Nucleus accumbens neurons are innately tuned for rewarding and aversive taste stimuli, encode their predictors, and are linked to motor output

The nucleus accumbens (NAc) is a key component of the brain's reward pathway, yet little is known of how NAc cells respond to primary rewarding or aversive stimuli. Here, naive rats received brief intraoral infusions of sucrose and quinine paired with cues in a classical conditioning paradigm while the electrophysiological activity of individual NAc neurons was recorded. NAc neurons (102) were typically inhibited by sucrose (39 of 52, 75%) or excited by quinine (30 of 40, 75%) infusions. Changes in firing rate were correlated with the oromotor response to intraoral infusions. Most taste-responsive neurons responded to only one of the stimuli. NAc neurons developed responses to the cues paired with sucrose and quinine. Thus, NAc neurons are innately tuned to rewarding and aversive stimuli and rapidly develop responses to predictive cues. The results indicate that the output of the NAc is very different when rats taste rewarding versus aversive stimuli.     

Rapid calorie metering inad lib rats

Adult wistar rats of either sex housed in individual cages and fedad lib were used to investigate the effects of calorie content and taste on ingestion. A 15 min single-choice solution (water, quinine, sucrose, saccharin) test as well as 1 h mixed diet (stock diet, sucrose-mixed, saccharin-mixed, quinine-mixed) tests were used. Calorically-rich sucrose either in solution or in mixed-diet was preferentially ingested. Tasteper se has not influenced calorie intake, as intake on sweet saccharin (3.4 ±0.4 cal) or on bitter quinine (3.3 ±0.6 cal) was similar to intake on stock diet (4.8 ±0.8 cal). Increased 5 min intake on sucrose solution (4.6 ±0.1 ml) over intake of other test solutions (saccharin 3.5 ±0.2 ml, quinine, 1.2 ±0.3 ml, water 2.1 ±0.1 ml) and calorie intake suppression on 1 h stock diet immediately following sucrose solution intake, indicate rapid calorie metering, probably based on fast-acting specific gustatory signals     

PLCbeta2-independent behavioral avoidance of prototypical bitter-tasting ligands

Using a brief-access taste assay, we show in the present report that although phospholipase C beta2 knockout (PLCbeta2 KO) mice are unresponsive to low- and midrange concentrations of quinine and denatonium, they do significantly avoid licking higher concentrations of these aversive compounds. PLCbeta2 KO mice displayed no concentration-dependent licking of the prototypical sweetener sucrose but were similar to wild-type mice in their responses to citric acid and NaCl, notwithstanding some interesting exceptions. Although these findings confirm an essential role for PLCbeta2 in taste responsiveness to sucrose and to low- to midrange concentrations of quinine and denatonium in mice as previously reported, they importantly suggest that higher concentrations of the latter two compounds, which are bitter to humans, can engage a PLCbeta2-independent taste transduction pathway.     

Striatal-Enriched Protein Tyrosine Phosphatase Controls Responses to Aversive Stimuli: Implication for Ethanol Drinking

The STriatal-Enriched protein tyrosine Phosphatase (STEP) is a brain-specific phosphatase whose dysregulation in expression and/or activity is associated with several neuropsychiatric disorders. We recently showed that long-term excessive consumption of ethanol induces a sustained inhibition of STEP activity in the dorsomedial striatum (DMS) of mice. We further showed that down-regulation of STEP expression in the DMS, and not in the adjacent dorsolateral striatum, increases ethanol intake, suggesting that the inactivation of STEP in the DMS contributes to the development of ethanol drinking behaviors. Here, we compared the consequence of global deletion of the STEP gene on voluntary ethanol intake to the consumption of an appetitive rewarding substance (saccharin) or an aversive solution (quinine or denatonium). Whereas saccharin intake was similar in STEP knockout (KO) and wild type (WT) littermate mice, the consumption of ethanol as well as quinine and denatonium was increased in STEP KO mice. These results suggested that the aversive taste of these substances was masked upon deletion of the STEP gene. We therefore hypothesized that STEP contributes to the physiological avoidance towards aversive stimuli. To further test this hypothesis, we measured the responses of STEP KO and WT mice to lithium-induced conditioned place aversion (CPA) and found that whereas WT mice developed lithium place aversion, STEP KO mice did not. In contrast, conditioned place preference (CPP) to ethanol was similar in both genotypes. Together, our results indicate that STEP contributes, at least in part, to the protection against the ingestion of aversive agents.     

Self-selection of Dietary Threonine in the Rat and the Effect of Taste Stimuli on its Selection

The possible regulation of dietary threonine intake and the effect of a palatable (sodium saccharin) or aversive (quinine sulfate) taste stimulus on the manner of threonine selection were investigated in rats using a self-selection feeding method. Weanling rats were offered the choice of two diets differing only in threonine content for 2 weeks. Both weight gain and food consumption in the rats offered the choice of diets were quite comparable to each other and were the same as those in rats fed one diet containing sufficient threonine. Threonine intake of the self-selecting rats ranged from 0.43 to 1.90% of the food consumed. The threonine concentrations in the plasma and brain of the self-selecting rats increased proportionally with the threonine intake. When rats were offered a choice of two diets containing various amounts of threonine with taste materials, i.e., sodium saccharin or quinine sulfate, neither the dietary threonine nor their growth were ever affected.

These results indicate clearly that rats have an ability to regulate threonine intake to meet their requirement for the l-amino acid, and the threonine selection is not influenced significantly by the dietary addition of palatable or aversive taste materials.     

Glossopharyngeal nerve transection does not alter taste reactivity to sucrose conditioned to be aversive

Glossopharyngeal nerve (GL) transection in rats is known to markedly reduce gaping, a stereotypical aversive oromotor behavior, in response to intraorally delivered quinine. In this experiment we tested whether GL transection would reduce gaping in response to an otherwise palatable stimulus (sucrose) but conditioned to be aversive. Sprague-Dawley rats were implanted with intraoral cannulae. Five received bilateral transection of the GL and five served as sham-operated controls. Water-deprived rats were presented with 0.3 M sucrose for 15 min immediately followed by an injection of 0.15 M LiCl on three occasions. Rats were then habituated to the taste reactivity chamber and intraoral fluid infusion for 3 days, and tested on day 4 with a 1 ml infusion (1 min) of 0.3 M sucrose. All rats drank negligible amounts of sucrose by the third conditioning session and there were no differences in sucrose intake between the groups. There were no significant differences in gapes, or any other measured oromotor response, to sucrose between GL-transected and sham-operated rats. These results show that the GL is not a necessary afferent limb for gaping in response to conditionally aversive taste compounds.     

Relationships between taste reactivity and intake in the neurologically intact rat

To investigate the differences between short-term intake testsand taste reactivity responses to tastes, rats received 1-minintraoral infusions of a variety of tastants delivered at therate of 1 ml/min. Oral responses were videotaped and analysedin terms of the sequence and number of ingestive and aversivetaste reactivity response components evoked. Intake was alsomeasured. The number of rats displaying ingestive taste reactivitycomponents and the mean number of ingestive components displayedper rat elicited by sucrose and NaCl increased with increasingconcentration. Intake was high across all concentrations. HClinfusions elicited alternation between ingestive and aversiveresponse components. The number of rats displaying aversivetaste reactivity response components and the mean number ofaversive response components displayed per rat, elicited byQHCl, increased with increasing concentration, while both intakeand the median latency to reject QHCl decreased (ExperimentI). To determine whether other tastes judged bitter by humanswould elicit a quinine-like taste reactivity response in therat, sucrose octa-acetate (SOA), quinine sulfate (QS) and caffeine(CAF) stimuli were examined. Both QS and CAF infusions elicitedan increased number of aversive response components with increasingconcentration, and intake decreased. SOA infusions elicitedalternation between ingestive and aversive response componentsfollowed by a display of solely aversive components, and bothintake and median latency to reject the infusions decreasedsignificantly with increased concentration (Experiment II).Experiment I demonstrated that hypertonic NaCl infusions elicitingestive response components, while short-term intake testsshow that hypertonic NaCl is rejected and is thus inferred tobe aversive. Rats received prolonged infusions of hypertonicNaCl solutions at the rate of 1 ml/min until fluid was seenon the surface of the chamber, indicating rejection. Prolongedinfusions of hypertonic NaCl solutions elicited an initial displayof solely ingestive response components followed by an abruptshift to solely aversive response components and active fluidrejection. Higher concentrations elicited this shift soonerthan lower ones (Experiment III). The results suggest that patternsof taste reactivity response components are good predictorsof intake duration.     

Suppression of S-methylglutathione-induced tentacle ball formation by peptides and nullification of the suppression by TGF-beta in Hydra

Tentacle ball formation (TBF) in Hydra elicited by S-methylglutathione (GSM) was modulated by a number of biologically active peptides. Hydra fed on Artemia, which had been hatched in a common salt solution supplemented with LiCl and ZnCl(2), easily induced TBF in response to GSM after pretreatment with trypsin. After Hydra were treated with 100 pg/ml trypsin for 10 min, the response to GSM (TBF) was sensitively suppressed by acidic fibroblast growth factor and other biologically active peptides for >10 h. Various peptides, but not transforming growth factor beta (TGF-beta), suppressed GSM-induced TBF in a specific pattern for each peptide. However, TGF-beta was unique in that it did not suppress the response to GSM, but nullified the suppressive effect of other peptides. Only active TGF-beta nullified the suppressive effect of the peptides, and the latent form of TGF-beta neither suppressed GSM-induced TBF nor nullified the suppressive effect of other peptides. Members of the TGF-beta family suppressed GSM-induced TBF. These results indicate that all peptides examined, except for TGF-beta suppressed the response to GSM in a manner specific to each peptide. This assay system would be useful in identification of biologically active peptides.     

μ-Opioid modulation in the rostral solitary nucleus and reticular formation alters taste reactivity: evidence for a suppressive effect on consummatory behavior

The neural control of feeding involves many neuromodulators, including the endogenous opioids that bind μ-opioid receptors (MORs). Injections of the MOR agonist, Damgo, into limbic and hypothalamic forebrain sites increase intake, particularly of palatable foods. Indeed, forebrain Damgo injections increase sucrose-elicited licking but reduce aversive responding (gaping) to quinine, suggesting that MOR activation may enhance taste palatability. A μ-opioid influence on taste reactivity has not been assessed in the brain stem. However, MORs are present in the first-order taste relay, the rostral nucleus of the solitary tract (rNST), and in the immediately subjacent reticular formation (RF), a region known to be essential for consummatory responses. Thus, to evaluate the consequences of rNST/dorsal RF Damgo in this region, we implanted rats with intraoral cannulas, electromyographic electrodes, and brain cannulas aimed at the ventral border of the rNST. Licking and gaping elicited with sucrose, water, and quinine were assessed before and after intramedullary Damgo and saline infusions. Damgo slowed the rate, increased the amplitude, and decreased the size of fluid-induced lick and gape bouts. In addition, the neutral stimulus water, which typically elicits licks, began to evoke gapes. Thus, the current results demonstrate that μ-opioid activation in the rNST/dorsal RF exerts complex effects on oromotor responding that contrast with forebrain effects and are more indicative of a suppressive, rather than a facilitatory effect on ingestion.     

Tamoxifen produces conditioned taste avoidance in male rats: An analysis of microstructural licking patterns and taste reactivity

Estrogen receptor activation has been shown to reduce body weight and produce a conditioned reduction in food intake in male rats that is putatively mediated by estradiol's suggested aversive effects. Evidence has shown that the selective estrogen receptor modulator tamoxifen used in the prevention and treatment of breast cancer may also produce changes in food intake and body weight, which are known to impact cancer development and survival. The purpose of the present study was to examine whether tamoxifen produces a conditioned reduction in intake similar to estradiol by producing a conditioned aversion. A one bottle lickometer test was used to examine conditioned changes in sucrose drinking, while the taste reactivity test was used to measure rejection reactions, which serve to index aversion in rats. A backward conditioning procedure that consisted of 3 conditioning days and one vehicle test day was used to examine conditioned changes in 0.3 M sucrose intake and taste reactivity. Our results show that tamoxifen produced a conditioned reduction in sucrose drinking in a one bottle fluid intake test that was similar to the effects produced by estradiol (positive control); however, no active rejection reactions were produced by either tamoxifen (1 and 10 mg/kg) or estradiol. The present results suggest that tamoxifen, at the doses used in the present study, acts as an estrogen receptor agonist to regulate food intake and that the conditioned reduction in intake produced by tamoxifen and estradiol reflects conditioned taste avoidance rather than conditioned taste aversion.     

Acute 3rd-ventricular amylin infusion potently reduces food intake but does not produce aversive consequences

In this study, a conditioned taste aversion (CTA) paradigm was used to assess the possibility that 3rd-ventricular (i3vt) administration of the pancreatic hormone amylin produces aversive consequences that secondarily reduce food intake independently of the normal regulation of energy balance. After 1-h daily access to water for 7 days, rats were given 1-h access to a 0.15% saccharin solution, followed immediately by i3vt amylin (100 pmol) in one group (n=7) and i3vt CSF vehicle in another (n=7). As positive control for the formation of a CTA, a third group of seven rats received intraperitonial (i.p.) lithium chloride (LiCl). Saline was given i.p. to a fourth group (n=7) as control for i.p. LiCl. As expected, the LiCl rats exhibited a marked aversion to the saccharin in a subsequent two-bottle intake test. In contrast, although the 100 pmol i3vt amylin dose is substantially higher than that required to reduce food intake, no evidence of a CTA was observed in the rats that had received i3vt amylin. In summary, these data are consistent with the conclusion that acute i3vt amylin infusion does not reduce food intake by producing aversive consequences.     

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.     

味觉刺激引起大鼠臂旁核神经元抑制性反应

本研究以轻度麻醉的大鼠为对象,应用细胞外微电极记录技术,观察并分析了脑桥臂旁核抑制性味觉神经元的自发活动及其对NaCl、HCl、盐酸奎宁（quinine HCl,QHCl））和蔗糖等四种基本味觉刺激的反应。共分析了18个具有自发活动的抑制性味觉神经元,自发放电频率分布在0．2～5．5Hz之间,平均放电频率（2．15±0．31）Hz。18个神经元中,1个神经元对单一味觉刺激呈抑制性反应,其余17个神经元对两种或两种以上的基本味觉刺激发生抑制性反应,且抑制具有潜伏期短、持续时间较长等特征。抑制持续时间5～80S,部分神经元表现为后抑制效应。根据神经元对四种基本味觉刺激呈抑制性反应的程度,将其分为NaCl优势神经元（n=8）,HCl优势神经元（n=3）,QHCl优势神经元n=3）和蔗糖优势神经元n=4）。其中NaCl优势神经元的反应谐宽最高（0．945）。这些神经元对欣快或厌恶刺激的区别能力较低。结果提示,在脑桥臂旁核存在对味觉刺激起抑制性反应的神经元,这些味觉神经元可能在味觉的调制及对欣快和厌恶刺激的编码中发挥重要的作用。     

Antennal gustatory perception and behavioural responses in Trissolcus brochymenae females

Trissolcus brochymenae (Hymenoptera: Platygastridae) is an egg parasitoid that could be used to control stink bugs like Murgantia histrionica (Heteroptera: Pentatomidae), a pest of brassicaceous crops. Before laying their eggs, parasitoid females spend considerable time examining the substrate with their antennae, which are also used during feeding and mating behaviours. This suggests that contact chemoreception plays a prominent role in many aspects of parasitoid ecology. Therefore, we examined the sensitivity of antennal taste sensilla chaetica to several classical tastants including those that are appetitive or aversive. In addition we studied the taste preferences of wasps when presented with these chemicals alone or mixed. The sensilla chaetica of T. brochymenae responded to serial concentrations of sucrose, salts, and quinine, but no concentration-dependent effect was observed when testing sinigrin, a secondary metabolite found in many brassicaceae. However, both sinigrin and quinine inhibited responses to 0.1 M sucrose when mixed with this sugar. Behavioural taste preference assays confirmed that wasps showed a dose dependent preference for sucrose over agarose. In addition, a behavioural avoidance of sucrose solutions containing quinine was observed. This effect was not observed when sinigrin was used as a feeding deterrent. In the two-choice tests the wasp did not discriminate between sucrose solutions mixed with salts and sucrose alone. Further no preference for salts or sinigrin compared to agarose alone was observed. This work represents the first step towards the identification of gustatory receptor neurons implicated in the detection of different types of chemical cues in egg parasitoids.     

Central mechanisms of roles of taste in reward and eating

Taste is unique among sensory systems in its innate association with mechanisms of reward and aversion in addition to its recognition of quality, e.g., sucrose is sweet and preferable, and quinine is bitter and aversive. Taste information is sent to the reward system and feeding center via the prefrontal cortices such as the mediodorsal and ventrolateral prefrontal cortices in rodents and the orbitofrontal cortex in primates. The amygdala, which receives taste inputs, also influences reward and feeding. In terms of neuroactive substances, palatability is closely related to benzodiazepine derivatives and beta-endorphin, both of which facilitate consumption of food and fluid. The reward system contains the ventral tegmental area, nucleus accumbens and ventral pallidum and finally sends information to the lateral hypothalamic area, the feeding center. The dopaminergic system originating from the ventral tegmental area mediates the motivation to consume palatable food. The actual ingestive behavior is promoted by the orexigenic neuropeptides from the hypothalamus. Even palatable food can become aversive and avoided as a consequence of a postingestional unpleasant experience such as malaise. The neural mechanisms of this conditioned taste aversion will also be elucidated.     

Changes in the activity of the taste receptor apparatus and the preference for a sodium chloride solution in newborn rats receiving capsaicin]

Taste receptor organ activity and preference of sodium chloride solution in rats with deficit of substance P (SP) were studied. Total impulse activity of chorda tympani nerve of 7-8 week old rats was recorded under nembutal anesthesia. The taste responses to four solutions (sucrose, quinine sulfate, sodium chloride and citric acid) were decreased in rats injected with capsaicin in comparison with rats injected with vehicle. The rats injected with capsaicin preferred water to sodium chloride (two-bottle technique). On the contrary the rats injected with capsaicin preferred the salt solution. These data together with previous studies show the important role of peptide SP in taste receptor activity and "salt appetite".     

Effects of isoproterenol treatment on gustatory neural responses in three inbred strains of mice

1. Treatment of a beta-agonist, isoproterenol, for 5 days reduced chorda tympani responses to sucrose by about 40% of the control without affecting responses to other taste stimuli, such as NaCl, HCl and quinine HCl, in balb CrSlc mice whereas such reduction of sucrose responses was not observed in C57BL/6-CrSlc and C3H/HeSlc mice, although in the latter two strains long-lasting off-responses to quinine HCl appeared after the treatment. 2. In BALB mice, the magnitude of reduction of sucrose responses by isoproterenol increased with prolonging the treatment from 1 to 5 days, although it reached almost its maximum level by the 3 days treatment. 3. BALB mice with the removal of the submandibular glands showed slightly greater control responses of the chorda tympani nerve to sucrose than BALB mice with the sham-operation or the removal of the sublingual glands, and showed no significant reduction of sucrose responses by isoproterenol treatment. 4. These results suggest that isoproterenol probably did not act directly on sweetener receptors of taste cell membranes but affect them through the submandibular salivary system.     

Anatomical dissociation of melanocortin receptor agonist effects on taste- and gut-sensitive feeding processes

Injections of the melanocortin 3/4 receptor (MCR) agonist melanotan II (MTII) to a variety of brain structures produces anorexia, suggesting distributed brain MCR control of food intake. We performed a detailed analysis of feeding behavior (licking microstructure analysis) after a range of MTII doses (0.005 nM to 1 nM) was targeted to the forebrain (third ventricle, 3V) or hindbrain (fourth ventricle, 4V) regions. MTII (0.1 nM and 1 nM) delivered to the 3V or 4V significantly reduced 0.8 M sucrose intake. The anorexia was mediated by reductions in the number of licking bursts in the meal, intrameal ingestion rate, and meal duration; these measures have been associated with postingestive feedback inhibition of feeding. Anorexia after 3V but not 4V MTII injection was also associated with a reduced rate of licking in the first minute (initial lick rate) and reduced mean duration of licking bursts; these measures have been associated with taste evaluation. MTII effects on taste evaluation were further explored: In experiment 2, 3V MTII (1 nM) significantly reduced intake of noncaloric 4 mM saccharin and 0.1 M and 1 M sucrose solutions, but not water. The anorexia was again associated with reduced number of licking bursts, ingestion rate, meal duration, initial lick rate, and mean burst duration. In experiments 3 and 4, brief access (20 s) licking responses for sweet sucrose (0.015 M to 0.25 M) and bitter quinine hydrochloride (0.01 mM to 1 mM) solutions were evaluated. Licking responses for sucrose were suppressed, whereas those for quinine solutions were increased after 3V MTII, but not after 4V MTII injections (0.1 nM and 1 nM). The results suggest that multiple brain MCR sites influence sensitivity to visceral feedback, whereas forebrain MCR stimulation is necessary to influence taste responsiveness, possibly through attenuation of the perceived intensity of taste stimuli.     

Effects of melanin-concentrating hormone on licking microstructure and brief-access taste responses

The effects of intracerebroventricular application of melanin-concentrating hormone (MCH) on licking for sucrose, quinine hydrochloride (QHCl), and water solutions were evaluated in two experiments. In experiment 1, rats received 90-min access to sucrose and water solutions after MCH or vehicle microinjection to the third ventricle (3V). MCH increased intake largely through increases in the rate of licking early in the meal and in the mean duration of lick bursts, suggesting an effect on gustatory evaluation. Therefore, in experiment 2, brief access tests were used with a series of sucrose and QHCl concentrations to behaviorally isolate the effects of intracerebroventricular MCH on gustatory evaluation. MCH uniformly increased licking for all sucrose solutions, water, and weak concentrations of QHCl; however, it had no effect on licking for the strongest concentrations of QHCl, which were generally avoided under control conditions. Thus MCH did not produce nonspecific increases in oromotor activity, nor did it change the perceived intensity of the tastants. We conclude that MCH enhanced the gain of responses to normally accepted stimuli at a phase of processing after initial gustatory detection and after the decision to accept or reject the taste stimulus. A comparison of 3V NPY and MCH effects on licking microstructure indicated that these two peptides increased intake via dichotomous behavioral processes; although NPY suppressed measures associated with inhibitory feedback from the gut, MCH appeared instead to enhance measures associated with hedonic taste evaluation.     

Induction of salivary gurmarin-binding proteins in rats fed gymnema-containing diets.

Gymnema sylvestre, a tropical plant, contains gurmarin that selectively suppresses sucrose responses of the chorda tympani nerve in rats and mice. We investigated preference for taste solutions and saliva composition in rats fed a diet containing this plant (gymnema diet). Preference for 0.01 M sucrose and a mixture of 0.03 M sucrose and 0.03 mM quinine-HCl significantly decreased at 1-2 days after the start of the gymnema diet and subsequently returned closely to the control levels within about a week. There was no significant change in preference for NaCl, monosodium glutamate and quinine-HCl during feeding trials. Submandibular saliva of rats fed the gymnema diet for 4 and 14 days showed an inhibitory effect on immunoreaction between gurmarin and antigurmarin serum. Analyses using electrophoresis and affinity chromatography indicated that the saliva contains gurmarin binding proteins with molecular weights of 15, 16, 45, 60 and 66 kDa. These results suggest that reduction of preference for sucrose was probably caused by gurmarin contained in the gymnema diet and subsequent restoration of the preference may be due to suppression of the effect of gurmarin by salivary gurmarin-binding proteins induced by the gymnema diet.