Intracerebroventricular injections of leucine are sufficient to suppress food intake, but it remains unclear whether brain leucine signaling represents a physiological signal of protein balance. We tested whether variations in dietary and circulating levels of leucine, or all three branched-chain amino acids (BCAAs), contribute to the detection of reduced dietary protein. Of the essential amino acids (EAAs) tested, only intracerebroventricular injection of leucine (10 μg) was sufficient to suppress food intake. Isocaloric low- (9% protein energy; LP) or normal- (18% protein energy) protein diets induced a divergence in food intake, with an increased consumption of LP beginning on day 2 and persisting throughout the study (P < 0.05). Circulating BCAA levels were reduced the day after LP diet exposure, but levels subsequently increased and normalized by day 4, despite persistent hyperphagia. Brain BCAA levels as measured by microdialysis on day 2 of diet exposure were reduced in LP rats, but this effect was most prominent postprandially. Despite these diet-induced changes in BCAA levels, reducing dietary leucine or total BCAAs independently from total protein was neither necessary nor sufficient to induce hyperphagia, while chronic infusion of EAAs into the brain of LP rats failed to consistently block LP-induced hyperphagia. Collectively, these data suggest that circulating BCAAs are transiently reduced by dietary protein restriction, but variations in dietary or brain BCAAs alone do not explain the hyperphagia induced by a low-protein diet.
- branched-chain amino acids
- protein restriction
- food intake
although the study of ingestive behavior historically has focused largely on the regulation of energy homeostasis, the consumption of adequate amounts of protein, specifically essential amino acids (EAAs), is also central to health and survival. Therefore, it seems likely that physiological systems exist to ensure sufficient protein intake. Alterations in dietary protein content can have profound effects on food intake, with diets high in protein suppressing food intake and diets moderately low in protein increasing food intake (2, 11, 18, 23, 38, 39). Furthermore, when given the choice between diets that differ in protein content, many species will self-select between diets to ensure adequate consumption of protein (16, 24, 27), often at the expense of carbohydrate and fat (9, 19, 31).
Despite these behavioral observations, the mechanism regulating protein intake is largely unknown (21). Recent work has focused on the branched-chain amino acid (BCAA) leucine as a potential protein signal. Intracerebroventricular injections of leucine suppress food intake and regulate key signaling systems (mTOR/AMPK) within hypothalamic neurons, while increased dietary leucine content reproduces the anorectic effects of a high-protein diet (3, 5, 23, 29). While these data demonstrate that administration of excess leucine either to the diet or the brain is sufficient to suppress food intake, it remains unclear whether physiological fluctuations in circulating or brain leucine actually contribute to the regulation of dietary protein intake or selection. Here, we use the adaptive hyperphagia that occurs in response to a low-protein diet to test the physiological role of dietary leucine and BCAAs in the detection of protein restriction.
MATERIALS AND METHODS
Animals and diets.
Male Sprague-Dawley (SD) rats (5–7 wk of age; Charles River Laboratories, Wilmington, MA) were used in all experiments. Rats were adapted to single housing in standard shoe-box cages with a 12:12-h light-dark cycle for at least 7 days before the study. Standard rodent chow (4.1 kcal/g, 28.5% of energy from protein, 13.5% of energy from fat, 58% of energy from carbohydrate; no. 5001, LabDiet, St. Louis, MO) was provided ad libitum unless otherwise noted. Experimental diets (seven variations in total protein, leucine, or BCAA content) were formulated and produced by Research Diets (New Brunswick, NJ) using casein and/or synthetic amino acids as the protein source (Table 1). Diets were made isocaloric via concurrent changes in carbohydrate content, while maintaining a constant fat content. The control (20% casein) and low-protein (10% casein) diets were chosen to maintain consistency with previous work from our laboratory and the laboratory of Dr. Roy Martin (23, 37, 38). These previous experiments demonstrate that a 10% casein diet reliably produces hyperphagia in rats, and the goal of the current work was to determine whether reductions in leucine in the diet or brain contribute to this hyperphagia. All experiments were approved by the Institutional Animal Care and Use Committee at Pennington Biomedical Research Center.
For cannulation of the third ventricle (22, 40), rats were anesthetized via chamber induction of 3% isoflurane in oxygen and were maintained on 3% isoflurane in oxygen via a non-rebreathing mask. The dorsal cranium was prepped for aseptic surgery, and a dorsal midline incision extending from the frontal sinuses to just caudal to the occipital protuberance was retracted laterally, allowing blunt dissection and debridement of the dorsal subcutaneous fascia and connective tissue. Bone screws (0–80 × 1/8, Plastics One; Roanoke, VA) were placed surrounding the cannulation site to serve as anchors for the methacrylate placed later. A 22-gauge stainless-steel cannula (Plastics One) or CMA 11 guide cannula (CMA Microdialysis, Kista, Sweden) was implanted after drilling (engraving cutter, 0.8-mm diameter; Dremel, Racine, WI), an entry site at coordinates −2.2 mm (Y) from bregma and −7.5 mm (Z) from the skull surface. Dental methacrylate was placed around the cannula and extended to anchor points at the previously placed bone screws. A 28-gauge obdurator or dummy was inserted in the cannula lumen, and the exposed skull was covered by partially closing the skin incision. Cannulation of the lateral ventricle followed an identical procedure, except for the cannula-targeted coordinates [−0.9 mm (X), −1.5 mm (Y) from bregma; −3.0 mm (Z) from the skull surface]. All rats received 5 mg/kg body wt carprofen (Pfizer, New York, NY) subcutaneously preoperatively and 24 h postoperatively as an analgesic. Before further study, all rats were given 1 wk to recover from surgery. During this time, food intake, body weight, and behavior were observed daily. Rats that failed to regain body weight postsurgery were excluded from the experiment. To verify correct cannula placement, an ink solution was injected through the cannula at euthanasia to confirm correct localization to the ventricle. The third ventricle was selected for acute intracerebroventricular amino acid injection because leucine was previously shown (23) to reduce food intake following third ventricular application. The lateral ventricle was chosen for chronic intracerebroventricular amino acid infusion for two reasons: 1) the lateral ventricle is a large structure to accommodate the constant flow of fluid and large amino acid doses, and 2) because the cerebrospinal fluid (CSF) flows from lateral via third to the fourth ventricle, amino acid infusion into the lateral ventricle maximizes the potential exposure of the brain to amino acids, while also delivering amino acids into the third ventricle. Finally, surgical implantation of a microdialysis guide cannula into the third ventricle was performed, as described above, and animals were given at least 1 wk to recover from surgery.
Experiment 1: effects of normal and LP diets on food intake.
To confirm low-protein-induced hyperphagia, rats (7–9/group) maintained on a chow diet were transitioned (during the mid-light cycle) to either a control, normal-protein diet (NP, 20% casein) or a low-protein diet (LP, 10% casein) diet, and daily food intake was recorded for 7 days. After 7 days on the diet, NP rats weighed 209.3 ± 4.0 g, while LP rats weighed 199.6 ± 3.5 g (means ± SE).
Experiment 2: comparative effects of select amino acids on food intake.
To determine whether leucine is unique in its capacity to regulate food intake, multiple cohorts of SD rats were implanted with indwelling third ventricle cannulas, as described above. Rats were maintained on standard rodent chow and were fasted 24 h prior to injection. Individual amino acids were injected as a single bolus (10 μg, 2 μl; 15–20 rats/group) 2 h prior to lights off, and 24-h intake of standard rodent chow was assessed. The rats were given at least 1 wk to recover between injections and were given a maximum of three injections. Average body weights at the time of injection ranged from 240 to 325 g. To test whether leucine was sufficient to suppress intake in rats consuming a LP diet, a group of rats were implanted with an intracerebroventricular cannula, and after 7 days of recovery, they were placed on LP diet for 4 days. Rats were injected intracerebroventricularly with either saline or leucine (10 μg, 2 μl; 10/group) in the fed state, and 24-h food intake was recorded.
Experiment 3: effects of NP and LP diets on plasma amino acids.
To correlate levels of food intake with changes in circulating amino acids, separate groups of rats maintained on a chow diet were transitioned (during the mid-light cycle) to either NP or LP diets, with individual groups (8/group) euthanized at day 0 (prior to dietary treatment), or days 1, 2, or 4 after the dietary change. Animals were killed in the fed state in the middle of the light cycle. Trunk blood was collected at death, and plasma was isolated to assess circulating amino acid levels.
Experiment 4: effects of NP and LP diets on brain amino acid-brain microdialysis.
Brain microdialysis was performed to assess LP-induced changes in brain amino acid concentrations. One week following implantation of the microdialysis guide cannula into the third ventricle, rats were placed into a sampling harness, and on the following day, they were transitioned (during the mid-light cycle) to the NP control diet. After 4 days of adaptation, half of the rats remained on the NP diet, while the other half were offered LP diet for 2 days (4 or 5/group). On day 2, food was removed 3 h prior to lights off, and the microdialysis probe was inserted. The CMA 11 microdialysis probes (CMA Microdialysis, Kista, Sweden; membrane length: 2 mm; membrane outer diameter: 0.24 mm) were flushed according to the manufacturer's recommendation prior to use and were inserted during brief (∼5 min) anesthesia on 3% isoflurane in oxygen. Perfusion was initiated as soon as the probe was implanted using CNS perfusion fluid (M Dialysis AB, Stockholm, Sweden) (in mmol/l): 147 NaCl, 2.7 KCl, 1.2 CaCl2, 0.85 MgCl2, at a flow rate of 1 μl/min. Perfusion began at food removal, and rats were given 3 h of recovery prior to lights off. At lights off, sample collection began, with samples collected every 30 min. One hour later (4 h of fasting), rats were offered 3.2 g of their respective NP or LP diet, which was consumed within 10 min in all animals. On a per gram basis, this test meal represents 13.5% of 24-h intake at day 2 for NP and 10.1% of 24-h intake for LP and was chosen to ensure that that all of the animals consumed equal amounts of food within the specified 10-min window, thus maintaining a similar energy intake but varying protein intake. Sampling continued for another hour in this postprandial period, at which point animals were killed, and trunk blood was collected to assess plasma amino acid levels. In summary, microdialysis samples were collected in 4-h fasted rats every 30 min for 1 h prior and 1 h following a test meal. Daily food intake and body weight were recorded throughout the 6-day protocol.
Experiment 5: effects of alterations in dietary amino acids on food intake.
To determine whether reduction in dietary leucine content is necessary or sufficient to induce hyperphagia on the LP diet, rats were exposed to one of five diets that altered dietary leucine content independently from other amino acids or total protein (12/group; starting body weight = 132 ± 2 g). The five diets were 1) normal-protein control diet with 20% casein (NP); 2) low-protein diet with 10% casein (LP); 3) a modified NP diet in which half of the protein was provided as casein and the other half was provided as free amino acids (LP+All). Total protein and amino acid balance were identical between the NP diet and the LP+All diet; 4) the LP+All diet with all free amino acids removed except leucine (LP+LEU). Thus, this diet was equivalent to the LP diet with the exception that leucine content was equivalent to NP; and 5) the LP+All diet with the free leucine removed (NP−LEU). Thus, this diet was equivalent to the NP and LP+All diets, except leucine levels were reduced to those of the LP diet. In summary, the LP+LEU diet and NP−LEU diets test whether the reduction of dietary leucine to LP levels is necessary or sufficient to induce hyperphagia. A precise description of these diets is included in Table 1. Rats were initially placed on the NP diet for 3 days before being randomly assigned to an experimental diet. Rats were maintained on these diets for 12 days, with food intake recorded daily.
Because rats seemed unresponsive to reductions in dietary leucine, a very similar experimental protocol was used to test whether the reduction of all three BCAAs (leucine, isoleucine, and valine) would be sufficient to induce hyperphagia. A second group of rats (12/group; starting body weight = 115 ± 2 g) was placed on one of five experimental diets consisting of the same control diets (NP, LP, LP+All), as well as diets that manipulate all three BCAAs in a manner similar to the changes in leucine above (LP+BCAA and NP−BCAA). Diets are described in Table 1. As above, all rats were initially placed on the NP diet for 3 days before being randomly assigned to one of five treatments. Rats were maintained on these diets for 7 days, with food intake recorded daily.
Experiment 6: effect of chronic intracerebroventricular amino acid infusion on LP-induced hyperphagia.
To assess whether elevations in brain amino acids would be sufficient to block LP-induced hyperphagia, rats consuming the NP or LP diet were chronically infused with saline or amino acids via an intracerebroventricular cannula. Indwelling lateral ventricular cannulas were implanted as described above. After 7 days of recovery, rats were briefly anesthetized (3% isoflurane in oxygen) to implant a subcutaneous osmotic minipump (Alzet Model 2002; Durect, Cupertino, CA), which was attached to the lateral ventricular cannula to allow direct brain infusion of amino acids. All rats received 5 mg/kg body wt carprofen (Pfizer, New York, NY) subcutaneously preoperatively as an analgesic. Minipumps contained either saline or increasing concentrations (5×, 10×, or 20×) of a cell culture-based amino acid mixture (MEM amino acids cat. no. 25-030-CI; Mediatech, Manassas, VA). Contents and dose rate are described in Table 2. Immediately following minipump surgery, rats were assigned to either NP or LP diet, producing five groups: NP+Saline, LP+Saline, LP+5×AA, LP+10×AA, and LP+20×AA (8–10/group). Daily food intake was recorded for 6 days.
Plasma and brain amino acid measures.
Plasma amino acids were measured using fluorometric HPLC via methods described previously (28, 41). Separation of the o-phthaldialdehyde amino acid derivatives was performed by gradient elution from a Supelcosil LC-18 column (15 cm × 4.6 mm, 3 μm; Sigma, St. Louis, MO).
Data were analyzed using the SAS software package (SAS V9; SAS Institute, Cary, NC) using a two-tailed t-test or ANOVA using the general linear model procedure. When experiment-wide tests were significant, post hoc comparisons were made using the LSMEANS statement with the PDIFF option, and represent least significant differences tests for preplanned comparisons. All data are expressed as means ± SE, with a probability value of 0.05 considered statistically significant.
Low-protein diet induces hyperphagia.
Consistent with previous experiments, rats consuming the LP diet exhibited a marked and persistent increase in food intake relative to rats consuming control, NP diet. The increase in food intake was first observed on day 2, and food intake remained elevated throughout the 7-day experiment (P < 0.01; Fig. 1).
Leucine uniquely suppresses food intake among tested essential amino acids.
Rats bearing third ventricular cannulas were fasted for 24 h, and then injected with a single amino acid (10 μg, 2 μl) or saline (n = 15–20/group). Tryptophan, methionine, lysine, threonine (all essential), and serine (nonessential) all had no effect on food intake (P > 0.05), but leucine produced a significant reduction in consumption of standard chow following the 24-h fast (P = 0.017; Fig. 2A), as shown previously (5, 23). Another group of rats were implanted with third ventricular cannulas, and following recovery, were placed on LP diet for 4 days. On day 4, rats were injected with either leucine (10 μg, 2 μl) or saline (n = 10/group) in the fed state, and 24-h food intake was recorded. Leucine injection in the LP animals significantly decreased food intake (P = 0.027; Fig. 2B). Taken together, these data suggest that elevations in brain leucine are sufficient to reduce food intake in rats consuming either chow or LP diet but that this anorectic effect is not a common feature of all EAAs.
LP diet alters plasma amino acid concentrations.
To determine whether changes in food intake in response to LP or NP diets are mirrored by changes in circulating leucine or other BCAAs, plasma amino acids were measured in rats consuming either NP diet or LP diet for either 1, 2, or 4 days. Following 1 day of the LP diet, plasma BCAAs arginine, lysine, methionine, phenylalanine, threonine, tryptophan (all EAAs), and asparagine were significantly decreased (P < 0.05; Table 3 and Fig. 3). Plasma histidine (EAA), glycine, and ornithine showed a tendency to decrease on LP at day 1 (0.05 < P < 0.1; Table 3). At day 2, plasma leucine was unaltered by the LP diet, whereas arginine, isoleucine, threonine, tryptophan, valine (all EAA), and taurine were significantly decreased (P < 0.05; Table 3 and Fig. 3). Plasma lysine and methionine (both EAA) showed a tendency to decrease on LP (P < 0.10), whereas alanine and glutamine were significantly elevated on LP diet at day 2 (P < 0.05; Table 3). At day 4, there was no difference in BCAAs between NP vs. LP (Fig. 3), with only plasma arginine, methionine, threonine, and tryptophan (all EAA) significantly decreased on LP (P < 0.05), whereas taurine showed a tendency to decrease (P < 0.10; Table 3). In contrast, alanine, glutamine, glycine, histidine, ornithine, and serine were significantly elevated on LP at day 4 (P < 0.05; Table 3), whereas glutamic acid showed a tendency to increase (P < 0.10; Table 3). These data demonstrate that reductions in dietary protein lead to transient reductions in plasma BCAAs that tend to normalize within a few days.
LP diets alter postprandial plasma and brain amino acids.
Brain levels of amino acids were measured on day 2 of LP feeding, with third ventricular microdialysis samples collected from 1 h before to 1 h after feeding of a controlled test meal. Day 2 was selected on the basis of prior evidence showing that food intake increases in LP animals on day 2. Rats used in the current study reproduce this effect, as food intake was significantly increased in LP rats on day 2 prior to sample collection (P < 0.05; Fig. 4A).
Preprandial brain BCAA levels were consistently lower in LP vs. NP rats, but this difference failed to reach statistical significance for any amino acid (P > 0.10, Table 4 and Fig. 4, B and C). However, postprandial brain levels of all BCAAs, alanine, asparagine, lysine, methionine, and threonine were significantly lower in LP vs. NP rats (P < 0.05; Table 4 and Fig. 4, B and C), whereas arginine, citrulline, and taurine showed a tendency to decrease (P < 0.10; Table 4). Measurement of amino acids in blood plasma samples collected at the end of the microdialysis experiment (1 h postprandial) revealed a significant decrease of BCAAs in LP rats (P < 0.05; Fig. 4D), as well as reduced methionine, threonine, and tryptophan (P < 0.05; Table 4). Only EAAs were reduced by LP, as nonessential amino acid (NEAA) concentrations did not differ between the NP and LP groups (Table 4 and Fig. 4D).
Leucine and BCAAs are neither necessary nor sufficient for the detection of low dietary protein.
While BCAAs fall in response to the LP diet, this reduction appears to be only transient compared with the persistent hyperphagia. Therefore, we directly tested whether reductions in dietary leucine or BCAA content were necessary or sufficient for the detection of protein restriction. As expected, the LP diet induced a significant increase in food intake compared with the NP diet or the LP+All diet (P = 0.001; Fig. 5). The LP+LEU diet induced hyperphagia to a similar extent as the standard LP diet despite containing normal levels of leucine. Similarly, food intake was not increased in the NP−LEU group despite the reduction in dietary leucine. These data suggest that the reduced leucine content in the LP diet is neither necessary nor sufficient to mediate hyperphagia on a LP diet.
To determine whether manipulation of all BCAAs would more readily alter food intake, we repeated the above experimental design in a separate group of rats, except that all three BCAAs were altered simultaneously. The results were similar. As before, rats exhibited hyperphagia in the LP diets compared with the normal protein diets (P < 0.01; Fig. 6), but reducing dietary BCAA levels is neither necessary nor sufficient to mediate hyperphagia on a LP diet.
Chronic central amino acid infusion does not blunt LP-induced hyperphagia.
To assess whether elevations in brain amino acids would be sufficient to block LP-induced hyperphagia, we chronically infused a mixture of amino acids into the brain of rats consuming the LP diet. Saline-treated LP rats consumed more food than saline-treated NP rats, as expected. However, central amino acid infusions were not sufficient to consistently suppress LP induced hyperphagia (Fig. 7).
A large number of studies indicate that variations in dietary protein alter food intake and body weight (21). The majority of these studies focus on the effect of high-protein diets to suppress food intake, increase energy expenditure, and decrease adiposity (2, 11, 18). Far fewer experiments have focused on the effects of low-protein diets, but the available data suggest that low-protein diets increase food intake and body adiposity (23, 38, 39). These data are consistent with the concept of “protein leveraging”, which suggests that protein intake is prioritized over energy (carbohydrate) intake (9, 16, 24, 31, 32). These data raise the obvious question of the identity of the protein signal. How do animals detect dietary protein content in general, and how do they avoid protein deficiency, in particular? Circulating amino acids represent an obvious candidate for this protein signal, and circulating BCAAs (leucine, isoleucine, and valine) seem well suited to act in the brain as signals of dietary protein intake. Leucine suppresses food intake when given intracerebroventricularly, directly impinges on hypothalamic neurons known to regulate food intake (NPY/AgRP and POMC), and influences signaling systems that are associated with feeding behavior (3, 5, 23, 29). However, while early studies indicated that increasing dietary BCAAs or leucine alone can reduce food intake (29), several recent experiments have failed to demonstrate an effect of leucine supplementation on food intake or hypothalamic signaling (15, 25, 45). In addition, the signaling systems thought to mediate leucine action (mTOR, AMPK) are also influenced by a wide variety of circulating and physiological signals associated with energy, but not protein, metabolism (5, 20, 43). Thus, the relevance of hypothalamic leucine signaling to a specific regulation of protein intake remains unclear. Therefore, we chose to address the role of leucine in a different context by focusing on the hyperphagia induced by a low-protein diet, a setting in which animals are protein-restricted but consume normal or elevated amounts of energy. The primary hypothesis being tested was that reduced dietary protein/leucine intake leads to decreases in circulating and brain leucine concentrations, with this fall of brain leucine providing a signal of protein restriction that triggers hyperphagia.
To determine whether leucine is unique in its capacity to regulate food intake, we first tested the relative efficacy of one nonessential (serine) and five essential amino acids to suppress food intake following intracerebroventricular injection. Because testing all EAAs was beyond the scope of our study, we chose amino acids representing an array of side chains. For instance, lysine is charged, threonine and serine are polar, and leucine, tryptophan, and methionine are hydrophobic. Methionine is sulfur-containing. Tryptophan is ringed and a precursor to serotonin. Serine, threonine, and methionine are glucogenic, while leucine and lysine are ketogenic, and tryptophan is both glucogenic and ketogenic. Serine serves as a precursor for biosynthesis of various protein and membrane lipid molecules and also interacts with N-methyl-d-aspartate receptor. On the basis of these observations, there is no obvious structural or metabolic reason to explain why leucine was the only amino acid to suppress food intake. However, a variety of studies in peripheral tissues and cell culture demonstrate that leucine is particularly effective at stimulating mTOR signaling (7, 8, 26, 30, 42). As such, we speculate that this signaling effect of leucine underlies its unique effects on food intake. However, because we only tested a subset of amino acids and did so at only a single dose, it remains possible that an, as yet, untested amino acid might also act in the brain to suppress food intake (13), or that a tested amino acid might work at a higher dose. Intracerebroventricular valine injection has been previously shown to have no effect on food intake (3–5). However, it remains possible that isoleucine would also exert effects on food intake, although isoleucine is substantially less potent at stimulating mTOR activity relative to leucine (30). Therefore, the primary conclusion is not that leucine is the only amino acid capable of regulating food intake following intracerebroventricular injection but that the suppression of food intake is not common to all EAAs. Importantly, the dose chosen (10 μg) would require a CSF volume of 25 ml to produce the 3-μM leucine concentration observed via microdialysis. Considering that the entire rat CSF volume is estimated to be around 0.1 ml, the leucine injection is producing concentrations roughly 250-fold higher than normally observed within the ventricular space, at least acutely. Therefore, even if higher doses of the tested amino acids were sufficient to suppress food intake, it is unclear whether such an effect would have any relevance to the physiological regulation of food intake.
We next tested whether fluctuations in circulating leucine correlated with changes in food intake in response to both normal- and low-protein diets. Our results show that plasma BCAAs, as well as the majority of other EAAs, decrease within 1 day of LP diet exposure (Fig. 3, Table 3). For BCAAs, this decrease is transient, however, with levels progressively increasing over time and normalizing relative to NP controls by day 4. In particular, plasma leucine levels were no longer significantly different by day 2. These changes in circulating BCAAs are consistent with the observed increases in food intake, as the hyperphagia is first observed on day 2, and food intake increases progressively through day 4. Interestingly, several other EAAs failed to normalize by day 4, including arginine, methionine, tryptophan, and, most notably, threonine. Finally, the pattern for NEAAs is much different, in that NEAAs are generally increased by the LP diet (14). Taken together, these data highlight the homeostatic regulation of blood amino acid levels in the face of reduced protein intake. LP diets induce relatively rapid decreases in liver and muscle protein synthesis and amino acid metabolism, which together serve to spare amino acids and buffer against dramatic decreases in their circulating levels (12, 33, 44). These data are also consistent with the hypothesis that reductions in circulating amino acids may represent a trigger for hyperphagia, although the rapid normalization of BCAAs levels despite persistent hyperphagia makes their specific role in the maintenance of hyperphagia unclear.
Amino acids enter the brain through diverse transport mechanisms. While brain levels tend to correlate with circulating levels, brain transporters are easily saturated, such that the brain is buffered against marked changes in circulating amino acids. Several amino acids also share a common transporter and, thereby, compete for uptake into brain. To directly assess changes in brain amino acid concentrations, microdialysis was used within the third ventricle in both the preprandial and postprandial state on day 2 of a LP diet, with day 2 chosen as the first day of hyperphagia in our model. Interestingly, preprandial brain amino acid levels were consistently lower in LP rats, but this difference was modest and did not reach statistical significance for any amino acid. However, consumption of an isocaloric NP or LP meal induced a clear postprandial difference in the CSF amino acids, primarily due to a postprandial increase in the NP controls that failed to occur in the LP group. Importantly, we observe clear differences in brain and blood levels of BCAA concentrations between NP- and LP-fed rats in the postprandial period, although it should be noted that plasma and brain amino acids were not simultaneously measured over time in the same animal.
Our data indicate that both circulating and brain concentrations of BCAAs are acutely reduced by the LP diet, particularly in the postprandial period. However, this effect is transient, and BCAAs levels in the blood normalized by day 4 despite persistent increases in food intake beyond day 4. This rapid normalization despite persistent hyperphagia raises considerable doubt regarding the role of BCAAs in the maintenance of hyperphagia on the LP diet. To more directly test the role of dietary leucine on LP-induced hyperphagia, isocaloric diets were used in which leucine was manipulated separately from total protein. Because our diets are 20% and 10% casein, respectively, we manipulated leucine levels within this dietary range, and, as such, tested whether reduced dietary leucine alone was sufficient to trigger hyperphagia, or if maintaining normal levels of dietary leucine was sufficient to block hyperphagia. Unexpectedly, the rats were unresponsive to the reduction of leucine, suggesting that reductions in dietary leucine alone are neither necessary nor sufficient to signal reduced dietary protein content. To test whether the combination of all BCAAs might alter food intake, the same experiment was repeated with the exception that all three BCAAs were manipulated. The result was the same, as the reduction of dietary BCAAs failed to trigger hyperphagia, and the maintenance of control levels failed to block hyperphagia. Importantly, rats on the control and the LP+All diet consumed very similar amounts of food in both studies, indicating that the inclusion of free amino acids in the diet did not markedly alter feeding behavior. Taken together, these data do not support the hypothesis that reduction in either leucine alone or the combination of all BCAAs represents a unique signal that triggers LP-induced hyperphagia. Importantly, our results are supported by previous work demonstrating that supplementing a LP diet with BCAAs had no effect on protein selection (1), and more recent work indicating that administration of leucine via drinking water, gavage, gastric infusion, or intraperitoneal injection all failed to suppress food intake (13, 15, 25, 45).
In light of these negative findings, it is tempting to speculate as to the composition of amino acids needed to trigger hyperphagia. One assumes that food intake would eventually increase if we continued to remove amino acids from the LP+All diet. Since plasma arginine, threonine, tryptophan, and methionine were all reduced in LP-fed animals on day 4 (Table 3), it seems possible that reduced dietary intake of these amino acids may collectively contribute to the LP-induced hyperphagia. However, it remains unclear whether the detection involves direct amino acid signaling within the brain, or alternatively whether the restriction of amino acids is detected elsewhere (gut, liver, muscle) and then secondarily signaled to the brain. We tested this question by determining whether direct brain infusion of a mixture of EAAs is sufficient to block LP-hyperphagia. While the highest dose tended to reduce food intake on the second and third day of infusion, this effect disappeared rapidly, and food intake was no different from saline-infused LP rats in any of the EAA-infused groups on days 4–6. As such, these data fail to support the hypothesis that increases in brain amino acids are sufficient to block LP-induced hyperphagia.
The work described here focused specifically on the hypothesis that alterations in brain leucine action might contribute to the hyperphagia induced by a LP diet. This hypothesis was based, in part, on our work and the work of others demonstrating that leucine suppresses food intake following intracerebroventricular injection, in part, on the evidence that leucine is particularly potent in activating intracellular signaling, and, in part, on the evidence that the brain is critically involved in the response to the dietary depletion of a single EAA. In this latter response, deprivation of a single dietary EAA activates the anterior piriform cortex (APC), specifically via reduction in the concentration of the limiting EAA in the APC (10). It should be mentioned, however, that this APC-dependent recognition of EAA deficiency produces a rejection (decreased intake) of a novel EAA-deficient meal, whereas in our work, the general restriction of dietary protein produces hyperphagia.
There is a variety of sites and mechanisms through which dietary amino acids could alter food intake, in particular, within the gastrointestinal (GI) tract, portal vein, or liver. Recent work demonstrates that intragastric infusion of glutamate, arginine, or lysine is sufficient to reduce food intake and that the effect of lysine requires an intact vagus nerve, while the effect of arginine and glutamate requires the brain stem area postrema (13). In this experiment, the amino acids are individually infused in excess as opposed to being removed from a mixed diet as in our current work, but it is also interesting to note that intragastric leucine was shown to have no effect. Amino acids in the intestine are well known to stimulate the release of anorexigenic gut peptides, such as cholecystokinin, glucagon-like peptide-1, and peptide YY, and it seems likely that the action of these hormones on vagal afferents or directly in the brain contributes to the anorectic effect of protein (6). However, these gut hormones are also induced by other macronutrients, suggesting that they are unlikely to represent unique signals of protein restriction. Amino acid sensors also exist within the gut and the hepatoportal circulation, leading to the activation of the vagus nerve (34–36), although hepatic vagal afferents do not seem essential to mediate the anorectic effects of a HP diet (17). Protein restriction also induces significant changes in hepatic metabolism (14), and thus, amino acid signaling within the liver may also contribute to the detection of protein restriction. Despite the evidence that amino acids impact hepatic function, gut hormone secretion, and vagal activity, the specific role of these mechanisms in mediating the response to dietary protein, particularly protein restriction, remain unclear (21).
Perspectives and Significance
Acute brain administration of leucine is sufficient to reduce food intake, and blood and brain levels of leucine acutely fall coincident with the initiation of hyperphagia in animals on LP diets. While these data suggest that leucine is positioned to regulate food intake during dietary protein restriction, reductions in dietary leucine or BCAA content had no effect on food intake, and chronic central amino acid infusion was not sufficient to block LP-induced hyperphagia. Taken together, we find no evidence that variations in dietary or brain BCAAs uniquely contribute to the hyperphagia induced by dietary protein restriction. These results do not indicate that BCAAs are unimportant. Instead, they suggest that the physiological mechanisms that trigger hyperphagia on a LP diet require reductions in additional amino acids beyond just the three BCAAs, and that this hyperphagia is not likely dependent on reductions in brain amino acid content.
This work was supported by a National Institutes of Health grant to C. D. Morrison (5R01DK081563). T. Laeger was supported by a research fellowship from the Deutsche Forschungsgemeinschaft LA 3042/2-1.
No conflicts of interest, financial or otherwise, are declared by the authors.
Author contributions: T.L., T.M.H., H.M., R.J.M., S.M.H., and C.D.M. conception and design of research; T.L., S.D.R., T.M.H., D.H.F., M.T., A.A., and C.D.M. performed experiments; T.L., S.D.R., T.M.H., D.H.F., M.T., A.A., S.M.H., and C.D.M. analyzed data; T.L., H.M., R.J.M., S.M.H., and C.D.M. interpreted results of experiments; T.L. and C.D.M. prepared figures; T.L., S.D.R., and C.D.M. drafted manuscript; T.L., S.D.R., T.M.H., H.M., S.M.H., and C.D.M. edited and revised manuscript; T.L., S.D.R., T.M.H., D.H.F., M.T., A.A., H.M., R.J.M., S.M.H., and C.D.M. approved final version of manuscript.
The authors would like to thank Dr. Michael Pellizon of Research Diets for assistance in diet formulation. The authors thank the staff of the PBRC Comparative Biology Core for their skillful assistance and excellent technical support. This project/work used facilities within the Animal Metabolism & Behavior Core, Genomics Core, and/or Cell Biology and Bioimaging Core at the Pennington Biomedical Research Center, which are supported, in part, by Centers of Biomedical Research Excellence (P20GM103528) and Nutrition Obesity Research Centers (P30DK072476) grants from the National Institutes of Health.
Present address: T. M. Henagan, Department of Nutrition Science, Purdue University, West Lafayette, IN 47907.
- Copyright © 2014 the American Physiological Society