The Journal of Neurophysiology Vol. 80 No. 4 October 1998, pp. 2102-2112
Copyright ©1998 by the American Physiological Society

Gustatory Responses of the Hamster Mesocricetus auratus to Various Compounds Considered Sweet by Humans

Vicktoria Danilova¹, Göran Hellekant¹, Jean-Marie Tinti², and Claude Nofre²

¹ Animal Health and Biomedical Sciences, The University of Wisconsin-Madison, Madison, Wisconsin 53706; and ² Université Claude Bernard, Faculté de Médecine Alexis Carrel, 69372 Lyon, France

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Danilova, Vicktoria, Göran Hellekant, Jean-Marie Tinti, and Claude Nofre. Gustatory responses of the hamster Mesocricetus auratus to various compounds considered sweet by humans. J. Neurophysiol. 80: 2102-2112, 1998. The taste of 30 compounds was studied in the golden hamster with three different methods: single-fiber recordings, two-bottle preference (TBP), and conditioned taste aversion (CTA) tests. On the whole, the results showed that the sense of taste in the hamster differs in many respects from that in humans because, of 26 tested compounds known as sweet to humans, 11 had no taste or tasted differently. The results also supported the notion that activity in S-fibers elicits liking and activity in Q- or H-fibers rejection. Specifically hierarchial cluster analysis of 36 single fibers from the chorda tympani proper nerve separated N-, H-, and S-clusters consisting of 11 sucrose-, 14 NaCl-, and 11 citric-best fibers. Ace-K, cyanosuosan, N-4-cyanophenyl-N'-cyanoguanidineacetate (CCGA), D-tryptophan, N-3,5-dichlorophenyl-N'-(S)--methylbenzylguanidineacetate (DMGA), saccharin, SC-45647, and suosan stimulated only the S-fibers, were significantly preferred in TBP tests, and generalized to sucrose in the CTA tests. Ethylene glycol stimulated the N-fibers in addition to the S-fibers. This explains its generalization to sucrose in CTA. Its toxicity may contribute to its rejection in TBP tests. Sodium cyclamate stimulated a few N- but no S-fibers, which may explain the nondiscriminatory TBP and CTA results. Glycine elicited its largest response in the S-fibers, although it also stimulated other fibers. The resulting mixed taste sensation may explain why it was not preferred in TBP, although it generalized to sucrose in the CTA. Alitame, aspartame, N-4-cyanophenylcarbamoyl-L-aspartyl-(R)--methylbenzylamine (CAM), N-4-cyanophenylcarbamoyl-(R, S)-3-amino-3-(3,4-methylenedioxyphenyl) propionic acid (CAMPA), N-(S)-2-methylhexanoyl-L-glutamyl-5-amino-2-pyridinecarbonitrile (MAGAP), N-1-naphthoyl-L-glutamyl-5-amino-2-pyridinecarbonitrile (NAGAP), NHDHC, superaspartame, and thaumatin were among the compounds considered sweet by humans that gave no response, were not discriminated in the TBP test, and gave no generalization in the CTA tests.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

The hamster was for 40 years a recurrent animal model in taste research (Carpenter 1956; Cummings et al. 1996; Fishman 1957; Formaker and Frank 1996; Frank 1972, 1973; Frank et al. 1987, 1988; Frank and Nowlis 1989; Hanamori et al. 1988; Hellekant and Roberts 1983; Hyman and Frank 1981; Noma et al. 1974; Ogawa et al. 1968, 1969; Rehnberg et al. 1990).

However, during these 40 years it also became increasingly evident that there are major differences in the sense of taste among mammalian species. Taste differences were especially documented in primate sweet taste where many differences can be related to the phylogeny (Glaser et al. 1995; Hellekant and Danilova 1996; Hellekant et al. 1997a; Nofre et al. 1996). However, between nonprimate species profound differences also exist in taste (cf. Jakinovich and Sugarman 1988 and Kare 1961), although these differences were not studied as systematically as in primates. Therefore one purpose here is to improve our understanding of differences in taste across species by testing in hamsters the taste of a number of compounds considered sweet by humans.

Another purpose of this study relates to data in primate research demonstrating that two hedonically opposed groups of taste fibers can be distinguished (Danilova et al. 1998; Hellekant et al. 1997a,b). One group, the S-fibers, responds most and best to compounds humans consider sweet, although by no means do all compounds sweet to humans evoke a response in these S-fibers. Activity in the other group, the Q-fibers, elicits a rejection to the compound that caused the activity. The final hedonical response is the result of these two opposing activities in the same way as sympathetic and parasympathetic activity determine the autonomic response. This study determines how applicable this idea is to the hamster by relating the neural with the behavioral responses to a large number of compounds.

	METHODS

Abstract Introduction Methods Results Discussion References

Chemicals

Table 1 lists the solutions used in the electrophysiological and behavioral experiments. Twenty-six of these compounds are sweet to humans, as indicated in the right column. Figure 1 presents the structure of some of the more unusual sweeteners.

View larger version (28K): [in this window] [in a new window]   FIG. 1. Structures of SC-45647, N-4-cyanophenylguanidineacetate (CGA), cyanosuosan, N-3,5-dichlorophenyl-N'-(S)--methylbenzylguanidineacetate (DMGA), N-4-cyanophenyl-N'-cyanoguanidineacetate (CCGA), TGC, N-4-cyanophenylcarbamoyl-(R,S)-3-amino-3-(3,4-methylenedioxyphenyl) propionic acid (CAMPA), N-(S)-2-methylhexanoyl-L-glutamyl-5-amino-2-pyridinecarbonitrile (MAGAP), N-1-naphthoyl-L-glutamyl-5-amino-2-pyridinecarbonitrile (NAGAP), N-4-cyanophenylcarbamoyl-L-aspartyl-(R)--methylbenzylamine (CAM), and N-4-azidophenyl-N'-diphenylmethylguanidineacetate (ADGA).

The compounds were dissolved in distilled water for the two-bottle preference (TBP) and conditioned taste aversion (CTA) experiments and in artificial hamster saliva for electrophysiological experiments. The composition of the artificial saliva was based on the composition of pilocarpine-stimulated natural hamster saliva (Rehnberg et al. 1992) and was made from 6.6 mM NaHCO₃, 15 mM KCl, 28 mM KHCO₃, pH 8.5. Because 10 mM quinine hydrochloride (QHCl) did not dissolve in the saliva, it was dissolved in distilled water. Precipitation also prevented the use of D- and L-asparagine over 24-h periods in the TBP tests.

Electrophysiology

The recordings were obtained from the chorda tympani proper (CT) nerve of nine golden hamsters, Mesocricetus auratus, of both sexes, 7- to 10-mo old. Anesthesia was initiated with 0.1 ml im Innovar followed by 0.1 ml pentobarbital sodium, 15 mg/ml im and then was maintained with pentobarbital intravenously as needed. The trachea was cannulated, and body temperature, heart, and respiratory rates were continuously monitored.

The right CT was dissected free from the point where it joins the lingual nerve to the tympanic bulla, where it was cut. Single-fiber impulses were recorded with an impulse-amplitude analyzer, which displayed adjustable upper and lower levels. It triggered a pulse when a nerve impulse exceeded the lower but not the upper level. These pulses were processed by an IBM-PC computer. The custom-made software controlled stimulus delivery and stored intervals between pulses together with information on the presented stimulus.

The solutions were delivered over the anterior part of the tongue by a computerized system described earlier (Hellekant and Roberts 1995). Each stimulation lasted for 5 s with 40-s rinsing time between stimulations. Hamster artificial saliva rinsed the tongue between stimulations (Rehnberg et al. 1992). The stimuli and rinses were maintained and delivered at constant temperature (34°C).

The spontaneous activity before each stimulation was deducted from the responses. We define here spontaneous activity as the impulse activity preceding the stimulation during rinsing of the tongue with artificial saliva for 5 s. A cluster was considered to be responsive to a stimulus if the nerve impulse rate during the first 5 s of stimulation was larger than 2 times SD of the spontaneous activity of the cluster. Cluster and multidimensional (MDS) analyses were performed on the responses of 36 fibers to 31 stimuli (SYSTAT for Macintosh, Version 5.2). The hierarchical cluster analysis measured the intercluster similarity with the use of the Pearson correlation coefficient, and the cluster analysis proceeded according to the average linkage method.

To facilitate comparisons with data from other species and studies, the responses of CT fibers to the four basic stimuli (NaCl, citric acid, QHCl, and sucrose) were used to categorize each fiber by its best stimulus (Frank 1973) and to calculate the breadth of responsiveness (H) for each fiber. H was calculated according to the formula H = Kp_i log p_i, where K is a scaling constant (1.6) and p_i is the proportional response each of the four basic stimuli (Smith and Travers 1979).

TBP experiments

A total of 28 adult hamsters of both sexes, housed individually, served as subjects. Each stimulus was tested with seven animals. Their intake of water and a stimuli was recorded during four consecutive 24-h periods with graduated cylinders switched at each measurement. In each hamster the preference ratio was calculated for each 24-h period as the amount of a tastant consumed divided by the total amount of liquid consumed from both cylinders. The mean daily preference ratios were then calculated for each hamster. Thus equal consumption of both liquids yields a preference ratio of 0.5; complete preference yields a preference ratio of 1.

Results were first assessed with the use of analysis of variance. Then a two-tailed t-test was used to compare mean preference ratios to 0.5, and, for all cases in the following text where significance is indicated, P values are <0.01.

CTA experiments

Twelve conditioned and eight control adult hamsters of both sexes served as subjects. Their consumption was measured as number of licks from 16 different bottles mounted in a carousel. Each lick broke an infrared beam positioned between the animal and the bottle in use and triggered one count by the computer. The first lick started a timer that presented the bottle for 30 s.

After a session in which the hamsters were trained to drink water through the apparatus, the animals were water deprived for 24 h and then offered 0.2 M sucrose solution (conditioning stimulus). After consumption of sucrose, 1.5 ml of 0.3 M LiCl (unconditioning stimulus) was injected intraperitoneal within 5-15 min. The control animals underwent the same routine as conditioned animals but without LiCl injections. After a 1-day recovery period we tested if the animals were successfully conditioned. If they drank as much sucrose as before first conditioning, we again injected them with LiCl. Two to four injections were necessary to produce conditioning.

The CTA test was carried without water deprivation. Each animal was tested with three cycles that each included distilled water, sucrose, and sweeteners presented at random.

The level of drinking differed between animals. To use the results from different animals, the consumption of sweeteners had to be normalized to a "standard." For each animal we considered the average number of licks of water during three cycles as the standard and assigned it the value 100%. Then for each hamster, consumption of all stimuli was expressed in percentage of this standard. The data were then averaged for the control and conditioned animals.

With the use of a t-test for independent samples we analyzed the difference in consumption by control and conditioned animals. A P < 0.01 value was considered a significant difference in the statistical analysis.

	RESULTS

Abstract Introduction Methods Results Discussion References

Single-fiber responses in hamster chorda tympani

Figure 2 shows an example of the nerve impulses recorded from single CT taste fiber HA94M25E classified as an S-fiber by the hierarchial analysis below. We present here only a part of the recordings from this unit. However, it is evident that sucrose, suosan, N-4-cyanophenyl-N'-cyanoguanidineacetate (CCGA), SC-45647, N-4-cyanophenylguanidineacetate (CGA), cyanosuosan, N-3,5-dichlorophenyl-N'-(S)-methylbenzylguanidineacetate (DMGA), and to some extent NaCl gave a response, but citric acid, QHCl, alitame, superaspartame, N-(S)-2-methylhexanoyl-L-glutamyl-5amino-2-pyridinecarbonitrile (MAGAP), N-1-naphthoylL-glutamyl-5-amino-2-pyridinecarbonitrile (NAGAP), N-4-cyanophenylcarbamoyl-L-aspartyl-(R)--methylbenzylamine (CAM), and sodium cyclamate did not.

View larger version (39K): [in this window] [in a new window]   FIG. 2. Recordings from the chorda tympani (CT) S-cluster fiber HA94M25E. Onset and offset of stimuli are shown as changes in bar code.

Figure 3 shows an overview of the responses to all compounds in all 36 CT fibers. The legend included in Fig. 3 relates the area of the circles with the nerve responses over the first 5 s of stimulation. Open circles indicate a suppression of the nerve activity during a stimulation. The fibers were ordered along the y-axis in groups, starting with the fibers responding best to NaCl, followed by the fibers mainly stimulated by citric acid and QHCl and finally a group of sucrose-best fibers.

View larger version (81K): [in this window] [in a new window]   FIG. 3. An overview of the response profiles of 36 single CT fibers with the use of topographical method. The area of the circles represents impulse activity over the first 5 s of stimulation. Open circles, inhibition; absence of a mark shows that data are missing. The stimuli were arranged along the x-axis in order of salty, sour, bitter, and sweet. The fibers were arranged along the y-axis in groups: NaCl-, acid-, and sucrose-best fibers.

The breadth of tuning (H) for the NaCl-best fibers was 0.23 (SE = 0.05), for citric acid-best 0.66 (SE = 0.04), and for sucrose-best 0.41 (SE = 0.07). This shows that the NaCl-best fibers are more narrowly tuned than the other types.

Hierarchical cluster analysis

The responses of the 36 CT fibers in Fig. 3 were subjected to a hierarchical cluster analysis. Our analysis distinguished three major clusters consisting of 11 S-fibers, 14 N-fibers, and 11 H-fibers. The result is represented as a dendrogram in Fig. 4 with the identity number of the fiber and response category on the basis of its best response to the four basic solutions listed on the left side.

View larger version (19K): [in this window] [in a new window]   FIG. 4. Result of hierarchical cluster analysis of the response profiles for 36 CT fibers. Intercluster similarity was measured with the Pearson correlation coefficient, and the cluster analysis proceeded according to the average linkage method. Number and response category of the fiber on the basis of its response to the 4 basic solutions are listed on the left. N, H, and S stand for NaCl-, citric acid-, sucrose-best fibers.

Average response profiles

Figure 5 shows the average response profiles of these three clusters. The stimuli were listed along the x-axis, and the average impulse activity measured over 5 s was plotted along the y-axis. The error bars illustrate the SE of these averages. We will present each cluster.

View larger version (30K): [in this window] [in a new window]   FIG. 5. Average response profiles of N-cluster (A), H-cluster (B), and S-cluster (C) of hamster CT fibers. Error bars are SE. Hatched columns, salts; open columns, acids; shaded columns, bitter compounds; black columns, sweeteners. Numbers within parentheses show number of fibers tested with each compound.

N-CLUSTER. This cluster included 14 units and was the largest one. Figure 5A shows that the N-cluster was characterized by strong responses to NaCl but also to ethylene glycol. These fibers did not respond to bitter compounds. Among the compounds that taste sweet to humans, only Na-cyclamate, ethylene glycol, and glycine elicited a strong responses in these fibers; responses to other stimuli were not significant. The average spontaneous activity of the 14 N-cluster fibers was 9.39 (SE = 1.92) imp/5 s.

H-CLUSTER. Figure 5B shows the average response profile of 11 units. Citric acid elicited the best response, and QHCl and L-asparagine gave significant responses. NaCl also stimulated this cluster, although responses were not strong. Most compounds sweet to humans did not stimulate the H-cluster fibers, with exception of ethylene glycol, glycine, and D-asparagine. The average spontaneous of the H-cluster was 4.02 (SE = 0.77) imp/5 s.

S-CLUSTER. Figure 5C shows massive responses to some but not all of the compounds sweet to humans. Thus all compounds from dulcin to glycine elicited a significant response according to our criteria for a response. The remaining sweeteners from NHDHC to MAGAP did not elicit any activity; responses to cyclamate and N-4-cyanophenylcarbamoyl - (R,S) - 3 - amino - 3 - (3,4 - methylenedioxyphenyl) propionic acid (CAMPA) were insignificant. The latter applies also to the activity recorded during stimulation with NaCl, citric acid, and L-asparagine. For all S-fibers the average spontaneous activity before stimulation was 14.2 (SE = 4.2) imp/5 s.

Multidimensional scaling

On the basis of a correlation matrix of the stimuli and to present the relationships between stimuli used, we performed multidimensional scaling. The spatial representation of the similarities among 30 stimuli is shown in Fig. 6. The stress value is 0.072. 

View larger version (24K): [in this window] [in a new window]   FIG. 6. Distribution of 30 stimuli in a 3-D space resulting from multidimensional scaling. Distribution was calculated with Pearson correlation coefficient between stimuli across 36 CT fibers. Kruskal stress value is 0.072.

Sucrose, suosan, cyanosuosan, CAMPA, CGA, TGC, dulcin, acesulfame-K, saccharin, SC-45647, CCGA, DMGA, D-tryptophan, and D-asparagine formed a tight group separate from the other stimuli. Another group included NaCl together with cyclamate and ethylene glycol. Citric acid, L-asparagine, and QHCl were segregated from sweeteners and sodium salts by dimension 2. The closeness between QHCl and citric acid is evident. The nine stimuli that elicited no response are scattered between the groups; eight of these compounds are sweet to humans.

Results of TBP tests

Figure 7 presents the results with TBP in hamsters with the compounds arranged in order of increasing intake. Except for L-tryptophan, all compounds used in the TBP tests taste sweet to humans. Some compounds were presented in two or three concentrations (see Table 1).

View larger version (67K): [in this window] [in a new window]   FIG. 7. Results of the two-bottle preference (TBP) test. Data were averaged for 7-9 hamsters individually tested during 4 consecutive days. Sweeteners eliciting responses in S-cluster fibers are in bold type. Error bars are SE. Line shows preference ratio level 0.5. *, significant difference from preference ratio 0.5.

On the basis of the intake all stimuli can be divided into three groups. One group was significantly rejected. It comprised ethylene glycol and L-tryptophan. The second group was neither preferred nor rejected as indicated by the statistical test showing that the intake of the two bottles did not differ from the 0.5 level. This suggests that hamsters do not discriminate between water on one side and MAGAP, N-4-azidophenyl-N'-diphenylmethylguanidineacetate (ADGA), superaspartame, cyclamate, aspartame, alitame, glycine, CAMPA, NAGAP, thaumatin, NHDHC, and CAM on the other.

To evaluate the possible influence of concentration, these stimuli were also tested at concentrations higher than in the electrophysiological experiments. Only one compound gave different results; 10× higher concentration of CAMPA was preferred over water.

The third group was significantly preferred and included saccharin, D-tryptophan, CGA, SC-45647, DMGA, CCGA, acesulfame-K, dulcin, sucrose, TGC, suosan, and cyanosuosan. The highlighted compounds are sweeteners that elicited a response in S-fibers. As can be seen, only two compounds that stimulate the S-fibers were shunned in the TBP tests.

Results of CTA tests

Figure 8 shows the extent of generalization between intake of sucrose and one of the compounds listed along the x-axis according to the procedure described under METHODS. The circles (control hamsters) and squares (conditioned hamsters) illustrate the number of licks for each compound expressed in percentage of number of licks when offered plain water. In the control group the means varied from 82% (for suosan) to 134% (for glycine).

View larger version (22K): [in this window] [in a new window]   FIG. 8. Consumption of sweeteners by conditioned and control groups of hamsters. Number of licks of each sweetener was normalized to consumption of water for each animal and expressed in percentages. Data were averaged for 8 control and 9-11 conditioned animals. Sweeteners eliciting responses in S-cluster fibers are in bold type. Error bars are SE. Significant difference between 2 groups: * P < 0.05, ** P < 0.01.

In Fig. 8 the stimuli were arranged from the most rejected to the least rejected compound by the conditioned hamsters. As expected the consumption of sucrose was most suppressed (13.9% of water intake) followed by all the compounds that elicited a response in the S-fibers (P < 0.01). As in previous figures these compounds are highlighted in Fig. 8.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

The aim of this study was to investigate how a number of compounds, known as sweet to humans, taste to the hamster. We used three different approaches in our attempts to elucidate this: single-fiber recordings from the CT nerve, TBP and CTA tests.

It seems that together these three techniques can provide us with information how a tastant tastes to an animal species. The electrophysiological method provides information as to whether a compound has taste or not. If a compound does not elicit any impulse activities in the taste nerve, it has no taste. If single-fiber responses are obtained, comparisons of these responses with those of other compounds reveal how similar their tastes are.

The TBP method shows how much the hamsters prefer the compounds over water but gives no information about the taste quality of the stimuli. On the other hand, our assumption is that compounds with a sucroselike taste are liked by hamsters.

The CTA technique here showed to what extent the hamsters generalized from their aversion to sucrose to the other stimuli. From this we conclude that the amount of suppression mirrors the taste similarities between the compound in question and sucrose.

Table 2 summarizes the results of all experiments and draws conclusions about the taste of each compound in the hamsters.

Compounds that did not stimulate CT fibers

The electrophysiology identified compounds that did not stimulate any of the CT fibers. These were L-tryptophan and eight sweeteners: alitame, aspartame, CAM, MAGAP, NAGAP, NHDHC, superaspartame, and thaumatin.

These sweeteners were neither preferred nor rejected over water in TBP experiments, and their consumption was not suppressed in CTA experiments. Thus we conclude that alitame, aspartame, CAM, MAGAP, NAGAP, NHDHC, superaspartame, and thaumatin in concentrations used here do not taste sweet for hamsters.

It is unlikely that these compounds might stimulate S-fibers in glossopharyngeal or superior laryngeal nerves. However, the population of sucrose-best fibers in these two nerves is small, and sucrose is a rather ineffective stimulus for the glossopharyngeal or superior laryngeal nerves (Dickman and Smith 1988; Hanamori et al. 1988). Hamster CT neurons provide the main information about sweeteners' and sodium salts' tastes, whereas NG neurons play a major role in description of bitter and acid stimuli (Frank and Nowlis 1989). Thus we can conclude that the results of the three methods are congruent and that these compounds are tasteless to hamsters.

Figure 7 shows that L-tryptophan was avoided in the TBP experiments, indicating an aversive taste sensation. It is possible that L-tryptophan stimulates Q-fibers in the glossopharyngeal nerve. Results of previous CTA experiments support this conclusion because hamsters show reciprocal generalization between 50 mM L-tryptophan and 1 mM QHCl (Yamamoto et al. 1988). Thus it is likely that L-tryptophan has to the hamster a slight quinine/acidlike taste, the same as it has to humans (Haefeli and Glaser 1990; Schiffman 1976).

Compounds that stimulated any CT fibers except S-cluster fibers

The rest of the stimuli elicited significant responses in single CT fibers. The electrophysiology distinguished stimuli that did not elicit responses in S-fibers from those which did. As mentioned we distinguished three clusters. This corroborates previous studies (Frank 1973, 1988), which showed similar clusters.

The compounds that did not stimulate the S-fibers included cyclamate and L-asparagine. Cyclamate, which was in form of sodium salt, stimulated the N-fibers, was neither preferred nor rejected in TBP test, and was not avoided in CTA experiments. The conclusion is supported by the finding that 25 mM cyclamate generalizes to NaCl in CTA (Nowlis et al. 1980). However it was also shown that 25 mM cyclamate generalizes to sucrose in CTA (Nowlis et al. 1980). When we tested 25 and 50 mM cyclamate in TBP tests, the animals did not prefer it. This indicates that, even if S-fibers would respond to higher concentrations of cyclamate, the responses of N-fibers of CT would determine the reaction of the whole organism. Thus hamsters may taste 10 mM cyclamate as similar to NaCl.

L-Asparagine stimulated only the H-fibers, and its consumption was not suppressed in CTA. We suggest that, to hamsters, L-asparagine tastes similar to acids.

Compounds that stimulated more than S-cluster fibers

The next group of compounds included sweeteners that stimulated more than one cluster: D-asparagine, glycine, and ethylene glycol. We suggest that different clusters of fibers provide information about different taste qualities.

D-Asparagine elicited responses in the S- and H-fibers. Unfortunately TBP data are missing, but the hamsters generalized from sucrose to D-asparagine, which indicates that it exerts a mixed sucrose- and acidlike taste.

Glycine stimulated fibers in every cluster. Thus it elicits a mixed taste. The suppression of its consumption in CTA suggests that its taste has a sucroselike component. This conclusion is supported by earlier studies showing reciprocal generalization between 0.1 M sucrose and 0.6 M glycine (Nowlis et al. 1980; Yamamoto et al. 1988). On the other hand the complexity of its taste is demonstrated by its generalization to HCl (Nowlis et al. 1980). Because neither preference nor rejection was recorded in the TBP experiments here, it is likely that the effects on S- and H-fibers balanced each other.

Ethylene glycol is a highly toxic compound. Consequently postingestive signals may explain the discrepancy between the generalization to sucrose in our CTA tests and the rejection in the TBP experiments. On the other hand, ethylene glycol stimulated both the N- and S-fibers (Fig. 3). It is possible that the strong response in N-fibers was the cause for the rejection because hamsters reject NaCl in TBP tests (Hettinger and Frank 1990). Ethylene glycol also generalized to sucrose in our CTA experiments. This can be explained by its substantial response in the S-fibers. Thus it has both a sucrose- and NaCl-like taste to hamsters.

Compounds that stimulated only S-cluster fibers

The last group of stimuli included compounds that stimulated only S-cluster fibers. Hamsters liked all these compounds in TBP experiments. The CTA experiments separated these sweeteners into two groups. The consumption of ace-K, cyanosuosan, CCGA, D-tryptophan, DMGA, saccharin, SC-45647, and suosan was suppressed, indicating that the hamsters generalized from sucrose to these compounds. Thus these compounds taste like sucrose.

The consumption of CGA, dulcin, and TGC was not suppressed in CTA. Although both dulcin and TGC evoked significant responses in the S-fibers, these responses were smaller than those of the other sweeteners (Fig. 5C). It is possible that we used concentrations of dulcin (1.6 mM) and TGC (0.16 mM) that were too low. CGA was the only compound that gave contradictory results. Because of several circumstances, one of these being lack of compound, we were unable to repeat the CTA tests for CGA.

CAMPA presents a special case. In CTA and TBP experiments with low concentration (28 µM), the hamsters showed no reaction. However, 10× higher concentration of CAMPA was preferred in the TBP. This is in contrast to all the other compounds tested with more than one concentration in TBP tests (Table 2). Furthermore, CAMPA at 28 µM stimulated four of nine S-cluster fibers, although the average response was not significant. MDS analysis positioned it together with stimuli that taste like sucrose. This indicates that the concentration of CAMPA used for electrophysiology was below the behavioral threshold for hamsters. At higher concentrations it might be sweet to hamsters.

These results further support the hypothesis that activity in S-fibers stimulates intake (Rehnberg et al. 1990) and activity in Q-fibers rejection. The results show that, by combining single-fiber nerve recordings and TBP and CTA methods, it is possible to gain insight on how an unknown compound may taste to an animal species. The results of all compounds can be explained by the assumption that activity in S-fibers will evoke a preference and activity in H-/Q-fibers rejection, whereas the behavioral effects of N-fiber activity are species related.

	FOOTNOTES

  Address for reprint requests: G. Hellekant, Animal Health and Biochemical Sciences, The University of Wisconsin-Madison, 1655 Linden Dr., Madison, WI 53706-1581.

  Received 29 December 1997; accepted in final form 3 June 1998.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• CARPENTER, J. A. Species differences in taste preferences. J. Comp. Physiol. Psychol. 49: 139-144, 1956.[Medline]
• CUMMINGS, T. A., DANIELS, C., KINNAMON, S. C. Sweet taste transduction in hamster: sweeteners and cyclic nucleotides depolarize taste cells by reducing a K⁺ current. J. Neurophysiol. 75: 1256-1263, 1996.[Abstract/Free Full Text]
• DANILOVA, V., HELLEKANT, G., ROBERTS, R., TINTI, J. M., AND NOFRE, C. Behavioral and single chorda tympani taste fiber responses in the common marmoset, Callithrix jacchus jacchus. Ann. NY Acad. Sci. In press.
• DICKMAN, J. D., SMITH, D. V. Response properties of fibers in the hamster superior laryngeal nerve. Brain Res. 450: 25-38, 1988.[Medline]
• DUBOIS, G. E., WALTERS, D. E., SCHIFFMAN, S. S., WARWICK, Z. S., BOTH, B. J., PECORE, S. D., GIBES, K., CARR, B. T., BRANDS, L. M. Concentration-response relantionships of sweeteners. In: Sweeteners: Discovery, Molecular Design, and Chemoreception, edited by D. E. Walters, F. T. Orthoefer, and G. E. DuBois Washington, DC: American Chemical Society, 1991, p. 261-276.
• FISHMAN, I. Y. Single fiber gustatory impulses in rat and hamster. J. Cell. Comp. Physiol. 49: 319-334, 1957.
• FORMAKER, B. K., FRANK, M. E. Responses of the hamster chorda tympani nerve to binary component taste stimuli: evidence for peripheral gustatory mixture interactions. Brain Res. 727: 79-90, 1996.[Medline]
• FRANK, M. Taste responses of single hamster chorda tympani nerve fibers. In: Olfaction and Taste, edited by D. Schneider. Stuttgart, Germany: Wissenschaftliche, 1972, vol. 4, p. 287-293.
• FRANK, M. E. An analysis of hamster afferent taste nerve response functions. J. Gen. Physiol. 61: 588-618, 1973.[Abstract/Free Full Text]
• FRANK, M. E., BIEBER, S. L., SMITH, D. V. The organization of taste sensibilities in hamster chorda tympani nerve fibers. J. Gen. Physiol. 91: 861-896, 1988.[Abstract/Free Full Text]
• FRANK, M. E., NOWLIS, G. H. Learned aversions and taste qualities in hamsters. Chem. Senses 14: 379-394, 1989.[Abstract/Free Full Text]
• FRANK, M. E., PATEL, S. J., AND HETTINGER, T. P. Analysis of tastes in mixtures by hamsters. In: Olfaction and Taste, edited by S. D. Roper and J. Atema. New York: NY Acad. Sci., 1987, vol. 10, p. 293-295.
• GLASER, D., TINTI, G. M., NOFRE, C. Evolution of the sweetness receptor in primates. I. Why does alitame taste sweet in all prosimians and simians, and aspartame only in Old World simians? Chem. Senses 20: 573-584, 1995.[Abstract/Free Full Text]
• HAEFELI, R. J., GLASER, D. Taste responses and thresholds obtained with primary amino acids in humans. Lebensm.-Wiss. Technol. 23: 523-527, 1990.
• HANAMORI, T., MILLER, I. J. JR, SMITH, D. V. Gustatory responsiveness of hamster glosspharyngeal nerve. J. Neurophysiol. 60: 478-498, 1988.[Abstract/Free Full Text]
• HELLEKANT, G., DANILOVA, V. Species differences toward sweeteners. Food Chem. 56: 323-328, 1996.
• HELLEKANT, G., DANILOVA, V., NINOMIYA, Y. The primate sense of taste: Behavioral and single chorda tympani and glossopharyngeal nerve fiber recordings in the rhesus monkey, Macaca mulatta. J. Neurophysiol. 77: 978-993, 1997a.[Abstract/Free Full Text]
• HELLEKANT, G., NINOMIYA, Y., DANILOVA, V. Taste in chimpanzees. II. Single chorda tympani fibers. Physiol. Behav. 61: 829-841, 1997b.[Medline]
• HELLEKANT, G., ROBERTS, T. W. Study of the effect of gymnemic acid on taste in hamster. Chem. Senses 8: 195-202, 1983.[Abstract/Free Full Text]
• HELLEKANT, G., ROBERTS, T. W. Whole nerve and single fiber recordings in non-human primates. In: Experimental Cell Biology of Taste and Olfaction, edited by A. I. Spielman, and J. G. Brand Boca Raton, FL: CRC, 1995, p. 277-290.
• HETTINGER, T. P., FRANK, M. E. Specificity of amiloride inhibition of hamster taste responses. Brain Res. 513: 24-34, 1990.[Medline]
• HYMAN, A. M., FRANK, M. E. Effects of binary taste stimuli on the neural activity of the hamster chorda tympani. J. Gen. Physiol. 76: 125-142, 1981.[Abstract/Free Full Text]
• JAKINOVICH, W. JR, SUGARMAN, D. Sugar taste reception in mammals. Chem. Senses 13: 13-31, 1988.[Abstract/Free Full Text]
• KARE, M. R. Comparative aspects of the sense of taste. In: Physiological and Behavioral Aspects of Taste, edited by M. R. Kare, and B. P. Halpern Chicago, IL: Univ. of Chicago Press, 1961, p. 6-15.
• NOFRE, C. AND TINTI, J. M., inventors; Univ. Claude Bernard Lyon, assignee. Sweetening agents. US Patent 4,645,678. Feb. 24, 1987.
• NOFRE, C. AND TINTI, J. M., inventors, assignee. Sweetening agents derived from L-aspartic or L-glutamic acid. US Patent 5,196,540. Mar. 23, 1993a.
• NOFRE, C. AND TINTI, J. M., inventors, assignee. Sweetening agents derived from L-glutamic acid. US Patent 5,272,272. Dec. 21, 1993b.
• NOFRE, C., TINTI, J. M. Sweetness reception in man: the multipoint attachment theory. Food Chem. 56: 263-274, 1996.
• NOFRE, C., TINTI, J. M., AND CHATZOPOULOS OUAR, F., inventors; Univ. Claude Bernard Lyon, assignee. Glycine and -alanine derivatives as sweetening agents. US Patent 4,877,895. Oct. 31, 1989.
• NOFRE, C., TINTI, J. M., AND CHATZOPOULOS OUAR, F., inventors; Univ. Claude Bernard Lyon, assignee. Sweetening agents. US patent 4,921,939. May 1, 1990.
• NOFRE, C., TINTI, J. M., GLASER, D. Evolution of the sweetness receptor in primates. II. Gustatory responses of non-human primates to nine compounds known to be sweet to man. Chem. Senses 21: 747-762, 1996.[Abstract/Free Full Text]
• NOMA, A., SATO, M., TSUZUKI, Y. Taste effectiveness of anomers of sugars and glycosides as revealed from hamster taste responses. Comp. Biochem. Physiol. A Physiol. 48: 249-262, 1974.
• NOWLIS, G. H., FRANK, M. E., PFAFFMANN, C. Specificity of acquired aversions to taste qualities in hamsters and rats. J. Comp. Physiol. Psychol. 94: 932-942, 1980.[Medline]
• OGAWA, H., SATO, M., YAMASHITA, S. Multiple sensitivity of chorda tympani fibres of the rat and hamster to gustatory and thermal stimuli. J. Physiol. (Lond.) 199: 223-240, 1968.[Abstract/Free Full Text]
• OGAWA, H., SATO, M., YAMASHITA, S. Gustatory impulse discharged in response to saccharin in rats and hamsters. J. Physiol. (Lond.) 204: 311-329, 1969.[Abstract/Free Full Text]
• PAUL, T. Der Süssungsgrad der Süsstoffe. Z. Untersuch. Nahrungs Genussmittel. 35: 137-152, 1922.
• PETERSEN, S. AND MÜLLER, E. Über eine neue Gruppe von Süsstoffen. Chemische Berichte. 31-38: 1948.
• REHNBERG, B. G., HETTINGER, T. P., FRANK, M. E. The role of sucrose-sensitive neurons in ingestion of sweet stimuli by hamsters. Physiol. Behav. 48: 459-466, 1990.[Medline]
• REHNBERG, B. G., HETTINGER, T. P., FRANK, M. E. Salivary ions and neural taste responses in the hamster. Chem. Senses 17: 179-190, 1992.[Abstract/Free Full Text]
• SCHIFFMAN, S. S. Taste of dipeptides. Physiol. Behav. 17: 523-535, 1976.[Medline]
• SHALLENBERGER, R. S. Taste Chemistry. London: Blackie, 1993.
• SMITH, D. V., TRAVERS, J. B. A metric for the breadth of turning of gustatory neurons. Chem. Sense Flavor. 4: 215-229, 1979.
• TINTI, J. M., NOFRE, C., DUROZARD, D. Studies on sweeteners requiring the simultaneous presence of both NO₂/CN and COO^- groups. Naturwissenschaften 68: 148, 1981.
• TINTI, J. M., NOFRE, C., PEYTAVI, A. M. Interaction of suosan with the sweet taste receptor. Z. Lebensm. Unters. Forsch. 175: 266-268, 1982.
• YAMAMOTO, T., MATSUO, R., KIYOMITSU, Y., KITAMURA, R. Taste effects of "umami" substances in hamsters as studied by electrophysiological and conditioned taste aversion techniques. Brain Res. 451: 147-162, 1988.[Medline]

0022-3077/98 $5.00 Copyright ©1998 The American Physiological Society