INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

SURVIVAL OF ENDOTHERMS IN the cold critically depends on their ability to sustain elevated levels of heat production for long periods. It can be achieved through the development of nonshivering thermogenesis (NST), a process that generates heat independently from contractile activity and that is considered to be the main characteristic of cold acclimation. In small mammals, NST has been mostly associated with brown adipose tissue (BAT), a specialized heat-producing tissue. However, the presence of NST can also be demonstrated in young cold-acclimated (CA) birds devoid of BAT (2). This mechanism contributes to the increased thermogenic capacity (4) and improved cold endurance of these birds (3). Skeletal muscle and liver are both major contributors to NST in CA ducklings (10). Fatty acids (FAs) play a major role in avian NST by serving both as substrates for respiration and/or as uncouplers of mitochondrial oxidative phosphorylation (6), as well as by activating ATP-consuming sarcoplasmic reticulum Ca²⁺ cycling through their long-chain acyl metabolites (13). The capacity to supply FA to thermogenic tissues could therefore comprise a number of controlled and limiting steps for cold thermogenesis, but this aspect has not yet been investigated in CA ducklings.

Mobilization from the sites of storage, i.e., adipose tissue, may be the first step. For instance, it is known in the rat that the fueling of BAT NST requires an increased FA supply from white adipose tissue, resulting from an activation of the sympathetic nervous system as well as increased sensitivity of adipocyte response to glucagon (29). Glucagon also has marked lipolytic activity in birds (review in Ref. 5). On the basis of the ultrastructural changes of adipose tissue induced by cold acclimation in ducklings (adipocytes becoming multilocular and smaller in size), it was suggested that an intense lipolysis may take place in the cold (2), but direct experimental data are lacking.

Tissue uptake of FA can also be modulated by the activity of endothelial lipases, lipoprotein lipase (LPL) and hepatic lipase, all of which are rate-limiting step enzymes enabling the extraction of FA from triacylglycerol (TG) and phospholipids of lipoproteins (18). It is known that BAT NST is fueled by a marked LPL upregulation (28). To our knowledge, possible changes in LPL or hepatic lipase activities have not been investigated in CA birds.

Intracellular transport of FA may also represent an important step for fueling cold thermogenesis. Within the cells, FA-binding proteins (FABPs), which are small (12-16 kDa) and abundant cytoplasmic proteins (15, 24, 26), have been involved in the cytoplasmic trafficking of FA and hydrophobic molecules. Although their physiological function has yet to be unequivocally established, FABPs may be an important and limiting determinant of FA transport (31) to membranes and organelles for energy storage or expenditure. A high content of FABP is indeed found in oxidative muscles of rat (24), and an increased FABP content is associated with increased FA oxidation in skeletal muscle of cold-exposed fish (21). Given the potentially high flux of FA in avian NST, resulting from their importance both as a fuel and as an intracellular signal for the activation of cellular thermogenic processes, FABPs may play a key role in the modulation of tissue thermogenic capacity. If the latter is true, one would expect their activity to be increased in the cold in thermogenic tissues.

The aim of this work was to investigate the capacity to supply lipids to thermogenic tissues (skeletal muscle and liver) of CA ducklings with respect to thermoneutral (TN) controls. The experiments therefore focused on several metabolic pathways, including in vitro lipolysis of adipose tissue fragments, plasma transport of lipids by lipoproteins, in vitro tissue FA uptake from TG and phospholipids associated with lipoproteins, and FA intracellular binding capacity of FABPs.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. Twelve male Muscovy ducklings (Cairina moschata) were obtained from a trade stock breeder (Ets Grimaud). One week after hatching, six control chicks were kept for 4-5 wk at thermoneutrality (TN, 25°C), and six CA chicks were exposed for 4-5 wk to cold (4°C) with the same photoperiod (8:16-h light-dark cycle) according to Barré et al. (2). They were fed ad libitum, and their food intake was measured daily. Ducklings (fasted overnight) were killed by decapitation, and tissues were quickly drained of blood before sampling. The subcutaneous fat deposit (around the leg) was immediately used for lipolytic activity analysis. For biochemical assays, heart ventricle, the red inner part of gastrocnemius muscles (12), pectoralis muscle, subcutaneous fat, and liver were sampled and stored at 80°C for periods not exceeding 4 wk before use. For the FABP purification procedure, muscles from the legs were sampled. Blood samples were collected with 0.01% (wt/vol) EDTA and spun for 10 min at 2,500 g for metabolite analysis and determination of lipoprotein profiles. Plasma samples were stored at 20°C. Finally, as an index of body adipose tissue mass, the distinctive subcutaneous fat deposit located around the legs and the visceral fat deposit located around the gizzard were weighed.

In vitro adipose tissue lipolysis. Subcutaneous adipose tissue was quickly cut into small pieces and incubated for 15 min at 39°C in a 10 mM glucose Krebs-Ringer buffer with 4% (wt/vol) BSA. After washing, several tubes, each containing 0.3 g of adipose tissue in 3 ml of the same buffer with differing concentrations of glucagon (0, 0.1, 0.5, and 1 ng/ml), were incubated for 1 h at 39°C under agitation in the presence of 95% (vol/vol) O₂ and 5% (vol/vol) CO₂. The released glycerol was determined as the difference between the final and initial content. Lipolytic activity was expressed as nanomoles of glycerol released per 100 milligram adipose tissue per hour. Preliminary experiments showed that the lipolytic rate was linear over the whole incubation duration.

Chemical composition and characterization of plasma lipoproteins. Plasma lipoproteins were isolated by density (d)-gradient ultracentrifugation as described by Hermier et al. (16). A discontinuous gradient (2.5 ml d = 1.006 g/ml, 2.5 ml d = 1.019 g/ml, 2 ml d = 1.063 g/ml, 2 ml d = 1.240 g/ml) was made from a stock solution of d = 1.240 g/ml KBr in 0.15 M NaCl containing 0.01% (wt/vol) EDTA. A fifth layer made of dissolved KBr in fresh plasma (3 ml d = 1.210 g/ml) was introduced within the gradient. Samples were then centrifuged at 288,000 g for 48 h at 15°C using a swinging-bucket rotor SW 41TI in a Beckman Optima L-60 ultracentrifuge. In accordance with the visual localization of lipoprotein bands along the density gradient, three different fractions corresponding to the very low density lipoproteins (VLDL, d < 1.013 g/ml), low-density lipoproteins (LDL, d = 1.013-1.046 g/ml), and high-density lipoproteins (HDL, d = 1.046-1.130 g/ml) were sampled. They were immediately dialyzed for 24 h at 4°C in a saline buffer (pH 7.4) containing 0.01% EDTA before freezing at 20°C. Free cholesterol (FC), total cholesterol (TC), TG, and phospholipid contents were then determined in each lipoprotein fractions using commercially available kits (see Materials). Protein content in lipoprotein fractions and tissues was determined by the bicinchoninic acid method using BSA as standard.

In vitro LPL and hepatic lipase assays. Assays were all carried out simultaneously according to Herpin and Lefaucheur (17) with minor modifications. Heart (1 g/8 ml), white adipose tissue (2 g/6 ml), gastrocnemius and pectoralis muscles (2 g/8 ml), and liver (2 g/8 ml) were homogenized in a 33 mM NH₄Cl-NH₄OH (pH 8.2) buffer, containing 2.5 U heparin/ml, and cooled on ice for 45 min. Homogenates were then spun at 700 g at 4°C for 10 min. For substrate activation, mixed plasma from CA and TN ducklings (85% vol/vol) with Intralipid emulsion (15% vol/vol as a source of TG) was incubated for 30 min at 39°C. [1-¹⁴C]triolein (0.37 MBq) was incorporated in the substrate solution by sonication (5 min, 100 W) with a Labsonic 1510 apparatus. Incubation media (2.5 ml) containing 32% (vol/vol) of homogenate supernatant and activated substrate, 12% (vol/vol) of 100 mM Tris · HCl (pH 8.2) buffer, and 24% (vol/vol) of BSA solution (20% wt/vol) (pH 8.2) were incubated for 1 h at either 0°C (basal activity) or 39°C (maximal activity) in a shaking bath. Lipase activity was stopped by ice cooling the incubation medium, and released nonesterified fatty acids were subsequently determined by measuring the content of [1-¹⁴C]oleic acid released during the 30 min of incubation at 39°C. For [1-¹⁴C]oleic acid extraction, 6 ml of methanol-chloroform-heptane 14:12:10 (vol/vol/vol) were added to 0.8 ml incubation medium. The tubes were vortexed, and 1 ml of 0.5 M NaOH was added. After centrifugation at 1,500 g for 15 min, [1-¹⁴C]oleic acid released in the upper phase was counted in a scintillation counter. To express activities as microequivalents of FAs released per hour and per gram of tissue or per organ, quantitation of released FAs in the incubation medium was carried out in parallel by a titrimetric method after Dole's extraction using palmitic acid as the standard. Lipase activity was obtained as the difference between maximal and basal activities determined in duplicated assays.

Assays of in vitro intracellular FA binding. Cytosols from skeletal muscle and liver were obtained by ultracentrifugation at 105,000 g for 60 and 90 min, respectively, as described by Paulussen et al. (26). Cytosolic proteins were delipidated with the Lipidex 1000 procedure at 37°C (15). Assays for FA binding were performed by two different protocols. In protocol 1, after incubation with 20 nM [1-¹⁴C]oleic acid, 8 mg of cytosolic protein were filtered on a calibrated Sephadex G75 chromatography column (1.5 × 100 cm) equilibrated with 150 mM KCl, 10 mM KH₂PO₄, and 0.02% NaN₃ (pH 7.4) as described by Glatz et al. (15) and Miller et al. (24). The FA-binding capacity was assessed as the amount of [1-¹⁴C]oleic acid bound to the eluted fractions corresponding to proteins weighing 12-18 kDa. The column was regenerated after each binding assay by washing with a (0.5%) BSA solution. The FA-binding capacity was expressed as picomoles oleic acid bound per milligram cytosolic protein. In protocol 2, after dealbuminization by filtration with a 30-kDa molecular mass membrane, delipidated cytosolic protein samples were subjected to the Lipidex 1000 assay (15, 33). Briefly, 20 µg cytosolic proteins were incubated for 30 min at 37°C with 1 µM [1-¹⁴C]oleic acid in a buffer containing 10 mM KH₂PO₄, 100 µM Triton X-100, and 0.02% NaN₃ (pH 7.4). An ice-cold Lipidex 1000 mixture was added to the medium and incubated for 30 min at 0°C. Finally, after centrifugation, the FA-binding capacity was measured in 100 µl of supernatant and was expressed as picomoles oleic acid bound per milligram cytosolic protein. Control values were obtained by measuring the radioactivity in the supernatant of the incubation medium in which the volume of protein samples had been substituted by an equivalent volume of Tris buffer. Control values were subtracted from the binding measure with the protein samples.

Purification of intracellular FABP from duckling skeletal muscle. The purification of duckling FABP was an important step required to obtain specific antibodies allowing us to measure the cytosolic concentration of FABP.

The protocol of FABP purification was based on three different steps allowing the elimination of contaminant proteins and the enrichment of small proteins with high FA-binding capacity. In the first step of purification, cytosolic proteins obtained from leg skeletal muscle were precipitated by 70% ammonium sulfate saturation and spun for 30 min at 14,000 g. The supernatant was dialyzed overnight against 0.01 M Tris · HCl (pH 7.4) buffer and lyophilized. In the second step, lyophilized proteins were resuspended with distilled water and applied to a Sephadex G75 gel filtration chromatography column. The (12-18 kDa) protein fractions (containing FA-binding activity in the binding assays) were lyophilized, dialyzed as described above, and incubated for 25 min with [1-¹⁴C]oleic acid (typically 5 nmol for 2 mg protein). In the third step, the protein mixture (4 ml at 4 mg/ml) was submitted to preparative isoelectric focusing using a Multiphor 1217 LKB apparatus. Sephadex G75 (4% wt/vol) with pH 3-10 ampholines (5% vol/vol) was used for gel migration on a refrigerated plate (24 × 10 cm). H₃PO₄ (1 N) and NaOH (1 N) were used for anodic and cathodic solutions, respectively. Proteins were submitted to migration for 18 h (8 W; 1,500 V) at 4°C, and the pH gradient of the gel was determined with a pH meter. The gel was then divided among several minicolumns and was washed with the previous Sephadex G75 buffer. Radioactivity bound to eluted protein fractions was counted with a 2100 TR Packard counter. Radioactive fractions were lyophilized and then filtered on a Sephadex G75 chromatography column to remove ampholines. The characteristics of the purified intracellular FABP were assessed by SDS-PAGE at pH 8.8 with a 10-22.5% gradient gel as described by Laemmli (20) and also by the Lipidex 1000 binding assay (15, 33). Production of polyclonal antibodies against duckling FABP and purification of IgG fraction were performed as described by Paulussen et al. (26).

Quantitation of muscle FABP content. Muscle FABP content was assessed by immunoblotting of skeletal muscle cytosols with a Bio-Rad miniprotean apparatus using the specific antibodies raised against duckling FABP. The FABP-IgG complexes were detected by autoradiography after incubation of the blots with ¹²⁵I-labeled protein-A (0.15 MBq) in a phosphate buffer. Quantitation was performed using a Vernon densitometer.

Materials. Ultrafree-MC filter units with a membrane of 30-kDa molecular mass limit and nitrocellulose sheets (0.2 µm) were obtained from Millipore, St. Quentin en Yvelines, France. FA-free BSA was purchased from Boehringer Mannheim, Meylan, France. [1-¹⁴C]triolein (1.85 MBq/mmol) and [1-¹⁴C]oleic acid (2.03 GBq/mmol) were purchased from Isotopchim, Ganagobie-Peyruis, France and from Dositek, Orsay, France, respectively. ¹²⁵I-labeled protein-A (30 MBq/ml) was purchased from Amersham, Les Ulis, France. The pH 3-10 ampholines were obtained from Serva, Paris, France. Autoradiography films (Kodak X-OMAT X-AR5) were purchased from Sigma, L'Isles d'Abeau Chesnes, France. Other chemicals were of analytic grade. Nonesterified and TG plasma levels were determined with the diagnostic Nefa C and the TG N kits supplied by Wako-Unipath, Dardilly, France; phospholipids and TC were determined with the peroxidase aminoantipyrine (PAP) phospholipids and PAP cholesterol kits from BioMérieux, Lyon, France; and, finally, FC and glycerol were determined with the biochemical analysis kits from Boehringer Mannheim. Cholesterol ester (CE) concentrations were determined according to the classical formula CE = (TC  FC) × 1.67.

Statistical calculations. The effect of cold acclimation (TN vs. CA) on different parameters was assessed using Student's t-test and covariance analysis to exclude the effect of tissue mass in the analysis of enzymatic activities. ANOVA was used to test the effect of increasing doses of glucagon on lipolysis. Values are presented as means ± SE.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Body and organ masses. At the time of experiment, CA and TN ducklings had similar body mass (Table 1). However, heart (+26%, P < 0.01) and gastrocnemius muscle (+11%, P < 0.05) were heavier in CA than in TN ducklings. No significant change in mass occurred in pectoralis muscle and liver. As judged from the lower mass of subcutaneous (17%, P < 0.05) and visceral (18%, P < 0.05) fat deposits, the stored body fat was decreased in CA ducklings. This occurred despite an increased food intake in the cold (+37%, P < 0.01) (Table 1).

In vitro adipose tissue lipolysis. The basal lipolytic activity of adipose tissue fragments was higher (130 ± 14 vs. 202 ± 9 nmol glycerol · 100 mg¹ · h¹, +55%, P < 0.01) in CA than in TN ducklings (Fig. 1). Glucagon exerted a powerful dose-dependent lipolytic effect in both groups of ducklings. However, the lipolytic response to increasing doses of glucagon (0.1, 0.5, and 1 ng/ml in incubated medium) was higher (+500, +140, and +150%, respectively, P < 0.01) in CA than in TN ducklings. Interestingly, to obtain the same relative stimulation of lipolysis (around +170% over basal), a dose of 0.5 ng/ml was needed in TN controls whereas only 0.1 ng/ml was required in CA ducklings.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Basal and glucagon-induced lipolysis in adipose tissue fragments from thermoneutral (TN) or cold-acclimated (CA) ducklings. Values are means ± SE. Compared with TN ducklings: ** P < 0.01.

Plasma metabolite levels and chemical composition of lipoproteins. Plasma levels of nonesterified FA (+57%) and glycerol (+31%) were higher (P < 0.05) in CA than in TN ducklings (Table 2), suggesting the enhancement of adipose tissue lipolysis during cold acclimation. TG plasma level was not significantly altered by cold acclimation.

Cold acclimation did not affect the chemical composition of VLDL, LDL, and HDL in ducklings (Table 3). The plasma levels of LDL and HDL were similar in both groups of birds. By contrast, the plasma level of VLDL was depressed (54%, P < 0.05) in CA ducklings.

In vitro LPL and hepatic lipase activities. Total endothelial lipase activities expressed per organ (Table 4) were higher in adipose tissue and heart than in liver, and much higher than in skeletal muscles. They increased during cold acclimation in red gastrocnemius muscle (+80%, P < 0.05), heart (+34%, P < 0.01), and liver (+55%, P < 0.05), whereas no change occurred in pectoralis muscle and adipose tissue. As indicated by covariance analysis, the effect of cold acclimation on heart LPL activity was mainly accounted for by the higher heart mass of CA ducklings. Cold acclimation had no significant effect (P > 0.60) on the specific activity (per gram of tissue) of the heart LPL. By contrast, in gastrocnemius and liver, exclusion of organ mass did not suppress the effect of cold acclimation.

In vitro intracellular FA binding. Assays for intracellular FA binding were performed by two different but complementary procedures. First, chromatography assays revealed the presence of three FA-binding peaks in the eluted cytoplasmic proteins from gastrocnemius muscle (Fig. 2) and liver (data not shown). The first peak was mainly attributable to albumin, the second was specifically linked to proteins weighing 12-18 kDa, and the third corresponded to unbound FA. The 12- to 18-kDa FA-binding capacity was increased with cold acclimation in both red gastrocnemius muscle (+46%, P < 0.01) and liver (+74%, P < 0.05) (Table 5). However, due to the hydrophobic nature of FA, chromatography assays have several disadvantages, such as possible variations in specific and nonspecific FA binding. Therefore, a second protocol using the Lipidex 1000 resin was performed (15, 33). This latter procedure did not define the molecular mass of the proteins involved in the FA binding. The results showed that FA-binding capacities were higher in red gastrocnemius muscle (+36%, P < 0.01) and liver (+54%, P < 0.05) from CA ducklings (Table 5).

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Fatty acid (FA)-binding assays by Sephadex G75 chromatography. Elution profiles were obtained with samples of 8 mg cytosolic protein from gastrocnemius muscle of TN (A) or CA (B) ducklings. FA-binding capacity was assessed as amount of radioactive FAs bound to 12- to 18-kDa fractions.

Purification and quantitation of intracellular FABP from duckling skeletal muscle. Our procedure resulted in the isolation of a protein fraction showing high specific FA-binding capacity. SDS-PAGE analysis performed in parallel indicated the progressive purification of an intracellular protein with an apparent molecular mass of 15.4 kDa (Fig. 3) and with high FA-binding capacity (Table 6). The purification procedure yielded 1.3% of total FA-binding activity present in the cytosols of duckling muscle. This FABP is thus likely to be involved in the FA-binding activity found in the 12- to 18-kDa fraction with the Sephadex G75 chromatography binding assays.

View larger version (75K): [in this window] [in a new window]   Fig. 3.   Analysis by gel electrophoresis of different protein samples obtained after 3 steps of purification of skeletal muscle FA-binding protein (FABP) in ducklings. Lane 1, calibration proteins: phosphorylase B (90.0 kDa), bovine serum albumin (67.0 kDa), ovalbumin (43.0 kDa), carbonic anhydrase (30.0 kDa), soybean trypsin inhibitor (20.1 kDa), -lactalbumin (14.4 kDa); lane 2, proteins from whole cytosol (60 µg); lane 3, proteins (40 µg) from first step of FABP purification procedure: 70% ammonium sulfate precipitation; lane 4, proteins (20 µg) from second step of FABP purification procedure: Sephadex G75 gel chromatography; lane 5, pure FABP (2 µg and 15.4 kDa) from third step of purification procedure: preparative isoelectric focusing.

Antibodies raised against duckling FABP were specific for the 15.4-kDa purified protein, as indicated by the single band obtained by immunoblotting of skeletal muscle (Fig. 4) and heart (data not shown) cytosols. There was a linear correlation between the FABP content and the antigen-IgG-¹²⁵I-protein-A complex, allowing quantitative determination of FABP in duckling skeletal muscle (Fig. 4). In both groups of birds, results showed a high, intermediary, and low FABP content in heart, gastrocnemius, and pectoralis muscles, respectively (Table 7). Cold acclimation increased only the FABP content (+37%, P < 0.05) in the cytosolic fraction of the red gastrocnemius muscle of ducklings (Fig. 4 and Table 7). Interestingly, immunoblotting of liver cytosol with our antibodies did not show any immunoreactive protein (data not shown).

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Calibration curve of immunoblot assays and immunoblot assays for quantitation of 15.4-kDa FABP in muscle cytosol (10 µg of proteins) from TN or CA ducklings. Calibration proteins: soybean trypsin inhibitor (20.1 kDa), -lactalbumin (14.4 kDa).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The present results have shown that in CA ducklings there is a coordinated enhancement of in vitro FA supply to thermogenic tissues occurring at several levels, including adipose tissue, liver, and skeletal muscle. In addition, the present data describe the isolation and purification of duckling intracellular FABP, which is more abundant in oxidative skeletal muscle after cold acclimation.

Cold-induced enhancement of adipose tissue lipolysis. Despite an increased food intake, CA ducklings exhibited a lower adiposity than TN controls, indicating that the cold-induced increase in energy intake did not completely compensate for the increased energy expenditure required for thermoregulation. Similar observations have been made in CA rats (27). These results agree with the observations that lipids may be an important energetic substrate in cold-exposed birds (review in Ref. 22).

The present results show that both basal and glucagon-induced lipolysis of adipose tissue fragments were enhanced by cold acclimation. Actual rates of basal and glucagon-stimulated lipolysis found in ducklings are comparable to those reported in adult chickens (8). The present evidence for a functional difference between CA and TN ducklings is consistent with previous morphological observations of adipose tissue (1). Indeed, relative to that of TN counterparts, adipose tissue of CA birds contains smaller multilocular adipocytes surrounded by more capillaries with abundant microvilli. These have been considered to be indicative of increased lipolytic activity (2). Furthermore, adipose tissue blood flow was shown (10) to be much higher in cold-exposed CA ducklings than in TN ducklings, whether exposed to thermoneutral conditions (+330%) or to cold (+152%). The increase in blood flow to adipose tissue in response to exogenous glucagon was also higher in CA ducklings (11). These changes in blood flow may reflect a more intense rate of lipid mobilization induced by either cold exposure or glucagon, on account of the major influence of blood flow on FA mobilization (7) and in association with increased plasma glucagon levels in CA ducklings (5). It is therefore likely that a higher stimulation of adipose tissue lipolysis also takes place in vivo in CA ducklings. It may therefore account for the higher plasma levels of nonesterified FA and glycerol observed in CA relative to TN ducklings and for the higher thermogenic effect of glucagon injection observed in CA ducklings (11). The possibility that part of the observed differences in lipolytic activity, both in the basal and glucagon-stimulated rates, is related to the fact that adipocytes are smaller in CA ducklings (2) cannot be dismissed. However, such size differences cannot account for the apparent change in sensitivity to glucagon. Isolated white adipocytes from CA rats also showed an enhanced responsiveness to glucagon (29). The enhancement of the adipose tissue response to glucagon may take place at the receptor level. An increased density of glucagon receptors has indeed been reported in CA rats (32), and the same phenomenon may be present in ducklings, although the latter has yet to be investigated. Future studies should also address whether adipocyte release of adenosine, a powerful inhibitor of lipolysis, is affected by cold acclimation and therefore whether any such changes can contribute to the observed difference in glucagon sensitivity in CA ducklings. In rats, however, differences in the lipolytic capacity observed between cold-exposed and control rats could not be explained solely by a difference in the accumulation of extracellular adenosine (29).

Increased tissue uptake of FAs by endothelial lipases. The present results show that there was an increased in vitro activity of endothelial lipases in red gastrocnemius muscle and liver of CA ducklings relative to TN controls. Gastrocnemius muscle showed a higher specific (per gram muscle) LPL activity than pectoralis muscle, possibly related to the higher proportion of slow oxidative (SO, 47%) and fast oxidative glycolytic fibers (FOG, 53%) in gastrocnemius (2) than in pectoralis muscle, the latter containing a high proportion of fast glycolytic fibers (FG) and FOG fibers (unpublished data). In mammals, higher LPL activity is also found in skeletal muscles rich in SO fibers (23). Part of the cold-induced rise in LPL activity may be related to the slight increase in the proportion of slow oxidative fibers observed in CA ducklings (12). Because LPL activity is generally high in tissues that are recruited in the cold (28), the present results suggest that skeletal muscles rich in SO fibers may play an important role in duckling thermogenesis during cold acclimation. This may also apply to other species lacking BAT. Similarly, a specific increase in the LPL activity of red skeletal muscles, rich in SO fibers, has also been reported in CA piglets devoid of BAT, while there was no change in white muscles, rich in FG and FOG fibers (17). The increased hepatic lipase activity in the cold is consistent with the higher oxidative capacity of this tissue in CA ducklings (1), suggesting a possible role for liver in cold-induced thermogenesis (10).

By contrast with the increased endothelial lipase activities in CA ducklings, there were no major changes in circulating lipoproteins, apart from the lowering of VLDL, which may be accounted for by the increased LPL activity. Indeed, in mammals, VLDLs are mainly catabolized by LPL in adipose tissue, heart, and skeletal muscles. In ducklings, however, this was not accompanied by concomitant changes in LDL plasma levels, the main products of VLDL catabolism (18). Overall, the present results indicate that there were no obvious adaptive changes in the capacity to transport lipids associated with circulating lipoproteins in CA ducklings.

Cold-induced increase in intracellular FA binding by FABPs. A major finding of the present study is the concomitant increase in the capacity to extract nonesterified FA from TG-rich lipoproteins and the intracellular FA-binding capacity in red skeletal muscles from CA ducklings. Values of intracellular FA-binding capacity reported here are highly comparable to those found in rat tissues (24). Because intracellular binding capacity, assessed by two different but complementary assays, was related to a low-molecular mass (12-18 kDa) protein fraction, it is tempting to suggest that it is accounted for by specialized FABPs similar to those described in mammals (15, 24, 26). It was indeed possible to purify a cytosolic protein with high capacity to bind FA, the protein having a molecular mass of 15.4 kDa and an intramyocyte concentration similar to mammalian muscle FABPs (26). The increased FABP content in muscle cytosols of CA ducklings may well account for the cold-induced rise in FA-binding capacity of the 12- to 18-kDa protein fraction. It cannot be excluded that a part of the cold-induced increase in muscle FABP content may be related to the increased proportion of SO fibers in CA ducklings (12), given that in rats higher FABP content was found in red skeletal muscles, rich in SO fibers (24). The amplitude of the increase in FA-binding capacity is, however, much larger than the change in fiber typing, suggesting that there are specific mechanisms responsible for the upregulation of duckling FABP. By contrast, in FG- and FOG-rich pectoralis muscle, no change in FABP content was observed after cold acclimation, suggesting that additional FABP may only be required in the most oxidative fibers preferentially recruited in the cold. Enhanced thermogenic capacity of SO fibers has already been observed in CA ducklings (12).

The increase in FABP concentration may allow red skeletal muscle to deal with the greater supply of FA induced by both the enhanced lipolysis in adipose tissue and the increased LPL activity. The combined effects of increased FA supply from adipose tissue, enhanced tissue FA uptake by LPL, increased intramuscular transport by FABP (present study), and increased muscle oxidative capacity (1) and blood flow (10, 11) may thus favor lipid oxidation and thermogenesis in red skeletal muscles of CA ducklings. Similarly, increased FABP content paralleled increased lipid oxidation in muscle of cold-exposed fish (21). In the thermogenic BAT of cold-acclimated rats, a dramatic increase in the heart-type FABP was also observed (9) in conjunction with the stimulation of the thermogenic capacity of this tissue.

In liver, the increased FA uptake through hepatic lipase was also positively linked to an increase in the (12-18 kDa) FA-binding capacity, which may be related to a putative hepatic FABP by analogy with mammals (15). As judged from the immunoblotting assays, this putative hepatic FABP appears to be distinct from that found in skeletal muscle, given that it was not recognized by the same antibodies. As reported in rodents, the enhanced hepatic lipase activity (30) and possibly the (12-18 kDa) FA-binding capacity of CA ducklings may be related in part to their higher food intake. It may also be indicative of hepatic thermogenic capacity, as suggested previously by the high hepatic oxidative capacity found in CA ducklings (1). In any case, both parameters suggest an increased intrahepatic trafficking of FA toward either resynthesis of TG or local oxidation, which may parallel an increased aerobic metabolism as indicated by the higher hepatic blood flow observed in CA ducklings (11).

It is concluded that in young growing ducklings, cold acclimation is associated with adaptive changes at several levels of lipid metabolism. Changes include an increased lipolytic activity of adipose tissue, a higher tissue uptake of plasma TG-derived FA, and a higher intracellular capacity to bind FA associated with an increased content of a 15.4-kDa FABP in skeletal muscle. These responses suggest a coordinated increase in FA supply to the intracellular sites of oxidation in thermogenic tissues.

Perspectives

Despite there being no direct evidence, a hormonal control of LPL activity may occur in tissues of CA chicks. Indeed, several hormones are known to modulate LPL gene expression and activity (review in Ref. 14). By analogy with results in mammals showing a positive regulation of LPL activity by catecholamines (23), the increased sympathetic activity found in skeletal muscles of cold-exposed birds (19) may play a part in the hormonal stimulation of muscle LPL activity. Further studies are required to clarify the endocrine control involved in CA ducklings.

In cells, FABPs, as potential modulators of FA (and their acyl metabolites) trafficking, may play a role in their intracellular signalling. They may, for instance, influence the cellular thermogenic processes modulated by FAs (6, 13) and therefore could conceivably intervene in the control of skeletal muscle NST. Further studies are required to clarify this putative role.