Bitter substances suppress afferent responses to an appetitive mixture: Evidence for peripheral integration of chemosensory stimuli

The processes that lead from detection of chemicals, transduction, and coding with the appropriate message to initiate ingestion of a palatable meal or to reject a potentially noxious substance are poorly understood in vertebrates owing to the complex organization of the taste system. As a first step in elucidating the cellular basis of the behavioral differences elicited by appetitive stimuli and bitter compounds, we recorded from the afferent nerves conveying peripheral chemosensory information to the CNS in the head of the leech, Hirudo medicinalis. Superfusion of the chemosensory region of the lip of Hirudo with a mixture of NaCl (150 mM) and arginine (1 mM), an appetitive solution that elicits ingestion, increased the neuronal activity in the afferent cephalic nerves, for example (Zhang X, Wilson RJ, Li Y, Kleinhaus AL. 2000. Chemical and thermal stimuli have short‐lived effects on the Retzius cell in the medicinal leech. J Neurobiol 43:304–311.). In the present paper we show that superfusing the lip with quinine or denatonium reduced the basal neural activity in the afferents. Furthermore, these bitter substances in the appetitive solution counteracted the increased activity the appetitive solution evoked in the cephalic nerves. Thus, the neural activity evoked by the application of appetitive and aversive stimuli to the chemosensory area of the lip paralleled the opposite behavioral responses to the same chemicals. The results suggest that individual leech taste cells possess receptors for both types of stimuli. Therefore, the leech may be a good model system in which to study peripheral taste events in cells that may possess multiple receptors and transduction mechanisms that interact to integrate information. © 2001 John Wiley & Sons, Inc. J Neurobiol 49: 255–263, 2001     

Chemical and thermal stimuli have short-lived effects on the retzius cell in the medicinal leech

During the appetitive phase of feeding, hungry leeches detect a prey by the integration of signals perceived by different sensory systems. Earlier reports suggested that chemical or thermal sensory stimulation of the lip was associated with increased afferent activity in cephalic nerves connecting the lip to the central nervous system. These authors further suggested that this activity was relayed to Retzius cells in segmental ganglia, which then released serotonin to initiate and control all aspects of feeding behavior. In this study, we show that chemosensory or thermal activation of the lip lasting for at least 5 min produces a distinct signal in the cephalic nerves consisting of action potentials of low amplitude. These small amplitude signals are clearly distinguishable from the large action potentials evoked by mechanosensory stimuli applied to the same area of the lip. Both types of sensory stimuli also evoke an increase in the firing frequency of the Retzius cells in segmental ganglia. However, the response recorded in the nerves and the Retzius cells during a maintained stimulus is not constant but decreases with an exponential time course. These results agree with our earlier observations on a semi-intact feeding preparation in which we showed that the firing frequency of the Retzius cell decreased as soon as the leech began to ingest its meal. Therefore, our data provide further evidence suggesting that it is unlikely that heat or chemical cues maintain the Retzius cell in an active state throughout the consummatory phase of feeding.     

Chemical and thermal stimuli have short‐lived effects on the Retzius cell in the medicinal leech

During the appetitive phase of feeding, hungry leeches detect a prey by the integration of signals perceived by different sensory systems. Earlier reports suggested that chemical or thermal sensory stimulation of the lip was associated with increased afferent activity in cephalic nerves connecting the lip to the central nervous system. These authors further suggested that this activity was relayed to Retzius cells in segmental ganglia, which then released serotonin to initiate and control all aspects of feeding behavior. In this study, we show that chemosensory or thermal activation of the lip lasting for at least 5 min produces a distinct signal in the cephalic nerves consisting of action potentials of low amplitude. These small amplitude signals are clearly distinguishable from the large action potentials evoked by mechanosensory stimuli applied to the same area of the lip. Both types of sensory stimuli also evoke an increase in the firing frequency of the Retzius cells in segmental ganglia. However, the response recorded in the nerves and the Retzius cells during a maintained stimulus is not constant but decreases with an exponential time course. These results agree with our earlier observations on a semi‐intact feeding preparation in which we showed that the firing frequency of the Retzius cell decreased as soon as the leech began to ingest its meal. Therefore, our data provide further evidence suggesting that it is unlikely that heat or chemical cues maintain the Retzius cell in an active state throughout the consummatory phase of feeding. © 2000 John Wiley & Sons, Inc. J Neurobiol 43: 304–311, 2000     

Anatomical pathways connecting lip sensory structures and central nervous system in hirudinid leeches visualized by carbocyanine dyes and laser scanning confocal microscopy

Chemoreception inHirudo medicinalis is thought to be mediated by ciliated cells grouped in sensory structures, the sensilla, arranged in bands on the animal's dorsal lip (Elliott, 1986; Zipseret al., 1994). Furthermore, chemical and/or thermal stimulation of the dorsal lip in reduced preparations evokes changes in the electrical activity of the cephalic nerves that connect the head with the central nervous system. However, the complete trajectory by which the sensory afferents teach the cerebral ganglia has not been demonstrated anatomically. In this study, we traced these pathways following retrograde and/or anterograde transport of carbocyanine dyes (DiI, DiA and DiD) in the cephalic nerves ofHirudo medicinalis and a closely related species,Macrobdella decora. While information regardingMacrobdella's chemoreception is scarce, the two species show some differences with regard to their chemical preferences. Dyes were applied to the sensillar structures along the dorsal lip, or to the cut ends of individual cephalic nerves in fixed preparations that included the lip and attached nerves with or without the head ganglia. After a two week incubation, specimens were mounted and imaged using a confocal microscope.The results show that the axons of the sensory neurons in the sensilla project through the four pairs of cephalic nerves. The sensillar projections are however more numerous in the dorsal nerves than they are in the ventral ones. In addition, the organization of the sensillar bands, the morphology of the pathways and the sensory structures themselves appear to be identical forHirudo andMacrobdella and therefore the behavioral differences in response to appetitive stimuli cannot be readily explained by differences in morphology.     

Optical detection of neuromodulatory effects of conditioned taste aversion in the pond snail Lymnaea stagnalis

Multiple site optical recording was used to analyze the neural activity changes caused by conditioned taste aversion (CTA) training in the pond snail Lymnaea stagnalis. In response to electrical stimulation of the median lip nerve, which transmits chemosensory signals of appetitive taste to the central nervous system, we optically detected large numbers of spikes in several parts of the buccal ganglion. The effects of CTA training on the spike responses were examined in two areas of the ganglion where the most active neural responses occurred. In one area (termed Area I) that included the N1 medial (N1M) cells, a class of central pattern generator interneurons involved in feeding behavior, the number of spikes in a period 1500-2000 ms after median lip nerve stimulation was significantly reduced in conditioned animals compared to control animals. In another area (termed Area II) positioned between buccal motoneurons, the B3 and B4CL (cluster) cells, the evoked spike responses were unaffected by CTA training. These results, taken together with our previous results indicating an enhancement of an inhibitory input to the N1M cells during CTA, suggest that an appetitive taste signal transmitted to the N1M cells through the median lip nerves is suppressed during CTA, resulting in a decrease of the feeding response.     

Input and output organization of cerebral neurones responding to tactile and taste stimuli applied to the lip of Helix pomatia L

Mechano- and chemoafferent responsiveness as well as outputs of identified cerebral neurones were investigated by electrophysiological methods in Helix pomatia L. the axonal projections of the identified cells were studied by intracellular staining. The studied neurones proved to be unipolar, their main axon branches were found in ipsilateral lip nerves. They could be divided into several groups according to their spontaneous activity, input and output organization and the selectivity of their responses to different tactile and taste stimuli applied to the lip. The activity of most of the neurones could be influenced by both ipsi- and contralateral inputs. They receive afferent input mostly through the medial lip nerves and their efferent information is transferred to the periphery mainly through the pair of inner lip nerves. There were seven neurones among the identified cells which responded selectively to taste stimuli identified in previous behavioural tests as phagostimulants. They can be considered as elements of the cerebral system regulating taste discrimination and feeding.     

Control of leech swimming activity by the cephalic ganglia

We investigated the role played by the cephalic nervous system in the control of swimming activity in the leech, Hirudo medicinalis, by comparing swimming activity in isolated leech nerve cords that included the head ganglia (supra- and subesophageal ganglia) with swimming activity in nerve cords from which these ganglia were removed. We found that the presence of these cephalic ganglia had an inhibitory influence on the reliability with which stimulation of peripheral (DP) nerves and intracellular stimulation of swim-initiating neurons initiated and maintained swimming activity. In addition, swimming activity recorded from both oscillator and motor neurons in preparations that included head ganglia frequently exhibited irregular bursting patterns consisting of missed, weak, or sustained bursts. Removal of the two head ganglia as well as the first segmental ganglion eliminated this irregular activity pattern. We also identified a pair of rhythmically active interneurons, SRN1, in the subesophageal ganglion that, when depolarized, could reset the swimming rhythm. Thus the cephalic ganglia and first segmental ganglion of the leech nerve cord are capable of exerting a tonic inhibitory influence as well as a modulatory effect on swimming activity in the segmental nerve cord.     

Coding of sweet,bitter, and umami tastes: different receptor cells sharing similar signaling pathways

Mammals can taste a wide repertoire of chemosensory stimuli. Two unrelated families of receptors (T1Rs and T2Rs) mediate responses to sweet, amino acids, and bitter compounds. Here, we demonstrate that knockouts of TRPM5, a taste TRP ion channel, or PLCbeta2, a phospholipase C selectively expressed in taste tissue, abolish sweet, amino acid, and bitter taste reception, but do not impact sour or salty tastes. Therefore, despite relying on different receptors, sweet, amino acid, and bitter transduction converge on common signaling molecules. Using PLCbeta2 taste-blind animals, we then examined a fundamental question in taste perception: how taste modalities are encoded at the cellular level. Mice engineered to rescue PLCbeta2 function exclusively in bitter-receptor expressing cells respond normally to bitter tastants but do not taste sweet or amino acid stimuli. Thus, bitter is encoded independently of sweet and amino acids, and taste receptor cells are not broadly tuned across these modalities.     

Morphology of chemosensory organs required for feeding in the leech Hirudo medicinalis

Sensilla that line the upper edge of the lip in the leech Hirudo medicinalis and that contain chemoreceptors required for feeding were examined in the scanning and transmission electron microscopes. The sensilla include two size-classes of ciliated button-like mounds--one about 35 microns in diameter and another about 10 microns in diameter. The larger sensilla are at the center of unpigmented patches of skin which are visible in the light microscope, while the smaller sensilla have not been previously described as distinct structures. Electron microscopy, though not light microscopy, shows that the lip sensilla differ markedly from the segmental sensilla of the leech, which have been shown to mediate mechanoreception and photoreception. In particular, the chemosensory lip sensilla contain multiciliated cells with cilia of a uniform length, whereas the segmental sensilla contain uniciliated cells with long, whip-like cilia, as well as multiciliated cells with short, stiff cilia. Thus, the two types of sensilla differ morphologically as well as functionally. In addition to the ciliated sensilla along the upper lip, structures consisting of a short, club-like process surrounded by granular material were observed inside the mouth. These structures may also be chemosensory organs.     

Diverse bitter stimuli elicit highly similar patterns of Fos-like immunoreactivity in the nucleus of the solitary tract

Previous studies have demonstrated that oral stimulation with quinine elicits Fos-like immunoreactivity in the first-order gustatory nucleus, the NST, with a different topographic distribution than sucrose or citric acid. However, it is unknown whether the quinine pattern is unique to this alkaloid or common across bitter stimuli with different chemical structures. Indeed, recent physiological experiments suggest that taste receptor cells and primary afferent neurons may exhibit selectivity for various bitter tastants. The present investigation compared the distribution of FLI in NST following stimulation with three bitter chemicals: QHCl, denatonium and propylthiouracil, stimuli that evoked Ca(2+) currents in almost entirely different sets of receptor cells. The results demonstrate that the quinine pattern is not idiosyncratic but instead generalizes to the other two tastants. Although it remains possible that intermingled but different NST neurons are activated by these stimuli, these data suggest that a specialized region in the NST is preferentially involved in processing a common aspect of bitter tastants. In contrast to citric acid, quinine, denatonium and propylthiouracil all elicited vigorous oromotor rejection responses, consistent with our earlier hypothesis that the medial third of the NST may be an afferent trigger zone for oromotor rejection.     

Chemosensory stimuli in feeding behavior of the leechHirudo medicinalis

The involvement of chemotherapy stimuli in the feeding behavior of the blood-sucking leech Hirudo medicinalis was investigated using a behavioral feeding test in which test solutions were encased in a highly permeable membrane and presented to the leech. Whole human blood or plasma at ambient temperature elicited the complete sequence of feeding behavior: probing, attachment, biting and ingestion. Spring water, 300 mM sucrose, or dialyzed plasma did not elicit any of these responses. Spring water warmed to 38 degrees C elicited probing and transient attachment but not ingestion. Thus, appropriate chemical stimuli were necessary for complete feeding behavior. A chemically defined artificial blood mix, containing the major components of low molecular weight found in blood, elicited all aspects of leech feeding behavior. Eliminating either NaCl or arginine from the mix resulted in complete loss of effectiveness. Moreover, a solution containing only NaCl (150 mM) and arginine (90 microM) was also an effective feeding stimulus. Thus, appropriate chemical stimuli are sufficient for complete feeding behavior. Neither NaCl nor arginine alone induced feeding although NaCl alone elicited probing. Sensory detection of blood was localized to a region of the dorsal lip that contains structures composed of ciliated, bipolar neurons, which are likely candidates as chemoreceptors. Surgical ablation of this region of the skin resulted in complete loss of ability to alert to, orient toward and ingest blood, while sham-operated controls fed normally. Substitution with other ions revealed specificity, with respect to both the cation and the anion, in the response to NaCl. Of the inorganic and organic cations tested, only Li+ substituted effectively for Na+. Of the inorganic and organic anions tested, only Br- was as effective as Cl-. Thus, the requirement for NaCl in leech feeding represents more than simply an ionic strength requirement or a requirement for Na+ ions and bears similarities to the chemosensory detection of NaCl in other species. Substitution with other amino acids and analogues for arginine revealed marked specificity in the feeding response to this compound as well. D-arginine at concentrations of up to 1000-fold greater than the effective threshold for L-arginine did not elicit ingestion, nor did other common L-amino acids, including the other basic amino acids histidine and lysine. Of the arginine analogues tested, only homoarginine and canavanine (in which all three functional groups of arginine are unchanged) were effective feeding stimulants.     

The selective serotonin reuptake inhibitor paroxetine does not alter consummatory concentration-dependent licking of prototypical taste stimuli by rats

Serotonin and the 5HT(1A) receptor are expressed in a subset of taste receptor cells, and the 5HT(3) receptor is expressed on afferent fibers innervating taste buds. Exogenous administration of the selective serotonin reuptake inhibitor, paroxetine, has been shown to increase taste sensitivity to stimuli described by humans as sweet and bitter. Serotonergic agonists also decrease food and fluid intake, and it is possible that modulations of serotonin may alter taste-based hedonic responsiveness; alternatively, or in combination, serotonin may interact with physiological state to impact ingestive behavior. In this study, the unconditioned licking of prototypical taste stimuli by rats in brief-access taste tests was assessed following paroxetine administration (0.3-10 mg/kg intraperitoneal). We also measured sucrose licking by rats in different deprivation states after paroxetine (5 mg/kg). In neither experiment did we find any evidence of an effect of paroxetine on licking relative to water to any of the taste stimuli in the brief-access test at doses that decreased food intake. However, in some conditions, paroxetine decreased trials initiated to tastants. Therefore, a systemic increase in serotonin via paroxetine administration can decrease appetitive behavior in brief-access tests but is insufficient to alter taste-guided consummatory behavior.     

“Tasting” the airway lining fluid

Specialized epithelial cells of the respiratory tract have been termed "solitary chemosensory cells" based upon the expression of components of the canonical sweet, umami and bitter taste transduction pathway, or "brush cells" based upon their characteristic morphological feature, i.e. an apical, brush-like tuft of rigid, villin containing microvilli. Cells defined by these criteria might not match one-to-one, and a generally accepted terminology is still lacking. With respect to cellular shape, ultrastructure, expression of elements of the taste transduction cascade, innervation and synapse formation, and effects evoked upon their stimulation, it appears that chemosensory/brush in the upper respiratory tract (nasal respiratory mucosa, vomeronasal duct, auditory tube), in the olfactory mucosa, in the larynx, in the lower airways (trachea, bronchi) and in the alveolar region (rat only) each represent distinct groups. Still, they have in common to monitor the chemical composition of the mucosal lining fluid. They serve as sentinels detecting bacterial colonization or the presence of other harmful components in the mucosal lining fluid, leading to the initiation of avoidance reflexes and/or local defense mechanisms which are adapted to their anatomical localization. Free nerve endings are also responsive to inhaled irritants and further work will be needed to discriminate between the contributions of such nerve endings and chemosensory cells in chemical monitoring and defense initiation. Interestingly, there is first emerging evidence that respiratory chemosensory cells may respond to more than one canonical taste quality so that they, in analogy to polymodal nociceptors, may serve as polymodal chemosensors of potentially dangerous signals.     

The pharmacology and signaling of bitter, sweet, and umami taste sensing

Over the last decade, many of the molecular components that mediate the transduction of taste signaling have been elucidated. The chemosensory receptors for taste have been identified as G protein-coupled receptors (GPCRs) and ion channels that are expressed on the surface of highly specialized taste sensory cells. Tastant molecules act as agonists, binding to and stabilizing active conformations of receptors, resulting in the initiation of signal transduction cascades. Taste signaling, therefore, should be amenable to the methods of pharmacology. This review focuses on the GPCR-mediated signaling of bitter, sweet, and umami tastes and emphasizes the opportunities for pharmacologic evaluation.     

Patch clamp recording of the responses to three bitter stimuli in mouse taste cells.

Although several pathways of bitter taste signal transduction have been proposed in taste cells, these mechanisms have not been elucidated in detail. To investigate the diversity of responses to bitter stimuli, we recorded the electrophysiological responses to quinine, denatonium and naringin using whole-cell patch clamp technique in isolated taste cells of C57BL/6J mice. Ten mM quinine induced depolarizing response under the current clamp mode, and inward current response under the voltage-clamp mode (holding potential -80 mV) using both K+ (with pseudo intracellular solution) and Cs+ (K+ was substituted by Cs+ in the pseudo intracellular solution) pipettes. However, when the K+ pipette was used, the membrane conductance was suppressed and activated in succession. On the other hand, the membrane conductance was only activated when the Cs+ pipette was used. Half to one mM denatonium induced depolarizing response under the current clamp mode, and outward current response under the voltage clamp mode with both pipettes. Using these pipettes, the membrane conductance was activated or suppressed in the individual case. Naringin-induced responses were not detected in these measurements. These electrophysiological recordings suggest that multiple transduction mechanisms are involved in bitter taste perception in mouse taste cells.     

Glutamate as a transmitter in the sensory pathway from prostomial lip to serotonergic Retzius neurons in the medicinal leechHirudo

The involvement of glutamate in putative ingestive sensory pathways affecting the excitability of serotonergic Retzius neurons (RZ) in the leech CNS was investigated with a pharmacological approach. Exposure of the prostomial lip to 150mm NaCl and 1mm arginine produced excitatory as well as inhibitory responses in RZ found in the reproductive segments, while only excitatory responses were elicted in standard midbody RZ. Antagonists of glutamatergic receptors of the kainate/quisqualate type effectively inhibited chemosensory dependent excitation of RZ. Antagonists of glutamatergic receptors of theN-methyld-aspartate type were ineffective in this regard. Cephalic nerve stimulation, like chemical stimulation of the lip, produced segment-specific responses in midbody RZ. Both the polysynaptic and monosynaptic components of the excitatory response of standard midbody RZ following cephalic nerve stimulation were inhibited in the presence of the kainate/quisqualate antagonist DNQX. These data suggest a role for glutamate as a transmitter in the neural circuitry from receptors of the leech prostomial lip to serotonergic RZ.     

Identification and characterization of human taste receptor genes belonging to the TAS2R family

The sense of taste is a chemosensory system responsible for basic food appraisal. Humans distinguish between five primary tastes: bitter, sweet, sour, salty and umami. The molecular events in the perception of bitter taste are believed to start with the binding of specific water-soluble molecules to G-protein-coupled receptors encoded by the TAS2R/T2R family of taste receptor genes. TAS2R receptors are expressed at the surface of taste receptor cells and are coupled to G proteins and second messenger pathways. We have identified, cloned and characterized 11 new bitter taste receptor genes and four new pseudogenes that belong to the human TAS2R family. Their encoded proteins have between 298 and 333 amino acids and share between 23 and 86% identity with other human TAS2R proteins. Screening of a mono-chromosomal somatic cell hybrid panel to assign the identified bitter taste receptor genes to human chromosomes demonstrated that they are located in chromosomes 7 and 12. Including the 15 sequences identified, the human TAS2R family is composed of 28 full-length genes and 16 pseudogenes. Phylogenetic analyses suggest a classification of the TAS2R genes in five groups that may reflect a specialization in the detection of specific types of bitter chemicals.     

Central representation and functional connections of afferent and efferent pathways of Helix pomatia L. lip nerves

Localization and distribution of cerebral neurones sending axons into the three pairs of Helix pomatia lip nerves were investigated by the method of retrograde axonal NiCl2 transport. Using electrophysiological technics (extracellular recordings) the dependence of lip nerve's activity on inputs of other lip nerves was studied after application of various types of stimuli to the lip of semi-intact preparations. All lip nerves have neuronal representation in each lobe of the cerebral ganglia but in different proportions. Labelled neurones were located mainly on the ventral surface of the cerebral ganglia, most of them projecting to the medial, the least to the inner lip nerve. Lip nerves differ from each other according to the proportions of neurones of various size. They share in the axons of large (55-70 microns) and medium sized (30-40 microns) neurones in the order inner greater than outer greater than medial and medial greater than outer greater than inner lip nerve, respectively. Most neurones projecting to different nerves are located in discrete groups. According to the electrophysiological results the medial lip nerve has the most prominent afferent, while the inner one has the strongest efferent activity. Both the afferent and efferent activities of the outer lip nerve proved to be the least significant compared to the other lip nerves. Contralateral cerebral connections play an important role in the sensory information processing. The sensory input of a given nerve usually activates the contralateral member of another pair of lip nerves. Mechano- and chemo-afferent pathways have almost the same properties but there are some differences in latencies and other parameters.     

Octopamine mediates rapid stimulation of protein kinase A in the antennal lobe of honeybees

In the honeybee octopamine mediates mechanisms of arousal that interfere with the appetitive proboscis extension response to food-indicating chemosensory stimuli. This study demonstrates that injections of octopamine or cyclic adenosine monophosphate (cAMP) into the primary chemosensory neuropil of the honeybee, the antennal lobe, evokes a rapid and transient activation of cAMP-dependent protein kinase (PKA). Other monoamines detectable in the antennal lobe, dopamine and serotonin, do not affect the level of PKA activity. Stimulation of the bees' antenna with the appetitive stimulus water or sucrose solution in vivo also causes a short-term activation of PKA in the antennal lobe. The increased PKA activity can be detected immediately (0.5 s) after stimulation but reverts to the basal level within 3 s. This effect can be abolished by monoamine depletion with reserpine. Since octopamine is the only monoamine that stimulates PKA, it appears to mediate the PKA activation after sucrose stimulus and may contribute to the processing of this chemosensory input. © 1995 John Wiley & Sons, Inc.     

Bitter taste stimuli induce differential neural codes in mouse brain

A growing literature suggests taste stimuli commonly classified as "bitter" induce heterogeneous neural and perceptual responses. Here, the central processing of bitter stimuli was studied in mice with genetically controlled bitter taste profiles. Using these mice removed genetic heterogeneity as a factor influencing gustatory neural codes for bitter stimuli. Electrophysiological activity (spikes) was recorded from single neurons in the nucleus tractus solitarius during oral delivery of taste solutions (26 total), including concentration series of the bitter tastants quinine, denatonium benzoate, cycloheximide, and sucrose octaacetate (SOA), presented to the whole mouth for 5 s. Seventy-nine neurons were sampled; in many cases multiple cells (2 to 5) were recorded from a mouse. Results showed bitter stimuli induced variable gustatory activity. For example, although some neurons responded robustly to quinine and cycloheximide, others displayed concentration-dependent activity (p<0.05) to quinine but not cycloheximide. Differential activity to bitter stimuli was observed across multiple neurons recorded from one animal in several mice. Across all cells, quinine and denatonium induced correlated spatial responses that differed (p<0.05) from those to cycloheximide and SOA. Modeling spatiotemporal neural ensemble activity revealed responses to quinine/denatonium and cycloheximide/SOA diverged during only an early, at least 1 s wide period of the taste response. Our findings highlight how temporal features of sensory processing contribute differences among bitter taste codes and build on data suggesting heterogeneity among "bitter" stimuli, data that challenge a strict monoguesia model for the bitter quality.