Orosensory Detection of Dietary Fatty Acids Is Altered in CB1R−/− Mice

Obesity is one of the major public health issues, and its prevalence is steadily increasing all the world over. The endocannabinoid system (ECS) has been shown to be involved in the intake of palatable food via activation of cannabinoid 1 receptor (CB1R). However, the involvement of lingual CB1R in the orosensory perception of dietary fatty acids has never been investigated. In the present study, behavioral tests on CB1R−/− and wild type (WT) mice showed that the invalidation of Cb1r gene was associated with low preference for solutions containing rapeseed oil or a long-chain fatty acid (LCFA), such as linoleic acid (LA). Administration of rimonabant, a CB1R inverse agonist, in mice also brought about a low preference for dietary fat. No difference in CD36 and GPR120 protein expressions were observed in taste bud cells (TBC) from WT and CB1R−/− mice. However, LCFA induced a higher increase in [Ca2+]i in TBC from WT mice than that in TBC from CB1R−/− mice. TBC from CB1R−/− mice also exhibited decreased Proglucagon and Glp-1r mRNA and a low GLP-1 basal level. We report that CB1R is involved in fat taste perception via calcium signaling and GLP-1 secretion.


Introduction
Due to the abundance of food resources in the modern era, the Western diet is comprised of more than 40% of fat, thereby contributing to the increase in the prevalence of obesity that is associated with a number of pathologies (type 2 diabetes mellitus, hypertension, cancer, and others).
The taste modalities represent an essential factor involved in food intake. It is now well established that obese subjects exhibit higher spontaneous preference for fat than lean subjects [1,2]. Recent studies have proposed the existence of a sixth taste modality dedicated to the orosensory perception of dietary fat. The CD36 (cluster of differentiation 36) has been suggested to act as lingual lipid receptor [3]. The binding of a fatty acid to lingual CD36 in taste bud cells (TBC) leads to modifications in the membrane potential and to an increase in free intracellular calcium concentrations, [Ca 2+ ] i , followed by the release of neurotransmitters [4,5]. These gustatory signals are transmitted from the oral cavity, through the cranial nerve IX (lingual branch of the glossopharyngeal), to the nucleus of the solitary tract (NST) [6]. The NST is connected to different brain areas associated with food intake, rewarding, memory, and processes integrating visceral signals [7,8]. Hence, the integration of the gustatory signals in brain triggers a behavioral and metabolic response [8]. GPR120 (G protein-coupled receptor 120) has also been proposed to play a role in fat-related regulation of satiation [8,9]. Nevertheless, GPR120 does not seem to have a major role in oral fat detection. Indeed, contradictory results have been reported for behavioral tests in GPR120 −/− mice [10][11][12]. However, the implication of GPR120 in the release of the incretin hormone glucagon-like peptide-1 (GLP-1) has been highlighted in mouse taste bud cells [9,13]. Thus, CD36 is likely to play a role in the detection of dietary fatty acids in TBC, whereas GPR120 would be implicated in the modulation of postprandial fat taste sensitivity.
The presence of GLP-1 and its receptor in the gustatory mucosa has been demonstrated [9,14,15], suggesting that taste bud cells may modulate taste perception in an autocrine or a paracrine manner. Indeed, linoleic acid (LA) has been reported to induce GLP-1 release in human TBC in a GPR120-dependent manner [9]. Martin et al. [13] have suggested that GLP-1 is locally active and might affect the basic functions in mouse taste buds. Besides, Shin et al. [15] showed that local GLP-1 signaling could enhance sweet-taste sensitivity, supporting the existence of a paracrine mechanism for the regulation of taste function.
Being largely expressed in the central and peripheral nervous system, CB 1 R has also been detected in TBC, and it has been shown that the activation of these receptors by cannabinoids enhances sweet taste [7]. However, the involvement of lingual CB 1 R in fat taste perception has never been investigated. Considering that the increase in dietary fat intake plays an important role in the prevalence of obesity, the present investigation was designed to assess whether the activation of CB 1 R in TBC is associated with the altered orosensory perception of dietary lipids in CB 1 R −/− and wild type (WT) mice.

Ethical Approval
French guidelines for the use and care of laboratory animals were followed, and the experimental protocols were approved by the regional animal ethic committee of the University of Burgundy. In vivo studies were performed on male C57BL/6J wild type (WT) mice (Janvier Labs, Le Genest Saint Isle, France) and CB 1 R -/mice (generous gift from Dr. James Pickel, National Institute of Mental Health, Bethesda, MD, USA) with a C57BL/6J background. Animals were individually housed in a controlled environment (constant temperature and humidity, dark period from 19:00 to 7:00). The mice had free access to standard regular chow and tap water during the experiments, unless otherwise specified.

Two-Bottle Preference Tests
After being deprived of water for 6 h, mice were offered simultaneously two bottles, containing either control or experimental solution for 12 h. To minimize bias due to textural properties, the two solutions contained 0.3% xanthan gum (w/v, Sigma, Saint Quentin-Fallavier, France), whereas the experimental solutions were added with either 0.2% rapeseed oil (w/v, Fleur de Colza, Lesieur, France) or 0.2% linoleic acid (w/v, LA, Sigma). At the end of each test, the intake of control and experimental solutions was recorded by weighing the feeders/bottles. The experiments were repeated two times, independently.
In parallel, two groups (n = 5 each) of 10 WT mice, treated daily with rimonabant (SR141716, 10 mg/kg of body weight, Sanofi, Paris, France) or vehicle (0.1% DMSO/0.025% Tween 80 in 0.9% NaCl), by an intraperitoneal injection for 26 days, were subjected to the same two-bottle preference test. Food intake and weight were monitored during the experiment.

Licking Tests
The CB 1 R −/− (n = 9) and WT (n = 9) mice were deprived of food and water for 6 h before the test. The mice were conditioned to choose between a palatable (4% sucrose) and a control solution. Once the mice were conditioned, they were randomly subjected to two-bottle test, containing either control (0.3% xanthan gum) or a test solution (0.3% xanthan gum + 0.2% linoleic acid, LA). The number of licks, motivated by each bottle, were recorded using computer-controlled lickometers (Med Associates, Fairfax, VT, USA). Data were analyzed for 5 min from the first lick.

Papillae and Taste Buds Isolation
The mice were anesthetized with 2% isoflurane gas, and then sacrificed by cervical dislocation. Taste bud cells (TBC) were isolated according to previously published procedure [4]. In brief, lingual epithelium was separated from connective tissues by enzymatic dissociation (elastase and dispase mixture, 2 mg/mL each in Tyrode buffer: 120 mM NaCl, 5 mM KCl, 10 mM HEPES, 1 mM CaCl 2 , 10 mM glucose, 1 mM MgCl 2 , 10 mM Na pyruvate, pH 7.4). Samples were frozen immediately in liquid nitrogen and stored at −80 • C (not exceeding one month) until RNA extraction, or lysed in a buffer for Western blot analyses. For real-time qPCR and Western blot, each point corresponds to a pool of TBC from four mice.

Real-Time qPCR
Total RNA from CB 1 R −/− and WT TBC (n = 6) was extracted by using TRIzol method according to the manufacturer's recommendations (Invitrogen, Cergy-Pontoise, France). After purification, mRNA was resuspended in RNase free water. The samples were then analyzed and quantified using Traycell (Hellma Analytics, Müllheim, Germany). Samples having a purity (A 260 /A 280 ) between 1.80 and 2.00 were retained for the rest of the experiment. mRNA (500 ng) was reverse-transcribed into cDNA using M-MLV reverse transcriptase (Invitrogen) in a 20 µL of reaction volume containing 5 µL mRNA, 1 µL random primer (100 ng/µL) (Invitrogen), 0.4 µL dNTP (25 mM), 8.6 µL RNase-free water. After incubation for 5 min at 65 • C, 2 µL 5× First-Strand Buffer, 1 µL DTT, 1 µL M-MLV (200 UI) and 1 µL RNaseOUT were added, and the samples were incubated for 1 h at 42 • C and then for 15 min at 70 • C. Real time qPCR reactions were performed on 10 ng cDNA in a 20 µL of reaction volume in triplicates with a StepOnePlus (Life Technologies, Saint-Aubin, France) device with the use of SYBR green PCR Master Mix (Life Technologies, Saint-Aubin, France). For each gene, a standard curve was established from five cDNA dilutions (50 ng to 0.05 ng per well) and used to determine the PCR efficiency. Forward and reverse primer sequences used for amplification were 5 -GGACACATGAAGTCATCTTTGCCT-3 and 5 -CAAGCCCTGGAAGGAAGTGAAGGA-3 for Glp-1r (NM_021332), 5 -TGCTGAAGG GACCTTTACCAGTGA-3 and 5 -GCCTTTCACCAGCCAAGCAATGAA-3 for Gcg (NM_008100), and 5 -TTCTTTGCAGCTCCTTCGTT-3 and 5 -ATGGAGGGGAATACAGCCC-3 for β-actin (NM_007393). The amplicon size for Glp-1r is 107 bp and is located in the exon 11 and 12; for Gcg is 85 bp and is located in the exon 4; for β-actin is 149 bp and is located in the exon 1 and 2. Real time qPCR reactions were performed with a denaturing step of 95 • C for 10 min, followed by 40 cycles of 95 • C for 15 s and 60 • C for 1 min. The primer specificity was checked using the melt curves. The PCR efficiency was calculated as follow 10 −1/slope − 1. The parameters for GLP-1r were as follows: slope −3.241, y intercept 33.848, R 2 0.98, and PCR efficiency 1.03; Gcg: slope −3.527, y intercept 29.296, R 2 0.92, and PCR efficiency 0.92; and β-actin: slope −2.837, y intercept 29.296, R 2 0.996, and PCR efficiency 1.25. The comparative 2 −∆∆CT method was used for relative quantification.

Tissue Culture of TBC and GLP-1 Release
Papillae from WT and CB 1 R −/− mice were isolated and incubated at 36 • C. The incubation media contained either 33 µM fatty acid-free BSA alone (control group) or 200 µM linoleic acid (LA) mixed and vortexed with 33 µM fatty acid-free BSA. After 2 h of incubation, the media were collected, and the active GLP-1 release was measured by ELISA (Millipore S.A.S., Molsheim, France). As the secretion of GLP-1 by TBC is very low, to be sure to detect active GLP-1 in the incubation medium, 10 pM of pure GLP-1 was systematically added in each experimental well, but not in standard curve, according to the manufacturers' recommendations. The dipeptidyl peptidase 4 (DPP4) inhibitor (0.1%, Millipore) was added to the medium to prevent GLP-1 degradation.

Measurement of Ca 2+ Signaling
TBC were freshly isolated from mouse tongues as described by Dramane et al. [26]. The cells were cultured onto 24-well plates, containing RPMI-1640 medium, supplemented with 10% fetal calf serum, 2 mM glutamine, 50 µg/mL penicillin-streptomycin, and 20 mM HEPES, and incubated overnight at 37 • C. The next day, the supernatant was discarded. The cells were then incubated with Fura-2/AM (Invitrogen) at 1 µM for 30 min at 37 • C in loading buffer which contained the following: 110 mM NaCl, 5.4 mM KCl, 25 mM NaHCO 3 , 0.8 mM MgCl 2 , 0.4 mM KH 2 PO 4 , 20 mM Hepes, 1.2 mM CaCl 2 , 10 mM Glucose; pH 7.4. After adding the test molecules into the wells, the changes in intracellular free Ca 2+ (F 340 /F 380 ) were monitored under the Nikon microscope (TiU) by using S-fluor 40× oil immersion objective. NIS-Elements software was used to record the images. The microscope was equipped with Lucas EM-CCD (Andor Technology, Gometz-le-châtel, France) camera for real-time recording of 16-bit digital images. The dual excitation fluorescence imaging system was used to analyze individual cells. The changes in intracellular free Ca 2+ were expressed as ∆Ratio, calculated as the difference between F 340 and F 380 . All test molecules were added in small volumes with no interruption in recordings. For Ca 2+ signaling experiments, the fatty acid was dissolved in ethanol (0.1%, v/v) and added into the experimental cuvette.

Statistics
Results are expressed as means ± SEM. The significance of differences between groups was evaluated with GraphPad Prism (GraphPad Software, La Jolla, CA, USA) using two-tailed Student's t-test or two-way ANOVA with Bonferroni correction. A p value of less or equal 0.05 was considered to be statistically significant.

The Absence of CB 1 R Gene Induces a Low Preference for Fatty Solutions Independently of Postprandial Factors
CB 1 R −/− mice displayed a significant decrease in the preference for the fatty solutions (rapeseed oil and LA) compared to wild type mice (Figure 1a,b). Hormonal post-ingestive regulatory feedback greatly influences metabolism and appetite and, consequently, behavior during long-term two-bottle preference tests [27] under CB 1 R activation [28]. Therefore, we performed short-term licking tests to measure fat preference limiting post-ingestive cues. LA was chosen among other LCFAs because it showed the best results in two-bottle preference tests. As shown in Figure 1c, the number of licks for LA was significantly higher in WT mice than in CB 1 R −/− mice, confirming the low preference in CB 1 R −/− mice for fatty solutions. t-test or two-way ANOVA with Bonferroni correction. A p value of less or equal 0.05 was considered to be statistically significant.

The Absence of CB1R Gene Induces a Low Preference for Fatty Solutions Independently of Postprandial Factors
CB1R −/− mice displayed a significant decrease in the preference for the fatty solutions (rapeseed oil and LA) compared to wild type mice (Figure 1a,b). Hormonal post-ingestive regulatory feedback greatly influences metabolism and appetite and, consequently, behavior during long-term two-bottle preference tests [27] under CB1R activation [28]. Therefore, we performed short-term licking tests to measure fat preference limiting post-ingestive cues. LA was chosen among other LCFAs because it showed the best results in two-bottle preference tests. As shown in Figure 1c, the number of licks for LA was significantly higher in WT mice than in CB1R −/− mice, confirming the low preference in CB1R −/− mice for fatty solutions.

Treatment with Rimonabant Induces a Low Preference for Fat Solutions and Does Not Alter Feeding Behavior
The mice treated with rimonabant for 26 days exhibited the same behavior as CB1R −/− mice, i.e., a low preference for fatty solutions (Figure 2a,b). Interestingly, the mice treated with rimonabant did not show reduced food intake or body weight ( Figure 2c). As the rimonabant induces an early and transient effect, we monitored these parameters for 5 days before starting two-bottle preference tests. Thus, no bias interfered with the behavioral experiments.

Treatment with Rimonabant Induces a Low Preference for Fat Solutions and Does Not Alter Feeding Behavior
The mice treated with rimonabant for 26 days exhibited the same behavior as CB 1 R −/− mice, i.e., a low preference for fatty solutions (Figure 2a,b). Interestingly, the mice treated with rimonabant did not show reduced food intake or body weight (Figure 2c). As the rimonabant induces an early and transient effect, we monitored these parameters for 5 days before starting two-bottle preference tests. Thus, no bias interfered with the behavioral experiments.

CD36 and GPR120 Protein Expressions Are Not Altered in TBC of CB1R −/− Mice
CB1R gene invalidation did not interfere with CD36 and GPR120 protein expression in taste buds from CB1R −/− (Figure 3). α-Gustducin, a marker of type II TBC, remained stable ( Figure 3). It seems that the low preference for fatty solutions observed in CB1R −/− mice was not due to altered expression of CD36 and GPR120.  3.3. CD36 and GPR120 Protein Expressions Are Not Altered in TBC of CB 1 R −/− Mice CB 1 R gene invalidation did not interfere with CD36 and GPR120 protein expression in taste buds from CB 1 R −/− (Figure 3). α-Gustducin, a marker of type II TBC, remained stable ( Figure 3). It seems that the low preference for fatty solutions observed in CB 1 R −/− mice was not due to altered expression of CD36 and GPR120.

CD36 and GPR120 Protein Expressions Are Not Altered in TBC of CB1R −/− Mice
CB1R gene invalidation did not interfere with CD36 and GPR120 protein expression in taste buds from CB1R −/− (Figure 3). α-Gustducin, a marker of type II TBC, remained stable ( Figure 3). It seems that the low preference for fatty solutions observed in CB1R −/− mice was not due to altered expression of CD36 and GPR120.

CB 1 R Gene Invalidation Induces a Decrease in Proglucagon and GLP-1r mRNA and Basal GLP-1 Level
Proglucagon and GLP-1r mRNA levels were significantly lower in CB 1 R −/− TBC than WT TBC (Figure 4a). According to previously published data [9], LA induces the release of GLP-1 from mouse TBC. To measure the release of active GLP-1, the mouse TBC were incubated for 2 h in an oxygenized medium containing anti-DPP4, to prevent GLP-1 degradation, and exposed, or not, to 200 µM LA. In CB 1 R −/− mice TBC, GLP-1 release in the culture medium was significantly lower than that in WT TBC, in both basal and LA-stimulated conditions. As expected, LA induced a small, but significant, release of active GLP-1 in culture medium of WT TBC (Figure 4b).

CB1R Gene Invalidation Induces a Decrease in Proglucagon and GLP-1r mRNA and Basal GLP-1 Level
Proglucagon and GLP-1r mRNA levels were significantly lower in CB1R −/− TBC than WT TBC (Figure 4a). According to previously published data [9], LA induces the release of GLP-1 from mouse TBC. To measure the release of active GLP-1, the mouse TBC were incubated for 2 h in an oxygenized medium containing anti-DPP4, to prevent GLP-1 degradation, and exposed, or not, to 200 μM LA. In CB1R −/− mice TBC, GLP-1 release in the culture medium was significantly lower than that in WT TBC, in both basal and LA-stimulated conditions. As expected, LA induced a small, but significant, release of active GLP-1 in culture medium of WT TBC (Figure 4b). Each value corresponds to the GLP-1 released by cultured TBC. We independently reproduced the results twice, by using, each time, TBC from three mice. We observed identical results, and we pooled them. Each point represents values as mean ± SD (n = 6).

Both LA and Cannabinoids Induce CB1R-Dependent Ca 2+ Responses in TBC
In mouse TBC, LCFA evokes increases in Ca 2+ signaling [26]. As expected, LA triggered a higher rise in [Ca 2+ ]i in WT TBC [6] (Figure 5a,f) than that in CB1R −/− TBC (Figure 5b,f). Similarly, arachidonyl-2′-chloroethylamide (ACEA, a specific CB1R agonist), induced a rise in [Ca 2+ ]i in WT TBC (Figure 5f). However, ACEA triggered a significantly lower increase in [Ca 2+ ]i in CB1R −/− TBC than that in WT TBC (Figure 5f). Finally, the combination of LA and ACEA also induced a strong rise in [Ca 2+ ]i in WT TBC (Figure 5c,f) corresponding to the combined effect of the two molecules that was not apparent in CB1R −/− TBC (Figure 5d,f). Interestingly, when tested on CB1R −/− TBC, ACEA still triggered a rise in [Ca 2+ ]i (Figure 5f). Furthermore, blockade of TRPV1 with a specific antagonist, A784168, in CB1R −/− TBC curtailed the action of ACEA on calcium response (Figure 5e). Each value corresponds to the GLP-1 released by cultured TBC. We independently reproduced the results twice, by using, each time, TBC from three mice. We observed identical results, and we pooled them. Each point represents values as mean ± SD (n = 6).

Both LA and Cannabinoids Induce CB 1 R-Dependent Ca 2+ Responses in TBC
In mouse TBC, LCFA evokes increases in Ca 2+ signaling [26]. As expected, LA triggered a higher rise in [Ca 2+ ] i in WT TBC [6] (Figure 5a,f) than that in CB 1 R −/− TBC (Figure 5b,f). Similarly, arachidonyl-2 -chloroethylamide (ACEA, a specific CB 1 R agonist), induced a rise in [Ca 2+ ] i in WT TBC (Figure 5f). However, ACEA triggered a significantly lower increase in [Ca 2+ ] i in CB 1 R −/− TBC than that in WT TBC (Figure 5f). Finally, the combination of LA and ACEA also induced a strong rise in [Ca 2+ ] i in WT TBC (Figure 5c,f) corresponding to the combined effect of the two molecules that was not apparent in CB 1 R −/− TBC (Figure 5d,f). Interestingly, when tested on CB 1 R −/− TBC, ACEA still triggered a rise in [Ca 2+ ] i (Figure 5f). Furthermore, blockade of TRPV1 with a specific antagonist, A784168, in CB 1 R −/− TBC curtailed the action of ACEA on calcium response (Figure 5e).

CB1R Blockade Significantly Decreases Ca 2+ Responses Triggered by LA, AEA, and ACEA in WT TBC
To further explore the role of CB1R on calcium signaling, we preincubated WT TBC with a specific CB1R inverse agonist rimonabant. As observed in Figure 6, rimonabant significantly abrogated [Ca 2+ ]i responses, induced by LA, AEA, and ACEA (Figure 6c), corroborating the previous results observed in CB1R −/− TBC.

CB 1 R Blockade Significantly Decreases Ca 2+ Responses Triggered by LA, AEA, and ACEA in WT TBC
To further explore the role of CB 1 R on calcium signaling, we preincubated WT TBC with a specific CB 1 R inverse agonist rimonabant. As observed in Figure 6, rimonabant significantly abrogated [Ca 2+ ] i responses, induced by LA, AEA, and ACEA (Figure 6c), corroborating the previous results observed in CB 1 R −/− TBC.

AEA-Induced Ca 2+ -Signaling Is PLC Dependent
Finally, in order to validate our model, we investigated the downstream CB1R pathway elicited by AEA treatment. Here, we used DB-cAMP, a cAMP analog and a phosphodiesterase inhibitor, and U73122, a phospholipase C inhibitor. We observed that DB-cAMP did not significantly alter the Ca 2+ response induced by AEA. Conversely, a pretreatment with U73122 significantly decreased the Ca 2+ response, indicating that the Ca 2+ response triggers a PLC-dependent Ca 2+ signaling (Figure 7).

AEA-Induced Ca 2+ -Signaling Is PLC Dependent
Finally, in order to validate our model, we investigated the downstream CB 1 R pathway elicited by AEA treatment. Here, we used DB-cAMP, a cAMP analog and a phosphodiesterase inhibitor, and U73122, a phospholipase C inhibitor. We observed that DB-cAMP did not significantly alter the Ca 2+ response induced by AEA. Conversely, a pretreatment with U73122 significantly decreased the Ca 2+ response, indicating that the Ca 2+ response triggers a PLC-dependent Ca 2+ signaling (Figure 7).

AEA-Induced Ca 2+ -Signaling Is PLC Dependent
Finally, in order to validate our model, we investigated the downstream CB1R pathway elicited by AEA treatment. Here, we used DB-cAMP, a cAMP analog and a phosphodiesterase inhibitor, and U73122, a phospholipase C inhibitor. We observed that DB-cAMP did not significantly alter the Ca 2+ response induced by AEA. Conversely, a pretreatment with U73122 significantly decreased the Ca 2+ response, indicating that the Ca 2+ response triggers a PLC-dependent Ca 2+ signaling (Figure 7).   [19] demonstrated that CB 1 R is responsible for increased food intake, induced by an endocannabinoid agonist. Later on, Yoshida et al. [7] revealed that endocannabinoids enhanced the gustatory responses to sweet tastants via CB 1 R. Indeed, the activation of endocannabinoid system (ECS) appears to be associated with hyperphagia and a preference for palatable food. Interestingly, CB 1 R are also expressed in a subset of taste bud cells [7]. We report here that CB 1 R −/− mice displayed no preference for fat solutions compared to WT mice. The same behavior was also observed when WT mice were treated with rimonabant, a CB 1 R blocker, confirming the role for this receptor in the detection of dietary lipids. We have employed LA as a candidate for LCFA because this fatty acid is abundantly present in Western food; however, it is possible that the saturated fatty acids like palmitic acid (PA) might also initiate the same gustatory response. Indeed, we have shown previously that LA and PA triggered the same increases in [Ca 2+ ] i in mouse taste bud cells [6].

Williams and Kirkham
In the present study, for behavioral experiments, we used whole body knockout mice for CB 1 R, and it is possible that the hypothalamic cannabinoid system, via the dopaminergic area, might be involved in fat taste preference [28]. Nonetheless, we sought to elucidate cellular mechanisms in the modulation of fat preference. We first tested the hypothesis whether there is an alteration in CD36 and GPR120 protein in TBC of CB 1 R −/− mice. In our study, CD36 and GPR120 protein expressions were not altered by the absence of CB 1 R, suggesting that the absence of preference for fatty solutions may be due to altered downstream signaling. Moreover, we checked the delivery of linoleic acid under both conditions, and we observed identical uptake of exogenous fatty acid.
Previous studies indicated that both CD36 and GPR120 activation by a LCFA triggered mobilization of [Ca 2+ ] i from the intracellular endoplasmic reticulum Ca 2+ pool during fat taste perception [9,29]. In our study, we show, for the first time, that LA-mediated increase in [Ca 2+ ] i was altered when CB 1 R was inactivated by rimonabant or by the absence of CB 1 R. In addition, the CB 1 R agonist ACEA also increased calcium flux per se in TBC, albeit with lower potency than LA. However, the effect of ACEA was maintained in TBC from CB 1 R −/− mice, raising the possibility that the increase in [Ca 2+ ] i could be mediated by the receptors other than CB 1 R, for example, TRPV1. Indeed, it has been shown that activation of TRPV1 by endocannabinoids induces calcium signaling [30,31]. Besides, blockade of TRPV1 with A784168 totally abolished [Ca 2+ ] i response induced by ACEA, indicating that the residual calcium signal observed in CB 1 R −/− TBC with ACEA may be due to TRPV1 activity. Furthermore, it appears that the CB 1 R-coupled downstream signaling is PLC-dependent, in accordance with the observations of De Petrocellis et al. [32]. However, it remains to be elucidated in future whether anadamide, employed in the present study, activates the Gβγ subunit of CB 1 R, and activates PLC via PI-3-kinase pathway. As a whole, our data indicate that CB 1 R may play a crucial role in fat taste perception by modulating calcium signaling.
As previously described, GLP-1 −/− mice have reduced taste responses to dietary fat, suggesting that orosensory detection of LCFA could be associated to the secretion of lingual GLP-1 [13]. Data reported herein showed that the secretion of active GLP-1 induced by LA is strongly decreased in CB 1 R −/− mice suggesting the existence of a link between CB 1 R signaling and GLP-1 production. Hence, CB 1 R activation may stimulate proglucagon and GLP-1r production and, therefore, modulate perception threshold of LCFA. Further investigations are needed to explore the possibility whether GLP-1 secretion is stimulated via [Ca 2+ ] i signaling in TBC or by other mechanisms [33].
In conclusion, the present report shows that CB 1 R influences fat taste perception via regulating calcium signaling in TBC. It is proposed that CB 1 R activation induces a [Ca 2+ ] i response that strengthens fat perception, that is mediated by CD36. Activation of ECS could, thereby, increase sensory stimuli relaying palatability of foods and, ultimately, stimulate food intake. The physiopathological relevance of such a regulatory pathway is supported by the fact that ECS tone is increased in obesity. Hence, the ECS seems to emerge as a key modulator of oral sweet and fat detection and may represent a potential target for developing new anti-obesity strategies or, conversely, for enhancing food intake in the case of loss of appetite as it occurs in cachexia.