 Simplified schematic diagram of coupling between hypoxia, glycolysis and vasodilation showing proposed actions of hypoxia, insulin and propranolol

At the sarcolemma, hypoxia, propranolol and insulin are believed to reduce non-esterified fatty acid (FFA) uptake. Hypoxia was hypothesized to promote GLUT4 translocation and increase GLUT4 cell-surface content and, thus, increase insulin sensitivity and glucose uptake. Propranolol antagonizes β-adrenoceptors, which might reduce vascular tone. Intracellularly, hypoxia accelerates anaerobic metabolism and increases lactate production. SR, sarcoplasmic reticulum; PDH, pyruvate dehydrogenase.

bath glucose concentration of 11 mM. Furthermore, dichloroacetate, an agent that reduces lactate generation by activating pyruvate dehydrogenase, abolished hypoxiainduced lactate production and increased coronary artery diameters. In the present study, we found no relationship between lactate concentration and diameter changes, which leaves little opportunity for lactate to play a causal role in hypoxic coronary artery dilation. We also found no differences in the hypoxic-induced dilation of coronary arteries between organ bath glucose concentrations of 2 and 20 mM. This is in contrast with findings in the rabbit aorta [24] where the reduction in contractile tone at low Po2 was less pronounced when the tissue was exposed to 25 mM glucose than with 5 mM glucose. The rate of glucose uptake during ischaemia is regulated at multiple sites, such as the glucose transporter hexokinase activity, the rates of glucose 6-phosphate activity, various points along the glycolytic pathway and the ability of the mitochondria to oxidize pyruvate and cytosolically produced NADH [28]. In the present study, addition of 1.4 × 10−7 M insulin had no effect on coronary artery interstitial glucose concentration during hypoxia. Others [29] have found that physiological concentrations of insulin (10−12 –10−9 M) increase glucose transport in rat aortic smooth muscle cells by 8–15 % and up to 60 % with supraphysiological concentrations (10−9 –10−6 M). In the same study [29], GLUT4 was shown, but not quantified, by antibody staining in rat C


vascular smooth muscle cells. In another study [30], GLUT4 was quantified in rats and, although aortic tissue contained less GLUT4 than soleus muscle, insulin (1.4 × 10−7 M)-stimulated 2-deoxyglucose transport increased more in the aorta than in soleus. In the present study, we found that GLUT4 was abundant in porcine myocardium, but sparse in coronary arteries. Even the addition of insulin before hypoxia did not change interstitial glucose. Thus our data did not show any significant insulin sensitivity in porcine coronary arteries. Our finding of opposite directional changes in dialysate glucose concentration during hypoxia in the presence of 2 and 20 mM glucose could not be explained by GLUT4 activity. The lactate/pyruvate ratio, a redox system in equilibrium with the cytoplasmic pool of free NADH/ NAD [31], was equally affected by both glucose concentrations. Glycogen breakdown represents an energy source for glycolysis but, in contrast with many other types of smooth muscle, vascular smooth muscle does not utilize glycogen to any significant extent [32].

Insulin and β-adrenoceptor antagonism In combination, and independent of glucose concentration, insulin and propranolol resulted in a significant coronary artery constriction during reoxygenation. We have only indirect evidence to suggest that this constriction was related to changes in metabolism, but the fact that the lactate/pyruvate ratio increased much more


with insulin than with the combination of insulin and propranolol indicates a profound modulating effect of βadrenoceptor antagonism on smooth muscle metabolism. Propranolol in the organ bath had no effect on coronary artery interstitial glucose concentration. In the canine myocardium, propranolol increases glucose extraction, a change that continues after coronary artery ligation [5].

Energetics An artery sustaining a tonic contraction must continually renew its energy stores. If pyruvate replaces glucose as substrate, the metabolism shifts to being virtually all oxidative, and contraction can no longer be sustained in the absence of oxygen in smooth muscle from rabbit urinary bladder [33]. Intracellular energy stores were not determined in the present study, but others have found that GIK supplementation approximately doubled myocardial PCr [34]. In phenylephrine precontracted femoral arteries, hypoxia decreased PCr concentration, increased intracellular phosphate, but had no effect on ATP concentration [17]. In rabbit aorta, a fall in PCr concentration could be correlated with the contractile changes observed during hypoxia [24]. In another rabbit aorta study [35], levels of MLC (myosin light chain) phosphorylation decreased during hypoxia in noradrenaline-activated segments, whereas MLC levels were unchanged in KCl-contracted segments. The authors [35] hypothesize that there is a low-affinity ATPor PCr-dependent reaction upstream of MLC kinase that links oxidative energy production and force. We found that insulin significantly increased the coronary artery interstitial lactate/pyruvate ratio during hypoxia but with no influence on the vessel diameter. We did not measure Po2 in the different layers of the coronary arteries and it could be argued that the region from where the microdialysis catheter is reporting represents both oxygenated areas and an anoxic core region in the middle of the wall [36]. However, the application of an organ bath oxygen tension of 35 mmHg is in line with myocardial Po2 after experimental coronary occlusion in open chest dogs [37] and indicates severe hypoxia. An explanation for the lack of a relationship between hypoxic coronary artery dilation and overall changes in glycolysis could be due to differences in the suggested two functionally separate compartments of glycolytic flux within vascular smooth muscle cells: aerobic glycolysis related to Na+ /K+ -ATPase and oxidative metabolism in response to increases in smooth muscle tension [8,9]. Reduced activity in both compartments could influence contractility, but in opposite directions: (i) inhibition of Na+ /K+ -ATPase increases internal sodium thus promoting sodium-calcium exchange with an increase in calcium and contractility [7]; and (ii) decreased oxidative metabolism could result in a reduction of energy supply

for contractility. The present study was not designed to distinguish between the effects in each of these two compartments, but it is intriguing to assume that hypoxia induces a complex effect on glycolysis by involving both compartments and preventing simple relations between glycolytic substances and contractility.

Conclusions and perspective The present findings suggest that insulin is able to induce marked changes in arterial glycolysis, but treatment with insulin and/or propranolol is not associated with enhanced coronary vasodilation during hypoxia. Hyperglycaemia has been associated with increased morbidity and mortality after myocardial infarction and the administration of GIK has been reported to reduce mortality [1,2]. In the present study, insulin significantly increased lactate/pyruvate ratio during hypoxia, indicating a profound increase in anaerobic glycolysis. We can only speculate that similar metabolic changes take place in the smaller coronary vessels during GIK administration, which may help to maintain coronary perfusion during prolonged periods of ischaemia. Further studies are required to address the significance of our findings in vivo and in diseased arteries.

ACKNOWLEDGMENTS We thank Margit Nielsen for expert technical help. The study was supported by The Danish Heart Foundation (no. 02-01-3-47-22989, Grosserer A. V. Lykfeldt og Hustrus Legat), Lundbeckfonden, the Novo Nordisk Foundation, the Danish Medical Research Council, and Stipendium Cappellanium.

REFERENCES 1 Malmberg, K. (1997) Prospective randomised study of intensive insulin treatment on long term survival after acute myocardial infarction in patients with diabetes mellitus. DIGAMI (Diabetes Mellitus, Insulin Glucose Infusion in Acute Myocardial Infarction) Study Group. Br. Med. J. 314, 1512–1515 2 Sodi-Pallares, D., Ponce, de Leon, Bisteni, A. and Medrano, G. A. (1969) Potassium, glucose, and insulin in myocardial infarction. Lancet i, 1315–1316 3 BHAT research group (1982) A randomized trial of propranolol in patients with acute myocardial infarction. I. Mortality results. JAMA, J. Am. Med. Assoc. 247, 1707–1714 4 Roberts, R., Rogers, W. J., Mueller, H. S. et al. (1991) Immediate versus deferred β-blockade following thrombolytic therapy in patients with acute myocardial infarction. Results of the Thrombolysis in Myocardial Infarction (TIMI) II-B Study. Circulation 83, 422–437 5 Opie, L. H. and Thomas, M. (1976) Propranolol and experimental myocardial infarction: substrate effects. Postgrad. Med. J. 52 (Suppl. 4), 124–132 6 Egert, S., Nguyen, N. and Schwaiger, M. (1999) Contribution of α-adrenergic and β-adrenergic stimulation to ischemia-induced glucose transporter (GLUT) 4 and GLUT1 translocation in the isolated perfused rat heart. Circ. Res. 84, 1407–1415

C


219

220


7 Opie, L. H. (1998) The Heart. Physiology, from Cell to Circulation, 3rd edition, Lippincott-Raven, New York 8 Paul, R. J. (1983) Functional compartmentalization of oxidative and glycolytic metabolism in vascular smooth muscle. Am. J. Physiol. 244, C399–C409 9 Campbell, J. D. and Paul, R. J. (1992) The nature of fuel provision for the Na+ ,K+ -ATPase in porcine vascular smooth muscle. J. Physiol. (Cambridge, U.K.) 447, 67–82 10 Oltman, C. L., Kane, N. L., Gutterman, D. D., Bar, R. S. and Dellsperger, K. C. (2000) Mechanism of coronary vasodilation to insulin and insulin-like growth factor I is dependent on vessel size. Am. J. Physiol. Endocrinol. Metab. 279, E176–E181 11 Taggart, M. J. and Wray, S. (1998) Hypoxia and smooth muscle function: key regulatory events during metabolic stress. J. Physiol. (Cambridge, U.K.) 509, 315–325 12 MacLean, D. A., Sinoway, L. I. and Leuenberger, U. (1998) Systemic hypoxia elevates skeletal muscle interstitial adenosine levels in humans. Circulation 98, 1990–1992 13 Dart, C. and Standen, N. B. (1993) Adenosine-activated potassium current in smooth muscle cells isolated from the pig coronary artery. J. Physiol. (Cambridge, U.K.) 471, 767–786 14 Miura, H., Wachtel, R. E., Loberiza, Jr, F. R. et al. (2003) Diabetes mellitus impairs vasodilation to hypoxia in human coronary arterioles: reduced activity of ATPsensitive potassium channels. Circ. Res. 92, 151–158 15 Foy, R. A., Shimizu, S. and Paul, R. J. (1997) The effects of hypoxia on pHi in porcine coronary artery endothelium and smooth muscle. A novel method for measurements in endothelial cells in situ. Circ. Res. 80, 21–27 16 Shimizu, S., Bowman, P. S., Thorne, III, G. and Paul, R. J. (2000) Effects of hypoxia on isometric force, intracellular Ca2+ , pH, and energetics in porcine coronary artery. Circ. Res. 86, 862–870 17 Leach, R. M., Sheehan, D. W., Chacko, V. P. and Sylvester, J. T. (2000) Energy state, pH, and vasomotor tone during hypoxia in precontracted pulmonary and femoral arteries. Am. J. Physiol. Lung Cell Mol. Physiol. 278, L294–L304 18 Oltman, C. L., Kane, N. L., Gutterman, D. D., Bar, R. S. and Dellsperger, K. C. (2000) Mechanism of coronary vasodilation to insulin and insulin-like growth factor I is dependent on vessel size. Am. J. Physiol. Endocrinol. Metab. 279, E176–E181 19 Schroeder, Jr, C. A., Chen, Y. L. and Messina, E. J. (1999) Inhibition of NO synthesis or endothelium removal reveals a vasoconstrictor effect of insulin on isolated arterioles. Am. J. Physiol. 276, H815–H820 20 Hasdai, D., Rizza, R. A., Holmes, Jr, D. R., Richardson, D. M., Cohen, P. and Lerman, A. (1998) Insulin and insulin-like growth factor-I cause coronary vasorelaxation in vitro. Hypertension 32, 228–234 21 De Sousa, P., Cherian, G., Thomas, J. and Thulesius, O. (1991) Coronary artery constriction is enhanced with nicotine and propranolol, particularly after endothelial damage. Clin. Physiol. 11, 143–152 22 Weisbrod, C. J., Minson, C. T., Joyner, M. J. and Halliwill, J. R. (2001) Effects of regional phentolamine on hypoxic vasodilatation in healthy humans. J. Physiol. (Cambridge, U.K.) 537, 613–621

23 Vanoverschelde, J. L., Janier, M. F., Bakke, J. E., Marshall, D. R. and Bergmann, S. R. (1994) Rate of glycolysis during ischemia determines extent of ischemic injury and functional recovery after reperfusion. Am. J. Physiol. 267, H1785–H1794 24 Namm, D. H. and Zucker, J. L. (1973) Biochemical alterations caused by hypoxia in the isolated rabbit aorta. Correlation with changes in arterial contractility. Circ. Res. 32, 464–470 25 Frøbert, O., Mikkelsen, E. O. and Bagger, J. P. (1999) The influence of transmural pressure and longitudinal stretch on K+ - and Ca2+ -induced coronary artery constriction. Acta Physiol. Scand. 165, 379–385 26 Buhl, E. S., Jessen, N., Schmitz, O. et al. (2001) Chronic treatment with 5-aminoimidazole-4-carboxamide-1-β-dribofuranoside increases insulin-stimulated glucose uptake and GLUT4 translocation in rat skeletal muscles in a fiber type-specific manner. Diabetes 50, 12–17 27 Frøbert, O., Mikkelsen, E. O., Bagger, J. P. and Gravholt, C. H. (2002) Measurement of interstitial lactate during hypoxia-induced dilatation in isolated pressurised porcine coronary arteries. J. Physiol. (Cambridge, U.K.) 539, 277–284 28 Lopaschuk, G. D. and Stanley, W. C. (1997) Glucose metabolism in the ischemic heart. Circulation 95, 313–315 29 Standley, P. R. and Rose, K. A. (1994) Insulin and insulin-like growth factor-1 modulation of glucose transport in arterial smooth muscle cells: implication of GLUT-4 in the vasculature. Am. J. Hypertens. 7, 357–362 30 Banz, W. J., Abel, M. A. and Zemel, M. B. (1996) Insulin regulation of vascular smooth muscle glucose transport in insulin-sensitive and resistant rats. Horm. Metab. Res. 28, 271–275 31 Ishida, Y., Takagi, K. and Urakawa, N. (1984) Tension maintenance, calcium content and energy production of the taenia of the guinea-pig caecum under hypoxia. J. Physiol. (Cambridge, U.K.) 347, 149–159 32 Paul, R. J. (1989) Smooth muscle energetics. Annu. Rev. Physiol. 51, 331–349 33 Wendt, I. R. (1989) Effects of substrate and hypoxia on smooth muscle metabolism and contraction. Am. J. Physiol. 256, C719–C727 34 Opie, L. H., Bruyneel, K. and Owen, P. (1975) Effects of glucose, insulin and potassium infusion on tissue metabolic changes within first hour of myocardial infarction in the baboon. Circulation 52, 49–57 35 Coburn, R. F., Moreland, S., Moreland, R. S. and Baron, C. B. (1992) Rate-limiting energy-dependent steps controlling oxidative metabolism-contraction coupling in rabbit aorta. J. Physiol. (Cambridge, U.K.) 448, 473–492 36 Pittman, R. N. and Duling, B. R. (1973) Oxygen sensitivity of vascular smooth muscle. I. In vitro studies. Microvasc. Res. 6, 202–211 37 Radvany, P., Maroko, P. R. and Braunwald, E. (1975) Effects of hypoxemia on the extent of myocardial necrosis after experimental coronary occlusion. Am. J. Cardiol. 35, 795–800

Received 8 January 2004/10 March 2004; accepted 7 April 2004 Published as Immediate Publication 7 April 2004, DOI 10.1042/CS20040006

C
