Fish may use lipoproteins instead of albumin-bound fatty acids to fuel endurance exercise, but lipoprotein kinetics have never been measured in ectotherms. In vivo bolus injections of labeled very-low-density lipoproteins (3H-VLDL labeled in vivo from donor fish) and continuous infusions of Intralipid (3H-labeled artificial emulsion) were used to investigate the effects of prolonged exercise (6 h at 1.5 body length/s) and heparin (600 U/kg) on the turnover rate of circulating triacylglycerol (TAG) in rainbow trout. We hypothesized that swimming would stimulate TAG turnover rate to fuel working muscles and that heparin would reduce flux by releasing lipoprotein lipase (LPL) from endothelial cells. Results from both tracer methods show that the baseline TAG turnover rate of trout ranges from 24 to 49 μmol TAG·kg−1·min−1 and exceeds all values measured to date in endotherms. More important, this high resting turnover rate is not stimulated during swimming, because it can already cover several times the energy requirements of locomotion. The fact that heparin causes a 50% decrease in baseline TAG turnover rate suggests that fish LPL must be bound to the endothelium for normal tissue uptake of fatty acids supplied by lipoproteins, as in mammals. We propose that the high resting TAG turnover rate of rainbow trout could be needed by ectotherms for rapid restructuring of membrane phospholipids. The continuous tracer infusion method implemented here could be a versatile tool to investigate the potential role of lipoproteins in providing fatty acids for rapid homeoviscous adaptation.
- lipoprotein lipase
- prolonged exercise
- oxidative fuel
- fish metabolism
- lipid kinetics
- homeoviscous adaptation
- tracer methodology
- Oncorhynchus mykiss
circulating lipids play a critical role in energy distribution during sustained exercise in vertebrates (26, 30, 41). They are transported as nonesterified fatty acids (NEFA) or esterified fatty acids in triacylglycerol (TAG) and phospholipids (PL): two critical components of lipoproteins. Because the relative abundance of nonesterified and esterified fatty acids differs among vertebrates, different species may show a variety of strategies to shuttle energy from lipid reserves to working muscles. Even though mammals can use significant amounts of lipoproteins (17, 29), they transport most lipids to working muscles as NEFA-albumin complexes (39), and their NEFA flux is strongly stimulated by exercise (6, 46). In contrast, fish do not increase NEFA flux during prolonged swimming (5), possibly because they use lipoproteins to fuel locomotory muscles. Recent evidence supporting this view includes 1) plasma lipoprotein concentration of sockeye salmon changes dramatically over the course of migration (23), 2) rainbow trout have particularly high lipoprotein levels [∼4 times mammalian values (2)] that account for 92% of total plasma lipids, and 3) endurance swimming activates lipoprotein lipase (LPL) in trout red muscle (25).
Paradoxically, the strong activation of trout LPL elicited by prolonged swimming (in red muscle) or by heparin (in plasma) is not accompanied by significant changes in lipoprotein concentration or composition (25). However, this does not mean that the TAG turnover rate of fish is unaffected by such treatments, because flux and concentration often vary independently (14). TAG turnover rate has never been measured in fish, and only rarely in mammals (3, 15, 18, 28, 32). This is probably because lipoproteins are normally not a major fuel for working mammalian muscles (39) [although their role can increase after a high-fat diet (17)], and the in vivo measurement of TAG turnover rate by tracer kinetics is technically challenging (4). Exercise and heparin administration have numerous metabolic consequences in fish (25, 30), but their effects on TAG turnover rate are unknown. In mammals, heparin frees LPL from its active sites on endothelial cells to plasma. In this study, we have developed novel in vivo tracer methods for measuring TAG turnover rate in fish. The first approach uses bolus injection of very-low-density lipoproteins (VLDL) labeled in vivo from donor fish, a natural substrate for fish LPL, but impossible to produce in large quantities. The second approach uses continuous infusion of an exogenously labeled Intralipid emulsion: an easy to prepare, artificial substrate hydrolyzed by trout LPL. Both substrates were labeled with tri-[3H]-oleate and successfully used as tracers in previous mammalian studies of lipoprotein kinetics (20, 21, 31, 35, 40). Our goals were to characterize the effects of endurance swimming and heparin administration on the TAG turnover rate of rainbow trout. We hypothesized that endurance swimming would stimulate TAG turnover rate to provide more fuel to working muscles and that heparin would reduce it by releasing LPL in plasma, thereby preventing normal lipoprotein uptake.
MATERIALS AND METHODS
Adult, sexually immature female rainbow trout, Oncorhynchus mykiss, were obtained from Linwood Acres Trout Farm (Campbellcroft, Ontario, Canada) in September 2006. They were kept in a 1,300-liter flow-through holding tank in dechloraminated, well-oxygenated water at 13°C under a 12:12-h light-dark photoperiod. The same water quality and photoperiod were used during all the measurements. Animals were acclimated to these conditions for at least 55 days before the experiments that were carried out between November 2006 and March 2007. In the holding tank, routine swimming speed was <0.3 body length per second (BL/s). Trout were fed floating fish pellets (Martins Mills, Elmira, ON, Canada) once a day for 5 days/wk (or an average of ∼0.45% of body weight/day). They were fasted for 72 h before measurements to eliminate circulating chylomicra (42). All experimental procedures were approved by the Animal Care Protocol Review Committee and adhered to guidelines established by the Canadian Council on Animal Care for the use of animals in research.
Preparation of tri-[3H]-oleate emulsion for measurement of TAG kinetics.
An Intralipid emulsion (Sigma, St. Louis, MO, USA) labeled with tri-[3H]oleate was used to prepare the tracer for bolus injection and continuous infusion experiments. A stock solution of labeled TAG was made by drying 450 μCi of tri-[9,10(n)-3H]oleate (Amersham, Buckinghamshire, UK) (22.5 μCi/ml) under N2 and resuspending in heptane. It was stored anaerobically at −20°C for up to 2 mo. Before each experiment, an emulsion was freshly prepared by sonicating an aliquot of the tri-[3H]oleate stock solution with Intralipid (20%) and Cortland saline, as described previously (1). The fatty acid composition and 3H content of the different lipid classes in the emulsion are presented in Table 1.
These experiments were designed to measure baseline TAG kinetics in resting rainbow trout, using VLDL labeled in vivo from donor fish as tracer. A batch of donor fish (body mass 462 ± 49 g, n = 9) was implanted with single dorsal aorta catheters and was used to produce the tracer for bolus injections. Surgery was performed under buffered ethyl-N-aminobenzoate sulfonic acid anaesthesia (MS-222; Sigma) following the procedure of Haman and Weber (13), but using 13 mM sodium citrate as anticoagulant (<0.2 ml, pH 7.7). The animals were allowed to recover for 24 h in opaque Plexiglas chambers (25). They were then infused with 8 ml of tri-[3H]-oleate emulsion at 1 ml/h. Twelve hours after the end of infusion, plasma was harvested from the donor animals, and the different lipoprotein fractions (HDL, LDL, and VLDL; see Table 2) were separated, following the procedure of Magnoni and Weber (25). This allowed us to collect enough labeled VLDL to perform four bolus injection experiments. The nature of the labeled VLDL particles was assessed by agarose gel electrophoresis (Paragon System, Beckman Coulter, Fullerton, CA). This procedure showed that VLDL migration was unaffected by label incorporation and that 92% of the activity was in TAG (see Table 2).
Animals used for bolus injections were implanted with single catheters as described above (body mass 410 ± 14 g, n = 4). At time 0, a bolus of radiolabeled VLDL (total activity of ∼500,000 dpm in 4 ml or ∼60,000 dpm/μmol TAG) was injected through the catheter in resting fish, and blood samples were collected at different times after injection. Plasma was immediately separated and stored at −80°C for subsequent determination of radioactivity (Beckman Coulter CS6500 scintillation counter; Palo Alto, CA) and TAG concentration. Concentration was measured by spectrophotometry using a kit (TR0 100 from Sigma) after validation by gas chromatography. All TAG measurements with the kit were corrected for plasma glycerol.
Continuous infusion experiments.
The purpose of these experiments was twofold: 1) to measure baseline TAG kinetics using an alternative tracer method to bolus injection, and 2) to quantify the effects of heparin and endurance swimming on TAG kinetics (bolus injection is not suited to follow changes in flux over time). Animals were implanted with two dorsal aorta catheters (13) and allowed to recover at rest for 24 h, either in opaque Plexiglas chambers (heparin group; body mass 374 ± 35 g, n = 8) or in a swim tunnel (swimming group, body mass 435 ± 54 g, n = 8) (masses were not different between heparin and swimming groups: t-test P = 0.32). A priming dose of tri-[3H]-oleate equivalent to 3 h of infusion was administered before starting infusions to ensure that isotopic steady state was reached in less than 2 h. For all blood samples collected during continuous infusions, hematocrit did not decrease significantly throughout the experiments (25.4 ± 2.0% before, and 21.6 ± 1.9% at the end of the experiments, P = 0.322).
For heparin experiments, the labeled emulsion was infused at 1 ml/h (206 dpm·min−1·g−1) in resting fish for 4 h, using a calibrated syringe pump (Harvard Apparatus, South Natick, MA). A bolus of 600 U heparin/kg (Hepalean, Organon, Toronto, ON, Canada) was injected 2 h after starting infusions. Blood samples (0.3 ml each) were drawn every 30 min throughout the heparin experiments. For swimming experiments, the emulsion was infused at 1 ml/h for 10 h in animals placed in a modified Blazka-type swim tunnel: for 2 h at rest, 6 h swimming at 1.5 BL/s (∼70% Ucrit, or moderate exercise intensity), and 2 h of recovery from exercise. Characteristics of the swim tunnel and environmental conditions during exercise were previously described (25). Blood samples were collected in 1 mg/ml EDTA, centrifuged at 5,000 g for 10 min at 13°C, and separated plasma was stored at −80°C before analyses.
Intralipid emulsion and plasma lipoprotein analyses.
Fatty acid composition was determined: 1) for the different lipid classes of the Intralipid emulsion (TAG, NEFA, and PL), and 2) for each lipid class of the different plasma lipoproteins (HDL, LDL, and VLDL). Because of blood volume restrictions, plasma fatty acids were analyzed in a separate group of fish (458 ± 25 g, n = 12) than used for flux measurements by bolus injection and continuous infusion. The different plasma lipoprotein fractions were first separated by ultracentrifugation (25). Lipids of all the samples (emulsion and plasma lipoproteins) were extracted with 2:1 chloroform:methanol (vol/vol) (11) and centrifuged (2,000 g for 10 min). Pellets were discarded, and supernatants were filtered before adding 0.25% KCl. After shaking, the mixture was centrifuged (2,000 g for 10 min) before discarding the aqueous phase and drying the organic phase under N2. Lipids extracted from the Intralipid emulsion and from the plasma lipoprotein fractions were loaded on solid-phase extraction columns (Supelclean; Sigma) to separate TAG, NEFA, and PL by sequential elution. Fatty acid composition was measured by gas chromatography using heptadecanoic acid (17:0) as an internal standard as detailed previously (25). Individual fatty acids were identified by determining exact retention times with authentic standards (Sigma-Aldrich).
Calculations and statistical analyses.
For bolus injections, the turnover rate of TAG was calculated as the dose injected divided by the surface area under the specific activity decay curve (see Fig. 1). To calculate this surface area, the decay curve was fitted with a multiexponential function using nonlinear regression (SigmaPlot 9.0). The fitted function was integrated between time 0 and the time when 5 or 10% of maximum activity was reached: a procedure commonly used because it corrects for label recycling (44, 45). Maximum activity was calculated as the dose injected divided by the volume of the rapidly mixing pool estimated at 5% of total body volume. For continuous infusions, the turnover rate of TAG was calculated as TAG infusion rate divided by TAG-specific activity (37).
Statistical differences were assessed using one-way ANOVA or one-way ANOVA with repeated measures (RM ANOVA). When significant changes were detected by ANOVA, the Holm-Sidak method was used to determine which means were different from baseline. Values are expressed as means ± SE. A level of significance of P < 0.05 was used in all the tests.
Fatty acid composition of Intralipid emulsion and plasma.
Characteristics of the 3H-labeled Intralipid emulsion used for measuring TAG kinetics by continuous infusion are presented in Table 1. It shows the fatty acid composition and relative distribution of total radioactivity in different lipid classes. The vast majority of fatty acids (95%) and radioactivity (90%) were found in TAG. Together, PL and NEFA only accounted for 5% of all the fatty acids and 10% of all the activity in the emulsion. The lipid composition of trout plasma is presented in Table 3. The great majority of fatty acids in rainbow trout plasma were found in PL and TAG (58% and 37%, respectively), particularly as PUFA (66% for PL and 49% for TAG). The fatty acid concentration for the three lipid classes and the ratio between phospholipids and TAG within each lipoprotein fraction are presented in Table 4. Results show that 50% of the fatty acids in trout VLDL are found in TAG, but that this percentage is smaller for other lipoproteins. The PL/TAG ratio is inversely proportional to lipoprotein size, and, therefore, VLDL has the lowest ratio among all lipoprotein fractions (0.9).
TAG kinetics measured by bolus injection of tri-[3H]-oleate VLDL.
Intralipid tri-[3H]oleate emulsion was infused in donor fish to produce labeled lipoproteins for the measurement of TAG kinetics by bolus injection in other fish. Label incorporation in lipoproteins is presented in Table 2. Twelve hours after infusion, most of the tri-[3H]oleate activity was incorporated in VLDL (77%), whereas the remaining activity was equally shared between HDL and LDL. Moreover, 92% of the activity incorporated in VLDL was found in the TAG of this lipoprotein fraction, with minor amounts in PL and NEFA. Therefore, labeled VLDL from the donor fish were used for bolus injection experiments.
Figure 1 shows the decay curve for mean TAG specific activity (TAG SA) after a bolus injection of labeled VLDL. It was fitted with the sum of three exponential functions (R2 = 0.999) to calculate the rate of appearance of TAG (Ra TAG) by integrating the surface area under the curve. Specific activity decreased sharply for 4 h after injection and then more gradually. Plasma TAG concentration remained stable throughout the experiments (see Fig. 1, inset, where time had no effect, P > 0.05). The Ra TAG calculated from the decay curve was 24.0 ± 7.5 μmol·kg−1·min−1 (when the curve was integrated until specific activity reached 10% of the maximum) and 49.1 ± 17.8 μmol·kg−1·min−1 (when the curve was integrated until specific activity reached 5% of the maximum).
Effect of heparin on TAG turnover.
Figure 2 shows the effect of heparin administration on TAG metabolism measured by continuous infusion of a tri-[3H]-oleate emulsion in resting trout. Plasma TAG concentration decreased from a baseline value of 1.3 ± 0.1 to 0.9 ± 0.1 μmol/ml ∼1–2 h after heparin injection (P < 0.001) (Fig. 2A). Heparin caused a significant increase in TAG SA from 5,777 to 11,832 dpm/μmol after 2 h (P < 0.001) (Fig. 2B). The higher variability in TAG-specific activity compared with TAG concentration and flux reflects differences in body size because the same infusion rate (40,207 dpm/min) was used for all fish. The rate of appearance of TAG was significantly decreased from a baseline value of 24.4 ± 5.5 μmol·kg−1·min−1 1 h after heparin administration (P < 0.001) (Fig. 2C).
Effect of swimming on TAG turnover.
Figure 3 shows the effect of swimming on TAG kinetics measured by continuous infusion of a tri-[3H]-oleate emulsion in trout. Six hours of sustained swimming at 1.5 BL/s did not alter plasma TAG concentration from resting values (Fig. 2A; P > 0.05) or TAG SA (Fig. 2B; P > 0.05). Therefore, the high rate of appearance of plasma TAG measured at rest was not significantly affected by prolonged exercise, even after 6 h of endurance swimming (Fig. 2C; P > 0.05).
Bolus injection vs. continuous infusion.
To increase confidence in the estimates of TAG kinetics, two separate tracer methods were used: bolus injection and continuous infusion. Both methods yielded similar values ranging from 24 to 49 μmol·kg−1·min−1. For the group of fish measured by bolus injection of tri-[3H]-oleate VLDL, values of 24.0 ± 7.5 and 49.1 ± 17.8 μmol·kg−1·min−1 were obtained (depending on the recycling correction applied). Both estimates obtained by bolus injection fell within the range of values measured by continuous infusion. The two groups of resting animals measured by continuous infusion of tri-[3H]-oleate emulsion (preheparin and preexercise) had TAG turnover rates of 25.0 ± 3.3 and 42.6 ± 2.6 μmol·kg−1·min−1, respectively.
This study provides the first measurements of TAG turnover rate in fish and shows that baseline lipoprotein metabolism is particularly active in this group of vertebrates. Both tracer methods reveal that rainbow trout support very high TAG turnover rates, even at rest. Such high baseline turnover rates can cover all the fuel requirements of locomotion (see calculations below), and they are not stimulated by endurance swimming (Fig. 3). Results also show that heparin-induced release of LPL in the circulation causes a 50% inhibition of TAG turnover rate (Fig. 2). The continuous infusion method (using an easily prepared, artificial substrate) and bolus injection method (using a difficult-to-produce, natural substrate) provide similar estimates of flux. The continuous Intralipid infusion method implemented here is a new tool for in vivo studies of fish lipoproteins that allows flux measurements under non-steady-state conditions.
Endurance swimming does not stimulate high resting TAG turnover rate.
The baseline TAG turnover rates of rainbow trout range from 25 to 42 μmol·TAG kg−1·min−1. This study characterizes the response of female trout, but the changes in lipid metabolism reported here probably also reflect those of males because sexually immature animals were measured. However, further studies will be needed to confirm this. Unexpectedly, the resting values of this ectothermic animal (after 72 h of fasting) exceed all fluxes measured to date in endotherms (fed, fasting, or exercising). Previous studies report postprandial TAG turnover rates of 18 μmol·kg−1·min−1 in resting or exercising dogs (38), and lower fluxes of 5 and 1.5 μmol TAG·kg−1·min−1 in rats fasted for 12 or 42 h, respectively (3, 40). Taken together, these mammalian results suggest that the effects of mass-specific metabolic rate on TAG turnover rate are dwarfed by those of fasting (body mass ratio of dog/rat = 60).
Label recycling is difficult to estimate experimentally, and we did not attempt to quantify it in this study. However, it is important to note that any recycling would lead to underestimating true TAG turnover rate (because it would erroneously increase surface area under the specific activity decay curve for bolus injection experiments and increase plateau-specific activity in continuous infusion experiments). For bolus injection, we have applied a commonly used recycling correction by interrupting curve integration when specific activity reaches 5% of its maximal value (44, 45). We have repeated calculations using a value of 10% that would assume unusually high recycling rates.
To determine whether circulating lipoproteins could fuel the locomotory muscles of swimming trout, we have calculated the theoretical metabolic rate needed to oxidize all fatty acids supplied by the TAG turnover rate measured in this study (24 μmol TAG·kg−1·min−1 or 72 μmol FA·kg−1·min−1 = lowest value measured in this study). Assuming that energy metabolism is only supported by lipid oxidation and that TAG is entirely made of trioleate, this theoretical metabolic rate would be 1,872 μmol O2·kg−1·min−1 (if each oleate requires 26 O2 for oxidation and if the contribution of glycerol is ignored). Because the real metabolic rate of a swimming trout is ∼109 μmol O2·kg−1·min−1 (7), we can determine that only 6% of TAG turnover rate is necessary to support exercise. Therefore, it is clear that resting TAG turnover rate is high enough to provide several times the energy needed by working muscles, and this explains why TAG turnover is not stimulated during swimming (Fig. 3). Why then does lipoprotein metabolism remain so active in a resting animal? We propose that high TAG turnover rates (this sudy) and high lipolytic rates (5) are fundamental features of ectotherm metabolism that allow the restructuring of membrane phospholipids to be synchronized with frequent changes in body temperature. To preserve normal membrane fluidity (16), adequate homeoviscous adaptation may depend on the rapid supply of lipoprotein-bound fatty acids with different chain length and degree of saturation.
TAG turnover rate is reduced by heparin.
What are the mechanisms responsible for the decrease in TAG turnover rate observed after heparin administration? Heparin is known to release LPL into plasma from its natural location, bound to endothelial proteoglycans, because of heparin's very high affinity for this enzyme (33). This effect of heparin could therefore explain the decrease in TAG turnover rate (Fig. 2), because tissue uptake of fatty acids from lipoproteins mainly relies on bound LPL (12). In mammals, this effect of heparin on LPL has been previously invoked to explain declines in fatty acid supply from lipoproteins (9), and in the uptake of a TAG-rich emulsion by the heart (1). Therefore, both fish and mammals seem to depend on the natural presence of LPL bound to the vascular endothelium for proper delivery of lipoprotein fatty acids to tissues. Even in mammals, the exact mechanism of action for LPL is still controversial, and several models have been proposed (27). In addition, LPL-independent mechanisms of TAG uptake by tissues cannot be excluded.
Measuring TAG turnover rate in fish: bolus injection vs. continuous infusion.
Investigating the metabolism of circulating lipids has been hindered by the lack of suitable methods to measure lipoprotein turnover rate. The limited information presently available has only been obtained for mammals, and almost exclusively by bolus injection, a method requiring a single catheter and small amounts of labeled substrate (21, 22, 34, 43). Unfortunately, these advantages come with significant limitations. Bolus injection must be used under steady-state conditions and only provides a single value of flux per experiment. In addition, flux calculations from bolus injection experiments depend on the surface area under the specific activity decay curve, a value difficult to estimate accurately because of label recycling. The more versatile method of continuous infusion can provide multiple flux measurements under non-steady-state conditions but requires the surgical placement of two catheters and larger amounts of labeled substrate. Such technical difficulties explain why we were only able to find two studies of lipoprotein kinetics that rely on continuous infusion in mammals (3, 40) and none in fish. For this first investigation of TAG turnover rate in fish, we have used both methods in an attempt to exploit their respective advantages. Developing a new continuous infusion technique for fish lipoproteins was greatly simplified by adapting a known double catheterization procedure specifically designed for trout (13).
Results show that bolus injection of endogenously labeled trout VLDL (Table 2) and continuous infusion of a labeled Intralipid emulsion provide similar high estimates of TAG turnover rates in trout, using two different substrates. With a 400-g fish having a blood volume of 20 ml and a hematocrit of 25%, we can calculate that the TAG pool in VLDL is 10 μmol (for a VLDL-TAG concentration of 2 μmol/ml plasma). With an average TAG turnover rate of 36 μmol·kg−1·min−1, this trout has a clearance rate of 18 ml plasma·kg−1·min−1.
A comparison of Tables 2 and 3 reveals similarities and differences in the fatty acid composition of TAG between the Intralipid emulsion and trout plasma. Both had the same percentage of saturated fatty acids (15%), but their largest difference was probably the presence of more linoleic acid in the emulsion (55%) than in plasma (7%). Another important similarity relevant to our experiments was that emulsion and plasma showed no significant difference in TAG concentration (9 vs. 8 μmol FA/ml). However, the emulsion had a much higher percentage of total FA in TAG than trout plasma (95% vs. 37%). Nevertheless, Intralipid particles appear to mimic the metabolic behavior of trout VLDL (Table 4), even though the PL/TAG ratios of these two substrates are different. This may be possible because artificial Intralipid emulsions are known to acquire natural apolipoproteins from circulating HDL and VLDL, thereby becoming suitable substrates for LPL in vivo (1, 8, 10, 19, 36). Finally, the significant decrease in TAG turnover rate induced by heparin (Fig. 2) shows that the continuous infusion method proposed here is sensitive enough to monitor biologically relevant changes in the TAG turnover rate of fish.
Perspectives and Significance
This first study of lipoprotein kinetics in an ectotherm reveals that baseline TAG turnover rate is higher in rainbow trout than in any endotherm measured to date. It shows that resting TAG turnover rate is not stimulated by endurance swimming because it is already high enough to cover all the fuel requirements of exercise. Results suggest that rainbow trout need to maintain a high TAG turnover rate at all times to cope with fluctuations in environmental temperature by rapid restructuring of membrane phospholipids. In rainbow trout, the inhibition of TAG turnover by heparin suggests that LPL must be bound to the endothelium for normal tissue uptake of fatty acids from lipoproteins. The continuous infusion method implemented here is a new versatile tool to investigate the potential role of lipoproteins in homeoviscous adaptation.
This work was supported by a Natural Sciences and Engineering Research Council Discovery Grant to J. M. Weber.
We thank Bill Fletcher for his invaluable help in animal care.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society