Autho

r's

pers

onal

co

py

Invited critical review

Oxidative mechanisms at rest and during exercise

Edouard Ghanassia, Jean-Frédéric Brun ⁎, Jacques Mercier, Eric Raynaud

INSERM, ERI 25, F-34000, Montpellier, FranceCHU Montpellier, Hôpital Lapeyronie, Service de Physiologie Clinique, F-34000, Montpellier, France

Université Montpellier1, EA 701, F-34000, Montpellier, France

Received 13 December 2006; received in revised form 27 March 2007; accepted 4 April 2007Available online 14 April 2007

Abstract

Carbohydrates (CHO) and lipids provide the amount of energy required for physical and chemical reactions inside the human body. Thevarious constraints the body has to resolve explain the use of these two substrates, catabolized via distinct pathways to one common final reaction.In the classic model, three main organs/tissues for substrate fluxes (liver, adipose tissue and skeletal muscle) and one organ regulating mainreactions by adaptation of hormonal secretions (endocrine pancreas) are described. From this point of view, the only interactions between CHOand lipid metabolisms are mediated by glycaemic changes via insulin/glucagon ratio (IGR). However, according to recent advances, this conceptseems to have a limited validity as it does take into account neither the many other interactions between CHO and lipid metabolism that are likelyto occur in addition to the coarse control by IGR, nor the long-term regulation of energy balance, whose description began with the discovery ofleptin. Moreover, it does not include the effects of energy expenditure.

Therefore, this review focuses on three topics: (i) describe interactions between CHO and lipid metabolism at the level of each tissue and organimplied, via hormonal signaling as well as direct action of nutrients, (ii) integrate fluxes of substrates and signals between those tissues at rest in aglobal view of the metabolism taking into account short-term and long-term regulating factors and (iii) describe separately, to avoid confusion orextrapolation, the short-term and long-term influence of exercise on these regulation loops.© 2007 Elsevier B.V. All rights reserved.

Keywords: Metabolism; Substrates; Exercise; Skeletal muscle; Adipose tissue; Liver

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22. Adipose tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.1. NEFA flux regulation balance between lipolysis and esterification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32.2. Adipose tissue heterogeneity: subcutaneous and visceral adipocytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42.3. Adipose tissue: an endocrine gland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Clinica Chimica Acta 383 (2007) 1–20www.elsevier.com/locate/clinchim

Abbreviations: α2R/βR, α2 receptors/β receptors; ACC, acetyl-CoA carboxylase; ALBP, adipocyte–lipid binding protein; AMPK, AMP kinase; ANP, atrialnatriuretic peptide; ASP, acetylation-stimulating protein; AT, adipose tissue; AT-LPL, adipose tissue-specific lipoprotein lipase; BNP, brain natriuretic peptide; CHO,carbohydrates; CPT-1, carnitine palmitoyl transferase-1; DAG, diacylglycerol; EGP, endogenous glucose production; ET, endurance training; FABP, fatty acid bindingprotein; FAT, fatty acid translocase; G6P, glucose-6-phosphate; GLP-1, glucagon-like peptide-1; HSL, hormone-sensitive lipase; IATG, intra-adipocyte triglyceride;IGR, insulin/glucagon ratio; IHTG, intrahepatic triglyceride; IMTG, intramuscular triglyceride; LPL, lipoprotein lipase; NEFA, non-esterified fatty acid; PKC, proteinkinase-C; RER, respiratory exchange ratio; SCAT, subcutaneous adipose tissue; TG, triglyceride; VAT, visceral adipose tissue.⁎ Corresponding author. Nom Prénom, Service de Physiologie Clinique, CERAMM, Hôpital Lapeyronie, 34295 Montpellier cedex 5, France. Tel.: +33 4 67 33 82

84; fax: +33 4 67 33 89 86.E-mail address: drjfrbrun@dixinet.com (J.-F. Brun).

0009-8981/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.cca.2007.04.006

Autho

r's

pers

onal

co

py

3. Pancreas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54. The liver. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

4.1. Intrahepatic lipid metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64.2. Regulation of EGP: an adipo-hepato-insular loop?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64.3. An adipo-hepato-insular loop of long-term regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

5. Skeletal muscle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75.1. Substrates flux and storage: glycogen and intramuscular triglycerides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85.2. Intramuscular interactions: Randle upside-down . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85.3. AMP kinase: at the crossroad of modulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95.4. Leptin, de novo lipogenesis and thermogenesis: not so independent interactions? . . . . . . . . . . . . . . . . . . . . . . . . 9

6. From virtuous to vicious circles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97. Physical activity: breaking vicious circles… or bypassing them? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118. Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

8.1. Same actors playing different parts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118.2. An overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128.3. Of catecholamines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128.4. Pancreas: of insulin secretion blockade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128.5. Adipose tissue: of lipids mobilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138.6. The liver as a store and a source of CHO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138.7. Muscle: consumer and regulator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

9. Breaking or bypassing the circles: effects of training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159.1. Training and exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

10. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

1. Introduction

Species' survival depends on their ability to cope withenvironmental constraints. Energy homeostasis, the balancebetween caloric intake and energy expenditure, is not anexception.

Energy is provided by three nutrients: proteins, carbohy-drates (CHO) and lipids, the two latter providing more than 80%of total caloric supply. They enter distinct metabolic pathwaysproducing one common metabolite, acetyl-CoA, which, bymitochondrial oxidation through the Krebs cycle and electrontransport in the respiratory chain, results in synthesis of ATP.These substrates fluxes provide the amount of energy requiredfor physical and chemical reactions allowing the wholebiological processes.

Any physiological system tends to reach the lowest potentialenergetic state [1]. So, one can wonder why two distinctsubstrates, catabolized via distinct pathways to one commonfinal reaction, are indeed used. In fact, the complexity of thissystem is explained by the various constraints the organism hasto resolve.

The first constraint is about energy availability: energydemand is continuous whereas caloric intake is discontinuous.So, mechanisms developed, allowing not only the use of thesesubstrates for ATP synthesis (glycolysis, β-oxidation) but alsotheir storage and releasing (glycogenogenesis, glycogenolysis,esterification of fatty acids, lipolysis) and even endogenousproduction of glucose from non-carbohydrate precursors(gluconeogenesis).

The second constraint is about glucose. Most organs andtissues are able to independently use CHO and lipids. However,

brain and red blood cells cannot use lipid oxidation and aremainly dependent upon CHO. Therefore lipids tend to be usedas the main energy source in order to spare glucose when thelatter is poorly available (post-absorptive state) whereas CHOtend to be used when they are abundant (post-prandial state) inorder to store lipids whose availability is crucial in the post-absorptive state.

The third constraint is about plasma availability which mustbe maintained in a range of values allowing survival, too high ortoo low a rate resulting in short- or long-term complications.

In order to describe factors regulating substrate metabolism,researches focused on identification of tissues and organs whereall the reactions take place and where regulating factors aresynthesized. In the “classic” conception, three main organs/tissues for substrate fluxes (liver, adipose tissue and skeletalmuscle) and one organ regulating main reactions by adaptationof hormonal secretions (endocrine pancreas) are described.According to this model:

(i) The liver is the only organ capable of ensuring constantavailability of glucose for it is able to store it (glycogen-ogenesis), to release it into circulation (glycogenolysis)and to produce it from non-carbohydrate substrates(gluconeogenesis).

(ii) Adipose tissue is considered as a storehouse, capable ofensuring lipid availability, releasing non-esterified fattyacids during the post-absorptive state and storing themafter esterification as triglycerides during the post-prandial state.

(iii) Skeletal muscle is the main energy-consumer, using lipidsduring the post-absorptive state and CHO during the post-

2 E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

prandial state. It can also store glucose as muscularglycogen to provide energy during exercise, particularlywhen extra-muscular energy supply does not achieve fastenough energy demand.

In all these tissues, metabolism is modulated by enzymescatalysing limiting steps. The activity of these enzymes isregulated by intermediary metabolites and other factors such aspH or body temperature. But the main factor is certainly theinsulin/glucagon ratio (IGR), insulin and glucagons being thetwo main hormones produced by the endocrine pancreas.

From this point of view, the only interactions between CHOand lipids metabolisms are mediated by IGR, i.e., by glycaemicchanges.

In the second half of 20th century, most researches focusedon CHO metabolism, opening an « All Glucose » era.Understanding pathogenesis of type 2 diabetes in a therapeuticperspective was the leitmotiv. It is also the time when theconcept of insulin resistance was born. One must notice that,although insulin effects on lipid metabolism were alreadyknown, insulin resistance has been defined as a defect of insulinaction on carbohydrate metabolism [2].

However, in the early sixties, Sir Philip John Randleintroduced a concept audacious enough to be the matter ofmany debates almost half a century later. According to histheory, now demonstrated in resting skeletal muscle, an increasein NEFA availability causes a decrease in muscular carbohy-drate utilization [3]. So, he suggested that blood glucose-induced variation in IGR is not the only regulating factor andthat many other interactions between CHO and lipid metabo-lism are likely to occur, anticipating the current concept ofnutrient sensing.

Moreover, the “classic” model does not take into accountlong-term regulation of energy balance, of which the descriptionbegan with the discovery of leptin. At molecular levels, manyadvances were made, but their integration in a physiologicalperspective is not clear yet. So, from our point of view, it seemsthat, in addition to the coarse control by IGR, many secondaryregulation loops ensure an adaptation finest than sole hormonalmodulation to adapt the various constraints. Two essentialcharacteristics to those loops can be described. First, acting as «virtuous circles » reinforcing a given metabolic situation, anyperturbation of the latter would turn them into « vicious circles »at the basis of a pathological state even if they are primarilyphysiological processes (for example, ketogenesis in responseto insulinopenia in type 1 diabetes is a physiological responseresulting in a pathological state). Next, those are open loops. So,exogenous factors are likely to interfere and turn back thesevicious circles to virtuous ones.

Finally, this model does not take into account the other sideof the energy balance: energy expenditure. Description of thephysiological adaptation to acute exercise suffers of lack ofprecision due to extrapolations from phenomena described atrest. In the same way, training effects on metabolism at rest arepoorly studied and data from literature also shows manyextrapolations. Consequently, we try to develop three points inthis review.

(i) Describe interactions between CHO and lipids metabo-lism at the level of each implied tissue and organ, viahormonal signalization as well as direct action ofnutrients.

(ii) Integrate fluxes of substrates and signals between thosetissues at rest in a global view of the metabolism takinginto account short-term and long-term regulating factors.

(iii) Describe separately, to avoid confusion or extrapolation,the short-term and long-term influence of exercise onthese regulations loops.

2. Adipose tissue

Adipose tissue (AT) is body's energy store. The rate ofNEFA released into circulation depends on the balance betweentheir esterification and lipolysis. For a long time, it wasconsidered a mere storehouse, releasing or storing NEFAaccording to the energetic status. However, more precisedescription of lipolysis, growing knowledge about metabolicheterogeneity of adipocytes according to their topography and,above all, evidence for an endocrine role of AT have made it asthe human body's most voluminous gland.

2.1. NEFA flux regulation balance between lipolysis andesterification

Lipolysis depends on the balance between lipolytic andantilipolytic systems whose central key enzyme is hormone-sensitive lipase (HSL) activity [4]. It catalyzes hydrolysis ofintraadipocyte triglycerides (IATG) to produce NEFA andglycerol. NEFA are transferred to the adipocyte membrane byadipocyte–lipid binding protein (ALBP) whose association toHSL seems crucial to normal lipolysis [5]. NEFA leave throughpassive or facilitated diffusion via specific carriers whereasglycerol is exported through facilitated diffusion only [6].ALBP is an enzymatic cofactor of both lipase and esteraseactivity of HSL, which prevents HSL activity negative feedbackfrom NEFA.

Antilipolytic factors are mainly insulin and catecholamines(via α2-receptors). Lipolytic factors are catecholamines (via β-receptors) and natriuretic peptides ANP and BNP (whose roleseems relevant only during exercise) [7]. Other factors wereidentified but they do not seem physiologically relevant inhumans.

NEFA esterification is under control by lipoprotein–lipase(LPL) activity. LPL is a circulating and ubiquitary enzymecatalyzing the hydrolysis of VLDL to produce NEFA. TheseNEFA enter the cell through passive or facilitated diffusion viafatty acid-binding protein (FABP) or fatty acid translocase(FAT/CD36) [8]. Then, they are activated to acyl-CoA by acyl-CoA synthase. Glucose enters the cell through facilitateddiffusion via GLUT-4 transporters and provides glycerol troughglycolysis. Three glycerols binding to alcohol function of acyl-CoA results in formation of one TG.

LPL regulation deserves a particular attention: insulinstimulates AT-LPL isoform (as part of an anabolic pathwayresulting in energy storage) and inhibits other tissues' LPL

3E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

isoform (as part of a catabolic pathway resulting in ATPproduction) [9]. Moreover, more recent studies demonstratedthat NEFA plasma rate exerts a negative feedback on AT-LPLisoform [10]. De novo lipogenesis, though, does not seem to bephysiologically relevant at rest in human [11].

During the post-absorptive state, NEFA release occursthrough HSL activation caused by an increase in catechola-mines [12] and a decrease in insulinaemia resulting in a reducedLPL activation and, consequently, a reduced NEFA esterifica-tion. During the post-prandial state, NEFA release is reduced forHSL is inhibited through an increase in insulinaemia [13] whoseinduces a reduction in the lipolysis-stimulating action ofcatecholamines [14]. Hyperinsulinaemia also induces NEFAesterification through the activation of AT-LPL isoform.

Besides this fine hormonal regulation, glucose and NEFAplasma rates are likely to have a modulating effect. Negativefeedback modulation of AT-LPL isoform activity by NEFAresults in a balance between the plasma pools of VLDL andNEFA and the pool of IATG. Recent studies also demonstratethat an increase in blood glucose per se can inhibit whole-bodyLPL activity [15].

Many points remain to be clarified, however. FABP- andFAT/CD36-dependent transport of NEFA inside adipocytes isstill poorly studied. A residual lipolysis in a HSL-deficientmouse model suggests the existence of other lipolytic effectorssuch as TAG lipase [16], but their relevance in humanphysiology remains to be established. A paracrine factornamed acetylation-stimulating protein (ASP) was also recentlydiscovered. It stimulates glucose uptake and inhibits HSL,favouring NEFA esterification. Its secretion is stimulated byinsulin and circulating chylomicrons and also seems correlatedto total body fat but in vivo studies of its regulation are needed[17].

Moreover, this is still an incomplete model for it postulatesAT to be a homogenous tissue considering metabolism orsensitivity to regulating factors. As a matter of fact, a

physiological heterogeneity of adipocytes depending on theirtopography is now well established.

2.2. Adipose tissue heterogeneity: subcutaneous and visceraladipocytes

In the late sixties, Jean Vague already noticed that metabolicabnormalities in obese patients are linked to an androidrepartition of AT [18]. Since then, Reaven and others developedthis concept which led to the current definition of the metabolicsyndrome where waist circumference is considered as theessential criterion [19].

In order to clarify, we distinguish in this review subcuta-neous adipose tissue (SCAT) and visceral adipose tissue (VAT).In fact, differences between superficial and deep SCAT are alsodescribed but these differences seem less relevant in aphysiological perspective (Fig. 1).

We also limit to processes described in human AT. Varioustechniques are used, from freshly isolated mature adipocyte tomore or less prolonged incubation of AT explants. Each methodhas its own limits and the available data must be interpretedconsidering specificities of each protocol.

Three main metabolic differences can be described betweenSCAT and VAT. First, VAT accounts for 10–15% of total bodyfat in healthy men [20]. Next, VAT is drained by portal veindirectly to the liver whereas a major part of SCAT is drained bysystemic circulation. Finally, insulin and catecholaminessensitivity is highly different between visceral and subcuta-neous adipocytes.

VAT is less insulin-sensitive and more catecholamine-sensitive than SCAT [21]. As a result, lipolysis is increasedand esterification decreased in VAT compared to SCAT [22].Insulin-induced stimulation of LPL, affinity for ASP and NEFAuptake [23,24] are increased in SCAT compared to VAT. Bycontrast, insulin-induced glucose uptake and de novo lipogen-esis are increased in VAT compared to SCAT [25,26]. This

Fig. 1. Adipose tissue secretions, regulating factors and the VAT/SCAT cycle (NEFA: non-esterified fatty acids; VAT: visceral adipose tissue; SCAT: subcutaneousadipose tissue).

4 E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

mechanism causes an increase in VAT NEFA esterificationcompensating the increased lipolysis [27].

This difference is well explained concerning catecholaminesand depends on the α2R (antilipolytic)/βR (lipolytic) ratio,favouring SCAT lipogenic tendency and VAT lipolytic tendency[28,29].

Then, a SCAT/VAT cycle can be pictured. Lipogenic andantilipolytic SCAT is likely to counterbalance less lipogenic andlipolytic VAT. This balance would depend on the SCAT/VATratio and could regulate NEFA flux, the latter not depending ontotal body fat but on this ratio. Consequently, a significant gainof fat mass would not interfere with NEFA availability if theSCAT/VAT ratio is maintained. If we assume that the level ofNEFA flux to the liver or in systemic circulation is associated tometabolic disturbances, then a dissociation of this ratio cancause either protective or pathological effects. In addition, mostNEFA released by VAT, drained by portal vein, are taken up bythe liver. So, their contribution to the systemic pool of NEFAgetting to skeletal muscle seems reduced [12].

2.3. Adipose tissue: an endocrine gland

More than a storehouse, AT plays a key role in metabolismregulation. NEFA have direct effects on hepatic and muscularsubstrate metabolism and indirect effects through insulinsecretion. Those effects have, in turn, consequences on ATmetabolism. Therefore, some regulation loops can be pictured.Moreover, the discovery of new regulating factors secreted byAT, such as leptin, drastically changed the classical view of ATas a “mere” storehouse. Since then, many factors were identified(leptin, adiponectin, resistin, IL-6, TNF-α, visfatin, …). Two ofthem were more intensely studied and are likely to interfere withenergy homeostasis: leptin and adiponectin.

Leptin is a circulating hormone almost exclusively producedby AT and acting through the long-chain isoform of its receptorOb-Rb. It has direct effects on central nervous system, mainlythrough hypothalamus, inhibiting food intake and increasingenergy expenditure, mainly via the autonomic nervous system[30]. Peripheral effects were also described. At the whole-bodylevel, leptin increases insulin sensitivity, enhances glycogenstorage and decreases intramuscular and intrahepatic TG poolwith a decrease in NEFA esterification and an increase in lipidoxidation [31]. In the liver, it potentializes insulin-inducedsuppression of endogenous glucose production (EGP) [32]. Inskeletal muscle, it increases glucose utilization, thermogenesisand de novo lipogenesis [33,34]. It also directly inhibits insulinsecretion [35]. Its secretion is stimulated by insulin, supposedlythrough the latter's effect to increase glucose utilization andoxidative glucose metabolism in adipocytes [36] whereas it isinhibited by ghrelin, catecholamines and NEFA [37], with themagnitude of this response depending on total body fat. Morenoticeable is the “asymmetric” regulation of leptin secretion byfood intake: caloric restriction causes a more important decreasein leptin secretion than would be expected in regard to thedecrease of total body fat or hormonal responses [38]. So, morethan an adiposity regulator, leptin seems to be an under-nutrition-preventing factor.

Adiponectin is an adipocytokine specific of mature adipo-cyte and is the most abundant transcript [39]. Two isoforms ofadiponectin receptor were identified in skeletal muscle and inthe liver, respectively [40]. Its plasma rate ranges from 5 to30 μg/ml and is correlated to insulin sensitivity [40] but, inopposition with leptin, it is negatively correlated with BMI [41].At the whole-body scale, it increases insulin sensitivity and lipidoxidation [42,43]. In skeletal muscle, it increases lipid oxidationand decreases the IMTG pool [44]. In the liver, it inhibits EGPthrough potentialization of insulin-induced suppression [40,45]but also through an insulin-independent mechanism. It alsoinhibits lipid oxidation and decreases the IHTG pool [40,45].Secretion is heterogeneous considering adipocytes topography.Leptin is strongly correlated to total body fat, although SCATcontributes slightly more to its production than VAT [46]. Bycontrast, adiponectin is significantly produced by SCAT morethan VAT [47].

3. Pancreas

Pancreas has been considered the key regulating organ ofCHO and lipid metabolism for a long time since Banting andBest discovered insulin. Its hormonal secretions act as a signalmodulating every reaction of substrate metabolism in responseto glycaemic variations. A decrease in blood glucose results in adecrease in IGR which stimulates a catabolic pathway ofreactions aiming at using energy stores whereas the oppositeoccurs with an increase in blood glucose.

Molecular mechanisms of insulin secretion were alsoextensively studied and other short- and long-term modulatingfactors were recently identified such as NEFA, incretins andadipocytokines. Similar to the balance between NEFA releaseand uptake by AT, insulin secretion is likely to be regu-lated by finest mechanisms than the only changes in bloodglucose.

Insulin secretion response to NEFA is biphasic. During thepost-prandial state, NEFA potentialize the effects of glucose oninsulin secretion whereas, if NEFA plasma rates are increasedbeyond 12 to 24 h, insulin secretion is inhibited [48]. In aphysiological perspective, this could mean that NEFA inducetheir own storage through a modulation of insulin secretion.However, during the post-absorptive state characterized by arelative lack of glucose, an increase in NEFA has no effect oninsulin secretion in subjects with normal glucose tolerance [49].Few research teams now seek to describe the mechanismsunderlying the effects of NEFA on insulin secretion. The recentdiscovery of the pancreatic receptor for NEFA called GPR40could be one of those [50].

Another indirect mechanism could imply the incretinglucagon-like peptide-1 (GLP-1), whose synthesis is stimulatedby glucose and lipid binding to gut receptor GPR120 [51]. It isof recent interest to study the role played by intestinal hormoneson substrate metabolism and this was emphasized with thetherapeutic use of GLP-1 in diabetes [52]. Describing theireffects on energy, glucid and lipid homeostasis could result in abetter understanding of the link between meal composition andpost-prandial metabolism.

5E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

At this point, we can already describe a short-term regulationloop between pancreas and AT during the post-prandial state(Fig. 2). NEFA potentialize the blood glucose-induced insulinsecretion, which inhibits lipolysis and stimulates LPL (modu-lated through negative feedback control by NEFA plasma rate).So, the decrease in NEFA causes a decrease in its effect oninsulin secretion. There is also a long-term regulation loopbetween pancreas and AT called adipo-insular axis (Fig. 2).Insulin increases leptin secretion which, in turn, inhibits insulinsecretion. Still in a long-term fashion, the lipogenic effect ofinsulin tending to increase total body fat also contributes toamplify leptin secretion rate. So these two loops act for the samepurpose: adapting insulin secretion to total body fat. As theadipo-insular axis is a long-term regulation loop, one canspeculate that it also regulates basal leptin concentration which,in turn, influences basal insulin concentration to adapt totalenergy stores.

4. The liver

Considering that the kidney plays a minor role, the liver isthe only organ capable of releasing glucose in systemiccirculation in order to ensure availability for brain and for redblood cells. During the post-prandial state, the liver can storeglucose as glycogen (glycogenogenesis). During the post-absorptive state, it can release glycogen into circulation asglucose (glycogenolysis) or synthesize glucose from glycogenicor nonglycogenic precursors (gluconeogenesis). Endogenousglucose production (EGP) is the flux of glucose produced andreleased into circulation from glycogenolysis and gluconeogen-esis and it is regulated, according to the classical model, by IGR.EGP regulation was more precisely described during the last10 years with a focus on various potential regulators such asCHO, lipids, insulin, glucagon, leptin and adiponectin. Theconcept of EGP autoregulation was also suggested in the early

nineties. Based on these elements, the liver also seems to be partof complex regulation loops whose key organ is no longer thepancreas.

4.1. Intrahepatic lipid metabolism

During the post-absorptive state, NEFA are the mainsubstrate for the intrahepatic triglycerides (IHTG) pool. Nearly30% are taken up by splanchnic tissues and directed toward β-oxidation [53] and, sometimes, ketogenesis. Allosteric regula-tions inhibit NEFA esterification and de novo lipogenesis: adecrease in citrate and insulin causes a decrease in acetyl-CoAcarboxylase-1 (ACC-1) and, thus, a decrease in malonyl-CoA.Consequently, gluconeogenesis is stimulated as an increase inacetyl-CoA and glucagon causes inhibition of pyruvatedehydrogenase.

During the post-prandial state, NEFA are still the main IHTGpool substrate but chylomicrons also have a contribution [54].De novo lipogenesis is no more physiologically relevant inthe liver than in AT. A very flexible system allows the liver toregulate IHTG pool and, consequently, VLDL secretiondepending on meal composition. A carbohydrate-rich mealincreases de novo lipogenesis whereas contribution of chylo-microns is decreased. The exact opposite is described with alipid-rich meal. Contribution of NEFA does not change: if thecritical increase in insulin following the carbohydrate-rich mealinhibits lipolysis, the decrease in circulating NEFA iscompensated through a negative feedback-activation of LPL.The same reaction occurs with a lipid-rich meal [55]. Thisstrong regulation of the NEFA flux getting to the liver is inaccordance with their role as regulating factors of livermetabolism.

4.2. Regulation of EGP: an adipo-hepato-insular loop?

EGP is stimulated by glucagon and catecholamines whereasit is inhibited by glucose and insulin. An increase in NEFAresults in an increase in EGP if insulin rates are low [56].Insulin-induced suppression of EGP was demonstrated to resultfrom a predominant decrease in glycogenolysis. Now, if insulinsecretion is normal, an increase in NEFA results in an increasein gluconeogenesis and a decrease in glycogenolysis whereasEGP remains equal. So, an increase in NEFA inhibits insulin-suppression of glycogenolysis [57,58]. NEFA effects on livercarbohydrate metabolism lie in their interaction with insulinaction. As NEFA also potentialize the effect of glucose oninsulin secretion, a short-term regulation loop between AT,pancreas and the liver can be pictured (Fig. 3).

During the post-absorptive state when insulin levels are low,NEFA can stimulate EGP and, in synergy with this flux ofglucose, tend to increase insulin secretion. The latter will, inturn, tend to inhibit EGP directly and through insulin-induceddown-regulation of lipolysis. Moreover, if blood glucose tendsto increase, this tends to decrease EGP.

During the post-prandial state, an increase in blood glucoseinhibits EGP and stimulates insulin secretion which, potentia-lized by a post-prandial increase in NEFA, results in an increase

Fig. 2. Long-term regulations between pancreas and adipose tissue: the adipo-insular axis (NEFA: non-esterified fatty acids).

6 E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

in the latter's esterification and, consequently, in a decrease intheir plasma rate and in their inhibiting effect on insulin-inducedsuppression of EGP.

Bergman and other authors even considers the direct effect ofinsulin on hepatocytes not fast enough to explain the metabolicchanges and suggests that insulin-induced suppression ismediated by the NEFA flux [59]. However, Fisher showed ina model of hepatic insulin-receptor-deficient mouse that insulin-induced suppression of EGP is inhibited despite a decrease inNEFA but a significant role of NEFA in the modulation of thisphenomenon cannot be excluded [60]. From our point of view,NEFA and insulin act in synergy and provide a fine-tuningregulation of EGP.

Hepatic autoregulation is thought to be aimed at compensat-ing an inappropriate increase in gluconeogenesis by a decreasein glycogenolysis [61]. In a physiological perspective, it seemsrather logical that the organism would spare glycogen if de novosynthesis of glucose is possible. However, if this mechanismwas reproduced in vivo, its regulation is still subject tocontroversy. For a growing number of authors, insulin wouldbe the key regulating factor. As a matter of fact, EGP isincreased when insulin levels are low. More recent studiesdemonstrated that a hypercaloric hyperglucidic diet resulting inan increase in glycogen stores can override hepatic autoregula-tion [62]. So, the latter appears to be a way to spare glycogenstores more than a fine-tuning regulator of EGP.

4.3. An adipo-hepato-insular loop of long-term regulation

Leptin and adiponectin increase hepatic lipid oxidation, tendto decrease IHTG pool and potentializes insulin-inducedsuppression of EGP [16,63]. Adiponectin also inhibits EGP inan insulin-independent way [64].

A regulation loop whose purpose would be the modulationof NEFA action and the protection against an increase in totalbody fat can be pictured (Fig. 4).

If total body fat is low, circulating leptin is expected to below and circulating adiponectin high. As they both increasehepatic lipid oxidation and decrease EGP, the latter does nottend to stimulate insulin secretion.

Considering increasing total body fat, the proportionbetween adiponectin and leptin is inversed. During the post-absorptive state, as adiponectin – which has a higher potentialof suppressing EGP than leptin – decreases, blood glucosetends to increase more, which results in higher insulin secretionbut leptin also inhibits insulin secretion and, thus, its lipogeniceffect. During the post-prandial state, when blood glucose andinsulinaemia increase, so does leptinaemia. This helps insulin tosuppress EGP without too much an increase. This loop, whichprevents hyperinsulinaemia, can truly be considered as aprotection against a vicious circle of increasing insulinaemia/increasing body fat.

5. Skeletal muscle

Skeletal muscle is the main energy user of the organism. Ittakes up 50% of NEFA during the post-absorptive state and 75–80% of glucose during the post-prandial state. So it can beconsidered as a major regulator of substrate metabolism andtheir plasma availability.

In healthy subjects, it has a unique ability to use the mainsubstrate (i.e. NEFA during the post-absorptive state andglucose during the post-prandial state) called ‘metabolicflexibility’ [65].

Mechanisms of this flexibility were extensively studied theselast 20 years. IGR and substrate availability do not seem to bethe major regulating factors as mechanisms of competition weredemonstrated to take place inside skeletal muscle.

In their previous article, Randle et al. demonstrated that, inthe resting muscle, an increase in lipid oxidation results in adecrease in glucose utilization, opening the debates around the

Fig. 3. Short-term regulations between adipose tissue, pancreas and liver: anadipo-hepato-insular loop (NEFA: non-esterified fatty acids; EGP: endogenousglucose production).

Fig. 4. Interactions between short-term and long-term regulation factors betweenpancreas, liver and adipose tissue: another adipo-hepato-insular loop (EGP:endogenous glucose production).

7E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

glucose/fatty acid cycle [3]. About a decade later, Winders and,later, McGarry demonstrated what would become ‘Randlereverse’: high glucose and insulin concentrations can suppressfatty acid oxidation [66].

From now on, we only refer to skeletal muscle as ‘muscle’but it has to be noticed that some of the reactions studied hereare aimed at storing energy for a potential exercise. Physiologyof exercising and trained muscle will be reviewed further.

5.1. Substrates flux and storage: glycogen and intramusculartriglycerides

NEFA getting to the muscle come from the plasma pool ofNEFA and hydrolysis of VLDL by muscular LPL. They enterthe muscular cell using FABP and FAT/CD36 transportersthrough facilitated diffusion. This flux is highly dependent upontheir plasma concentrations [67]. During the post-absorptivestate, when their plasma rate is high, 50% of circulating NEFAenter the muscle [68]. During the post-prandial state, insulinstimulates their uptake by FAT/CD36 [8]. However, in the sametime, insulin inhibits lipolysis and muscular LPL, significantlydecreasing their plasma availability and uptake from VLDL.Thus, this insulin-dependent intramuscular NEFA uptakeappears to be irrelevant, at least in the resting muscle.

Once inside, NEFA are activated to fatty acyl-CoA by acyl-CoA synthase. 50% of fatty acyl-CoA are directly oxidized andthe other 50% are directed toward esterification and enter theIMTG pool [69]. The latter received growing interest these lastfew years since evidence of a correlation between size of theintramuscular triglycerides (IMTG) pool and insulin resistance.However, no causal relationship could be established. IMTG aresubmitted to a relevant and constant turnover estimated at 29 h[69]. Like AT, the size of the IMTG pool depends on the balancebetween lipolysis and esterification, which is noticeablyregulated in normal conditions as fasting-induced increase inNEFA availability has no effect on its size [69]. Other factorsare known to influence IMTG content: insulin increases IMTGpool as it inhibits muscular lipolysis and stimulates esterifica-tion [70]. Leptin has opposite effects as it stimulates lipolysisand inhibits esterification [30]. In a physiological perspective,the IMTG pool size and turnover rate have no values inthemselves. They have to be considered in regard with lipidoxidation rate. As a matter of fact, the ratio between IMTGturnover and lipid oxidation is a significant determinant ofintramuscular NEFA availability and concentration. Thus,insulin and leptin would not have such an influence on thelatter since insulin simultaneously decreases this turnover andlipid oxidation rate whereas leptin has the exact opposite effect.However, the relative effects of these two hormones on eachside of the balance remain to be quantitatively studied. Bycontrast, catecholamines increase muscular lipolysis without aninfluence on lipid oxidation [71].

In pathophysiology, an increase in IMTG can be seen as anaccumulation resulting from a decreased ability of muscle tooxidize lipids but, in physiology, it can also be seen as anadaptation of muscle to repeated and increased needs in lipids[72].

Glucose enters the muscle mainly through an insulin-dependent mechanism: GLUT-4 translocation to cell surface.It can be potentialized by muscular contraction but this isdiscussed further as we currently only consider muscle in theresting state. A concentration-dependent transport also existsbut he is of minor physiological relevance. Once inside the cell,glucose is phosphorylated to glucose-6-phosphate (G6P) anddirected toward muscular glycogenogenesis or toward glyco-lysis. Glycogenogenesis depends on the balance betweenglycogen synthase and glycogen phosphorylase. Glycogen-ogenesis is stimulated by insulin, intracellular G6P concentra-tion, energy demand and intramuscular glycogen concentration.It is aimed at ensuring a sufficient flux of glucose duringexercise. It takes place during the post-prandial state whenglucose uptake generally exceeds the immediate needs of theresting muscle (excepted after depletion of the glycogen storesdue to a recent and intense exercise). Extrapolating this data,one can consider the ability of storing glycogen as another wayof regulating blood glucose.

5.2. Intramuscular interactions: Randle upside-down

In a recent review [73], Frayn noticed: “the coarse control ofreciprocal utilization of glucose and NEFA in the body isbrought about through insulin secretion and fine-tuning isprovided by skeletal muscle”. Since Randle's first publications,these mechanisms have been extensively studied and anelegant system of interactions suggesting that the relationshipbetween carbohydrate and lipid metabolism is indeed reci-procal and independent, even though the latter is now debated(Fig. 5).

The glucose/fatty acid cycle is now well established in vivoin human muscle. To explain the underlying mechanisms,Randle had hypothesized a system of allosteric interactions:excess of acetyl-CoA produced by β-oxidation inhibitspyruvate dehydrogenase and excess of citrate inhibits phospho-fructokinase. This results in a decrease in glycolysis, intracy-tosolic accumulation of G6P and, finally, a decrease inhexokinase and in glucose uptake [3]. Nevertheless, recentstudies, more than lipid oxidation incriminate NEFA concen-tration (more precisely acyl-CoA and diacylglycerol (DAG))which, through activation of an atypical protein kinase-C(PKC), results in serine/threonine phosphorylation of IRS-1[74]. This causes PI3K-dependent signal inhibition and, thus,inhibits GLUT-4 translocation. So, to summarize, an increasedconcentration of intramuscular acyl-CoA and DAG causesinsulin-dependent glucose uptake inhibition.

Winders', and later McGarry's description of ‘Randlereverse’ cycle brought a decisive element to the debate. Hedemonstrated that an increase in blood glucose and insulinaemiadecreased hepatic lipid oxidation by inhibiting their intrami-tochondrial transport [66]. Once energy demand is satisfied,glucose in excess is converted to malonyl-CoA by ACC, itselfactivated by citrate, glucose and insulin. Malonyl-CoA is themain precursor of de novo lipogenesis but it is also a powerfulallosteric inhibitor of carnitine palmitoyl transferase-1 (CPT-1)which catalyzes intramitochondrial transport of acyl-CoA,

8 E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

regarded as a key limiting step of lipid metabolism [75]allowing them to enter β-oxidation. Two isoforms of ACC areidentified [76]: ACC-1 is predominant in lipogenic tissues andhas a cytosolic localization. Cytosolic pool of malonyl-CoA isbelieved to serve the purpose of de novo lipogenesis [34]. ACC-2, predominant in muscle, has preferential juxtamitochondriallocalization close to CPT-1 and malonyl-CoA produced by thisisoform is believed to be the key inhibitor of muscular CPT-1,more sensitive to the inhibitor effect of malonyl-CoA than itslipogenic tissues homologue [34]. However, it has to be kept inmind that this inhibiting effect of malonyl-CoA is also observedin the liver and that recent studies demonstrated de novolipogenesis in muscle [34].

5.3. AMP kinase: at the crossroad of modulations

AMP kinase (AMPK) was named after its ability to beactivated in response to energy depletion (an increase in AMP/ATP ratio). To simplify, it stimulates catabolic reactions(leading to ATP production) and inhibits anabolic reactions(leading to ATP utilization). Its inhibiting effect on ACCactivation decrease acetyl-CoA conversion to malonyl-CoA andthe latter's inhibition of CPT-1, which leads to an increase infatty acyl-CoA flux and oxidation into mitochondria [77]. It alsoinhibits NEFA esterification and de novo lipogenesis [34].About CHO, it stimulates glucose uptake and glycolysis [77]and inhibits glycogenogenesis with phosphorylation of glyco-gen synthase [78].

5.4. Leptin, de novo lipogenesis and thermogenesis: not soindependent interactions?

These last few years, other studies also demonstrated thatAMP-independent mechanisms [79], such as hyperosmoticstress and leptin [80], can also activate AMPK. This suggeststhat the latter is not only the cell regulator of energy balance.Leptin production is increased during the post-prandial state,stimulated by insulin and glycolysis and its secretion rate is

proportional to total body fat. So, it is fully active in situationswhere substrates (lipid stores and circulating glucose) areabundant.

Leptin and adiponectin potentialize insulin-dependent glu-cose uptake and tend to lower IMTG pool size with aninhibition of NEFA esterification and a stimulation of lipidoxidation. Leptin effects on lipid metabolism need AMPKactivation and its effects on glucose metabolism need normalPI3K-dependent signaling [34].

So it seems that leptin plays a role in stimulatingthermogenesis. This is well established as well as the fact thatthese effects are not found after suppression of de novolipogenesis or lipid oxidation.

According to recent studies by Duloo et al., de novolipogenesis also occurs in muscle, yet a non-lipogenic tissue.They also demonstrated that hyperglycaemia causes an increasein IMTG and that this increase results from de novo lipogenesiswithout change in lipid oxidation but with a decrease in glucoseuptake and glycogenogenesis [34].

6. From virtuous to vicious circles

At this point, all these data are in accordance with a“modern” vision of substrate metabolism: a coarse control byIGR finely tuned by secondary regulating loops reinforcingcatabolic or anabolic effects depending on the nutritional state.Thanks to these loops, the constraints are matched with anoptimal management in terms of energy balance.

SCAT/VAT cycle determines a range of plasma NEFAavailability. Adipo-insular cycle, through short-term regulationby NEFA or long-term regulation by adipocytokines determinesa range of insulinaemia and, thus, leptinaemia. Extending thiscycle to the liver as an adipo-hepato-insular regulation loopemphasizes the synergy between IGR and nutrient sensing: in ashort-term perspective, it modulates CHO availability depend-ing on NEFA availability and vice versa. In a long-termperspective, it modulates total body fat. All these regulationloops are aimed at modulating substrate availability and, thus,

Fig. 5. Intramuscular interactions during rest in human skeletal muscle (NEFA: non-esterified fatty acids; DAG: diacylglycerol; PKC: protein kinase-C; ACC: acetyl-CoA carboxylase, CPT-1: carnitine palmitoyl transferase-1; IMTG: intramuscular triglyceride).

9E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

plasma concentrations. In muscle, Randle's cycle (the originalglucose/fatty acid cycle) and its counterpart, Randle Reverse,act as final regulators of the balance of substrates utilization asthey determine metabolic flexibility.

Substrate availability is the major point and this is why theterm ‘virtuous circle’ can be used: if the organism considers as apriority to maintain plasma concentrations of these substrates inphysiological ranges, it is not only a matter of protection againstcomplications such as hypoglycaemia, atherosclerosis ormicroangiopathy but also because… physiological concentra-tions are self-maintaining. In other words, as long as CHO andNEFA concentrations remain in a defined range of values, theorganism has a strong ability of regulation but would it fail for acertain amount of time, all these regulation loops would go onplaying their role of reinforcing current metabolic conditions asthey consider it is strongly ensured by IGR. So, self-maintainingpathological conditions yet just doing their job, they can indeedbe named “vicious circles”.

Unfortunately, the finer and more complex the mechanism,the more weak points, the more frequent the breakdown! Manyweak points certainly exist. Among them, VAT/SCAT ratio,insulin secretion and leptin sensitivity are the best known. Thisis linked to the second characteristic of these circles: these areopen loops! Independently of genetic mutations, environmentalfactors can interfere through any of these weak points.

For example (Fig. 6), if the VAT/SCAT ratio is increased,whatever the reason, the increase in NEFA availability causesan increase in insulin secretion directly and through an increasein EGP. NEFA availability can be normalized throughhyperinsulinaemia but this favours CHO uptake and utilizationin muscle and, consequently, causes a decrease in lipidoxidation during the post-absorptive state. At a given point,this leads to an increase in NEFA plasma concentrations evenduring the post-prandial state and, then, despite hyperinsuli-

naemia, muscle uses less CHO than it should, due to glucose/fatty acid cycle and more insulin is needed to maintain bloodglucose. Metabolic inflexibility and adipo-hepato-insular-induced hyperinsulinaemia already act as vicious circles. But,if NEFA concentrations remains high for more than 12 h,insulin secretion can be attenuated, though not leading toglucose tolerance trouble but increasing metabolic inflexibilitywith a net decrease in CHO utilization in the post-prandialstate. So, substrate plasma concentrations can still be regulated:the price is hyperinsulinaemia and, unfortunately, its cardio-vascular consequences. We will not detail further butconsequences are far more serious if insulin secretion cannotmatch the increasing demand as hyperglycaemia is up to occur,leading to the complete metabolic inflexibility with muscle notusing enough NEFA during the post-absorptive state because ofhyperglycaemia and not using enough CHO during the post-prandial state because of elevated NEFA concentration and, so,elevated NEFA concentrations and hyperglycaemia are self-maintained. As far as we know, this leads to another well-known vicious circle taking place in the pancreas calledglucotoxicity.

To summarize, metabolic flexibility depends on physiologi-cal substrate plasma availability and metabolic inflexibility canoccur when this availability reaches pathological values.Fortunately, as we said, these are open loops. It means that if,on one side, pathogenic factors can interfere, on the other side,so can therapeutic factors. If metabolic diseases therapy isaimed at restoring metabolic flexibility and physiologicalsubstrates concentrations, they could use factors such asnutritional intervention, drugs or physical activity. The secondpart of this review will be focused on the latter and themechanisms which can be potentially used to counteract theinitial pathogenic factors and turning vicious circles back tovirtuous ones.

Fig. 6. A vicious circle: effect of an excess of visceral adipose tissue on an adipo-hepato-insulo-muscular loop of regulation (NEFA: non-esterified fatty acids; VAT:visceral adipose tissue; SCAT: subcutaneous adipose tissue; EGP: endogenous glucose production).

10 E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

7. Physical activity: breaking vicious circles… or bypassingthem?

Physical activity also plays a key role in energy balanceregulation as it accounts for about 30% of total energyexpenditure.

Studies of physical activity and its effects on metabolismmust be cautiously considered. First, a distinction between theeffects of acute exercise and repeated exercise (training)inducing long-term adaptations must be made. Exercise mustbe defined by parameters such as intensity and duration and itincludes a time of recovery. A significant repetition of exercisesis defined by its frequency and induces long-term adaptationsinteracting with metabolism during exercise as well as duringthe resting state. So the body must face new constraintscomparing to the resting state.

First constraint is about chemical energy which must besupplied to muscle to be converted into mechanic energyallowing performance.

Second constraint is about blood glucose which must bemaintained in ranges allowing both sufficient supply to themuscle and prevention of hypoglycaemia and hyperglycaemia.

Third constraint is about muscular and hepatic glycogen,whose availability is highly correlated to performance andwhich must be spared as long as possible.

Fourth and last constraint lays on the fact that mobilization ofthe various energy sources is not equally fast. As energydemand increases and fast supply is needed, the body has to usethe most rapidly available substrates.

In addition to these constraints, Nature tends to optimize theenergetic cost of these constraints in term of energy cost, whichleads to long-term adaptations in order to get a compromisebetween performance and substrates use.

Considering the data of literature, it is noticeable that theyoften show discordant results. Two major facts are likely toexplain this: first, protocols designed for these studies usedifferent frequencies, durations, intensities, exercise types andsubjects are not always matched for their sex, age or trainingstatus. Next, diet habits, an important bias, are not always takeninto account. Finally, phenomena well established in the restingstate are sometimes extrapolated to the exercising state.

Except for particular situations, protein contribution toexercise-induced energy expenditure is negligible. So we willconsider CHO and lipids as the two main substrates duringexercise. Their respective contribution to total energy expen-diture must be finely regulated in order to face the variousconstraints. We will name this respective contribution substratebalance. Of course, regulating mechanisms differ from theresting state, as do constraints.

So, considering that intertissular interactions and substratesfluxes are, at least partially, different between rest and exercise,one can assume that an interference with regulation loops andvicious/virtous circles of the resting state can occur duringexercise and/or training.

Now, we will focus on the most recent advances concerningthis matter. First, we will describe the contribution of each organand tissue to the metabolic adaptation during exercise with the

specific influence of duration and intensity. Next, we willdescribe the long-term influence of training status on metabo-lism at rest and during exercise.

8. Exercise

8.1. Same actors playing different parts

Exercise implies circulatory, respiratory, hormonal andmetabolic adaptations in order to achieve sufficient supply ofenergy substrates and oxygen to mitochondria. In the rest of thetext, we only consider hormonal and metabolic responses.

Regarding metabolism, exercise has three different stages:starter, exercise itself (with a given duration and variousintensities) and recuperation.

CHO used during exercise are blood glucose, muscular andhepatic glycogen and, as described further, lactate. Lipids areplasma NEFA, VLDL, adipocytes TG and IMTG.

Duration and intensity are the two intrinsic parameters of agiven exercise which influence the substrate balance. Tosimplify, the relative contribution of lipids used increases withduration whereas the relative contribution of CHO increaseswith intensity [81]. Practically, because very intense exerciseshave a short duration, two main situations are observed: mildand moderate-intensity exercise whose duration can be quitelong and short high-intensity exercise.

Each individual has its own ability in term of exerciseintensity, expressed as a percent of maximal oxygen consump-tion capacity (VO2max

) which allows a more precise interindivi-dual comparison than absolute performance does. So, it isadmitted that intensity is mild under 35% VO2max

, moderatebetween 40% and 70% VO2max

and high above 75% VO2max. Of

course, this increase in intensity is a continuum and these valueswere set arbitrarily. Consequently, studies with the samehypothesis can get two different conclusions if one used 40%VO2max

and the other 60% VO2max.

One essential constraint, as we already said, is the absoluteneed to maintain an energetic substrates flux at least equal totheir utilization speed rate. Now, aerobic metabolism has alimited capacity, which defines VO2max

. Beyond a given intensity,substrates oxidation rates do not match energy demand. That iswhy a system which can produce energy from glycolysispyruvate exists. This system, capable of producing energy fasterthan substrate oxidation and in an oxygen-independent way isanaerobic metabolism. Its final metabolite is lactic acid (lactate),whose utilization at rest in negligible (except when tissuehypoxia occurs). Its significant production during high-intensityexercise can induce new interactions considering it is acarbohydrate and a weak acid. Coexistence of aerobic andanaerobic pathways also explains the different muscle fibretypes: “slow” type 1 fibres, which are mainly oxidative and“fast” type 2a and 2b fibres, which are mainly glycolytic.Nevertheless, considering the whole muscle, CHO metabolismenzymes are more abundant than lipid metabolism enzymes.

This is one more point about complementarity of substrates.Lipids stores are abundant but their mobilization is slow andthey provide energy only through oxidative metabolism.

11E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

Glucose stores are less abundant and the constraints ofmaintaining blood glucose (for survival) and of sparingglycogen stores (for performance) must be matched. However,their mobilization can be achieved faster. Finally, intramuscularsubstrates can be mobilized faster than extra-muscular ones,whatever their nature.

If we consider all these data, we partially explain the fact thatCHO are the main substrate during high-intensity exercise aswell as during exercise initiation when energy used for a“starter” effect must be quickly provided. By contrast, the factthat CHO are a minor substrate during mild- and moderate-intensity exercise reflects the organism ability to keep this“safety exit” in case of absolute need and to maintain, as long aspossible, the use of a substrate whose plasma concentration doesnot condition survival and whose stores are the most abundantand not correlated to performance, i.e. lipids.

Now we will discuss the factors regulating substrate balanceduring exercise and describe how they can allow adaptation toexercise conditions and match the different constraints. Diethabits, nutritional and training status will not be taken intoaccount yet in this part of the text. Influence of the latter will bediscussed further.

8.2. An overview

Two mechanisms resulting from exercise seems to initiatemetabolic adaptations: muscular contraction and catechola-mines production [81]. They cause a decrease in insulinsecretion and a variable increase in glucagon productionthrough neural afferences and α-adrenergic action.

Then, energy substrates availability is increased with anincrease in EGP and glucose uptake whose synchronizationleads to a remarkable maintain in blood glucose. Lipolysis,NEFA uptake and muscular glycogenolysis are also activated.IMTG metabolism during exercise is still an object ofcontroversy.

From this overview, two major points must be pointed out.First, this situation is physiological but the major differencewith the resting state is maintained blood glucose and anincrease in glucose uptake in spite of a decreased insulinsecretion. This implies the existence and action of CHOmetabolism regulators other than insulin. Next, the change inbalance of substrates during exercise also implies interactionsbetween producing and regulating tissues and a fine-tuningregulation inside muscle.

8.3. Of catecholamines

By which way catecholamine secretion and sympatheticnervous system are activated is still debated and probablymultifactorial. Whether this is a nervous signal coming frommuscle, or glycaemic changes or anticipation from centralnervous system, catecholamines are increased during exerciseand are considered a major regulator of metabolic adaptation[81].

They are catabolic hormones, reflecting an increased energydemand. They induce catabolic reactions such as lipolysis in AT

(with β-adrenergic signal overtaking α-adrenergic signal) and,more recently demonstrated, in muscle [82]. They also inhibitinsulin secretion through α-adrenergic signal [83] and stimulatehepatic and muscular glycogenogenesis. They also havedifferential effects depending on their concentration. As plasmaepinephrine rises, it induces first lipolysis, then glycogenogen-esis and finally suppresses insulin action [84]. Norepinephrinestimulates EGP. So, theoretically, they are responsible forsubstrate mobilization from their various sources, directly aswell as acting on pancreatic secretions. However, this has to bebalanced with more recent studies taking into account intensity.During mild- or moderate-intensity exercise, catecholamines arepoorly increased (up to fourfold the basal rate). So, in this case,they are not expected to have a significant effect on lipolysisand EGP. By contrast, during high-intensity exercise, their raterises up to 18-fold compared to resting state, which makes themmore likely to be physiologically relevant [85].

Moreover, they are quickly secreted is high and their half-lifeis short. Considering that, during high-intensity exercise, energymust also be quickly supplied, it is logical to find them exertinga direct effect where an indirect signaling pathway would taketoo much time.

8.4. Pancreas: of insulin secretion blockade

Insulin secretion is decreased during exercise throughLangerhans islets sympathetic innervation and increasedcatecholamines concentration [83]. This causes IGR to decreasewhich creates a “catabolic” climate where CHO and lipids canbe mobilized.

The antilipolytic effect of insulin is decreased and NEFAflux to the muscle is increased, matching muscular energydemand. This decrease in IGR also causes a decrease in EGPsuppression and glucose flux to the muscle is increased as well[86].

When exercise duration is high, although plasma NEFAconcentration is increased and blood glucose is maintained at aconstant rate, insulin secretion remains inhibited. In otherwords, not only can nutrient-regulation of insulin secretion beovertaken by insulin-secretion inhibiting factors during exercisebut other mechanisms, distinct from insulin, can mimic itseffects on intramuscular glucose uptake. Nevertheless, glucosestill increases insulin secretion as the latter appears lessdecreased after a carbohydrate intake before or during exercise[87].

The increase in glucagon is significant only during a high-duration or high-intensity exercise. Its regulation seems todepend mainly on glycaemic changes [88].

During recovery, catecholamines rapidly decrease andmaintained EGP causes blood glucose to be high, which leadsto an increase in insulin secretion and a decrease in glucagonsecretion [89,90]. This increase in IGR creates an “anabolicclimate” favouring lipids and CHO storing.

As exercise duration and/or intensity increases, the decreasein IGR appears to be linear. Thus, this variable can beconsidered as a predominant regulation system when intensityis moderate and duration is high. However, during high-

12 E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

intensity exercise, this decrease is less significant and hardlylinear, mainly due to a less marked decrease in insulin secretion[85].

This may be related to our discussion about catecholamines.As long as adrenergic signaling, in conjunction with otherfactors, can « take enough time » to be mediated throughpancreatic hormonal secretions to mobilize substrates whichthemselves take a significant amount of time to supply energy,this optimal management in term of energy can occur. Butwhenever energy supply speed rate does not match energydemand, catecholamines are then secreted, exerting less finelyregulated but more direct and, thus, faster effects.

8.5. Adipose tissue: of lipids mobilization

Lipolysis is increased during exercise, with NEFA releasedat a rate two- to fivefold increased compared to basal rate [53].NEFA esterification is simultaneously inhibited.

This can be explained with the decrease in the antilipolyticeffect of insulin and the predominantly lipolytic effect ofepinephrine. Recent studies have also demonstrated the lipolyticeffect of ANP and BNP in animal models and in human in vivo[7,91,92]. Their regulation remains to be studied but their effectis believed to be independent from nutritional status and tocompensate a decrease in catecholamines activity [91,92]. Otherfactors might play a role in HSL regulation during exercise butavailable data are still unclear and in vivo studies in human arelacking.

AT is not metabolically homogeneous. So, NEFA are mainlyprovided by SCAT. However, as VAT has an increased lipolyticactivity compared to SCAT [21–27], one can hypothesize that itis preferentially depleted during exercise. Moreover, anotherdistinction is now established: lipolysis activity during exerciseis higher in abdominal SCAT than in femoral or gluteal SCAT[53].

Duration of exercise influences lipolysis activity: after 4 h ofmoderate-intensity exercise, it can be increased by tenfoldcompared to basal rate [93]. By contrast, as intensityprogressively increases, lipolysis is unchanged but rate ofNEFA released into circulation is decreased. This was attributedto a decrease in AT blood flow due to adrenergic vasoconstric-tion [94].

So, NEFA availability seems regulated by the synergic actionof catecholamines, insulin and, but it has to be precise, ANP andBNP. But they are not the only lipidic substrate, as VLDL andIMTG are also likely to provide energy to muscle. This pointwill be discussed further.

8.6. The liver as a store and a source of CHO

EGP is increased during exercise, with a four- to fivefoldincrease compared to basal state [95]. As the result ofglycogenolysis and gluconeogenesis it directly depends onhepatic glycogen stores and gluconeogenic precursor's avail-ability. It is quite remarkable that, in most situations, EGP isfinely correlated to glucose uptake resulting in almost perfectlyconstant blood glucose.

As exercise duration increases, gluconeogenesis progres-sively becomes the major contributor to EGP, which can beviewed either as a way to spare glycogen stores or as a reactionto match glycogen depletion.

As exercise intensity increases, i.e., when energy must bequickly provided to mitochondria, glycogenolysis tends back tobe the major contributor to EGP. One must notice that, at thevery beginning of any exercise, when a fast energy supply isrequired for a short amount of time, glycogenolysis contributionto EGP predominates over gluconeogenesis.

During recovery, glycogen storage occurs and is correlated tototal energy expenditure and glycogen stores levels butregulating factors are still unknown.

Once again, other regulation loops can be pictured.During moderate-intensity exercise, sympathetic nervous

system and catecholamines both inhibit insulin secretion, whichresults in a rise in EGP through a decrease in insulin-inducedsuppression of gluconeogenesis and glycogenolysis. Catecho-lamines and decreased insulin secretion, as well as ANP andBNP secretion, result in lipolysis and NEFA flux to the livertends to increase gluconeogenesis. The resulting increase inblood glucose inhibits EGP directly and through a minor butsignificant stimulating effect on insulin secretion possiblypotentialized by NEFA. However, physiological relevance ofthis mechanism, significant at rest, remains to be demonstratedduring exercise.

As exercise duration increases, glucagon secretion gets tosignificant levels and, in synergy with NEFA flux, tends toincrease gluconeogenesis and decrease glycogenolysis. Acarbohydrate-induced increase in insulin secretion before orduring exercise tends to decrease EGP. Anyway, lipid andexogenous CHO availability results in hepatic glycogen storessparing. Most of these effects occur through pancreatichormones secretions and catecholamines do not play significantrole in CHO homeostasis.

During high-intensity exercise, IGR effects do not achieveenergy demand anymore and catecholamines, through theirimportant increase, become the main regulators. Marliss et al.showed a catecholamine-induced increase in EGP up toeightfold compared to basal state (with glycogenolysis asmajor contributor) but, as they inhibit glucose uptake, the latterwas only increased three- to fourfold compared to basal state[85]. This resulted in hyperglycaemia. During recovery, theassociation of a hyperglycaemia with a decrease in catechola-mines results in a net increase in insulinaemia. This restores ananabolic climate and, thus, glycogen storage, highly depleted atthis intensity.

Glucose homeostasis is well regulated during exercise butthe model we described considers muscle as a passive substratesextractor and user depending on extra-muscular signals. In fact,this remarkable coordination between EGP and glucose uptakeis not only due to catecholamines, sympathetic nervous systemor pancreatic hormones. Recent data suggested that lipidmetabolism, intramuscular interactions between CHO andlipids and other factors including IL-6 are part of a “workfactor” [96] modulating the balance between EGP and glucoseuptake.

13E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

8.7. Muscle: consumer and regulator

At the whole-body scale, hormonal responses and substratesmobilization are influenced by exercise intensity level. Thesame can be said at the muscle scale. During mild- to moderate-intensity exercise, lipid oxidation is the major catabolic reactionand IGR is the major regulating factor, especially if exerciseduration is high. At intensity lower than 30% VO2max

, energy isalmost exclusively provided by plasma NEFA and neitherintramuscular substrates nor blood glucose are used. Atintensity between 50% and 70% VO2max

, half the energy isprovided by fatty acids (25% from plasma NEFA and 25% fromIMTG) and the remaining is provided by CHO (12% frommuscular glycogen and 38% from blood glucose). At intensityabove 80% VO2max

, when catecholamine concentrations are highand IGR is less correlated to metabolic changes, 75% energy isprovided by CHO (60% from muscular glycogen and 15% fromblood glucose) and the remaining is provided by fatty acids[97].

This is in accordance with what we already said: as intensityincreases, energy provision switches from plasma to intramus-cular sources and from lipids to CHO.

In turn, muscle seems capable to produce a complex signalensuring an optimal flux of substrate and matching the variousconstraints. AMPK and lactate production play key roles in thiscoordination.

After 1 h of exercise at 45% VO2max, 60% NEFA released

during lipolysis enter muscle, 45% being directly oxidized and15% being esterified in IMTG pool. These data are influencedby exercise intensity and duration. Like at rest, NEFA uptake ishighly dependant on their concentration. However, beyond agiven step, this correlation is attenuated, which means thatfactors other than concentration are likely to limit lipidsmuscular uptake and utilization [98]. Intramuscular transport

and cytosolic translocation are not limiting steps, as demon-strated in various studies [97]. By contrast, CPT-1-dependentintramitochondrial transport of acyl-CoA is a key step [99]. Butalthough malonyl-CoA is decreased during exercise, thereseems to be a lack of correlation between its concentration andin CPT-1 activity [100]. AMPK inhibits ACC-2, whichcatalyzes malonyl-CoA synthesis from acetyl-CoA but AMPKseems to play a significant role only during deep glycogendepletion or prolonged exercise [101]. So malonyl-CoA is notlikely to be the key regulating factor of CPT-1 during exercise.

CPT-1 is also regulated by intramuscular rate of its maincofactor, carnitine. Studies at rest did not find a correlationbetween intramuscular carnitine concentration and lipid oxida-tion rate [102]. However, during high-intensity exercise, anincrease in glucose utilization results, through PDH activity, toan increase in acetyl-CoA and in acetyl-carnitine/free-carnitineratio (AC/FC) [103]. This AC/FC ratio is correlated to RER andnegatively correlated to CPT-1 activity and lipid oxidation rate.As glycogen depletion gets deeper, an increase in AMPKactivity, resulting in an increase in malonyl-CoA concentration,is likely to reinforce the AC/FC ratio inhibiting effect on CPT-1activity. By contrast, during moderate-intensity exercise,especially if it is prolonged, intramuscular CHO metabolismis decreased resulting in a decrease in AC/FC ratio and, thus, inCPT-1 inhibition, which allows lipid oxidation. Althoughmechanisms are different, this can be viewed as an equivalentof the ‘Randle Reverse’ during exercise: an increase in CHOutilization results in a decrease in lipid oxidation (Fig. 7).

IMTG metabolism during exercise is still extensivelystudied. Unfortunately, the many differences in subjectsrecruitment, exercise paradigm and AT study techniques limittheir interpretation. Nutritional status influences the IMTG poolsize as it is increased by fat-rich diet and decreased by CHO-richdiet [97]. IMTG are usually found close to mitochondria, which

Fig. 7. Intramuscular substrate interactions during exercise, reproduced from Kiens et al. (CL: citrate lyase; ACC: acetyl-CoA carboxylase; AMPK: AMP-kinase;PDH: pyruvate dehydrogenase: CPT: carnitine palmitoyl transferase: CAT: carnitine acyl transferase).

14 E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

suggests that they might play a role similar to glycogen as animmediately available substrate. In fact, it seems that trainingstatus, more than the exercise paradigm, influences IMTGutilization during exercise [104]. In most studies, IMTG poolsize does not change after a single bout of exercise [53]. Duringmoderate-intensity exercise, the effects of decreased insulinae-mia and increased lipid oxidation on IMTG pool arecompensated by the NEFA flux. During high-intensity exercise,lipid oxidation is decreased and increased lactate concentrationinhibits NEFA uptake. Catecholamines stimulate muscularlipolysis, which seems paradoxical. However, metabolismduring recovery should be taken into account as it tends toreplenish substrate stores proportionally to energy depletion.Whatever the hypotheses, studies without nutritional bias usingtechniques directly measuring IMTG turnover are crucial to ourunderstanding of IMTG roles during exercise.

IMTG and NEFA are not the only lipid substrates but VLDLswere so far considered a negligible energy source. However ifthis postulate were wrong, IMTG contribution would beoverestimated. Currently, recent studies have demonstratedthat VLDL processing is increased after 1 h of recovery and thatthis processing rate was influenced by exercise intensity andduration, and by the nutritional status (this study used fat-richdiet) [105]. In fact, VLDL seem to have a significant role duringrecovery. VLDL–NEFA uptake is under the control of muscularLPL and despite an increase in insulinaemia during recovery,LPL activity is increased in proportion to energy expenditure[106]. LPL activity is maximal 18 h after the end of exercise,which coincides with IMTG pool size nadir and is alsoinfluenced by exercise duration and intensity [107]. Therefore,two similar phenomena occur during exercise and recoverywhich demonstrate why the various regulation loops can bequalified as ‘open’. During exercise, an increased muscularglucose uptake coexists with a decreased insulinaemia as well asduring recovery, an increased muscular VLDL–NEFA uptakecoexists with an increased insulinaemia. So, if an increase ininsulinaemia is crucial to redirect blood glucose towardglycogenogenesis in order to replenish glycogen stores, theremaining glucose must match the vital constraint of maintain-ing blood glucose and cannot be used as energy source for therest of the body. Fatty acids would be expected to supply theenergy glucose cannot and, as an increase in insulinaemiainhibits lipolysis, this apparently paradoxical increase in LPLactivity, with mechanisms overriding insulin effects on LPLactivity, could allow NEFA–VLDL uptake and their use asenergy source.

Factors regulating CHO metabolism during exercise are alsodifferent compared to the resting state. Insulin secretion isinhibited, although blood glucose still has a small but significanteffect. The major fact is GLUT-4 translocation occurringthrough an insulin-independent pathway [108–110].

Muscular glycogen is another source of CHO and perfor-mance is conditioned by its disposal. There seems to be noevidence for glucagon receptors in muscle. Glucose uptake andutilization appears to be under the control of AMPK, activatedwhen AMP/ATP ratio increases or when muscular contractionand glycogen depletion occur [76].

During moderate-intensity exercise, NEFA flux achievesenergy demand, sparing glycogen stores. In this case, muscularcontraction is the only factor activating AMPK, and itsstimulation of glucose uptake and utilization is reduced.

However, as intensity is increased, muscular glycogen isdegraded in response to catecholamines and, as energy demandis increased, AMP/ATP ratio gets higher. AMPK activation ishighly stimulated, resulting in an increase in glucose uptake andutilization whereas accumulation of pyruvate and acetyl-CoAtends to inhibit lipid oxidation through an increase in AC/FCratio.

Moreover, at a given intensity termed ‘lactate threshold’,anaerobic metabolism contribution to energy supply becomessignificant [111]. This is not due, as previously believed, to a‘Pasteur-like effect’ (i.e. oxygen supply can no longer achievemitochondrial needs), but rather to an unbalance betweenmetabolic pathways [112]. Thus oxidative pathways can nolonger process all the pyruvate produced by carbohydratebreakdown and pyruvate is thus directed toward anaerobicconversion to lactate. Lactate itself can be either oxidized orrecycled as a gluconeogenic precursor into the liver (Coricycle). An increase in lactate concentration also results inmetabolic acidosis which inhibits lipolysis and muscular NEFAuptake. In fact, there is a basal lactate production and itsincrease starts from the beginning of exercise but, under thelactate threshold, its oxidation rate matches its production rate.From the lactate threshold, however, production predominatesover oxidation, resulting in an increase in lactate concentration[111]. As a result of all these interactions, during high-intensityexercise, anaerobic metabolism results in an increase in CHOutilization and in a decrease in NEFA release, uptake and, thus,utilization.

To summarize, during exercise, interactions between CHOand lipid metabolism do not seem reciprocal. While it isestablished that an increase in CHO utilization results in adecrease in lipid oxidation, the reverse (which would beRandle's cycle during exercise) is not. But this remains logicalas long as we consider the various constraints: at rest, absolutepriority is the struggle against hypoglycaemia, even if muscularglucose uptake must be inhibited, whereas during exercise,supplying glucose to the muscle is as much a priority as bloodglucose. This is certainly why the body developed an insulin-independent glucose uptake ability in order to spare muscularglycogen. So, in this perspective, the existence of a systeminhibiting glucose uptake would be counterproductive. In fact,muscle metabolism seems to privilege lipid more than CHOutilization and plasma more than muscular substrates as far astheir speed rate of supplying and oxidation can achieve energydemand.

9. Breaking or bypassing the circles: effects of training

Alternation of exercise and recovery states results inadaptations optimizing energy balance for a given performance.We only discuss here the effects of endurance training, morestudied than resistance training, even if there is a current interestin its utilization in diabetes and obesity.

15E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

After a description of training effects on substratesmetabolism, we will discuss interactions between trainingstatus and exercise paradigm. Then, as a conclusion, we will seehow training and exercise can interact with regulations loops atrest.

9.1. Training and exercise

Endurance training (ET) results in an increase in maximalpower output, also represented by VO2max

. ET also increases lipidoxidation and decreases catecholamines production whereashigh-intensity exercise causes an increase in CHO utilizationand catecholamines concentration [112]. So, interactionbetween training status and exercise intensity should probablybe taken into account to study ET effects on metabolicadaptations during exercise. The crossover concept, as devel-oped by Brooks and Mercier, integrates all those parameters[112]. Glucodependency is defined as the exercise intensityrange when CHO become the predominant substrates. Accord-ing to this concept the crossover point is set as the relativeintensity when this shift toward predominant CHO utilizationoccurs. ET was demonstrated to cause an increase in thecrossover point. For example, after a few weeks of ET,exercising at a given relative intensity results in a higherabsolute power output and an increased lipid oxidation. Thelatter is explained by an increase in mitochondrial mass,oxidative enzymatic activity and respiratory channel energyproduction. In other words, for an identical oxygen flux, trainedsubjects produce more ATP than untrained subjects. Conse-quently, this increase in ATP inhibits muscular glycogenogen-esis, phosphofructokinase, pyruvate-dehydrogenase and, thus,glycolysis which leads to a decrease in lactate production duringexercise [113,114].

ET's effects on catecholamines secretion is biphasic. Duringmoderate-intensity exercise, epinephrine is decreased in trainedsubjects [115] and reaches lipolysis-stimulating concentrations.

By contrast, during high-intensity exercise, catecholamineresponse is increased in trained subjects [116]. It has to benoted that these changes occur early, before circulatory ormitochondrial adaptations, and result in an increase in lipidoxidation for intensities higher than before ET [117].

Pancreas hormones secretions are also modified by ET,which lowers the decrease in insulin secretion and the increasein glucagon secretion. As these changes apparently favourpredominant CHO utilization over lipid oxidation, this wouldseem a paradox if an increase in tissue's sensitivities to thosevarious signals were not taken into account.

ET causes an increase in AT sensitivity to the lipolyticeffects of epinephrine during exercise, allowing maintenance ofan adapted NEFA flux despite a decrease in epinephrineconcentration and an inhibition of decrease in insulin secretion.ET's effects on the liver are discordant among the variousstudies with an apparent increase in EGP and gluconeogenesis.Considering muscle, ET causes an increase in GLUT-4 poolsize and translocation as well as in hexokinase activity [118–120]. IMTG pool size is also increased by ET. As remindedabove, a correlation between insulin resistance and IMTG poolsize is well established but a causative link is not. Thisobservation led to the debate about ‘the IMTG paradox’recently reviewed by Van Loon [104] as this correlation waspositive in sedentary, obese and diabetic subjects but negativein healthy trained athletes. This is no longer a paradox if onereasons with oxidative capacity as explained before. Moreprecisely, ET causes an increase in IMTG utilization and,consequently, in NEFA esterification. As NEFA flux isdecreased due to ET-induced decrease in catecholamines andin insulin secretion blockade, this increase in IMTG plays acompensating role. A simultaneous ET-induced increase inlipid oxidation, IMTG utilization and NEFA storing couldallow the use of IMTG as an alternative substrate to glycogenas it is quickly available with a well-coordinated turnover[104].

Fig. 8. Back from a vicious to a virtuous circle: effect of exercise and training on an adipo-hepato-insulo-muscular loop of regulation modified by an excess of visceraladipose tissue (NEFA: non-esterified fatty acids; VAT: visceral adipose tissue; SCAT: subcutaneous adipose tissue; EGP: endogenous glucose production).

16 E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

Glycogen metabolism is also a matter of current interest.Glycogen stores are the main factor determining perfor-mance and also play a role in regulating substrates' meta-bolism. As a matter of fact, when muscular glycogenconcentration is low, respiratory exchange ratio (RER) isdecreased, lipolysis and lipid oxidation are increasedwhereas CHO oxidation is unchanged [121]. This suggestsfor the existence of a signal linking muscular glycogenconcentration with lipid metabolism, which is actuallyunknown. AMPK would be a good candidate but it mainlystimulates CHO metabolism. Other signals, such as IL-6 orTNF-α, are likely to be part of a “work factor” as a signal-modulating metabolism according to glycogen availability morethan as a signal of contraction [96,122]. Practically, IL-6 isincreased during contraction and its rate of secretion isproportional to glycogen depletion. It contributes to an increasein EGP and a lipolytic activity of this cytokine has also beendemonstrated in vitro [122].

10. Conclusion

On the whole, considering that the above-described regula-tion loops are open, vicious or virtuous circles and that on theother hand exercise and training activate metabolic pathwaysthat are irrelevant at rest, exercise and training appear to be ableto compensate mechanisms and thus to turn those circles fromvicious to virtuous state.

For example, muscle contraction and training both increasemuscular glucose uptake, respectively, in a short-term and along-term perspective and, most important of all, in an insulin-independent fashion (Fig. 8).

Insulin resistance is defined as a lack of efficiency of a givenconcentration of insulin to normalize blood glucose. Thus, anymechanism that helps insulin to restore glucose homeostasisdecreases insulin resistance. This simple example shows that inorder to understand physiology or pathophysiology of meta-bolic diseases with their apparent paradox and in order to testdrug or non-drug treatments of these diseases, vicious andvirtuous circles should be taken into account. As long asphysiological conditions can be restored, even if the initialpathogenic mechanism is still at work, then acting on theseregulation loops could be beneficial. Well-designed studies onthe therapeutic use of physical activity, even if suggested sinceHippocrates, have been lacking until recently. The recentdevelopment of these studies has shown that exercise is able toreverse or delay vicious circles. On the basis of the mechanismsreviewed above, it can be assumed that an individualquantification of these regulation loops will allow targetingoptimal exercise parameters aiming at better restoring thevirtuous state.

References

[1] Leverve X. Homéostasie énergétique et métabolisme glucido–lipidique:pourquoi deux substrats différents? Cah Nutr Diét 2005;40:161–5.

[2] DeFronzo RA. Insulin secretion, insulin resistance, and obesity. Int JObes 1982;6(Suppl 1):73–82.

[3] Randle PJ, Garland PB, Hales CN, Newsholme EA. The glucose fatty-acid cycle. Its role in insulin sensitivity and the metabolic disturbances ofdiabetes mellitus. Lancet 1963;13:785–9.

[4] Stich V, Berlan M. Physiological regulation of NEFA availability:lipolysis pathway. Proc Nutr Soc 2004;63:369–74.

[5] Shen WJ, Liang Y, Hong R, et al. Characterization of the functionalinteraction of adipocyte lipid-binding protein with hormone-sensitivelipase. J Biol Chem 2001;276:49443–8.

[6] Kishida K, Kuriyama H, Funahashi T, et al. Aquaporin adipose, a putativeglycerol channel in adipocytes. J Biol Chem 2000;275:20896–902.

[7] Sengenes C, Verlan M, De Glisezinski I, Lafontan M, Galitzky J.Natriuretic peptides: a new lipolytic pathway in human adipocytes.FASEB J 2000;14:1345–51.

[8] Koonen DP, Glatz JF, Bonen A, Luiken JJ. Long-chain fatty acid uptakeand FAT/CD36 translocation in heart and skeletal muscle. BiochimBiophys Acta 2005;1736:163–80.

[9] Brunzell JD, Schwartz RS, Eckel RH, Goldberg AP. Insulin and adiposetissue lipoprotein lipase activity in humans. Int J Obes 1981;5:685–94.

[10] Ranganath LR, Beety JM, Wright J, Morgan LM. Nutrient regulation ofpost-heparin lipoprotein lipase activity in obese subjects. Horm MetabRes 2001;33:57–61.

[11] Hellerstein MK, Schwarz JM, Neese RA. Regulation of hepatic de novolipogenesis in humans. Annu Rev Nutr 1996;16:523–57.

[12] Nielsen S, Guo Z, Albu JB, Klein S, O'Brien PC, Jensen MD. Energyexpenditure, sex, and endogenous fuel availability in humans. J ClinInvest 2003;111:981–8.

[13] Jensen MD, Caruso M, Heiling V, Miles JM. Insulin regulation oflipolysis in nondiabetic and IDDM subjects. Diabetes 1989;38:1595–601.

[14] Stich V, Pelikanova T, Wohl P, et al. Activation of alpha2-adrenergicreceptors blunts epinephrine-induced lipolysis in subcutaneous adiposetissue during a hyperinsulinemic euglycemic clamp in men. Am J PhysiolEndocrinol Metab 2003;285:E599–607.

[15] Kovar J, Fejfarova V, Pelikanova T, Poledne R. Hyperglycemiadownregulates total lipoprotein lipase activity in humans. Physiol Res2004;53:61–8.

[16] Soni KG, Lehner R, Metalnikov P, et al. Carboxylesterase 3 (EC 3.1.1.1)is a major adipocyte lipase. J Biol Chem 2004;24(279):40683–9.

[17] Havel PJ. Update on adipocyte hormones: regulation of energy balanceand carbohydrate/lipid metabolism. Diabetes 2004;53(Suppl 1):S143–51.

[18] Vague J. The distinction of android and gynoid obesities as the base fortheir prognosis. Sem Hop 1955;31:3503–9.

[19] Alberti KG, Zimmet P, Shaw J. Metabolic syndrome—a new world-widedefinition. A Consensus Statement from the International DiabetesFederation. Diabet Med 2006;23:469–80.

[20] Abate N, Burns D, Peshock RM, Garg A, Grundy SM. Estimation ofadipose tissue mass by magnetic resonance imaging: validation againstdissection in human cadavers. J Lipid Res 1994;35:1490–6.

[21] Seidell JC, Bouchard C. Visceral fat in relation to health: is it a majorculprit or simply an innocent bystander? Int J Obes Relat Metab Disord1997;21:626–31.

[22] Meek SE, Nair KS, Jensen MD. Insulin regulation of regional free fattyacid metabolism. Diabetes 1999;48:10–4.

[23] Fried SK, Russell CD, Grauso NL, Brolin RE. Lipoprotein lipaseregulation by insulin and glucocorticoid in subcutaneous and omentaladipose tissues of obese women and men. J Clin Invest 1993;92:2191–8.

[24] Saleh J, Christou N, Cianflone K. Regional specificity of ASP binding inhuman adipose tissue. Am J Physiol 1999;276:E815–21.

[25] Lundgren M, Buren J, Ruge T, Myrnas T, Eriksson JW. Glucocorticoidsdown-regulate glucose uptake capacity and insulin-signaling proteins inomental but not subcutaneous human adipocytes. J Clin EndocrinolMetab 2004;89:2989–97.

[26] Virtanen KA, Lonnroth P, Parkkola R, et al. Glucose uptake and perfusionin subcutaneous and visceral adipose tissue during insulin stimulation innonobese and obese humans. J Clin Endocrinol Metab 2002;87:3902–10.

[27] Giorgino F, Laviola L, Eriksson JW. Regional differences of insulinaction in adipose tissue: insights from in vivo and in vitro studies. ActaPhysiol Scand 2005;183:13–30.

17E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

[28] Mauriege P, Galitzky J, Berlan M, Lafontan M. Heterogeneousdistribution of beta and alpha-2 adrenoceptor binding sites in humanfat cells from various fat deposits: functional consequences. Eur J ClinInvest 1987;17(2)(Apr.):156–65.

[29] Fried SK, Leibel RL, Edens NK, Kral JG. Lipolysis in intraabdominaladipose tissues of obese women and men. Obes Res 1993;1:443–8.

[30] Yildiz BO, Haznedaroglu IC. Rethinking leptin and insulin action:therapeutic opportunities for diabetes. Int J Biochem Cell Biol 2006;38:820–30.

[31] Barzilai N, Wang J, Massilon D, Vuguin P, Hawkins M, Rossetti L.Leptin selectively decreases visceral adiposity and enhances insulinaction. J Clin Invest 1997;100:3105–10.

[32] Rossetti L, Massillon D, Barzilai N, et al. Short term effects of leptin onhepatic gluconeogenesis and in vivo insulin action. J Biol Chem1997;272:27758–63.

[33] Kamohara S, Burcelin R, Halaas JL, Friedman JM, Charron MJ. Acutestimulation of glucose metabolism in mice by leptin treatment. Nature1997;389:374–7.

[34] Dulloo AG, Gubler M, Montani JP, Seydoux J, Solinas G. Substratecycling between de novo lipogenesis and lipid oxidation: a thermogenicmechanism against skeletal muscle lipotoxicity and glucolipotoxicity. IntJ Obes Relat Metab Disord 2004;28(Suppl 4):S29–37.

[35] Kieffer TJ, Habener JF. The adipoinsular axis: effects of leptin onpancreatic beta-cells. Am J Physiol Endocrinol Metab 2000;278:E1–14.

[36] Mueller WM, Stanhope KL, Gregoire F, Evans JL, Havel PJ. Effects ofmetformin and vanadium on leptin secretion from cultured rat adipocytes.Obes Res 2000;8:530–9.

[37] Margetic S, Gazzola C, Pegg GG, Hill RA. Leptin: a review of itsperipheral actions and interactions. Int J Obes Relat Metab Disord2002;26:1407–33.

[38] Wolfe BE, Jimerson DC, Orlova C, Mantzoros CS. Effect of dieting onplasma leptin, soluble leptin receptor, adiponectin and resistin levels inhealthy volunteers. Clin Endocrinol 2004;61:332–8.

[39] Berg AH, Combs TP, Scherer PE. ACRP30/adiponectin. An adipokineregulating glucose and lipid metabolism. Trends Endocrinol Metab2002;13:84–9.

[40] Yamauchi T, Kamon J, Waki H, et al. The fat-derived hormoneadiponectin reverses insulin resistance associated with both lipoatrophyand obesity. Nat Med 2001;7:941–6.

[41] Arita Y, Kihara S, Ouchi N, et al. Paradoxical decrease of an adipose-specific protein, adiponectin, in obesity. Biochem Biophys Res Commun1999;257:79–83.

[42] Fruebis J, Tsao TS, Javorschi S. Proteolytic cleavage product of 30-kDaadipocyte complement-related protein increases fatty acid oxidation inmuscle and causes weight loss in mice. Proc Natl Acad Sci U S A 2001;98(4):2005–10.

[43] Berg AH, Combs TP, Du X, Brownlee M, Scherer PE. The adipocyte-secreted protein Acrp30 enhances hepatic insulin action. Nat Med2001;7:947–53.

[44] Thamer C, Machann J, Tschritter O, et al. Relationship between serumadiponectin concentration and intramyocellular lipid stores in humans.Horm Metab Res 2002;34:646–9.

[45] Stefan N, Stumvoll M, Vozarova B, et al. Plasma adiponectin andendogenous glucose production in humans. Diabetes Care 2003;26:3315–9.

[46] Lefebvre AM, Laville M, Vega N, et al. Depot-specific differences inadipose tissue gene expression in lean and obese subjects. Diabetes1998;47:98–103.

[47] Lihn AS, Bruun JM, He G, Pedersen SB, Jensen PF, Richelsen B. Lowerexpression of adiponectin mRNA in visceral adipose tissue in lean andobese subjects. Mol Cell Endocrinol 2004;219:9–15.

[48] Grill V, Ovigstad E. Fatty acids and insulin secretion. Br J Nutr 2000;83(Suppl 1):S79–84.

[49] Stefan N, Stumvoll M, Bogardus C, Tataranni PA. Elevated plasmanonesterified fatty acids are associated with deterioration of acute insulinresponse in IGT but not NGT. Am J Physiol Endocrinol Metab 2003;284:E1156–61.

[50] Itoh Y, Hinuma S. GPR40, a free fatty acid receptor on pancreatic betacells, regulates insulin secretion. Hepatol Res 2005;33:171–3.

[51] Hirasawa A, Tsumaya K, Awaji T, et al. Free fatty acids regulate gutincretin glucagon-like peptide-1 secretion through GPR120. Nat Med2005;11:90–4.

[52] Meier JJ, Nauck MA. The potential role of glucagon-like peptide 1 indiabetes. Curr Opin Investig Drugs 2004;5(4)(Apr.):402–10.

[53] Jensen MD. Fate of fatty acids at rest and during exercise: regulatorymechanisms. Acta Physiol Scand 2003;178:385–90.

[54] Romanski SA, Nelson RM, JensenMD. Meal fatty acid uptake in adiposetissue: gender effects in nonobese humans. Am J Physiol EndocrinolMetab 2000;279:E455–62.

[55] Timlin MT, Barrows BR, Parks EJ. Increased dietary substrate deliveryalters hepatic fatty acid recycling in healthy men. Diabetes 2005;54:2694–701.

[56] Ferrannini E, Barrett EJ, Bevilacqua S, DeFronzo RA. Effect of fattyacids on glucose production and utilization in man. J Clin Invest 1983;72:1737–47.

[57] Boden G, Cheung P, Stein TP, Kresge K, Mozzoli M. FFA cause hepaticinsulin resistance by inhibiting insulin suppression of glycogenolysis.Am J Physiol Endocrinol Metab 2002;283:E12–9.

[58] Stingl H, Krssak M, Krebs M, et al. Stingl lipid-dependent control ofhepatic glycogen stores in healthy humans. Diabetologia 2001;44:48–54.

[59] Bergman RN. Non-esterified fatty acids and the liver: why is insulinsecreted into the portal vein? Diabetologia 2000;43:946–52.

[60] Fisher SJ, Kahn CR. Insulin signaling is required for insulin's direct andindirect action on hepatic glucose production. J Clin Invest 2003;111:463–8.

[61] Jenssen T, Nurjhan N, Consoli A, Gerich JE. Failure of substrate-inducedgluconeogenesis to increase overall glucose appearance in normalhumans. Demonstration of hepatic autoregulation without a change inplasma glucose concentration. J Clin Invest 1990;86:489–97.

[62] Clore JN, Helm ST, Blackard WG. Loss of hepatic autoregulation aftercarbohydrate overfeeding in normal man. J Clin Invest 1995;96:1967–72.

[63] Huang W, Dedousis N, Bandi A, Lopaschuk GD, O'Doherty RM. Livertriglyceride secretion and lipid oxidative metabolism are rapidly alteredby leptin in vivo. Endocrinology 2006;147:1480–7.

[64] Andreelli F, Foretz M, Knauf C, et al. Liver adenosine monophosphate-activated kinase-alpha2 catalytic subunit is a key target for the control ofhepatic glucose production by adiponectin and leptin but not insulin.Endocrinology 2006;147:2432–41.

[65] Saltin B, Gollnick PD. Skeletal muscle adaptability : significance formetabolism and performance. In: Peachey LD, Adrian R, Geiger SR,editors. Handbook of physiology. Section 10: skeletal muscle. Baltimore:Williams & Wilkins; 1983. p. 555–632.

[66] McGarry JD, Mannaerts GP, Foster DW. A possible role for malonyl-CoA in the regulation of hepatic fatty acid oxidation and ketogenesis.J Clin Invest 1977;60:265–70.

[67] Turcotte LP, Petry C, Kiens B, Richter EA. Contraction-induced increasein Vmax of palmitate uptake and oxidation in perfused skeletal muscle.J Appl Physiol 1998;84:1788–94.

[68] Jensen MD, Ekberg K, Landau BR. Lipid metabolism during fasting. AmJ Physiol Endocrinol Metab 2001;281:E789–93.

[69] Sacchetti M, Saltin B, Olsen DB, van Hall G. High triacylglycerolturnover rate in human skeletal muscle. J Physiol 2004;561:883–91.

[70] Hopp JF, Palmer WK. Effect of glucose and insulin on triacylglycerolmetabolism in isolated normal and diabetic skeletal muscle. Metabolism1991;40:223–5.

[71] Langfort J, Donsmark M, Ploug T, Holm C, Galbo H. Hormone-sensitivelipase in skeletal muscle: regulatory mechanisms. Acta Physiol Scand2003;178:397–403.

[72] Van Loon LJ, Goodpaster BH. Increased intramuscular lipid storage inthe insulin-resistant and endurance-trained state. Pflugers Arch 2006;451:606–16.

[73] Frayn KN. The glucose–fatty acid cycle: a physiological perspective.Biochem Soc Trans 2003;31:1115–9.

[74] Boden G, Laakso M. Lipids and glucose in type 2 diabetes: what is thecause and effect? Diabetes Care 2004;27:2253–9.

18 E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

[75] McGarry JD, Brown NF. The mitochondrial carnitine palmitoyltransfer-ase system. From concept to molecular analysis. Eur J Biochem1997;244:1–14.

[76] Ha J, Lee JK, Kim KS, Witters LA, Kim KH. Cloning of human acetyl-CoA carboxylase-beta and its unique features. Proc Natl Acad Sci U S A1996;93:11466–70.

[77] Merrill GF, Kurth EJ, Hardie DG, Winder WW. AICA riboside increasesAMP-activated protein kinase, fatty acid oxidation, and glucose uptake inrat muscle. Am J Physiol 1997;273:E1107–12.

[78] Hardie DG, Carling D. The AMP-activated protein kinase—fuel gauge ofthe mammalian cell? Eur J Biochem 1997;246:259–73.

[79] Carling D, Fryer LG, Woods A, Daniel T, Jarvie SL, Whitrow H.Bypassing the glucose/fatty acid cycle: AMP-activated protein kinase.Biochem Soc Trans 2003;31:1157–60.

[80] Minokoshi Y, Kim YB, Peroni OD, et al. Leptin stimulates fatty-acidoxidation by activating AMP-activated protein kinase. Nature 2002;415:339–43.

[81] Viru A. Plasma hormones and physical exercise. Int J Sports Med1992;13:201–9.

[82] Peters SJ, Dyck DJ, Bonen A, Spriet LL. Effects of epinephrine onlipid metabolism in resting skeletal muscle. Am J Physiol 1998;275:E300–9.

[83] Broadstone VL, Pfeifer MA, Bajaj V, Stagner JI, Samols E. Alpha-adrenergic blockade improves glucose-potentiated insulin secretion innon-insulin-dependent diabetes mellitus. Diabetes 1987;36:932–7.

[84] Clutter WE, Bier DM, Shah SD, Cryer PE. Epinephrine plasma metabolicclearance rates and physiologic thresholds for metabolic and hemody-namic actions in man. J Clin Invest 1980;66:94–101.

[85] Marliss EB, Vranic M. Intense exercise has unique effects on both insulinrelease and its roles in glucoregulation: implications for diabetes.Diabetes 2002;51(Suppl 1):S271–83.

[86] Wasserman DH, O'Doherty RM, Zinker BA. Role of the endocrinepancreas in control of fuel metabolism by the liver during exercise. Int JObes Relat Metab Disord 1995;19(Suppl 4):S22–30.

[87] Guezennec CY, Portero P, Satabin P, Seyfried D. Carbohydrate nutritionbefore and during exercise. Sci Mot 1988;5:39–45.

[88] Sotsky MJ, Shilo S, Shamoon H. Regulation of counterregulatoryhormone secretion in man during exercise and hypoglycemia. J ClinEndocrinol Metab 1989;68:9–16.

[89] Marliss EB, Simantirakis E, Miles PD, et al. Glucose turnover and itsregulation during intense exercise and recovery in normal male subjects.Clin Invest Med 1992;15:406–19.

[90] Sigal RJ, Fisher S, Halter JB, Vranic M, Marliss EB. The roles ofcatecholamines in glucoregulation in intense exercise as defined by theislet cell clamp technique. Diabetes 1996;45:148–56.

[91] Sengenes C, Moro C, Galitzky J, Berlan M, Lafontan M. Natriureticpeptides: a new lipolytic pathway in human fat cells. Med Sci (Paris)2005;21:29–33.

[92] Lafontan M, Moro C, Sengenes C, Galitzky J, Crampes F, Berlan M.An unsuspected metabolic role for atrial natriuretic peptides: thecontrol of lipolysis, lipid mobilization, and systemic nonesterified fattyacids levels in humans. Arterioscler Thromb Vasc Biol 2005;25:2032–42.

[93] Wolfe RR, Klein S, Carraro F, Weber JM. Role of triglyceride–fatty acidcycle in controlling fat metabolism in humans during and after exercise.Am J Physiol 1990;258:E382–9.

[94] Romijn JA, Coyle EF, Sidossis LS, et al. Regulation of endogenous fatand carbohydrate metabolism in relation to exercise intensity andduration. Am J Physiol 1993;265:E380–91.

[95] Winder WW. Control of hepatic glucose production during exercise. MedSci Sports Exerc 1985;17:2–5.

[96] Febbraio MA, Hiscock N, Sacchetti M, Fischer CP, Pedersen BK.Interleukin-6 is a novel factor mediating glucose homeostasis duringskeletal muscle contraction. Diabetes 2004;53:1643–8.

[97] Kiens B. Skeletal muscle lipid metabolism in exercise and insulinresistance. Physiol Rev 2006;86:205–43.

[98] Horowitz KF, Klein S. Lipid metabolism during endurance exercise. AmJ Clin Nutr 2000;72:558S–63S [Suppl].

[99] McGarry JD. Banting Lecture 2001: dysregulation of fatty acidmetabolism in the etiology of type 2 diabetes. Diabetes 2002;51:7–18.

[100] Odland LM, Heigenhauser GJ, Lopaschuk GD, Spriet LL. Humanskeletal muscle malonyl-CoA at rest and during prolonged submaximalexercise. Am J Physiol 1996;270:E541–4.

[101] Raney MA, Yee AJ, Todd MK, Turcotte LP. AMPK activation is notcritical in the regulation of muscle FA uptake and oxidation during low-intensity muscle contraction. Am J Physiol Endocrinol Metab 2005;288:E592–8.

[102] Roepstorff C, Vistisen B, Roepstorff K, Kiens B. Regulation of plasmalong-chain fatty acid oxidation in relation to uptake in human skeletalmuscle during exercise. Am J Physiol Endocrinol Metab 2004;287:E696–705.

[103] Van Loon LJ, Greenhaff PL, Constantin-Teodosiu D, Saris WH,Wagenmakers AJ. The effects of increasing exercise intensity on musclefuel utilisation in humans. J Physiol 2001;536:295–304.

[104] Koopman R, Manders RJ, Jonkers RA, Hul GB, Kuipers H, van Loon LJ.Intramyocellular lipid and glycogen content are reduced followingresistance exercise in untrained healthy males. Eur J Appl Physiol2006;96:525–34.

[105] Helge JW, Watt PW, Richter EA, Rennie MJ, Kiens B. Fat utilizationduring exercise: adaptation to a fat-rich diet increases utilization ofplasma fatty acids and very low density lipoprotein–triacylglycerol inhumans. J Physiol 2001;537:1009–20.

[106] Kiens B, Lithell H, Mikines KJ, Richter EA. Effects of insulin andexercise on muscle lipoprotein lipase activity in man and its relation toinsulin action. J Clin Invest 1989;84:1124–9.

[107] Kiens B, Richter EA. Utilization of skeletal muscle triacylglycerolduring postexercise recovery in humans. Am J Physiol 1998;275:E332–7.

[108] Douen AG, Ramlal T, Rastogi S, et al. Exercise induces recruitment of the“insulin-responsive glucose transporter”. Evidence for distinct intracel-lular insulin- and exercise-recruitable transporter pools in skeletal muscle.J Biol Chem 1990;265:13427–30.

[109] Cheatham B, Vlahos CJ, Cheatham L, Wang L, Blenis J, Kahn CR.Phosphatidylinositol 3-kinase activation is required for insulin stimula-tion of pp70 S6 kinase, DNA synthesis, and glucose transportertranslocation. Mol Cell Biol 1994;14:4902–11.

[110] Okada T, Kawano Y, Sakakibara T, Hazeki O, Ui M. Essential role ofphosphatidylinositol 3-kinase in insulin-induced glucose transport andantilipolysis in rat adipocytes. Studies with a selective inhibitorwortmannin. J Biol Chem 1994;269:3568–73.

[111] Pocock G, Richards CD. Human physiology: the basis of medicine.Oxford: Oxford University Press; 1999.

[112] Brooks GA, Mercier J. Balance of carbohydrate and lipid utilizationduring exercise: the “crossover” concept. J Appl Physiol 1994;76:2253–61.

[113] Henriksson J, Reitman JS. Time course of changes in human skeletalmuscle succinate dehydrogenase and cytochrome oxidase activities andmaximal oxygen uptake with physical activity and inactivity. ActaPhysiol Scand 1977;99:91–7.

[114] Mole PA, Oscai LB, Holloszy JO. Adaptation of muscle to exercise.Increase in levels of palmityl CoA synthetase, carnitine palmityltransfer-ase, and palmityl CoA dehydrogenase, and in the capacity to oxidize fattyacids. J Clin Invest 1971;50:2323–30.

[115] Deuster PA, Chrousos GP, Luger A, et al. Hormonal and metabolicresponses of untrained, moderately trained, and highly trained men tothree exercise intensities. Metabolism 1989;38:141–8.

[116] Kjaer M, Farrell PA, Christensen NJ, Galbo H. Increased epinephrineresponse and inaccurate glucoregulation in exercising athletes. J ApplPhysiol 1986;61:1693–700.

[117] Green HJ, Helyar R, Ball-Burnett M, Kowalchuk N, Symon S, FarranceB. Metabolic adaptations to training precede changes in musclemitochondrial capacity. J Appl Physiol 1992;72:484–91.

[118] Friedman JE, Sherman WM, Reed MJ, Elton CW, Dohm GL. Exercisetraining increases glucose transporter protein GLUT-4 in skeletal muscleof obese Zucker (fa/fa) rats. FEBS Lett 1990;268:13–6.

19E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20

Autho

r's

pers

onal

co

py

[119] Kern M, Wells JA, Stephens JM, et al. Insulin responsiveness in skeletalmuscle is determined by glucose transporter (Glut4) protein level.Biochem J 1990;270:397–400.

[120] Ren JM, Semenkovich CF, Gulve EA, Gao J, Holloszy JO. Exerciseinduces rapid increases in GLUT4 expression, glucose transport capacity,and insulin-stimulated glycogen storage in muscle. J Biol Chem1994;269:14396–401.

[121] Weltan SM, Bosch AN, Dennis SC, Noakes TD. Influence of muscleglycogen content on metabolic regulation. Am J Physiol 1998;274:E72–82.

[122] Steinacker JM, Lormes W, Reissnecker S, Liu Y. New aspects of thehormone and cytokine response to training. Eur J Appl Physiol 2004;91:382–91.

20 E. Ghanassia et al. / Clinica Chimica Acta 383 (2007) 1–20