Glucagon-like peptide-2 (GLP-2) is a nutrient-responsive neuropeptide that exerts diverse actions in the gastrointestinal tract, including enhancing mucosal cell survival and proliferation. GLP-2 stimulates mucosal growth in vivo with an increased rate of protein synthesis. However, it was unclear whether GLP-2 can directly stimulate protein synthesis. The objective was to test critically whether GLP-2 receptor (GLP-2R) activation directly stimulates protein synthesis through a PI 3-kinase-dependent Akt-mTOR signaling pathway. HEK 293 cells (transfected with human GLP-2R cDNA) were treated with human GLP-2 with/without pretreatment of PI 3-kinase inhibitor (LY-294002) or mTOR inhibitor (rapamycin). Results show that 1) GLP-2 specifically bound to GLP-2R overexpressed in the HEK cells with Ka = 0.22 nM and Bmax = 321 fmol/μg protein; 2) GLP-2-stimulated protein synthesis was dependent on the amount of GLP-2R cDNA and the dosage of GLP-2 and reached the plateau among 0.2–2 nM GLP-2; 3) GLP-2-stimulated protein synthesis was abolished by the PI 3-kinase inhibitor and mTOR inhibitor; and 4) GLP-2-mediated stimulation of phosphorylation on Akt and mTOR was dependent on the amount of GLP-2R cDNA transfected and the dosage of GLP-2. In addition, GLP-2-mediated action and signaling in regulation of protein synthesis were confirmed in mouse hippocampal neurons (expressing native GLP-2R). GLP-2 directly stimulated protein synthesis of primary cultured neurons in dosage-dependent, PI 3-kinase-dependent, and rapamycin-sensitive manners, which linked with activation of Akt-mTOR signaling pathway as well. We conclude that GLP-2R activation directly stimulates protein synthesis by activating the PI 3-kinase-dependent Akt-mTOR signaling pathway. GLP-2-stimulated protein synthesis may be physiologically relevant to maintaining neuronal long-term potentiation and providing secondary mediators (namely neuropeptides or growth factors).
- phosphatidylinositol 3-kinase
- mammalian target of rapamycin
glucagon-like peptide-2 (GLP-2) is a 33-amino acid peptide produced by intestinal endocrine L-cells and cerebral neurons. It exerts diverse actions involved mainly in control of gastrointestinal growth and function, especially mucosal epithelial cell survival, crypt cell proliferation, glucose and protein anabolism, and microcirculation (8, 12, 15). For example, under pathological condition, GLP-2 prevents gut atrophy induced from cancer chemotherapy or total parenteral nutrition-induced intestinal mucosal atrophy (6, 9) and improves inflammatory bowel disease or short bowel syndrome (2, 16, 19, 30, 32), whereas under physiological condition GLP-2 mediates refeeding-induced adaptation in the mouse gut (4). Recently, using an in vivo stable isotope tracer balance approach, we demonstrated that GLP-2 acutely stimulated intestinal protein anabolism via increased amino acid uptake and protein synthesis without effect on protein breakdown (15). However, GLP-2 receptor (GLP-2R) is localized to enteric neurons, enteroendocrine cells, and subepithelial myofibroblasts, but not enterocytes (5, 14, 22, 23, 33), suggesting that GLP-2-induced trophic actions on the epithelium are mediated indirectly. Our recent study indicates that GLP-2 directly modulates L-type calcium channel activity in primary neurons (28), which might be attributed to GLP-2-triggered release of neurotransmitters, hormones, or local growth factors from GLP-2R-expressing cells. To fulfill this function, however, de novo synthesis of protein must be initiated and sustained somehow. Thus, we hypothesized that GLP-2 might directly stimulate protein synthesis.
Akt is a key mediator in regulation of protein metabolism, cell survival and proliferation, and cell cycle progression (18). Animal studies in vivo have shown that GLP-2-mediated stimulation of gut growth is associated with activation of Akt signaling (4, 7, 8). However, it is unclear whether GLP-2R activation can directly activate the Akt signaling. It is well documented that growth factors and hormones (e.g., insulin) stimulate PI 3-kinase activity and activated its downstream effector (Akt). So we wanted to critically test whether GLP-2 can activate the Akt signaling via the PI 3-kinase pathway.
The mammalian target of rapamycin (mTOR) kinase plays a key role in control of protein synthesis and ribosome biogenesis (25, 26). mTOR can respond not only to growth stimuli but also to nutrient availability. The activity of mTOR can be fine-tuned by the PI 3-kinase downstream mediators and amino acids (e.g., leucine). The mTOR downstream targets [e.g., p70 S6 kinase (p70S6K1) and eukaryotic initiation factor 4E-binding protein-1 (4E-BP1)] are directly involved in the regulation of protein synthesis and cell proliferation by modulating initiation and elongation of translation and biogenesis of ribosomes. Note that increased rates of protein synthesis are a key feature of cell proliferation and growth. Studies in vivo indicate that mTOR signaling is activated in GLP-2-stimulated gut growth (7, 8, 11). Again, it is not clear whether GLP-2R-induced cellular action and metabolism are directly mediated through the mTOR signaling pathway. Therefore, our objective was to critically test whether GLP-2R activation directly stimulates de novo protein synthesis by activating the PI 3-kinase-Akt-dependent mTOR signaling pathway.
All experiments were approved by the Institutional Animal Care and Use Committee of Baylor College of Medicine. C57BL/6J mice (Jackson Laboratory, Bar Harbor, ME) were fed ad libitum (standard rodent diet no. 2920, Harlan Teklad) and given free access to water.
Primary culture of hippocampal neurons.
Neurons were obtained from 1-day-old neonatal mice and cultured as described previously (28). In brief, the hippocampus was dissected in F-12 medium, digested with L15-NeurobasalA medium containing 0.1% trypsin and 0.1% DNAase at 37°C for 30 min, triturated with a sterile pipette tip on ice, and centrifuged at 1,000 rpm at 4°C for 5 min. Dissociated cells were resuspended, seeded onto slides precoated with poly-l-lysine for 30 min, and cultured for 6 days in NeurobasalA medium supplemented with B27, 10% FBS, 10 ng/ml fibroblast growth factor-β (FGFβ), 0.5 mM glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 100 μg/ml Primocin (Amaxa). Cytosine-β-d-arabinofuranoside (10 μM; Sigma) was added 48 h postplating to inhibit proliferation of glial cells in the primary culture.
Hippocampal neurons cultured on glass coverslips on day 6 in vitro were exposed to 20 nM GLP-2 for 30 min to localize phosphorylation of Akt and mTOR as described previously (28). In brief, cells on coverslips then were fixed in 4% paraformaldehyde in PBS for 20 min, permeabilized in PBS containing 0.1% Triton X-100, and blocked in 10% normal donkey serum for 1 h. Cells were incubated with primary antibodies [guinea pig PGP9.5, Millipore Intl.; rabbit phospho-Akt (Ser473) and phospho-mTOR (Ser2448); Cell Signaling Technology, Danvers, MA]. After washing, cells on coverslips were incubated with secondary donkey anti-rabbit IgG conjugated with 488 and donkey anti-guinea pig IgG conjugated with Cy3 (Jackson ImmunoResearch Laboratories, West Grove, PA) with TOPRO-3 for nuclear counterstaining. Images were captured using a confocal laser-scanning microscope (Zeiss LSM 510).
Cell culture and transfection.
Human embryonic kidney (HEK) 293 cells were maintained in Dulbecco's modified Eagle's medium (high glucose) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin at 37°C, 5% CO2. Cells were transiently transfected with a full-length coding sequence of human GLP-2R at 60∼70% confluence for 24 h using FuGENE 6 reagent (Roche, Indianapolis, IN). Each transfection was carried out in triplicate and performed independently three times.
Intact cell binding assay.
To determine GLP-2-specific binding using a modified protocol (28), we incubated HEK 293 cells (after transient transfection with human GLP-2R cDNA for 24–36 h) in 24-well plates of 0.2 ml each with 0–0.5 nM 125I-labeled human GLP-2 (1–33; Phoenix Pharmaceuticals, Burlingame, CA) with or without 0.5–10 μM unlabeled human GLP-2 (1–33; American Peptide, Sunnyvale, CA) at 37°C for 2 h. The cells were washed in cold 50 mM Tris buffer (pH 7.4) and centrifuged at 18,000 rpm for 15 min at 4°C. Bound tracer in pellets was quantified by a gamma counter, and protein mass in pellets was estimated by the BCA protein assay (Pierce, Rockford, IL). All assays were performed in triplicate, and experiments were repeated three times. The specific binding was calculated by a difference in DPM between total binding and nonspecific binding (in presence of unlabeled GLP-2). The data for the GLP-2 specific binding were simulated using a one-site binding hyperbola nonlinear regression model (P < 0.01, R2 = 0.99; SigmaPlot, San Jose, CA), and the binding affinity was estimated by a disassociation constant (Kd).
Protein synthesis analysis.
GLP-2-stimulated protein synthesis was measured using [3H]phenylalanine (Phe) incorporation assay (29). HEK 293 cells (∼60% confluence) were transiently transfected with full-length cDNA of human GLP-2R for 24 h using FuGENE 6 in a six-well culture plate. Human GLP-2R was cloned in the laboratory previously (14), and pcDNA 3.1 was the vector control for transfection studies. After being starved for 4 h in serum-free medium, transfected HEK 293 cells and hippocampal neurons cultured on day 6 in vitro were treated with GLP-2 (0, 0.2, 2, 20, and 200 nM) for 4 h plus a prior administration of PI 3-kinase inhibitor (LY-294002 at 50 μM) or mTOR inhibitor (rapamycin at 50 μM) for 45 min. Or, after a different amount of GLP-2R cDNA was transfected, HEK 293 cells were treated with GLP-2 at 20 nM. Human serum album (protein carrier) and insulin (100 nM) were used as negative and positive controls, respectively, in GLP-2 stimulation studies. Each treatment was carried out in triplicate and performed independently three times.
The stimulated cells were pulsed with 2 μCi/ml [3H]Phe (Amersham Bioscience, Piscataway, NJ) for 4 h before harvest. After washing with ice-cold PBS, the cells were scraped into 0.5 ml of PBS, lysed, precipitated in 0.5 ml of 20% trichloroacetic acid for 30 min on ice, and centrifuged at 10,000 g for 15 min at 4°C. The pellets (precipitated proteins) were washed twice with 10% trichloroacetic acid and solubilized in 1 ml of 0.3 N NaOH for 1 h. An aliquot (400 μl) was taken to determine the incorporated radioactivity by liquid scintillation counter (Beckman LS 3801, Fullerton, CA), whereas aliquots of 100 μl were used for protein content by a BCA protein assay kit (Pierce). Incorporation rate of [3H]Phe into total protein was expressed as disintegrations per minute per microgram protein mass and considered as protein synthesis.
Cell signaling studies.
After reaching ∼60% confluence, the cells were starved overnight in serum-free medium but containing insulin (1.72 μM). The next day, the cells were pretreated with or without PI 3-kinase inhibitor LY-294002 or mTOR inhibitor rapamycin at 50 μM for 45 min and then treated with human serum album (as negative control) or GLP-2 at 1 or 20 nM for 15 or 30 min. (GLP-2 was dissolved in 0.1% human serum albumin in saline; LY-294002 was dissolved in ethanol; rapamycin was dissolved in water.) pcDNA 3.1 was used as vector control in transfection studies. In addition, hippocampal neurons cultured on day 6 in vitro were starved in serum-free medium for 4 h and then treated with GLP-2 (0, 2, 20, and 200 nM) for 30 min. Each treatment was carried out in triplicate and performed independently three times.
After treatment, HEK 293 cells and hippocampal neurons were harvested on ice in radioimmunoprecipitation assay buffer (50 mmol/l Tris·HCl at pH 7.4, 1% NP-40, 0.25% Na-deoxycholate, 150 mmol/l NaCl, 1 mmol/l EDTA, 1 mmol/l phenylmethylsulfonyl fluoride, 1 μg/ml aprotinin, 1 μg/ml leupeptin, 1 μg/ml pepstatin, 1 mmol/l sodium orthovanadate, 1 mmol/l sodium fluoride) and centrifuged at 10,000 g for 15 min at 4°C. Protein concentration was determined by BCA protein assay (Pierce) using bovine serum albumin as a standard. Samples were boiled at 100°C for 10 min in 2× sample buffer. An equal amount of protein (100 μg per sample) was loaded and electrophoresed in running buffer on a 7.5–12% Tris-glycine sodium dodecyl sulfate-polyacrylamide gel. After the sodium dodecyl sulfate-polyacrylamide gel electrophoresis, proteins were transferred to a nitrocellulose membrane. After blocking, the membrane was incubated with primary antibodies. The membranes were blocked and probed with an appropriate primary antibody (Cell Signaling Technology, Danvers, MA): phospho-Akt (Ser473, 1:1,000; Thr308, 1:1,000), phospho-mTOR (Ser2448, 1:750), phospho-p70S6K1 (Thr389, 1:500), phospho-4E-BP1 (Thr37/46, 1:500), β-actin (1:2,000), total Akt (1:1,000), total mTOR (1:1,000), total p70S6K1 (1:500), and total 4E-BP1 (1:500). After washing, the membrane was incubated with anti-rabbit IgG horseradish peroxidase-conjugated secondary antibody (1:3,000; Bio-Rad Laboratories, Hercules, CA) and reacted with ECL-Plus chemiluminescent detection horseradish peroxidase reagents (Amersham Biosciences, Piscataway, NJ). Western blotting images were scanned and analyzed on a Storm 860 PhosphorImager (GE Healthcare, Fairfield, CT). To quantify site-specific protein phosphorylation of interest, we first normalized protein abundance in densitometry with a loading control (β-actin) and then calculated a densitometry ratio of phosphorylated protein to the total protein and finally expressed the ratio as a percentage of the control group.
Independent experiments (n = 3) were performed for each treatment. Data were analyzed by ANOVA (SAS v. 9.1; SAS Institute, Cary, NC). General linear model was performed with the SAS procedure. Data are expressed as means ± SE. A P value < 0.05 or 0.01 was considered statistically significance.
GLP-2R is functionally expressed in transfected HEK 293 cells.
Since cell lines expressing endogenous GLP-2R have not been identified, we chose to transiently express human GLP-2R in HEK 293 cells as an in vitro model for molecular analysis of GLP-2 action and signaling. Our laboratory has previously cloned a functional, full-length coding sequence of human GLP-2R (14) and overexpressed it transiently in HEK 293 cells. To characterize if transfected HEK 293 cells functionally and specifically responded to GLP-2 (a native ligand), we determined the GLP-2 binding curve by using an intact cell binding assay. Transfected HEK 293 cells showed a saturable and specific binding to the 125I-labeled human GLP-2 (Fig. 1). By regression analysis, the specific binding at the top plateau, i.e., the maximal number of binding sites (Bmax) was estimated at 321 fmol/μg protein and the dissociation constant (Kd) at 0.22 nM. This saturation binding indicated that the affinity of human GLP-2 to GLP-2R expressed on HEK 293 cells was high and similar to high-affinity binding of 125I-Tyr34 human GLP-2 (1–34) in COS cells stably transfected with rat GLP-2R cDNA (Kd = 0.57 nM) (21) or 125I-human GLP-2 (1–33) in the mouse hippocampal neurons (Kd = 0.48 nM) (28). Therefore, our in vitro cell culture model could be used for functional analysis of GLP-2 action.
GLP-2-mediated stimulation of protein synthesis depends on GLP-2R cDNA amount and GLP-2 dosage.
We wanted to critically test whether GLP-2 directly and specifically stimulated de novo synthesis of protein in transfected HEK 293 cells. First, we determined whether GLP-2 (at 20 nM)-stimulated protein synthesis was dependent on the amount of GLP-2R cDNA transfected. Figure 2A shows that protein synthesis (indicated by the [3H]Phe incorporation rate) increased quadratically with increasing amounts of GLP-2R cDNA transfected in HEK 293 cells [compared with that in the control of pcDNA3.1 (2 μg)-transfected HEK 293 cells] and reached the peak among 0.5–1.0 μg of human GLP-2R cDNA, approximately equivalent to copy numbers of 2.5–5.0 × 1011/105 HEK 293 cells. Interestingly, if overtransfected with a higher amount of the GLP-2R cDNA, the [3H]Phe incorporation rate decreased, and this might be attributed to an artifact, nonspecific off-target, inhibitory effect of GLP-2R cDNA transfected too much. Data suggested that GLP-2-stimulated protein synthesis depended on the amount of GLP-2R cDNA transfected. This characteristic (of GLP-2-stimulated protein synthesis depending on the receptor copy number per se) not only determined the optimal amount (1.0 μg) of GLP-2R cDNA for transient transfection but also probably pointed to the physiological relevance of GLP-2-mediated protein synthesis (see discussion).
Then, we wanted to determine whether GLP-2-mediated stimulation of protein synthesis was ligand dosage dependent. Under the optimal amount (1.0 μg) of GLP-2R cDNA-transfected HEK 293 cells, the [3H]Phe incorporation rate was determined at a range of GLP-2 concentrations. It is shown in Fig. 2B that GLP-2-stimulated de novo synthesis of protein was dependent on the concentration of GLP-2; it increased significantly within 0.2–20 nM and reached the plateau among 0.2–2 nM. Data suggested again that GLP-2 stimulated de novo synthesis of protein in a dosage-dependent manner.
GLP-2-mediated stimulation of protein synthesis is abolished under inhibition of PI 3-kinase or mTOR kinase.
Previous in vivo studies have shown that chronic GLP-2 stimulated intestinal protein synthesis with increased phosphorylation of Akt and mTOR. Thus, we wanted to critically test whether GLP-2-mediated stimulation of protein synthesis was acutely and directly modulated by the PI 3-kinase-dependent Akt-mTOR signaling. It is shown in Fig. 2B that GLP-2-mediated stimulation of protein synthesis (i.e., increase in [3H]Phe incorporation rate) was abolished when pretreated with 50 μM LY-294002 (a PI 3-kinase inhibitor). Moreover, the basal rate of [3H]Phe incorporation (in the absence of GLP-2) was decreased (by 58%) in the presence of LY-294002, indicating that activation of PI 3-kinase signaling is required for protein synthesis in our in vitro cultured cells. Furthermore, the GLP-2-mediated increase in the rate of [3H]Phe incorporation was blunted when pretreated with 50 μM rapamycin (a mTOR inhibitor) (Fig. 2C). Interestingly, the basal rate of [3H]Phe incorporation was decreased (by 15%) in the presence of rapamycin, suggesting that activation of mTOR signaling is also required for protein synthesis in our cultured cells. These data suggested that GLP-2 acutely and directly stimulated protein synthesis, and this was PI 3-kinase dependent and mTOR sensitive.
GLP-2 activates PI 3-kinase-dependent akt signaling.
Note that GLP-2-stimulated protein synthesis was assessed by 4-h incorporation, which could not be discriminated from enhanced cell proliferation if any. Thus, we wanted to define if GLP-2 acutely (within minutes) activated the PI 3-kinase-dependent signaling. It is well documented that PI 3-kinase activity can be indicated by phosphorylation of Akt on Ser473 and Thr308 (two key stimulatory sites).
Representative images of Western blotting and quantitative results of three experimental units each are shown in following figures. The abundance ratio of p-Akt Ser473 to Akt (i.e., phosphorylation) increased in the presence of 20 nM GLP-2 for 20 min with increasing amount of GLP-2R cDNA transfected in HEK 293 cells (Fig. 3A) and reached the plateau among 0.5–2.0 μg of the construct that was required for the maximal rate of protein synthesis (as shown above). [Note that multiple bands of Akt (i.e., Akt1) might be attributed to phosphorylated and unphosphorylated forms of Akt1 in the present study]. Importantly, GLP-2-acutely stimulation of phosphorylation of Akt on Ser473 was dependent on the amount of GLP-2R cDNA transfected. Thus, the optimal amount (1.0 μg) of GLP-2R cDNA was used for transient transfection in the present study. Moreover, GLP-2-stimulated phosphorylation of Akt on Ser473 and Thr308 was ligand dosage dependent (Fig. 3B) and peaked within ∼20 nM GLP-2. Therefore, the optimal concentration (20 nM) of GLP-2 was used further to define the time course for its stimulation on Akt phosphorylation (Fig. 3C). Phosphorylation of Akt on Ser473 was increased as early as at 10 min, while phosphorylation on Thr308 was detected at 20 min posttreatment of GLP-2. This might have resulted from sequential phosphorylation of Akt, which is required for full activation of the enzyme. Moreover, GLP-2-induced stimulation of Akt phosphorylation on Ser473 was sustained for at least 60 min (Supplemental Fig. S1; supplementary materials are found in the online version of this paper at the Journal website). Thus, the data suggested that GLP-2 could stimulate phosphorylation of Akt in both time- and dose-dependent manners and that depended on the amount of GLP-2R cDNA, indicating that GLP-2-mediated stimulation on Akt phosphorylation was acute and specific.
We further characterized whether GLP-2-stimulated Akt phosphorylation was dependent on the PI 3-kinase pathway. In our optimized transfection (1 μg of GLP-2R cDNA), HEK 293 cells were treated with 0, 1, or 20 nM GLP-2 for 15 or 30 min in the absence or presence of 50 μM LY-294002. Results show in Fig. 4 that 1) GLP-2 at 20 nM, but not at 1 nM, stimulated phosphorylation of Akt on Ser473 with no difference between 15 and 30 min; 2) GLP-2 at 1 nM for 30 min stimulated phosphorylation of Akt on Thr308, while GLP-2 at 20 nM for 15 or 30 min stimulated it; and 3) GLP-2-mediated stimulation of Akt phosphorylation on Ser473 and Thr308 was completely abolished in the presence of the PI 3-kinase inhibitor. Therefore, GLP-2-mediated stimulation of Akt phosphorylation was through activating PI 3-kinase-dependent Akt signaling.
GLP-2 activates PI 3-kinase-dependent mTOR signaling.
As shown above, GLP-2-mediated stimulation of protein synthesis was sensitive to the presence of rapamycin (mTOR inhibitor), while mTOR has been identified as one of Akt downstream targets. Thus, we wanted to critically test whether GLP-2 could acutely activate mTOR signaling as a consequence of stimulating the PI 3-kinase-dependent Akt signaling pathway. Results show in Fig. 5 that 1) GLP-2 at 1 and 20 nM stimulated phosphorylation of mTOR on Ser2448 (a key stimulatory site of the kinase) with no difference between 15 and 30 min; 2) GLP-2 at 20 nM stimulated phosphorylation of p70S6K1 and 4E-BP1 (two key mTOR downstream targets) with no difference between 15 and 30 min, but the stimulation for 15 min was more profound at 20 nM GLP-2 compared with that at 1 nM; 3) the presence of 50 μM LY-294002 not only suppressed mTOR signaling (and Akt signaling as well) at the basal level but also partially blocked GLP-2-mediated activation of their signaling. Therefore, GLP-2-mediated enhancement of mTOR signaling was partially through activation of the PI 3-kinase-dependent Akt pathway as well.
GLP-2-mediated stimulation of protein synthesis in hippocampal neurons.
In addition to HEK 293-transfected cells, we wanted to test whether GLP-2-mediated stimulation of protein synthesis is physiologically relevant in the mouse hippocampal neurons (expressing native GLP-2R) that had been characterized and used in our electrophysiology study (28). Results are shown in Figs. 6 and 7. 1) GLP-2 increased de novo synthesis of protein in a dosage-dependent manner, and this reached the plateau among 20–200 nM (Fig. 6A). Interestingly, GLP-2-mediated stimulation in protein synthesis (by 35% for 20 nM GLP-2) was very comparable to a 40% increase for 100 nM insulin (Fig. 6B). 2) GLP-2-mediated stimulation of protein synthesis was abolished when pretreated not only with 50 μM LY-294002 but also with 50 μM rapamycin (Fig. 6, C and D). Notably, the basal rate of [3H]Phe incorporation in hippocampal neurons decreased by 25 and 40% in the presence of LY-294002 and rapamycin, respectively, suggesting that activation of Akt-mTOR signaling is required for protein synthesis in primary neurons.
GLP-2-mediated activation of akt-mTOR signaling in hippocampal neurons.
Finally, we wanted to test whether GLP-2 acutely and directly activates Akt-mTOR signaling in hippocampal neurons. After a 30-min treatment with 20 nM GLP-2, phosphorylation of Akt at Ser473 and mTOR at Ser2448 increased in hippocampal neurons (Fig. 7). It can be appreciated that phosphorylation of Akt and mTOR (in green in Fig. 7, A and B) was stronger in cytoplasm of neurons (in red by neuronal marker PGP9.5) treated with GLP-2 compared with the control. Quantitatively, phosphorylated protein abundance of Akt at Ser473, mTOR at Ser2448, or p70S6K1 at Thr389 increased in a GLP-2 dosage-dependent manner (Fig. 7, C and D). Note that GLP-2-increased phosphorylation only occurred at higher concentrations (20–200 nM) of GLP-2 in hippocampal neurons.
In the present study, we showed in both transfected HEK 293 cells and cultured hippocampal neurons that GLP-2 acutely and directly enhanced de novo synthesis of protein, and this was through activating the PI 3-kinase-dependent Akt-mTOR signaling pathway, which could provide us further insight into GLP-2-induced action and signaling at cellular and molecular levels. We speculate that GLP-2R-induced activation of the PI 3-kinase-dependent Akt-mTOR signaling pathway may play an important role in de novo synthesis of proteins, which is required for providing secondary mediators (namely neuropeptides or growth factors) and sustaining neuronal long-term potentiation to fulfill GLP-2-mediated anabolic action and signal transduction.
Previous in vivo studies have shown that GLP-2 chronically increases intestinal growth by stimulating cell proliferation and protein synthesis and acutely enhances L-type calcium channel activity. It is generally accepted that GLP-2-mediated action on epithelial cell proliferation is indirect through GLP-2R activation-mediated release of unidentified secondary mediators, namely vasoactive intestinal polypeptide (VIP), endothelial nitric oxide synthase (eNOS), glucagon, epidermal growth factor (EGF), and insulin-like growth factor I (IGF-I) (3, 4, 10, 13–15, 20). If GLP-2-mediated action is sustained, de novo synthesis of these secondary mediators might be essential. In addition, GLP-2-mediated stimulation of protein synthesis might be equally important for long-term potentiation in GLP-2R-positive neurons. It is well established that protein synthesis is required for establishment and maintenance of long-term potentiation underling learning and memory in the hippocampus (1, 17, 27). Therefore, we hypothesized that GLP-2 might acutely and directly stimulate de novo synthesis of protein.
Characterization of GLP-2R transient expression in HEK 293 cells.
To address our hypothesis, we employed a transient transfection approach. This is because cell lines expressing endogenous GLP-2R are not available, and primary culture of GLP-2R-expressive cells (namely enteric neurons, enteroendocrine cells, or subepithelial myofibroblasts) is a real challenge. After HEK 293 cells were transiently transfected with the full-length coding sequence of human GLP-2R cDNA (previously cloned in the laboratory), we further characterize their function and specificity. Our previous study (14) had shown that human GLP-2R cDNA was expressed on the plasma membrane of COS cells, possibly indicating appropriate folding and trafficking. In the present study, we further characterized the specificity and responsiveness of GLP-2R overexpressed in HEK 293 cells. It is clear that 125I-labeled human GLP-2 binding was saturable (Fig. 1), and GLP-2-mediated stimulation of protein synthesis was not only concentration dosage (ligand) dependent, but also the cDNA amount was (receptor) dependent (Fig. 2, A and B). By regression analysis of the GLP-2 binding curve, HEK 293 cells with human GLP-2R cDNA transient transfection had a comparable disassociation constant (Kd = 0.22 nM) compared with COS cells stably transfected with rat GLP-2R cDNA (Kd = 0.57 nM) and cultured hippocampal neurons (Kd = 0.48 nM) (21, 28). However, the maximal number of binding sites (Bmax = 321 fmol/μg whole cell protein) of HEK 293 cells was much higher than that of COS cells (Bmax = 1.8 fmol/μg plasma membrane protein) or hippocampal neurons (Bmax = 1.4 × 10−3 fmol/μg whole cell protein). Note that binding characteristics of human GLP-2 to rat GLP-2R, human GLP-2 to mouse GLP-2R, and human GLP-2 to human GLP-2R was determined in COS cells, hippocampal neurons, and HEK 293 cells, respectively. Thus, differences in binding characteristics might be attributed to the abundance of GLP-2R protein (for binding capacity) and the specificity of GLP-2-GLP-2R per se (for binding affinity). Thus, HEK 293 transfection cells, possessing comparable affinity and much higher binding capability, were binding specific and function responsive, probably indicating a suitable cell model for molecular analysis of GLP-2R action and signaling.
Optimal amount of GLP-2R transfected in HEK 293 cells.
To determine transient expression of GLP-2R in specific and functional responses to GLP-2 stimulus, we wanted to assess whether GLP-2-mediated stimulation of de novo protein synthesis was dependent on the amount of GLP-2R cDNA construct transfected. Theoretically, the amount of GLP-2R cDNA transfected could be related to its copy number if efficiency was constant for transfection into individual cells. Therefore, the optimal copy number of GLP-2R cDNA was estimated at roughly 2.5–5.0 × 1011/105 HEK 293 cells, i.e., 2.5–5.0 × 106 each cell for the maximal stimulation of protein synthesis. This estimation was calculated as follows: 0.5–1.0 μg (roughly 2.79–5.57 × 1011 copy number) of GLP-2R cDNA was transfected in 1 × 105 HEK 293 cells (in a 6-well plate), with assumption of 80% transfection efficiency achieved using the FuGENE 6 transfection reagent. Importantly, our data suggested that GLP-2-mediated stimulation of protein synthesis resulted from amount (probably copy number)-dependent GLP-2R activation. We speculate that GLP-2-mediated stimulation of protein synthesis depends not only on the concentration of GLP-2 (ligand) but also on the expression level of GLP-2R (receptor).
GLP-2 directly stimulated protein synthesis in a PI 3-kinase-dependent and rapamycin-sensitive manners.
We showed that GLP-2 acutely and directly stimulated de novo synthesis of protein in the present study. GLP-2 stimulated protein synthesis in ligand dosage (GLP-2)- and receptor (GLP-2R) cDNA amount-dependent manners, indicating that GLP-2R activation-mediated action on protein anabolism is not only kinetically specific but also pathophysiologically relevant. Due to higher binding capacity and affinity, rates of protein synthesis in HEK 293 transfection cells peaked within a lower range of GLP-2 concentrations (0.2–2 nM), and responded quadratically to its dosage. In contrast, rates of protein synthesis in hippocampal neurons reached plateaus within a higher range of GLP-2 concentrations (20–200 nM) and responded linearly to its dosage. Note that a decline in protein synthesis (in HEK 293 transfection cells) at the highest level of 200 nM GLP-2 might result from an artifact, nonspecific, off-target inhibitory effect on the molecular level. It is documented that GLP-2 secretes (from enteroendocrine L-cells) in response to food intake (31) and circulates at postprandial, high levels, probably stimulating protein synthesis at basal rates under the physiological status. Importantly, GLP-2-mediated stimulation of protein synthesis was completely abolished in the presence of a PI 3-kinase inhibitor. However, it was blunted in the presence of an mTOR inhibitor to a less substantial degree in HEK 293 transfection cells compared with hippocampal neurons. Thus, GLP-2-mediated anabolic action (on protein synthesis) follows a growth factor-mediated mitogenic, trophic pattern in a PI 3-kinase-dependent and rapamycin-sensitive manner.
GLP-2 acutely activated PI 3-kinase-dependemnt akt-mTOR signaling pathway.
It is well documented that the PI 3-kinase-dependent AKT-mTOR pathway can be activated by growth factors, hormones, and nutrients (amino acids) (18, 25, 26). Activation of PI 3-kinase initiates a cascade of events: phosphorylated Akt activates mTOR, which promotes the activation of translational activator p70S6K and the hierarchical phosphorylation of 4E-BP1. Activated p70S6K phosphorylates the ribosomal protein S6 and causes an increase in translation of 5′ TOP-containing mRNAs. Hyperphosphorylated 4E-BP1 release eIF4E, thereby allowing for cap-dependent translation to occur. Our data indicated that GLP-2-mediated stimulation of protein synthesis is linked to activation of the PI 3-kinase-dependent Akt-mTOR signaling pathway.
PI 3-kinases are involved in cellular functions (e.g., cell growth, proliferation, survival, and apoptosis) and metabolism (e.g., glucose uptake and protein synthesis), most of which is mediated by PI 3-kinase-activated Akt signaling (18). GLP-2 rapidly and dose-dependently phosphorylated Akt on Ser473 and Thr308, and this was blocked by inhibition of PI 3-kinase. As discussed above, PI 3-kinase activation was essential for GLP-2-induced stimulation of protein synthesis. LY-294002 (an inhibitor of PI 3-kinase) not only potently inhibited GLP-2-increased rates of [3H]Phe incorporation but also abolished GLP-2-induced activation of Akt signaling. Though pretreatment with LY-294002 suppressed protein synthesis and Akt phosphorylation at the basal level, GLP-2-mediated stimulation would still occur in the presence of LY-294002 if not majorly through a PI 3-kinase-dependent pathway. Therefore, GLP-2 stimulated protein synthesis through a PI 3-kinase-dependent pathway, and GLP-2-enhanced Akt phosphorylation might play an important role in protein synthesis.
mTOR is a key downstream effector of PI 3-kinase-dependent Akt, which regulates cell growth, proliferation, and survival, protein synthesis, and angiogenesis (25). Specifically, mTOR is essential for ribosomal biogenesis and protein translation. Two downstream targets of mTOR, namely 4E-BP1 and p70S6K, are important regulators of protein translation and cell proliferation. GLP-2-induced increase in [3H]Phe incorporation rate was blunted in the presence of rapamycin (an inhibitor of mTOR) (see above in discussion). GLP-2R activation stimulated the mTOR phosphorylation on Ser2448, thus activating p70S6K and inactivating 4E-BP1, which would be essential in the control of protein synthesis. In the present study, moreover, GLP-2 stimulated phosphorylation of p70S6K and 4E-BP1, and this was blocked by the PI 3-kinase inhibitor (LY-294002), indicating again that GLP-2-mediated activation of the mTOR signaling was PI 3-kinase dependent. Interestingly, GLP-2 (20 nM)-mediated activation of mTOR signaling, i.e., phosphorylation of mTOR, p70S6K, and 4E-BP1, occurred as early as 15 min post-GLP-2 treatment (Fig. 5), and GLP-2-mediated phosphorylation of Akt on Ser473 occurred earlier compared with that on Thr308 (Fig. 3C). Note that Akt can be phosphorylated on Thr308 by PDK1 (phosphoinositide-dependent kinase-1), and phosphorylated on Ser473 by mTORC2 (rictor-mTOR complex) (24, 26). Therefore, our data suggest that GLP-2-mediated activation of mTOR signaling might not be completely attributed to downstream action of Akt signaling.
In summary, GLP-2R activation stimulated protein synthesis and activated the mTOR signaling pathway, and this was partially mediated by the PI 3-kinase-dependent Akt signaling pathway. This metabolic action and signaling pathway of GLP-2 was not only defined in HEK 293 transfection cells but also established in cultured hippocampal neurons. GLP-2 acutely and dose-dependently stimulated protein synthesis in PI 3-kinase-dependent and mTOR-sensitive manners, suggesting that GLP-2-mediated activation of the PI 3-kinase-dependent Akt-mTOR signaling pathway was necessary for its stimulation of de novo synthesis of protein. We speculate that GLP-2-induced stimulation of protein synthesis might be physiologically relevant to providing secondary mediators [namely neuropeptides (VIP) or growth factors (EGF and IGF-I)] and maintaining neuronal long-term potentiation, which both require de novo synthesis of protein to fulfill GLP-2 signal transduction.
This work was supported by the USDA/ARS under Cooperative Agreement No. 6250-51000-043, National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-075489 and DK-084125, and National Natural Science Foundation of China Grant 30728016 (X. Guan).
No conflicts of interest, financial or otherwise, are declared by the authors.
We thank Douglas Burrin, Qiang Tong, Guangcheng Zhang, and Adam Gillum at Baylor College of Medicine for scientific and technical support.
This work is a publication of the United States Department of Agriculture/Agricultural Research Service (USDA/ARS) Children's Nutrition Research Center, Departments of Pediatrics and Medicine, Baylor College of Medicine, and Texas Children's Hospital, Houston, TX. The contents of this publication do not necessarily reflect the views or policies of the U.S. Department of Agriculture; nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.