6.2.1. Preference Tests

We used two-bottle preference tests ranging in time from 1 h to several weeks. Some of these tests were conducted in standard rat housing and others in our rat “hotel,” a modified cage arrangement that allowed for microstructure measurements of ingestive behavior. In the hotel we could study the number and size of ingestive bouts and rates of intake within a bout. In all experiments, animals were individually housed.

The preference tests were conducted in the standard manner with two fluid bottles placed on the home cage. The positions of the bottles were reversed daily to counteract any side preferences. Preference scores were calculated by dividing the consumption in one of the bottles by the total fluid intake.

The tests in the rat “hotel” were conducted in a similar manner. The illustration in Figure 6.1 describes the shoebox plastic cage where the eating and drinking ports are attached to the front and back, respectively. Eight such cages constituted the hotel. The eating port housed a 4 ounce glass jar which held up to 65 g of powdered food. When the rat entered the feeding compartment its head broke a beam between an infrared light emitting diode and a photodetector. Interruptions of the beam were recorded and stored in a microcomputer. In some of the hotels the food jar rested on a load beam which continuously recorded the weight of the jar. The correlation between duration of the beam break and amount of food consumed has averaged about 0.97 over many observations. A stainless steel rack which held two glass water bottles was attached to the back of the cage. Each of the two sipper tubes had contact circuits that counted the individual licks and stored the interlick intervals in the microcomputer. The data were recorded in 6 s intervals, which gave 13,800 data points for each of the three ingestive ports. The day/night cycle was 12/12 with the lights on at 07.00 h. A photoreceptor attached to the top of the cage rack was used to indicate the period when the lights were on. At the end of each daily testing period the data were saved and transferred to an analysis computer for processing. On each of the water bottles and food jars were bar codes that could be read by a scanner attached to an electronic balance, which enabled rapid determination of the amount of food and liquid consumed (Smith and Bigbie, 1997). The data analysis utilized Windows-based customized software that allowed for the determination of number of eating and drinking bouts, their durations and the intervals between them (inter-bout intervals). In addition, we determined the mean number of licks per second (or volume per second) during each drinking bout. In order to attain this measure we divided the number of licks in each bout by the length of the bout. If the rat licked without pause during a bout, this value was approximately 6 s⁻¹. Our overall rate measure reflected the number of pauses that occurred within a bout.

FIGURE 6.1

The “rat hotel” was fabricated from a standard shoebox plastic cage with cutouts in the front for the feeding port and the back for the drinking ports. When the rat entered the food port an infrared beam was broken which allowed for the (more...)

This latter measure has proved to be quite interesting to us in our studies of sucrose intake. It is well known that when the rat is presented with a two-bottle preference test between various concentrations of sucrose and water, an inverted U-shaped curve is obtained for the sweetened solution with the highest sucrose intake at about 0.25 M (Richter and Campbell, 1946; Smith and Wilson, 1988). It is not correct to refer to this peak as the “most preferred” sucrose concentration. Collier and colleagues (Collier and Bolles, 1968) have clearly shown in two-bottle preference tests between concentrations of sucrose ranging from 1.0 M and lower, the rats always drink more of the higher over the lower concentrations.

In short-term preference tests (lasting only a few minutes) or sham-feeding tests across a range of sucrose concentrations, rats always drink more of the higher than the lower concentrations and do not show a peak at 0.25 M (Smith, 2000). In other words, when ingestion is under the control of primary oral sensory input, because postingestive factors have been minimized, the sweeter the better. Our measure of “licks per second during a drinking bout,” correlates highly with other taste tests that eliminate or minimize postingestive factors (Spector and Smith, 1984; Smith, 2000). A comparison of our rate measure with (a) ultra short-term taste tests (to be described in Section 6.2.3), (b) electrophysiological recordings from the greater superficial petrosal (GSP) branch of the facial nerve, and (c) sham-feeding tests can be seen in Figure 6.2. These electrophysiological and behavioral measurements have been equated by fitting them with Beidler’s taste equation (i.e., R = CKRs/1 + CK) where R is the magnitude of response to concentration C, Rs is the maximum response, and K is the association constant (inverse of the concentration which produces a response, half the magnitude of the maximum response) which is considered a measure of the affinity of the receptor for the ligand (Smith, 1988).

FIGURE 6.2

Each of the four measurements described here was taken for a variety of sucrose concentrations. Licking in the Davis Rig (Davis), consumption by a rat in a sham-feeding test (Sham), electrical activity in the GSP nerve, and overall rate of licking (Rate) (more...)

Because of these similarities seen in Figure 6.2, we have concluded that our measure of “licks per second during a sucrose bout” from data collected in the rat hotel is also a measure of taste. Each drinking bout is like a short-term test and by study of the details of licking within a bout one can make inferences regarding gustatory experience. We will use this rate measure later in this chapter to infer that the rats “taste” corn oil in a similar manner since their rate of licking in a bout increases as the concentration of the oil is increased in a manner similar to the findings from sucrose testing.

6.2.2. Conditioned Taste Aversion Tests

We also used conditioned taste aversion (CTA) procedures induced by injections of lithium chloride. As such, we could condition an aversion to one substance and test the effects on preferences for other substances, allowing us to infer similarities and differences in the gustatory perceptions of the rats.

The CTAs referred to in this chapter have all been induced by presenting the conditioned stimulus (CS) solution for 10 min followed immediately by an injection of LiCl. The LiCl concentration was 0.6 M, delivered in doses of 5 μL/g of body weight. Nachman and Ashe (1973) recommended this dose for producing a reliable flavor aversion after only one pairing. Control groups were run by injecting rats with isotonic saline at the same dose level following their ingestion of the CS. In order for the rats to drink on conditioning day, they were subjected to a water deprivation regimen for 5 days before the conditioning day. After an overnight water deprivation on the first day of this schedule the rats received water for 60 min followed by a second overnight deprivation regimen. On the following 4 days they received water for 30 min, 20 min and 2 days of 10 min drinking. Using this procedure ensured that all rats would drink the CS solution on conditioning day prior to the lithium chloride injection. The CS is described later in this chapter for each experiment. The pairing of the CS and the LiCl injection was always followed by 24 h of food and water ingestion and the preference tests were initiated at the end of this period when the rats were no longer water deprived. The postconditioning flavor aversion was described with a preference score (i.e., CS Intake/(CS Intake + Solvent Intake)).

When the CS was a mixture of substances, we tested the substances separately to allow inferences about the salient components of the mixture.

6.2.3. Davis Rig Tests

Davis Rig tests were used to take short-term measurements of licking that is based primarily on orosensory stimulation while minimizing the effects of postingestive factors. The Davis Rig has been described in detail elsewhere (Smith et al., 1992; Smith, 2001). Basically it is a small rat-testing cage with a drinking port at the front end. The rat had access to the drinking port only when a motor-driven shutter is opened. Behind the shutter was a tray that held eight drinking tubes with stainless steel sipper tubes. The tray could be moved so that only one of the sipper tubes was available to the rat at a given time. The positioning of a single tube was controlled by a reversible motor and a heliport located under the tray. A computer program controlled the position of the tray and the opening and closing of the shutter. Contact circuits on each of the tubes allowed for measurement of individual licks. A list of interlick intervals was compiled, allowing for subsequent analysis of drinking bursts and clusters of licks. In addition, the latency of the first lick after the shutter was opened was recorded. This latter measurement proved to be very important in our analysis of sucrose intake. We found that the latency to lick was inversely related to the concentration of the sucrose, leading to the suggestion that the rats could detect the presence of sucrose by olfaction at higher concentrations (Rhinehart-Doty et al., 1994). This latency measurement is used later in this chapter to denote the role of olfaction in the detection of corn oil.

In our experiments with corn oil/sweetener ingestion, we typically ran the Davis Rig with the following parameters:

1. About 1 min after the rat was placed in the chamber the shutter opened. Following the first lick the shutter remained open for 30 s.
2. The shutter was then closed for 30 s while the tray moved to another preselected position.
3. The shutter was then opened and this procedure was followed until access to all eight tubes had been accomplished.
4. On any trial if the rat failed to lick on the tube, the 30 s timer would not start. If no licking occurred for 180 s, the shutter was closed for 30 s and the presentation of the next tube followed.

If a rat licked quickly after the shutter opened, it could complete the session in slightly over 8 min, having a total of only 240 s of licking. This Davis Rig procedure yields a short-term ingestive measure, minimizing the possibility of much postingestive contribution to the intake. The rig allows for the opportunity of automatically measuring the intake of as many as eight different compounds during one short-term test.

6.2.4. Preparation of the Solutions

We chose corn oil as our prototypical fat stimulus and we have most frequently presented it blended with water that was sweetened with sucrose (SU) or a mixture of glucose and saccharin (G + SA). The corn oil (CO) (Mazola Brand) was blended with the sweetened solution by the addition of 5 mL of Tween-80. The sweetened water, corn oil, and Tween-80 were mixed in an industrial blender for 2 min. For example, we mixed 85.6 g of sucrose in 840 mL of deionized water and with the addition of 5 mL of Tween-80; we blended this with 160 mL of corn oil. We called this a 16% corn oil mixture in a 0.25 M sucrose solution. The phase separation of the oil from the water occurs fairly rapidly and our measurements of this have been described elsewhere (Smith et al., 2000). Therefore, in this chapter all of the oil/water mixtures are described as the concentration when they were mixed, acknowledging that the concentration of the oil consumed by the rat diminished over time to about half the original concentration by the end of a 24 h period.

6.2.5. Animal Subjects

All rats used in these experiments were Sprague-Dawley. All were males except in a few cases as noted. In the few cases where females were used, females did not differ from the males in the intake or preference measures. In all cases the rats ranged in body weight between 325 and 550 g, except where noted.

6.2.6. Statistical Analyses

The analyses reported in this chapter were performed by appropriate t-tests and ANOVAs. For the sake of simplicity, F, t, and p values are not reported. Rather, outcomes of statistical tests are reported as significant if they reached the 5% level. Post hoc tests used the Fisher’s least significant difference (LSD) test.

6.3.1. Background

Careful observations made by Sclafani and colleagues (Lucas and Sclafani, 1990) compared the intake and weight gain of rats given a sweet/fat mixture (35% corn oil blended with 8% sucrose solution), a sweet solution (8% sucrose) or a fat mixture (35% corn oil blended with water). They found that this mixture of sucrose and corn oil resulted in enhanced caloric intake and a significant increase in body weight. In order to determine if the increased intake of the sucrose–fat mixture was the result of the sweetness, the group that had received only the corn oil mixture was subsequently tested with a 0.2% sodium saccharin solution rather than water mixed with the corn oil. The intake of this sweetened-fat solution over the fat alone increased immediately. Further evidence that the palatability of the sweet-fat solutions played a major role in the enhanced intake was shown in two-bottle preference tests between the oil alone vs. the oil plus sweetener. It was demonstrated that saccharin–oil was preferred over oil, sucrose–oil was preferred over oil and that sucrose–oil was preferred over saccharin–oil mixtures. Since it is well known that sucrose is preferred over saccharin (Collier and Novell, 1967) it seems quite likely that the increase of intake of these two sweet-oil solutions is the result of taste.

6.3.2. Gustatory Cues Play a Role in the Ingestion of a Corn Oil with Glucose + Saccharin Mixture

In our laboratory we have collected similar data where we blended the corn oil with a mixture of glucose and saccharin (Smith, 2004). Valenstein and colleagues demonstrated that when presented with a glucose and saccharin solution (30 g glucose + 1.25 g saccharin in 1 L of water) rats drank excessive amounts of this mixture (Valenstein et al., 1967). In fact, some rats exceeded their own body weight in intake during a 24 h period. They suggested that this solution could be used as a vehicle for getting rats to ingest substances that they normally may not consume in large quantities. In our laboratory we have also replicated this extremely high intake of the G + SA mixture. In fact, we found that when we presented the glucose and saccharin solutions in separate bottles, rats will alternately drink from both bottles, mixing the glucose and saccharin solutions in appropriate quantities. (Smith et al., 1976, 1980, 1982; Smith and Foster, 1980). Hence, we used the G + SA solution as a vehicle for the corn oil. We divided the experiment into three phases. For the first phase, 30 male rats were divided into three groups that were given 24 h ingestion tests for 56 days. The first group received only water and powdered Purina Chow (PC) throughout the testing period. The second group received food and water for 7 days and then received the G + SA mixture for the remaining 49 days. The third group was treated like the second group for the first 14 days (a week of food and water followed by a week of food, water, and G + SA) and then in their third week of testing, 2% corn oil was blended with the G + SA mixture. In the fourth week the corn oil concentration was doubled to 4%. This doubling procedure was continued until the eighth week when the concentration of the corn oil was 64%.

The results of these manipulations can be seen in Figure 6.3. In the upper left panel it can be seen that the total caloric intake from the food, glucose, and corn oil (only group 3 had corn oil) rose as expected as the concentration of the corn oil was increased. The difference between the corn oil group and the other two groups became statistically significant at week 4 when the concentration of the corn oil had reached 4%. Body weight for the corn oil group began to significantly increase after the third week of testing when they first consumed the 2% corn oil mixture (lower left panel). As can be seen in the upper right panel, caloric intake from the PC was significantly lower for the second group (G + SA) as compared to the food–water control group (W) because of the calories from the glucose in the solution. The caloric intake from PC for the corn oil group dropped significantly as the concentration of the oil increased over the 8-week period. However, the rats were still consuming about 50 cal a day from the PC when the oil concentration was at 64%. This was not enough reduction to compensate for the increase in calories from the solution as can be seen in the bottom right panel. Hence, the total calories went up over the course of these observations as a result of ingesting the sweetened corn oil mixture.

FIGURE 6.3

Fluid and powdered PC were measured daily in these rats over an 8-week period. The first group received nothing but chow and water (W) over the time of testing. Rats in the second group received food and water the first week and then a solution of glucose (more...)

We wanted to take some measurements of the “taste” of the sweet–fat mixture of corn oil with G + SA to determine if this blend was more palatable to the rat than the sweetener alone. In order to get a more detailed measure of this intake, we repeated the procedure for the group that received the incremental increases in corn oil mixed with the G + SA solution. In this replication we collected the data in our rat “hotel” where we could get a measurement of rate of eating and drinking during an ingestive bout. These eight male rats received first week of PC and water, a second week of PC, water, and the G + SA mixture, and in their third week of testing they received the G + SA solution with corn oil added. The concentrations of corn oil were doubled each week until it increased to 64%. Similar to the weight increase reported in the earlier study, the rats in this group had a 35% increase in body weight over the 8-week period.

As can be seen in Figure 6.4, the overall rate of drinking during about increased significantly as the concentration of the corn oil went from 2% to 64%. This measure does not reflect any marked increase in the local lick rate (usually around 6 s⁻¹) but points out that the rat is licking more steadily without an excessive amount of pausing. As mentioned earlier in this chapter, when dealing with sucrose ingestion, we inferred that this was a measure of the taste-related motivational response of the rat.

FIGURE 6.4

Licks per second within a drinking bout are plotted as a function of increasing the corn oil concentration that is blended in the G + SA solution starting in the third week of testing. During the first week of testing the rats received only water and (more...)

6.3.3. Taste, Textural, and Olfactory Factors Influence the Ingestion of a Corn Oil and Sucrose Mixture

We conducted a study somewhat like the earlier study with G + SA where we mixed sucrose with the corn oil. In this study we divided 30 male rats into three groups. One group received no corn oil, a second group received 4% corn oil, and the third group was given 32% corn oil. Over several days sucrose was added to the solutions in a concentration series including 0.25, 0.5, and 1.0 M. The calories from food and fluid can be seen in Figure 6.5. It can be seen from this figure that when no fat was available (left panel) the total caloric adjustment was quite good. When the solution contained 32% corn oil the rats failed to decrease PC intake and ingested significantly more daily calories. In Figure 6.6 the proportion of total daily calories from food is plotted. The analysis of these data yielded a significant difference between the corn oil groups, a significant difference across the sucrose concentrations, and a significant interaction. Post hoc comparisons show that from 0.25 to 1.0 M there was a significant decrease in food calories for both the 0% and 4% corn oil groups. In contrast, the group receiving 32% corn oil showed no decrease in calories from PC as the sucrose concentration increased.

FIGURE 6.5

Mean calories from chow and sucrose solutions are shown for sucrose concentrations of 0, 0.25, 0.50, and 1 M and corn oil concentrations of 0%, 4%, and 32%.

FIGURE 6.6

Proportion of calories from PC are given for each of the sucrose and corn oil combinations indicated in the figure.

When ingesting higher concentrations of sucrose without added fat, rats adjusted their total caloric intake much more efficiently than when given either the corn oil–G + SA mixture or the corn oil–sucrose (Smith, 2000). In fact, in a study where we gave Fisher 344 rats high concentrations of sucrose for a lifetime, they gained very little extra weight. In that longitudinal study, as caloric intake increased with the increase in sucrose concentration, the rats lowered their chow calorie intake appropriately. Rats that were given 1 M sucrose took 50% of their calories from the solution and showed no marked increase in calories over this 28-month span (Smith et al., 1992). As was seen in Sclafani’s data and from our present data, if corn oil is added to the sweetener the rats do not appropriately reduce their chow caloric intake as they increase their calories from the sweet-fat solutions. The result is a significant increase in body weight.

The orosensory factors that influence the ingestion of corn oil involve not only taste, but also texture and olfaction. We conducted several experiments designed to further understand why the mixture of sweet and corn oil results in such large caloric intake. Using a CTA design, we used a mixture of sucrose and corn oil as the CS (Smith et al., 2000). Forty-two male and female rats (we found no differences between the sexes) were subjected to mild water deprivation and on the conditioning day were given a 0.25 M sucrose solution blended with 16% corn oil. Twenty-two were then given the LiCl injection and 20 control animals received the saline injection. Forty-eight hours later all animals were given a 90 min preference test between the sucrose–corn oil mixture and water (SUCO vs. W). It can be seen from Figure 6.7 that the LiCl-injected rats showed a profound aversion to the SU + CO mixture. The following day all rats were given a preference test between sucrose and water (SU vs. W). The preference for sucrose in the lithium-injected rats was significantly reduced, but it was not profound. The third preference test was between the corn oil and water (CO vs. W). Here again, there was a significant aversion to the corn oil. The final test was between corn oil and sucrose (SU vs. CO). As can be seen, the saline injected rats consumed both sucrose and corn oil, but the lithium-injected animals significantly avoided the corn oil. We inferred that the salient feature of the sucrose–corn oil mixture was the corn oil. It could be argued that it is simply easier to get an aversion to corn oil than it is to sucrose. To test for this possibility, we conditioned two more groups of rats to sucrose (N = 17) or to corn oil (N = 20). Rats conditioned with sucrose showed a profound aversion to sucrose and rats conditioned with corn oil demonstrated a profound aversion to corn oil. The magnitude of these two aversion scores was not different, so it does not appear that it is easier to get an aversion to corn oil than an aversion to sucrose.

FIGURE 6.7

Preference scores are plotted comparing LiCl- and NaCl-injected rats for preference tests between corn oil + sucrose vs. water (SUCO vs. W), sucrose vs. water (SU vs. W), corn oil vs. water (CO vs. W), and sucrose vs. corn oil (SU vs. CO). For all four (more...)

We have found that a good measure of a taste aversion is how long it lasts (Spector et al., 1981). We tested the extinction of the aversion to sucrose–corn oil. Twenty-eight male rats were given the 0.25 M sucrose and corn oil mixture for 10 min on conditioning day. All received an injection of LiCl following the drinking period. Forty-eight hours later, a 90 min preference test was initiated and these tests were repeated for the next 9 days. The preference test for the first group was between the sucrose–corn oil mixture and water (SU + CO); for second group the test was between sucrose and water (SU); for the third group the test was between corn oil and water (CO) and the rats in the fourth group received only water (W) during the test. The extinction curves can be seen in Figure 6.8. The rats that received corn oil alone or corn oil with sugar showed little, if any, extinction. The aversion to sucrose was extinguished by the third day of testing. The water group consumed about 25 mL of water during each test. Again we concluded from this more stringent measure of CTA that the corn oil is the salient feature of the sucrose–corn oil mixture.

FIGURE 6.8

The intake of solutions is plotted over the 9 days after an aversion was induced by pairing a solution of sucrose (SU) and corn oil (CO) with an injection of LiCl. When tested with water (W), the mean daily intake was about 25 mL and it did not change (more...)

The next test that we performed was to conduct the CTA protocol in the reverse order from the previous approach. We hypothesized that if corn oil were the salient feature of the sucrose–corn oil mixture, then if we used corn oil as the CS, we would get a good aversion to the mixture of sucrose and corn oil. In contrast, if we used sucrose as the CS, we would get little or no aversion to the sucrose–corn oil mixture. Twenty-four rats were divided into two groups with one group receiving 0.25 M sucrose as the CS and the other group receiving 16% corn oil as the CS. The 90 min daily preference tests conducted for both groups were between the sucrose–corn oil mixture and water. The rats that received the corn oil as the CS showed a profound aversion to the mixture across 5 days of testing and the group that received sucrose as the CS initially showed a mild aversion to the mixture that extinguished completely by the fifth day of testing.

It is not clear why the corn oil would be the salient feature in the sucrose–corn oil mixture. The rat could be discriminating the corn oil by taste, texture, or olfaction. In order to test the possibility that texture plays a major role, we performed an experiment to see if the rat could discriminate a sucrose–corn oil mixture from sucrose mixed with mineral oil. We hypothesized that the lubricity of corn oil and mineral oil would be quite similar to the rat and would result in the same textural sensations. In this experiment, 20 rats were given a sucrose–corn oil mixture as the CS in a CTA design. After drinking the mixture for 15 min, half of the rats received an injection of LiCl and the other half got a saline injection. On both of the following 2 days the rats received a 1 h two-bottle preference test between a sucrose–corn oil mixture and a blend of sucrose with mineral oil. The sucrose concentration was 0.25 M and both oils were 16%. The control rats drank equal amounts of the two solutions, but the LiCl-injected rats consumed significantly more of the mineral oil blend than of the corn oil. These data are consistent with the data of Greenberg and Smith (1996) in their procedure with sham-fed rats. They found that rats took an equal amount of the two solutions, but could easily discriminate between mineral and corn oil in subsequent tests.

6.3.4. Linoleic Acid (L) Is Detected in the Mouth but Is Not the Salient Feature of a Linoleic Acid–Sucrose Mixture

Gilbertson and others have proposed that linoleic, oleic, and other fatty acids are cleaved from corn oil in the rat’s oral cavity as the result of lingual lipase being secreted from Von Ebner’s gland. The rat would then get a gustatory sensation from the fatty acids (Gilbertson, 1998, 1999; Gilbertson et al., 1997, 1988a,b, 2005). We postulated that if we mixed linoleic acid with our sweeteners we should obtain results that are similar to what we found for mixtures of corn oil with sucrose or the glucose and saccharin solutions. The recordings made by Gilbertson showed that the taste cells were sensitive to concentrations of linoleic acid as low as 10 μM. Our first experiment was designed to see if rats could discriminate linoleic acid from water and then to determine the behavioral gustatory sensitivity to this fatty acid. We prepared the solution by dissolving 28 μL of linoleic acid in 1 mL of ethanol using a vortex mixer for 1 min. This solution was mixed into 4 L of deionized water. This linoleic mixture served as the CS in a CTA design. Following our standard water deprivation procedure we gave 20 male rats the linoleic acid solution for 15 min. Immediately after the CS period was over, we gave 10 of the rats a LiCl injection and the other 10 received an injection of NaCl. The preference test was initiated 24 h later and lasted for 1 h each day. An hour after the preference tests, the rats were given a water supplement for an hour in order to maintain good hydration. Rather than pairing the linoleic acid with water during the preference tests, the preference tests were between linoleic acid and ethanol solution (1 mL of ethanol in 4 L of water). The concentration of the linoleic acid was reduced each day of preference testing from 28 to 3.5 μM as can be seen in Figure 6.9. When compared to the NaCl-injected rats, the aversion was significant for the concentration of 28, 14, and 10.5 μM and not significant for the three lower concentrations. By this measurement, the threshold for detection would lie between 10.5 and 7 μM, a value that agrees with the electrophysiological findings of Gilbertson (Gilbertson, 1998, 1999; Gilbertson et al., 1997, 1998a,b, 2005).

FIGURE 6.9

These rats were conditioned to avoid linoleic acid by pairing a 28 μM solution with a LiCl injection When tested subsequently, they showed a significant aversion to concentrations of 28, 14, and 10.5 μM linoleic acid but no aversion to (more...)

In the next experiment we mixed the linoleic acid with the G + SA solution to see if the fatty acid would be a salient feature of that mixture as the corn oil proved to be in the previous experiments. Twenty male rats were given our standard water deprivation procedure and a mixture of linoleic acid in the G + SA solution was given for 15 min as the CS on conditioning day. After a 15 min period of drinking, half of the rats were given an injection of LiCl and the other half served as a control group receiving the saline injection. After a 24 h period with food and water available, a series of seven daily preference tests were administered in the following order: Day 1, G + SA + linoleic acid vs. water; Day 2, glucose vs. water; Day 3, saccharin vs. water; Day 4, linoleic acid vs. water; Day 5, G + SA vs. water; Day 6, G + linoleic acid vs. water; Day 7, saccharin + linoleic acid vs. water.

As can be seen in Figure 6.10, the rats developed an aversion to all combinations of the mixture of glucose, saccharin, and linoleic acid. Unlike the results from the experiments with corn oil mixed with the sweetened solutions (see Figure 6.7), linoleic acid does not appear to be the salient feature of the solution when it is mixed with G + SA. However, the rats do show a significant aversion to the linoleic acid when it is tested without the G + SA, indicating that they can detect it in the mixture.

FIGURE 6.10

Linoleic acid (L) mixed with a glucose plus saccharin solution (G + SA + L) was used as the CS in a CTA design with half receiving a LiCl injection and the other half receiving a NaCl injection after drinking the solution. Seven preference tests were (more...)

Our next experiment was to see if the rats could detect the linoleic acid when it was mixed with a high concentration of sucrose. In this experiment, 20 male rats were subjected to the standard water deprivation. On conditioning day they all received a 15 min CS period with a mixture of 0.25 M sucrose and 28 μM concentration of linoleic acid. Half the rats received the LiCl injection and the other half were injected with saline. Twenty-four hours later the first of three daily preference tests were given. These two-bottle tests lasted 1 h and the choices were as follows: Day 1, 0.25 M sucrose + 28 μM linoleic acid vs. water; Day 2, 0.25 M sucrose vs. water; Day 3, 28 μM linoleic acid vs. water.

In order to ensure that the small amount of ethanol that was a necessary addition in the solutions containing linoleic acid was not a factor, the same quantity of ethanol (4.3 mM) was added to the sucrose on Day 2 and the water in all of the preference tests. The results of this experiment can be seen in Figure 6.11. The preference in the LiCl-injected group is significantly lower than in the NaCl-injected group on all three tests. Once again we showed that the linoleic acid was not the salient feature of the sucrose–linoleic acid mixture (CS). However, we infer that the rat could taste the linoleic acid in the mixture. We failed to run extended testing in these two experiments which may have revealed a more rapid extinction of the sweeteners than with the linoleic acid as we had seen previously with the corn oil mixtures.

FIGURE 6.11

Linoleic acid mixed with sucrose (SU + L) was used as the CS in a CTA design with half of the rats receiving a LiCl injection and the other half receiving a NaCl injection after drinking the CS solution. Three preference tests were run comparing the solution (more...)

6.3.5. Ethanol Is Detected by the Olfactory System

Because ethanol was used as the solvent for mixing the linoleic acid, we next tested to see if rats could detect the alcohol at the concentrations that we used. The ethanol was tested by adding 0.25 mL of ethanol to 1 L of deionized water (a concentration of 4.3 mM). Twenty male rats were given the standard water deprivation for our CTA design. The CS on conditioning day was 4.3 mM solution of ethanol. Half the rats received a LiCl injection and the other half received the saline injection. Twenty-four hours later, the first of seven daily 1 h preference tests was administered. The fluids available in the preference test were: Day 1, 4.3 mM ethanol vs. water; Day 2, 4.3 mM ethanol vs. water; Day 3, 3.4 mM ethanol vs. water; Day 4, 2.6 mM ethanol vs. water; Day 5, 1.7 mM ethanol vs. water; Day 6, 0.9 mM ethanol vs. water; Day 7, 0.43 mM ethanol vs. water.

It can be seen in Figure 6.12, there is a clear aversion to the ethanol solution at all but the lowest concentrations. The ANOVA for these data gave a significant main effect for LiCl vs. NaCl injection, across the concentrations and the interactions. The Fisher LSD comparisons were significant for all but the 0.43 mM concentration. The sensitivity of the rats to such low concentrations of ethanol was so unusual to us that we replicated this finding two additional times. Because of the controls that we used in the previous experiments with linoleic acid (which was mixed with ethanol) we still conclude that the rats can taste linoleic acid, but we recommend that care be given in behavioral experiments where ethanol is used as a vehicle. It was not clear at this point if the rats discriminated the ethanol from water in these observations by gustation or by olfaction. Our next experiment was an effort to clarify this issue.

FIGURE 6.12

Ethanol at a concentration of 4.3 mM was used as the CS in a CTA design. After drinking the ethanol solution, half of the rats received a LiCl injection and the other half were given a NaCl injection. The ensuing two-bottle preference tests were given (more...)

For testing the role of olfaction in ethanol detection, 28 male rats were divided into two groups of 14 each. One group was subjected to a surgical procedure where the olfactory bulbs were ablated (OLFFx) and the second group served as sham-operated controls (SHAM). All rats were given the standard CTA water deprivation procedure and on conditioning day they received 4.3 mM solution of ethanol as the CS for 15 min. The two groups were subdivided into two groups of seven each. Seven rats from the OLFx group received a LiCl injection (OLFx-LiCl) and the other seven received saline injections (OLFx-SHAM). The 14 sham-operated rats were also divided into two groups of seven each. Half of these sham-operated rats received the LiCl injection (SHAM-LiCl) and the other seven received a saline injection (SHAM-NaCl). Twenty-four hours later they were given a 1 h two-bottle preference test between ethanol and water. The results of this preference test can be seen in Figure 6.13. The sham-operated rats showed a significant aversion to the ethanol solution, but the OLFx group exhibited no aversion. We concluded that the rats were detecting the low concentrations of ethanol as seen in the previous experiments by olfaction and not by taste.

FIGURE 6.13

Half of a group of rats were subjected to bilateral olfactory bulb ablations and the other half were sham-operated. After recovery, ethanol at a concentration of 4.3 mM was used as the CS in a CTA design. After drinking the ethanol solution, half of the (more...)

We repeated this last experiment using 22 μM linoleic acid as the CS in a CTA design. Fourteen male rats had their olfactory bulbs ablated and an additional 14 were sham-operated. Half of each of these groups were given LiCl and half saline after drinking the linoleic acid mixture for 15 min. The following day they were given a single 1 h two-bottle preference test between 22 μM linoleic acid and 4.3 mM ethanol. The results from this manipulation can be seen in Figure 6.14. Olfactory bulb ablations had no effect on the discrimination of linoleic acid from water in this CTA design. In contrast to the results from the ethanol test above, we concluded that linoleic acid is detected by the gustatory, not the olfactory, system. This finding is in agreement with the work of Pittman and his colleagues (McCormack et al., 2006; Pittman et al., 2006, 2007, 2008) and Stratford and her colleagues (Stratford et al., 2006).

FIGURE 6.14

Half of a group of rats were subjected to bilateral olfactory bulb ablations and the other half were sham-operated. Following recovery from surgery, linoleic acid at a concentration of 28 μM was used as the CS in a CTA design. After drinking the (more...)

6.3.6. Brief Access Studies with the Davis Rig

“The control of fat intake is the result of interactions between fat palatability under orosensory control and the satiating and metabolic effects of fat under the control of postingestive factors. The relative importance of these factors in the control of intake is unknown” (Greenberg and Smith, 1996). The work described thus far in this chapter from our laboratory does little to separate the role of orosensory factors from the postingestive factors since our preference tests have been at least an hour in length. It is clear that by the end of an hour, postingestive factors are playing a significant role in the ingestion of fat (Davis et al., 1995). However, most of the early research has emphasized texture and olfaction as the major contributors to the palatability of fat and have ignored the role of gustation. The work described here and in other chapters in this volume supports the contribution of the sense of taste proper in influencing fat intake.

Davis and colleagues (Davis et al., 1995) have thoroughly described the microstructure of corn oil drinking during 30 min tests. Depending on the concentration of the corn oil, the postingestive factors begin to have an effect after a few minutes of licking. Taste tests with the sham-feeding procedure essentially eliminate postingestive feedback and allow for an explicit study of the orosensory factors in fat ingestion. However, Davis’ findings would indicate that very short-term taste tests could also allow for the study of orosensory factors while minimizing postingestive feedback. A second advantage to these short-term tests in the Davis Rig is that there was little time for the phase separation of the corn oil from the water in the solutions. Finally, it is possible to measure the latency of the first lick after the Davis Rig shutter is opened, an important dependent variable that is not normally available in the standard two-bottle preference tests. If the rat shows some discrimination of the solutions before taking a lick, it is likely that olfaction is contributing to the oral sensations (Rhinehart-Doty et al., 1994). Thus, we conducted several experiments with short-term ingestive tests using the Davis Rig.

Six rats were placed on our water restriction regimen and trained to lick on the water spouts in the Davis Rig. A range of concentrations of corn oil was mixed as described earlier and loaded in seven of the tubes. The rats were tested with 0%, 1%, 2%, 4%, 8%, 16%, and 32% corn oil. Three daily tests were conducted and the average number of licks per second was calculated. These data are plotted in Figure 6.15. The rate of licking increased as a function of the concentration of the corn oil in a similar manner as we found in the hotels earlier.

FIGURE 6.15

The number of licks per second as a function of corn oil concentration is plotted here. It can be seen that the rats are licking at a maximum rate when the concentration of the corn is 4% and higher.

In the spirit of an experimental design used by P. T. Young (Young, 1966) to study the effectiveness of binary mixtures to stimulate licking in rats, we then made 30 solution combinations of sucrose and corn oil in order to test the lick rate as concentrations of both were increased. The concentrations of corn oil and sucrose used can be seen in Figure 6.16. When no fat was in the solution the lick rate rose as the concentration of sucrose was increased as was previously shown in Figure 6.2. When only 4% corn oil was added to the sucrose solutions, the rats were rapidly approaching their maximum rate of licking. In these very short-term drinking tests there is little probability that postingestive factors play any role in the rate of licking. We concluded that almost any combination of sucrose and fat is highly palatable to the rat.

FIGURE 6.16

Licks per second are plotted as a function of the concentration of corn oil (x-axis) that has been blended with the concentrations of sucrose indicated in the box on the right-hand side.

With the Davis Rig we replicated some of the CTA experiments where two-bottle preference tests were used to measure the strength of the aversion. When using the Davis Rig, our dependent variables would be the number of lick made in 30 s and the latency to the first lick.

After 10 rats were trained to lick in the apparatus, we filled all of the tubes with 0.25 M sucrose mixed with 16% corn oil. The rats had the opportunity to lick on all eight tubes for 30 s each with a 30 s delay between the presentations of each of the eight tubes. Five of the rats then received an injection of LiCl and the other five received the saline injection. The following day the tubes were filled as follows: Tubes 1 and 5 contained water; Tubes 2 and 6 contained the sucrose–corn oil mixture; Tubes 3 and 7 contained 25 M sucrose; Tubes 4 and 8 contained 16% corn oil.

The tubes were presented to the rats in order of 1–8. This presentation was repeated on the second day of testing. The results can be seen in Figure 6.17. The LiCl-injected rats licked very little when tubes were presented that contained corn oil when it was presented with or without sucrose. Quite like the results from the earlier two-bottle preference tests, the conditioned rats displayed no aversion to the sucrose. Because of the brevity of this test, we concluded that this discrimination of the corn oil was the result of an orosensory cue. When we measured the latency to the first lick for the LiCl-injected group we found that the latency to lick was significantly longer for the two tubes that contained the corn oil (Figure 6.18). This would give some evidence that the rats were sniffing at the opening and not licking until later in the 30 s period. As can be seen in this figure, the variation was quite high and the sample size was quite small for this observation.

FIGURE 6.17

Taste aversion to a mixture of sucrose and corn oil (SU + CO) presented in the Davis Rig is plotted here. When sucrose plus corn oil or corn oil alone was presented in one of the drinking tubes there was a significant difference between LiCl- and NaCl-injected (more...)

FIGURE 6.18

Latency to the first lick for the solutions indicated is presented here. If the solution presented contains corn oil, this latency is significantly longer than when either water or sucrose is presented.

We conducted another CTA experiment with the sucrose–corn oil mixture to test for a role of texture in the detection of corn oil. In this experiment, 22 rats were given training to lick in the Davis Rig. On conditioning day, the rats received the mixture of 0.25 M sucrose and 16% corn oil in each of the eight tubes. After drinking it, half of the rats were given an injection of LiCl and the other half received a saline injection. On the following day three tubes were filled with the sucrose–corn oil mixture, three with a sucrose–mineral oil mixture and the other two tubes with water. They were presented in random order. From Figure 6.19 it can be seen that on Day 1 the LiCl-injected rats developed a strong aversion to both corn and mineral oil as compared to water drinking. However, over the next 5 days, the aversion to mineral oil extinguished much faster than it did to the corn oil. The rats were significantly discriminating the corn oil from the mineral oil on the second day of testing. The saline injected rats showed no significant change in licking over the 6 days if testing, with about the same number of licks on the corn oil and the mineral oil tubes. These saline injected rats did lick significantly less on the water tube during the 30 s tests than to both corn oil and mineral oil.

FIGURE 6.19

In this CTA experiment, corn oil was the CS followed by an injection of either LiCl or NaCl. There were subsequently tested with corn oil, mineral oil, or water in the testing tubes in the Davis Rig. It can be seen on the first day of testing that there (more...)

Earlier we presented data showing that by using a mixture of glucose, saccharin, and corn oil as the CS in a CTA design, we could condition an aversion to any combination of these three compounds as long as it contained the corn oil. It was not clear if this conditioning was the result of orosensory factors only or if some postingestive factor played a role. Furthermore, we could not rule out olfaction as part of the orosensory sensation. We repeated that conditioning design by measuring the flavor aversion in a short-term test in the Davis Rig. Thus, we could minimize the role of postingestive factors and by measuring the latency to the first lick in the Davis Rig we could get some idea if olfaction played a role in that discrimination. Fourteen rats were trained to lick on the eight tubes in our Davis Rig. On conditioning day all eight tubes were filled with a mixture of glucose, saccharin, and corn oil (16% corn oil blended with the standard G + SA mixture). Following this drinking session, eight rats received the LiCl injection and the other six received a saline injection. The following day the tubes were filled as follows: Tube 1, water; Tube 2, glucose + saccharin + corn oil; Tube 3, glucose; Tube 4, saccharin; Tube 5, corn oil; Tube 6, glucose + saccharin; Tube 7, glucose + corn oil; and Tube 8, saccharin + corn oil.

The mean number of licks in each of the 30 s presentations can be seen in Figure 6.20. There was some aversion generalized to the saccharin (Tube 4) and to the glucose + saccharin (Tube 6), but the strong aversions were seen to the tubes that contained the corn oil. It was not surprising that we saw no aversion to the glucose since we previously have shown that the salient feature of the glucose and saccharin mixture is the saccharin. The brevity of this test allowed us to conclude that these aversions were most likely the result of orosensory factors. What we found most interesting was the latency to the first lick when the shutter in the Davis Rig opened. If the tube contained corn oil this latency was significantly longer as can be seen in Figure 6.21. The rats were detecting the aversive characteristics of the CS prior to taking their first lick.

FIGURE 6.20

These rats were conditioned with glucose (G) + saccharin (SA) + corn oil (CO) and tested with the solutions indicated in the Davis Rig. Differences in the number of licks that were statistically significant between LiCl- and NaCl-injected groups are indicated (more...)

FIGURE 6.21

These rats were conditioned with glucose (G) + saccharin (SA) + corn oil (CO) and tested with the solutions indicated in the Davis Rig. Differences in the latency to the first lick that are statistically significant between LiCl- and NaCl-injected groups (more...)

For the final experiment to be reported in this chapter, we replicated the previous experiment using linoleic acid in place of corn oil in the mixtures. As can be seen in Figure 6.22 using G + SA + linoleic acid as the CS in the CTA design, the aversion generalized to all of the components of this mixture. The important result here is that the rats displayed an aversion to the linoleic acid when it was presented alone. It is unfortunate that we did not run extinction trials here since it would have been interesting to see which components extinguished most rapidly. The latency to make the first lick was not different for either the tube containing the G + SA + linoleic acid or the linoleic acid alone. There was no sign that the rats were detecting the linoleic acid by olfaction.

FIGURE 6.22

These rats were conditioned with glucose (G) + saccharin (SA) + linoleic acid (L) and tested with the solutions indicated in the Davis Rig. Differences in the number of licks that are statistically significant between LiCl- and NaCl-injected groups are (more...)