Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2013 Jul 29;591(Pt 22):5611–5621. doi: 10.1113/jphysiol.2013.258947

Specific amino acids inhibit food intake via the area postrema or vagal afferents

Josua Jordi 1,2,3, Brigitte Herzog 1,3, Simone M R Camargo 1,3, Christina N Boyle 2,3, Thomas A Lutz 2,3, François Verrey 1,3
PMCID: PMC3853499  PMID: 23897232

Abstract

To maintain nutrient homeostasis the central nervous system integrates signals that promote or inhibit eating. The supply of vital amino acids is tuned by adjusting food intake according to its dietary protein content. We hypothesized that this effect is based on the sensing of individual amino acids as a signal to control food intake. Here, we show that food intake was most potently reduced by oral l-arginine (Arg), l-lysine (Lys) and l-glutamic acid (Glu) compared to all other 17 proteogenic amino acids in rats. These three amino acids induced neuronal activity in the area postrema and the nucleus of the solitary tract. Surgical lesion of the area postrema abolished the anorectic response to Arg and Glu, whereas vagal afferent lesion prevented the response to Lys. These three amino acids also provoked gastric distension by differentially altering gastric secretion and/or emptying. Importantly, these peripheral mechanical vagal stimuli were dissociated from the amino acids’ effect on food intake. Thus, Arg, Lys and Glu had a selective impact on food processing and intake suggesting them as direct sensory input to assess dietary protein content and quality in vivo. Overall, this study reveals novel amino acid-specific mechanisms for the control of food intake and of gastrointestinal function.

Key points

  • Proteins are more satiating than fats or lipids. Proteins are built by the 20 proteogenic amino acids.

  • Here, we identified l-arginine, l-lysine and l-glutamic acid as the most potent anorectic amino acids in rats.

  • l-Arginine and l-glutamic acid require intact neurons in the area postrema to inhibit food intake, whereas l-lysine requires intact afferent fibres of the vagus nerve. All three mediate their effect by the blood stream.

  • All three amino acids induce gastric distension by delaying gastric emptying and inducing secretion. However, the gastric phenotype does not mediate the anorectic response.

  • These results unravel amino acid-specific mechanisms regulating digestion and eating behaviour and thereby contribute to the understanding of nutrient sensing in vivo.

Introduction

Survival of an organism depends on its capability to adjust eating to its nutritional requirements. The brain ultimately controls this balance by integrating diverse signals related to the nutritional status of the body and the ingested meal both in terms of its energy and macronutrient content (Morton et al. 2006). Meal-borne signals can be sensed centrally in specialized brain areas or be relayed from the periphery by abdominal vagal afferents, all finally leading to meal termination. The induction of these satiating signals might be triggered directly by circulating nutrients, indirectly by hormones released from enteroendocrine cells or via mechanical stimuli such as gastric distension as discussed in several reviews (Morton et al. 2006; Woods, 2009; Fromentin et al. 2012). In many cases, these mechanisms were studied for their role in energy homeostasis, whereas their macronutrient specificity was subject to less extensive research.

Protein intake is a necessity to supply the organism with essential amino acids and species as diverse as insects, fish, rodents and humans adjust food intake according to its dietary protein content (Anderson et al. 2004; Morrison et al. 2012). The sensory mechanisms eliciting this behaviour are however debated. Taste, smell, sight and texture of a meal have been suggested to help in identifying its protein content based on learned associations, but experimental alterations of these cues did not prevent the anorectic effect of a high protein diet in rodents (L’Heureux-Bouron et al. 2004). A physiological sensory mechanism detecting the protein content of a meal might therefore originate from the gut or be post-absorptive for instance in peripheral nerves or directly in the brain. Indeed, the digestion of proteins to amino acids in the gastrointestinal tract stimulates vagal and endocrinal signals including CCK, GLP-1 and PYY release (Blom et al. 2006; Tome et al. 2009). However, total vagotomy did not interfere with the anorectic effect of a high protein diet (L’Heureux-Bouron et al. 2003) and the specificity of gastrointestinal hormones for protein intake is debatable as lipids and carbohydrates induce similar hormonal responses (Morrison et al. 2012). An alternative possibility is that amino acids may act directly on nutrient sensing neurons. Indeed, it was shown that injection of the amino acid leucine directly into the brain ventricles inhibits food intake via modulation of mTOR signalling (Cota et al. 2006; Morrison et al. 2007; Blouet & Schwartz, 2012). However, the role of this effect was challenged by the observation that physiological changes in plasma Leu concentration did not inhibit food intake even though circulating amino acids have access to the brain (Laviano et al. 2006; Zhang et al. 2007; Potier et al. 2009; Nassl et al. 2011; Morrison et al. 2012; Purpera et al. 2012). Importantly, all other 19 individual amino acids were so far not studied. Hence, the sensory mechanisms tuning ingestion of food to the dietary protein content remains elusive.

Here, we hypothesize that specific individual amino acids act as a physiological signal to control food intake. First, we systematically investigated which individual amino acid given by oral gavage had a specific impact on food intake in rats. We showed that Arg, Lys and Glu were significantly more active than all other proteogenic amino acids at an isomolar dose. To localize their respective sensory mechanism, we performed surgical invalidations that demonstrated in the case of Arg and Glu the involvement of area postrema (AP) neurons, and in the case of Lys of vagal afferents, respectively. Interestingly, in the gut all three amino acids induced gastric distension by specifically inducing secretion or delaying gastric emptying. However, these peripheral mechanical stimuli were not linked to the control of food intake as intravenously applied Arg and Lys inhibited food intake, but did not modulate gastrointestinal function. These results demonstrate for the first time the importance of Arg, Lys and Glu for the independent adaptation of digestive functions and eating behaviour probably as a direct sensory input to evaluate dietary protein content and quality.

Methods

Ethical approval

All procedures for rat handling and experimental interventions were according to the Swiss Animal Welfare laws, approved by the Kantonale Veterinäramt Zürich and conform to the principles of UK regulations.

Animal care

Male Wistar rats (Janvier labs, CS 4105 Le Genest Saint Isle, Saint Berthevin Cedex 53941, France) were individually housed in wire-mesh hanging cages (room temperature 21 ± 1°C, artificial 12/12 h light/dark cycle, water ad libitum, rat chow-3430 Kliba Nafag, Kaiseraugst, Switzerland).

Application and killing

Intragastrically administered l-amino acids (freshly prepared, Sigma-Aldrich, Buchs, Switzerland) were applied in 2 ml tap-water. The single dose selected was 6.7 mmol kg−1 for the screen based on comparison to other previous publications (Bialik et al. 1989; Ng & Anderson, 1992). The plasma amino acid changes measured after individual amino acid administration (Supplementary Table 2) are in the range observed after the intake of a high protein diet (Peters & Harper, 1987). Intravenously administered l-amino acids were dissolved in Ringer lactate and adjusted to pH 7.2. Animals were killed using pentobarbital (i.p., 100 mg kg−1; Kantonsapotheke, Zürich, Switzerland) in combination with isoflurane (5%) for quicker induction.

Measurement of food and water intake

Food/water intake was measured manually by weighing food containers/water bottles or using an automated system (BioDAQ, Research Diets, New Brunswick, NJ, USA). A meal was defined by the intake of at least 0.25 g chow, and the intermeal interval criterion was set to 10 min.

Blood analysis

Blood was collected into EDTA-coated tubes (Sarsted, Sevelen, Switzerland), inverted and centrifuged (1000 g, 10 min, 4°C). Plasma amino acid concentration were measured by the Functional Genomics Center Zurich using high pressure liquid chromatography, plasma ions by enzymatic detection with Picollo Renal Functional Panel (Abaxis, Darmstadt, Germany) and plasma glucose by AccuCheck Aviva (Roche, Basel, Switzerland). Gastric secretion was previously shown to correlate with base excess in blood, an effect termed alkaline tide (Rune, 1966; Niv et al. 1993).

Phenol red quantification

Phenol red (Sigma) was quantified as described by others (Tsurugizawa et al. 2009). Phenol red recovery was 57.5 ± 0.6% (Supplementary Fig. S3A), respectively 74.3 ± 0.9% (Supplementary Fig. S3B) of the administered oral dose with no significant differences between treatment groups, thereby enabling comparison.

cFOS immunohistochemistry

This time point after gavage was chosen, because cFOS expression is strongest 90–120 min post-stimulation (Watts et al. 2006). Animals were anaesthetized with pentobarbital (no isoflurane) and perfused transcardially with ice-cold phosphate buffer (0.1 m), followed by 4% paraformaldehyde in phosphate buffer. After removal, brains were kept in paraformaldehyde for 2 h to achieve proper tissue fixation. Following incubation in 20% sucrose solution (in phosphate buffer, 48 h, 4°C) brains were snap frozen in hexane. Coronal sections (20 μm) were cut in a cryostat (CM 3050 Leica, Germany) throughout the brain. Every slice was thaw mounted on microscopic glass slides (SuperFrost Plus Faust, Switzerland). For the detection of cFos expression, frozen sections were air-dried at room temperature for 1 h and rehydrated in phosphate-buffered saline. Unspecific binding was blocked by 2 h incubation in 1.5% normal donkey serum. The primary antibody (1:5000, rabbit anti cFos, Ab-5, Calbiochem, EMD Millipore, 290 Concord Road, 01821 Billerica, MA, USA) was applied for 48 h at 4°C. Sections were incubated with the secondary antibody (1:10,000, biotinylated donkey antirabbit, Jackson 711-065-152, ImmunoResearch Laboratories, Inc., 872 West Baltimore Pike West Grove, PA, USA) for 2 h at room temperature. After incubation in ABC (Vectastain-Elite ABC Kit; Vector Laboratories, 3, Accent Park, Bakewell Road Orton Southgate, Peterborough, PE2 6XS, England), followed by 0.05% DAB solution (in 0.05 m Tris-HCl with 0.009% H2O2 and colour enhancement with 0.04% NiCl2.6H2O and 0.08% CoCl2.6H2O), the sections were dehydrated in graded alcohols, cleared in xylene and fixed with entellan. The rat brain atlas of Swanson was used to localize the cFOS expressing neurons (Swanson, 2004). In three adjacent sections, cFos-positive cells were counted manually approximately at bregma −13.76 in the AP and the nucleus of the solitary tract (NTS) by a treatment-blinded investigator.

Area postrema lesion surgery

AP lesion surgery was conducted similar to a thermal lesion approach described previously, except that the AP was removed by vacuum aspiration to reduce damage in the NTS (Lutz et al. 1998). Surgery success and specificity was verified functionally and histologically. The functional test was based on the absence of amylin-induced satiation, which depends on an intact AP, and the presence of CCK-induced satiation, which depends on intact vagal afferents (Lutz et al. 1998; Ruttimann et al. 2009). Sixteen hour food-deprived rats were i.p. injected with amylin (10 μg kg−1; Bachem AG, Bubendorf, Switzerland), CCK-8 (4 μg kg−1; Bachem AG) or saline, respectively in a cross-over design. AP-lesioned animals that did not reduce 30 min post-CCK application food intake by at least 30% were excluded due to a damaged vagus nerve (six of 24 animals; food intake: sham NaCl 4.2 ± 0.3 g, sham CCK 2.1 ± 0.4 g, AP lesion NaCl 2.9 ± 0.3 g, AP lesion CCK 0.9 ± 0.3 g). AP-lesioned animals that reduced 2 h post-amylin application food intake by more than 15% were excluded due to an incomplete AP lesion (seven of 24 animals; food intake: sham NaCl 8.6 ± 0.6 g, sham amylin 6.5 ± 0.5 g, AP lesion NaCl 7.6 ± 0.6 g, AP lesion amylin 8.3 ± 0.7 g). Histological validation of the surgery included a test for the anterograde labelling capacity of vagal afferents to the NTS (Ruttimann et al. 2009). Rats were i.p. injected with 1 mg fluorogold (Fluorochrome, Denver, CO, USA) in 1 ml saline 3 days before their brains were excised and fixed by a pH switch protocol described by Khan et al. (2007). Brain sections were cut as described above, slides air-dried and fixed with glycerol. An observer blind to surgery quantified the extent of the AP lesion and counted fluorogold-labelled neurons in the NTS. Only animals with histologically confirmed AP lesions and similar numbers of fluorogold-positive NTS neurons as sham-lesioned animals were included for the final statistical analysis.

Capsaicin vagotomy

The local vagal capsaicin treatment was performed as previously described (Raybould and Tache, 1988). Surgery success and specificity was verified functionally and histologically. The functional test was the absence of CCK-induced satiation, as described above. Capsaicin treatment was considered successful if CCK reduced eating 30 min post-application by less than 30% compared to controls (three of 13 animals were excluded; food intake: sham NaCl 4.8 ± 0.2 g, sham CCK 2.5 ± 0.3 g, capsaicin-treated NaCl 3.9 ± 0.4 g, capsaicin-treated CCK 3.7 ± 0.3 g). Histological validation of the surgery included a test for the anterograde labelling capacity of vagal afferents to the NTS as described above. Only animals with histologically confirmed fluorogold staining in the AP (as injection control) and decreased fluorogold staining in the NTS (due to damaged vagal afferents) were included into the final statistical analysis.

Conditioned taste aversion

Water access was restricted during the 14 days of the conditional taste aversion test. Importantly animals received access to water or saccharine specific to the experimental day as specified below for a 30 min period at dark onset. In the first 4 days of the experiment, rats received access to two water bottles to adapt animals to the drinking schedule. On day 5, rats received access to two bottles of saccharine (0.3% in tap water; Sigma). Saccharine intake was equal between experimental groups on day 5 thereby showing equal preference for saccharine intake. Following the 30 min access to saccharine, amino acids (i.g., 6.7 mmol kg−1; i.v. 2 mmol kg−1) or LiCl (i.p., 76 mg kg−1; Sigma), which was used as positive control, were administered. This conditioning paradigm was repeated on days 8 and 11; in between, i.e. on days 6–7, 9–10 and 12–13, respectively, water was offered as described for days 1–4. On day 14, rats were presented one bottle of water and one bottle of 0.3% saccharine for 30 min at dark onset. On all test days, rats were offered water 4 h after dark onset for a 45 min period every day to ensure proper hydration. Liquid intake was measured throughout the entire experiment by weighting the bottles.

Images

Stomach images were captured with a digital camera (D50, Canon), brain slice images with the microscope (Axio Imager 2, Zeiss). Images were not altered except for minor adjustments in brightness and contrast. Fluorogold images were black/white inverted for better visibility.

Statistics

Rats were randomly allocated into treatment and/or surgical group. The application order was randomized. Results are presented as mean ±s.e.m. The data were analysed using Graph Pad Prism 5.0.

Results

Identification of the most potent oral anorectic amino acids

To identify the most potent anorectic individual amino acid, we administered isomolar doses of all 20 proteogenic amino acids individually to rats via the physiological gastrointestinal route and measured subsequent food intake. The amino acids were administered by gavage at an equal dose of 6.7 mmol kg−1, corresponding in the case of Glu to 1 g kg−1 and 27% of average daily Glu intake. Surprisingly the anorectic response induced by individual amino acids did not follow a rank order according to side chain structure, energy content, nutritional necessity (essential) or role as neurotransmitter (-precursor), but rather revealed a unique role for Arg, Lys and Glu (Fig. 1A). Each of these three amino acids significantly reduced food intake in the first hour after administration, whereas all other amino acids had no significant effect compared to water control (Fig. 1A). In the long term only Arg and Lys decreased food intake, whereas Glu-treated animals compensated for the initial decrease in food intake in the following 48 h (Fig. 1B). Consistent with previous studies, oral Trp had a small effect on food intake which, however, was statistically not significant when testing for multiple comparisons (Fig. 1A; unpaired two-tailed t test, P= 0.03; Ng and Anderson, 1992); the other aromatic amino acids Phe or Tyr did not inhibit food intake as previously shown (Bialik et al. 1989). Meal pattern analysis indicated that oral Arg, Lys and Glu administration reduced eating only by an effect on the first meal size while the latency to meal initiation, meal number, meal duration or time between meals did not differ (Fig. 1C–F). Additional control experiments excluded that pH, osmolarity or the feeding state, hence the presence of nutrients, altered the anorectic effect of Lys (Supplementary Fig. S1); dose–response studies indicated a plateauing of the anorectic effect at 4.7 to 5.4 mmol kg−1 (Supplementary Fig. S2A–C). Drinking behaviour was not altered by Arg and Lys, but Glu reduced water intake in a dose-dependent manner (Supplementary Fig. S2D–F). These findings are striking because they identify Arg, Lys and Glu as the individual amino acids with the strongest anorectic effect.

Figure 1. Intragastric Arg, Lys and Glu inhibited food intake most potently of all proteogenic amino acids.

Figure 1

Open in a new tab

After 12 h food deprivation rats were gavaged with isomolar doses of individual amino acids (6.7 mmol kg−1, which in the case of Glu corresponds to 1 g kg−1 and to 27% of the daily average Glu intake) and their subsequent food intake measured for 1 (A), 2, 4, 24 and 48 h (B); n= 12, mean ±s.e.m.; (A) unpaired one-way ANOVA, Dunnett post-test; (B) unpaired two-way ANOVA, Bonferroni post-test; *P < 0.05, **P < 0.01, ***P < 0.001. (C–F) After 16 h food deprivation rats were gavaged with individual amino acids (6.7 mmol kg−1) and their subsequent meal pattern analysed in an automated BioDaq system; n= 10, mean ±s.e.m.; unpaired one-way ANOVA, Dunnett post-test; *P < 0.05.

Neuronal pathways mediating the anorectic effect of Arg, Lys and Glu

Eating behaviour is ultimately controlled by the brain, which integrates a large array of sensory inputs projecting to specific areas (Berthoud, 2002). Therefore we investigated where in the brain Arg, Lys and Glu triggered neuronal activity by assessing the number of cells expressing cFOS, a neuronal activity marker induced by increases in intracellular Ca2+ (Watts et al. 2006). Upon gavage with these three amino acids, an increased number of cFOS-positive cells were detected in the NTS and the AP; two areas known to mediate the effect of specific anorectic signals (Fig. 2). Importantly, the AP is not protected by the blood–brain barrier such that it may directly sense blood-borne signals, and the NTS is the main projection site of vagal afferents (Morton et al. 2006). In contrast, the nucleus accumbens or the hypothalamus (arcuate nucleus, paraventricular nucleus of the hypothalamus, lateral hypothalamic area) did not display increased cFOS expression at this time point. To discriminate whether the anorectic effects of Arg, Lys and Glu were mediated via the AP or vagal afferents respectively, we surgically lesioned both routes. The AP was vacuum aspirated, whereas vagal afferents were chemically lesioned by local subdiaphragmatic capsaicin application. Surgery specificity and success was validated functionally and histologically (Fig. 3A). Arg and Glu lost their anorectic effect in AP-lesioned animals, but not in capsaicin-treated animals (Fig. 3B and C). Lys, on the other hand, showed an opposite response to the surgical interventions because its anorectic effect was dependent on intact vagal afferents but not on intact AP (Fig. 3). These findings suggest that the anorectic response to Arg and Glu is induced centrally in the AP, whereas the Lys response is relayed from the periphery to the NTS by abdominal vagal afferents.

Figure 2. Intragastric Arg, Lys and Glu induced neuronal activity in the brainstem.

Figure 2

Open in a new tab

After 16 h food deprivation, rats were gavaged with isomolar doses (6.7 mmol kg−1) of individual amino acids and transcardially perfused with paraformaldehyde 2 h later. Animals had no access to food between gavage and transcardial perfusion. Different brain areas were analysed for cFOS expression. A, representative images showing cFOS-positive cells in the AP and the NTS (located –13.76 from bregma). Scale bar, 100 μm. Quantification of cFOS-positive cells in the AP (B) and NTS (C); n= 6–8, mean ±s.e.m.; unpaired one-way ANOVA, Dunnett post-test; **P < 0.01, ***P < 0.001. 4V, 4th ventricle; AP, area postrema; NTS, nucleus of the solitary tract; DMX, dorsal motor nucleus vagus nerve; GR, gracile nucleus.

Figure 3. AP mediated the anorectic effect of Arg and Glu, whereas abdominal vagal afferents are necessary for the anorectic effect of Lys.

Figure 3

Open in a new tab

(A) Histological surgery validation showed complete removal of the AP and an intact vagus nerve, visualized by retrograded tracing of i.p. injected fluorogold to the NTS (primary vagal projection site), in AP-lesioned animals. Capsaicin treatment lesioned the vagus nerve and consequently less fluorogold reached the NTS in respective animals. The AP showed fluorogold labelling due to the absence of the blood–brain barrier. Location −13.76 from bregma. Scale bar 100 μm. After 16 h food deprivation AP-lesioned (B), capsaicin-treated (C) and the respective sham animals were gavaged with individual amino acids (6.7 mmol kg−1) and their subsequent food intake measured for 1 h; n= 7–11, mean ±s.e.m.; repeated measures two-way ANOVA, Bonferroni post-test; *P < 0.05, **P < 0.01, ***P < 0.001. 4V, 4th ventricle; AP, area postrema; NTS, nucleus of the solitary tract; DMX, dorsal motor nucleus vagus nerve; GR, gracile nucleus.

Satiation effects are dissociated from gastrointestinal actions

Furthermore, vagal afferents fire in response to gastric distension, which may contribute to meal-ending satiation (Woods, 2009). Apart from the ingestion of food and water, gastric distension can also be due to gastric secretion or delayed gastric emptying, the latter described to depend mainly on the gastric caloric content (Camilleri, 2006). Arg, Lys and Glu increased gastric distension 0.5 h after their application, an effect that coincided with the time of decreased food intake (Fig. 4A). The increase in gastric volume by Arg and Lys administration appeared to be mediated by acidic gastric secretion as the alkaline tide increased plasma albumin and decreased plasma Cl− concentration (Supplementary Table S1). Gastric emptying was delayed by Lys and Glu, as measured by phenol red retention in the stomach (Supplementary Fig. S3A). For all three amino acids, wet weight of the stomach and/or caecum was still increased 1.5 h after amino acid administration (Fig. 4B). These findings demonstrate that specific amino acids differentially modulate gastrointestinal function, an important finding in light of the caloric focus in this field (Camilleri, 2006). We next tested whether the gastric and anorectic responses to Arg, Lys and Glu are functionally linked by bypassing the gastrointestinal lumen through intravenous amino acid administration. The dose of 2 mmol kg−1 was selected, because it led to similar changes in plasma amino acid concentration as the oral dose (6.7 mmol kg−1) given by gastric gavage (Supplementary Table S2). Intravenous Arg, Lys and Glu reduced food intake similar as after oral application, but only intravenous Glu had an effect on gastric function (Fig. 5, Supplementary Fig. S3B, Table S2). However, in AP-lesioned animals Glu delayed gastric emptying, but did not inhibit food intake (Fig. 3; Supplementary Fig. S4). These data suggest that all three amino acids Arg, Lys and Glu, inhibit food intake through circulation and that their effect on gastrointestinal function can be dissociated from their anorectic effect.

Figure 4. Arg, Lys and Glu induced gastric distension by distinct mechanisms.

Figure 4

Open in a new tab

A, after 16 h food deprivation rats were gavaged with an isovolumic (2 ml) dose of individual amino acids (6.7 mmol kg−1); 30 min later their stomach was excised and weighted; n= 6, mean ±s.e.m.; unpaired one-way ANOVA, Dunnett post-test; *P < 0.05, ***P < 0.001. Representative images are shown; scale bar, 4 mm. B, after 16 h food deprivation rats were gavaged with individual amino acids (6.7 mmol kg−1); 30 min later received access to 3 g rat chow and 1.5 h post-administration the gastrointestinal tract was excised and the wet weight measured; n= 6, mean ±s.e.m.; unpaired two-way ANOVA, Bonferroni post-test; *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 5. Intravenous Arg, Lys and Glu inhibited food intake, but only intravenous Glu delayed gastric emptying.

Figure 5

Open in a new tab

A, after 16 h food deprivation rats were anaesthetized, received RLA or amino acid (2 mmol kg−1) injection into the lateral tail vein, were gavaged with water or amino acid (HK, LK) and their subsequent food intake was measured for 1 h; n= 9–11, mean ±s.e.m.; unpaired one-way ANOVA, Dunnett post-test; *P < 0.05, **P < 0.01, ***P < 0.001. B, after 16 h food deprivation rats were anaesthetized, received RLA or amino acid (2 mmol kg−1) injection into the lateral tail vein, were gavaged with water or amino acid (HK, LK), 30 min later their stomach was excised and weighted; n= 6, mean ±s.e.m.; unpaired one-way ANOVA, Dunnett post-test; *P < 0.05, ***P < 0.001. C, after 16 h food deprivation rats were anaesthetized, received RLA or amino acid (2 mmol kg−1) injection into the lateral tail vein, were gavaged with water, 30 min later received access to 3 g rat chow and 1.5 h post-administration the gastrointestinal tract was excised and the wet weight measured; n= 6, mean ±s.e.m.; unpaired two-way ANOVA, Bonferroni post-test; *P < 0.05. HK, high l-lysine dose (6.7 mmol kg−1 Lys); LK, low l-lysine dose (2 mmol kg−1 Lys); RLA, Ringer lactate.

Finally, we tested if the three amino acids cause visceral discomfort. Therefore, we conducted a conditioned taste aversion test after Arg, Lys and Glu administration by the intragastric and intravenous route respectively. Oral Glu treatment did not alter the innate preference to saccharine of rats, whereas oral Arg and Lys did induce conditioned taste aversion (Fig. 6A). If Arg and Lys were given intravenously, animals did not reduce their saccharine intake (Fig. 6B). These data suggest that the gastrointestinal function changes caused by oral Arg and Lys, but not by intravenous Arg and Lys, induced conditioned taste aversion. Importantly the anorectic effect of Arg and Lys was equally prominent after oral or intravenous application suggesting limited importance of visceral discomfort for their anorectic effect.

Figure 6. Conditioned taste aversion test following different routes of amino acid administration.

Figure 6

Open in a new tab

A, rats received an amino acid (6.7 mmol kg−1) or water gavage during the conditioning days. B, rats received an amino acid (2 mmol kg−1) or RLA injection into the lateral tail vein and were gavaged with water or amino acid (HK) during the conditioning days. i.p. injected LiCl is the positive control. n= 6 (LiCl n= 3), mean ±s.e.m.; unpaired one-way ANOVA, Dunnett post-test; *P < 0.05, **P < 0.01, ***P < 0.001. HK, high l-lysine dose (6.7 mmol kg−1 Lys); RLA, Ringer lactate.

Discussion

Our study establishes a novel role for amino acids Arg, Lys and Glu in the control of food intake and gastrointestinal function. This expands the existing concept of gastrointestinal function modulation through the caloric content of meals by adding specific amino acids as novel calorie independent signals (Camilleri, 2006). By inducing gastric secretion these amino acids may favour acid-mediated denaturation of proteins and, by delaying gastric emptying, extend the time for enzymatic protein digestion. Hence, we suggest that the gastrointestinal tract recruits appropriate digestive capacity to cope with a high protein load based on specific amino acid detection. Future work should more systematically test the effect of individual amino acids on gastrointestinal function with an emphasis on translating animal findings to humans. Gastric distension is a mechanical stimulus, transmitted by vagal afferent to the NTS that may contribute to the control of food intake (Davis et al. 1993; Schwartz, 2006; Woods, 2009). Our experiments did, however, not support such a nutrient unspecific gut–brain axis and showed a clear dissociation of the gastric and the behavioural responses to Arg, Lys and Glu as schematically represented in Fig. 7. In agreement, gastric distensions in balloon inflation studies did not inhibit food intake in rats and humans (Oesch et al. 2006).

Figure 7. Schematic representation of main findings.

Figure 7

Open in a new tab

Arg, Lys and Glu were identified as the most potent oral anorectic amino acids. After gastric delivery, they specifically stimulated gastric secretion and/or delayed gastric emptying. After absorption only Glu inhibited gastric emptying. Importantly all three amino acids also inhibited food intake when administered intravenously but by a different neuronal mechanism. Lys is detected by vagal afferents projecting to the NTS, whereas Arg and Glu centrally in the AP – a brain area not protected by the blood barrier. Thus, Arg, Lys and Glu selectively affected food processing and intake suggesting a direct sensory input to assess dietary protein content and quality. AP, area postrema; NTS, nucleus of the solitary tract.

Our data highlight the importance of circulating amino acids for the control of food intake and we identified Arg, Lys and Glu as the most effective ones, at least at isomolar doses. Naturally, a complete dose–response relationship for all 20 amino acids would be desirable; nevertheless, the single dose approach used here enabled the identification of the most potent anorectic amino acids. All three induced satiation by reducing first meal size, but not meal duration, unlike classical satiation-inducing peptides such as CCK or amylin, which proportionally reduce meal size and duration (Woods, 2009). Interestingly, the observed amino acid specificity is in marked contrast to the previously emphasized unique role of the amino acid Leu, which inhibits food intake by regulating mTOR signalling when administered directly to the third brain ventricle (Cota et al. 2006; Morrison et al. 2007). Indeed, consistent with others who showed that physiological changes in plasma Leu concentration do not lead to a reduction of food intake, our data also do not support a role for peripheral Leu in food intake control (Laviano et al. 2006; Zhang et al. 2007; Nassl et al. 2011; Morrison et al. 2012; Purpera et al. 2012). Importantly, so far no systematic study on the effect of the 20 proteogenic amino acids had been published (Fig. 1A). Here, we unravel an amino acid specificity that can be interpreted either as an evolutionary reductive strategy to detect the few amino acids that most accurately indicated protein supply or as a strategy to detect protein quality (Lamb, 2012). Supporting the latter possibility, high protein diets of different sources differentially alter satiation in rodents and humans (Anderson et al. 2004; Faipoux et al. 2006; Pichon et al. 2008).

Interestingly two different afferent pathways respond specifically to one of the candidate amino acids. This redundancy may ensure accurate detection of protein intake in vivo. Others have already proposed the existence of a Lys sensor located on vagal afferents innervating the hepatic portal vein. However, the molecular identity of the sensor remained obscure (Torii & Niijima, 2001). Similarly the molecular mechanism of Arg and Glu sensing by AP neurons is unknown, but other studies previously reported neurons able to sense specific amino acids (Cota et al. 2006; Karnani et al. 2011; Blouet & Schwartz, 2012). We identified here three novel amino acids as potent modulators of short-term food intake and suggest their neuronal site of action. They can contribute to the anorectic effect of a high protein diet as Arg, Lys and Glu plasma concentrations were shown to raise proportionally to the protein content of a meal (Peters & Harper, 1987). Analogously dietary Arg supplementation was shown to reduce body weight gain in rats (Fu et al. 2005). Based on the aversive response rats exhibit to diets depleted from essential amino acids, we expect a diet depleted from Arg to be rejected but we are currently not aware of such a study (Morrison et al. 2012). One can speculate that a diet supplemented or partially restricted by one of the three candidate amino acids may mimic the effect of a high or of a low protein diet on feeding behavior. Taken together we propose that the central nervous system detects dietary protein content and quality to control food intake by sensing specific circulating amino acids, in particularly Arg and Glu via the AP and Lys via vagal afferents (Fig. 7).

Acknowledgments

We want to thank Daniel Tomé for discussions, Daria Stöcker, Annika Donauer and Myrtha Arnold for technical advice and the Functional Genomics Center Zurich their service.

Glossary

4V

4th ventricle

Ala

l-alanine

AP

area postrema

Arg

l-arginine

CCK

cholecystokinin

DMX

dorsal motor nucleus vagus nerve

GLP1

glucagon-like peptide 1

Glu

l-glutamic acid

GR

gracile nucleus

HK

high l-lysine dose (6.7 mmol kg−1 Lys)

Leu

l-leucine

LK

low l-lysine dose (2 mmol kg−1 Lys)

Lys

l-lysine

mTOR

mammalian target of rapamycin

NTS

nucleus of the solitary tract

Phe

l-phenylalanine

PYY

peptide YY

Trp

l-tryptophan

Tyr

l-tyrosine

Additional information

Competing interests

All authors declare no competing financial, personal, or professional interest that could be construed to have influenced this paper.

Author contribution

All authors contributed to study design, data interpretation and manuscript revision. B.H. contributed to data collection. J.J. performed all experiments and surgeries and wrote the paper with the help of T.A.L. and F.V. All aspects of the project were supervised by T.A.L. and F.V. All authors approved the final version of the manuscript and all authors qualifying for authorship are listed.

Funding

J.J. and this work were supported by the Zurich Center of Integrative Human Physiology (ZHIP). The laboratories of F.V. and T.A.L. are supported by Swiss NSF grant 31-130471/1 and 31-138246, respectively and the National Centre of Competence in Research (NCCR) Kidney.CH (FV).

Supplementary material

Supplementary Table 2

Supplementary Fig. S3A

Supplementary Fig. S3B

Supplementary Fig. S1

Supplementary Fig. S2A–C

Supplementary Fig. S2D–F

Supplementary Table S1

Supplementary Table S2

Supplementary Fig. S4

tjp0591-5611-SD1.docx (218.7KB, docx)

References

  1. Anderson GH, Tecimer SN, Shah D, Zafar TA. Protein source, quantity, and time of consumption determine the effect of proteins on short-term food intake in young men. J Nutr. 2004;134:3011–3015. doi: 10.1093/jn/134.11.3011. [DOI] [PubMed] [Google Scholar]
  2. Berthoud HR. Multiple neural systems controlling food intake and body weight. Neurosci Biobehav Rev. 2002;26:393–428. doi: 10.1016/s0149-7634(02)00014-3. [DOI] [PubMed] [Google Scholar]
  3. Bialik RJ, Li ET, Geoffroy P, Anderson GH. Route of delivery of phenylalanine influences its effect on short-term food intake in adult male rats. J Nutr. 1989;119:1519–1527. doi: 10.1093/jn/119.10.1519. [DOI] [PubMed] [Google Scholar]
  4. Blom W, Lluch A, Stafleu A, Vinoy S, Holst JJ, Schaafsma G, Hendriks HF. Effect of a high-protein breakfast on the postprandial ghrelin response. Am J Clin Nutr. 2006;83:211–220. doi: 10.1093/ajcn/83.2.211. [DOI] [PubMed] [Google Scholar]
  5. Blouet C, Schwartz GJ. Brainstem nutrient sensing in the nucleus of the solitary tract inhibits feeding. Cell Metab. 2012;7:579–587. doi: 10.1016/j.cmet.2012.10.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Camilleri M. Integrated upper gastrointestinal response to food intake. Gastroenterology. 2006;131:640–658. doi: 10.1053/j.gastro.2006.03.023. [DOI] [PubMed] [Google Scholar]
  7. Cota D, Proulx K, Smith KA, Kozma SC, Thomas G, Woods SC, Seeley RJ. Hypothalamic mTOR signaling regulates food intake. Science. 2006;312:927–930. doi: 10.1126/science.1124147. [DOI] [PubMed] [Google Scholar]
  8. Davis JD, Smith GP, Miesner J. Postpyloric stimuli are necessary for the normal control of meal size in real feeding and sham feeding rats. Am J Physiol Regul Integr Comp Physiol. 1993;265:R888–R895. doi: 10.1152/ajpregu.1993.265.4.R888. [DOI] [PubMed] [Google Scholar]
  9. Faipoux R, Tome D, Bensaid A, Morens C, Oriol E, Bonnano LM, Fromentin G. Yeast proteins enhance satiety in rats. J Nutr. 2006;136:2350–2356. doi: 10.1093/jn/136.9.2350. [DOI] [PubMed] [Google Scholar]
  10. Fromentin G, Darcel N, Chaumontet C, Marsset-Baglieri A, Nadkarni N, Tome D. Peripheral and central mechanisms involved in the control of food intake by dietary amino acids and proteins. Nutr Res Rev. 2012;25:29–39. doi: 10.1017/S0954422411000175. [DOI] [PubMed] [Google Scholar]
  11. Fu WJ, Haynes TE, Kohli R, Hu J, Shi W, Spencer TE, Carrol RJ, Meininger CJ, Wu G. Dietary l-arginine supplementation reduces fat mass in Zucker diabetic fatty rats. J Nutr. 2005;135:714–721. doi: 10.1093/jn/135.4.714. [DOI] [PubMed] [Google Scholar]
  12. Karnani MM, Apergis-Schoute J, Adamantidis A, Jensen LT, de Lecea L, Fugger L, Burdakov D. Activation of central orexin/hypocretin neurons by dietary amino acids. Neuron. 2011;72:616–629. doi: 10.1016/j.neuron.2011.08.027. [DOI] [PubMed] [Google Scholar]
  13. Khan AM, Ponzio TA, Sanchez-Watts G, Stanley BG, Hatton GI, Watts AG. Catecholaminergic control of mitogen-activated protein kinase signalling in paraventricular neuroendocrine neurons in vivo and in vitro: a proposed role during glycemic challenges. J Neurosci. 2007;27:7344–7360. doi: 10.1523/JNEUROSCI.0873-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Lamb RF. Amino acid sensing mechanisms: an Achilles heel in cancer. FEBS J. 2012;279:2624–2631. doi: 10.1111/j.1742-4658.2012.08659.x. [DOI] [PubMed] [Google Scholar]
  15. Laviano A, Meguid MM, Inui A, Rossi-Fanelli F. Role of leucine in regulating food intake. Science. 2006;313:1236–1238. doi: 10.1126/science.313.5791.1236b. author reply. [DOI] [PubMed] [Google Scholar]
  16. L’Heureux-Bouron D, Tome D, Rampin O, Even PC, Larue-Achagiotis C, Fromentin G. Total subdiaphragmatic vagotomy does not suppress high protein diet-induced food intake depression in rats. J Nutr. 2003;133:2639–2642. doi: 10.1093/jn/133.8.2639. [DOI] [PubMed] [Google Scholar]
  17. L’Heureux-Bouron D, Tome D, Bensaid A, Morens C, Lacroix M, Huneau JF, Fromentin G. Preabsorptive factors are not the main determinants of intake depression induced by a high-protein diet in the rat. Physiol Behav. 2004;81:499–504. doi: 10.1016/j.physbeh.2004.02.029. [DOI] [PubMed] [Google Scholar]
  18. Lutz TA, Senn M, Althaus J, Del Prete E, Ehrensperger F, Scharrer E. Lesion of the area postrema/nucleus of the solitary tract (AP/NTS) attenuates the anorectic effects of amylin and calcitonin gene-related peptide (CGRP) in rats. Peptides. 1998;19:309–317. doi: 10.1016/s0196-9781(97)00292-1. [DOI] [PubMed] [Google Scholar]
  19. Morrison CD, Xi X, White CL, Ye J, Martin RJ. Amino acids inhibit Agrp gene expression via an mTOR-dependent mechanism. Am J Physiol Endocrinol Metab. 2007;293:E165–E171. doi: 10.1152/ajpendo.00675.2006.. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Morrison CD, Reed SD, Henagan TM. Homeostatic regulation of protein intake: in search of a mechanism. Am J Physiol Regul Integr Comp Physiol. 2012;302:R917–R928. doi: 10.1152/ajpregu.00609.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Morton GJ, Cummings DE, Baskin DG, Barsh GS, Schwartz MW. Central nervous system control of food intake and body weight. Nature. 2006;443:289–295. doi: 10.1038/nature05026. [DOI] [PubMed] [Google Scholar]
  22. Nassl AM, Rubio-Aliaga I, Sailer M, Daniel H. The intestinal peptide transporter PEPT1 is involved in food intake regulation in mice fed a high-protein diet. PLoS One. 2011;6:e26407. doi: 10.1371/journal.pone.0026407. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Ng LT, Anderson GH. Route of administration of tryptophan and tyrosine affects short-term food intake and plasma and brain amino acid concentrations in rats. J Nutr. 1992;122:283–293. doi: 10.1093/jn/122.2.283. [DOI] [PubMed] [Google Scholar]
  24. Niv Y, Abu-Avid S, Neumann G. Further applications of blood gas analysis to gastric acidity determination. Clin Chim Acta. 1993;215:9–19. doi: 10.1016/0009-8981(93)90244-x. [DOI] [PubMed] [Google Scholar]
  25. Oesch S, Ruegg C, Fischer B, Degen L, Beglinger C. Effect of gastric distension prior to eating on food intake and feelings of satiety in humans. Physiol Behav. 2006;87:903–910. doi: 10.1016/j.physbeh.2006.02.003. [DOI] [PubMed] [Google Scholar]
  26. Peters JC, Harper AE. Acute effects of dietary protein on food intake, tissue amino acids, and brain serotonin. Am J Physiol Regul Integr Comp Physiol. 1987;252:R902–R914. doi: 10.1152/ajpregu.1987.252.5.R902. [DOI] [PubMed] [Google Scholar]
  27. Pichon L, Potier M, Tome D, Mikogami T, Laplaize B, Martin-Rouas C, Fromentin G. High-protein diets containing different milk protein fractions differently influence energy intake and adiposity in the rat. Br J Nutr. 2008;99:739–748. doi: 10.1017/S0007114507831709. [DOI] [PubMed] [Google Scholar]
  28. Potier M, Darcel N, Tome D. Protein, amino acids and the control of food intake. Curr Opin Clin Nutr Metab Care. 2009;12:54–58. doi: 10.1097/MCO.0b013e32831b9e01. [DOI] [PubMed] [Google Scholar]
  29. Purpera MN, Shen L, Taghavi M, Munzberg H, Martin RJ, Hutson SM, Morrison CD. Impaired branched chain amino acid metabolism alters feeding behaviour and increases orexigenic neuropeptide expression in the hypothalamus. J Endocrinol. 2012;212:85–94. doi: 10.1530/JOE-11-0270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Raybould HE, Tache Y. Cholecystokinin inhibits gastric motility and emptying via a capsaicin-sensitive vagal pathway in rats. Am J Physiol Gastrointest Liver Physiol. 1988;255:G242–G246. doi: 10.1152/ajpgi.1988.255.2.G242. [DOI] [PubMed] [Google Scholar]
  31. Rune SJ. Comparison of the rates of gastric acid secretion in man after ingestion of food and after maximal stimulation with histamine. Gut. 1966;7:344–350. doi: 10.1136/gut.7.4.344. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Ruttimann EB, Arnold M, Hillebrand JJ, Geary N, Langhans W. Intrameal hepatic portal and intraperitoneal infusions of glucagon-like peptide-1 reduce spontaneous meal size in the rat via different mechanisms. Endocrinology. 2009;150:1174–1181. doi: 10.1210/en.2008-1221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Schwartz GJ. Integrative capacity of the caudal brainstem in the control of food intake. Philos Trans R Soc Lond B Biol Sci. 2006;361:1275–1280. doi: 10.1098/rstb.2006.1862. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Swanson LW. Brain Maps III. Structure of the rat brain/an atlas with printed and electronic templates for data, models and schematics. 3rd edn. Amsterdam: Elsevier Academic Press; 2004. [Google Scholar]
  35. Tome D, Schwarz J, Darcel N, Fromentin G. Protein, amino acids, vagus nerve signalling, and the brain. Am J Clin Nutr. 2009;90:838–843. doi: 10.3945/ajcn.2009.27462W. [DOI] [PubMed] [Google Scholar]
  36. Torii K, Niijima A. Effect of lysine on afferent activity of the hepatic branch of the vagus nerve in normal and L-lysine-deficient rats. Physiol Behav. 2001;72:685–690. doi: 10.1016/s0031-9384(01)00426-7. [DOI] [PubMed] [Google Scholar]
  37. Tsurugizawa T, Uematsu A, Nakamura E, Hasumura M, Hirota M, Kondoh T, Uneyama H, Torii K. Mechanisms of neural response to gastrointestinal nutritive stimuli: the gut-brain axis. Gastroenterology. 2009;137:262–273. doi: 10.1053/j.gastro.2009.02.057. [DOI] [PubMed] [Google Scholar]
  38. Watts AG, Khan AM, Sanchez-Watts G, Salter D, Neuner CM. Activation in neural networks controlling ingestive behaviors: What does it mean, and how do we map and measure it. Physiol Behav. 2006;89:501–510. doi: 10.1016/j.physbeh.2006.05.025. [DOI] [PubMed] [Google Scholar]
  39. Woods SC. The control of food intake: behavioral versus molecular perspectives. Cell Metab. 2009;9:489–498. doi: 10.1016/j.cmet.2009.04.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Zhang Y, Guo K, LeBlanc RE, Loh D, Schwartz GJ, Yu YH. Increasing dietary leucine intake reduces diet-induced obesity and improves glucose and cholesterol metabolism in mice via multimechanisms. Diabetes. 2007;56:1647–1654. doi: 10.2337/db07-0123. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

tjp0591-5611-SD1.docx (218.7KB, docx)

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES