INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

OVER THE PAST several decades many studies have examined the metabolic responses of fish white muscle to high-intensity, exhaustive exercise together with the pattern of metabolite recovery (16). These studies have led to the development of a model of fuel selection during exhaustive exercise based on hydrolysis of high-energy phosphates [i.e., creatine phosphate (CrP) and ATP] and "anaerobic" glycolysis leading to lactate accumulation. Furthermore, it has been demonstrated that there is a temporal shift in fuel selection during exhaustive exercise from an initial hydrolysis of CrP (7, 8) to an activation of glycogenolysis and glycolysis (25). As a result, exhaustion in rainbow trout is characterized by a 40 to 60% decrease in white muscle ATP and CrP concentrations and up to a 90% decrease in muscle glycogen concentrations with reciprocal and stoichiometric increases in inosine monophosphate (IMP), free creatine (Cr), and lactate, respectively (e.g., Ref. 39).

During recovery, pathways must be activated to resynthesize ATP, CrP, and glycogen in preparation for another possible bout of exercise. To this end, trout experience an excess postexercise O₂ consumption (EPOC) (34), in part, representing a stimulation of oxidative phosphorylation for ATP production during recovery. The tricarboxcylic acid (TCA) cycle supplies reducing equivalents for mitochondrial oxidative phosphorylation through the utilization of acetyl-CoA. Acetyl-CoA can be produced either from the decarboxylation of pyruvate via pyruvate dehydrogenase (PDH) or through -oxidation of lipid fuels. Amino acids can also supply substrate for the TCA cycle and support ATP production, but it is believed that the contributions of protein oxidation to metabolism are low and can be ignored, particularly during exercise (41). Therefore, the two major fuel sources available to trout white muscle during recovery are the accumulated lactate from glycolysis and lipid fuels. The complete oxidation of a small amount of lactate (4 to 6 µmol/g wet tissue), through the activation of PDH, could yield adequate ATP to support recovery in white muscle. However, there is accumulating circumstantial evidence that suggests the majority of accumulated lactate in trout white muscle is spared from an oxidative fate (20, 45) and retained as the substrate for in situ glyconeogenesis (25, 35).

Traditionally, lipids have not been considered an important fuel during exhaustive exercise and recovery; however, there is mounting evidence that suggests many species rely on lipid oxidation in muscle to fuel recovery (15, 38). In the rainbow trout, Wang et al. (39) showed that immediately after exhaustive exercise, there were decreases in free carnitine and increases in acetyl-carnitine and acetyl-CoA concentrations in white muscle. Accumulation of acetyl groups during recovery points to an activation of oxidative phosphorylation during recovery. Decreases in free carnitine, accompanied by the accumulation of short-chain acyl-carnitine, suggested that lipid was the source of acetyl groups. Decreases in white muscle total lipid concentrations (20) and decreases in plasma nonesterified fatty acids (NEFA) (7) during postexercise recovery in trout further support the use of lipids in fueling recovery from exhaustive exercise.

If we accept this scenario that -oxidation may contribute to ATP production during postexercise resynthesis of CrP, ATP, and glycogen, then the question arises as to the underlying mechanism that regulates fuel selection during recovery. Moyes et al. (24) demonstrated that NEFA oxidation inhibited pyruvate oxidation when isolated trout white muscle mitochondria were incubated simultaneously with pyruvate and NEFA. These workers speculated that this inhibition of carbohydrate oxidation in the presence of NEFA was due to allosteric inhibition of PDH, providing a mechanism by which carbohydrate can be spared at the expense of lipid oxidation. Furthermore, in higher vertebrates, recent evidence has implicated malonyl-CoA in regulating lipid oxidation in skeletal muscle (32). Malonyl-CoA is the first committed step in the de novo synthesis of fatty acids and has been shown in muscle to allosterically regulate carnitine palmitoyltransferase-1 (CPT 1), the enzyme responsible for catalyzing the rate-limiting transfer of fatty acids to carnitine for uptake by mitochondria (1). There remains considerable debate surrounding the regulatory role of malonyl-CoA in fuel selection in muscles of different species (32, 43).

The objectives of the present research were to determine the metabolic fuels oxidized during recovery in trout white muscle to support synthesis of CrP, ATP, and glycogen. Specifically, we measured the activation state of PDH and glycogen synthase (GS) and changes in oxidative metabolites (e.g., acetyl-CoA) and glycolytic intermediates in an attempt to isolate whether lipid or carbohydrate was oxidized during recovery in trout white muscle. Furthermore, we measured changes in intramuscular triacylglycerol (IMTG) and plasma NEFA in an attempt to determine whether endogenous or exogenous lipids were oxidized during recovery. Insights into the control of lipid and carbohydrate oxidation were gained through measurements of malonyl-CoA and estimates of free ADP and AMP (ADP_f and AMP_f, respectively).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal Care

Experimental Protocol

Arterial blood and white muscle were terminally sampled at rest, immediately after exhaustive exercise, and at 0.25-, 0.5-, 1-, 2-, 4-, 8-, 16-, and 32-h recovery. Resting fish were kept in the acrylic boxes for at least 48 h before sampling. For exhaustive exercise, individual fish were transferred from their acrylic box to a 150-liter circular tank filled with water at experimental temperature and manually chased to exhaustion [5 min; similar protocol to Wang et al. (39)]. Upon exhaustion, identified by no further response to manual stimulation, trout were returned to their individual boxes and sampled at the preassigned recovery times. At sampling, trout were terminally anesthetized by adding 0.5 g/l MS-222 to their surrounding water from a neutralized stock solution. During the onset of anesthesia, 3 ml of arterial blood were drawn into an ice-cold gas-tight Hamilton syringe through the DA catheter. Plasma was immediately separated from blood cells by centrifugation at 16,000 g for 10 s. A portion (300 µl) of the plasma was deproteinized in 600 µl of 1 M HClO₄, and the remaining plasma (~1.5 ml) was frozen in liquid nitrogen.

At complete anesthesia (~1 min), the fish were removed from the water, and a white muscle sample was excised from between the lateral line and dorsal fin with a scalpel. The muscle samples were immediately freeze-clamped between two aluminum blocks cooled in liquid N₂, and all samples were stored under liquid N₂ for later analysis. White muscle sampling took less than 10 s.

Analytic Techniques

The active fraction of PDH (PDH_a) was measured in muscle homogenates using a modified technique of Putman et al. (29). Briefly, muscle (30 to 50 mg) was homogenized in 15 times its wet weight in a buffer containing (in mM) 200 sucrose, 50 KCl, 5 MgCl₂, 5 EGTA, 50 Tris · HCl, 50 NaF, 5 dichloroacetic acid, and 0.1% Triton X-100 at pH 7.5. Homogenates were immediately frozen in liquid N₂ until analysis on the same day. To assay for PDH activity, homogenates were thawed on ice, and 60-µl aliquots of homogenate were incubated in duplicate at 10°C in an assay buffer containing (in mM) 144 Tris · HCl, 0.72 EDTA, 1.44 MgCl₂, 3 NAD⁺, 1 CoA-SH, and 1 thiamine pyrophosphate at pH 7.5. The reaction was initiated by the addition of 1 mM pyruvate, and 200-µl aliquots of the incubation media were sampled at 2, 4, and 6 min, except those tissues from the exhausted fish that were sampled at 1, 2, and 3 min because of the high-PDH activity. Tissue blanks were also run with homogenates incubated in the same buffer, but with the addition of deionized water instead of pyruvate. The reaction was stopped by the addition of each aliquot to 40 µl of 0.5 M HClO₄. After 5 min at room temperature, each aliquot was neutralized with 1 M K₂CO₃, centrifuged for 3 min at 16,000 g, and stored at 80°C until analysis of acetyl-CoA. PDH activity determined in the presence of pyruvate was corrected for PDH activity in the blank, and a regression between acetyl-CoA production and time was used to calculate reaction rates.

Total PDH activities (PDH_tot) were assayed on a separate group of fish taken from the same stock. Briefly, muscle was homogenized in a similar buffer as described for PDH_a with the addition of 10 mM D-glucose, 10 mM CaCl₂, and 4 U/ml sulfate-free hexokinase. Homogenates were immediately frozen in liquid N₂, thawed on ice, and incubated at 10°C for 30 min before samples were incubated in assay buffer as described above for PDH_a. The percent mole fraction of PDH transformation was determined by dividing PDH_a by PDH_tot.

An aliquot of lyophilized muscle was used for the determination of GS activity. Briefly, 5 to 10 mg of dry muscle were homogenized at 25°C in 200 µl of buffer containing (in mM) 50 imidazole · HCl, 100 KF, 10 EDTA, and 60% (vol/vol) glycerol at pH 7.5. Homogenates were then diluted with 800 µl of the above buffer without glycerol and homogenized further at 0°C. Total GS (GS_tot) and the active fraction (GS_a) were determined at saturating and physiological concentrations of glucose 6-phosphate (glu 6-P), respectively. The GS assay measured the incorporation of glucose from UDP-glucose into glycogen with the subsequent analysis of liberated UDP. For GS_tot activity (high glu 6-P), 100-µl aliquots of homogenate were incubated with 450 µl of buffer containing 50 mM imidazole · HCl, 2 mM EDTA, 0.2% (wt/vol) glycogen, 0.02% (wt/vol) BSA, 0.5 mM dithiothreitol, and 10 mM glu 6-P at pH 7.5. For GS_a, 100-µl aliquots of homogenate were incubated with 450 µl of buffers of the same composition as for GS_tot, except glu 6-P concentrations were adjusted to reflect those measured in vivo in white muscle (see Table 2 in RESULTS). The reactions for GS_a and GS_tot were initiated by the addition of 8 mM UDP-glucose and incubated at 10°C for 45 min. The reaction was stopped by the addition of 60 µl 0.5 M HCl, and after 10 min on ice, samples were neutralized with 60 µl 0.5 M KOH, centrifuged at 20,000 g for 5 min at 4°C, and the supernatant was assayed for free UDP. Free UDP was assayed in a buffer containing (in mM) 20 Tris · HCl, 30 KCl, 4 MgCl₂, 0.02% (wt/vol) BSA, 0.4 phospho(enol)pyruvate, 0.2 NADH, and 5 U/ml lactate dehydrogenase following the oxidation of NADH after the addition of 3 U/ml pyruvate kinase. The percent mole fraction of GS activation was determined by dividing GS_a by GS_tot.

For the determination of muscle glycogen, ~20 mg of lyophilized muscle were digested in 1 ml of 30% KOH at 100°C. Glycogen was isolated as described by Hassid and Abraham (12), and free glucose was determined after digestion with amyloglucosidase (2). IMTG was determined by measuring total glycerol spectrophotometrically after transmethylation with tetraethylammonium hydroxide (20% aqueous solution) (15).

For the extraction of metabolites from white muscle, aliquots of lyophilized muscle (~20 mg) were weighed into borosilicate tubes; 1 ml of 1 M HClO₄ was added and homogenized for 20 s at 0°C using a Virtis handishear homogenizer at the highest speed. Homogenates were transferred to 1.5-ml centrifuge tubes, centrifuged for 5 min at 20,000 g at 4°C, and the supernatant was neutralized with 3 M K₂CO₃. These extracts were assayed spectrophotometrically for ATP, CrP, Cr, pyruvate, lactate, glu 6-P, fructose 6-P (fru 6-P), glycerol 3-phosphate (gly 3-P), and glycerol (2). Acetyl-CoA, free CoA (CoA-SH), acetyl-, free-, and total carnitine were assayed on neutralized extracts according to radiometric methods (4). Short-chain fatty acyl carnitine was estimated by subtracting acetyl-carnitine and free carnitine from total carnitine. Long-chain fatty acid carnitines were determined on digested white muscle pellets after HClO₄ extraction. Briefly, white muscle pellets were suspended in 1 M HClO₄, vortexed, and centrifuged at 4,800 g for 5 min at 4°C. The washed pellet was then digested in 0.5 ml of 0.5 M KOH for 1.5 h at 50°C, neutralized with 0.25 ml of 1 M HCl, and centrifuged for 10 min at 20,000 g at 4°C. The supernatant was then assayed for free carnitine as described above.

Malonyl-CoA was extracted from lyophilized white muscle in 15 times its weight of 0.5 M HClO₄ containing 50 µM dithioerythritol and 10 µg/ml isobutyl-CoA as an internal standard. After homogenization for 20 s at 0°C at the highest speed of the Virtis homogenizer, homogenates were centrifuged at 20,000 g for 10 min at 4°C, and 200 µl of supernatant were transferred to a borosilicate vial. Extract pH was adjusted to 4 or 5 with 17.5 µl of 4 M NaOH while being vortexed. Supernatants were transferred to autosample vials containing 20 µl of 1 M MOPS (pH 6.8), and final pH of the sample was determined using pH paper: pH was always <5. Autosample vials containing the tissue extracts were immediately placed into a refrigerated autosampler (4°C; WISP 601; Waters, Mississauga, Ontario, Canada). Malonyl-CoA was separated by reverse-phase HPLC using a modified method originally described by Demoz et al. (6). Briefly, 50-µl aliquots of extract were automatically injected onto a Kromasil-octadecyl silane (ODS) column [25 × 0.46 cm; 100 Å ODS, 5 µm; Chromatography (CSC) Sciences, Montreal, Quebec, Canada] fitted with a guard column packed with the same material. An elution gradient, set up by a Waters 660 controller, was used to separate the CoA esters. Solvent A was 100 mM sodium phosphate and 75 mM sodium acetate in ultrapure deionized water (pH 4.2), and solvent B was the same as A except in 30% CH₃CN. The gradient was as follows: 0 min, 90% A; 10 min, 60% A; 17.6 min, 10% A. Baseline condition was established again after 8 min of washing with 90% A. The elution was carried out at ambient temperature, and the flow rate was 1.5 ml/min. Absorbance measurements were made at 254 nm on a photodiode array detector (Waters). Resulting peaks were manually identified by comparison of retention times to standards of known composition, and peaks were quantified by comparison with the internal standard.

Plasma lactate and glycerol were analyzed on deproteinized plasma, and triacylglycerol (TAG) was analyzed on plasma using spectrophotometric methods (2). Plasma total NEFA was analyzed using a Wako NEFA C assay kit (WAKO Chemicals, Osaka, Japan). To determine the fatty acid profiles, plasma NEFA samples were methylated using a modification of the methods of Lepage and Roy (18) and separated using gas chromatography. Briefly, 150 µl of frozen plasma, along with 15 µg of heptadecanoic acid (internal standard), were added to 5 ml methanol-acetyl chloride mixture (50:1) in siliconized vials with tight-fitting teflon-lined caps. Vials were maintained at 25-26°C in a Reacti-Therm dry block and placed on a rotating stir plate for 45 min. Methylation was stopped by the addition of 3 ml of 6% K₂CO₃ followed by the addition of 400 µl of hexane. Tubes were shaken and centrifuged at 2,000 g for 10 min, and 300 µl of the upper hexane layer, containing the methyl esters, were removed and placed into 2-ml borosilicate vials with tight-fitting teflon-lined caps. Hexane was then evaporated under N₂ gas, and the methyl esters were redissolved in 50 µl of CS₂. Methyl esters dissolved in CS₂ were stored under N₂ at 25°C until analysis. One microliter of CS₂ containing methyl esters was injected into a gas chromatograph (3400 star gas chromatograph; Varian, Mississauga, Canada) fitted with a flame-ionization detector. Fatty acid methyl esters were separated on a DB-1 capillary column (30-m × 0.25-mm ID, 0.25 µm film; Chromatographic Specialities, Brockville, Ontario, Canada) using a temperature gradient from 100 to 300°C increasing at a rate of 5°C/min and H₂ carrier gas. Unknown fatty acid methyl esters were identified by comparing their retention times with those of a known standard, and the fatty acid methyl esters were quantified by comparison with the internal standard.

Calculations

Data Presentation and Statistical Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In response to manual chasing, trout swam vigorously for the first 1-2 min; thereafter swimming slowed for the remaining 5 min of exercise. Complete exhaustion was characterized by the lack of an avoidance response to >20 s of handling.

Muscle Metabolites

Adenylates and CrP. Muscle [ATP] decreased by ~65% due to the exercise regime and remained lower than resting values for greater than 2 h postexercise (Fig. 1). Exhaustive exercise caused a 75% decrease in [CrP] that was restored to resting concentrations within 15 min (Fig. 1). Decreases in [CrP] were mirrored by stoichiometric increases in [Cr] that remained higher than resting values for >1 h (Table 1). The calculated [ADP_f] and [AMP_f] increased immediately after exhaustive exercise, but they returned to resting values or lower by 15 min postexercise. The ATP/ADP_f ratio followed the same pattern as [ADP_f] and [AMP_f], decreasing immediately after exhaustive exercise and then recovering to resting values by 15 min (Table 1).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   White muscle ATP (A) and creatine phosphate (CrP; B) concentrations at rest (R) and during 32-h recovery from exhaustive exercise. Vertical bar represents 5 min of exhaustive exercise. Data are means ± SE; n = 8 for each point except at time 0 where n = 7. *Significant differences from rest.

GS. The maximal GS_tot activity was similar at rest and throughout the recovery period at 15.1 ± 0.3 nmol · g wet tissue¹ · min¹ (n = 72), except at time 0 where GS_tot was significantly lower at 11.8 ± 0.6 nmol · g wet tissue¹ · min¹ (n = 7). The activation state of GS (% of GS in the "a" form) increased from ~40% at rest to almost 90% during the bout of exhaustive exercise and remained transformed for >8 h recovery (Fig. 2).

View larger version (14K): [in this window] [in a new window]   Fig. 2.   White muscle glycogen synthase activation state at rest and during 32-h recovery from exhaustive exercise. *Significant differences from rest. See Fig. 1 legend for other details.

Glycogen, glycolytic intermediates, and IMTG. White muscle [glycogen] decreased by 85% during exhaustive exercise, and this decrease was mirrored by stoichiometric (1:2) increases in muscle [lactate] (Fig. 3). Muscle glycogen took between 8 and 16 h to return to values that were not statistically different from resting concentrations, although they were still nonsignificantly lower at 32 h. [Lactate] recovered to resting values within 4 h.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   White muscle glycogen (A) and lactate (B) concentrations at rest and during 32-h recovery from exhaustive exercise. *Significant differences from rest. See Fig. 1 legend for other details.

PDH. Total PDH activity was 167.6 ± 8.9 nmol · g wet tissue¹ · min¹ (n = 7). Exhaustive exercise caused a 50-fold increase in PDH_a in trout white muscle and fully transformed PDH into the active state (Fig. 4). After the activation of PDH at exhaustion, there was a dramatic decrease in PDH_a transformation and activity, back to resting values, within the first 15 min postexercise.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   White muscle pyruvate dehydrogenase (PDH) activity and active (PDHa) mole fraction at rest and during 32-h recovery from exhaustive exercise. *Significant differences from rest. See Fig. 1 legend for other details.

Acetyl group accumulation and carnitine. Muscle [CoA-SH] did not change significantly after exercise and throughout the postexercise period (Fig. 5) and constituted ~90% of the total CoA pool within the muscle. Muscle [acetyl-CoA] increased by 1.6-fold at 15-min recovery and remained elevated compared with resting values for >2 h (Fig. 5).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   White muscle acetyl-CoA and free CoA (CoA-SH) concentrations at rest and during 32-h recovery from exhaustive exercise. *Significant differences from rest. See Fig. 1 legend for other details.

View larger version (29K): [in this window] [in a new window]   Fig. 6.   White muscle acetyl-carnitine (Acetyl), free carnitine (Free), short-chain (SC) fatty acyl carnitine, and total carnitine (Total) concentrations (A) and long-chain (LC) fatty acyl carnitine (B) at rest and during 32-h recovery from exhaustive exercise. *Significant differences from rest. See Fig. 1 legend for other details.

Malonyl-CoA. Muscle [malonyl-CoA] did not change due to the exhaustive exercise, but it increased gradually to approximately twice the resting levels at 2 and 4 h (Fig. 7). Subsequently, [malonyl-CoA] returned to resting values by 8 h.

View larger version (13K): [in this window] [in a new window]   Fig. 7.   White muscle malonyl-CoA concentrations at rest and during 32-h recovery from exhaustive exercise. *Significant differences from rest. See Fig. 1 legend for other details.

Plasma Metabolites

Total [NEFA], measured by enzymatic analysis, was not affected by exhaustive exercise, but it decreased within the first 15 min postexercise, remained depressed for 1 h, and then returned to resting values (Table 4). Analysis of plasma NEFA by gas chromatography (GC) (Table 4) yielded changes in total [NEFA] that followed a similar trend to that observed by enzymatic analysis, but the concentrations were four- to sixfold higher by GC. At rest, palmitic acid (16:0) accounted for ~24% of the total [NEFA], whereas unsaturated 18, 20, and 22 carbon NEFA comprised 20, 15, and 26%, respectively, of the total [NEFA]. The decreases in total [NEFA] observed by enzymatic and GC analysis during recovery were made up of significant decreases in palmitoleic acid (16:1) and unsaturated 18 carbon fatty acids, plus nonsignificant decreases in many of the others (Table 4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study examined the effects of 5-min exhaustive exercise on white muscle metabolism in trout and monitored the recovery of muscle metabolites for up to 32 h. On the basis of substrate depletion and enzyme activities, we have demonstrated that CrP hydrolysis, glycolysis, and oxidative phosphorylation of carbohydrate fuels are responsible for ATP production during exhaustive exercise in trout white muscle. Immediately postexercise, there is a dramatic shift in substrate utilization from phosphagen and carbohydrate during exercise to lipid during recovery. Furthermore, this substrate shift from carbohydrate to lipid oxidation during recovery occurs in the presence of a high concentration of carbohydrate substrate (lactate). Our data argue against lactate oxidation during recovery [classical O₂ debt hypothesis (13)] and add further evidence to the growing idea that recovery metabolism is supported by lipid oxidation (15, 24, 39).

ATP Production During Exercise

PDH is the rate-limiting enzyme that regulates the entry of glycolytically derived pyruvate into the TCA cycle and oxidative metabolism (42); therefore, PDH controls the oxidative utilization of carbohydrate fuels. PDH activity is regulated by both product inhibition (NADH and acetyl-CoA) and by reversible covalent modification (phosphorylation and dephosphorylation). The transformation of PDH between active PDH_a and inactive PDH_b is regulated by the relative activities of PDH kinase, which phosphorylates and thus deactivates PDH, and the activity of PDH phosphatase, which dephosphorylates and thus activates PDH (42). PDH kinase is stimulated by elevated ratios of acetyl-CoA to CoA-SH, ATP to ADP_f, and NADH to NAD⁺ and is inhibited by elevated pyruvate concentrations (38). PDH phosphatase is stimulated by elevated Ca²⁺ concentration (42).

At the onset of exercise, Ca²⁺ release from the sarcoplasmic reticulum probably acted as the initial cue to activate PDH in the trout white muscle. Subsequently, an accumulation of pyruvate (Table 2) from high glycolytic flux and a decrease in ATP/ADP_f (Table 1) likely acted to maximally stimulate PDH during exhaustive exercise. There was no change in acetyl-CoA/CoA-SH ratio at exhaustion (Fig. 5), and in a very similar study in our laboratory, the redox potential (NADH/NAD⁺) of white muscle cytoplasm did not change during a comparable exercise regime (39). The constant acetyl-CoA/CoA-SH ratio and redox state at exhaustion indicate that there were no strong inhibitory forces acting on PDH kinase and thus PDH transformation. If PDH was fully transformed into PDH_a for the 5 min of exhaustive exercise, oxidative phosphorylation of pyruvate could contribute up to 13 µmol ATP/g wet tissue in addition to the ~80 µmol ATP/g wet tissue supplied by ATP, CrP, and glycogen.

The activation of PDH at exhaustion also challenges the dogma that lactate accumulation during high-intensity exercise is due to "anaerobiosis" (8). In fact, accumulating evidence in human muscle suggests that lactate accumulation is due to a mismatched balance between the activities of glycogen phosphorylase, which sets the upper limit for glycogen entry into glycolysis, and the activity of PDH (28). The same, yet accentuated, explanation of lactate production can be applied to trout white muscle. White muscle of fish has a very low mitochondrial content (<2% volume density) (14) and very likely has a low copy number of PDH. Maximal activation of PDH in trout white muscle is not sufficient to accommodate the rate of pyruvate production by glycolysis during exhaustive exercise, thus resulting in lactate accumulation. Clearly, more research is needed to clarify the role of PDH in ATP and lactate production in trout white muscle during exhaustive exercise.

Recovery Metabolism in White Muscle

Glycogen synthesis is regulated, in part, by the activation of the rate-limiting enzyme GS, which catalyzes the addition of glucose from UDP-glucose to glycogen (36). The transformation of GS between active GS_a and inactive GS_b is regulated by the relative activities of a GS kinase and a GS phosphatase. Phosphorylation by a kinase inactivates GS, and dephosphorylation by a phosphatase activates GS. In human muscle, GS phosphatase is primarily stimulated by an increase in glu 6-P and inhibited by high ATP and glycogen concentrations (27).

The large decreases in white muscle glycogen observed at the end of exercise (Fig. 3) and the accumulation of glu 6-P during the first 4 h of recovery were probably the main allosteric regulators for the stimulation of GS phosphatase activity and therefore activation of GS. As a result, within 30 min postexercise, ~90% of the GS was transformed into GS_a (relative to a resting level of ~40%) and remained elevated compared with resting values, for >8 h postexercise (Fig. 2). During this period, there was a 53% increase in white muscle glycogen and a return to resting values for white muscle lactate (Fig. 3). However, our data suggest further that the maximal activity of GS_a (14.4 nmol · g wet tissue¹ · min¹) may limit the rate of recovery in trout white muscle. At the maximal GS activities measured during recovery in trout muscle, it would take >12 h for glycogen to be resynthesized, similar to the pattern of glycogen recovery observed (Fig. 3).

The slow recovery observed in the present study is probably due to the confinement of the trout immediately after exercise. Milligan et al. (21) have shown that rainbow trout swum at 0.9 body lengths/s during recovery restore white muscle glycogen and lactate concentrations to resting values within 2 h postexercise vs. >4 h needed in the present study and others (7, 20, 39). There is mounting evidence that cortisol release during the postexercise period in confined trout may be responsible for prolonged recovery times. If cortisol levels are kept low, either by allowing the fish to swim slowly during the postexercise period (21) or by pharmacological blockade of cortisol synthesis or release (10), trout white muscle carbohydrate and acid-base status recovers at an accelerated rate. However, the precise mechanism behind the action of cortisol during recovery remains elusive. Given the apparent limiting activity of GS in trout white muscle demonstrated in the present study, there might be a yet uninvestigated link between cortisol release and GS activity.

At the onset of recovery, the main purpose of ATP production shifts from providing energy for actin-myosin cycling during the exercise to providing energy for the resynthesis of metabolites. During recovery, ATP must be rapidly generated to resynthesize CrP (~31 µmol ATP/g wet tissue needed within 15-min recovery) followed by a slower synthesis of ATP and glycogen (>2 and >8 h, respectively). From the present study, ATP synthesis from IMP (39) would require ~10 µmol ATP/g wet tissue, and glycogen synthesis from lactate would require between 19 and 30 µmol ATP/g wet tissue. The total ATP required to fuel recovery in the present study was calculated to be between 60 and 70 µmol ATP/g wet tissue.

Fate of Lactate During Recovery

Lactate oxidation during recovery would require the sustained transformation of PDH into PDH_a as well as a maintained catalytic rate. Within 15 min postexercise, PDH is nearly fully transformed into its inactive form (PDH_b). This rapid transformation into PDH_b is probably due to increases in acetyl-CoA/CoA-SH (see Fig. 5) and ATP/ADP_f (Table 1) ratios acting to increase PDH kinase activity, resulting in greater PDH phosphorylation and inactivation of PDH. On the sole basis of the transformation state of PDH, ~6 µmol · g wet tissue¹ · min¹ of pyruvate could be decarboxylated by PDH during the first 4 h of recovery, allowing enough lactate oxidation to provide ATP for recovery. Under most exercise conditions (e.g., Refs. 28, 29, 42), the catalytic rate of PDH is equal to the transformation state. However, in the present study, it seems likely that the catalytic rate of PDH would be far lower than indicated by transformation because of significant product inhibition (acetyl-CoA from lipid oxidation; Fig. 5) and the low substrate concentration (pyruvate) observed during recovery (Table 2). Therefore, the rate of lactate oxidation in vivo would be far less than required to supply ATP for recovery. However, these regulatory effects of low pyruvate and elevated acetyl-CoA on PDH have only been documented in mammalian muscle (e.g., Ref. 42) and await confirmation in fish muscle.

Lipid Oxidation During Recovery

The rate-limiting step in muscle lipid oxidation is the CPT 1-catalyzed binding of NEFA, especially long-chain NEFA, to carnitine for the subsequent transfer of fatty acyl carnitines into the mitochondria. Recent evidence implicates CPT 1 as the main point of regulation of lipid oxidation through the interactions with malonyl-CoA. Malonyl-CoA is an intermediate in the de novo synthesis of fatty acids and has been demonstrated in rat muscle to negatively regulate CPT 1 and thus lipid metabolism (33) and to further contribute to the regulation of the glucose-fatty acid cycle (30). However, malonyl-CoA does not equally regulate CPT 1 in all organisms. In human muscle, malonyl-CoA participates in the regulation of fuel selection at rest, but it does not appear to be important for fuel selection during exercise (26). In the trout white muscle, malonyl-CoA concentrations were low at rest and increased between 2 and 4 h postexercise (Fig. 7). It is paradoxical that there were increases in malonyl-CoA during a period characterized by an increase in lipid oxidation. Two possibilities exist to explain these increases in malonyl-CoA during recovery. First, malonyl-CoA may not be an important modulator of CPT 1 in trout white muscle during recovery. Elevated malonyl-CoA may represent an increased elongation of short-chain fatty acids in an attempt to maintain elevated concentrations of long-chain fatty acids for mitochondrial oxidation (32). Second, the delayed increase in malonyl-CoA may indicate that the majority of the costs of recovery are met within the 2-h recovery and that subsequently lipid oxidation is allosterically inhibited by increasing malonyl-CoA. Further research is needed to clarify the role of malonyl-CoA in trout muscle during recovery.

Further support for lipid oxidation during the early stages of recovery is provided by the general decreases in IMTG that were significant at 1 h postexercise (38% reduction; Table 2). IMTG hydrolysis yields three NEFA for -oxidation and one for glycerol. If the decrease in IMTG represents complete oxidation of TAG containing palmitic acid (16:0), it could supply 1.8 mmol ATP/g wet tissue, 21-fold more ATP than required for recovery metabolism. Thus it is likely that in addition to increased oxidation of fatty acids during recovery, there is probably an increase in TAG-NEFA cycling between the muscle and other tissues, both contributing to the EPOC observed in juvenile rainbow trout (34). Utilization of endogenous lipid during recovery is supported further by the results of Milligan and Girard (20) who showed large, highly variable decreases in trout white muscle total lipid concentrations after exhaustive exercise that persisted through 6-h recovery. The significant accumulation of gly 3-P and generally depressed white muscle glycerol concentration (Table 2) suggest that glycerol liberated by TAG breakdown enters glycolysis and may contribute to glycogen resynthesis.

NEFA released into the bloodstream from adipose tissue may also be used during recovery for ATP synthesis. Plasma total NEFA concentration, determined using enzymatic analysis, decreased during the first 15 min and remained depressed for up to 1 h postexercise. These decreases in plasma NEFA concentration were primarily due to significant decreases in palmitoleic (16:1) and unsaturated 18 carbon fatty acids, although many of the others also tended to decrease (Table 4). Furthermore, these decreases in plasma NEFA were not associated with a change in plasma TAG (Table 3), indicating that esterification of NEFA into TAG does not occur in the extracellular fluid during recovery. The decrease in plasma NEFA observed during recovery was probably due to the combined effects of increased NEFA uptake from the plasma in addition to decreased NEFA release from adipose tissue.

The mobilization of NEFA from adipose tissue is determined by the relative activities of two opposing nonequilibrium reactions: lipolysis of stored TAG and reesterification of NEFA into TAG. This substrate cycling between NEFA and TAG constitutes a means of rapidly adjusting substrate flux without extreme activation or inactivation of any one reaction (40). Stimulation of hormone-sensitive lipase (HSL), by the characteristic mobilization of catecholamines into trout plasma after exhaustive exercise (e.g., Ref. 21), would be expected to shift NEFA-TAG cycling toward NEFA production, thus resulting in a release of NEFA into circulation (40). However, high plasma lactate concentrations, such as those observed during the postexercise period in trout (Table 3), have been demonstrated to inhibit HSL in human adipose tissue (3, 40). Inhibition of HSL would result in reduced NEFA release from adipose tissue. Increased muscle uptake of NEFA coupled with decreased NEFA release from adipose tissue may explain the reduction in plasma NEFA concentrations.

The distribution of NEFA within plasma, as determined by GC, is similar to distributions observed by other researchers who employed the same methylation technique (11); however, there are unresolved differences in plasma NEFA concentrations when analyzed using enzymatic analysis vs. GC. Trout, similar to many teleost fish, are unique in that they have high concentrations of high-density lipoprotein in their plasma [HDL; 15 mg/ml (5) vs. ~2 mg/ml in mammals (19)], and in addition to albumin, they use HDL as a fatty acid transport protein. The chemical methylation process involved in GC may liberate NEFA from HDL and therefore yield higher plasma [NEFA], whereas HDL-bound NEFA may be undetectable by the enzymatic method. If this analytic possibility is true, it suggests that HDL may be the major fatty acid binding protein in trout plasma (similar to carp) (17), accounting for ~80% of the total NEFA carrying capacity of the plasma. This analytic possibility deserves further attention.

The present study provides evidence that during recovery, the majority of the ATP needed to synthesize CrP, ATP, and glycogen is provided for by lipid oxidation. NEFA, especially long-chain NEFA, from both exogenous and endogenous sources are taken up by the mitochondria of white muscle via a carnitine-dependent transport mechanism (CPT 1) and oxidized by -oxidation yielding acetyl-CoA. Acetyl groups are accumulated in the muscle postexercise and support ATP production through increased TCA cycle flux and oxidative phosphorylation. Increases in malonyl-CoA during recovery do not appear to limit fatty acid oxidation, but they may represent elongation of fatty acids for mitochondrial oxidation. Lactate is saved from an oxidative fate during recovery by a rapid transformation of PDH into its inactive form, in addition to product inhibition, thus providing further evidence that glycogen synthesis is likely the major fate of lactate during recovery.

Perspectives