TRPM5 is critical for linoleic acid-induced CCK secretion from the enteroendocrine cell line, STC-1

Abstract

Fatty acid-induced stimulation of enteroendocrine cells leads to release of the hormones such as cholecystokinin (CCK) that contribute to satiety. Recently, the fatty acid activated G protein-coupled receptor GPR120 has been shown to mediate long-chain unsaturated free fatty acid-induced CCK release from the enteroendocrine cell line, STC-1, yet the downstream signaling pathway remains unclear. Here we show that linoleic acid (LA) elicits membrane depolarization and an intracellular calcium rise in STC-1 cells and that these responses are significantly reduced when activity of G proteins or phospholipase C is blocked. LA leads to activation of monovalent cation-specific transient receptor potential channel type M5 (TRPM5) in STC-1 cells. LA-induced TRPM5 currents are significantly reduced when expression of TRPM5 or GPR120 is reduced using RNA interference. Furthermore, the LA-induced rise in intracellular calcium and CCK secretion is greatly diminished when expression of TRPM5 channels is reduced using RNA interference, consistent with a role of TRPM5 in LA-induced CCK secretion in STC-1 cells.

ingested fat, specifically in the form of free fatty acids (FFAs), activates enteroendocrine cells (EECs) in the proximal small intestine, which initiates a number of physiological functions including satiety, potentiating insulin release, and delaying gastric emptying (7, 9, 14). FFA-induced satiety is mainly mediated by secretion of satiety peptides including cholecystokinin (CCK) and glucagon-like peptide-1 (GLP-1) by specialized EECs (type I) present in preabsorptive sites in the intestine (6, 7, 14, 24, 28). Recently, the fatty acid-activated G protein-coupled receptor, GPR120, has been shown to mediate long-chain unsaturated FFA-induced CCK and GLP-1 release from the enteroendocrine cell line, STC-1, yet the downstream signaling pathway from GPR120 activation to CCK secretion remains unclear (9, 26). Several groups have shown that FFA-induced CCK release in STC-1 cells is dependent on increased intracellular calcium. Moreover, removal of extracellular calcium and/or application of the L-type voltage-gated calcium channel (VGCC) blocker nicardipine significantly reduced the FFA-induced intracellular calcium rise and CCK secretion in STC-1 cells (23, 26). Thus, it is reasonable to postulate from these observations that polyunsaturated fatty acids (PUFAs) may depolarize STC-1 cells by activating GPR120 and initiating membrane depolarization thereby activating L-type VGCCs leading to an elevation of intracellular calcium concentration and CCK secretion. The link between GPR120 and the development of the receptor potential (i.e., depolarization) remains unknown.

One candidate for producing the FFA-induced receptor potential in nutrient-responsive cells is transient receptor potential channel type M5 (TRPM5). TRPM5 is a calcium-activated, nonselective, monovalent cation-permeable channel that is highly expressed in taste receptor cells (TRCs) and is required for sweet, bitter, and umami taste transduction (16, 20, 21, 25, 30). TRPM5 in TRCs appears to be involved in the final step of a signaling cascade that culminates in membrane depolarization. While TRPM5 is unequivocally involved in taste transduction, activity of TRPM5 channels has never been recorded in native cells in response to taste stimuli. Other than TRCs, TRPM5 mRNA has been expressed in various tissues like olfactory epithelium, lung epithelium, and, importantly, in chemosensory cells found in gut, including STC-1 cells (1, 11, 15). While TRPM5 has been implicated in chemosensory transduction in STC-1 cells, it has not been studied directly in these cells.

In the present study, we show that the prototypical long-chain FFA stimulus, linoleic acid (LA), depolarizes STC-1 cells via activation of TRPM5. LA-induced depolarization is significantly reduced when GDPβS and/or U73122 is used to block G proteins and PLC, respectively, suggesting that depolarization in response to LA depends on a G protein-PLC pathway. Transfection of small interfering RNA (siRNA) specific to TRPM5 showed significant reduction in LA-induced intracellular calcium rise as well as CCK secretion from STC-1 cells, suggesting that TRPM5 plays a crucial role in fatty acid-induced CCK release. Finally, we showed that LA-induced TRPM5 currents were significantly reduced when siRNA specific to GPR120 was transfected in STC-1 cells consistent with the interpretation that TRPM5 is activated downstream of GPR120.

MATERIALS AND METHODS

Cell culture.

The murine enteroendocrine cell line STC-1, a generous gift from Dr. D. Hanahan (University of California, San Francisco), was derived from intestinal endocrine tumor developed in double transgenic mouse expressing insulin promoter linked to simian virus 40 large T antigen and polyoma virus small T antigen. STC-1 cells were cultured using Dulbecco's modified Eagle's medium containing 10% (vol/vol) fetal bovine serum. STC-1 cells were plated on glass coverslips 24–48 h before calcium imaging and electrophysiological studies.

Intracellular calcium imaging.

STC-1 cells were loaded with Fura-2-AM (5 μM; Molecular Probes, Eugene, OR) for 1 h in Tyrode's with 10% pluronic acid at 37°C in the dark. Cells were then rinsed and placed in medium for 30 min to allow de-esterification of acetoxymethyl ester group from Fura-2. The coverslips were mounted onto the chamber (RC-25F, Warner Instruments, Hamden, CT), placed on an inverted microscope (Nikon, Eclipse TS100, Japan), and perfused continuously with Tyrode's. Cells were illuminated with a 100-W xenon lamp, and excitation wavelengths (340/380 nm) were delivered by a monochromator (Bentham FSM150, Intracellular Imaging, Cincinnati, OH) at a rate of 20 ratios per minute. Fluorescence was measured by a CCD camera (pixelFly, Cooke, MI) coupled to a microscope and controlled by imaging software (Incyt Im2, Intracellular Imaging). The ratio of fluorescence (340 nm/380 nm) was directly converted to calcium concentrations using a standard curve generated for the imaging system using Fura-2 calcium imaging calibration kit (Molecular Probes). Fatty acids and other compounds were applied extracellularly with a bath perfusion system at a flow rate of 4 ml/min permitting complete exchange of the extracellular solution in <20 s.

Solutions.

Standard extracellular saline solution (Tyrode's) contained (in mM) 140 NaCl, 5 KCl, 1 CaCl₂, 1 MgCl₂, 10 HEPES, 10 glucose, and 10 Na pyruvate; pH 7.40 adjusted with NaOH; 310 mosM. Sodium-free saline (sodium-free Tyrode's) contained (in mM) 280 mannitol, 5 KCl, 1 CaCl₂, 1 MgCl₂, 10 HEPES, 10 glucose; pH 7.40 adjusted with TrisOH; 310 mosM (adjusted with mannitol). The 60 mM Na⁺ Tyrode's contained (in mM) 50 NaCl, 180 mannitol, 5 KCl, 1 CaCl₂, 1 MgCl₂, 10 HEPES, 10 glucose, and 10 Na pyruvate. The 10 mM Na⁺ Tyrode's contained (in mM) 280 mannitol, 5 KCl, 1 CaCl₂, 1 MgCl₂, 10 HEPES, 10 glucose, and 10 Na pyruvate. Calcium-free saline (calcium-free Tyrode's) contained (in mM) 140 NaCl, 5 KCl, 1 EGTA, 1 MgCl₂, 10 HEPES, 10 glucose, and 10 Na pyruvate; pH 7.40 adjusted with NaOH; 310 mosM. A K⁺-based intracellular solution was used for measurement of membrane potential and contained (in mM) 140 K gluconate, 1 CaCl₂, 2 MgCl₂, 10 HEPES, 11 EGTA, 3 ATP, and 0.5 GTP; pH 7.2 adjusted with KOH; 310 mosM. A Cs⁺-based intracellular solution was used for recording TRPM5 currents and contained (in mM) 140 Cs acetate, 1 CaCl₂, 2 MgCl₂, 10 HEPES, 11 EGTA, 3 ATP, and 0.5 GTP; pH 7.2 adjusted with TrisOH; 310 mosM. U-73122, an inhibitor of PLC, U-73343, the inactive analog of U-73122, and bombesin were purchased from Sigma (St. Louis, MO). GDPβS, an inhibitor of G protein activation, was obtained from EMD Biosciences (La Jolla, CA). The LA stock was made in 100% ethanol, and working solutions were made from stock solutions immediately before use.

Electrophysiological recording and analysis.

We performed whole cell patch-clamp recording to measure membrane potential (current-clamp mode) and/or to record FFA-induced currents (voltage-clamp mode) in STC-1 cells with an Axopatch-200B amplifier (Molecular Devices, Sunnyvale, CA). Patch pipettes were fabricated from borosilicate glass on a Flaming-Brown micropipette puller (model P-97; Sutter Instrument, Novato, Ca) and subsequently fire polished on a microforge (model MF-9, Narishige) to a resistance of 5–8 MΩ. Commands were delivered and data were recorded using pCLAMP software (version 10, Molecular Devices) interfaced to an Axopatch-200B amplifier with a DigiData 1322A analog-to-digital board. Data were filtered online at 1 kHz and sampled at 2–4 kHz.

Membrane potential measurement.

The membrane potential (V_M) of STC-1 cells was recorded continuously before, during, and after fatty acid (LA) application using the current-clamp mode of the amplifier while the cell was held at its zero current level (i.e., at rest). For recording FFA-induced changes in membrane potential, LA was applied by bath application in most cases. In other experiments, rapid, focal application of fatty acids was used and produced similar results. To determine the ionic dependence of FFA-induced changes in membrane potential, V_M was recorded in three different extracellular solutions including Tyrode's, sodium-free Tyrode's, and calcium-free Tyrode's.

TRPM5 current recording.

FFA-induced TRPM5 currents in STC-1 cells were recorded using the voltage-clamp mode. Typical inward currents were recorded at a holding potential of −100 mV. Fatty acid was applied focally from a pipette positioned near the cell and delivered by a PicoSpritzer III (Parker Hannifin, Cleveland, OH) controlled by the data acquisition and analysis software. Ramp protocols from −100 mV to +100 mV (500-ms duration) were used to generate instantaneous current-voltage relationship of LA-induced TRPM5 current in various solutions.

siRNA construction and transfection.

For siRNA experiments, we used Silencer predesigned siRNA targeted against GPR120 and TRPM5, respectively (Ambion, Austin, TX) according to the manufacturer's directions. As a negative control, we used the Silencer control siRNA kit (Ambion). The concentration of siRNA used for transfection was 25–200 nM. STC-1 cells were reverse transfected with siRNA complexes using Lipofectamine 2000 (Invitrogen, Carlsbad CA) transfection reagent 24–48 h before functional assays were performed (patch-clamp studies, calcium imaging, and CCK measurement) or quantitative PCR.

Quantitative RT-PCR.

RNA was isolated using Tri Reagent (Molecular Research Center, Cincinnati, OH) according to manufacturer's instructions. For real-time quantification of mRNA, we used two-tube RT-PCR. First-strand cDNA was synthesized using an iScript kit from Bio-Rad (Hercules, CA). The RNA concentration was 50 ng per 20 μl reaction volume. The SmartCycler from Cepheid (Sunnyvale, CA) was used to follow the PCR reaction in real time. The HotMaster Taq DNA polymerase kit (5 Prime, Gaithersburg, MD ) was used for the PCR reaction [final concentration: 1× reaction buffer, 5 mM Mg²⁺, 200 μM dNTPs, 300–900 nM forward and reverse primers (for GPR120), 60 nM for GAPDH and 1.25 U/μl HotMaster Taq]. Two microliters cDNA per reaction were amplified using a two-step PCR (6 s denaturation, 60 s annealing and extension). We used a TaqMan detection system, and primer pairs for the target sequences were multiplexed with the primer pairs for the housekeeping gene, GAPDH, for comparison of expression levels in STC-1 cells. Primers and probes for GPR120 (sense/antisense; 5′-CTGCACATTGGATTGGC-3′/5′-TCTGGTGGCTCTCGGAGTAT-3′; 624 −640/783–802) and GAPDH (sense/antisense: 5′-TGCACCACCAACTGCTTAG-3′/5′-GGATGCAGGGATGATGTTC-3′; 487–496/636–654) were designed using Oligo 6.0 Primer Analysis software (Molecular Biology Insights, Cascade, CO). GPR120 probe (5′-CTCTTCCGCGAGGCTTTCGTGATCTGT-3′; 738–764) was labeled at the 5′-end with FAM as the reporter dye and BHQ-1 at the 3′-end as the quencher dye. The GAPDH probe (5′-ATCACGCCACAGCTTTCCAGAGGG-3′; 595–618) was labeled with ROX as the reporter and BHQ-2 as the quencher (Integrated DNA Technologies, Coralville, IA). For detection of TRPM5, a predesigned Taqman Gene Expression Assay (Applied Biosystems, Foster City, CA) was used.

CCK secretion from STC-1 cells.

For CCK measurement studies, STC-1 cells were plated in 24-well plates at a density of 1 × 10⁵ cells/cm². Before LA stimulation, STC-1 cells were rinsed twice with Tyrode's solution. Cells were then incubated with 30 μM LA containing Tyrode's solution at 37°C for 30 min. The supernatant was collected and centrifuged at 4°C for 5 min at 1,000 g to remove cell debris and stored at −20°C until assayed. CCK concentration was measured by specific CCK octapeptide (26–33) fluorescent enzyme immunoassay kit according to the manufacturer's directions (Phoenix Pharmaceuticals, Belmont, CA).

Statistical analysis.

The significant effects of all the treatments were determined by unpaired Student's t-test (α = 0.05) compared with their controls as described in the text. Data are presented as means ± SE, unless otherwise indicated.

RESULTS

LA depolarizes and elicits a rise in intracellular calcium in STC-1 cells.

In the present study, we have used ratiometric calcium imaging and patch-clamp recording to explore the signaling pathway of PUFA-induced hormone secretion in STC-1 cells. As a first approach we used the prototypical PUFA stimulus, LA (30 μM), to elicit changes in intracellular calcium measured by a high-speed ratiometric calcium imaging system (InCyt2-IM; Intracellular Imaging) using the ratiometric dye Fura-2-AM. Similar to the responses shown by others (9), LA (30 μM) elicits a robust and reversible increase in intracellular calcium (349.0 ± 10.7 nM; n = 114; Fig. 1, A and D) in STC-1 cells.

Fig. 1.

LA-induced intracellular calcium rise is dependent on extracellular [Na⁺] ([Na⁺]_out) and [Ca²⁺] ([Ca⁺]_out). LA depolarizes STC-1 cells, and this depolarization is dependent on [Na⁺]_out but not on [Ca²⁺]_out. A: bath application of LA (30 μM) induces increase in intracellular calcium concentration ([Ca²⁺]_in) in STC-1 cells. B and C: LA (30 μM)-induced rise in [Ca²⁺]_in in the absence of [Na⁺]_out (B) and [Ca²⁺]_out (C). D: mean ± SE responses (rise in [Ca²⁺]_in) of LA (30 μM) in Tyrode's (control), calcium-free Tyrode's, and sodium-free Tyrode's. E: bath application of LA (30 μM) induces depolarization in STC-1 cells. Cells were held at 0 current level (e.g., at its resting potential). F and G: LA (30 μM)-induced depolarization in the absence of [Na⁺]_out (F) and [Ca²⁺]_out (G). H: mean ± SE responses (depolarization) of LA (30 μM) in Tyrode's, calcium-free Tyrode's, and sodium-free Tyrode's. NS, not significant.

We then carried out a series of imaging experiments to investigate the dependence of the LA-induced changes in intracellular calcium on extracellular sodium, extracellular calcium, and intracellular calcium. In the absence of extracellular calcium, the LA-induced rise in intracellular calcium was significantly reduced (55.0 ± 5.3 nM; P < 0.001; n = 56; Fig. 1, C and D). This significant reduction in intracellular calcium rise in the absence of extracellular calcium is similar to that shown by others (23). Our data are consistent with the interpretation that LA also causes release of calcium from intracellular stores. Thus, calcium from both intracellular and extracellular sources appears to contribute to the fatty acid responses in STC-1 cells. Interestingly, removal of extracellular sodium also caused a significant decrease in the LA-induced rise in intracellular calcium (95.1 ± 10.9 nM; n = 58; Fig. 1, B and D), suggesting that sodium ions influx may be involved in fatty acid transduction.

We also explored the effect of LA on membrane potential of STC-1 cells using patch-clamp recording in the whole cell current-clamp configuration. STC-1 cells were held at zero current level to determine the resting membrane potential. In 13 cells the recorded resting membrane potential was −60.0 ± 2.5 mV on average. LA (30 μM) applied by bath perfusion elicited a large and reversible depolarization of 68 ± 2.4 mV (n = 13; Fig. 1E). Importantly, the time courses of both the LA-induced calcium rise (Fig. 1A) and the membrane depolarization (Fig. 1E) were comparable using these two approaches.

To determine the ionic dependence of LA-induced membrane depolarization, we performed a series of experiments in the absence of extracellular sodium or calcium ions. Removal of extracellular calcium ions did not alter the 30 μM LA-induced depolarization compared with control [extracellular Ca²⁺ concentration ([Ca²⁺]_out) = 0: 68 ± 1.7 mV, n = 7; control: 68 ± 2.4 mV, n = 12; Fig. 1, G and H]. On the other hand, when extracellular sodium was removed, the LA-induced depolarization was significantly reduced (9 ± 3.3 mV; n = 6; P < 0.001; Fig. 1, F and H). These results may indicate that sodium entry is necessary for depolarization in STC-1 cells upon LA application and that calcium influx is downstream of sodium entry (depolarization).

Involvement of G proteins and phospholipase C in LA-induced depolarization.

GPR120, a long-chain PUFA-activated G protein-coupled receptor (GPCR), has been shown to play a key role in the PUFA-induced calcium rise and eventual hormone (GLP-1, CCK) release from STC-1 cells (9, 26). To confirm the role of G proteins in the signaling pathway, we examined the effect of blocking G protein activation with the reversible blocker GDPβS (1 mM) on LA-induced depolarization. In electrophysiological experiments, GDPβS was included in the intracellular solution and membrane potential was recorded 2–3 min after the whole cell configuration was achieved. LA-induced depolarization was significantly reduced in the presence of GDPβS (7 ± 2.6 mV; n = 8; P < 0.001; Fig. 2, A, D, and E) compared with control (68 ± 2.4 mV; n = 12; Fig. 2, A and E) in STC-1 cells. Next, we determined the involvement of PLC in LA signaling. Depolarization in response to LA was significantly reduced in the presence of the PLC blocker U73122 (3 μM, 15-min pretreatment; 4 ± 4.2 mV; n = 5; P < 0.001; Fig. 2, B and E) but was not affected by treatment with the inactive analog U73343 [3 μM, 15-min pretreatment; 57 ± 6.7 mV; n = 5; not significant (NS); Fig. 2, C and E]. To determine the role of PLC in the intracellular calcium rise in response to LA using calcium imaging, STC-1 cells were pretreated with 3 or 5 μM U73122 for 30 min. U73122-treated STC-1 cells showed a profound reduction in intracellular calcium changes in response to LA compared with controls (data not shown).

Fig. 2.

Current-clamp studies in STC-1 cells held at their zero current level showing that the depolarization in response to linoleic acid (LA; 30 μM) is greatly reduced upon blocking the activity of G proteins and PLC. A–D: LA (30 μM)-induced depolarization in control (A), U73122 (3 μM)-treated (B), U73343 (3 μM, the inactive analog of U73122)-treated (C), and GDPβS (1 mM)-treated (D) STC-1 cells. E: mean ± SE responses of 30 µm LA in control conditions, GDPβS treatment, U73122 treatment, and U73343 treatment in STC-1 cells.

LA activates sodium-dependent inward currents in STC-1 cells.

In whole cell voltage-clamp experiments (V_HOLD= −100 mV), rapid and focal application of 30 μM LA induced large-amplitude inward currents in STC-1 cells (2,010.5 ± 379.3 pA; n = 16; Fig. 3A). In separate current-clamp experiments, we found that rapid and focal application of 30 μM LA depolarized the cell (Fig. 3A, inset) with a similar time course. Removal of extracellular sodium abolished inward currents in response to LA (48.6 ± 15.2 pA; n = 10; P < 0.001; Fig. 3, B and C). Removal of extracellular calcium ions, however, did not alter the inward current in response to LA (data not shown). Next we did a series of experiments to investigate the permeability properties of LA-induced inward currents. In standard experimental conditions (i.e., total monovalent cation concentrations equal on both sides of the membrane), the current-voltage relationship of LA-induced currents yields a reversal potential of ≈ 0 mV (Fig. 3D), consistent with the interpretation that LA activates a nonselective monovalent cation-permeable pathway. Under these conditions (Tyrode's outside; cesium-based intracellular solution), the inward current is carried by sodium and the outward current is carried by cesium ions. Consistent with this conclusion, changes in extracellular sodium ions led to predictable changes in the reversal potential of LA-induced currents that were closely predicted by Goldman-Hodgkin-Katz (GHK) equation for sodium, potassium, and cesium-permeable conductance (Fig. 3, D and E), assuming that the relative permeability of all three ions is equal. Furthermore, partial substitution of intracellular cesium ions with mannitol shifted the reversal potential of LA-induced currents in a predictable manner (data not shown).

Fig. 3.

Rapid and focal application of LA (30 μM) elicited monovalent cation-permeable currents. A and B: LA (30 μM)-induced inward currents [holding potential (V_H) = −100 mV] in Tyrode's (A) and sodium-free Tyrode's (B). A, inset: rapid and focal application of LA induces depolarization with time course similar to the activation of inward currents. C: mean ± SE responses (inward currents) of LA (30 μM) in Tyrode's and sodium-free Tyrode's. D: current-voltage relationship of LA (30 μM)-induced currents at 10, 60, and 140 mM [Na⁺]_out. E: comparison between theoretical and measured reversal potentials (E_rev) ± SE of LA (30 μM)-induced currents at extracellular sodium concentrations of 10, 60, and 140 mM. Reversal potentials shown were corrected for liquid junction potential.

TRPM5 mediates LA-induced responses.

The calcium-activated, monovalent cation-selective channel TRPM5 has been shown to play a critical role in the taste transduction pathway of sweet, bitter, and umami stimuli where it apparently acts downstream of taste GPCR activation (16, 30). Recently, we have shown that LA activates TRPM5 currents in mouse taste cells, the first detailed report showing the activation of tastant-induced TRPM5-like currents in native taste cells (17). TRPM5, as well as GPCR taste receptors and other G protein signaling elements, is expressed in STC-1 cells (4, 16, 21). On the basis of our results that LA-activated currents in STC-1 cells are primarily monovalent cation selective, we investigated specific involvement of TRPM5 channels in LA-activated inward current by using the TRPM5 channel blocker triphenylphosphine oxide (TPPO) (17, 19). In whole cell voltage-clamp experiments, we compared LA-induced inward currents in control STC-1 cells and in TPPO-treated (100 μM, 2-min pretreatment) STC-1 cells. TPPO treatment significantly reduced the LA-induced inward current (374.3 ± 84.4 pA; n = 9; P = 0.04; Fig. 4, A and B) compared with untreated STC-1 cells (1,631.5 ± 521 pA; n = 10; Fig. 3, A and B). To confirm that TPPO was acting at TRPM5, we also examined the effect of TPPO on currents elicited by the bitter stimulus, denatonium benzoate (DB, 10 mM). TPPO significantly reduced DB-induced inward currents (162 ± 32.6 pA; n = 6; P < 0.001; Fig. 4B) compared with control DB responses (1,176.6 ± 102.6 pA; n = 6; Fig. 4B). Interestingly, TPPO-mediated inhibition of inward currents in response to both LA and DB is qualitatively and quantitatively very similar (Fig. 4, A and B).

Fig. 4.

LA (30 μM)-induced responses are mediated by transient receptor potential channel type M5 (TRPM5). A: LA-induced and denatonium benzoate (DB)-induced inward currents in the presence of TRPM5 antagonist triphenylphosphine oxide (TPPO; 100 μM). B: mean ± SE responses (inward currents) to LA (30 μM) and DB (10 mM) in control conditions and during TPPO (100 μM) treatment. C and D: LA (30 μM)-induced inward currents in negative control small interfering RNA (siNEG)-treated (C) and siTRPM5-treated (D) STC-1 cells. E: mean ± SE responses (relative inward current to untransfected cells) of LA (30 μM) in siNEG-treated and siTRPM5-treated STC-1 cells. F and G: LA (30 μM) and bombesin (2 μM)-induced intracellular calcium rise in siNEG-treated (F) and siTRPM5-treated (G) STC-1 cells. H: mean ± SE responses (rise in [Ca²⁺]_in) to LA (30 μM) and bombesin (2 μM) in siNEG-treated and siTRPM5-treated STC-1 cells. I: relative expression of TRPM5 mRNA in siTRPM5 transfected STC-1 cells compared with siNEG-transfected STC-1 cells assayed by quantitative real-time PCR (Q-PCR).

To further validate the involvement of TRPM5 in the LA-induced activation of STC-1 cells, we transfected STC-1 cells either with siRNA specific to TRPM5 (siTRPM5) or with a nonsense sequence (control, siNEG) and examined inward currents in response to 30 μM LA. Inward currents in response to LA were significantly reduced in siTRPM5-treated cells (346.9 ± 75.0 pA, n = 10; P < 0.001; Fig. 4, D and E) compared with siNEG-treated cells (1,925.7 ± 285.1 pA, n = 18; Fig. 4, C and E). There was no significant difference found in the magnitude of inward currents in response to 30 μM LA between siNEG-treated and untransfected STC-1 cells (n = 16; Fig. 4E). Quantitative real-time PCR in STC-1 cells confirmed that siTRPM5 treatment significantly reduced TRPM5 mRNA by ∼80% compared with siNEG-treated cells (P = 0.038; Fig. 4I). There was no statistical difference found between TRPM5 mRNA levels in siNEG-treated and untransfected STC-1 cells.

Next, we examined the effect of TRPM5 knockdown on LA-stimulated calcium signaling since we hypothesized, on the basis of our data, that much of the rise in intracellular calcium was downstream of TRPM5 activation. The LA-induced rise in intracellular calcium concentration was significantly reduced in siTRPM5-treated STC-1 cells (95.0 ± 13.8 nM; n = 71; P < 0.001; Fig. 4, G and H) compared with siNEG-treated STC-1 cells (355.0 ± 31.2 nM; n = 84; Fig. 4, F and H). However, intracellular calcium rise in response to bombesin (2 μM), which does not involve the activation of TRPM5, was unaffected in all the examined cells (siNEG, 348.7 ± 20.7 nM, n = 84; siTRPM5, 318.0 ± 16.2 nM, n = 71, NS; Fig. 4, F, G, and H), confirming that the knockdown of TRPM5 did not affect other signaling components in the cell.

TRPM5 is activated downstream of GPR120.

In STC-1 cells, GPR120, and apparently not GPR40, mediates PUFA-induced intracellular calcium rise (9). Next, we investigated whether activation of TRPM5 is downstream of activation of GPR120 receptors in response to LA. We transfected STC-1 cells either with siRNA specific to GPR120 (siGPR120) or nonsense siRNA (siNEG) and compared the magnitude of LA-activated TRPM5 currents. TRPM5 currents in response to LA were significantly reduced in siGPR120-treated STC-1 cells (562.1 ± 77.8 pA; n = 17; P < 0.05; Fig. 5, B and C) compared with siNEG-treated STC-1 cells (2,025.3 ± 337.2 pA; n = 15; Fig. 5, A and C). We did not observe any significant difference in LA-induced TRPM5 currents between untransfected cells (2,010.4 ± 379.3 pA; n = 16) and siNEG-treated (2,025.3 ± 337.2 pA; n = 15) STC-1 cells. As shown in Fig. 5D, siGPR120 treatment significantly (P < 0.05) reduced GPR120 mRNA by ∼90% compared with siNEG-treated cells. There was no statistical difference found between GPR120 mRNA levels in siNEG-treated and untransfected STC-1 cells. Furthermore, we examined the effect of GPR120 knockdown on LA-induced intracellular calcium changes since the rise in intracellular calcium is an essential signal for initiating satiety hormone secretion in STC-1 cells. The LA-induced rise in intracellular calcium was significantly reduced in siGPR120-treated STC-1 cells (127.2 ± 17.0 nM; n = 88; P < 0.05; Fig. 5, F and G) compared with siNEG-treated STC-1 cells (355.0 ± 31.2 nM; n = 84; Fig. 5, E and G), an effect also seen by others (9). The bombesin-induced rise in intracellular calcium was unaffected by siGPR120 treatment in all cells examined (siNEG, 348.7 ± 20.7 nM, n = 84; siGPR120, 306.3 ± 15.7 nM, n = 88; Fig. 5, E, F, and G), showing that responsiveness in other pathways remained intact.

Fig. 5.

LA-induced activation of TRPM5 channels is downstream of G protein-coupled receptor GPR120. A and B: LA (30 μM)-induced inward currents in siNEG-treated (A) and siGPR120-treated (B) STC-1 cells. C: mean ± SE responses (relative inward current to untransfected cells) to LA (30 μM) in siNEG-treated and siGPR120-treated STC-1 cells. D: relative expression of GPR120 mRNA in siGPR120-transfected STC-1 cells compared with siNEG-transfected STC-1 cells assayed by Q-PCR. E and F: LA (30 μM) and bombesin (2 μM)-induced intracellular calcium rise in siNEG-treated (E) and siGPR120-treated (F) STC-1 cells. G: mean ± SE responses (rise in [Ca²⁺]_in) to LA (30 μM) and bombesin (2 μM) in siNEG-treated and siGPR120-treated STC-1 cells.

TRPM5 mediates LA-induced CCK release from STC-1 cells.

To link the activation of TRPM5 to the eventual release of CCK in STC-1 cells, we performed quantitative measurements of CCK secretion using commercially available ELISA assays in siNEG-treated (control) and siTRPM5-treated (TRPM5 knockdown) STC-1 cells. As shown in Fig. 6, siTRPM5-treated STC-1 cells showed significantly diminished LA-induced CCK secretion (P = 0.027) compared with siNEG-treated STC-1 cells. Reduced LA-induced CCK release in siGPR120-treated STC-1 cells confirms the previous finding that GPR120 mediates fatty acid-induced CCK release in STC-1 cells (26).

Fig. 6.

TRPM5 mediates LA-induced cholecystokinin (CCK) secretion. Mean ± SE responses (normalized CCK secretion) following 30 μM LA application (30 min) in siNEG-, siTRPM5-, and siGPR120-treated STC-1 cells.

DISCUSSION

FFAs are known to be involved in numerous physiological functions; relatively little is known of how FFAs activate cells or the specific elements of the transduction pathway. Gilbertson et al. (5) provided the first evidence that FFAs activate chemosensory cells (i.e., TRC) by interacting with delayed rectifying potassium (DRK) channels. More recently, several additional fatty acid-responsive proteins have been identified that may play a role in initiating fatty acid transduction. These include the fatty acid-binding protein CD36 (12) and several GPCRs including GPR40, GPR41, GPR43, GPR84, and GPR120 that have been identified as receptors for FFAs and shown to mediate various FFA-induced regulatory functions in different tissues (2, 3, 8, 10, 13, 22, 27, 29). Of particular interest in the digestive system is GPR120, which is abundantly expressed in intestine and functions as the receptor for unsaturated long-chain FFAs (9).

PUFAs stimulate STC-1 cells causing the release of satiety hormones like CCK by activating the GPCR (i.e., GPR120) and elevating intracellular calcium concentrations (26). Several groups have shown that intracellular calcium rise is mediated largely by calcium influx through VGCCs and is a necessary and sufficient step for FFA-induced satiety hormone release (18, 26). However, the details of the intracellular signaling cascade initiated by PUFAs in STC-1 cells that culminates in CCK release have not been studied in great detail. In the present manuscript we have attempted to unravel the PUFA-initiated signaling cascade that leads to the intracellular calcium rise and eventually to CCK secretion in STC-1 cells. In the initial calcium imaging experiments in STC-1 cells, we found that the removal of extracellular sodium ions and extracellular calcium ions significantly reduced the LA-induced intracellular calcium rise. This led us to conclude that sodium entry is required for the LA-induced rise in intracellular calcium. Subsequently, we showed that LA depolarized STC-1 cells and that this depolarization was strictly dependent on extracellular sodium ions but not on extracellular calcium ions. These results in conjunction with our calcium imaging results demonstrate that the LA-mediated depolarization of STC-1 cells is dependent on sodium entry and provides the stimulus (i.e., depolarization) required for the activation of VGCCs, which, in turn, allows the influx of calcium ions needed for CCK release. Furthermore, our data showed that LA-induced depolarization is profoundly reduced by blocking G protein activation and PLC activity by GDPβS and U73122, respectively, thus confirming the role of GPCRs in PUFA signaling in STC-1 cells.

In the present study we also showed for the first time that LA initiates transient receptor potential (TRP)-like rapid inward currents in STC-1 cells. LA-induced inward currents were abolished upon removal of extracellular sodium ions, which is consistent with the dependence of LA-induced depolarization and intracellular calcium rise on extracellular sodium ions. LA-induced inward currents were not affected by altering extracellular calcium ions. Ion substitution studies confirmed that the LA-induced currents were carried largely, if not exclusively, by monovalent cations as the reversal potential for this current was closely predicted by the GHK equation based on the assumption that all monovalent cations were equally permeable (Fig. 3D).

Next, we attempted to discover molecular identity of transmembrane proteins that underlie LA-induced Na⁺-dependent inward currents. STC-1 cells express multiple bitter-activated GPCRs (members of the T2R GPCR family) and signaling components identified in type II taste receptor cells including α-gustducin, PLC-β2, and TRPM5. TRPM5, an ion channel found in several types of chemosensory cells, has been shown to play an instrumental role in bitter, sweet, and umami taste transduction (20, 25, 30). From heterologous expression studies in human embryonic kidney cells, TRPM5 was shown to be activated by intracellular calcium ions and appears largely selective for monovalent cations (16, 21). While the role of TRPM5 in bitter taste transduction is unequivocal, TRPM5 currents in response to bitter compounds have never been recorded in native cells. Our results showed for the first time that application of a bitter compound, denatonium benzoate, activated extracellular sodium-dependent inward currents (presumably via TRPM5). Because the LA-induced TRP-like inward currents were selective for monovalent cations, we investigated the possibility of TRPM5 involvement in LA-induced signaling in STC-1 cells. TPPO, a specific antagonist for TRPM5 (19), significantly reduced LA-induced and, importantly, denatonium benzoate-induced inward currents. We saw very similar levels of inhibition of LA-induced and denatonium benzoate-induced inward currents upon application of TPPO. These findings suggested that LA, like bitter stimuli, may also be activating TRPM5 channels to allow the entry of sodium ions, providing the depolarizing receptor potential. Transfection of siRNA specific to TRPM5 in STC-1 cells showed diminished LA-induced inward currents, which gave us additional verification that LA activates TRPM5 ion channels. Furthermore, transfection of siRNA specific to TRPM5 in STC-1 cells showed diminished intracellular calcium rise and CCK secretion in response to LA. From these studies, we show that TRPM5 provides the basis for the fatty acid-induced receptor potential necessary for the opening of VGCCs to elicit intracellular calcium rise needed for CCK secretion and provide some of the first data on TRPM5 currents activated by ligands in native cells. Furthermore, we discovered that linoleic acid-activated TRPM5 currents were greatly reduced upon knockdown of GPR120 expression. These results clearly implicate that TRPM5 is activated downstream of GPR120, which is also supported by the electrophysiological experiments where LA-induced depolarization was profoundly inhibited when G proteins or PLC activity was blocked. Finally, our results confirm and support the previous findings about GPR120 as a primary receptor in the fatty acid-induced intracellular calcium rise and CCK secretion (9).

Previous studies have clearly established that bitter substances stimulate CCK secretion from STC-1 cells by opening voltage-gated calcium channels (4). In addition, in the present study, we observed that bitter substances (such as DB) clearly elicit TPPO (specific TRPM5 channel blocker)-sensitive inward currents (that may produce receptor potential to open voltage-gated calcium channels) in STC-1 cells. Thus, it is plausible that bitter substances by acting on bitter receptors activate similar intracellular pathways [PLC, inositol trisphosphate (IP₃), and TRPM5 dependent] culminating in CCK release. However, further studies will be required to investigate this possibility.

Finally, the pathway (as shown in Fig. 7) underlying LA (PUFA)-induced CCK release appears to involve the activation of GPR120 and subsequently PLC. The by-products of PLC activation [IP₃ and diacylglycerol (DAG)] either directly or indirectly lead to the activation of TRPM5 and the concomitant sodium influx through these channels, which is responsible for the development of the LA-induced receptor potential. The receptor potential, in turn, activates VGCCs, producing the requisite calcium influx needed for CCK release. The data showing that the activation of TRPM5, presumably by intracellular calcium, mediates the LA-induced receptor potential that eventually leads to the release of the satiety hormone CCK, present a more complete picture of the transduction pathway for fatty acids in the gut. Having elucidated the cellular and molecular underpinnings of this pathway may be important for designing novel approaches that may contribute to the eventual control of fat intake.

Fig. 7.

A model showing LA transduction pathway in STC-1 cells. Fatty acids (FAs) bind and activate G protein-coupled receptors linked to the stimulation of a specific isoform of phospholipase C (PLCβ2) and the generation of two second messengers, diacylglycerol (DAG) and inositol trisphosphate (IP₃). IP₃ binds to IP₃ receptors (IP₃R) on the endoplasmic reticulum and causes the release of intracellular calcium. The intracellular calcium released from stores activates a monovalent cation channel, TRPM5, whose activation provides the FA-induced depolarizing receptor potential via cation influx. This depolarization opens voltage-gated calcium channel (VGCCs), leading to an intracellular calcium rise and eventual release of satiety hormones like glucagon-like peptide-1 (GLP-1) and CCK.

GRANTS

The project described was partially supported by award no. R01 DK-059611 (T. A. Gilbertson) from the National Institute of Diabetes and Digestive and Kidney Diseases. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Diabetes and Digestive and Kidney Diseases or the National Institutes of Health. Additional support was provided by project no. 630 from the Utah Agricultural Experiment Station and from International Flavors & Fragrances.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

B.P.S., P.L., D.R.H., and T.A.G. conception and design of research; B.P.S., P.L., T.Y., and D.R.H. performed experiments; B.P.S., T.Y., D.R.H., and T.A.G. analyzed data; B.P.S., T.Y., D.R.H., and T.A.G. interpreted results of experiments; B.P.S. and T.A.G. prepared figures; B.P.S. and D.R.H. drafted manuscript; B.P.S., P.L., T.Y., D.R.H., and T.A.G. edited and revised manuscript; B.P.S., P.L., T.Y., D.R.H., and T.A.G. approved final version of manuscript.