Amino acids differ in their capacity to stimulate GLP-1 release from the perfused rat small intestine and stimulate secretion by different sensing mechanisms.

The aim of this study was to explore individual amino acid-stimulated GLP-1 responses and the underlying stimulatory mechanisms, as well as to identify the amino acid-sensing-receptors involved in amino acid-stimulated GLP-1 release. Experiments were primarily based on isolated perfused rat small intestines, which have intact epithelial polarization allowing discrimination between luminal and basolateral mechanisms as well as quantitative studies of intestinal absorption and hormone secretion. Expression analysis of amino acid sensors on isolated murine GLP-1 secreting L-cells was assessed by qPCR. We found that L-valine powerfully stimulated GLP-1 secretion but only from the luminal side (2.9-fold increase). When administered from the vascular side, L-arginine and the aromatic amino acids stimulated GLP-1 secretion equally (2.6-2.9 fold increases). Expression analysis revealed that Casr expression was enriched in murine GLP-1 secreting L-cells, whereas Gpr35, Gprc6a, Gpr142, Gpr93 (Lpar5) and the umami taste receptor subunits Tas1r3 and Tas1r1 were not. Consistently, activation of GPR35, GPR93, GPR142 and the umami taste receptor with specific agonists or allosteric modulators did not increase GLP-1 secretion (P>0.05 for all experiments), whereas vascular inhibition of CaSR reduced GLP-1 secretion in response to luminal infusion of mixed amino acids. In conclusion, amino acids differ in their capacity to stimulate GLP-1 secretion. Some amino acids stimulated secretion only from the intestinal lumen, while other amino acids exclusively stimulated secretion from the vascular side, indicating that amino acid-stimulated GLP-1 secretion involves both apical and basolateral (post-absorptive) sensing mechanisms. Sensing of absorbed amino acids involves CaSR activation as vascular inhibition of CaSR markedly diminished amino acid stimulated GLP-1 release.


INTRODUCTION
Ingestion of dietary proteins and amino acids has long been known to stimulate the secretion of the appetite-inhibiting and blood glucose lowering gut hormone, glucagon-like peptide 1 (GLP-1) (1-7). Ingested protein thereby contributes to postprandial metabolic and appetite control in animals and humans. Currently, GLP-1 mimetics are being used for the treatment of obesity and type 2 diabetes (8)(9)(10). An alternative approach would be to enhance the endogenous GLP-1 secretion (as seen after gastric bypass surgery) to mimic the beneficial effects of GLP-1, potentially with fewer side effects. As dietary proteins are believed to elicit the greatest appetite-suppressive effect of all macronutrients (11), possibly by stimulating the release of various anorectic hormones including peptide YY (PYY), cholecystokinin (CCK), and GLP-1 (6,7,(12)(13)(14)(15), and as a recent study has shown that a protein preload was able to improve the efficacy of the dipeptidyl peptidase-4 (DPP-4) inhibitor to increase incretin concentrations, slow gastric emptying, and reduce postprandial glycaemia in patients with type 2 diabetes (16), strategies based on the mechanisms underlying these effects could be valuable in the treatment of type 2 diabetes and obesity. In this context, a detailed knowledge of the mechanisms involved in the sensing of dietary peptides and amino acids by the enteroendocrine cells is essential. However, our knowledge of the mechanisms whereby the individual amino acids stimulate GLP-1 secretion is limited.
Sensing of amino acids by enteroendocrine cells is believed to occur by three modes of actions: 1) by electrogenic amino acid uptake by different amino acid transporters, 2) by intracellular metabolism, which may result in cell depolarization by closure of ATP-sensitive potassium channels, or 3) by activation of different nutrient receptors, including G-protein coupled receptors (GPCRs), which then initiate intracellular signaling cascades leading to release of calcium from intracellular stores or increased cyclic adenosine monophosphate (cAMP) production, both of which may result in hormone release (Fig. 1).
Some amino acids are known to be transported in a sodium-coupled manner, which might depolarize cell membranes, followed by voltage-dependent calcium entry (Fig. 1). In addition, several GPCRs including the calcium-sensing receptor (CaSR), GPR142, GPR93, GPR35, GPRC6A (GPCR, Class C, group 6, subtype A), and the umami taste receptor (Tas1R1/Tas1R3) have been demonstrated in various static in Figure 1. Sensing of amino acids by enteroendocrine L-cells. Sensing of amino acids by enteroendocrine L-cells is believed to occur by three modes of action: 1) through electrogenic amino acid and peptide uptake, which may result in cell depolarization leading to opening of voltage-gated calcium channels (Ca 2 þ V ) and hormone secretion (as indicated by red arrows), and 2) intracellular metabolism of amino acids and peptides may lead to closure of ATP-sensitive potassium channels (K ATP ), which depolarizes the cell membrane again leading to opening of voltage-gated calcium channels, calcium influx, and hormone release (indicated by green arrows), or 3) absorbed amino acids may stimulate nutrient receptors, including G-protein coupled receptors (GPCRs), which then initiate intracellular signaling cascades leading to release of calcium from intracellular stores or increased cyclic adenosine monophosphate (cAMP) production, both of which may result in hormone release (indicated by blue arrows). vitro models to bind and respond to dietary peptides and amino acids accompanied by release of gastrointestinal hormones like CCK, glucose-dependent insulinotropic polypeptide (GIP), GLP-1, and PYY (17)(18)(19)(20)(21)(22)(23)(24)(25).
Several amino acids have been demonstrated to be potent stimulators of GLP-1 release in vivo and in vitro (4,5,(26)(27)(28)(29)(30), however, detailed characterizations of relative efficacy of the individual amino acids are rare, and important aspects of the responsible mechanisms remain unknown, since neither studies in vivo nor conventional in vitro studies can discriminate between stimulation of GLP-1 secretion from the luminal or vascular side of the gut. In addition, further identification of the repertoire of receptors responsible for sensing of the individual amino acids is needed. The aim of this study was therefore to identify amino acid-sensing receptors involved in GLP-1 secretion, and to assess the relative stimulatory potential of the individual amino acids as well as to elucidate the responsible stimulatory mechanisms. We hypothesized that the underlying mechanisms would differ with some amino acids stimulating GLP-1 secretion from the luminal side and others from the vascular side of the intestine. We also speculated that the aromatic amino acids would be the most powerful secretagogues, as these have been shown to activate both CaSR and GPR142, which both stimulate secretion of GLP-1 in vivo and in vitro (18,25,31).

Ethical Considerations
Studies were conducted with permission from the Danish Animal Experiments Inspectorate (2018-15-0201-01397) and the local ethical committee (EMED, P18-336) in accordance with the guidelines of Danish legislation governing animal experimentation (1987) and the National Institute of Health.

Animals
Male Wistar rats ($220 g) were obtained from Janvier (Le Genest-Saint-Isle, France), housed two to four rats per cage, and kept on a 12:12-h light/dark cycle with ad libitum access to water and standard chow. Rats were allowed at least 1 wk of acclimatization before the experiments.
GLU-Venus mice were derived from an in-house breed, originally generated at University of Cambridge (32). They were housed 2-8 per cage with free access to standard rodent chow under a 12:12-h light/dark cycle.

Isolation and Perfusion of the Proximal Rat Small Intestine
Nonfasted rats ($220 g) were anesthetized with a subcutaneous injection of hypnorm/midazolam (0.0158 mg fentanyl citrate þ 0.5 mg fluanisone þ 0.25 mg midazolam/100 g). The proximal part of the small intestine was surgically isolated by ligation of the blood supply and removal of the colon and the distal half of the small intestine as described previously (23). A plastic tube was inserted into the lumen of the intestine and the intestine was gently flushed with heated (37 C) isotonic saline to remove luminal contents. A catheter was placed into the superior mesenteric artery and the intestine was perfused with heated (37 C) and oxygenated (95% O 2 and 5% CO 2 ) perfusion buffer (pH 7.4) at a flow rate of 7.5 mL/min using a single-pass perfusion system (UP100, Hugo Sachs Harvard Apparatus, Germany).
The venous effluent was collected every minute via a draining catheter inserted into vena portae. As soon as proper flow was apparent, the rats were euthanized by perforation of the diaphragm. The intestine was perfused for 25 min before initiation of the experimental protocol for stabilization. Each protocol started with a baseline period followed by addition of various test substances applied either into the intestinal lumen or intra-arterially through the tube inserted in the lumen or through the catheter in the upper mesenteric artery. Effluent samples were immediately placed on ice and stored at À20 C until analysis.
Total amino acid concentrations were measured using a colorimetric L-amino acid assay kit from Abcam (Cat. No. ab65347, Cambridge, UK) following provided instructions. The sensitivity of the assay is reported to be 40 μmol/L. The assay detects all L-amino acids except glycine.

Isolation of Murine L-Cells and Quantitative PCR Expression Analysis
L-cells were isolated from intestinal segments of 30 male C57BL/6JRj transgenic GLU-Venus mice expressing the fluorescent Venus protein driven by the proglucagon promoter (32) as previously described (34). Cells were sorted by fluorescence activated cell sorting (FACS) using a BD FACSAria II (BD Biosciences, Palo Alto, CA) yielding a GLU-Venus positive or GLU-Venus negative fraction of cells. The RNA was extracted using NucleoSpin RNA XS kit (Cat. No. 740902, Macherey-Nagel, Germany). The cDNA was generated using QuantiTect Whole Transcriptome kit (Cat. No. 207043, Qiagen, Germany). Expression was analyzed using customdesigned 384-well quantitative PCR plates from Lonza, Denmark. For the analysis of Gprc6a, Casr, Gpr35, and Gpr142, 10 mice were pooled in each run of quantitative PCR (qPCR) (n = 3). The sequences used for the receptors and housekeeping genes are published elsewhere (35). Expression of Gprc6a, Casr, Gpr35, and Gpr142 is shown as an expression relative to a mean of three housekeeping genes: 18S ribosomal RNA (Rn18s), tyrosine 3-monooxygenase (YWHAZ), and hypoxanthine phosphoribosyltransferase 1 (HPRT1). Relative expression of Lpar5, Tas1r1, and Tas1r3 was analyzed from another dataset published elsewhere (36). The relative copy number was calculated according to the formula: Relative copy number = (2 À (CT target À CT DNA )/NF) Â C, where CT target = the CT value of the receptor in the cDNA sample. CT DNA = the CT value of the receptor in a genomic DNA sample containing all assayed genes. NF is a GeNorm derived normalization factor using all genes with CT values below 35 in all samples. C is an arbitrary constant dependent on the DNA concentration, in this case C = 11,585, consequently CT values of 35 are on average equal to one relative copy. Undetectable targets were assigned a CT value of 40 (35).

Immunohistochemistry
CaSR and GLP-1 immunoreactivity was tested in archival paraffin-embedded specimens of human small intestine (jejunum, n = 3). The tissue blocks were sectioned on a microtome (section thickness 5 μm) and sections were placed on glass slides. The sections were dewaxed and subjected to antigen retrieval by boiling for 15 min in citrate buffer of pH 6. Next, sections were pretreated with 2% (w/v) bovine serum albumin (BSA) in phosphate buffered saline (PBS), all washing steps were carried out in PBS. The sections were then incubated overnight at 4 C with a mixture of the primary antibodies against GLP-1 and CaSR. Antibodies were inhouse rabbit GLP-1 antibody 2135 and mouse CaSR antibody (19347, Abcam, Cambridge, UK), both diluted 1:2,500 in PBS with 2% BSA. On day 2, slides were incubated with a mix of Alexa 488-labeled donkey anti-mouse antibody (1:200) and Alexa 568-labeled donkey anti-rabbit (1:200) for 1 h, washed, mounted with Dako fluorescence mounting medium (Agilent, Santa Clara, CA), and coverslipped. The slides were then examined using an Axioscope 2 plus microscope (Zeiss, Jena, Germany) and images were taken using a CoolSNAP camera (Photometrics, Tucson, AZ).

Calculations and Statistical Analysis
Hormone concentrations (pmol/L to fmol/mL) and amino acid concentrations (mmol/L) in the venous effluents are presented as means ± SE. As perfusion flow was constant throughout the experiments, the actual hormone secretion rate (fmol/min) can be calculated by multiplying with the flow rate (7.5 mL/min).
To test for statistical significance, mean values from the test period (based on 10 consecutive minutes) were compared with mean values from the baseline period (10 min before administration of test stimulant) using Student's t test or one-way ANOVA followed by Dunnett multiple comparisons test, as indicated in figure legends. Statistical testing and construction of graphs was done in GraphPad Prism version 8.0 (GraphPad Software, La Jolla, CA). Adobe CC software (San Francisco, CA) was used for illustrations. P values <0.05 were considered statistically significant.

Individual Amino Acid-Stimulated GLP-1 Secretion
To assess potential differences in amino acid-stimulated GLP-1 secretion, we initially administered individually all amino acids except L-tyrosine into the lumen of the perfused rat's small intestine at a concentration of 50 mmol/L (L-tyrosine was not soluble at the desired concentration), and measured GLP-1 concentrations in the venous effluent (each graph can be seen in Supplemental Fig. S1; see https://doi. org/10.6084/m9.figshare.13607333, n = 4 for all groups). For these experiments, bombesin (BBS) was used as a positive control for GLP-1 secretion and was administered by the end of the experiments (Supplemental Fig. S1).
We next infused the same amino acids intra-arterially (at 20 mmol/L) (each experiment can be seen in Supplemental Fig. S1, n = 4 for all groups). In these experiments, L-arginine served as a positive control for GLP-1 secretion, as vascular infused L-arginine powerfully stimulated GLP-1 secretion in pilot studies (data not shown, n = 6). Thus, the L-cell amino acid responses thereby intentionally could be compared with that induced by L-arginine. Compared with intraluminal administration, more amino acids increased GLP-1 secretion when infused intra-arterially, and the GLP-1 responses observed were in general greater, suggesting that amino acids predominantly stimulate GLP-1 secretion by basolateral activation of the intestinal L-cells. Of particular note, vascular but not luminal infusion of L-arginine and L-tryptophan resulted in 2.9-and 2.7-fold increases in GLP-1 secretion (P < 0.05, Fig. 2B1, n = 4), indicating that these amino acids stimulate GLP-1 secretion mainly through postabsorptive mechanisms. This stands in contrast to L-valine (and L-glutamine), which only stimulated GLP-1 secretion when given intraluminally (Fig. 2B1,  n = 4). Thus, amino acids stimulate secretion both by activation of luminal sensors/uptake mechanisms and by postabsorptive mechanisms (i.e., by basolateral uptake or receptor activation).
L-Phenylalanine stimulated GLP-1 secretion from both sides of the intestine (P < 0.05, Fig. 2, A and B), but the vascular response was considerably higher (2.6-vs. 1.9-fold higher than baseline, P = 0.0005 vs. P = 0.0145) (Fig. 2, A and B, n = 4). As the luminally induced GLP-1 response may reflect a postabsorptive "intra-arterial" response, we decided to evaluate the absorption rate of luminal infused amino acids.
Luminally infused L-valine is efficiently absorbed in contrast to L-arginine and L-tryptophan.
To validate whether these amino acids were all detected by the L-amino acid assay to a similar extent, we tested the recoveries of the individual amino acids used in the perfusion experiments. As in the perfusion experiments, amino acids were prepared in perfusion buffer. At a concentration range between 0.375 and 3 mmol/L, the recovery of all amino acids was measured (Supplemental Fig. S2). At a concentration of 3 mmol/L, the majority of amino acids could be detected with a recovery close to 100%, however, some amino acids could not be detected at this concentration (serine, threonine, glutamate, aspartate, proline, cysteine, and lysine). At a concentration range lower than 3 mmol/L, the recoveries were far from 100% except for histidine (Supplemental Fig. S2).
Inhibition of L-type amino acid transporter 2 does not affect the absorption of L-valine or L-valine-mediated GLP-1 secretion.
As inhibiting amino acid transport was unsuccessful in this experimental setup, we next continued investigating the amino acid sensors involved in GLP-1 secretion.

Expression of Amino Acid Sensors in Murine and Human Enteroendocrine L-Cells
To examine which receptors are involved in sensing of amino acids by the intestinal L-cells, we investigated the expression of selected amino acid sensors in enteroendocrine L-cells. Intestinal cells obtained from duodenum, jejunum, and ileum of transgenic GLU-Venus mice were sorted using fluorescence-activated cell sorting (FACS) giving rise to GLU-Venus positive (GLP-1 producing L-cells) and GLU-Venus negative (intestinal cells not expressing the proglucagon promotor) cell populations. In all intestinal segments (duodenum, jejunum, and ileum), the expression of Casr was enriched in GLP-1 secreting L-cells compared with neighboring intestinal cells (GLU-Venus negative cells) (Fig. 3, A-C), supporting that CaSR may be involved in stimulation of GLP-1 secretion. Gpr35 and Gpr142 were likewise expressed in GLP-1 secreting L-cells (GLU-Venus positive cells), however, the expression was not enriched compared with neighboring cells in neither of the intestinal segments examined (Fig. 3, A-C). Gprc6a was not detected in either GLU-Venus negative or GLU-Venus positive cells, suggesting that this receptor is not involved in sensing of amino acids by intestinal cells in duodenum, jejunum, and ileum of mice. Gpr93 (Lpar5) and umami taste receptor subunit, Tas1r3, were both expressed in GLP-1 secreting L-cells (GLU-Venus positive cells), however, the expression was not enriched compared with neighboring GLU-Venus negative cells in the upper part ($10 cm) of the proximal mouse small intestine (32) (Fig. 3D). The expression of the other subunit of the umami taste receptor, Tas1r1, was very low both in GLP-1 secreting L-cells as well as in the neighboring intestinal cells (GLU-Venus negative cells) (Fig. 3D).
To examine whether CaSR was likewise expressed on human endocrine L-cells, human jejunal tissue was immunohistochemically co-stained for CaSR and GLP-1. CaSR and GLP-1 frequently colocalized with few GLP-1 positive cells without CaSR and many CaSR positive cells without GLP-1 (Fig. 3E, n = 3), suggesting that CaSR is also involved in other functions in the gastrointestinal tract.

Amino Acid Sensors Involved in the Stimulation of GLP-1 Secretion
To further investigate which receptors are involved in amino acid-stimulated GLP-1 secretion, as well as their localization (apical or basolateral) on the intestinal L-cells, we stimulated the perfused intestine with specific candidate receptor agonists both intra-arterially and intraluminally.

Tas1R1/Tas1R3
However, further studies with an inhibitor of Tas1R1, like Gurmarin, are needed to further clarify the involvement of Tas1R1/Tas1R3 in L-glutamic acid-stimulated GLP-1 secretion.
Inhibition of calcium-sensing receptor with NPS2143 decreases the GLP-1 response to mixed amino acids but not to L-arginine.
Opening of ATP-sensitive potassium channels does not affect amino acid-induced GLP-1 secretion.
As a large amount of absorbed amino acids will be used by the intestinal cells for metabolism, a mechanism of amino acid sensing could also involve increased intracellular ATP/ ADP ratio leading to closure of ATP-sensitive potassiumchannels (K ATP -channels) and subsequently membrane depolarization (Fig. 1). To examine this, we administered the K ATP -channel opener, diaxozide (250 μmol/L), while stimulating with luminal infused amino acids (Vamin; 51 mg/mL) (Fig. 4G). Opening of K ATP -channels neither appeared to affect the basal GLP-1 secretion nor the GLP-1 response to mixed amino acids (baseline-subtracted values; Vamin 2: 11.2 ± 2.6 vs. Vamin 2 þ diaxozide: 11.7 ± 1.3 pmol/L, P = 0.9161, n = 6) (Fig. 4G), suggesting that amino acids do not stimulate GLP-1 secretion through intracellular metabolism followed by closure of K ATP -channels.
Inhibition of phospholipase C decreases the GLP-1 response to mixed amino acids.
To investigate further the intracellular pathway involved in amino acid-induced GLP-1 secretion, we infused a phospholipase C (PLC) inhibitor (U73122; 10 μmol/L), while stimulating with a mixture of luminal infused amino acids (Vamin). Inhibition of PLC decreased amino acid induced GLP-1 secretion significantly when subtracting baseline secretion (baseline-subtracted values; Vamin 2: 11.2 ± 2.6 vs. Vamin 2 þ U73122: 2.2 ± 2.7 pmol/L, P < 0.05), n = 6) (Fig. 4H), suggesting that activation of PLC leading to formation of diacylglycerol and inositol triphosphate is involved in the mechanisms of amino acid stimulated GLP-1 release. This furthermore supports the involvement of CaSR in amino acid stimulated GLP-1 secretion, as CaSR is mainly found to be Ga q -coupled, thereby leading to activation of PLC (46,47).

DISCUSSION
Approximately 95% of dietary protein is absorbed in the small intestine (48). Of these, a considerable amount is used by the intestinal cells for metabolism (e.g., oxidation and protein synthesis), while the remainder will be released into the portal system (49,50). As amino acids that are not metabolized by the intestinal cells will be transported to the interstitium to diffuse into the circulation, the local concentration of amino acids at the basolateral surface of the intestinal cells may reach high levels after a protein-rich meal.
In this study, the absorption of amino acids in the proximal rat small intestine was measured when stimulating with luminally infused amino acids. When administering individual amino acids intraluminally at a concentration of 50 mmol/L (corresponding to the total intraluminal concentration in the upper small intestine after a high protein meal) (51)(52)(53), the amino acid concentration measured in the venous effluent increased from 0.5 mmol/L to $2 mmol/L (except for L-tryptophan and L-arginine). In humans, plasma total amino acid concentration at fasting is around 2 mmol/L and may increase up to 5 mmol/L after meal intake (51,54).
The finding that L-tryptophan and L-arginine were not efficiently absorbed could explain why these amino acids did not stimulate GLP-1 secretion when infused intraluminally ( Fig. 2A), although both powerfully stimulated GLP-1 release from the vascular side of the gut (Fig. 2B). The underlying reason of the differences in absorption rates of the individual amino acids warrants further investigation, but it is likely to be influenced by differences in absorption capacity of different transporters involved and/or differences in transporter expression along the small intestine as well as detection issues. However, in a physiological setting, regional differences in transporter abundance are presumably of less importance for the total rate of amino acid absorption as uptake of di-and tripeptides through peptide transporter 1 (PepT1) mainly is responsible for the amino acid absorption, since PepT1 transports several amino acid monomers for each turnover of the transporter (55). The dipeptides and tripeptides are thought to be cleaved into single amino acids intracellularly, and will be transported as such across the basolateral membrane (56).
The mechanisms of L-arginine-and L-tryptophan-stimulated GLP-1 release most likely involve postabsorptive sensing mechanisms, as vascularly infused L-arginine and Ltryptophan powerfully stimulated GLP-1 release. The sensing mechanisms of L-arginine stimulated GLP-1 release remains to be established, but is unlikely to involve CaSR activation as inhibition of CaSR had no effect on vascular L-argininestimulated GLP-1 secretion.
The fact that L-phenylalanine both stimulated GLP-1 secretion when infused into the intestinal lumen and when infused through the vascular supply indicate that L-phenylalanine is a powerful stimulator of GLP-1 release. However, the luminally induced response could also reflect a postabsorptive basolateral response, as L-phenylalanine was efficiently absorbed (reaching a vascular concentration of 1 mmol/L). Furthermore, since L-phenylalanine is reported to activate CaSR (18,31,57), the stimulatory mechanism possibly involves postabsorptive CaSR-mediated sensing.
An important finding in this study was that L-valine is a powerful stimulator of GLP-1 release when infused into the lumen but not when administered intra-arterially. This suggests that an apical sensing mechanism, possibly coupled to sodium uptake, is involved. However, we were not able to identify the specific mechanism as blocking L-type amino acid transporters (SLC7A5, SLC7A8, SLC43A1, and SLC43A2) with BCH did not affect L-valine absorption, highlighting the complexity of the amino acid transporter systems with several transporters capable of transporting the same amino acids (perhaps some unknown as well).
Limitations of this study include that we were unable to include L-tyrosine in our analysis, since it was insoluble in isotonic saline and in our perfusion buffer at the desired concentrations. Furthermore, we did not test whether specific activation of GPRC6A, which preferentially binds the basic amino acids L-arginine and L-ornithine (58,59), affected GLP-1 secretion from the perfused rat intestine like previously demonstrated in the murine colonic GLUTag cell line (24). However, as recent findings in mice in vivo questions the involvement of GPRC6A in GLP-1 secretion (27), and as Gprc6a was not expressed in murine intestinal cells (either in GLU-Venus positive or GLU-Venus negative cells), GPRC6A is most likely not involved in amino acid stimulated GLP-1 release. Moreover, the antagonists currently available for GPRC6A are the CaSR positive allosteric modulator, Calindol, and the CaSR negative allosteric modulator, NPS2143, which both have a 30-fold higher potency of binding to CaSR than GPRC6A (60,61), making it difficult to differentiate between GPRC6A-and CaSR-mediated responses.
The expression of Gpr142, Gpr35, Gpr93 (Lpar5), Tas1R1, and Tas1R3 was not enriched in murine GLP-1 secreting Lcells (GLU-Venus positive cells) compared with neighboring intestinal cells (GLU-Venus negative cells) in line with the finding that activation of these receptors with specific agonists or allosteric modulators did not increase GLP-1 secretion from the perfused rat small intestine. However, the involvement of GPR142, GPR35, GPR93, and Tas1R1/Tas1R3 in secretion of other gut hormones, like GIP and CCK, deserves further investigation, as previous studies have demonstrated an important role e.g., for GPR142 in stimulation of GIP secretion in mice (25) and GPR93 in stimulation of CCK secretion from murine intestinal STC-1 cells (19).

CONCLUSIONS
Amino acids differ markedly in their capacity to stimulate GLP-1 release and the mechanism of stimulation differs between the individual amino acids. Some amino acids only stimulated GLP-1 secretion when infused through the vascular supply (L-arginine and L-tryptophan), whereas others only stimulated secretion when infused intraluminally (L-valine and L-glutamine), suggesting that amino acids and therefore protein meals stimulate GLP-1 secretion both through apical/ absorptive and through postabsorptive mechanisms (by activation of basolateral CaSR and perhaps unidentified receptors). Amino acid sensing by GPR35, GPR93, GPR142, and Tas1R1/Tas1R3 does not seem to be involved in GLP-1 release (at least in the perfused proximal rat small intestine, which is the relevant part of the intestine to study as digested protein is rapidly absorbed before reaching the more distal part of the small intestine). Furthermore, closure of ATP-sensitive potassium channels as a result of intracellular amino acid metabolism does not seem to be involved in amino acid-stimulated GLP-1 release. Rather, activation of PLC may be involved in the mechanisms of amino acid-stimulated GLP-1 release as inhibition of PLC decreased mixed amino acid-induced GLP-1 release.
Finally, our findings suggest that dietary supplementation with L-valine may serve as a potential strategy to increase GLP-1 secretion.