Recognition by Rats of Binary Taste Solutions and Their Components.

This behavioral study investigated how rats conditioned to binary mixtures of preferred and aversive taste stimuli, respectively, responded to the individual components in a conditioned taste aversion (CTA) paradigm. The preference of stimuli was determined based on the initial results of 2 bottle preference test. The preferred stimuli included 5mM sodium saccharin (Sacc), 0.03M NaCl (Na), 0.1M Na, 5mM Sacc + 0.03M Na, and 5mM Sacc + 0.2mM quinine hydrochloride (Q), whereas the aversive stimuli tested were 1.0M Na, 0.2mM Q, 0.3mM Q, 5mM Sacc + 1.0M Na, and 5mM Sacc + 0.3mM Q. In CTA tests where LiCl was the unconditioned stimulus, the number of licks to the preferred binary mixtures and to all tested preferred components were significantly less than in control rats. No significant difference resulted between the number of licks to the aversive binary mixtures or to all tested aversive components. However, when rats pre-exposed to the aversive components contained of the aversive binary mixtures were conditioned to these mixtures, the number of licks to all the tested stimuli was significantly less than in controls. Rats conditioned to components of the aversive binary mixtures generalized to the binary mixtures containing those components. These results suggest that rats recognize and remember preferred and aversive taste mixtures as well as the preferred and aversive components of the binary mixtures, and that pre-exposure before CTA is an available method to study the recognition of aversive taste stimuli.


Introduction
Most foods and beverages do not have pure, but have complex tastes. Previous studies determined whether animals, including humans, are capable of recognizing individual components of taste mixtures (McBurney and Gent 1979;Erickson and Covey 1980;Erickson 1982;McCutcheon and Brown 1983;Smith and Theodore 1984;Frank et al. 2003). The results of some of the previous psychophysiological studies, however, were contradictory in reporting that components of tested mixtures were identifiable (McBurney and Gent 1979) or had unique tastes that were not predictable from the taste of the binary mixture (Erickson and Covey 1980;Erickson 1982). In previous conditioned taste aversion (CTA) studies, investigators determined whether rodents could recognize the components of binary taste mixtures (Smith and Theodore 1984;Frank et al. 2003). The results indicated that both Long Evans hooded rats (Smith and Theodore 1984) and golden hamsters (Frank et al. 2003) recognized the components of the tested binary taste mixtures via the CTA paradigm. However, in their experiments, it was unclear whether the rodents conditioned to the aversive binary taste stimuli recognized the aversive CS and their components. For example, rats did not readily ingest quinine even when water deprived (Smith and Theodore 1984), and the CTA to quinine was weakly expressed when quinine was mixed with NaCl in hamsters (Frank et al. 2003). These facts complicated an analysis of the pattern of generalization when animals were negatively conditioned to an aversive taste stimulus. It is unclear, therefore, whether animals conditioned to an aversive binary taste stimulus (the CS) can recognize the individual components. To overcome the suppression of aversive taste stimuli intake, the "learned safety" paradigm brought about by "pre-exposure" to the taste stimulus was employed. Here, familiar and safe taste stimuli are consumed more than a novel one (Siegel 1974;Domjan 1977;Lin et al. 2012). Rats pre-exposed to conditioned taste stimulus (the CS) showed a significant decrease in the degree of avoidance-response as compared with non-pre-exposed controls (McLaurin et al. 1963;Farley et al.1964;Kalat and Rozin 1973). Previous studies (Lubow and Moore 1959;Lubow 1973) also reported a "latent inhibition" defined as a decrement in learning performance resulting from the non-reinforced pre-exposure of the to-be-conditioned stimulus. These reports demonstrated that a preexposed taste stimulus is drunk more than a non-pre-exposed one. To allow the rats to drink more aversive solutions, we conducted a "pre-exposure" experiment.
In this study, we investigated whether rats recognized the components of binary mixtures (the CS) when they were conditioned to aversive as well as preferred mixtures. Two bottle preference tests were conducted to determine the preference for binary taste solutions and their components. We also investigated how rats responded not only to the preferred components, but also to the aversive stimuli when they were conditioned to preferred and aversive binary taste solutions. And we also investigated whether rats conditioned to a pure taste stimulus generalized to their mixtures reciprocally.

Material and methods
Experiment 1: 48-h 2-bottle preference tests: taste stimuli versus DW Twenty adult male Wistar/ST rats (180-210 g, aged 8 weeks; Slc:Wistar/ST, Chubu Kagaku Shizai, Aichi, Japan) were tested. Rats were housed in individual cages each containing 2 drinking bottles and could access food (MF, Oriental Yeast Co. Ltd., Tokyo, Japan) freely. The colony was maintained on a 12/12-h light/dark cycle. Taste stimuli included: 0.3-10.0 mM sodium saccharin (Sacc), 0.03-1.0 M sodium chloride (Na), 0.1-0.3 mM quinine hydrochloride (Q), the binary mixture of 5 mM Sacc and either 0.03-1.0 M Na (Sacc + Na), and the binary mixture of 5 mM Sacc and either 0.1-0.3 mM Q (Sacc + Q). Distilled water (DW) at 25 °C was the solvent for the taste stimuli. Two test bottles, one filled with a taste stimulus and the other with DW, were placed in each cage. The rats were allowed free access to the 2 bottles for 48 h. The position of the bottles was switched after 24 h for each 48 h test. All bottles were presented randomly. The volume of intake in the 2 bottles was measured both before and after each 48 h test, and a preference score was calculated according to the following formula: Preference score (%) = 100 × (volume of the taste stimulus ingested / total ingested volume from the 2 bottles). The volume of intake from these 2 bottles was analyzed using the 2-tailed t-test with statistical significance set at P < 0.05.
Experiment 2: lick test: recognition of the components in the preferred binary taste mixtures by the CTA test Forty adult male Wistar/ST rats (180-210 g, aged 8 weeks) were tested. Rats were housed in individual cages with free access to food. The colony was maintained on a 12/12-h light/dark cycle. The rats were evenly distributed among 2 groups according to the specific conditioned stimulus (CS) (n = 20) tested. The CS included the binary mixtures 5 mM Sacc + 0.03 M Na (Na group) and 5 mM Sacc + 0.2 mM Q (Q group). Each CS and test stimulus was dissolved in 25 °C DW. Each experimental group was divided into conditioned and control groups (n = 10 for each group). The rats were trained for 5 days in plastic cages. Each cage had a round opening with sensors of a lickometer through which the rat could lick a taste stimulus contained within a plastic bottle. The number of licks were detected by a photo sensor (see Sako et al. 2004). The rats of each group were placed in plastic cages and given free access to DW for 10 min per day. On the sixth day, all groups were presented the CS for 10 min. Subsequently, the conditioned groups were given an intraperitoneal (IP) injection of 0.15 M lithium chloride (LiCl; 2% of body weight) as an unconditioned stimulus. The control groups were given an IP injection of physiological saline (2% of body weight) instead of LiCl. The next day was a recovery day in which the rats were presented only DW. On the eighth day, the number of licks that occurred per 10 s for each test stimulus was counted subsequent to the first lick of each stimulus via the lickometer. In the Na group, DW, 5 mM Sacc, 0.03 M Na, 0.1 M Na, 1.0 M Na, and 5 mM Sacc + 0.03 M Na were tested. For the Q group, DW, 5 mM Sacc, 0.2 mM Q, 0.3 mM Q, and 5 mM Sacc + 0.2 mM Q were tested. The interval between each test solution was 30 s. DW was offered between each test stimulus to avoid the effect of the previous stimuli. All bottles were presented randomly. The number of licks was also quantified by expressing it as a mean licking suppression rate (MLS) based on the ratio of the number of licks to each taste stimulus shown by the conditioned group to that of the control group, as shown by the following formula: MLS (%) = {1 − (licks per 10 s in conditioned group)/(licks per 10 s in control group)} × 100. The number of licks per 10 s in the conditioned and control rats that occurred to each stimulus presentation was analyzed using the 2-tailed t-test with the level of statistical significance set at P < 0.05.

Experiment 3-1: lick test: recognition of the components in the aversive binary taste mixtures by the CTA test in naive rats
Forty adult male Wistar/ST rats (180-210 g, aged 8 weeks) were tested. The rats were evenly distributed among 2 groups according to the specific CS (n = 20) tested. The CSs included the binary mixture 5 mM Sacc + 1.0 M Na (Na-N group) and 5 mM Sacc + 0.3 mM Q (Q-N group). Each experimental group was divided into conditioned and control groups (n = 10 for each group). By using a similar procedure as in Experiment 2, the number of licks in all groups were counted and analyzed. In the Na-N group, DW, 5 mM Sacc, 0.03 M Na, 1.0 M Na, and 5 mM Sacc + 1.0 M Na were tested. For the Q-N group, DW, 5mM Sacc, 0.2 mM Q, 0.3 mM Q, and 5 mM Sacc + 0.3 mM Q were tested. Experiment 3-2: lick test: recognition of the components in the aversive binary taste mixtures by the CTA test in pre-exposure rats Forty adult male Wistar/ST rats (180-210 g, aged 8 weeks) were tested.
The rats were evenly distributed among 2 groups to pre-exposure to 1.0 M Na (Na-Px group) and 0.3 mM Q (Q-Px group). Rats pre-exposed to the aversive components, Na-Px and Q-Px, respectively, were allowed free access to DW and to their aversive stimulus on alternate days for 8 days. DW was used in order to correct for the influence of dehydration. By using a similar procedure as in Experiment 2, the number of licks in all groups was counted and analyzed. 1.0 M Na (Na-Px group) and 0.3 mM Q (Q-Px group), respectively, as well as DW were used as taste solutions for the training. The rats were trained for 5 days and were given 1.0 M Na (Na-Px group) or 0.3 mM Q (Q-Px group) for 5 min prior to DW for 5 min per day. On the sixth day, each group was divided into conditioned and control groups (n = 10 for each group). Both groups were presented the binary aversive mixture, 5 mM Sacc + 1.0 M Na (Na-Px group) and 5 mM Sacc + 0.3 mM Q (Q-Px group) as a CS for 10 min. For the Na-Px group, DW, 5 mM Sacc, 0.03 M Na, 1.0 M Na, and 5 mM Sacc + 1.0 M Na were tested, whereas for the Q-Px group, DW, 5 mM Sacc, 0.2 mM Q, 0.3 mM Q, and 5 mM Sacc + 0.3 mM Q were tested.
Experiment 4: lick test: recognition of binary mixtures in rats pre-exposed to aversive pure CS by the CTA test Forty adult male Wistar/ST rats (180-210 g, aged 8 weeks) were tested.
The rats were evenly distributed among 2 groups to pre-exposure to the 1.0 M Na (Na group) and 0.3 mM Q (Q group). Each group, pre-exposed to an aversive CS (1.0 M Na or 0.3 mM Q) was allowed free access to DW and to the CS using a similar procedure as in Experiment 3-2. The rats were trained for 5 days and were given 1.0 M Na (Na group) or 0.3 mM Q (Q group) for 5 min prior to DW for 5 min per day. On the sixth day, the rats were presented 1.0 M Na (Na group) or 0.3 mM Q (Q group) as a CS for 10 min. In the Na group, DW, 5 mM Sacc + 0.03 M Na, 5 mM Sacc + 1.0 M Na, and 1.0 M Na were tested whereas for the Q group, DW, 5 mM Sacc + 0.2 mM Q, 5 mM Sacc + 0.3 mM Q, and 0.3 mM Q were tested.
All experiments were run according to the "Guidlines for the Proper Conduct of Animal Experiments (Science Council of Japan; 2006)" and "The Animal Care Guidelines of Asahi University". All experimental protocols were also approved by "The Animal Care and Ethics Committee of Asahi University" (Approval Nos: 11-010, 11-014, 13-010, 15-004 and 16-014).

Experiment 1: 48-h 2 bottle preference tests: taste stimuli versus DW
The volume of intake to 0.3-10.0 mM Sacc, 0.03 M Na, and 0.1 M Na was significantly greater than that to DW (P < 0.001 and P < 0.01, t-test), whereas the volume of intake to 0.3 M Na, 1.0 M Na, and 0.1-0.3 mM Q was significantly less than that to DW (P < 0.001, t-test) ( Figure 1A-C). The volume of intake to the 4 binary mixtures, 5 mM Sacc + 0.03 M Na, 5 mM Sacc + 0.1 M Na, 5 mM Sacc + 0.1 mM Q, and 5 mM Sacc + 0.2 mM Q was significantly greater than that to DW (P < 0.001, t-test). In contrast, the volume of intake Asterisk marks indicate significant differences between the volume of intake for taste stimuli and parallel DW (**P < 0.01, ***P < 0.001, t-tests).
to the 3 other binary mixtures, 5 mM Sacc + 0.3 M Na, 5 mM Sacc + 1.0 M Na, and 5 mM Sacc + 0.3 mM Q was significantly less than that to DW (P < 0.001, t-test) ( Figure 1D and E).
Experiment 2: lick test: recognition of the components of the preferred binary taste mixtures by the CTA test When 5 mM Sacc + 0.03 M Na was used as the conditioned stimulus (CS), the number of licks to 5 mM Sacc + 0.03 M Na, 5 mM Sacc, 0.03 M Na, and 0.1 M Na in the conditioned group was significantly less than in the control group (P < 0.001, t-test). However, no significant difference was observed between the number of licks to 1.0 M Na and DW in either the conditioned or control groups (Figure 2A).
When 5 mM Sacc + 0.2 mM Q was used as the CS, the number of licks to 5 mM Sacc + 0.2 mM Q and 5 mM Sacc in the conditioned group was significantly less than in the control group (P < 0.001, t-test). However, no significant difference was observed between the number of licks to 0.2 mM Q, 0.3 mM Q, and DW in either the conditioned or control groups ( Figure 2B).

Experiment 3-1: lick test: recognition of the components in aversive binary taste mixtures by the CTA test in naive rats
When 5 mM Sacc + 1.0 M Na was used as the CS, the number of licks to 5 mM Sacc, 0.03 M Na in the conditioned group was significantly less than in controls (P < 0.05, t-test). However, no significant difference was observed between the number of licks to the 5 mM Sacc + 1.0 M Na, 1.0 M Na, and DW in either the conditioned or control groups (Na-N group in Figure 3).
When 5 mM Sacc + 0.3 mM Q was used as the CS, the number of licks to 5 mM Sacc in the conditioned group was significantly less than in controls (P < 0.001, t-test). However, no significant difference was observed between the number of licks to the 5 mM Sacc + 0.3 mM Q, 0.2 mM Q, 0.3 mM Q, and DW in either the conditioned or control groups (Q-N group in Figure 3). Experiment 3-2: lick test: recognition of the components in the aversive binary taste mixtures by the CTA test in pre-exposure rats When 5 mM Sacc + 1.0 M Na was used as the CS, the number of licks to all tested stimuli, except DW, in the conditioned group was significantly less than in the control group (P < 0.001, t-test) (Na-Px group in Figure 3).
When 5 mM Sacc + 0.3 mM Q was used as the CS, the number of licks to all tested stimuli, except DW, in the conditioned group was also significantly less than in the control group (P < 0.001, t-test) (Q-Px group in Figure 3).
Experiment 4: lick test: recognition of binary mixtures in rats pre-exposed to aversive pure CS by the CTA test When 1.0 M Na was used as the CS, the number of licks to all tested stimuli, except DW, in the conditioned group was significantly less than in the control group (P < 0.001, t-test) (Na group in Figure 4). When 0.3 mM Q was used as the CS, the number of licks to all tested stimuli, except DW, in the conditioned group was also significantly less than in the control group (P < 0.001, t-test) (Q group in Figure 4).

Discussion
This CTA study investigated whether rats conditioned to preferred and aversive binary taste mixtures, respectively, recognized the respective components and vice versa.
In experiment 1 prior to the lick tests using the CTA technique, 48-h 2 bottle preference tests were performed to determine the preference of the rats to the test stimuli that included binary taste mixtures as the conditioned stimuli (CS). From these results, 5 mM Sacc + 0.03 M Na and 5 mM Sacc + 0.2 mM Q were chosen as preferred, and 5 mM Sacc + 1.0 M Na and 5 mM Sacc + 0.3 mM Q were selected as aversive CSs. Preferences for the single pure taste stimuli used as test stimuli in the CTA experiments were also determined. Previous studies described the preferences of rats for pure Sacc, Na, and Q. An investigation of the influence of the preference for Sacc by food deprivation showed that in 2 bottle preference tests (48 h), the volume of intake of 0.1% (about 4.9 mM) Sacc was greater than that for deionized water in both food intake and food deprived rats (Strouthes 1971). A study of the preference for Sacc in 2 bottle preference tests (48 h) post adrenalectomy revealed that sham-operated rats preferred 0.008-0.5% (about 0.39-24.4 mM) Sacc rather than water (Silva 1977). Both of these prior reports support our results of the preference of rats for Sacc. Further, the preference of rats for Na and Q was also previously reported (Weiner and Stellar 1951;Pfaffmann 1978;Contreras and Kosten 1983). Preference for the binary mixture of sweet and bitter stimuli in Sprague-Dawley rats that were selectively bred for high versus low Sacc intake (Nachman 1959) was also shown (Dess 2000). This previous study did not use Sacc like our current report, but selected sucrose as the sweetener. Preference for the binary mixture of sucrose + Q became less depending upon the concentration of the Q component tested in both bred rat strains which supports our current results of the preference of rats for mixtures of sweet and bitter stimuli. This previous study and our present results may suggest that the aversive of bitter stimulus is suppressed by pleasantness of sweetener.
In experiment 2, we examined how rats conditioned to preferred taste mixtures responded to the individual taste stimuli, including the components of the CSs, by comparing the number of licks between the conditioned and control rats. When 5 mM Sacc + 0.03 M Na or 5 mM Sacc + 0.2 mM Q was used as the CS, the number of licks to all the preferred components in the conditioned groups was significantly less than in controls. These results suggest that rats conditioned to the preferred CSs recognize the respective preferred components. Previous studies reported that rodents were able to recognize the components of binary and ternary mixtures (Smith and Theodore 1984;Frank et al. 2003;Grobe and Spector 2008). The present results are consistent with these previous studies in that rats recognized the preferred components of the tested binary taste mixtures and clearly avoided the CSs.
In experiment 3-1, the aversive taste mixtures were also presented as CSs. The results indicated that no significant differences occurred in the number of licks to the aversive CSs between the conditioned and control groups. Since the number of licks for both the conditioned and control groups was minimal, this experiment might not have examined the recognition of the aversive components of the taste mixtures as the results were likely depended upon the quantity  of licks to the CSs. A previous study reported that Q was to have been used as a CS, but pilot experiments indicated that rats would not readily ingest Q even when water deprived (Smith and Theodore 1984). Another investigation reported that a CTA to Q was barely learned and weakly expressed when quinine was mixed with NaCl (Frank et al. 2003). It is thus difficult to know whether animals conditioned to aversive taste stimuli are able to recognize both the aversive CS and the aversive components; therefore, in experiment 3-2, we investigated how rats pre-exposed and conditioned to aversive taste mixtures responded to the individual components. In this experiment, the number of licks was compared between the conditioned and control groups which were pre-exposed to the aversive components of the CSs. This procedure was based on the "latent inhibition" and "learned safety" strategies. "Latent inhibition" is defined as the decrement in learning performance resulting from nonreinforced pre-exposure of the CS (Lubow and Moore 1959;Lubow 1973;Lubow 2009). "Learned safety" is a situation where safe stimuli are familiar and are consumed in greater quantity than novel taste stimuli due to attenuation of neophobia (Kalat and Rozin 1973;Siegel 1974;Domjan 1977). In previous experiments, the preference score to Sacc used as the CS in habituated rats to Sacc was significantly greater than that for the non-habituated rats (McLaurin et al. 1963). Further, Sacc habituated rats showed a significant decrease in the degree of the conditioned Sacc avoidance response as compared to non-habituated controls (Farley et al. 1964). It was also previously reported that when Sprague-Dawley rats were presented 0.1 mM Q or 0.5% Sacc for 4 consecutive days, the number of licks increased after the initial test day (Lin et al. 2012). These reports indicate that pre-exposure of taste stimuli allows the preference to these stimuli to increase.
In our results of experiment 3-2, the rats were pre-exposed to the aversive components prior to conditioning. To examine the effect of this pre-exposure, we compared the number of licks to 5 mM Sacc + 1.0 M Na or 5 mM Sacc + 0.3 mM Q which was used as the aversive CS in the pre-exposed control groups to those in the naive control; the results indicated that the number of licks to all tested aversive stimuli in the pre-exposed control groups was significantly greater than for the naive controls (P < 0.001, t-tests). However, no significant differences occurred in the number of licks to the preferred components between the pre-exposed control and naive control groups (P > 0.05, t-tests), which suggests that pre-exposure caused an increase in the number of licks to the aversive stimuli. Furthermore, the pre-exposed conditioned rats avoided all tested preferred and aversive taste stimuli contained in the CSs.
Our present study did not require the plural pairing of the CS and the unconditioned stimulus (US) since a single pairing was sufficient to acquire the CTA. This result indicates that pre-exposure of the CS does not entirely mitigate the CTA as suggested by previous reports (McLaurin et al. 1963;Farley et al. 1964). Further, aversive stimuli may be learned easier than the preferable ones used as CSs in the previous studies.
It is possible that a mixture of taste stimuli could result in an interaction between the stimuli depending upon the specific combination. For example, bitterness suppressed sweetness in the gustatory periphery when bitter and sweet stimuli were applied as a binary mixture (Formaker and Frank 1996;Formaker et al. 1997;Frank et al. 2005;Talavera et al. 2008;Tokita and Boughter 2012). However, in control rats, the number of licks to 5 mM Sacc + 0.2 mM Q was significantly greater than that to 0.2 mM Q (P < 0.001, t-test, see Figure 2B). Furthermore, when rats pre-exposed to 0.3 mM Q were conditioned to 5 mM Sacc + 0.3 mM Q, they generalized to 5 mM Sacc (see Figure 3 Q-Px group). From these results, it is likely that even if such an interaction occurred in the present study, it did not affect the behavioral response reported.
The results of experiments 3-1 and 3-2 suggest that rats have the ability to acquire a CTA to both preferred and aversive CSs even if components of the CSs are preferred or aversive taste stimuli.
We demonstrated that the CTA to binary taste mixtures generalized to their components in experiment 3. Furthermore, in experiment 4, we also conducted a reverse experiment to investigate whether the CTA to a pure taste stimulus generalized to the mixture containing the pure stimulus. This experiment is important to make clear whether the generalizations are reciprocal and whether mixtures are perceptually unique or are merely a composite of the components. The results suggest that the generalizations are reciprocal, and that the taste quality of binary mixtures are likely a composite of the components.
Our study showed that rats acquire a CTA to both preferred and aversive CSs and recognize and remember both the preferred and aversive components of binary mixtures. We also found that the CTA technique used along with the pre-exposure paradigm were useful methods to examine the taste quality of aversive taste stimuli.

Conclusion
The number of licks to the preferred components of 2 preferred binary mixtures in rats negatively conditioned to these mixtures was significantly less than in control rats. When 2 aversive binary mixtures were used as the CSs in naive rats, no significant differences occurred in the number of licks to the aversive components or to the 2 aversive binary mixtures between the conditioned and control groups. After the rats were pre-exposed to the aversive components but prior to conditioning, the number of licks to all the aversive tested stimuli in rats negatively conditioned to the aversive mixtures was significantly less than in control rats. Rats conditioned to components of the aversive binary mixtures generalized to the binary mixtures containing those components. These results suggest that rats are able to acquire a CTA to the preferred and aversive CSs and can recognize and remember both the preferred and aversive components of preferred and aversive binary mixtures.