Bitter Taste Receptor Agonists Elicit G-Protein-Dependent Negative Inotropy in the Murine Heart

The FASEB Journal • Research Communication

Bitter taste receptor agonists elicit G-protein-dependent negative inotropy in the murine heart

Simon R. Foster,* Kristina Blank,† Louise E. See Hoe,‡ Maik Behrens,† Wolfgang Meyerhof,† Jason N. Peart,‡ and Walter G. Thomas*,1 *School of Biomedical Sciences, University of Queensland, Brisbane, Queensland, Australia; †Department of Molecular Genetics, German Institute of Human Nutrition (DIfE) Potsdam- Rehbrücke, Nuthetal, Germany; and ‡Grifﬁth Health Institute, Grifﬁth University, Gold Coast, Queensland, Australia

ABSTRACT G-protein-coupled receptors (GPCRs) The activation of G-protein-coupled receptors are key mediators in cardiovascular physiology, yet (GPCRs) converts extracellular stimuli into intracellu- frontline therapies for heart disease target only a small lar signaling to mediate a broad range of cellular and fraction of the cardiac GPCR repertoire. Moreover, physiological responses, as well as the senses of smell, there is emerging evidence that GPCRs implicated in taste, and vision. GPCRs represent the largest receptor taste (Tas1r and Tas2rs) have specific functions beyond superfamily in the human genome (1), and on the the oral cavity. Our recent description of these recep- strength of more than 4 decades of research, continue tors in heart tissue foreshadows a potential novel role to be prime therapeutic targets. The effect of GPCR- in cardiac cells. In this study, we identified novel targeted drugs is particularly evident in the treatment agonist ligands for cardiac-Tas2rs to enable the func- of cardiovascular disease, in which GPCR signaling is tional investigation of these receptors in heart tissue. modulated to control hypertension, stimulate inotropy Five Tas2rs cloned from heart tissue were screened in heart failure, and prevent thrombosis (2). Nonethe- less, current cardiovascular therapeutics target only a against a panel of 102 natural or synthetic bitter com- small fraction of the cardiac GPCR repertoire, predom- pounds in a heterologous expression system. We iden- inantly focusing on the adrenergic and angiotensin tified agonists for Tas2r108, Tas2r137, and Tas2r143 receptor systems. that were then tested in Langendorff-perfused mouse GPCR expression profiling analyses in human and hearts (from 8-wk-old male C57BL/6 mice). All Tas2r mouse tissues suggest that the heart expresses Ͼ150 agonists tested exhibited concentration-dependent ef- different GPCRs (3–5), reinforcing the notion that there fects, with agonists for Tas2r108 and Tas2r137, leading are many as yet unexplored, potential GPCR targets for ϳ to a 40% decrease in left ventricular developed the modulation of the heart and cardiovascular physiol- pressure and an increase in aortic pressure, respec- ogy. Moreover, despite representing more than half of the tively. These responses were abrogated in the presence GPCR repertoire, taste and odorant receptor genes have ␣ ␤␥ of G i and G inhibitors (pertussis toxin and gallein). been excluded from these previous analyses, but are This study represents the first demonstration of pro- attracting more attention, as they are increasingly being found Tas2r agonist-induced, G protein-dependent ef- identified in cells and tissues outside those of the mouth fects on mouse heart function.—Foster, S. R., Blank, and nose (reviewed in ref. 6). For example, the type 1 K., See Hoe, L. E., Behrens, M., Meyerhof, W., Peart, taste receptors (Tas1rs) are family C GPCRs that respond J. N., Thomas, W. G. Bitter taste receptor agonists elicit to sugars and amino acids and have been implicated in G-protein-dependent negative inotropy in the murine nutrient sensing and hormone secretion in the brain, gut, heart. FASEB J. 28, 4497–4508 (2014). www.fasebj.org and skeletal muscle (7–11). Type 2 taste receptors (Tas2rs) are a family of ϳ30 highly variable, single-exon genes that mediate bitter taste. Tas2rs also have a broad Key Words: Tas2r ⅐ GPCR ⅐ Langendorff ⅐ cardiovascular tissue expression profile beyond the gustatory system, system most notably throughout the gastrointestinal tract and in the airways (12–16). We have recently provided the first evidence that members of both the Tas1r and Tas2r gene Abbreviations: AoP, aortic pressure; DMEM, Dulbecco’s families are expressed in rodent and human heart tissue, modified Eagle’s medium; DMSO, dimethyl sulfoxide; EDP, with a subset of the Tas2rs up-regulated in the heart end diastolic pressure; FBS, fetal bovine serum; FLIPR, fluo- rometric imaging plate reader; GPCR, G-protein-coupled receptor; HEK293T, human embryonic kidney 293T; HSV, 1 Correspondence: School of Biomedical Sciences, Univer- herpes simplex virus; LVDP, left ventricular developed pres- sity of Queensland, St. Lucia, QLD 4072, Australia. E-mail: sure; PTX, pertussis toxin; SST, somatostatin; SysP, systolic [email protected] pressure; Tas1r, type 1 taste receptor; Tas2r, type 2 taste doi: 10.1096/fj.14-256305 receptor; TRPM5, transient receptor potential cation chan- This article includes supplemental data. Please visit http:// nel, subfamily M, member 5 www.fasebj.org to obtain this information.

0892-6638/14/0028-4497 © FASEB 4497

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. during starvation (17). These initial findings led us to cloned into pcDNA5/FRT (Invitrogen, Carlsbad, CA, USA) in speculate that cardiac-expressed Tas2rs respond to local a cassette containing the first 45 aa of the rat somatostatin nutrient status in the heart, although the precise func- type 3 receptor as an amino terminus cell membrane– tional role of these receptors in the cardiac setting is yet to targeting signal (24) and a carboxyl-terminal herpes simplex be thoroughly investigated. virus (HSV) glycoprotein D epitope, which does not interfere with receptor signaling and can be used for immunocyto- A major challenge to the successful broader physio- chemistry (25). The coding regions of the remaining cardiac- logical and pharmacological characterization of these expressed Tas2rs (identified in (17) were cloned from neo- GPCRs is the paucity of experimental tools for the natal rat heart cDNA, using gene-specific primers including rodent Tas2rs—most important among them, the lack EcoRI and NotI restriction sites into the pcDNA5/FRT soma- of selective agonist and antagonist ligands. Understand- tostatin (SST) HSV vector backbone (Table 1). Cloned full- ably, most studies directed at agonist identification length Tas2rs were confirmed by sequencing. have focused on the human TAS2Rs (18–20). Culmi- Human embryonic kidney 293T (HEK293T) cells (American Type Culture Collection, Manassas, VA, USA) were maintained nating with comprehensive compound screening with ϩ heterologous expression systems, almost all of the hu- in Dulbecco’s modified Eagle’s medium (DMEM) 10% fetal bovine serum (FBS). HEK293T cells that were modified to man TAS2R repertoire has now been effectively “deor- ␣ stably express the chimeric G-protein subunit G 16/gust44 (26) phanized” (21). Incidentally, the only definitive agon- were used for compound screening experiments. Cells were ist–receptor pairings for rodent Tas2rs are denatonium grown in a humidified incubator in 95% O2 and 5% CO2 at benzoate, with the mouse receptor Tas2r108 (previ- 37°C. ously designated mT2R8), and cycloheximide, with All compounds used in ligand screening were designated mouse Tas2r105 (mT2R5) (22) and its rat homologue as bitter-tasting in human psychophysical tests or were se- Tas2r105 (previously rT2R9) (20). The lack of rodent lected based on existing knowledge (21). A complete com- Tas2r ligand data is an important consideration, given pound list, including the concentrations tested in the initial that there is often substantial amino acid sequence screening, is provided in Supplemental Table S1. Analytical divergence between homologous human and rodent grade compounds were purchased from Sigma-Aldrich (St. bitter taste receptor genes, which may result in func- Louis, MO, USA), except for amarogentin (Chromadex, Irvine, CA, USA), or were isolated (19, 21). The compounds tionally distinct receptors. There are precedents in the were dissolved in Ca2ϩ assay buffer (10 mM d-glucose, 130 taste receptor literature where bitter ligand-mediated mM NaCl, 5 mM KCl, 10 mM HEPES, and 2 mM CaCl2,pH effects have been attributed to Tas2r function, most 7.4) or in dimethyl sulfoxide (DMSO), with the final DMSO prominently in the airways and gut (12, 16, 23). How- concentration not exceeding 0.1% (v/v), to minimize cellular ever, without specific pharmacological tools or appro- toxicity. Many of the bitter compounds are pharmacologically priate genetically modified mouse models, there re- active or are hydrophobic or amphiphilic and therefore could mains some controversy around the Tas2r dependence interfere with cellular calcium responses. Accordingly, pilot experiments were performed before screening in untrans- of the observed effects. ␣ In this study, we used an agonist screen, with the fected HEK293T G 16/gust44 cells, to ensure that the concen- specific purpose of identifying novel activating agonists trations tested were below those that elicited nonspecific calcium responses (21). for the Tas2rs that are expressed in cardiac tissue. We For Langendorff perfusion experiments, pertussis toxin then characterized the effect of putative Tas2r ligands on (PTX) was purchased from List Biological Laboratories the spectrum of cardiac functional parameters by using (Campbell, CA, USA) and gallein from Tocris Biosciences the Langendorff isolated heart preparation. Accordingly, (Bristol, UK). we demonstrated for the first time an agonist-dependent function of Tas2rs in the mouse heart. Calcium mobilization assay

␣ HEK293T G 16/gust44 cells were plated on poly-d-lysine– MATERIALS AND METHODS coated, 96-well, black-walled plates (50,000 cells/well) and were transfected the next day with a plasmid encoding a cardiac-expressed Tas2r, empty vector/mock plasmid as the Plasmids, cell culture, and reagents negative control or with a well-characterized taste receptor as the positive control (rat Tas2r105 or human TAS2R16, which Plasmid constructs encoding rat Tas2r105, Tas2r121, and robustly respond to cycloheximide and salicin, respectively, Tas2r126 were generated by using a published cloning strat- ref. 20). Ca2ϩ mobilization experiments were performed 24 h egy (20, 21). In brief, the Tas2r coding sequence was sub- after transfection, using an automated ﬂuorometric imaging

TABLE 1. Tas2r cloning primer sequences containing 5= EcoRI and 3= NotI restriction sites

Gene symbol Reference sequence Sequence

Tas2r108 NM_001024686.1 F: AGTGGCGAATTCATGCTCTGGGAACTGTATGCATTTG R: TGGCGGCCGCCGTTACTAACGAAATCCCGCAATTTGTAG Tas2r137 NM_001025149.1 F: GTGGCGAATTCATGTTGGGATTCACTGAAGGGATATTTC R: TGGCGGCCGCCTGAAGCAAAGAACCTCTTAGATCCAGG Tas2r143 NM_001025061.1 F: AGTGGCGAATTCATGCCCTCCACACCCACATT R: TGGCGGCCGCCAAATCTCATCTTCAGGGCCTTTCTC

F, forward; R, reverse.

4498 Vol. 28 October 2014 The FASEB Journal ⅐ www.fasebj.org FOSTER ET AL.

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. plate reader (FLIPR; Molecular Devices, Sunnyvale, CA, cated. G-protein inhibitor concentrations were based on previ- USA), as described previously (21). In brief, cells were loaded ous work in our laboratory or from the related literature (27, 30, with Fluo-4 AM dye (2 ␮M; Molecular Probes) in serum-free 31). PTX-treated hearts received 100 ␮g/kg intraperitoneally 24 DMEM containing 2.5 mM probenecid for1hat37°C and h before heart isolation. For gallein treatment, the hearts were washed in Ca2ϩ assay buffer. Changes in intracellular Ca2ϩ infused with 10 ␮M gallein for 10 min before bitter ligand levels were measured during automatic application of bitter infusion, and both were then administered throughout the compounds in the FLIPR. A second application of 100 nM experiment. Coronary flow, aortic pressure (AoP), and left somatostatin 14 (Bachem, Torrance, CA, USA), which acti- ventricular pressure signals were recorded on a 4-channel Ma- vates the endogenous somatostatin receptor type 2, was cLab data acquisition system (ADInstruments, Castle Hill, NSW, included in all assay protocols, to assess cell viability. Ligand Australia). Left ventricular pressure was digitally processed using screening experiments were performed in duplicate wells in Labchart 7 (ADInstruments), to yield left ventricular developed the absence (i.e., mock transfected with pcDNA5/FRT SST pressure (LVDP), systolic pressure (SysP), EDP, and heart rate. HSV) and presence of the Tas2rs, for comparison. Data analysis Langendorff-perfused heart model All data are presented as means Ϯ sem and were analyzed in All animal procedures were performed with specific approval GraphPad Prism 6.0 (Graph-Pad Software Inc., San Diego, CA, from Griffith University Animal Ethics Committee (MSC/04/ USA). Ligand screening was performed in duplicate with recep- 13/AEC) and follow the Australian Code of Practice for the tor-dependent responses determined by subtracting the calcium Care and Use of Animals for Scientific Purposes. Hearts were response of empty vector–transfected cells from Tas2r-trans- isolated from 8-wk-old, wild-type male C57/BL6 mice and fected cells for a given compound with the Screenworks 3.2 perfused (27, 28). Briefly, the mice were anesthetized with 60 software package (Molecular Devices). In subsequent experi- mg/kg pentobarbital sodium administered intraperitoneally, ments to validate the concentration dependence of Tas2r acti- a thoracotomy was performed, and the hearts were excised vation, the agonist-mediated Ca2ϩ response was quantified by ⌬ into ice-cold perfusion fluid. The aorta was cannulated on a using the F/F0 method to calculate the change in calcium shortened and blunted 21-gauge needle and perfused at a fluorescence signal intensity (⌬F) relative to the baseline fluo-

constant pressure of 80 mmHg with modified Krebs-Henseleit rescence signal (F0). Concentration-response data were fitted buffer (120 mM NaCl, 22 mM NaHCO3, 4.7 mM KCl, 1.2 mM with a 4-parameter nonlinear regression (with variable slope) to ϭ MgCl2, 2.5 mM CaC12, 1.2 mM KH2PO4,11mMd-glucose, 2 calculate the EC50. For perfused heart experiments (n 6), data mM pyruvate, and 0.5 mM EDTA). The perfusate was equili- were analyzed by 2-way ANOVA and Dunnett’s multiple com- brated in 95% O2 and 5% CO2 at 37°C (pH 7.4, Po2 600 parisons test, with ligand concentration and treatment as vari- mmHg). All perfusate delivered to the heart was passed ables. Values of P Ͻ 0.05 were considered significant. through an in-line 0.22 ␮m Sterivex-HV filter cartridge (Mil- lipore, Bedford, MA, USA) to continuously remove micropar- ticulates. The left ventricle was vented with an apical polyethylene drain, and the hearts were instrumented for functional RESULTS measurements, as described below. They were then immersed in warmed perfusate inside a water-jacketed bath maintained Rat Tas2r agonist screen at 37°C. The temperature of the perfusion fluid was moni- tored by a needle thermistor at the entry into the aortic To identify novel agonists, 5 Tas2rs (Tas2r108, -121, -126, cannula, and the temperature of the water bath was assessed -137, and -143) that we previously identified in rat cardiac with a second thermistor probe. Temperature was recorded tissue (17), were screened against a panel of 102 natural using a three-channel TH-8 digital thermometer (Physitemp and synthetic bitter compounds, with a fluorescence- Instruments, Clifton, NJ, USA). For assessment of isovolumic 2ϩ function, fluid-filled balloons constructed of polyvinyl chlo- based Ca mobilization assay. The initial compound ride plastic film were inserted into the left ventricle via the screen was performed in duplicate for each Tas2r at a mitral valve. The balloons were connected to a P23 XL single near-maximal concentration based on analogous pressure transducer (Viggo-Spectramed, Oxnard, CA, USA) experiments on human TAS2Rs (18, 19, 21). Positive ϩ by fluid-filled polyethylene tubing, permitting continuous control bitter taste receptor–dependent Ca2 mobiliza- measurement of left ventricular pressure. The balloon vol- tion responses are shown in Supplemental Fig. S1A (10 ume was increased to an end diastolic pressure (EDP) of 5 mM salicin with human TAS2R16 and 400 ␮M cyclohex- mmHg. Coronary flow was kept constant throughout, moni- imide with rat Tas2r105 vs. empty vector–transfected tored via a Doppler flow probe (Transonic Systems, Ithaca, HEK293T G␣ cells). Positive receptor-mediated NY, USA) placed in the aortic perfusion line and connected 16/gust44 to a T206 flowmeter (Transonic Systems). agonist responses were identified by comparison with After a 20 min equilibration at a constant perfusion pressure those of empty vector–transfected cells, and putative hits of 80 mmHg, the hearts were switched to ventricular pacing at a for each Tas2r were then validated for their concentration rate of 420–480 beats/min (S9 stimulator; Grass, Quincy, MA, dependence in further experiments, detailed below (Figs. 1 USA) depending on their intrinsic heart rate. This heart rate has and 2, Supplemental Figs. S1 and S2, and Table 2). been shown to yield optimal left ventricular performance in this model (29), and by controlling heart rate (and removing it as a Tas2r108 responds to a range of compounds tested variable), subsequent assessment of contractile function is sim- ϩ plified. Hearts were then switched to constant flow, where the In the initial Ca2 mobilization experiments, Tas2r108- flow was initially set to obtain an aortic perfusion pressure of 80 mmHg. Hearts were stabilized for a further 10 min before transfected cells responded to a range of different com- experimentation. Baseline measurements were made, and in- pounds, varying from typical GPCR-mediated transients creasing ligand concentrations were infused at 5-min intervals. (e.g., amarogentin, cromolyn, and sodium thiocyanate) to All drugs were infused through a 0.22 ␮m filter at not more than more irregular biphasic traces (e.g., chlorpheniramine) 3% of coronary flow, to achieve the final concentrations indi- (Fig. 1 and Supplemental Fig. S2). In subsequent valida-

NEGATIVE INOTROPY OF BITTER LIGANDS IN THE HEART 4499

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. Figure 1. Multiple agonists activate rat taste receptor Tas2r108 in vitro in a concentration-dependent manner. Tas2r108- ␣ 2ϩ transfected HEK293T G 16/gust44 cells responded to a range of different bitter compounds in initial FLIPR-based Ca mobilization experiments. Left panels: representative Tas2r108-dependent Ca2ϩ responses (corrected for empty vector responses) after agonist stimulation. Concentrations were as follows: amarogentin, 300 ␮M; azathioprine, 300 ␮M; chloramphenicol, 1 mM; chlorpheniramine, 100 ␮M; colchicine, 3 mM; cromolyn, 10 mM; denatonium saccharide, 3 mM; saccharin, 10 mM; sodium thiocyanate, 3 mM; and yohimbine, 200 ␮M. Right panels: concentration-dependent responses for rat Tas2r108 agonists (meansϮsem, nϭ2 performed in triplicate).

tion experiments, 10 of the candidate agonists gave small It is important to note that most bitter agonists tested were ϩ yet reproducible concentration-dependent Ca2 responses active at these Tas2rs in the micromolar-to-millimolar con- in Tas2r-transfected cells (Fig. 1). centration range, consistent with levels in analogous experiments on the human TAS2Rs (21). No agonists were iden- Tas2r137 and Tas2r143 are activated by a single tiﬁed and validated for Tas2r121 and Tas2r126. compound Novel Tas2r agonists exert concentration-dependent In contrast to the broadly responsive Tas2r108 receptor, effects in perfused mouse hearts Tas2r137 and Tas2r143 responded to a single bitter agonist tested in the compound screen (Fig. 2). In both cases, the The best candidates identiﬁed as novel Tas2r agonists initial hits for Tas2r137 (sodium benzoate) and Tas2r143 in the screen were then tested on intact mouse hearts. ϩ (arbutin) elicited concentration-dependent Ca2 mobiliza- In this regard, the Langendorff-perfused heart provides tion responses in further validation experiments. an ideal model to test the effects of ligand infusion on

4500 Vol. 28 October 2014 The FASEB Journal ⅐ www.fasebj.org FOSTER ET AL.

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. a steady increase in AoP that was most obvious at the highest agonist concentrations tested (Fig. 3). Sodium benzoate–infused mouse hearts displayed minor concentration-dependent effects on contractile parameters (LVDP, SysP, and EDP) and a pronounced biphasic effect on AoP. At low concentrations, sodium benzoate increased the AoP (peaking at 30 ␮M), but at concentrations above 1 mM, the AoP dropped substantially below baseline levels (Fig. 4). Arbutin infusion had a modest effect on cardiac function, which became apparent only at the highest concentrations tested (Fig. 5). Because of these small responses, arbutin was not pursued in subsequent experiments aimed at testing G-protein dependence of bitter ligands in the heart.

Tas2r agonist-induced changes in heart function are G-protein dependent

Figure 2. Tas2r137 and Tas2r143 are activated selectively by a 2ϩ Having demonstrated that the novel Tas2r agonists single compound. Left panels: receptor-dependent Ca re- elicit inotropic effects in the perfused heart model, we sponses after stimulation with 10 mM sodium benzoate for sought next to characterize these responses in terms of Tas2r137 (A) and 30 mM arbutin for Tas2r143 (B). Right panels: putative agonists elicited concentration-dependent the G-protein-mediated signaling pathways. As has been responses (meansϮsem, nϭ2 performed in triplicate). mentioned, this is a critical step toward the delineation of Tas2r function beyond the mouth, especially given the uncertainty over ligand selectivity and specificity. To this cardiac function, as it enables simultaneous recording end, mice were pretreated with PTX or gallein, to inhibit ␣ ␤␥ of multiple functional parameters. Hearts were isolated the G i and G G-protein subunits, respectively (27, 30, from 7- to 8-wk-old mice (21.8Ϯ0.2 g body weight; 31). Neither of these inhibitors affected the basal nϭ66 across all groups), and the baseline functional cardiac functional parameters (Table 3 and data not parameters EDP, LVDP, and coronary flow were con- shown). stant across treatment groups (Table 3). Sodium thiocyanate–mediated effects on mouse Three agonists were chosen from the earlier screen hearts were G-protein dependent, as the decreases in for testing in the whole hearts, as representative ago- LVDP and SysP were fully abrogated by PTX and gallein nists for specific cardiac-expressed Tas2rs: sodium thio- treatment (Fig. 3). In addition, the sodium thiocyanate– cyanate for Tas2r108, sodium benzoate for Tas2r137, dependent increase in AoP appeared to be attenuated ␣ ␤␥ and arbutin for Tas2r143. Interestingly, sodium thiocy- with G i and G inhibition, although the difference anate, sodium benzoate, and, to a lesser extent, arbutin, was not statistically significant. exhibited concentration-dependent effects on cardiac G-protein inhibition also had a marked influence on function. Specifically, sodium thiocyanate infusion elic- the cardiac effects of sodium benzoate infusion. Pre- ited a 30 to 40% decrease in LVDP and SysP, as well as treatment with PTX and gallein both blocked the

TABLE 2. Summary of Tas2r-dependent Ca2ϩ mobilization in agonist screening

Putative hits Concentration EC50 Physiological/therapeutic Tas2r (n) dependent Bitter compound (mM) application Reference

Tas2r108 24 10 Amarogentin 0.15 Cell proliferation inhibitor 70–72 Azathioprine 0.17 Immunosuppressive 73, 74 Chloramphenicol 0.33 Antibacterial 75, 76 Chlorpheniramine 0.08 Antihistamine 77 Colchicine 0.41 Anti-inﬂammatory, gout 78 Cromolyn 1.04 Allergic rhinitis/asthma 79 Denatonium saccharide 0.52 Aversive agent 80 Saccharin 5.69 Artiﬁcial sweetener 81 Sodium thiocyanate 6.68 Chemical synthesis 46 Yohimbine 0.20 Erectile dysfunction 44 Tas2r121 – – – – – Tas2r126 1 – – – – Tas2r137 1 1 Sodium benzoate 3.66 Food preservative, urea cycle 49, 82 disorders Tas2r143 1 1 Arbutin 2.46 Skin-lightening agent 83, 84

2ϩ Putative receptor-dependent hits were validated for concentration dependence. Estimated EC50 is based on the Ca mobilization assay responses. In many cases, these ligands have been implicated in or associated with physiological processes and disease. Relevant references are listed. See Supplemental Table S1 for complete list of tested compounds.

NEGATIVE INOTROPY OF BITTER LIGANDS IN THE HEART 4501

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. TABLE 3. Baseline functional parameters of Langendorff-perfused hearts measured after 30 min of stabilization

Ligand and treatment n Body weight (g) EDP (mmHg) LVDP (mmHg) Flow (ml/min)

Sodium thiocyanate Untreated 6 22.0 Ϯ 0.8 3.0 Ϯ 0.7 110.4 Ϯ 7.2 3.8 Ϯ 0.5 PTX 6 21.3 Ϯ 0.5 5.8 Ϯ 1.3 90.4 Ϯ 6.3 2.5 Ϯ 0.2 Gallein 6 22.5 Ϯ 0.5 5.5 Ϯ 1.1 103.6 Ϯ 4.8 3.0 Ϯ 0.3 Sodium benzoate Untreated 6 21.6 Ϯ 0.6 4.2 Ϯ 0.6 109.3 Ϯ 12.6 2.7 Ϯ 0.3 PTX 6 21.3 Ϯ 0.4 5.9 Ϯ 0.6 119.7 Ϯ 6.9 3.1 Ϯ 0.1 Gallein 6 21.6 Ϯ 0.8 7.2 Ϯ 1.4 98.3 Ϯ 10.0 2.8 Ϯ 0.3 Chloroquine Untreated 6 21.3 Ϯ 0.2 4.3 Ϯ 0.5 112.0 Ϯ 8.1 2.7 Ϯ 0.2 PTX 6 22.3 Ϯ 0.6 4.9 Ϯ 1.0 110.4 Ϯ 9.9 2.6 Ϯ 0.2 Gallein 6 22.9 Ϯ 0.5 6.2 Ϯ 0.8 100.6 Ϯ 3.8 3.2 Ϯ 0.4 Arbutin: untreated 4 21.9 Ϯ 0.3 5.5 Ϯ 0.6 93.0 Ϯ 5.5 2.8 Ϯ 0.1

biphasic effect on AoP exerted by sodium benzoate Langendorff perfused-heart experiments for compari- infusion (Fig. 4). son, particularly given the recently described role of this compound in airway relaxation (12, 32). Chloro- Chloroquine has profound effects on cardiac quine infusion had profound concentration-dependent contractility via a G-protein-independent mechanism effects on cardiac function, including a dramatic increase in AoP at low concentrations (0.1–10 ␮M) Although it did not activate any of the Tas2rs in the followed by a reduction that was coupled to an increase agonist screen, we also included chloroquine in our in EDP. In addition, chloroquine-infused mouse hearts

Figure 3. Sodium thiocyanate exerts concentration-dependent and G-protein-dependent effects on perfused mouse hearts. Normalized grouped data from sodium thiocyanate–infused mouse hearts for different cardiac parameters show a consistent decrease in LVDP and SysP, and a steady increase in AoP that was more pronounced at the highest agonist concentrations. ␣ ␮ ␤␥ ␮ Inhibition of G i (PTX, 100 g/kg i.p. 24 h before heart isolation) and G (gallein, 10 M gallein infused 10 min before bitter agonist infusion and throughout experiment) abrogated the sodium thiocyanate–induced decreases in LVDP and SysP (meansϮse, nϭ6/treatment, relative to untreated control hearts, 2-way ANOVA and Dunnett’s multiple comparisons tests, with

agonist concentration and treatment as the variables). Pint, interaction; Pconc, concentration; Ptrt, treatment.

4502 Vol. 28 October 2014 The FASEB Journal ⅐ www.fasebj.org FOSTER ET AL.

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. Figure 4. Sodium benzoate exerts concentration-dependent effects on AoP that are G-protein dependent. Normalized grouped data from sodium benzoate–infused mouse hearts showed minor effects on the measures of developed pressure (LVDP, SysP, and EDP) and a biphasic effect on AoP, with an initial increase followed by a pronounced decrease at high agonist ␣ ␮ ␤␥ ␮ concentrations. Inhibition of G i (PTX, 100 g/kg i.p. 24 h before heart isolation) and G (gallein, 10 M gallein infused 10 min before bitter agonist infusion and throughout the experiment) abrogated the sodium benzoate–induced biphasic effect on AoP (meansϮsem, nϭ6/treatment, expressed relative to untreated control hearts, 2-way ANOVA and Dunnett’s multiple

comparisons tests, with agonist concentration and treatment as variables). Pint, interaction; Pconc, concentration; Ptrt, treatment.

displayed reductions in LVDP and SysP that culminated In contrast to the sodium benzoate– and sodium in a complete loss of cardiac contractility at 1 mM. thiocyanate–elicited effects, the chloroquine-mediated Interestingly, this effect was reversible on cessation of effects on mouse heart function were not modulated by ligand infusion, generally within 5 min (Fig. 6A, B). G-protein inhibition, using either PTX or gallein (Fig. 6C). Together, these data suggest that chloroquine exerts its effects via a GPCR-independent mechanism.

DISCUSSION

We recently reported the expression of a subset of “taste receptors” in cardiac cells and tissues (17), although the lack of available pharmacological tools (i.e., agonists) prevented the further characterization of Tas2rs in heart. In this study, as the necessary first step to elucidating their function, we performed a compound screening and successfully identified novel activating ligands for 3 of the 5 heart-expressed Tas2rs. Using the Langendorff retrograde-perfused heart model, we provide the first demonstration of clear, concentration- and G-protein-dependent effects of novel Tas2r agonists on LVDP and AoP. These compel- Figure 5. Arbutin exerts minor concentration-dependent ling data represent the initial evidence that Tas2r effects on perfused mouse hearts. Grouped data from arbu- activation may have important negative inotropic ef- tin-infused mouse hearts show minor effects on LVDP, SysP, fects on mouse heart. EDP, and AoP that are most apparent at the highest agonist Taste GPCRs were first identified as the molecular concentrations tested (nϭ4, meansϮsem). mediators of taste, but are also expressed in multiple

NEGATIVE INOTROPY OF BITTER LIGANDS IN THE HEART 4503

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. Figure 6. Chloroquine-mediated effects on mouse heart function are not G-protein dependent. A, B) Representative traces for the duration of agonist infusion for LVDP (A) and AoP (B). Solid circles indicate the start of a given concentration of chloroquine. Open circles indicate the pause in agonist infusion to enable the changing of syringes on the infusion pump. C) ␣ ␮ ␤␥ ␮ Inhibition of G i (PTX, 100 g/kg i.p. 24 h before heart isolation), and G (gallein, 10 M gallein infused 10 min before bitter agonist infusion and throughout the experiment) had no effect on the pronounced concentration-dependent effects of chloroquine infusion on cardiac function (meansϮsem, nϭ6/treatment, expressed relative to untreated control hearts, 2-way

ANOVA and Dunnett’s multiple comparisons tests, with agonist concentration and treatment as variables). Pint, interaction; Pconc, concentration; Ptrt, treatment.

cells and tissues outside of the gustatory system. How- ligand binding profiles do not necessarily translate to ever, especially for the Tas2 GPCR family, the definitive the rodent clade of Tas2rs. Indeed, to date, only 2 elucidation of receptor-dependent function in tissues rodent Tas2rs (Tas2r105 and Tas2r108 in the current beyond the mouth has proven challenging (reviewed in nomenclature) have been effectively deorphanized (20, ref. 6). In many cases, the primary difficulty has been in 22). Accordingly, after identifying several Tas2rs in relating findings from readily controlled, cell-based, rodent heart tissue, we cloned and screened 5 “cardiac” Tas2r overexpression studies to unequivocal demon- Tas2rs against a panel of bitter agonists, as has been stration of Tas2r involvement in complex heteroge- done for human TAS2Rs (21). We determined the neous tissue responses. Moreover, most taste receptor validity of new agonist compounds on the basis of 2 research has focused on the human receptors, and the criteria: first, there was a clear receptor-dependent

4504 Vol. 28 October 2014 The FASEB Journal ⅐ www.fasebj.org FOSTER ET AL.

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. ϩ Ca2 transient on stimulation that was absent in the arbutin activates alternate signaling pathways down- corresponding empty vector–transfected cells, and sec- stream of Tas2r143 in the heart, with consequences ond, the response was robust, reproducible, and con- unrelated to contractility. Whatever the merits of this centration dependent. According to these criteria, we idea, a lack of effect of arbutin on heart function has have identified multiple agonists for rat Tas2r108 and also recently been reported in the context of cardiac single compounds that activate each of Tas2r137 and ischemia–reperfusion injury (43). Tas2r143. Our data are consistent with studies on In considering the Tas2r agonists that we have iden- human TAS2Rs, which are classified as broadly tuned tified, it is important to acknowledge that they may have (e.g., TAS2R14, which can be activated by Ͼ40 chemi- other molecular targets in addition to Tas2rs. For cally diverse compounds) or highly selective (21, 33). example, the 10 structurally diverse ligands identified These rodent Tas2r agonists, which we identified, for Tas2r108 include anti-inflammatory, immunosup- were active in the micromolar-to-millimolar range in pressive, antibiotic, and antihistaminergic agents (Ta- ϩ the Ca2 mobilization assay, entirely consistent with ble 2). Similarly, yohimbine is an agonist for Tas2r108 ␣ previously reported concentrations for activation of the but is also an 2-adrenoreceptor antagonist with affinity human TAS2Rs (21). Our identification of novel Tas2r for several other family A GPCRs (44, 45). Sodium agonists was based on a limited screening of a boutique thiocyanate is used in chemical synthesis (46), but was library of 102 natural and synthetic compounds. Future previously used for the treatment of essential hyperten- studies with larger and more chemically diverse com- sion, with noted cardiac side effects, including angina, pound sets are likely to identify additional, more potent coronary artery occlusion, and congestive heart failure agonists, particularly for Tas2r121 and Tas2r126, for (47, 48). Sodium benzoate is a food preservative that which we did not find activating ligands. Moreover, use may also have a therapeutic application for the treat- of other second-messenger signaling readouts (beyond ment of urea cycle disorders (49, 50), and arbutin is ϩ Ca2 mobilization as was used in this study), is likely to used as a topical skin lightening agent (51). So, al- reveal further Tas2r ligands and potentially to signal though we have identified some ligands that activate pathway-specific agonists (so-called biased agonists; ref. Tas2rs, we must remain cognizant of their potential 34) and antagonists. off-target activities; equally, previously described ac- We next sought to investigate the cardiovascular tions for these ligands, particularly in the heart, should effects of these agonist compounds ex vivo in retrograde be reinterpreted in light of our identification of these perfused hearts. Intriguingly, all agonists tested exerted compounds as agonists at Tas2rs. concentration-dependent functional effects in the We also observed a marked chloroquine-mediated heart, albeit on different cardiac parameters. Specifi- decrease in cardiac function, which is interesting, given cally, sodium thiocyanate elicited a pronounced de- that chloroquine is an agonist for human TAS2R3 and crease in left ventricular contractility, sodium benzoate is often used in functional studies as an archetypal predominantly affected AoP, and arbutin had a minor bitter agonist (12, 21, 32). However, chloroquine also effect on heart contractility. Our most critical finding has multiple intracellular targets and is used as a was to demonstrate G-protein-dependent cardiovascu- lysosomal inhibitor (underscoring its use in the treat- lar actions of these new Tas2r agonists. The changes in ment of malaria; ref. 52), as well as a nicotinic cholin- LVDP (sodium thiocyanate) and AoP (sodium benzo- ergic antagonist (53) and a proapoptotic anticancer ate) seen in control hearts were completely abrogated agent (54). In the murine perfused hearts, chloroquine ␣ ␤␥ by treatment with G i and G inhibitors. Both of these infusion completely abolished cardiac contractility, ap- inhibitors, PTX and gallein, have been extensively used parently consistent with observations from lung func- to clarify the G-protein signaling pathways for cardiac tional studies (12, 32). Notably, this effect was revers- GPCRs (27, 35–38). Moreover, PTX has been shown to ible on washout of ligand, suggesting that the overall block the effects of taste receptor activation in vitro viability of the cardiac tissue was not compromised. In (39), and gallein has similarly been used to demon- stark contrast to the data from mouse airway smooth strate G-protein-dependent signaling of taste receptors muscle cells (32), gallein and PTX pretreatment had in isolated mouse airways (32). In this regard, our data no effect on the chloroquine response in perfused ␣ ␤␥ support a role for G i and G subunits as mediators of hearts, questioning whether chloroquine acts via a Tas2r-dependent signaling in the heart. These data are GPCR, specifically a Tas2r. This is supported by the lack also consistent with the canonical Tas2r signal transduc- of effect of chloroquine on Tas2r-transfected cells in tion pathway, whereby receptors couple to an inhibitory our compound screening and also recent evidence that G␣ protein (gustducin) to mediate a decrease in cyclic chloroquine exerts its effect by decrease-modulating ϩ ϩ nucleotides, whereas the ␤␥ subunits activate PLC␤2to Ca2 signaling and Ca2 sensitivity via a G-protein- increase inositol 1,4,5-trisphosphate and intracellular independent mechanism (55). In addition, long-term ϩ Ca2 and subsequently open transient receptor poten- chloroquine treatment has been associated with several tial cation channel, subfamily M, member 5 (TRPM5) instances of human cardiomyopathy, suggestive of non- (40–42). specific effects (56, 57). Consequently, we would cau- Interestingly, not all Tas2rs seem to mediate robust tion against attributing the chloroquine-induced de- modulation of cardiac function. In our heterologous crease in cardiac function (and in other settings, e.g., expression system, arbutin is clearly an agonist for bronchodilation) to Tas2rs, until the precise mecha- Tas2r143 (a robustly expressed receptor in heart; ref. nism is resolved. 17), yet we did not observe an effect of arbutin infusion There are notable parallels in the potential molecu- in perfused mouse hearts. One explanation may be that lar processes that underpin cardiac contractility and

NEGATIVE INOTROPY OF BITTER LIGANDS IN THE HEART 4505

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. taste receptor activation in the tongue. Classic Tas2 REFERENCES ϩ receptor activation in the tongue results in Ca2 - activated, TRPM5-mediated membrane depolarization 1. Lagerstrom, M. C., and Schioth, H. B. (2008) Structural diversity (58). Interestingly, multiple members of the TRP chan- of G protein-coupled receptors and significance for drug discovery. Nat. Rev. Drug Discov. 7, 339–357 nel family are found in the cardiovascular system, with 2. DeWire, S. M., and Violin, J. D. (2011) Biased ligands for better recent evidence suggesting that TRPM5 and its close cardiovascular drugs: dissecting G-protein-coupled recep- functional homologue TRPM4 are expressed in cardi- tor pharmacology. Circ. Res. 109, 205–216 omyocytes (59, 60). TRPM4 has also been identified in 3. Hakak, Y., Shrestha, D., Goegel, M. C., Behan, D. P., and Chalmers, D. T. (2003) Global analysis of G-protein-coupled the cardiac conduction system (61), where it has been receptor signaling in human tissues. FEBS Lett. 550, 11–17 implicated in human conduction disorders (62, 63). 4. Moore-Morris, T., Varrault, A., Mangoni, M. E., Le Digarcher, Similarly, the sarcoplasmic/endoplasmic reticulum A., Negre, V., Dantec, C., Journot, L., Nargeot, J., and Couette, ϩ Ca2 -ATPase (SERCA) pump that is critical in seques- B. (2009) Identification of potential pharmacological targets by ϩ tering Ca2 during cardiac excitation–contraction cou- analysis of the comprehensive G protein-coupled receptor rep- 2ϩ ertoire in the four cardiac chambers. Mol. Pharmacol. 75, 1108– pling (64), has also been implicated in Ca clearance 1116 in taste receptor cells in the tongue (65). Last, voltage- 5. Regard, J. B., Sato, I. T., and Coughlin, S. R. (2008) Anatomical 2ϩ 2ϩ profiling of G protein-coupled receptor expression. Cell 135, gated L-type Ca (Cav1.2) channels initiate Ca in- flux and contraction in cardiomyocytes (66), and inhibi- 561–571 6. Foster, S. R., Roura, E., and Thomas, W. G. (2014) Extrasensory tion of these channels was essential for bitter ligand- perception: odorant and taste receptors beyond the nose mediated bronchodilation (32). Evidently, the potential and mouth. Pharmacol. Ther. 142, 41–61 ϩ modulation of Ca2 -regulating pumps and channels in 7. Jang, H. J., Kokrashvili, Z., Theodorakis, M. J., Carlson, O. D., mediating the cardiac effects downstream of Tas2r activa- Kim, B. J., Zhou, J., Kim, H. H., Xu, X., Chan, S. L., Juhaszova, M., Bernier, M., Mosinger, B., Margolskee, R. F., and Egan, J. M. tion in the heart is an area worthy of future investigation. (2007) Gut-expressed gustducin and taste receptors regulate As a final comment, the experiments using PTX and secretion of glucagon-like peptide-1. Proc. Natl. Acad. Sci. U. S. A. gallein suggest that the effects of certain ligands on 104, 15069–15074 cardiac function are G protein-dependent. Further 8. Mace, O. J., Lister, N., Morgan, E., Shepherd, E., Affleck, J., Helliwell, P., Bronk, J. R., Kellett, G. L., Meredith, D., Boyd, R., experiments are needed to unambiguously demon- Pieri, M., Bailey, P. D., Pettcrew, R., and Foley, D. (2009) An strate that they are Tas2 receptor-dependent. Such energy supply network of nutrient absorption coordinated by experiments have proven quite challenging within the calcium and T1R taste receptors in rat small intestine. J. Physiol. taste receptor field, given that there are no known 587, 195–210 selective antagonists for these rodent Tas2rs. To date, 9. Margolskee, R. F., Dyer, J., Kokrashvili, Z., Salmon, K. S. H., Ilegems, E., Daly, K., Maillet, E. L., Ninomiya, Y., Mosinger, B., there have been very few reports of antagonists for and Shirazi-Beechey, S. P. (2007) T1R3 and gustducin in gut human TAS2Rs, and even these compounds are not sense sugars to regulate expression of Naϩ-glucose cotrans- entirely specific (67, 68). Moreover, the broad tuning porter 1. Proc. Natl. Acad. Sci. U. ;S. A. 104, 15075–15080 of some Tas2rs and the expression of multiple receptor 10. Ren, X., Zhou, L., Terwilliger, R., Newton, S. S., and de Araujo, I. E. (2009) Sweet taste signaling functions as a hypothalamic genes in the heart may limit the utility of traditional glucose sensor. Front. Integr. Neurosci. 3, 12 gene knockout studies. To this end, it may be possible 11. Wauson, E. M., Zaganjor, E., Lee, A. Y., Guerra, M. L., Ghosh, to take advantage of the genomic clustering of the A. B., Bookout, A. L., Chambers, C. P., Jivan, A., McGlynn, K., Tas2r to simultaneously knock out multiple receptors. Hutchison, M. R., Deberardinis, R. J., and Cobb, M. H. (2012) The G protein-coupled taste receptor T1R1/T1R3 regulates Alternatively, as has been done for one TAS2R, generation mTORC1 and autophagy. Mol. Cell 47, 851–862 of humanized mice expressing tissue-specific human taste 12. Deshpande, D. A., Wang, W. C. H., McIlmoyle, E. L., Robinett, GPCRs may be a profitable approach in the future (69). K. S., Schillinger, R. M., An, S. S., Sham, J. S. K., and Liggett, This study represents significant progress toward the S. B. (2010) Bitter taste receptors on airway smooth muscle delineation of the functional role of taste receptors in bronchodilate by localized calcium signaling and reverse ob- struction. Nat. Med. 16, 1299–1304 the heart. For the first time, using the newly identified 13. Finger, T. E., Böttger, B., Hansen, A., Anderson, K. T., Alimo- putative Tas2r agonists, sodium thiocyanate and so- hammadi, H., and Silver, W. L. (2003) Solitary chemoreceptor dium benzoate, we have demonstrated pronounced cells in the nasal cavity serve as sentinels of respiration. Proc. and distinct effects on cardiac function that were Natl. Acad. Sci. U. S. A. 100, 8981–8986 14. Rozengurt, N., Wu, S. V., Chen, M. C., Huang, C., Sternini, C., crucially G-protein dependent. The data provide cor- and Rozengurt, E. (2006) Colocalization of the alpha-subunit of roborating evidence for rodent Tas2r-dependent func- gustducin with PYY and GLP-1 in L cells of human colon. Am. J. tion beyond the mouth. Future experiments will focus Physiol. Gastrointest. Liver Physiol. 291, G792–G802 on demonstrating unequivocally that the observed re- 15. Shah, A. S., Ben-Shahar, Y., Moninger, T. O., Kline, J. N., and Welsh, M. J. (2009) Motile cilia of human airway epithelia sponses to bitter ligands are attributable to specific are chemosensory. Science 325, 1131–1134 Tas2rs in the heart. 16. Wu, S. V., Rozengurt, N., Yang, M., Young, S. H., Sinnett-Smith, J., and Rozengurt, E. (2002) Expression of bitter taste receptors This work was supported by project grants awarded to of the T2R family in the gastrointestinal tract and enteroendo- W.G.T. from the Australian National Health and Medical crine STC-1 cells. Proc. Natl. Acad. Sci. U. S. A. 99, 2392–2397 Research Council (NHMRC; 1024726) and the National 17. Foster, S. R., Porrello, E. R., Purdue, B., Chan, H.-W., Voigt, A., Heart Foundation of Australia (G-12B-6532); an Australian Frenzel, S., Hannan, R. D., Moritz, K. M., Simmons, D. G., Molenaar, P., Roura, E., Boehm, U., Meyerhof, W., and Thomas, Postgraduate Award scholarship to S.R.F.; a scholarship from W. G. (2013) Expression, regulation and putative nutrient- the National Heart Foundation of Australia to L.S.H.; and a sensing function of taste GPCRs in the heart. PLoS One 8, e64579 Future Fellowship from the Australian Research Council to 18. Behrens, M., Brockhoff, A., Kuhn, C., Bufe, B., Winnig, M., and J.N.P. Meyerhof, W. (2004) The human taste receptor hTAS(2)R(14)

4506 Vol. 28 October 2014 The FASEB Journal ⅐ www.fasebj.org FOSTER ET AL.

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. responds to a variety of different bitter compounds. Biochem. G. I., Bernstein, D., Baker, A. J., and Conklin, B. R. (2000) Biophys. Res. Commun. 319, 479–485 Conditional expression of a Gi-coupled receptor causes ventric- 19. Brockhoff, A., Behrens, M., Massarotti, A., Appendino, G., and ular conduction delay and a lethal cardiomyopathy. Proc. Natl. Meyerhof, W. (2007) Broad tuning of the human bitter taste Acad. Sci. U. S. A. 97, 4826–4831 receptor hTAS2R46 to various sesquiterpene lactones, clero- 38. Zheng, M., Zhang, S. J., Zhu, W. Z., Ziman, B., Kobilka, B. K., dane and labdane diterpenoids, strychnine, and denatonium. J. and Xiao, R. P. (2000) beta 2-adrenergic receptor-induced p38 Agric Food. Chem. 55, 6236–6243 MAPK activation is mediated by protein kinase A rather than by 20. Bufe, B., Hofmann, T., Krautwurst, D., Raguse, J.-D., and Gi or gbeta gamma in adult mouse cardiomyocytes. J. Biol. Chem. Meyerhof, W. (2002) The human TAS2R16 receptor mediates 275, 40635–40640 bitter taste in response to ␤-glucopyranosides. Nat. Genet. 32, 39. Ozeck, M., Brust, P., Xu, H., and Servant, G. (2004) Receptors 397–401 for bitter, sweet and umami taste couple to inhibitory G protein 21. Meyerhof, W., Batram, C., Kuhn, C., Brockhoff, A., Chudoba, E., signaling pathways. Eur. J. Pharmacol. 489, 139–149 Bufe, B., Appendino, G., and Behrens, M. (2010) The molecular 40. Huang, L. Q., Shanker, Y. G., Dubauskaite, J., Zheng, J. Z., Yan, receptive ranges of human TAS2R bitter taste receptors. Chem. W. T., Rosenzweig, S., Spielman, A. I., Max, M., and Margolskee, Senses 35, 157–170 R. F. (1999) Ggamma13 colocalizes with gustducin in taste 22. Chandrashekar, J., Mueller, K. L., Hoon, M. A., Adler, E., Feng, receptor cells and mediates IP3 responses to bitter denatonium. L., Guo, W., Zuker, C. S., and Ryba, N. J. (2000) T2Rs function Nat. Neurosci. 2, 1055–1062 as bitter taste receptors. Cell 100, 703–711 41. Yan, W. T., Sunavala, G., Rosenzweig, S., Dasso, M., Brand, J. G., 23. Jeon, T. I., Zhu, B., Larson, J. L., and Osborne, T. F. (2008) and Spielman, A. I. (2001) Bitter taste transduced by PLC-beta(2)- SREBP-2 regulates gut peptide secretion through intestinal dependent rise in IP3 and alpha-gustducin-dependent fall in bitter taste receptor signaling in mice. J. Clin. Invest. 118, cyclic nucleotides. Am. J. Physiol. Cell Physiol. 280, C742–C751 3693–3700 42. Zhang, Y., Hoon, M. A., Chandrashekar, J., Mueller, K. L., Cook, 24. Ammon, C., Schafer, J., Kreuzer, O. J., and Meyerhof, W. (2002) B., Wu, D., Zuker, C. S., and Ryba, N. J. P. (2003) Coding of Presence of a plasma membrane targeting sequence in the sweet, bitter, and umami tastes: different receptor cells sharing amino-terminal region of the rat somatostatin receptor 3. Arch. similar signaling pathways. Cell 112, 293–301 Physiol. Biochem. 110, 137–145 43. Broskova, Z., Drabikova, K., Sotnikova, R., Fialova, S., and Knezl, 25. Roosterman, D., Roth, A., Kreienkamp, H. J., Richter, D., and V. (2013) Effect of plant polyphenols on ischemia-reperfusion Meyerhof, W. (1997) Distinct agonist-mediated endocytosis of injury of the isolated rat heart and vessels. Phytother. Res. 27, cloned rat somatostatin receptor subtypes expressed in insuli- 1018–1022 noma cells. J. Neuroendocrinol. 9, 741–751 44. Lebret, T., Herve, J. M., Gorny, P., Worcel, M., and Botto, H. 26. Ueda, T., Ugawa, S., Yamamura, H., Imaizumi, Y., and Shimada, (2002) Efficacy and safety of a novel combination of L-arginine S. (2003) Functional interaction between T2R taste receptors ␣ glutamate and yohimbine hydrochloride: a new oral therapy for and G-protein subunits expressed in taste receptor cells. J. erectile dysfunction. Eur. Urol. 41, 608–613 Neurosci. 23, 7376–7380 45. Millan, M. J., Newman-Tancredi, A., Audinot, V., Cussac, D., 27. Peart, J. N., and Gross, G. J. (2006) Cardioprotective effects of Lejeune, F., Nicolas, J. P., Coge, F., Galizzi, J. P., Boutin, J. A., acute and chronic opioid treatment are mediated via different Rivet, J. M., Dekeyne, A., and Gobert, A. (2000) Agonist and signaling pathways. Am. J. Physiol. Heart Circ. Physiol. 291, antagonist actions of yohimbine as compared to fluparoxan at H1746–H1753 alpha(2)-adrenergic receptors (AR)s, serotonin (5-HT)(1A), 28. Reichelt, M. E., Willems, L., Hack, B. A., Peart, J. N., and Headrick, 5-HT(1B), 5-HT(1D) and dopamine D(2) and D(3) receptors: J. P. (2009) Cardiac and coronary function in the Langendorff- significance for the modulation of frontocortical monoaminer- perfused mouse heart model. Exp. Physiol. 94, 54–70 gic transmission and depressive states. Synapse 35, 79–95 29. Headrick, J. P., Peart, J., Hack, B., Flood, A., and Matherne, 46. Schwan, A. L. (2001) Sodium thiocyanate. In Encyclopedia of G. P. (2001) Functional properties and responses to ischaemia- reperfusion in Langendorff perfused mouse heart. Exp. Physiol. Reagents for Organic Synthesis, John Wiley & Sons, Ltd, New York 86, 703–716 47. Bacq, Z. M., Charlier, R., Philippot, E., and Fischer, P. (1949) 30. Bonacci, T. M., Mathews, J. L., Yuan, C., Lehmann, D. M., Malik, The action of sodium thiocyanate on cardiac output. Br. J. S., Wu, D., Font, J. L., Bidlack, J. M., and Smrcka, A. V. (2006) Pharmacol. Chemother. 4, 162–167 Differential targeting of Gbetagamma-subunit signaling with 48. Beamish, R. E., and Adamson, J. D. (1945) Treatment of small molecules. Science 312, 443–446 essential hypertension with sodium thiocyanate. Can. Med. Assoc. 31. Lehmann, D. M., Seneviratne, A. M., and Smrcka, A. V. (2008) J. 53, 236–242 Small molecule disruption of G protein beta gamma subunit 49. Enns, G. M., Berry, S. A., Berry, G. T., Rhead, W. J., Brusilow, signaling inhibits neutrophil chemotaxis and inflammation. S. W., and Hamosh, A. (2007) Survival after treatment with Mol. Pharmacol. 73, 410–418 phenylacetate and benzoate for urea-cycle disorders. N. Engl. J. 32. Zhang, C.-H., Lifshitz, L. M., Uy, K. F., Ikebe, M., Fogarty, K. E., Med. 356, 2282–2292 and ZhuGe, R. (2013) The cellular and molecular basis of bitter 50. Nair, B. (2001) Final report on the safety assessment of benzyl tastant-induced bronchodilation. PLoS Biol. 11, e1001501 alcohol, benzoic acid, and sodium benzoate. Int. J. Toxicol. 33. Levit, A., Nowak, S., Peters, M., Wiener, A., Meyerhof, W., 20(Suppl. 3), 23–50 Behrens, M., and Niv, M. Y. (2014) The bitter pill: clinical drugs 51. Parvez, S., Kang, M., Chung, H. S., Cho, C., Hong, M. C., Shin, that activate the human bitter taste receptor TAS2R14. FASEB J. M. K., and Bae, H. (2006) Survey and mechanism of skin 28, 1181–1197 depigmenting and lightening agents. Phytother. Res. 20, 921–934 34. Kenakin, T., and Christopoulos, A. (2013) Signalling bias in new 52. Slater, A. F. (1993) Chloroquine: mechanism of drug action and drug discovery: detection, quantification and therapeutic im- resistance in Plasmodium falciparum. Pharmacol. Ther. 57, 203–235 pact. Nat. Rev. Drug Discov. 12, 205–216 53. Ballestero, J. A., Plazas, P. V., Kracun, S., Gomez-Casati, M. E., 35. Piao, L., Fang, Y. H., Parikh, K. S., Ryan, J. J., D’Souza, K. M., Taranda, J., Rothlin, C. V., Katz, E., Millar, N. S., and Elgoyhen, Theccanat, T., Toth, P. T., Pogoriler, J., Paul, J., Blaxall, B. C., A. B. (2005) Effects of quinine, quinidine, and chloroquine on Akhter, S. A., and Archer, S. L. (2012) GRK2-mediated inhibi- alpha9alpha10 nicotinic cholinergic receptors. Mol. Pharmacol. tion of adrenergic and dopaminergic signaling in right ventric- 68, 822–829 ular hypertrophy: therapeutic implications in pulmonary hyper- 54. Lakhter, A. J., Sahu, R. P., Sun, Y., Kaufmann, W. K., Androphy, tension. Circulation 126, 2859–2869 E. J., Travers, J. B., and Naidu, S. R. (2013) Chloroquine promotes 36. Meens, M. J., Mattheij, N. J., van Loenen, P. B., Spijkers, L. J., apoptosis in melanoma cells by inhibiting BH3 domain-mediated Lemkens, P., Nelissen, J., Compeer, M. G., Alewijnse, A. E., and PUMA degradation. J. Invest. Dermatol. 133, 2247–2254 De Mey, J. G. (2012) G-protein betagamma subunits in vasore- 55. Tan, X., and Sanderson, M. J. (2014) Bitter tasting compounds laxing and anti-endothelinergic effects of calcitonin gene-re- dilate airways by inhibiting airway smooth muscle calcium oscilla- lated peptide. Br. J. Pharmacol. 166, 297–308 tions and calcium sensitivity. Br. J. Pharmacol. 171, 646–662 37. Redfern, C. H., Degtyarev, M. Y., Kwa, A. T., Salomonis, N., 56. Stas, P., Faes, D., and Noyens, P. (2008) Conduction disorder Cotte, N., Nanevicz, T., Fidelman, N., Desai, K., Vranizan, K., and QT prolongation secondary to long-term treatment Lee, E. K., Coward, P., Shah, N., Warrington, J. A., Fishman, with chloroquine. Int. J. Cardiol. 127, e80–e82

NEGATIVE INOTROPY OF BITTER LIGANDS IN THE HEART 4507

Downloaded from www.fasebj.org by (2001:388:608c:4800:e578:c55:a347:3276) on September 12, 2018. The FASEB Journal Vol. ${article.issue.getVolume()}, No. ${article.issue.getIssueNumber()}, pp. 4497-4508. 57. Tonnesmann, E., Kandolf, R., and Lewalter, T. (2013) Chloro- 71. Pal, D., Sur, S., Mandal, S., Das, A., Roy, A., Das, S., and Panda, quine cardiomyopathy: a review of the literature. Immunophar- C. K. (2012) Prevention of liver carcinogenesis by amarogentin macol. Immunotoxicol. 35, 434–442 through modulation of G1/S cell cycle check point and induc- 58. Perez, C. A., Huang, L. Q., Rong, M. Q., Kozak, J. A., Preuss, tion of apoptosis. Carcinogenesis 33, 2424–2431 A. K., Zhang, H. L., Max, M., and Margolskee, R. F. (2002) A 72. Saha, P., Mandal, S., Das, A., and Das, S. (2006) Amarogentin transient receptor potential channel expressed in taste recep- can reduce hyperproliferation by downregulation of Cox-II and tor cells. Nat. Neurosci. 5, 1169–1176 upregulation of apoptosis in mouse skin carcinogenesis model. 59. Dietrich, A., and Gudermann, T. (2011) TRP channels in the Cancer Lett. 244, 252–259 cardiopulmonary vasculature. Adv. Exp. Med. Biol. 704, 781–810 73. Elion, G. (1989) The purine path to chemotherapy. Science 244, 60. Watanabe, H., Murakami, M., Ohba, T., Takahashi, Y., and Ito, 41–47 H. (2008) TRP channel and cardiovascular disease. Pharmacol. 74. Tiede, I., Fritz, G., Strand, S., Poppe, D., Dvorsky, R., Strand, D., Ther. 118, 337–351 Lehr, H. A., Wirtz, S., Becker, C., Atreya, R., Mudter, J., Hildner, 61. Demion, M., Bois, P., Launay, P., and Guinamard, R. (2007) K., Bartsch, B., Holtmann, M., Blumberg, R., Walczak, H., Iven, TRPM4, a Ca2ϩ-activated nonselective cation channel in mouse H., Galle, P. R., Ahmadian, M. R., and Neurath, M. F. (2003) sino-atrial node cells. Cardiovasc. Res. 73, 531–538 CD28-dependent Rac1 activation is the molecular target of 62. Kruse, M., Schulze-Bahr, E., Corfield, V., Beckmann, A., azathioprine in primary human CD4ϩ T lymphocytes. J. Clin. Stallmeyer, B., Kurtbay, G., Ohmert, I., Brink, P., and Pongs, O. Invest. 111, 1133–1145 (2009) Impaired endocytosis of the ion channel TRPM4 is 75. Falagas, M. E., Grammatikos, A. P., and Michalopoulos, A. associated with human progressive familial heart block type I. J. (2008) Potential of old-generation antibiotics to address current Clin. Invest. 119, 2737–2744 need for new antibiotics. Expert Rev. Anti Infect. Ther. 6, 593–600 63. Liu, H., El Zein, L., Kruse, M., Guinamard, R., Beckmann, A., 76. Park, J.-Y., Kim, K.-A., and Kim, S.-L. (2003) Chloramphenicol is Bozio, A., Kurtbay, G., Mégarbané, A., Ohmert, I., Blaysat, G., a potent inhibitor of cytochrome P450 isoforms CYP2C19 and Villain, E., Pongs, O., and Bouvagnet, P. (2010) Gain-of-func- CYP3A4 in human liver microsomes. Antimicrob. Agents Che- tion mutations in TRPM4 cause autosomal dominant isolated mother. 47, 3464–3469 cardiac conduction disease/clinical perspective. Circ. Cardiovasc. 77. Assanasen, P., and Naclerio, R. M. (2002) Antiallergic anti- Genet. 3, 374–385 inflammatory effects of H1-antihistamines in humans. Clin. 64. Inesi, G., Prasad, A. M., and Pilankatta, R. (2008) The Ca2ϩ Allergy Immunol. 17, 101–139 ATPase of cardiac sarcoplasmic reticulum: physiological role 78. Cocco, G., Chu, D. C., and Pandolfi, S. (2010) Colchicine in and relevance to diseases. Biochem. Biophys. Res. Commun. 369, clinical medicine: a guide for internists. Eur. J. Intern. Med. 21, 182–187 503–508 65. Iguchi, N., Ohkuri, T., Slack, J. P., Zhong, P., and Huang, L. (2011) 79. Storms, W., and Kaliner, M. A. (2005) Cromolyn sodium: fitting Sarco/endoplasmic reticulum Ca2ϩ-ATPases (SERCA) contribute an old friend into current asthma treatment. J. Asthma 42, 79–89 to GPCR-mediated taste perception. PLoS One 6, e23165 80. Klein-Schwartz, W. (1991) Denatonium benzoate: review of 66. Shaw, R. M., and Colecraft, H. M. (2013) L-type calcium efficacy and safety. Vet. Hum. Toxicol. 33, 545–547 channel targeting and local signalling in cardiac myocytes. 81. Swithers, S. E., Martin, A. A., and Davidson, T. L. (2010) Cardiovasc. Res. 98, 177–186 High-intensity sweeteners and energy balance. Physiol. Behav. 67. Brockhoff, A., Behrens, M., Roudnitzky, N., Appendino, G., 100, 55–62 Avonto, C., and Meyerhof, W. (2011) Receptor agonism and 82. Krebs, H. A., Wiggins, D., Stubbs, M., Sols, A., and Bedoya, F. antagonism of dietary bitter compounds. J. Neurosci. 31, 14775– (1983) Studies on the mechanism of the antifungal action 14782 of benzoate. Biochem. J. 214, 657–663 68. Slack, J. P., Brockhoff, A., Batram, C., Menzel, S., Sonnabend, 83. Lim, Y. J., Lee, E. H., Kang, T. H., Ha, S. K., Oh, M. S., Kim, C., Born, S., Galindo, M. M., Kohl, S., Thalmann, S., Ostopovici- S. M., Yoon, T. J., Kang, C., Park, J. H., and Kim, S. Y. (2009) Halip, L., Simons, C. T., Ungureanu, I., Duineveld, K., Bologa, Inhibitory effects of arbutin on melanin biosynthesis of alpha- C. G., Behrens, M., Furrer, S., Oprea, T. I., and Meyerhof, W. melanocyte stimulating hormone-induced hyperpigmentation (2010) Modulation of bitter taste perception by a small mole- in cultured brownish guinea pig skin tissues. Arch. Pharm. Res. cule hTAS2R antagonist. Curr. Biol. 20, 1104–1109 32, 367–373 69. Mueller, K. L., Hoon, M. A., Erlenbach, I., Chandrashekar, J., 84. Maeda, K., and Fukuda, M. (1996) Arbutin: mechanism of its Zuker, C. S., and Ryba, N. J. P. (2005) The receptors and coding depigmenting action in human melanocyte culture. J. Pharma- logic for bitter taste. Nature 434, 225–229 col. Exp. Ther. 276, 765–769 70. Kesavan, R., Potunuru, U. R., Nastasijevic, B., T, A., Joksic, G., and Dixit, M. (2013) Inhibition of vascular smooth muscle cell Received for publication May 21, 2014. proliferation by Gentiana lutea root extracts. PLoS One 8, e61393 Accepted for publication June 23, 2014.

4508 Vol. 28 October 2014 The FASEB Journal ⅐ www.fasebj.org FOSTER ET AL.