The effects of intracranial transforming growth factor (TGF)-β3 on spontaneous motor activity and energy metabolism were examined in rats. After injection of TGF-β3 into the cisterna magna of the rat, spontaneous motor activity decreased significantly for 1 h. The intracranial injection of TGF-β3 produced an immediate decrease in respiratory exchange ratio (RER). No significant changes were observed in energy expenditure. TGF-β3 induced a significant increase in total fat oxidation and a decrease in total carbohydrate oxidation. Furthermore, the serum substrates associated with fat metabolism were significantly altered in rats injected with TGF-β3. Both lipoprotein lipase activity in skeletal muscle and the concentration of serum ketone bodies increased, suggesting that the increase in fat oxidation caused by TGF-β3 may have occurred in the liver and muscle. Intracranial injection of TGF-β3 appeared to evoke a switch in the energy substrates accessed in energy expenditure. These results suggest that the release of TGF-β3 in the brain by exercise is a signal for regulating energy consumption.
- spontaneous motor activity
- respiratory exchange ratio
- energy metabolism
we have previously reported that intracranial administration of cerebrospinal fluid (CSF) from exercise-exhausted rats produced a decrease in spontaneous motor activity, whereas CSF from sedentary rats had no such effect (19). The substance that regulates the urge for motion in response to exercise exhaustion was identified as transforming growth factor-β (TGF-β) (23, 24). Growing evidence indicates that accumulation of the active forms of TGF-β2 and/or TGF-β3 in the brain is involved in the fatigue induced by exercise (19, 20,23).
The sensation of fatigue in the brain, however, may not merely be an inconvenience but may constitute a physiological defense mechanism against total exhaustion. If this is so, the active forms of TGF-β found in the brain may function positively to prevent peripheral exhaustion and to enhance recovery.
The brain may detect changes in the normal levels of the constituents of the blood, such as the concentration ratio of tryptophan to branched-chain amino acids (4, 6, 7, 11, 12, 26-29), which can then act as a specific signal to increase the organism's sensitivity to fatigue. However, little is known about the counteractive effects of the central nervous system (CNS) on peripheral energy metabolism.
After exercise exhaustion, the respiratory exchange ratio is usually lower for a more extended period than it is before exercise, although oxygen consumption returns readily to its preexercise level (5,32). This suggests that the sensation of fatigue affects not only spontaneous motor activity but also energy metabolism, i.e., it enhances fat oxidation to conserve glucose. It seems reasonable to suppose that the substances released in the brain that accompany fatigue may regulate energy metabolism and induce the restoration of energy resources.
We have reported that the intracranial administration of TGF-β3 suppresses spontaneous motor activity in mice without substantial exercise loading (20) and may affect peripheral energy metabolism. The TGF-βs represent a multifunctional family of cytokines with three closely related isoforms: TGF-β1, TGF-β2, and TGF-β3. These isoforms are expressed in several cell types of the CNS, including neurons, astrocytes, and microglia (9). Unsicker et al. (36) reported that TGF-β2 and TGF-β3 mRNAs are present in all brain areas, including the cerebral cortex, hippocampus, striatum, cerebellum, and brain stem. In this study, we used TGF-β3 to represent the TGF-β isoforms for the following reasons. We have found that TGF-β2 and TGF-β3 suppress spontaneous motor activity equally, and a considerably higher dose of TGF-β1 is required to exert a suppressive effect equal to that required for either TGF-β2 or TGF-β3 (20). TGF-β2 and TGF-β3 are ubiquitously abundant in the rat brain (36), and we have confirmed that TGF-β1 and TGF-β2 levels do not change in the brain, even when total TGF-β levels increase (unpublished data). In the present study, we demonstrate that injection of TGF-β3 into the brain alters peripheral energy metabolism to resemble that induced during exercise exhaustion.
MATERIALS AND METHODS
Animals and diets.
Male Sprague-Dawley rats (8 wk old, Japan Charles River, Yokohama, Japan) were used in the present study. All animals were maintained on an inverse 18:6-h light-dark cycle (light on for 18 h, and light off for 6 h) for 1 wk to make them active during the experimental time. They were individually housed in standard cages (25 × 38 × 17.5 cm; one rat per cage) under controlled conditions. Room temperature and humidity were regulated at 22 ± 0.5°C and 50%, respectively. During the study period, rats were given free access to water and a high-fat (30%) diet, containing 21 g/kg protein, 30 g/kg fat, and 49 g/kg carbohydrate. All animals were treated humanely as outlined in the National Research Council's Guide for the Care and Use of Laboratory Animals (Kyoto University Animal Care Committee according to NIH #86-23, revised 1985).
Intracranial injection (brain implantation of a guide cannula).
TGF-β3 was purchased from R&D Systems (Minneapolis, MN). TGF-β3 (40 ng), dissolved in 40 μl of saline containing 0.5 mM HCl and 0.1% bovine serum albumin (BSA), was injected into the brains of rats. An equal volume of vehicle was used as a control. Soon after rats were purchased, they were anesthetized with 1 mg/kg pentobarbital sodium (Wako Pure Chemical Industries, Osaka, Japan), fixed onto a stereotaxic apparatus, and implanted with a permanent 23-gauge guide cannula for sample injection into the cisterna magna. Each cannula was inserted 2.5 mm posterior to the lambda and 8.5 mm deep, and it inclined anteriorly at an angle of 45° to the skull surface. The cannula was secured to the skull with dental cement and then plugged with a cap. After implantation of the guide cannula, the rats were allowed to recover for 3 days before measurements were made of spontaneous motor activity, oxygen consumption (V˙o 2), and CO2 production (V˙co 2).
Determination of spontaneous motor activity.
The spontaneous motor activity of each rat was examined with a Supermex (Muromachi Kikai, Tokyo, Japan) for 1 h after intracranial injection of TGF-β3. This apparatus surveys the measurement area with multiple lenses that detect the infrared radiation emitted by animals. Motor activity was assessed as a single count when the animal moved from one region of the measurement area, which was optically divided by the multiple lenses, to a neighboring region. The rats were sedentary and not subjected to any exercise before measurements were made. The cages used for measurement were completely novel environments for the rats.
Assessment of metabolic rate.
The respiratory exchange ratio (RER) was measured by indirect calorimetry. Rats were fasted overnight on the day before the experiments, and food was provided for 1.5 h immediately before the experiments. Rats were placed in the chamber individually before the experiment for 1 h to maintain RER at a constant value. After the injection of TGF-β3, RER was measured for ∼1 h. To determine whether the effects of TGF-β3 on the metabolic rate were specific, 1 μg of thyrotropin-releasing hormone (TRH) (Research Biochemicals Int, Natick, MA) dissolved in 40 μl of saline was also injected into the brains of rats as a positive control.
A specific laboratory-made open-circuit calorimeter was used. The gas analyzer consisted of six acrylic chambers, CO2 and O2 analyzers (model RL-600, AlcoSystem), and a switching system (model ANI6-A-S, AlcoSystem) to sample gas from each metabolic chamber. Each chamber had a 204-cm2 floor and a height of 12.7 cm. Air flow (3 l/min) was circulated and monitored by a mass-sensitive flowmeter. Air from each chamber was sampled for 60 s. During the last second, the concentrations of O2 and CO2 were measured 100 times, and the mean values were used to calculate V˙o 2 and RER. The data were then processed mathematically using a specific algorithm.V˙o 2 andV˙co 2 were calculated from changes in gas content (percent) and air flow (l/min) by use of differential gas analyzers. The RER was calculated as the ratio ofV˙co 2 toV˙o 2. Carbohydrate and fat consumption rates were computed from V˙o 2and the respiratory quotient according to the theory of Frayn (13).
Analysis of serum samples.
Blood was collected from severed neck veins, and serum was isolated by centrifugation and stored at −80°C until analysis. Serum glucose was measured using the glucose oxidase method combined with mutarotase by use of glucose AR-II and a commercial kit (Wako Pure Chemical Industries). Serum free fatty acids (FFA) were measured by an acyl-CoA synthetase and acyl-CoA oxidase method using NEFA C (Wako Pure Chemical Industries). Serum triglycerides were assayed by the glycerol-3-phosphate oxidase method with the triglyceride G test (Wako Pure Chemical Industries). Serum lactic acid was measured by the lactate oxidase method using Determiner LA (Kyowa Medics, Tokyo, Japan). Serum ketone bodies were measured using a ketone test (Sanwa Chemical Institute, Nagoya, Japan). For the catecholamine assay, serum samples were purified with aluminum oxide by the method of Anton and Sayre (1). Serum samples (100 μl) containing 10% Na2S2O5 (50 μl/ml) and 3,4-dihydroxybenzylamide (40 ng/ml) as the internal standards were added to 100 μl of 2 M Tris · HCl buffer, pH 8.6, and aluminum oxide (100 mg/ml). The mixture was shaken in a microtube mixer for 10 min, the supernatant was removed, and the aluminum oxide was washed twice with methanol and distilled water. Epinephrine and norepinephrine were eluted with 60 μl of 0.5 N HCl. The eluate was assayed using an HPLC-electrochemical detector (37). Serum insulin and leptin were measured using a Mercodia Rat Insulin Kit (Mercodia, Uppsala, Sweden) and a Morinaga Rat Leptin Kit (Morinaga, Yokohama, Japan), respectively.
Measurements of muscle lipoprotein lipase activity and serum glycerol concentrations.
Gastrocnemius and soleus muscle samples were dissected away from visible fat. Samples (3–10 mg) of skeletal muscle (gastrocnemius) were ground in liquid nitrogen and incubated (in duplicate) in 200 μl of Krebs-Ringer solution, 0.1 M Tris · HCl buffer (pH 8.4) containing 1 g/100 ml of BSA and 2.5 U (50 mg/l) of heparin (35), with gentle shaking at 28°C. After 40 min, the tissue was removed from the medium by centrifugation for 5 min, and the supernatant was used for the measurement of lipoprotein lipase (LPL) activity with an LPL activity kit including a nonfluorescent substrate emulsion that becomes intensely fluorescent upon interaction with LPL (Roar Biomedical, New York, NY). The total protein content of each sample was measured by the Bradford method with Coomassie brilliant blue solution. LPL activity was assessed relative to the protein content of the tissue, and the values are expressed as the ratio of each group to the nontreatment group. Serum glycerol was determined by the ultraviolet method by use of F-kit glycerol (J. K. International, Tokyo, Japan).
All data are expressed as means ± SE. Statistical analysis of differences between preinjection and postinjection measurements in the same group was performed with repeated-measures ANOVA test. Dunnett's test was used for post hoc analysis. The means of more than three groups measured at the same point in time were compared by one-way ANOVA followed by Dunnett's test.
Effects of TGF-β3 on spontaneous motor activity in rats.
The spontaneous motor activity of the rats was determined with Supermex for 1 h after the injection of TGF-β3 by detecting the movement of infrared radiation emanating from the animal every 5 min. The spontaneous motor activity was gradually suppressed 20 min after injection in rats treated with TGF-β3 (Fig.1 A). Figure 1 Bshows the total spontaneous motor activity for 1 h. Total counts were calculated by adding up all counts over 5-min periods. The administration of TGF-β3 significantly suppressed the activity of the rat.
Effects of TGF-β3 on metabolic rate.
Figure 2 shows the changes in metabolic rate in rats injected intracranially with vehicle, TGF-β3, or TRH. The high-fat diet maintained RER at ∼0.85 before sample injection. An injection of 40 ng of TGF-β3 significantly lowered RER compared with that in the same rat before injection. TGF-β3 significantly reduced the RER of the rat by 7 min after injection, and its effect was maintained for ≥1 h after injection. The same volume of vehicle injected as a control reduced the RER slightly. During the experiment, rats were deprived of access to food for 2 h and, therefore, RER was naturally reduced.
Intracranial injection of TRH elevated thermogenesis andV˙o 2 in rats (14, 21). Food restriction before the experiment decreased RER, whereas TRH injection may increase RER. The decrease in RER was least in the TRH group.
V˙o 2 rose abruptly after the injection of each sample from time 0 to 7 min. Intracranial injection of TRH significantly increased V˙o 2, and the increase was maintained for ∼40 min after injection. It has been reported previously that injection of TRH into the brain increasesV˙o 2 (14). Therefore, this result demonstrates that the injected samples reached the intracerebral location responsible for the regulation ofV˙o 2. V˙o 2was similar in the TGF-β3 and vehicle groups (Fig.3). These results suggest that intracranial TGF-β3 does not affect energy consumption. Carbohydrate oxidation decreased gradually after injection with both TGF-β3 and vehicle (Fig. 4 A). We inferred that this reduction in carbohydrate oxidation was caused by food restriction before the experiment. Injection of TRH tended to increase carbohydrate oxidation for ∼40 min. Intracranial TGF-β3 injection significantly increased fat oxidation for ≥1 h compared with preinjection values (Fig. 4 B). Injection with either TRH or vehicle tended to increase fat oxidation, but not significantly. Figure5 shows the rates of carbohydrate and fat oxidation from time 0 to 28 min and from 28 to 56 min, compared with each oxidation value for 28 min before injection. In the TGF-β3 group, the rate of increase in fat oxidation was higher than in the vehicle group (Fig. 5 B). However, the rate of change in carbohydrate oxidation was similar in the TGF-β3 and vehicle groups.
Changes in serum energy substrates and concentrations of hormones after injection with TGF-β3.
Serum concentrations of glucose and lactic acid, which are affected by carbohydrate oxidation, had not changed 14 and 28 min after injection in either the TGF-β3 or vehicle group (Fig.6, A and B). The concentrations of the serum parameters associated with fat oxidation (FFA, triglycerides, and ketone bodies) were measured (Fig. 6,C-E). The concentrations of serum FFA in the vehicle group had decreased significantly at 28 min compared with those at 14 min, but no change was observed in the TGF-β3 group. Therefore, the serum FFA concentrations in the TGF-β3 group were apparently higher than those in the vehicle group. Triglyceride concentrations in the TGF-β3 group were lower than those in the vehicle group at 14 min. The concentration of ketone bodies increased significantly in the TGF-β3 group. These results suggest that intracranial injection of TGF-β3 significantly facilitated fat oxidation and tended to restrict carbohydrate oxidation, which corresponds to the metabolic condition after exercise.
No significant changes in blood serum concentrations of catecholamines, insulin, or leptin were observed (Fig.7, A-D). Epinephrine concentration at 14 min had increased in the TGF-β3-treated rats compared with the vehicle group (Fig.7 A). Insulin concentration in the vehicle group increased at 14 min compared with the value before injection, but insulin levels in the TGF-β3 group did not change (Fig. 7 C). No significant change was observed in concentrations between the TGF-β3 and vehicle groups (Fig. 7 B). The concentration of leptin was similar in both groups (Fig. 7 D).
Changes in muscle LPL activity and serum glycerol concentration after injection of TGF-β3.
To determine whether TGF-β3 caused an increase in fat oxidation in skeletal muscle, LPL activity was estimated. LPL activity in skeletal muscle, especially the gastrocnemius, is shown in Fig.8 A. In the gastrocnemius, LPL activity rose significantly 28 min after the injection of TGF-β3. In the vehicle group, there was no significant increase in LPL activity. Similar effects of TGF-β3 were also observed in the soleus muscle, although there was no significant difference compared with the vehicle group or before injection (data not shown).
Serum glycerol concentrations tended to increase after injection with TGF-β3, although the difference was not significant (Fig.8 B).
In previous studies, we have demonstrated that physical exercise causes an increase in TGF-β3 levels in the mouse brain and that intracranial TGF-β3 injection induces the suppression of spontaneous motor activity (19, 20). These studies indicated that TGF-β3 may be associated with the induction of central fatigue during exercise.
In the present study, we investigated the effects of intracranial injection of TGF-β3 on the peripheral metabolism in rats. Intracranial injection of TGF-β3 suppressed spontaneous motor activity in rats and decreased the RER (Figs. 1 and 2), similar to the corresponding experimentally induced effects in mice. The injection of TGF-β3 did not cause any toxicity or abnormal behavior in the rats. One day after injection of TGF-β3, both the spontaneous motor activity and the RER of every rat were restored to normal (data not shown).
Rats were fed a high-fat diet, which influences the RER. There was the dispersion in RER value of each rat fed the commercial diet; however, the RER value stabilized between 0.8 and 0.9 by the high-fat diet. We used the high-fat diet because RER was more stable, and it facilitated the comparison between pre- and posttreatment. Also, we used it in previous experiments and wanted to be able to compare these data to those generated previously in our laboratory. We have been assured that there were no differences in the effect of TGF-β3 on spontaneous motor activity and RER of rats between use of a high-fat diet and a commercial diet.
In general, the blood-brain barrier is highly permeable to water, carbon dioxide, oxygen, and most lipid-soluble substances and is almost impermeable to plasma proteins and most non-lipid-soluble large organic molecules. Accordingly, we speculated that the permeability of TGF-β into the blood-brain barrier might be very low or almost none. Therefore, we can derive from this premise that TGF-β3 released from the brain directly affects spontaneous motor activity and the energy status of the animal. This implies that the effects of TGF-β3 on peripheral tissues with energy are mediated through the CNS.
Overall, V˙o 2 was not altered by the injection of TGF-β3 (Fig. 3). In this experiment, TRH was injected into the rat brain as a positive control on the basis of the report by Griffiths et al. (14) that acute or chronic injection of a TRH analog (RX-77368) and TRH itself stimulatedV˙o 2 in rats (14). They demonstrated that intracranial TRH markedly increased the metabolic rate in rats without any apparent effect on physical activity. TRH is suggested to have a physiological role in the control of these phenomena via centralized activity. Hence, TRH injection is an appropriate procedure with which to verify cannula placement. The injection of either sample may cause a spike inV˙o 2 in both groups from immediately after injection to 14 min after injection. Therefore, it would appear that the injection of either sample may cause some stress to rats. However, any handling effects were apparent for only the first 14 min, as indicated by the fact that the V˙o 2 of vehicle animals returned to basal levels at this time. After this time, the differences observed should be strictly treatment related.
Short and Sedlock (32) reported that postexercise RER decreased for ≥1 h in humans. They reported that RER values were at or below baseline throughout much of the recovery period but thatV˙o 2 immediately reverted to the baseline levels calculated before exercise. It has also been shown in rats that RER values decrease 10 min after exercise and thatV˙o 2 is restored to basal levels (33). Intriguingly, the peripheral energy status induced by intracranial injection of TGF-β3 appears to correspond to that induced by exercise. Our findings imply that the intracranial injection of TGF-βs may replicate energy metabolism after exercise.
Fat oxidation was facilitated by intracranial TGF-β3 injection (Fig. 4 B). Rats injected with TGF-β3 exhibited increased serum FFA and ketone body concentrations and decreased serum triglyceride concentrations compared with the vehicle group, which suggests that their lipid metabolism was elevated (Fig. 6,C-E). FFA concentrations in the TGF-β3 group were higher than in the control group. Usually, fatty acids are seldom supplied from the triacylglycerol originally presented in blood plasma. Most FFAs oxidized during exercise are supplied from the triacylglycerol stored in adipose tissue and muscle (8, 16). Because the only product of hydrolysis that appears in the blood is glycerol, we examined whether TGF-β3 actually induced hydrolysis by measuring serum glycerol. As shown in Fig. 8 B, serum glycerol concentrations tended to increase after intracranial injection of TGF-β3. From these results, it is reasonable to suppose that TGF-β3 enhances FFA delivery to the muscles. Serum triglyceride concentrations had decreased by 15 min (Fig. 6 D). It is inferred that an increase in LPL activity in skeletal muscle caused the decrease in serum triglycerides, which seems to be a consistent and reasonable proposition.
There were no differences in the metabolic parameters associated with carbohydrate oxidation between the rats injected with vehicle and those injected with TGF-β3 (Fig. 6, A and B). However, the metabolic parameters associated with fat oxidation changed after injection of TGF-β3. Furthermore, intracranial injection of TGF-β3 significantly increased LPL activity in skeletal muscle at 28 min after injection (Fig. 8 A). Serum ketone body concentrations increased, indicating that lipid oxidation was enhanced in the liver. This is because ketone bodies produced in the liver are more easily taken up by skeletal muscle as an energy resource. We anticipate that TGF-β3 causes an increase in fat oxidation in both the liver and the muscle. It has been reported that exercise induces LPL activity in skeletal muscle (22, 31), which would be necessary to burn the fat efficiently. It is interesting that intracranial injection of TGF-β3 caused LPL activity in skeletal muscle in the same way as exercise.
Leptin levels in rats injected with TGF-β3 did not change. It has been reported that leptin production is not changed by short-term exercise (12, 30). We can suggest that our observation in this study is similar, because no changes in leptin were seen during short-term exercise.
Hwa and colleagues (17, 18) reported that a single intracerebroventricular injection of leptin increased energy expenditure while reducing the respiratory quotient in a dose-dependent manner. They demonstrated that leptin regulates the energy balance via multiple mechanisms, which was certainly mediated by the regulation of both food intake and energy metabolism (17,18). Although both leptin and TGF-β augment fat oxidation, our results suggest that intracranial injection of TGF-β3 does not directly affect the peripheral leptin levels. How does intracranial TGF-β increase fat oxidation? Minokoshi et al. (25) showed that the intrahypothalamic injection of leptin increases the fatty acid oxidation by activating the 5′-AMP-activated protein kinase (25). Careful consideration of our results regarding LPL activity leads us to infer that intracranial injection of TGF-β3 induced an increase in fat oxidation via the sympathetic nervous system, just as leptin works through the hypothalamic-sympathetic nervous system. However, at the present time, we cannot know whether both TGF-β and leptin modulate fat oxidation via the same neural pathway. Further studies are required to clarify whether there is an interaction between leptin and TGF-βs in the CNS.
As shown in Fig. 7 A, epinephrine concentration tended to increase at 14 min after treatment with TGF-β3. Serum epinephrine concentrations also increase after intense exercise. Tadjore et al. (34) reported that plasma epinephrine concentrations in rats increased after prolonged swimming. Cooper et al. (10) demonstrated that blood levels of epinephrine increase gradually after high- and low-intensity exercise in humans. In our study, however, the turnover rate of serum catecholamines was rapid; we suspected that catecholamines had already been recovered. Serum insulin levels increased in the vehicle group, whereas they did not change in the TGF-β3 group (Fig. 7 C). Injection of TGF-β3 may suppress an increase in insulin to restrict fat storage. However, the reasons for the increase in insulin levels in the vehicle group are unclear.
The muscle LPL activity was increased by TGF-β; however, at this time there is no significant change in norepinephrine levels to explain this increase (Fig. 8 A). Furthermore, despite the fact that LPL activity no longer increased after 56 min, fat oxidation kept increasing. This might suggest that the activation of LPL is not critical in this experiment after 56 min. Although we have not studied the fat oxidation in liver, we anticipated that the lipid oxidation also increased in liver because of the increase of ketone bodies, and that it might contribute to continuing fat oxidation after 56 min.
The metabolic changes induced by the injection of TGF-β3 are very similar to the state of energy metabolism after physical exercise. Because energy expenditure did not change, intracranial injection of TGF-β3 may cause a switch of energy substrates. During prolonged exercise, utilization of energy substrates shows a gradual transition from carbohydrate to fat. Other studies have reported that there is a significant substrate shift toward fat oxidation after high-intensity exercise (3, 38). This is a common phenomenon in exercise physiology, but the complete mechanism of the switch in energy substrates has not been clarified. TGF-β3 that is released in the brain during exercise may increase the rate of fat oxidation to conserve glucose. We also reported that the changes shown by electroencephalogram after intracranial injection of TGF-β were consistent with those after exercise (2). This suggested that the increase in TGF-β level in the brain is partly relevant to the change of neuronal activity after exercise. It seems reasonable that TGF-β3 released in the brain during exercise suppresses spontaneous motor activity to encourage rest and causes an alteration in the energy substrates of the peripheral system.
We thank Wataru Mizunoya for technical assistance in the measurements of glycerol and LPL activity. We gratefully acknowledge Dr. G. Lynis Dohm of East Carolina University for critical reading and valuable comments.
Address for reprint requests and other correspondence: T. Fushiki, Laboratory of Nutrition Chemistry, Division of Food Science and Biotechnology, Graduate School of Agriculture, Kyoto Univ. Oiwakecho, Kitashirakawa, Sakyo, Kyoto, Japan 606-8502. (E-mail:).
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.
May 28, 2002;10.1152/ajpendo.00094.2001
- Copyright © 2002 the American Physiological Society