amélioration de la sensibilité à l’insuline.
Introduction
Lorsque j’ai débuté ma thèse, notre laboratoire possédait une lignée de souris Lhs KO (lignée de C. Holm, Suède) et des tests préliminaires in vitro avaient montré une activité enzymatique de la LHS diminuée chez les souris haplodéficientes. J’ai donc été en charge de valider cette observation et nous avons ensuite débuté un travail exploratoire de description du phénotype des souris LHS+/-. Les premières approches ont été réalisées en régime standard. Nous avons caractérisé le défaut lipolytique associé à l’haplodéficience de la LHS. L’investigation de la sensibilité à l’insuline des animaux LHS+/- en régime standard a montré une amélioration de la tolérance à l’insuline (résultats non présentés). Cependant, la différence entre les souris LHS+/- et contrôles était très faible et a nécessité l’emploi d’un grand nombre d’animaux pour être statistiquement significative. Nous avons donc imaginé que le phénotype des souris LHS+/- pourrait être révélé par un régime enrichi en graisses. Des souris mâles B6D2/F1 soumises à un régime à 45% kCal de lipides (high fat diet : HFD) pendant 12 à 16 semaines, ont développé une obésité comparable à celle des animaux contrôles tandis que les souris Lhs KO, comme décrit dans la littérature (Harada et al., 2003), ont résisté à la prise de poids. Par une approche de fluxomique, nous avons étudié les flux d’AG dans ce modèle et montré que les souris LHS+/- présentaient un turn-over global des AG diminué. Nous avons également observé que les souris LHS+/-, même soumises à un régime gras montraient toujours une amélioration de la tolérance à l’insuline par rapport aux animaux contrôles. De plus, les animaux LHS+/- ont présenté une augmentation de l’utilisation du 2-désoxyglucose par le TA et les muscles révélant une augmentation de la sensibilité à l’insuline. Des données de calorimétrie indirecte viennent compléter le phénotype et nous informent sur une augmentation de l’utilisation du glucose comme substrat énergétique préférentiel chez les souris LHS+/-. Ce travail montre qu’une réduction des capacités lipolytiques peut modifier les flux lipidiques et influencer positivement la sensibilité à l’insuline.
Article
La diminution des capacités lipolytiques du
tissu adipeux par l’haplodéficience de la
lipase hormono-sensible diminue les flux
d’acides gras et améliore l’insulino-sensibilité
Amandine Girousse1,2, Geneviève Tavernier1,2, Laurent Monbrun1,2, Véronic
Bézaire1,2, Cédric Moro1,2, Aurélie Waget2,3, Balbine Roussel1,2, Bénédicte
Prunet Marcassus4, Sylvie Caspar-Bauguil1,2,5, Cecilia Holm6, Jean Galitzky2,7,
Thierry Sulpice4, , Rémy Burcelin2,3, and Dominique Langin1,2,5.
From 1 Inserm, U858, Obesity Research Laboratory, Team 4, Rangueil Institute
of Molecular Medicine, BP 84225, F-31432 Toulouse Cedex 4, France
2 University of Toulouse, IFR150, Paul Sabatier University, France
3 Inserm, U858, Team 2, Rangueil Institute of Molecular Medicine, Toulouse,
France 4Physiogenex, Prologue Biotech ,Rue Pierre et Marie Curie,Labège-
Innopole, France
5 Laboratory of clinical biochemistry, Toulouse University Hospitals, France 6 Department of Experimental Medical Science, Lund University, Lund, Sweden 7 Inserm, U858, Team 1, Rangueil Institute of Molecular Medicine, Toulouse,
France
Hormone-Sensitive Lipase Haploinsufficiency Reduces
Fatty Acid Fluxes and Improves Insulin Sensitivity
Amandine Girousse1,2, Geneviève Tavernier1,2, Laurent Monbrun1,2, Véronic Bézaire1,2,
Cédric Moro1,2, Bénédicte Prunet Marcassus4, Aurélie Waget2,3, Balbine Roussel1,2,
Sylvie Caspar Bauguil1,2,5, Jean Galitzky2,7,Cecilia Holm6, Thierry Sulpice4, Rémy
Burcelin2,3, and Dominique Langin1,2,5.
From 1Inserm, U858, Obesity Research Laboratory, Team 4, Rangueil Institute of Molecular Medicine, BP 84225, F-31432 Toulouse Cedex 4, France
2 University of Toulouse, IFR150, Paul Sabatier University, France
3 Inserm, U858, Team 2, Rangueil Institute of Molecular Medicine, Toulouse, France 4Physiogenex, Prologue Biotech ,Rue Pierre et Marie Curie,Labège-Innopole, France 5 Laboratory of clinical biochemistry, Toulouse University Hospitals, France
6 Department of Experimental Medical Science, Lund University, Lund, Sweden 7 Inserm, U858, Team 1, Rangueil Institute of Molecular Medicine, Toulouse, France
Running title: Altered lipid metabolism in HSL+/- obese mice
Correspondance: E-mail: dominique.langin@inserm.fr, Phone: +33(0)5 61325628; Fax: +33(0)561325623
Abstract
Hormone-sensitive lipase (HSL) is one of the enzymes catalyzing acylglycerol degradation in adipose tissue (AT). Expression and activity of AT HSL are decreased in obesity and insulin resistance. To determine metabolic fluxes and insulin sensitivity with an inhibition of AT lipolysis comparable to the defect reported in human obesity, HSL heterozygous mice (HSL+/- ) fed high fat diet were investigated. Blunted lipolytic capacities due to HSL haploinsufficiency was associated with global free fatty acid turn-over reduction and improvement of insulin sensitivity measured in vivo without modification of body weight and fat mass. This enhancement was not attributable to a modulation of AT inflammation. Treatment with a specific HSL inhibitor attenuated insulin resistance in high fat diet-fed mice. This work demonstrates that a reduction in lipolytic capacity can reshape lipid fluxes and improve insulin sensitivity without modification of fat mass.
INTRODUCTION
White adipose tissue (WAT) ensures the control of energy stores according to the nutritional status. In the fed state, under the influence of insulin, WAT stores excess energy as triacylglycerols (TG) in the lipid droplet of adipocytes. When energy is needed between meals or during physical exercise, WAT delivers fuel to the peripheral organs in the form of fatty acids (FA). Lipolysis is the process by which stored TG are released as non esterified FA (NEFA) (Lafontan and Langin, 2009). It involves different players such as lipases, co- lipases and also proteins that coat the lipid droplet. It is now largely accepted that the enzymatic breakdown of TG is initiated by adipose triglyceride lipase (ATGL) and leads to the formation of diacylglycerols (DG) that are in turn hydrolyzed by hormone-sensitive lipase (HSL) (Bezaire et al., 2009; Zimmermann et al., 2004). Of note, HSL also shows TG hydrolase activity. The final step of this catabolic process is the hydrolysis of monoacylglycerols by monoglyceride lipase leading to the release of one molecule of glycerol and three molecules of FA. In human AT, HSL mRNA and protein content correlates with maximal stimulated lipolysis (Large et al., 1998). Reduction of WAT fat cell lipolysis and HSL activity has been observed in obesity (Large et al., 1999). Moreover, ATGL and HSL
expression are decreased in WAT of obese and insulin-resistant patients (Jocken et al., 2007; Langin et al., 2005). The defective HSL expression has also been observed in cultures of differentiated preadipocytes from obese individuals highlighting the primary nature of this impairment (Langin et al., 2005). The increase in fat mass seen in obesity is accompanied by increased plasma FA levels (Opie and Walfish, 1963) and FA have been postulated to play a critical role in the development of insulin resistance (McGarry, 1992). However, the influence of variation in fat cell lipolysis on insulin sensitivity remains elusive. From human data, it remains unclear whether decreased HSL expression and activity favours the development of obesity through retention of FA within adipocytes or is a protective mechanism to avoid excess NEFA release and consequent insulin resistance and metabolic abnormalities.
The role of HSL has been studied in knock-out (KO) mice (Haemmerle et al., 2002; Harada et al., 2003; Osuga et al., 2000; Wang et al., 2001). The major phenotypic features of HSL KO mice are male sterility, DG accumulation in various tissues, blunted stimulated WAT lipolysis and, surprisingly, resistance to diet-induced obesity. The unsuspected impairment of fat development during high-fat feeding was hypothetically attributed to a defect in adipocyte differentiation (Harada et al., 2003; Zimmermann et al., 2003). The absence of HSL could alter the release of PPARγ ligands necessary for complete fat cell differentiation. Furthermore, WAT of HSL KO mice have been described as hypermetabolic with enhanced fat oxidation capacity (Strom et al., 2008). Because of their resistance to diet-induced obesity, HSL KO mice are not a relevant model to study the impact of impaired HSL activity seen in human obesity.
HSL haploinsufficient (HSL+/-) mice showed reduced WAT lipase activity. These preliminary results prompted us to hypothesize that obese HSL+/- mice could appropriately mimic the HSL defect observed in obese subjects. HSL+/- mice had impaired HSL expression and enzymatic activity, and blunted stimulated lipolysis. When fed a high-fat diet, HSL+/- mice became as obese as WT control mice. Lipid metabolism studied dynamically in vivo revealed a global reduction of FA fluxes in HSL+/- mice. The slowdown of FA metabolism was accompanied by an improvement of insulin sensitivity.
RESULTS
Functional consequences of HSL haploinsufficiency in high fat diet-fed mice
HSL+/- mice were generated by mating WT and HSL-/- mice. HSL mRNA expression was 50% lower in WAT of HSL+/- mice compared to WT mice (Fig.1A). mRNA expression of ATGL, its co-activator CGI-58 and PLIN1 were similar in WT and HSL+/- mice suggesting that no compensatory mechanism occurred as a result of the reduction in HSL expression (Fig. 1A). In vitro hydrolase activities against a cholesterol ester (Fig. 1B) and a TG (Fig. 1C) were both reduced (-60% and -45%, respectively) in WAT of HSL+/- mice indicating that reduced expression of HSL had the expected impact on cognate enzymatic activities. Reduced HSL activity was associated with blunted in vitro and in vivo β-adrenergic stimulated lipolysis (Fig. 1D and E). However, basal lipolysis was not altered. Therefore, decreased expression of HSL alters lipolytic function at the cellular level as well as in vivo.
Body weight and fat mass in HSL heterozygous mice fed high-fat diet
On a standard chow diet, the growth curves of WT and HSL+/- mice were similar (Fig. 2A). Next, we examined the consequences of diminished HSL function in mice fed a high-fat diet. As previously reported (Harada et al., 2003), HSL-/- mice were resistant to diet-induced obesity whereas WT and HSL+/- mice gained weight at a comparable rate and became obese (Fig. 2B). After 12 weeks of high-fat diet, fat mass was assessed by quantitative nuclear magnetic resonance imaging. WT and HSL+/- presented similar weight and fat mass while fat mass of HSL-/- mice was two third lower (Fig. 2C). There was no difference in WAT TG content between HSL+/- and WT mice (523 ± 21 vs 547 ± 15 nmol/mg tissue in WT and HSL+/- mice, respectively). Accordingly, plasma levels of leptin were identical in the two genotypes (Table SI). Food intake was similar in HSL+/- and WT mice (Fig. 2D). Energy
adipocyte area was not modified in HSL mice compared to control mice (Fig. 2G). The data show that decreased expression of HSL does not influence fat mass and WAT morphology during diet-induced obesity.
Metabolic plasma parameters in HSL heterozygous mice
Metabolic plasma parameters from HSL+/- and WT mice fed a high fat diet are presented in Table SI. Fasting plasma glucose and insulin were not different between HSL+/- and WT mice. The decreased expression of HSL does not affect basal fasted parameters of glucose homeostasis. Fasted plasma NEFA, glycerol, TG and total cholesterol were similar in both strains of mice.
Modification of fatty acid fluxes in HSL heterozygous mice
As no alteration of metabolic parameters was observed in steady state measurements, we determined FA fluxes in high fat diet-fed mice by stable perfusion of radiolabelled palmitate. Global tracer clearance that represents exit of the radioactive tracer from the blood compartment was markedly decreased in HSL+/- mice indicating that partial HSL depletion reduces peripheral FA uptake. Global FA oxidation represented by radioactive water measured in plasma was also reduced (Fig. 3A). Total radioactive FA storage, deduced from these measurements, was decreased in HSL+/- mice. Global FA turnover estimated through the evolution of plasma radioactive palmitate isotopic dilution (which is influenced by the clearance and dilution by cold and radioactive FA released from WAT in the fasted state) was in turn reduced in HSL+/- mice compared to WT mice. Tissue-specific radioactive FA incorporation in the TG pool was evaluated and showed reduction in WAT and heart of HSL+/- mice whereas esterification in soleus muscle was not affected (Fig. 3B). Total TG content was not affected in WAT of HSL+/- mice and whereas it was decreased in soleus and heart (Fig. 3C). Hence, the decreased lipolytic capacity in WAT induced by partial HSL deficiency provokes a compensatory diminution in FA uptake and storage, especially in WAT. These changes take place without influencing total WAT mass.
HSL haploinsufficiency improves insulin sensitivity and glucose tolerance in obese mice
Increased levels of plasma FA are known deleterious factors for insulin sensitivity. We therefore postulated that the alteration of FA fluxes could influence peripheral insulin sensitivity in HSL+/- mice. Insulin and glucose tolerance tests performed on high fat diet-fed mice revealed that partial HSL depletion improved insulin and glucose tolerance in vivo (Fig. 4A and 4B). Plasma measurements of adiponectin and palmitoleate, two insulin sensitizers produced by adipocytes, did not show any difference between HSL+/- and WT mice (Table SI). In order to gain further insights into the origin of the global improvement of insulin tolerance in HSL+/- mice, in vivo glucose utilization was determined in various tissues by 2- desoxy-D-[3H] glucose infusion. Insulin-stimulated glucose utilization was increased in soleus (oxidative) muscle and showed a tendency to increase in biceps femoris (glycolytic) muscle (p=0.10) (Fig. 4C). An increase was also observed in WAT. Skeletal muscles and WAT may therefore participate in the global improvement of insulin tolerance seen in HSL+/- mice. Interestingly, respiratory quotient was increased in HSL+/- mice suggesting a shift from FA to glucose as substrate (Fig. 4D). Furthermore, substrate oxidation measured ex vivo in soleus muscle was increased for glucose but not for oleate in these mice (Fig. 4E). We wished to confirm that HSL haploinsufficiency protects against the development of insulin resistance in another model. To that end, WT and HSL+/- mice were fed a high fructose diet for 48 weeks. While no difference in body weight was observed between the genotypes (Fig. S1A), HSL+/- mice were more insulin sensitive than WT mice (Fig. S1B). Together these data suggest that the improvement in global insulin sensitivity is linked to the modification of FA metabolism due to partial HSL deficiency in WAT.
HSL haploinsufficiency does not influence adipose tissue inflammation in obese mice
As WAT inflammation may cause insulin resistance, we investigated WAT macrophages and inflammatory molecules. The number of macrophages in the stroma-vascular fraction of WAT did not differ between HSL+/- and WT mice fed a high fat diet (Fig. S2A). Accordingly, mRNA levels of macrophage markers were not different (Fig. S2B). Similarly, gene expression of
inflammatory markers was similar in WAT from HSL+/- and WT mice (Fig. S2C). Therefore, the resistance to obesity-induced insulin resistance of HSL+/- mice is not linked to a decrease in WAT inflammation compared to WT mice.
Pharmacological HSL inhibition also improves insulin sensitivity of obese mice
Diet-induced obese mice chronically treated with a specific HSL inhibitor, BAY, showed improved insulin tolerance (Fig 4F) without modification of body weight (Fig 4G).
DISCUSSION
During the lipolytic process in WAT, FA are produced by the coordinated action of lipases. Obesity and insulin resistance are associated with decreased expression and activity of HSL correlated with blunted stimulated lipolysis (Jocken et al., 2007; Large et al., 1999). This defect seems to be a primary event as it is observed in non-obese first-degree relatives to obese subjects and ex vivo in differentiated preadipocytes from obese subjects (Hellström et al., 1996; Langin et al., 2005). Resistance to catecholamine-induced lipolysis has also been shown in vivo in obese subjects from childhood to adulthood (Bougnères et al., 1997; Enoksson et al., 2000). The physiological significance of lipolysis and HSL defects in obesity may be seen in two ways. A lipolytic defect could contribute to the development of obesity through impairment in the mobilization of fat stores. Alternatively, the defect may protect against excessive FA release and ensuing deleterious action of FA on insulin sensitivity (Samuel et al., 2010). To address these clinically relevant questions, we produced a mouse model similar to the human pathological condition, i.e. an obese mouse with reduced HSL activity in WAT. Phenotyping of the animals revealed that a partial defect in HSL expression and lipolysis did not have a major influence on fat mass and, hence, that chronic inhibition of FA release from WAT did not contribute to the development of obesity. However, it provoked a global slowdown of FA metabolism leading to an improvement of insulin tolerance (Fig. 4H).
The development of fat mass was not compromised in HSL+/- mice fed high fat diet. Surprisingly, complete HSL deficiency leads to a resistance to diet-induced obesity (Harada et al., 2003). It may be hypothesized that the presence of an active allele is sufficient to compensate for the defect in adipogenesis, e.g. through production of signalling lipolytic by- products (Kraemer and Shen, 2006). The identical fat mass in HSL+/- and WT mice was supported by similar food intake, energy expenditure and leptin levels in the two genotypes. Adipocyte size was not modified by HSL haploinsuficiency. In accordance with the observation that larger adipocytes show increased basal release of FA (Wueest et al., 2009), no modification in basal lipolysis was seen in HSL+/- mice. The normal development of WAT in a condition of decreased FA release raised questions on the dynamics of lipid fluxes. Using a fluxomics approach, we showed that HSL+/- mice presented altered global FA turnover, decreased WAT lipolysis being linked to reduced FA esterification in WAT. The new equilibrium of FA metabolism at a slower rate may explain the preservation of fat mass in HSL+/- mice. The data indicate that the HSL and consequent stimulated lipolysis defect seen in human obesity is unlikely to participate in the excessive development of WAT.
The data also shed new light on the mechanisms of lipid-induced insulin resistance. To date, there is a general agreement that FA cause deleterious effects on insulin signalling in peripheral organs. The mechanisms of lipid-induced insulin resistance have partly been unravelled (Samuel et al., 2010). The working models are based on overload of lipids from exogenous sources, i.e., dietary FA for high fat diet or FA produced by lipoprotein lipase- mediated hydrolysis of TG during lipid and heparin infusion (Samuel et al., 2010; Schenk et al., 2008). To date, little is known on the effect of FA released by WAT lipolysis on the modulation of insulin sensitivity. Here, we show that, when WAT lipolysis is diminished, a slowdown in FA metabolism is likely to contribute to the improvement of insulin tolerance in the obese state. This happens in a situation when FA and TG steady-state plasma levels remain unchanged. HSL haploinsufficiency in obese mice was associated with improved insulin and glucose tolerance. Results from insulin and glucose tolerance tests were
with insulin tolerance test, in vivo insulin-stimulated glucose uptake was increased in WAT and skeletal muscles of HSL+/- mice. Accumulation of TG in skeletal muscle of sedentary people is associated with impaired insulin-stimulated glucose metabolism (Galgani et al., 2008). The reduced TG level observed in soleus muscle of HSL+/- could therefore be involved in the improved metabolic performance of this tissue. The direct involvement of altered FA fluxes in the improvement of insulin resistance of obese mice is also supported by the lack of modifications in other potential contributors to the control of insulin sensitivity(Schenk et al., 2008). Plasma levels of two molecules with insulin-sensitizing properties, adiponectin, an established adipocytokine, and palmitoleate, a novel lipokine, were identical in HSL+/- and WT mice (Cao et al., 2008; Shetty et al., 2009). Increase in WAT macrophage number and expression of inflammatory molecules has been proposed to contribute to obesity-induced insulin resistance (Bourlier and Bouloumie, 2009). However, neither macrophage number nor gene expression of macrophage markers and inflammatory factors were modified in high fat diet-fed HSL+/- mice indicating that chronic inhibition of lipolysis, unlike acute stimulation, does not modify the content of macrophages in WAT (Kosteli et al., 2010).