Cellular energetics in the preconditioned state: protective role for phosphotransfer reactions captured by 18O-assisted 31P NMR.

Cell survival is critically dependent on the preservation of cellular bioenergetics. However, the metabolic mechanisms that confer resistance to injury are poorly understood. Phosphotransfer reactions integrate ATP-consuming with ATP-producing processes and could thereby contribute to the generation of a protective phenotype. Here, we used ischemic preconditioning to induce a stress-tolerant state and (18)O-assisted (31)P nuclear magnetic resonance spectroscopy to capture intracellular phosphotransfer dynamics. Preconditioning of isolated perfused hearts triggered a redistribution in phosphotransfer flux with significant increase in creatine kinase and glycolytic rates. High energy phosphoryl fluxes through creatine kinase, adenylate kinase, and glycolysis in preconditioned hearts correlated tightly with post-ischemic functional recovery. This was associated with enhanced metabolite exchange between subcellular compartments, manifested by augmented transfer of inorganic phosphate from cellular ATPases to mitochondrial ATP synthase. Preconditioning-induced energetic remodeling protected cellular ATP synthesis and ATP consumption, improving contractile performance following ischemia-reperfusion insult. Thus, the plasticity of phosphotransfer networks contributes to the effective functioning of the cellular energetic system, providing a mechanism for increased tolerance toward injury.

In the heart, coupling of energetics with contractile function is facilitated through phosphotransfer relays, catalyzed by creatine kinase, adenylate kinase, and glycolysis (11)(12)(13)(14)(15). Poor contractile performance in the failing myocardium is associated with deficits in phosphotransfer-dependent metabolic signaling (16 -19). Furthermore, disruption in phosphotransfer enzymes compromises the ability of heart muscle to respond to metabolic stress (20 -23). Alterations in cellular energy metabolism triggered by ischemic preconditioning, including a characteristic creatine phosphate overshoot, indicates that this protective process targets phosphotransfer reactions (24 -27). However, direct evidence demonstrating the protective role of phosphotransfer networks in the preconditioned state is still lacking.
Here, we demonstrate that ischemic preconditioning of heart muscle induces remodeling in cellular energy transduction, transfer, and utilization processes, thereby promoting preservation of energy metabolism. Post-ischemic contractile recovery was tightly associated with preconditioning-induced adjustment in metabolic flux through creatine kinase, adenylate kinase, and glycolytic systems. Thus, preconditioning shifts intracellular phosphotransfer networks into a stress-tolerant mode rendering heart muscle more resistant to metabolic injury.
Experimental Protocols-Control hearts were perfused for 45 min. Preconditioned hearts were perfused for 25 min and then subjected to two "conditioning" cycles, 5 min of ischemia plus 5 min of reperfusion each. Control post-ischemic hearts were perfused for 45 min and then subjected to 45 min of zero-flow normothermic ischemia followed by a 30 min-long reperfusion. Preconditioned post-ischemic hearts were preconditioned as described above and then subjected to 45 min of ischemia and 30 min reperfusion. Hearts in each of the four groups were labeled with 18 O at the end of respective protocols. 18 O Phosphoryl Labeling-Heart labeling with 18 O displays exponential kinetics with saturation occurring after 2 min. 18 O labeling was performed for 30 s within the initial linear phase using a K-H buffer with 30% of 18 O H 2 O (Isotec). Hearts were freeze-clamped, pulverized under liquid N 2 , and extracted in a solution containing 0.6 M HClO 4 and 1 mM EDTA (11,22). Protein content was determined with a DC Protein Assay kit (Bio-Rad). Extracts were neutralized with 2 M KHCO 3 and used to determine 18 O labeling by 31 P NMR spectroscopy. Metabolite levels were determined by 31 P and 1 H NMR spectroscopy. The percent-ages of recovery of 86 and 92% and of hydrolysis of 3 and 1% were measured for creatine phosphate and ATP, respectively.
NMR Spectroscopy-Extracts were processed for NMR spectroscopy as described (22). Samples were pre-cleaned for 1 h with Chelex 100 resin (Sigma) supplemented with internal NMR standards for 31 P and 1 H spectroscopy (1250 nmol of methylene diphosphonate and 50 nmol of 3-trimethylsilyl tetradeutero sodium propionate) and concentrated by vacuum centrifugation (Savant) to a volume of 0.3 ml. Concentrated extracts were filtered (centrifuge filter; 0.22 m, Milipore) and supplemented with 0.3 ml of D 2 O (Isotec). Samples were cleaned additionally with the Chelex resin by rotation at 4°C for 12 h. Extensive chelation was required to remove divalent cations, which otherwise reduce spectral resolution and render samples vulnerable to degradation.
High-resolution 31 P and 1 H NMR spectra were acquired, respectively, at 202.5 and 500 MHz on a Bruker 11 T spectrometer (Avance) in 5-mm tubes at ambient temperature. For 31 P spectra, signal accumulation (i.e. 9000 scans) was run without relaxation delay (acquisition time, 1.61 s) using a pulse width of 10 s (53°angle). Proton decoupling (WALTZ-16 decoupling at 3 kHz radiofrequency; pulse width of 506 s for 1 H) was applied during signal acquisition. The effect of NOE (nuclear Overhauser enhancement) on signal intensities was corrected based on factors derived from recordings on typical samples with and without decoupling. The effect of rapid pulsing on signal intensity was corrected using attenuation factors calculated from T 1 relaxation times of each signal, where times T 1 were determined in typical samples by the inversion-recovery technique. Free induction decays were Fouriertransformed after zero filling to 32 K and filtering with a line-broadening factor of 0.3 Hz. Phase and base line were automatically corrected, and peak integrals were determined with a built-in integration routine (Xwinnmr 2.5 software, Bruker). For 1 H spectra, 64 scans were accumulated under fully relaxed conditions (12.8 s relaxation delay) with a pulse width of 9 s (90°angle). Free induction decays were zero-filled to 32 K and Fourier-transformed without filtering. Phase and base line were manually corrected, and peak integrals were determined with a built-in integration routine.
Calculation of Phosphotransfer Rates- 18 O phosphoryl labeling allows measurement of synthesis, transfer, and consumption of high energy phosphoryl-carrying molecules (22, 28 -30). This procedure is based on the incorporation of one 18  Up to three 18 O atoms are incorporated in phosphoryls of ␥-ATP, creatine phosphate (CrP), ␤-ADP, ␤-ATP, and glucose-6-phosphate (G6P), although a maximum of four 18 O atoms can be incorporated into inorganic phosphate (P i ). In this way, the 18 O-phosphoryl labeling procedure detects only newly generated molecules containing 18 O-labeled phosphoryls reflecting net fluxes through individual phosphotransfer pathways (28 -31).
Incorporation of each 18 O induces isotope shifts of 0.0210, 0.0228, 0.0250, 0.0228, and 0.0228 ppm in the 31 P NMR spectrum of P i , ␥-ATP, CrP, ␤-ADP, and G6P, respectively. It should be noted that the isotope shift in the spectrum of ␤-ATP was different for bridging and nonbridging 18 O oxygens, 0.0170 and 0.0287 ppm, respectively. Moreover, G6P existed in an equatorial and an axial form, and therefore the 16 O and 18 O species of G6P were represented with two peaks corresponding to each of the two forms. During the integration procedure, bridging and nonbridging forms of ␤-ATP as well as equatorial and axial forms of G6P for particular 16 (30). Total cellular ATP turnover was estimated from the total number of 18 O atoms that appeared in phosphoryls of G6P, P i , ␥-ATP, CrP, ␤-ADP, and ␤-ATP (11,29). Creatine kinase phosphotransfer rate was determined from the rate of appearance of CrP species containing 18 O-labeled phosphoryls using pseudolinear approximation (29). Adenylate kinase phosphotransfer flux was determined from the rate of appearance of 18 O-containing ␤-phosphoryls in ADP and ATP (31), whereas glycolytic flux was determined from the rate of appearance of 18 O-labeled G6P (14).
Metabolite Levels-Tissue levels of G6P, ␣-glycerophosphate, and P i were determined according to internal standards from 31 P NMR spectroscopy (32), and ATP, ADP, CrP, creatine, alanine, and lactate were calculated from 1 H NMR spectra (33, 34). Metabolites were identified based on literature data, and assignments were confirmed by standard additions.
Statistical Analysis-Data are expressed as mean Ϯ S.E. analysis of variance with post hoc comparison of means was used for statistical analysis. A difference at p Ͻ 0.05 was considered significant. Correlations were performed by fitting data using linear or sigmoid functions with three parameters. The fidelity of fits was assessed based on R 2 value with variance calculated by the least square method. Statistical significance was calculated for linear fits, and p Ͻ 0.05 considered significant.

Preconditioning-induced Stress-tolerant State-Vigorous
myocardial performance was abruptly lost at the onset of ischemia, and could be only marginally recovered upon reperfusion indicating poor tolerance of heart muscle to ischemiareperfusion injury. Conditioning with two periods of brief (5-min) ischemia, prior to a prolonged ischemia-reperfusion insult increased cardiac tolerance to post-ischemic injury. In control hearts, work performance expressed as the rate pressure product and lusitropic properties reflected by the LVEDP were 35,000 Ϯ 1,500 mm Hg/s and 7.7 Ϯ 0.3 mm Hg, respectively (n ϭ 7). Preconditioning per se did not significantly disturb heart function, with rate pressure product and LVEDP maintained at 32,000 Ϯ 1,800 mm Hg/s and 9.4 Ϯ 0.2 mm Hg, respectively (n ϭ 7). However, preconditioning did provide significant protection against subsequent prolonged (45-min long) ischemia. On average, rate pressure product and LVEDP following 30 min of reperfusion were 6,100 Ϯ 2,200 mm Hg/s and 91 Ϯ 5 mm Hg in nonconditioned (n ϭ 7) versus 17,600 Ϯ 4,300 mm Hg/s and 53 Ϯ 7 mm Hg in preconditioned (n ϭ 7) postischemic hearts, respectively (p Ͻ 0.05). Thus, preconditioning induces a stress-tolerant state characterized by improved performance of the post-ischemic myocardium.
Remodeling of Intracellular Phosphotransfer in the Preconditioned State-Muscle performance critically depends on nucleotide homeostasis, optimal high energy phosphoryl transfer, and efficient ATP utilization (35,36). Here, the dynamics of phosphotransfer reactions were quantified by capturing the rates of 18 O incorporation in high energy phosphoryls using 18 O-assisted 31 P nuclear magnetic resonance spectroscopy ( 18 O/ 31 P NMR). Creatine kinase phosphotransfer rate was monitored in control and preconditioned myocardium through the appearance of phosphoryl species of CrP containing 18 (Fig. 1A). In control hearts, net creatine kinase phosphotransfer flux was vigorous, at 330 Ϯ 20 nmol⅐mg protein Ϫ1 ⅐min Ϫ1 (n ϭ 7; Fig. 1B), in line with the major energetic role of this enzyme in the heart (11,16). Following ischemia-reperfusion, creatine kinase flux dropped over 4-fold to 74 Ϯ 10 nmol⅐mg protein Ϫ1 ⅐min Ϫ1 (n ϭ 7, p Ͻ 0.05; Fig. 1B), indicating deficient phosphotransfer in the post-ischemic heart. In conditioned hearts, creatine kinase phosphotransfer flux was 409 Ϯ 5 nmol⅐mg protein Ϫ1 ⅐min Ϫ1 (n ϭ 7, p Ͻ 0.05; Fig. 1B) before ischemia, conditioning the myocardium for subsequent metabolic stress. Accordingly, creatine kinase flux upon reperfusion was better preserved in the preconditioned heart at 180 Ϯ 30 nmol⅐mg protein Ϫ1 ⅐min Ϫ1 (n ϭ 7), a value significantly higher than that of nonconditioned hearts (Fig. 1, A and B; p Ͻ 0.05). Consistent with alterations in creatine kinase flux, the levels of CrP were higher in preconditioned hearts before and after ischemia (Fig. 1C). Specifically, before ischemia, CrP was 30.4 Ϯ 1.6 versus 39.1 Ϯ 0.7 nmol⅐mg protein Ϫ1 (p Ͻ 0.05), and following ischemia-reperfusion CrP was 9.0 Ϯ 0.8 versus 18.8 Ϯ 2.5 nmol⅐mg protein Ϫ1 (p Ͻ 0.05) in nonconditioned (n ϭ 7) and pre-conditioned (n ϭ 7) hearts, respectively (Fig. 1C). Preconditioning also reduced the loss of total creatine following ischemia-reperfusion (52 Ϯ 3 versus 45 Ϯ 2 nmol⅐mg protein Ϫ1 , p Ͻ 0.05; n ϭ 7). In general, the efficiency of cellular phosphotransfer appeared improved by preconditioning, as indicated by the increase in the CrP/P i ratio (Fig. 1D). Before ischemia-reperfusion injury, the CrP/P i ratio was 1.09 Ϯ 0.07 and 1.67 Ϯ 0.07 in control and preconditioning hearts, respectively (p Ͻ 0.05). Following ischemia-reperfusion, the CrP/P i ratio was dramatically reduced to 0.15 Ϯ 0.02 in nonconditioned but was partially preserved, at 0.37 Ϯ 0.06, in preconditioned hearts (p Ͻ 0.05; n ϭ 7). Thus, preconditioning shifts the creatine kinase phosphotransfer system into a new, more efficient steady state.
In addition to its energy-producing role, glycolysis also has the capability of catalyzing spatially directed high energy phosphoryl transfer (15,37). Here, glycolytic phosphotransfer was monitored by measuring the rate of appearance of 18 O-labeled phosphoryl species in G6P (Fig. 3A) 18 O. B, creatine kinase (CK) flux, expressed in nmol of CrP produced, in control and preconditioned hearts prior to (C and P, respectively) and following (CR and PR, respectively) ischemia-reperfusion. C, cellular creatine phosphate (CrP) levels in C, P, CR, and PR hearts. D, CrP/P i ratio in C, P, CR, and PR hearts. ‡, indicates significant difference between C and P, CR, or PR hearts; *, indicates significant difference between CR and PR hearts.
Phosphotransfer Flux Supports Post-ischemic Recovery-The significance of phosphotransfer in supporting myocardial recovery was further indicated by the relationship observed between individual phosphotransfer enzyme fluxes and post-ischemic contractile function (Fig. 4). In particular, creatine kinase flux correlated closely with the performance of the myocardium following ischemia-reperfusion. The squared correlation coefficient (R 2 ) was 0.81 (n ϭ 6, p Ͻ 0.05) and 0.96 (n ϭ 6, p Ͻ 0.05) in control and preconditioned hearts, respectively (Fig. 4A), suggesting coupling of creatine kinase phosphotransfer with muscle performance. Flux through adenylate kinase also correlated well with contractile performance following ischemiareperfusion. R 2 was 0.77 (n ϭ 6, p Ͻ 0.05) and 0.83 (n ϭ 5, p Ͻ 0.05) in control and preconditioned hearts, respectively (Fig.  4B), indicating contribution of adenylate kinase-catalyzed ␤-phosphoryl energetics in support of cardiac function. Correlation between hexokinase flux and post-ischemic functional recovery was not significant in nonconditioned hearts (R 2 ϭ 0.57, n ϭ 6; p ϭ 0.08). However, it became significant in the preconditioned state (R 2 ϭ 0.68, n ϭ 6, p Ͻ 0.05; Fig. 4C), suggesting that glycolytic phosphotransfer was recruited by the conditioning process to augment post-ischemic contractile recovery. Thus, promotion of phosphotransfer reactions by ischemic preconditioning is associated with protection against ischemia-reperfusion injury.
The ATP consumption rate, monitored by the appearance of 18 O labeled species in P i , was also protected by ischemic preconditioning (Fig. 6A). Total myocardial ATPase activity, expressed as percentage of P i phosphoryls replaced by 18 18 O. B, hexokinase phosphotransfer flux expressed in nmol of nascent G6P in control and preconditioned hearts prior to (C and P, respectively) and following (CR and PR, respectively) ischemia-reperfusion. C, lactate levels in C, P, CR, and PR hearts. D, ␣-glycerophosphate levels in C, P, CR, and PR hearts. ‡, indicates significant difference between C and P, CR, or PR hearts; *, significant difference between CR and PR.
reduced from a base-line value of 79 Ϯ 2 and 76 Ϯ 2% to 22 Ϯ 2 and 39 Ϯ 4% in nonconditioned and preconditioned hearts, respectively (n ϭ 7, p Ͻ 0.05; Fig. 6B). There was a significant correlation (p Ͻ 0.05) between 18 O labeling of P i and contractile performance (Fig. 6C), indicating tight coupling between ATP hydrolytic activity occurring primarily at the actomyosin site and the functional performance of heart muscle. Both linear and sigmoid functions provided good fits of experimental data. Total myocardial ATP turnover was reduced by ischemia-reperfusion in nonconditioned hearts but was significantly protected by preconditioning (Fig. 6D). In control hearts, ATP turnover was 552 Ϯ 3 nmol⅐mg protein Ϫ1 ⅐min Ϫ1 (n ϭ 7), indicating that the total cellular ATP pool undergoes about 20 renewal cycles/ min. At base line, preconditioned hearts had similar ATP turnover rates, at 524 Ϯ 6 nmol⅐mg protein Ϫ1 ⅐min Ϫ1 (n ϭ 7). However, after the ischemic insult, ATP turnover was markedly preserved and was about 80% higher (305 Ϯ 33 versus 171 Ϯ 18 nmol⅐mg protein Ϫ1 ⅐min Ϫ1 , n ϭ 7) in preconditioned than in nonconditioned hearts (p Ͻ 0.05; Fig. 6D). Thus, preconditioning-induced protection of the phosphotransfer network is associated with maintained ATP synthesis and ATP consumption.
Preconditioning Improves Feedback Metabolic Communication-Cardiac contractile function depends not only on the delivery of high energy phosphoryls but also on removal of end-products of the ATPase reaction (38). Here, we took advantage of the unique feature of 18 O labeling to monitor concomitantly phosphotransfer flux and metabolite exchange between cellular compartments. Specifically, 18 O-labeled P i produced during ATP hydrolysis at an ATPase site must reach a distinct ATP production site to be incorporated into ␥-ATP (Fig. 7A). The efficiency of P i removal and the exchange rate between cellular ATPases and mitochondrial ATP synthase can, therefore, be monitored as the ratio of 18 O-labeled P i over 18  O]␥-ATP ratio was high, on the order of 0.84 Ϯ 0.01 and 0.83 Ϯ 0.01, indicating rapid P i removal and/or delivery to ATP production sites (Fig. 7A). Ischemia-reperfusion dramatically reduced the metabolically active P i pool (Fig. 7A) (n ϭ 7). This finding suggests that P i removal from cellular ATPases as well as communication between ATPases and ATP synthases are markedly compromised (Fig. 7A). Indeed, there was negative correlation between LVEDP, a parameter defining muscle relaxation, and the severity of decline in the [ 18 O]P i / [ 18 O]␥-ATP ratio, reflecting impaired intracellular P i mobility (Fig. 7B). Preconditioning, however, limited the decline in the [ 18 O]P i /[ 18 O]␥-ATP ratio and maintained it at a value (0.49 Ϯ 0.04, n ϭ 7) significantly higher (p Ͻ 0.05) than that recorded in nonconditioned hearts (Fig. 7A). Thus, preconditioning promotes removal of the end-products of ATP hydrolysis following ischemia-reperfusion, thereby improving the cardiac contraction-relaxation cycle.

DISCUSSION
Induction of the cytoprotective response is fundamental to cell survival under stress, and yet the mechanisms underlying acquired tolerance to metabolic injury are poorly understood. Here, we demonstrate by the novel approach of 18 O/ 31 P NMR analysis the importance of intracellular phosphotransfer pathways in producing a stress-tolerant cellular energetic phenotype. Preconditioning induced redistribution of high energy phosphoryl transfer through individual phosphotransfer reactions catalyzed by creatine kinase, adenylate kinase, and glycolytic enzymes, leading to an improved intracellular metabolic communication and preservation of cellular ATP synthesis and ATP consumption processes. Thus, the present study establishes that adaptive remodeling of the cellular energetic system is an integral mechanism in ischemic preconditioning-induced cell resistance to stress.
The energetic homeostasis of the cell and, consequently, energy-driven cellular functions require tight coordination between ATP utilization and ATP generation (38 -40). Such coordination is believed to be mediated by phosphotransfer relays composed of creatine kinase, adenylate kinase, and glycolytic enzymes that facilitate high energy phosphoryl delivery and removal of ATPase end-products (12,36,37,41). Here, phosphotransfer dynamics were dissected by 18 O/ 31 P NMR spectroscopy, which allowed simultaneous capture of net phosphotransfer flux through individual enzymes, as well as total ATP production and consumption rates in intact heart muscle. In nonconditioned hearts, a prolonged ischemia-reperfusion challenge produced a marked reduction in the rates of creatine kinase, adenylate kinase, and glycolysis-catalyzed phosphotransfer, contributing to energetic and contractile dysfunction of the post-ischemic myocardium. Phosphotransfer deficit further disrupted intracellular handling of P i , resulting in the accumulation of nascent P i at ATPase sites and impaired de- livery to mitochondrial ATP synthase. A restricted mobility of intracellular P i could be the result of functional entrapment and/or physical diffusional restriction due to the viscosity and high structural organization of the muscle cytosol (15,18). This would negatively impact ATPase activity responsible for efficient contraction-relaxation cycles and contribute to reduced ATP synthesis by oxidative phosphorylation (41). With shortterm hypoxia, the drop in creatine kinase flux is usually compensated for by increased adenylate kinase and/or glycolytic flux (22,37,42). However, with prolonged metabolic stress, uncompensated deficits in phosphotransfer enzymes do develop, as observed here and as previously reported in severe cardiac conditions such as heart failure and various forms of myocardial insufficiency (11, 16 -19). The deletion of genes encoding creatine kinase or adenylate kinase compromises the ability of a muscle to sustain cellular energetic economy under metabolic stress (14, 20 -23, 43). Thus, defective phosphotransfer networking, associated with a reduced rate of ATP turnover, precipitates poor myocardial recovery in the nonconditioned myocardium following ischemia-reperfusion.
Preconditioning improves myocardial post-ischemic contractile recovery and provides protection from metabolic injury (4,8,9,24). In the present study, by 18 O/ 31 P NMR analysis of myocardial phosphotransfer dynamics we found that preconditioning up-regulates creatine kinase phosphotransfer flux associated with higher levels of CrP and an improved CrP/P i ratio. This finding supports the notion that preconditioninginduced creatine phosphate overshoot is required for generation of the protective phenotype (24 -27, 44). In addition to the creatine kinase system, preconditioning also markedly increased glycolytic phosphotransfer associated with higher lactate and ␣-glycerophosphate levels. Previous studies have FIG. 5. Preconditioning protects ATP production in post-ischemic hearts. A, 18 O/ 31 P NMR spectra of 18 Olabeled ␥-ATP in control (CR, left panel) and preconditioned (PR, right panel) hearts following ischemia-reperfusion. 18 O incorporation induces an isotope shift of 0.0228 ppm in 31P NMR spectra. 16 18 O labeling of ␥-ATP, a measure of cellular ATP synthesis, in control and preconditioned hearts prior to (C and P, respectively) and following (CR and PR, respectively) ischemia-reperfusion. C, correlation between ATP synthesis expressed as 18 O labeling of ␥-ATP and cardiac performance expressed as the rate pressure product (RPP). Both linear and sigmoid functions provided tight fits of experimental data. D, cellular ATP levels in C, P, CR, and PR hearts. ‡, indicates significant difference between C and CR or PR hearts; *, significant difference between CR and PR hearts.
FIG. 6. Preconditioning improves ATP production in post-ischemic hearts. A, 18 18 O labeling of P i , a measure of the cellular ATPase rate, in control and preconditioned hearts prior to (C and P, respectively) and following (CR and PR, respectively) ischemiareperfusion. C, correlation between the ATPase rate, expressed as 18 O labeling of P I , and cardiac performance, expressed as the rate pressure product (RPP). Both linear and sigmoid functions provided tight fits of experimental data. R 2 , squared correlation coefficient. D, total ATP turnover expressed as the sum of 18 O incorporated into cellular high energy phosphoryls, in C, P, CR, and PR hearts. ‡, indicates significant difference between C and P, CR, or PR hearts; *, significant difference between CR and PR hearts. indicated increased glucose utilization and reduced glycogenolysis in preconditioning, leading to maintained glycolytic flux at reperfusion (25,45,46). Traditionally, glycolytic metabolism has been identified as an alternative source of myocardial ATP production. The present finding supports an additional role for glycolysis in transferring and distributing high energy phosphoryls in line with vigorous phosphoryl exchange rates in the glyceraldehyde-3-phosphate dehydrogenase/3-phosphoglycerate kinase system observed with 31 P NMR saturation transfer (47,48). In fact, increased phosphotransfer in the glyceraldehyde-3-phosphate dehydrogenase/3-phosphoglycerate couple could be responsible for improved intracellular P i trafficking (15,18) observed in preconditioning hearts. In contrast to the increased contribution of creatine kinase and glycolytic flux, adenylate kinase phosphotransfer in the preconditioned myocardium showed an apparent trend toward down-regulation. This finding may suggest increased competition with creatine kinase and glycolytic systems leading to redistribution of intracellular phosphotransfer flux. Thus, in the preconditioned mode, cellular phosphotransfer equilibrium would operate at a new steady state conditioning cells to withstand subsequent metabolic stress.
The remodeling of cellular metabolism has been considered as a contributor to metabolic defense (22,27,40). Apparently, the plasticity of cellular energetic dynamics materializes in a more efficient coupling of ATP turnover with cellular functions. Tight correlation was discovered here between metabolic fluxes through creatine kinase, adenylate kinase, and glycolytic phosphotransfer and contractile recovery of post-ischemic hearts. The close correlation between phosphotransfer flux and heart muscle performance in the post-ischemic state suggests that phosphotransfer reactions are an obligatory route channeling the flow of high energy phosphoryls. Thus, integrated cellular energetics contributes to the preconditioned phenotype by translating metabolic adaptation into increased stress tolerance.
The sequences of events leading to the development of a preconditioning state are still not fully understood. However, a number of signal transduction cascades, including protein kinases and ion channels, have been implicated as triggers and/or effectors of the protective phenotype (4, 8 -10). Rapid alterations in cellular energetics induced by ischemic preconditioning could trigger signal transduction events by changing the cellular phosphorylation potential. Indeed, there is a close relationship between creatine kinase phosphotransfer and the activity of protein kinase C, believed to be critical in early stages of preconditioning (49). Moreover, adenylate kinase phosphotransfer, through AMP-driven signaling, modulates the behavior of the AMP-activated protein kinase, a metabolic stress kinase (50). Furthermore, intracellular phosphotransfer reactions regulate the behavior of ATP-sensitive K ϩ (K ATP ) channels, an alarm mechanism setting membrane excitability in response to metabolic stress (15,42,(51)(52)(53)(54)(55). In turn, both protein kinase C and AMP-activated protein kinase regulate creatine kinase phosphotransfer rate and other metabolic pathways through phosphorylation of target proteins (49,56). Thus, the feedback communication between stress-sensitive metabolic and signal transduction events may be central to the generation of the preconditioned state.
In summary, this study has uncovered a homeostatic mechanism by which cells induce a preconditioned energetic state conferring increased tolerance toward injury. This is accomplished by a coordinated redistribution of high energy phosphoryl flux through phosphotransfer enzymes allowing more efficient communication of energetic signals and preservation of ATP generation and consumption processes. In this way, intracellular phosphotransfer reactions emerge as an essential component required for the development of an injury-tolerant state and could thereby serve as a target for regulating the cellular response to stress.