Research article

Gut bitter taste receptor signalling induces ABCB1 through a mechanism involving CCK

Tae-Il Jeon, Young-Kyo Seo, Timothy F. Osborne


T2Rs (bitter taste-sensing type 2 receptors) are expressed in the oral cavity to prevent ingestion of dietary toxins through taste avoidance. They are also expressed in other cell types, including gut enteroendocrine cells, where their physiological role is enigmatic. Previously, we proposed that T2R-dependent CCK (cholecystokinin) secretion from enteroendocrine cells limits absorption of dietary toxins, but an active mechanism was lacking. In the present study we show that T2R signalling activates ABCB1 (ATP-binding cassette B1) in intestinal cells through a CCK signalling mechanism. PTC (phenylthiocarbamide), an agonist for the T2R38 bitter receptor, increased ABCB1 expression in both intestinal cells and mouse intestine. PTC induction of ABCB1 was decreased by either T2R38 siRNA (small interfering RNA) or treatment with YM022, a gastrin receptor antagonist. Thus gut ABCB1 is regulated through signalling by CCK/gastrin released in response to PTC stimulation of T2R38 on enteroendocrine cells. We also show that PTC increases the efflux activity of ABCB1, suggesting that T2R signalling limits the absorption of bitter tasting/toxic substances through modulation of gut efflux membrane transporters.

  • ATP-binding cassette B1 (ABCB1) efflux transporter
  • bitter taste-sensing type 2 receptor (T2R)
  • cholecystokinin (CCK)
  • enteroendocrine cell
  • gut efflux membrane transporter


Because many dietary toxins are sensed by mammals as tasting ‘bitter’, a taste aversion to bitterness has evolved to prevent the ingestion of potentially toxic substances from the diet. A family of single-exon GPCRs (G-protein-coupled receptors) called T2Rs (bitter taste-sensing type 2 receptors) are expressed in taste buds of the tongue and adjacent oral epithelium where they ‘sense’ bitter food components as agonists [1]. The expression of T2Rs has also been localized to the gastric and intestinal mucosa where their physiological function is less clear [2].

T2R signalling in cultured enteroendocrine cells results in secretion of gut polypeptide hormones such as CCK (cholecystokinin) and GLP-1 (glucagon-like peptide-1) [3]. Indeed, we have previously shown that T2R expression and signalling in cultured murine enteroendocrine cells and in the mouse intestine are regulated by the cholesterol-sensing SREBP-2 (sterol-regulatory-element-binding protein-2) protein, which results in cholesterol-regulated secretion of CCK [4]. These observations suggested that intestinal T2R activity and CCK secretion are regulated by dietary composition and, based on the known actions of CCK in limiting intestinal motility and food intake, we proposed that T2R-mediated CCK secretion is important in a gut-sensing mechanism to help limit the absorption of dietary-derived bitter-tasting toxins that escape taste aversion in the mouth [4]. We further hypothesized a more direct role for CCK in limiting absorption of bitter-tasting toxic substances through actively limiting their cellular uptake through gut efflux membrane transporters.

In the present paper we report that the bitter taste receptor T2R38 is a functional modulator of the ABCB1 (ATP-binding cassette B1) efflux transporter in both Caco-2 human intestinal cells and mouse intestine. ABCB1 is a member of the ABC transporters and is strategically expressed on the apical membrane of intestinal epithelial cells and is thus positioned in the expected location to limit absorption of toxic substrates contained in food. Known ABCB1 substrates include xenobiotics and many chemotherapeutic agents [5], and increased ABCB1 activity explains why many cancer cells become resistant to chemotherapy [6].

The results of the present study indicate that T2R signalling in enteroendocrine cells results in the secretion of CCK that stimulates enterocytes to produce more ABCB1. We show further that the T2R38 bitter agonist PTC (phenylthiocarbamide) is a substrate for ABCB1. Taken together, these results suggest a paracrine signalling mechanism whereby T2R-stimulated CCK secretion from enteroendocrine cells increases ABCB1-mediated efflux from the neighbouring intestinal epithelial cells.



Caco-2 cells provided by Dr H.M. Said (Department of Medicine and Physiology/Biophysics, University of California at Irvine, Irvine, CA, U.S.A.) and AGS-E gastric adenocarcinoma cells (stably expressing the CCK2 receptor) provided by Dr T.C. Wang (Department of Medicine, Columbia University Medical Center, New York, NY, U.S.A.) were maintained in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10% heat-inactivated FBS (fetal bovine serum) and antibiotics (penicillin/streptomycin) in an atmosphere of 5% CO2 at 37 °C. Puromycin (2 μg/ml) was used for selection in AGS-E cells [7].


The human ABCB1 promoter fragment (−1780 to +119) was amplified by PCR using human genomic DNA as a template, and cloned into the pGL3-Basic firefly luciferase reporter vector (Promega). The construct was verified by DNA sequencing.

siRNA (small interfering RNA) transfection in Caco-2 cells

The siRNA targeting human T2R38 (GenBank® accession number NM_176817) and a non-targeting siRNA control were purchased from Dharmacon. The cells were transfected for 48 h with 100 nM of each siRNA by using X-tremeGENE siRNA transfection reagent (Roche), according to the manufacturer's instructions, and then cells were incubated with PTC (Sigma–Aldrich) for 9 h.

RNA analysis

Total RNA was extracted from Caco-2 cells and intestine with TRIzol® (Invitrogen) and Qiagen RNeasy isolation kits and used for RT (reverse transcriptase)-qPCR (quantitative PCR) with an iQ5 real-time PCR detection system (Bio-Rad Laboratories). Primer sequences used in the present study are shown in Supplementary Table S1 (at mRNA levels were normalized to expression of GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA as a control and were calculated using the comparative threshold cycle method.


Cells and intestine homogenates were centrifuged at 1000 g for 5 min at 4 °C. Pellets were resuspended in extraction buffer [10 mM Tris/HCl (pH 8.6), 0.14 M NaCl, 1.6 mM MgCl2, 1% Nonidet P40 and protease inhibitor cocktail] and incubated on ice for 5 min, followed by centrifugation at 20000 g for 30 min at 4 °C. The supernatants were separated by SDS/PAGE (6.5% gel) and then transferred on to nitrocellulose membranes and subjected to immunoblotting using monoclonal antibodies against ABCB1 (C219, Covance), and β-actin (Sigma–Aldrich). The blots were visualized using ECL (enhanced chemiluminesence) detection reagents (GE Healthcare).

CCK secretion

Cells were rinsed with HBSS (Hanks balanced salt solution, Invitrogen) supplemented with 20 mM Hepes and 0.1% BSA, and 10 mM PTC dissolved in HBSS was then added to the culture. Cells were then incubated at 37 °C for 45 min. The medium was collected and centrifuged at 4 °C for 5 min at 1000 g to remove cell debris and the supernatants were stored at −20 °C. CCK was measured by ELISA (Phoenix Pharmaceuticals).

Transient transfection in AGS-E cells

AGS-E cells (2×105 cells/well) seeded in a 24-well plate were transfected with a luciferase reporter for the human ABCB1 promoter (−1780 to +119) using Lipofectamine™ 2000 reagent (Invitrogen). A pCMV (cytomegalovirus)-β-gal (β-galactosidase) expression construct was included in every transfection as a normalization control. At 24 h post-transfection, gastrin (1 μM) was added to the cells. After 3 h, cells were harvested and assayed for luciferase and β-gal activities.

Rhodamine 123 accumulation

Caco-2 cells (1×106 cells/well) seeded in six-well plates with sterile coverslips were pre-incubated for 9 h with 10 mM PTC, and then rhodamine 123 (10 μM) dissolved in HBSS was added to the cells after washing. After incubation for 1 h, the cells were thoroughly washed with HBSS. Coverslips were mounted using Vectashield (Vector laboratories), and images were acquired by an Axioskop inverted microscope using the AxioVision camera and software (Carl Zeiss). The fluorescence intensity of rhodamine 123 in cell lysates was measured at 500 nm (excitation) and 550 nm (emission) using the Gemini XPS fluorescence microplate reader (Molecular Devices).

Flow cytometric analysis

FACS analysis was performed with MDR1 (multidrug-resistance 1; an alternative name for ABCB1) shift assay kit (Millipore). Briefly, cells were incubated with PTC or VIN (vinblastine) at 37 °C for 10 min, followed by addition of the IgG2a or UIC2 monoclonal anti-ABCB1 antibody. After 15 min, cells were analysed using a FACSCalibur (BD Biosciences). Flow cytometric data were analysed with FlowJo software (Tree Star).

Gavage administration of PTC

All animal experiments were performed in accordance with accepted standards of animal welfare and with permission of the Sanford-Burnham Medical Research Institute at Lake Nona, IACUC (protocol 09-117). Male FVB mice (6 weeks old) were purchased from Taconic Farms and fed on a normal rodent chow diet and allowed to adapt for 2 weeks to a 12 h light/12 h dark cycle. A mixture of T2R agonists/bitter compounds (10 mM PTC, 10 mM denatoneum benzoate, 1.5 mM quinine and 5 mM D-[−]salicin) in sterilized DW (distilled water) was administered by oral gavage to mice and DW was delivered to the control group. The CCK receptor antagonist YM022 (250 μM) was administered by intraperitoneal injection 30 min before oral gavage of the BCM (bitter compounds mixture). Mice were killed by CO2 asphyxiation. Blood samples were collected from the heart after 1 h gavage, and the plasma CCK concentration was measured by ELISA (Phoenix Pharmaceuticals). For RNA and protein isolation, intestines were collected after 16 h gavage.

Statistical analysis

The results are presented as means±S.E.M. Differences between the means of the individual groups were assessed by one-way ANOVA with Dunnet's multiple comparison test and Student's t test; differences were considered significant at a P value<0.05. The statistical software package Prism 5.0 (GraphPad Software) was used for these analyses.


Although previous studies suggest that T2R signalling in the gut may limit absorption of bitter (and potentially toxic) dietary constituents, a mechanism for this protective response has been lacking. We hypothesized that it might involve membrane transporters that either limit influx or enhance efflux of dietary toxins. To evaluate this, we first surveyed several intestinal-derived cell lines and showed that the human epithelial colorectal cell line Caco-2 expresses both T2Rs and several intestinal efflux transporters. Therefore we treated Caco-2 cells with PTC, a bitter tastant that is an agonist for the receptor T2R38 [8]. Both the efflux transporter ABCB1 mRNA and protein levels were dramatically elevated by the addition of PTC in the Caco-2 cells (Figures 1A and 1B). ABCB1, a member of the ABC-type membrane transporters, is well known to limit the absorption of xenobiotics [9], and expression of other ABC efflux transporters known to be expressed in the intestine, including ABCC1 and ABCG2, were not affected by PTC (Figure 1A). There are two possible explanations for these findings (see below). ABCB1 gene expression is activated through xenobiotic agonists for the PXR (pregnane X receptor) nuclear receptor or T2R38 signalling. However, it is unlikely that the PTC induction of ABCB1 is through PXR because CYP3A4 (cytochrome P450, family 3, subfamily A, polypeptide 4), another PXR target gene involved in xenobiotic detoxification [10], was not induced. In contrast, PTC induction of ABCB1 is likely to be through T2R signalling in Caco-2 cells because induction was prevented when cells were pre-treated with a siRNA against T2R38 (Figures 1C and 1D). In addition, Caco-2 cells expressed Ca2+ signalling components [α-gustducin, PLCβ2 (phospholipase C β2) and TRPM5 (transient receptor potential cation channel subfamily M member 5)] (Supplementary Figure S1 at

Figure 1 ABCB1 is induced by PTC through T2R38 in Caco-2 cells

(A) Caco-2 cells were incubated with PTC (10 mM) for 9 h, and mRNA levels for intestinal transporters and T2R38 were analysed by RT-qPCR. OATP, organic anion-transporting polypeptide. (B) ABCB1 protein was analysed by immunoblotting. (C and D) Caco-2 cells were transfected with a siRNA targeting human T2R38 or a negative control siRNA (100 nM). At 48 h post-transfection, cells were treated with PTC for 9 h. (C) mRNA levels of T2R38 and ABCB1 were analysed by RT-qPCR. (D) ABCB1 protein was analysed by immunoblotting. 1, control; 2, siRNA T2R38; 3, siRNA negative control. In (A and C), results are means±S.E.M., n=3 for three separate experiments. *P value<0.05.

Because bitter agonist signalling increases intracellular Ca2+ levels and stimulates CCK and GLP-1 secretion from enteroendocrine cells [3,4] and intestinal efflux transporters are expressed in the enterocytes [11,12], we hypothesized that gut peptides secreted by T2R signalling might regulate ABCB1 through a paracrine mechanism involving induction of either CCK or GLP-1. This hypothesis predicts that Caco-2 is a heterogenous cell line with some cells that secrete gut peptides in response to bitter agonist signalling and some cells that express the functional peptide receptor. To address this, we first showed that CCK secretion from Caco-2 cells was stimulated by PTC treatment (Figure 2A). STC-1 and HEK-293T [HEK (human embryonic kidney)-293 cells expressing the large T-antigen of SV40 (simian virus 40)] cells were analysed as positive and negative controls respectively [3]. This result suggests that Caco-2 cells possess at least some properties attributed to enteroendocrine cells and, in support of this hypothesis, we showed that Caco-2 also expresses chromogranin A, a well-established marker for enteroendocrine cells [13] (Supplementary Figure S2 at We also analysed the expression of mRNAs encoding known receptors for both CCK and GLP in Caco2 cells by RT-qPCR. mRNA for the CCK2R (CCK2 receptor) was expressed in Caco-2 cells, whereas mRNAs for other receptors were at near background levels (Figure 2B). Thus Caco2 cells might secrete and respond to a CCK2R agonist such as CCK itself or the related peptide hormone gastrin [14]. To determine whether PTC activation of ABCB1 might be mediated by CCK2R signalling, we analysed the effects of YM022, a CCK2R antagonist, on the PTC-dependent induction of ABCB1. This inhibitor blunted the induction of ABCB1 by PTC (Figure 2C). Moreover, we demonstrated that gastrin directly induced expression of ABCB1 mRNA through its transcriptional activation (Figure 3). In order to determine whether ABCB1 is also regulated by T2R signalling in vivo, the plasma CCK level and intestinal ABCB1 expression was measured after delivery of BCM by oral gavage. Delivery of the BCM significantly increased the plasma levels of CCK as well as intestinal expression of ABCB1 mRNA and protein (Figure 4). Importantly, inclusion of the CCK receptor antagonist YM022 completely blocked the induction of intestinal ABCB1 expression by BCM. Taken together, these observations suggest that the intestinal efflux transporter ABCB1 is regulated by bitter-tasting agonists through a signalling mechanism mediated by CCK release from enteroendocrine cells in response to T2R38 signalling.

Figure 2 PTC induction of ABCB1 is mediated by CCK2R in Caco-2 cells

(A) Three cell lines were treated with different concentrations of PTC for 45 min at 37 °C. The medium was collected and analysed for CCK by ELISA. (B) mRNA levels for receptors of CCKs and GLPs in Caco-2 cells were analysed by RT-qPCR. N.D., not detected. (C) Caco-2 cells were pre-treated with the CCK2R antagonist YM022 (10 μM) for 2 h, and then PTC was added to the cultured cells for 9 h. The mRNA level for ABCB1 was analysed by RT-qPCR. Results are means±S.E.M., n=3 for three separate experiments.

Figure 3 Gastrin induces ABCB1 through its promoter activation

(A) Caco-2 cells were incubated with different concentration of gastrin as indicated in the Figure. (B) Caco-2 cells were incubated with gastrin (1 μM) in the presence or absence of YM022 (10 μM), and the mRNA level for ABCB1 was analysed by RT-qPCR. (C) AGS-E cells were transfected with the −1923 ABCB1 reporter, and then gastrin (1 μM) was added to the cells in the presence or absence of YM022. Luciferase activities were normalized to β-gal activity. Results are means±S.E.M., n=3 for three separate experiments. *P value<0.05. GAS, gastrin.

Figure 4 Bitter agonists increase the plasma CCK level and ABCB1 in mouse intestine

BCM (10 mM PTC, 10 mM denatoneum benzoate, 1.5 mM quinine and 5 mM D-[−]salicin) was introduced to FVB mice through oral gavage. The CCKR antagonist YM022 (250 μM) was administered 30 min before oral gavage of BCM. After 1 h of gavage, the plasma CCK concentration was measured by ELISA (A). Intestine was collected 16 h later, and both mRNA and protein levels for ABCB1 was analysed by RT-qPCR (B) and immunoblotting (C) respectively. Immunoblotting results are shown for three individual mice of each group. Results are means±S.E.M., n=4 for three separate experiments. 1, control; 2, BCM; 3, YM022; 4, YM022+BCM. TR, transferrin receptor.

If T2R signalling increases the functional activity of the ABCB1 transporter we reasoned that Caco-2 cells treated with PTC would accumulate less intracellular rhodamine 123, which is a fluorescent dye and is a known efflux substrate for ABCB1 [15]. Consistent with this prediction, PTC addition to Caco-2 cells significantly decreased rhodamine 123 accumulation (Figure 5A and Supplementary Figure S3 at Next, we sought to determine whether PTC is also a substrate for ABCB1. We treated cells with the UIC2 antibody which only interacts with ABCB1 when it is in the ATP-bound active conformation [16]. The results of the FACS assay in Figure 5(B) show that ABCB1 shifts into its active conformation when cells are treated with either vinblastine, a known substrate for ABCB1 [17] or PTC. Thus PTC is a functional modulator of ABCB1 activity in vivo.

Figure 5 PTC decreases rhodamine 123 accumulation and is a substrate of ABCB1

(A) PTC-treated cells were exposed to rhodamine 123 for 1 h, and the fluorescence intensity in cell lysates was analysed using a microplate reader. (B) Cells were incubated with the UIC2 antibody in the presence and absence (filled histogram) of PTC (black line) or VIN (grey line) as described in the Experimental section. UIC2 reactivity for ABCB1 was analysed by flow cytometry. Rh123, rhodamine 123; Con, control; PE, phycoerythrin.

Our results indicate that T2Rs and ABCB1 play important interconnected roles in intestinal host defence: potentially toxic bitter food substances are ‘sensed’ by T2Rs expressed in enteroendocrine cells, which secrete CCK in response. Then, CCK acts on enterocytes to limit absorption of the dietary-derived T2R agonist by increasing ABCB1 efflux activity. It is likely that a similar protection is afforded to limit many other orally ingested drug/toxic substances that are both agonists for other T2Rs and substrates for ABCB1 (Figure 6).

Figure 6 Model of the intestinal host defence system by T2R and efflux transporter

Toxic substance (i.e. PTC) activates T2R38 leading to the release of gut hormones. CCK released from the basal surface of the enteroendocrine cells increases, through paracrine signalling expression of the efflux transporter ABCB1 in the enterocytes, resulting in rapid excretion of bitter or dietary toxin. CCKR, CCK receptor; EEC, enteroendocrine cells; EC, enterocytes.


Tae-Il Jeon and Timothy Osborne designed the research. Tae-Il Jeon and Young-Kyo Seo performed the research. Tae-Il Jeon and Timothy Osborne wrote the paper.


This work was supported, in part, by the National Institutes of Health [grant numbers HL48044, DK71021 (to T.O.)].


We thank Helen Hobbs and Jonathan Cohen for helpful suggestions.

Abbreviations: ABCB1, ATP-binding cassette B1; BCM, bitter compounds mixture; CCK, cholecystokinin; CCK2R, CCK2 receptor; DW, distilled water; β-gal, β-galactosidase; GLP-1, glucagon-like peptide-1; HBSS, Hanks balanced salt solution; PTC, phenylthiocarbamide; PXR, pregnane X receptor; qPCR, quantitative PCR; RT, reverse transcriptase; siRNA, small interfering RNA; T2R, bitter taste-sensing type 2 receptor; VIN, vinblastine


View Abstract