Post-oral sugar detection rapidly and chemospecifically modulates taste-guided behavior

Abstract

Several recent studies have shown that post-oral sugar sensing rapidly stimulates ingestion. Here, we explored the specificity with which early-phase post-oral sugar sensing influenced ingestive motivation. In experiment 1, rats were trained to associate the consumption of 0.3 M sucrose with injections of LiCl (3.0 meq/kg ip, conditioned taste aversion) or given equivalent exposures to the stimuli, but in an unpaired fashion. Then, all rats were subjected to two brief-access tests to assess appetitive and consummatory responses to the taste properties of sucrose (0.01–1.0 M), 0.12 M NaCl, and dH₂O (in 10-s trials in randomized blocks). Intraduodenal infusions of either 0.3 M sucrose or equiosmolar 0.15 M NaCl (3.0 ml) were administered, beginning just before each test. For unpaired rats, intraduodenal sucrose specifically enhanced licking for 0.03–1.0 M sucrose, with no effect on trial initiation, relative to intraduodenal NaCl. Rats with an aversion to sucrose suppressed licking responses to sucrose in a concentration-dependent manner, as expected, but the intraduodenal sucrose preload did not appear to further influence licking responses; instead, intraduodenal sucrose attenuated trial initiation. Using a serial taste reactivity (TR) paradigm, however, experiment 2 demonstrated that intraduodenal sucrose preloads suppressed ingestive oromotor responses to intraorally delivered sucrose in rats with a sucrose aversion. Finally, experiment 3 showed that intraduodenal sucrose preloads enhanced preferential licking to some representative tastants tested (sucrose, Polycose, and Intralipid), but not others (NaCl, quinine). Together, the results suggest that the early phase-reinforcing efficacy of post-oral sugar is dependent on the sensory and motivational properties of the ingesta.

foods and fluids elicit a cascade of sensory and metabolic signals, as they are ingested, digested, and assimilated into the body. Among the many wisdoms of the body are its capacities to predict a food's biologically significant consequences and tailor ingestion to nutritional needs by flexibly integrating these oral and postoral sensory signals both within the meal and across meals (via learning). Within the meal, these two systems have been conventionally characterized as separate, yet counterbalanced. Accordingly, whereas the taste signals are critical for selecting and promoting the ingestion of palatable and presumptively nutritious foods, the post-oral sensory systems are thought to negatively feedback to inhibit intake as sufficient nutrient loads accumulate in the gastrointestinal (GI) tract. Then, the post-oral events that unfold in the minutes and hours after the meal are generally thought to provide the more qualitative feedback (e.g., nutrient type, metabolic efficacy, and illness-inducing), which confer lasting changes to the motivational significance of the earliest perceptible signal(s) in that sensory cascade, and, in effect, reinforce appetitive and consummatory responses toward the orosensory properties of foods with known beneficial outcomes in future meals. Indeed, there are extensive literature in support of each of these respective controls (e.g., 2, 10, 11, 14, and for reviews, see Refs. 43 and 50).

Interestingly, however, with the aid of temporally detailed behavioral measures, a number of recent studies have revealed a distinct “early phase” of positive post-oral feedback generated by some nutrients, like sugars, which rapidly stimulates ingestion and leads to an increase in overall meal size (e.g., 3, 28, 44, 46, 62, 63). The functional relevance of this response remains to be determined, but one interesting possibility is that it serves as an online motivational counterpart to the gustatory system. For while the gustatory system has clearly evolved to respond as expediently as possible to broad classes of biologically significant chemicals, there are oftentimes discrepancies among a food's taste and its nutrient content (or other post-oral consequences). Sugars and artificial sweeteners are a good example of this. Both classes of compounds bind with the same taste receptor and elicit a similar palatable sensation (60), yet only the former eventually yields energy. Similarly, many unpalatable foods contain sugars and starches, essential amino acids, and/or fats, as well as various micronutrients. So, on the one hand, the capacity for post-oral input to leverage ingestive motivation, regardless of the information being received from the oral receptors, could be a useful strategy to rapidly resolve some of these ambiguities and accommodate adequate nutrient intake. However, on the other hand, the harmful effects of an ingested food may be rather delayed from these early-phase signals and even from the meal at large; hence, increasing the amount of that substance consumed in a given meal could unwittingly put the animal in more danger.

This poses the important question of specificity: is the early-phase post-oral effect a general one, independent of the orosensory properties and/or hedonic significance of a stimulus, or is it somehow contingent upon these factors? Indeed, the answer to this question will help define the rules governing the motivation to ingest. Here, we conducted a series of three experiments to begin to assess precisely how post-oral sugar stimulation exerts its influence on ingestive behaviors elicited by taste compounds varying in sensory quality and/or hedonic value. The specific designs, behavioral tasks, and hypotheses, are introduced for each experiment, in turn, below.

Experiment 1. Effects of Intraduodenal Sucrose on Taste-Guided Behaviors to Sucrose and NaCl in Rats with or Without a Conditioned Taste Aversion to Sucrose

The purpose of this experiment was to begin to assess whether the early-phase sugar signal exerts a generic positive effect on ingestive behaviors or whether its influence depends on the orosensory and/or motivational properties of the substance being consumed. Additionally, we sought to determine whether the post-oral sugar stimulus works through consummatory and/or appetitive behavioral outputs. Toward each of these ends, we used a modified brief-access taste test paradigm. The brief access taste test involves the presentation of an array of taste stimuli in discrete 10-s trials in a serial randomized order across a given session. Importantly, the rat autonomously initiates each 10-s trial. As such, this test provides separate measures of appetitive responding in terms of the number of trials initiated and consummatory responding in terms of the number of licks elicited to each taste stimulus. Moreover, the short duration of these trials minimizes the potential for the post-oral effects of the taste stimulus to impact responding in a given trial. Thus, to examine the effect of an early-phase post-oral sugar stimulus on these measures of taste-guided behavior, a small (3.0 ml) sucrose bolus (or saline control) was infused directly into the duodenum, beginning just prior to the start of the brief-access taste test. The taste array included five sucrose concentrations and a nonsweet probe stimulus (0.12 M NaCl). For half of the rats, the hedonic significance of the sucrose stimuli was left intact (i.e., positive, unpaired group), while, for the remaining rats, the same sucrose stimulus was rendered aversive by pairing its consumption with the illness-inducing agent, lithium chloride (LiCl) [conditioned taste aversion (CTA) group] in the home cage prior to the brief-access taste test phase. As in previous studies (39, 40), infusions were delivered directly into the duodenum, rather than to the stomach, because various lines of evidence suggest that the chemoreceptors that subserve rapid, within-meal, effects, as well as flavor-nutrient learning, are postgastric (e.g., 2, 13, 34, 54).

We hypothesized that, consistent with previous studies (e.g., 3, 62, 63), intraduodenal sucrose would rapidly stimulate ingestive behavior, but that this experiment would extend those findings to show that post-oral sugar more effectively stimulates or maintains licking for certain types of taste stimuli (i.e., sucrose over NaCl) in unpaired rats. We further hypothesized that this influence of the post-oral sugar depends on the hedonic significance of the sucrose stimulus. In other words, intraduodenal sucrose is not expected to similarly enhance responding to oral sucrose in rats that have a CTA to sucrose. If anything, intraduodenal sucrose ought to further reduce appetitive and/or consummatory responses to oral sucrose under these conditions.

MATERIALS AND METHODS

Subjects.

Twenty naïve male Sprague-Dawley rats (∼8 wk old at the start of the experiment) were individually housed in environmentally enriched (with Rattle-A-Round, Otto Environmental) hanging shoebox cages in a climate-controlled colony room (lights on 0700; lights off 1900). All experimental procedures were carried out in the light phase. Rats had ad libitum access to pelleted chow (LabDiet 5001, PMI Nutrition) and deionized water (dH₂O) in the home cage, except as noted otherwise for training and testing purposes. Rats had ad libitum access to pelleted chow (Lab Diet 5001, PMI Nutrition) and deionized water (dH₂O) in the home cage, except as noted otherwise for training and testing purposes. All experimental protocols were approved by and conducted in accordance with the Florida State University Animal Care and Use Committee.

Surgery.

After an overnight fast, each rat was anesthetized with a ketamine:xylazine mixture (125 mg/kg:25 mg/kg im) and then laparotomized. An intraduodenal catheter (inside diameter = 0.64 mm, outside diameter = 1.19 mm; Dow Corning, Midland, MI) was introduced through a puncture wound in the greater curvature of the forestomach and then advanced through the pyloric sphincter. The catheter was attached to the intestinal wall ∼4 cm distal to the pyloric sphincter with a single stay suture and piece of Marlex mesh (Davol, Cranston, RI). The puncture wound in the stomach was closed around the free end of the catheter with a purse string suture and concentric serosal tunnel. The free end was then tunneled subcutaneously to an interscapular exit site, where it was exteriorized and connected to a Luer lock adapter, which was mounted in a harness worn by the rat at all times (Quick Connect Harness, Strategic Applications, Lake Villa, IL). Postoperative antibiotic (penicillin G Procaine 30,000 units in 0.1 cc sc) and analgesic (Ketoralac, 2 mg/kg sc) were administered immediately after surgery and once daily for 3 days thereafter to aid recovery. Rats were given a limited amount (∼10 g) of chow mash (∼50% powdered chow: 50% dH₂O) after surgery and then ad libitum access to chow mash for at least 2 days, before being gradually introduced back onto regular pelleted chow over the next few days. The intraduodenal catheter was routinely flushed with 0.5 ml of isotonic saline beginning 48 h after surgery to maintain its patency. Harness bands were adjusted daily to accommodate changes in body mass.

Stimuli.

Reagent-grade sodium chloride (NaCl) (BDH Chemicals), lithium chloride (LiCl) (Sigma Aldrich), and sucrose (Mallinckrodt) were mixed into solutions with dH₂O fresh each day as needed to the concentrations indicated in the Brief-access training and testing subsection.

Apparatus.

Brief-access taste tests were conducted in one of four identical Davis Rigs (Davis MS-160; DiLog Instruments, Tallahassee, FL). Each chamber comprised a wire mesh grid floor, three Plexiglas walls, and a stainless-steel front wall. An access slot was located in the center of the front wall, ∼2 mm above the grid floor. This access slot was opened and closed by a computer-controlled shutter on the exterior of the front wall. A motorized table, positioned proximal to the front wall, holds up to 16 bottles with sipper tubes connected to a contact lickometer. The accompanying software allowed the experimenter to control the order and amount of time that the rat has access to each test stimulus on the motorized table. Licks were registered through a high-frequency AC circuit, timestamped, and saved for subsequent analyses. Thus, in a typical situation, the computer program positioned the designated tube at the center access slot, the shutter opened, and the trial commenced when the rat made contact with the sipper spout (i.e., the first lick). After the set amount of time elapsed (in this case, 10 s), the shutter closed and the motorized table repositioned so that the next sipper tube in the schedule was in front of the center access slot and so on. A small fan was located above the center access slot to minimize odor cues. The intraduodenal infusion line consisted of Tygon tubing connected to pump (model 500; New Era Pump Systems, Farmingdale, NY) on one end and a single-channel swivel (Instech Solomon, Plymouth Meeting, PA) suspended in a counterbalance arm that was mounted just above the Davis Rig on the other end. A second segment of polyurethane tubing encased in a spring tether was attached to the swivel and connected to the rat's harness via a Luer lock connector; this arrangement allowed infusates to be covertly delivered directly to the small intestine, while still permitting the rat to move freely about the chamber.

Brief-access training and testing.

Upon recovery from surgery (∼14 days), rats were placed on a restricted food and water access schedule. On this schedule, rats were given access to dH₂O for 30 min each day in the Davis Rig, followed ∼30 min later by access to chow and an 8-ml supplement of dH₂O in the home cage for 3 h. Rats were maintained on this schedule for two 5-day blocks, separated by 2 days of ad libitum access to food and dH₂O in the home cage. During the first 2 days of the first 5-day block, rats were placed into the test apparatus and presented with dH₂O through a stationary sipper tube at the center slot for 30 min. This was followed by three daily sessions in which dH₂O was presented in 10-s stimulus trials, each separated by a 1-s dH₂O trial. This obligate 1-s dH₂O trial was presented between each 10-s test stimulus trial to allow the rat the opportunity to rinse its tongue between each test stimulus. Rinse trials were not included in the data analyses. An interstimulus interval of 7.5 s was interposed between each 10-s trial and 1-s rinse in the sequence. These sessions likewise lasted 30 min, during which time, the rat was free to initiate as many trials as possible. To acclimate the rats to the infusion system, the second block began with two 30-min sessions in which the rat was placed in the test apparatus and connected to the infusion line. The infusion pump started, and 2 min later, the center slot shutter opened, and access to dH₂O in 10-s trials began, as described above. The infusion pumps ran for an additional 2 min (4 min in total), but no infusates were administered. The third session in this block was the first of two pre-CTA brief-access tests. These test sessions were identical to the previous two, except that rats were infused with one of two intraduodenal infusates (3.0 ml, 0.75 ml/min for 4 min), beginning 2 min prior to the start of the brief-access tests. Half of the rats were given an intraduodenal infusion of 0.15 M NaCl prior to the first test and an ID infusion of equiosmolar 0.3 M sucrose prior to the second test. The remaining rats received the alternative order. On both tests, rats were presented with the same seven taste stimuli in the 10-s trials, in place of dH₂O. The taste stimuli included five concentrations of sucrose (0.01, 0.03, 0.1, 0.3, 1 M), 0.12 M NaCl (a nonsweet taste probe stimulus), and dH₂O. These tastants were presented in serial randomized order in blocks of seven trials (without replacement). All other session parameters were identical to the dH₂O sessions that preceded them. The two test sessions were separated by 1 day in which dH₂O only was presented in the 10-s trials (no intraduodenal infusates were administered). These pre-CTA tests were simply designed to accustom the rats to receiving intraduodenal infusates and licking for various target taste solutions. The lick responses and number of trials initiated on these tests were subsequently used to match CTA training groups (see below). Three rats were discontinued from the experiment during the pre-CTA brief-access testing phase due to intraduodenal catheter failure. After CTA training (see below), the rats were placed on the same food and water access schedule (as described above) and were reacquainted with taking dH₂O in 10-s trials in the Davis Rigs in two 30-min sessions. Then, the two brief-access tests were conducted again in an identical manner to that described above.

CTA training.

After pre-CTA brief-access acclimation testing, rats were allowed ad libutum access to chow and water in the home cage for two days, and then were placed on a restricted water-access schedule, in which dH₂O was presented for 15 min each morning and for 30 min ∼5 h later each afternoon for three consecutive days. All morning and afternoon fluid access sessions took place in the home cage for this phase. Intakes were measured to the nearest milliliter. After the third day, rats were divided into two CTA training groups, matched on 15-min dH₂O intake (on day 3), body weight (on day 3), mean number of trials initiated, and mean licks per 10-s trial on the pre-CTA brief-access acclimation tests. Then, on the fourth day, rats were given a CTA training session. Half of the rats were given 0.3 M sucrose in place of the usual morning dH₂O, while the other half was given dH₂O as usual. Immediately after this session, rats were injected with one of two agents (3 meq/kg 0.15 M LiCl or an equivalent volume of 15 M NaCl ip), according to their assigned treatment condition. Rats in the CTA group were injected with LiCl after consuming sucrose and saline after consuming dH₂O. Rats in the unpaired control group were injected with saline after consuming sucrose, and LiCl after dH₂O. Deionized water was presented as usual in the afternoon session. On the following day, rats were presented with the alternate solution (0.3 M sucrose or dH₂O) in the morning session and then injected with either LiCl or NaCl, as prescribed. Once again, dH₂O was presented as usual in the afternoon session. Together, these two sessions comprised a full CTA-conditioning trial. This two-session trial was repeated two more times over the next 7 days, each separated by 1–2 days in which dH₂O was presented in the morning session (no intraperitioneal injections). Deionized water was returned to the home cage in the afternoon of the last CTA training session for 4 days to allow the rats to access chow and water ad libitum before post-CTA brief-access testing began.

Data analyses.

Intakes on the three two-session (CS vs. dH₂O) CTA trials were compared across training groups in a mixed repeated-measures ANOVA, with stimulus and group as factors. Licking responses to the various taste stimuli presented in the brief-access taste tests were standardized against lick responses to the water stimulus in the same session as follows: Lick Score = mean licks to stimulus − mean licks to dH₂O.

The effects of the two intraduodenal infusates on post-CTA lick scores to the five sucrose solutions were analyzed in repeated-measures ANOVA with ID infusate and concentration as factors. Where appropriate, separate t-tests were used to compare the effects of intraduodenal sucrose vs. intraduodenal NaCl on lick scores for each taste stimulus in the array. Then, to assess the time course of the intraduodenal infusate effects on lick responses, repeated-measures ANOVAs were conducted on lick scores on trial blocks 1–3 for a given taste stimulus. These trial blocks were selected because 1) all unpaired rats completed at least three full trial blocks on both post-CTA tests and 2) these blocks spanned the approximate first third of session (completed, on average, by 9.66 min post-intraduodenal infusion). The number of trials initiated on the two post-CTA brief access test sessions were analyzed as a function of intraduodenal infusate in separate t-tests for each training group. In all statistical tests, a P ≤ 0.05 was considered significant. The Bonferroni procedure was used to correct for multiple comparisons.

RESULTS

CTA.

Figure 1 shows sucrose and dH₂O consumption for rats explicitly conditioned to avoid sucrose (CTA Group) relative to the controls (Unpaired Group); unpaired rats received equivalent exposures to sucrose and LiCl, but in an unpaired fashion, across the three training trials. Table 1 presents the full statistical outcomes of the corresponding three-way ANOVA. As expected, rats in the CTA group gradually reduced intake of sucrose across trials, completely avoiding the stimulus on the third and final trial. Rats in the Unpaired Group, on the other hand, maintained high levels of sucrose consumption across all three trials. Importantly, despite the fact that unpaired rats received LiCl injections immediately after the dH₂O sessions of each trial, intake of dH₂O did not drop across trials for this group. In fact, dH₂O remained quite stable for both groups across this phase.

Fig. 1.

A: means ± SE intake of 0.3 M sucrose that was paired with either an intraperitoneal injection of LiCl [conditioned taste aversion+ (CTA+), n = 9] or saline (UNPAIRED−, n = 8) on each trial. B: mean ± SE intake of dH₂O on the alternate sessions, after which the CTA rats were intraperitoneally injected with saline (−) and the UNPAIRED rats were intraperitoneally injected with LiCl. Asterisks (*) denote Bonferroni-corrected significant differences.

Table 1.

Experiment 1: CTA training and post-CTA brief access testing statistical outcomes

CTA Training
Group Effect	Stimulus Effect	Trial Effect	Trial × Group	Trial × Stimulus	Group × Stimulus	Group × Stimulus × Trial
F(1,15)=34.41, P = 0.00003	F(1,15)=0.003, P = 0.99	F(2,30)=63.59, P < 0.000001	F(2,30)=15.55, P = 0.00002	F(2,30)=25.30, P < 0.000001	F(1,15)=62.23, P < 0.000001	F(2,30)=13.77, P = 0.00006

PostCTA Brief Access Taste Tests
		Lick Scores ANOVA:Sucrose Concentrations
Group	Trials Initiated t-Tests: ID Sucrose vs. ID NaCl	ID Effect	Conc. Effect	ID × Conc.	Lick Scores t-Tests: ID Sucrose vs. ID NaCl
Unpaired (n = 8)	t(7) = 0.13, P = 0.90	F(1,7) = 7.08, P = 0.03	F(4,28) = 93.80, P < 0.00001	F(4,28) = 4.05, P = 0.01	0.12 M NaCl	t(7) = 0.22, P = 0.83
					0.01 M Suc.	t(7) = −0.57, P = 0.59
					0.03 M Suc.	t(7) = 4.61, P = 0.002
					0.1 M Suc.	t(7) = 2.33, P = 0.052
					0.3 M Suc.	t(7) = 2.43, P = 0.05
					1 M Suc.	t(7) = 2.36, P = 0.05
CTA (n = 9)	t(8) = 3.60, P = 0.007	F(1,6) = 0.07, P = 0.80	F(4,24) = 39.41, P < 0.00001	F(4,24) = 0.68, P = 0.61

Significant effects are denoted by italics.

Conc., concentration; ID, intraduodenal; M, molar; Suc., sucrose.

Post-CTA brief-access tests.

Unpaired rats evinced concentration-dependent licking for sucrose (Fig. 2A, Table 1). Interestingly, these lick responses were increased further by a brief intraduodenal infusion of 0.3 M sucrose, relative to an equivolume and equiosmotic infusion of 0.15 M NaCl. This difference was especially pronounced at the 0.03–1.0 M sucrose concentrations. Lick responses to the nonsweet tastant, 0.12 M NaCl, were not differentially affected by the two intraduodenal infusates, nor was the lowest concentration of sucrose. Despite clear effects on these relative consummatory responses for sucrose, intraduodenal infusions did not appear to impact appetitive responding for these test solutions, as measured in the total number of trials initiated (see Fig. 2B).

Fig. 2.

A: means ± SE lick responses (standardized against licks to dH₂O) of the UNPAIRED rats (n = 8) for five concentrations of sucrose and a single concentration of NaCl presented in 10-s trials (randomized order) on two brief access taste tests. Beginning just prior to the start of one test, an ID infusion of 0.3 M sucrose (ID sucrose) was administered at a rate of 0.75 ml/min for 4 min, for a total infusion volume of 3 ml. An equivalent volume of saline (ID NaCl, 0.15 M) was infused, beginning just prior to the start of the alternative brief access taste test session. For reference, the mean ± SD licks to dH₂O on the ID NaCl test was 35.92 ± 9.10 and on the ID Sucrose test was 24.67 ± 13.87 [t(7) = 2.63, P = 0.03]. B: corresponding mean ± SE trials initiated on each of these test sessions. Asterisks (*) denote Bonferroni-corrected significant differences. Carets (^) denote uncorrected significant differences. C and D: Means ± SE lick scores (standardized against licks to dH₂O) for a representative 0.1 M sucrose concentration and the 0.12 M NaCl stimulus to each of the first three trial blocks for the UNPAIRED group. Lick scores are plotted at the mean time point of the last trial in the respective block, relative to the last 2 min of the ID infusion period and start of the brief access taste test (time point 0).

Figure 2, C and D plot the lick responses to a representative sucrose stimulus (0.1 M) and 0.12 M NaCl, respectively, across each successive trial block for the unpaired rats. Clearly, intraduodenal sucrose more rapidly enhanced licking for sucrose than did intraduodenal NaCl in the approximate first third of the test session [ID effect: F(1, 7) = 7.16, P = 0.03], although this did not statistically vary as a function of trial block [ID × trial: F (2, 14) = 10.8, P = 0.37]. Similar patterns were observed for the highest four sucrose concentrations tested (not shown). By contrast, however, intraduodenal sucrose failed to enhance preferential licking for 0.12 M NaCl on any trial block [intraduodenal effect: F(1, 7) = 2.77, P = 0.14; intraduodenal × Trial: F(2, 14) = 0.13, P = 0.88].

As expected, rats previously conditioned to avoid 0.3 M sucrose in the home cage showed a dramatic reduction in licking for sucrose in the brief-access taste test (Fig. 3A, Table 1). Lick responses to sucrose showed an inverse concentration effect, with lick scores decreasing as the sucrose concentration increased. Although lick responses did not further vary as a function of intraduodenal infusate, intraduodenal sucrose significantly reduced the number of trials initiated by the CTA rats (see Fig. 3B, Table 1).

Fig. 3.

A: Means ± SE lick responses (standardized against licks to dH₂O) for five concentrations of sucrose and a single concentration of NaCl presented in 10-s trials (randomized order) in each of two brief access taste tests. CTA rats (n = 9) were trained to associate 0.3 M sucrose with the malaise-inducing agent LiCl (3 meq/kg; 0.15 M) in the home cage, prior to these tests. Beginning 2 min before the start of one test, an ID infusion of 0.3 M sucrose (ID sucrose) was administered at a rate of 0.75 ml/min for 4 min, for a total infusion volume of 3 ml. An equivalent volume of saline (ID NaCl, 0.15 M) was infused, just prior to the start of the alternative brief access taste test session. The black arrowhead marks the sucrose concentration used for CTA training. B: the corresponding mean ± SE trials initiated on each of these brief access test sessions. Asterisk (*) denotes a significant difference. For reference, the mean ± SD licks to dH₂O on the ID NaCl test was 68.76 ± 5.17 and on the ID sucrose test was 68.29 ± 4.48 [t(8) = 0.18, P = 0.86].

Experiment 2. Effects of Intraduodenal Sucrose on Ingestive and Aversive Oromotor Responses to Intraorally Delivered Sucrose in Rats with a CTA to Sucrose

In experiment 1, whereas intraduodenal sucrose rapidly and specifically enhanced licking for sucrose in a concentration-dependent manner in the Unpaired Group, it did not appear to affect consummatory responding to sucrose one way or another in the CTA group. That said, CTA rats took significantly fewer trials, and the intraduodenal sucrose bolus significantly depressed trial taking even further. In fact, some rats in this group failed to complete even one full trial block. Thus, the primary effect on appetitive responding may have preempted the emergence of any effects on consummatory responding. Therefore, experiment 2 employed a serial taste reactivity (TR) test to directly assess whether intraduodenal sucrose modifies consummatory oromotor responses to oral sucrose. This test paradigm takes advantage of the fact that rats elicit stereotyped oromotor reactions in response to stimulation of the oral cavity with a taste solution; these responses are thought to be a reflection of the tastant's hedonic value (e.g., 19–21, 51). As such, these responses can be categorized as “ingestive TR,” which comprises positive reactions, such as tongue protrusions and mouth movements, associated with ingestion, and “aversive TR,” which comprises negative reactions, such as gapes and chin rubs, associated with the rejection or removal of the solution from the oral cavity. In this test procedure, a tastant of interest (in this case, 0.3 M sucrose) is delivered directly into the oral cavity of the rat under experimenter-controlled conditions, and, as such, requires no appetitive action. As in experiment 1, rats were trained to associate sucrose with LiCl-induced malaise in the home cage prior to the test phase. During the TR tests, rats were first preloaded with a small volume (3.0 ml) of intraduodenal 0.3 M sucrose or ID 0.15 M NaCl. Then, beginning immediately after the ID infusion, brief intraoral (IO) infusions (0.5 ml, 30 s) of 0.3 M sucrose were infused every ∼3 min across the 15-min test session. Ingestive and aversive oromotor responses to IO sucrose were video recorded at each time point and later scored offline.

Subjects.

Naïve male Sprague-Dawley rats (n = 18, ∼11 wk of age at the start of the experiment) were maintained in conditions identical to that described for experiment 1. All experimental protocols were approved by and conducted in accordance with the Florida State University Animal Care and Use Committee.

Surgery.

After an overnight fast, each rat was surgically outfitted with an intraduodenal catheter and connecting harness, as described in experiment 1, except that isoflurane (5% for induction, 2–2.5% for maintenance) was used as the anesthetic agent instead of ketamine/xylazine. At the same time, bilateral IO cannulas were implanted according to the Grill and Norgren procedure, with some minor modifications (see Refs. 19 and 51). Rats were treated postoperatively with analgesic (carpofen, 5 mg/kg) and antibiotic (Baytril, 2.3 mg/kg) immediately after the surgery and once per day for 3 days thereafter. After surgery, rats were given chow mash to eat and then were gradually weaned back onto ad libitum chow, according to the schedule described in experiment 1. In some cases, rats did not consume enough of the chow to maintain/gain weight in the days after surgery; these rats were offered chow mash in addition to the pelleted chow in the home cage until weight gain had resumed. Beginning 48 h after surgery, harnesses were adjusted to accommodate changes in body mass, intraduodenal infusion lines were flushed with 0.5 ml of sterile dH₂O, and IO cannulas were cleared of chow and debris each day.

Apparatus.

The taste reactivity chamber was a cylindrical arena with clear Plexiglas walls and a clear Plexiglas floor. A dual-channel stainless-steel swivel (21G, Instech Solomon) was mounted in the chamber lid, with two separate infusion lines, each encased in a spring tether, one for connection to the IO cannula and one for connection to the intraduodenal catheter port. This arrangement allowed the rat to move freely about the chamber, while preventing the rat from accessing the lines and preventing the lines from being tangled. The alternate ends of the respective lines were run to two separate infusion pumps. A mirror was mounted at a 45° angle just below the chamber floor, and a digital video camera (Sony DSC-WX50 HD) was positioned facing the mirror on a tripod ∼35 cm away. Video files were viewed off line and in slow motion (frame by frame) by a trained observer, who was unaware of the stimulus or group assignment of the animal, using Sony Movie Studios HD software (v. 11) for scoring (see Data analyses).

Stimuli.

Molar concentrations of reagent-grade NaCl (BDH Chemicals), LiCl (Sigma), and sucrose (Mallinckrodt) were made fresh each day with dH₂O, as needed.

Pre-CTA taste reactivity acclimation and testing.

All rats were first acclimated to the TR chamber and IO infusions of dH₂O each day for three consecutive days. Chow and water bottles were removed from the home cage before the first TR acclimation session (20 h and 4 h, respectively). During the acclimation sessions, the rat was placed in the chamber and connected to the ID infusion and IO infusion lines. Three minutes later, the ID infusion pump was started, but no ID infusion was administered. The ID infusion pump ran for a total of 4 min. When that period had elapsed, the IO pump was started (time 0), delivering dH₂O to the rat via the IO cannula at a rate of 1 ml/min for a total of 30 s from when the rat initiated oromotor responding. The IO infusion was repeated five more times across the remaining 15 min of the session (i.e., a 30-s IO infusion of dH₂O was administered every 3 min for a total of six IO infusions). Then, after the final IO infusion, the rat was placed back into its home cage and, ∼30 min later, the water bottle was returned, and 3-h access to chow was permitted. The same procedure was repeated on each of the next 2 days. The 4th and 5th days were pretest days. Pretest sessions were run identically to the acclimation sessions, with two exceptions. First, prior to the start of each test session, a different intraduodenal infusion was delivered during the 4-min period that preceded the first IO infusion (3.0 ml, 0.75 ml/min). For half of the rats, 0.3 M intraduodenal sucrose was delivered before the first test, and 0.15 M intraduodenal NaCl was delivered before the second test; the other half of the rats received the same infusates, just in the opposite order. On both tests, all rats were given 0.3 M sucrose intraorally in each of the six 30-s infusions, instead of dH₂O. All rats were subjected to both tests so as to equate experience with each of the infusates prior to CTA training, but only responses on the first test were scored and analyzed (see below). One rat was discontinued from the experiment prior to the first TR test due to ID catheter failure. Data from two rats were excluded from pre-CTA TR test analyses because, in both cases, one of the six video files from the first TR test failed to save to the SD card. However, because this error in no way interfered with their experience, these rats were continued in the experiment (for post-CTA analyses, see below).

After CTA training (see below), rats were food and water restricted as above and reacclimated to the TR chambers and infusion procedures over three consecutive days. The fourth session served as the one and only post-CTA serial TR test session. During this test, all of the rats received six discrete IO infusions of 0.3 M sucrose (one 30-s infusion every 3 min over 15 min, as above), but half of the rats were preloaded with 3 ml of intraduodenal 0.3 M sucrose during the 4 min just prior to the start of the TR test (0.75 ml/min), and the other half of the rats were preloaded with ID 0.15 M NaCl (same rate and volume). Four rats were discontinued from the experiment during this phase due to intraduodenal catheter or IO cannula failures. Therefore, only the CTA training data from rats who completed the post-CTA test phase were included for those analyses.

CTA training.

After pre-CTA TR testing was complete and all rats had the opportunity to access food and water ad libitum in the home cage (2 days). Then, they were placed on a restricted water-access schedule, in which dH₂O was presented for 15 min each morning and for 30 min ∼5 h later each afternoon for three consecutive days in the home cage. Intakes were measured to the nearest milliliter. Then, for CTA training, 0.3 M sucrose was presented in place of water in the morning session, and, at the end of this session, all rats were injected with 2.0 meq/kg 0.15 M LiCl to induce visceral malaise. Deionized water was presented as usual in the afternoon session. On the following 2 days, rats received dH₂O in the morning and afternoon sessions. Then, on the seventh day, 0.3 M sucrose was presented again in the morning session followed by LiCl for a second training trial. Water bottles were replaced on the home cages ∼5 h after this second training trial, and the rats were given 3 days to access chow and water ad libitum before taste reactivity reacclimation began.

Data analyses.

An experimenter unaware of the infusate conditions of the rats watched each video in slow motion (i.e., frame by frame) and categorized and quantified the oromotor reflexive responses across each 30-s IO infusion. These oromotor reflexes have been previously described in detail (19, 51) and are generally divided into two categories: ingestive and aversive. Ingestive responses comprised tongue protrusions, mouth movements, paw licks, and lateral tongue protrusions. Paw licking counts were derived by multiplying the time spent paw licking in each 30-s video by 6. Aversive responses comprised gapes, chin rubs, forelimb flails, and head shakes. Cumulative ingestive and aversive responses were separately analyzed with mixed repeated-measures ANOVAs that included intraduodenal infusate and time point as factors. Intakes on the two CTA trials were compared with a t-test.

Results

Figure 4 presents ingestive and aversive TR to the IO sucrose deliveries prior to CTA training (pretest). Not surprisingly, rats evinced maximal ingestive responses, with little to no aversive responses, across the entire test session (statistical outcomes in Table 2). These responses were not affected by intraduodenal preload type. Although the results of experiment 1 predict that intraduodenal sucrose ought to enhance ingestive responses to IO sucrose, it is important to iterate that this test was conducted with a single suprathreshold concentration of sucrose and ingestive TR was at or near the response ceiling in this case.

Fig. 4.

Means ± SE total ingestive responses (tongue protrusions, mouth movements, lateral tongue protrusions, and paw licks; A) and corresponding mean ± SE total aversive responses (gapes, chin rubs, forelimb flails, headshakes; B) to each 30-s/0.5 ml IO infusion of 0.3 M sucrose after an ID preload of 0.3 M sucrose (n = 6) vs. 0.15 M NaCl (n = 7) tested after CTA training (post-CTA). For reference, unconditioned levels of ingestive and aversive responding to sucrose on across the first pretest session prior to CTA training are represented in the gray symbols (pretest) in A and B, respectively (n = 7 or 8 rats per ID infusate).

Table 2.

Experiment 2: Pre- and Post-CTA taste reactivity statistical outcomes

	Infusate Effect	Time Effect	Time × Infusate
PreCTA
Ingestive score	F(1,13) = 0.07, P = 0.80	F(5,65) = 1.49, P = 0.21	F(5,65) = 1.15, P = 0.34
Aversive score	F(1,13) = 0.31, P = 0.58	F(5,65) = 0.98, P = 0.44	F(5,65) = 1.30, P = 0.28
PostCTA
Ingestive score	F(1,11) = 4.06, P = 0.07	F(5,55) = 16.94, P < 0.0001	F(5,55) = 2.91, P = 0.02
Aversive score	F(1,11) = 2.55, P = 0.14	F(5,55) = 12.03, P < 0.0001	F(5,55) = 1.03, P = 0.41

Significant effects are denoted by italics.

As expected, all rats significantly reduced intake of the 0.3 M sucrose solution from trial 1 (16.08 ± 4.89 ml) to trial 2 (0.23 ± 0.26 ml) of the home cage CTA training phase [t(12)=11.82, P < 0.000001].

Ingestive and aversive TR to 0.3 M sucrose after CTA training are presented in Fig. 4 (post-CTA, with statistical outcomes in Table 2). In general, the CTA to sucrose was expressed as a gradual reduction in ingestive responses at each successive IO infusion. Aversive responses correspondingly increased, especially across the first three time points. Preloading the intestine with 0.3 M sucrose further augmented the gradual reduction in ingestive responses and increased, albeit nonsignificantly, aversive responses to IO sucrose across the test session.

Experiment 3A: Effects of Intraduodenal Sucrose on Taste-Guided Behaviors to Sucrose, NaCl, Polycose, and Intralipid

Although intraduodenal sucrose specifically enhanced licking for the sucrose solutions and did not appear to impact responding for 0.12 M NaCl in the Unpaired Group of experiment 1, it is important to take into consideration that only a single oral NaCl concentration was examined compared with an array of sucrose solutions. Considering the effect of intraduodenal infusion on responding to oral sucrose varied as a function of concentration, and, as such, varied in intensity and hedonic value, it remains possible that the selected NaCl probe stimulus was less salient on one or both of these dimensions. Moreover, although a nice feature of that design is the fact that the rat is responding to various types of taste solutions within a single test session, permitting the assessment of an ID infusate's effect on different categories of taste stimuli presented in the same “meal” or context, it remains possible that contrast effects among the taste properties of sucrose and NaCl overshadows any effects of the intraduodenal sucrose. Thus, experiment 3 was designed to answer the critical question: does intraduodenal sucrose only impact responding to an oral sugar solution or does it also affect responding to other taste stimuli? Rats were divided into two infusate groups (intraduodenal sucrose vs. intraduodenal NaCl) and then were presented with four separate brief-access taste test sessions. Three concentrations of a single taste stimulus were presented in each test. The taste stimuli included sucrose, NaCl, a representative maltodextrin (Polycose), which is palatable to rodents but qualitatively distinct from sucrose, and a representative fat stimulus, Intralipid, also known to be palatable to rodents (31, 47, 56, 57, 61).

MATERIALS AND METHODS

Subjects.

Naïve male Sprague-Dawley rats (∼10–17 wk of age at the start of the experiment) were maintained under identical conditions described for experiment 1. This experiment was run in three replicate phases. All experimental protocols were approved by and conducted in accordance with the Florida State University Animal Care and Use Committee.

Surgery.

As in experiment 1, each rat was surgically outfitted with an intraduodenal catheter and harness. With the exception of a different anesthetic agent (isoflurane, 5% for induction and ∼2–2.5% for maintenance), all other surgical procedures were identical to that described for experiment 1. Postoperatively, rats were treated with Baytril (2.3 mg/kg sc) and carpofen (5 mg/kg sc) and provided with chow mash (as above). Beginning 48 h after surgery, harnesses were adjusted to accommodate changes in body mass, and intraduodenal infusion lines were flushed with 0.5 ml of sterile dH₂O to maintain patency each day.

Stimuli.

Molar concentrations of reagent grade NaCl (0.03, 0.1, 0.3 M) and sucrose (0.03, 0.1, 0.3 M) (Sigma) and weight/volume (%) concentrations of Polycose (1, 3, 10%) (Ross Laboratories, Columbus, OH) were mixed fresh each day with dH₂O. Twenty percent Intralipid (Sigma) was diluted to the desired concentrations (volume: volume, %) with dH₂O (0.41, 1.3, 4.1%), fresh each day. The sucrose, Polycose, and Intralipid oral test solutions were equicaloric at each concentration.

Brief-access training and testing.

All brief-access training and testing was conducted in the same Davis Rigs described for experiment 1. As in experiment 1, rats were deprived of food and water and acclimated to consuming fluid in a 30-min session in the mornings, followed ∼30 min later by access to chow and a 10-ml dH₂O supplement for 3 h in the home cage in the afternoons. The same basic training regimen was used to acclimate rats first to licking for dH₂O at a stationary sipper tube and then to licking for dH₂O in discrete 10-s trials (with 1-s dH₂O rinses) in the 30-min morning sessions in the Davis Rigs, although because some rats failed to lick or initiate trials, supplementary dH₂O training days were administered on a case-by-case basis. No pretest sessions were administered in this experiment. Rats were then divided into two intraduodenal infusate groups (0.15 M NaCl, 0.3 M sucrose), matched on mean licks per trial, number or trials initiated, and body weights on the final dH₂O trial day. Both intraduodenal infusate groups were given four different brief-access taste tests, the order of which was counterbalanced. A different taste stimulus (three concentration array plus dH₂O) was presented in each test. The order of concentrations (and dH₂O) was randomized across blocks of four 10-s trials. As in experiment 1, intraduodenal infusions (3 ml, 0.75 ml/min) began 2 min prior to the start of the brief-access tests. The four test sessions were separated by 1–3 days, in which dH₂O only was presented in 10-s trials (i.e., no intraduodenal infusates were administered). Rats that failed to initiate enough trials to complete at least one full trial block were given a repeat test session at the end of testing. In total, 6 out of 28 rats were excluded from the experiment due to catheter failure, computer malfunction, or failure to take at least one full trial block on at least one test (including the repeat test).

Data analyses.

Lick scores (described in experiment 1) were used to adjust lick rates to dH₂O. The effects of the two intraduodenal infusates on lick scores were analyzed in separate repeated-measures ANOVAs for each stimulus array. Post hoc t-tests were used where appropriate. The number of trials initiated for each stimulus array were also compared across intraduodenal infusate groups in separate t-tests. The Bonferroni method was used to correct for multiple comparisons.

RESULTS

As in experiment 1, intraduodenal sucrose enhanced licking for sucrose concentrations, though, in this case, the difference just missed statistical significance (see Fig. 5A; Table 3). Likewise, consistent with experiment 1, intraduodenal sucrose failed to selectively enhance licking for NaCl concentrations. In fact, the patterns of lick responses to NaCl were identical across both intraduodenal infusate types. Interestingly, lick scores for both Polycose and Intralipid were significantly augmented by intraduodenal sucrose, relative to intraduodenal NaCl. Once again, intraduodenal infusate did not influence the number of trials initiated; that was true for all four test stimuli.

Fig. 5.

Means ± SE lick responses (standardized against licks to dH₂O) and the corresponding trials initiated for three concentrations of sucrose (A), NaCl (B), Polycose (C), and Intralipid (D) presented in 10-s trials in separate brief-access taste tests. Beginning just 2 min prior to the start of each test session, a small (3 ml; 0.75 ml/min) ID infusion of 0.3 M sucrose (n = 15) or 0.15 M NaCl (n = 12) was administered. The asterisk (*) denotes a Bonferroni-corrected significant difference. For reference, means ± SD licks to dH₂O on each test as a function of ID infusate group are as follows: Sucrose [ID NaCl: 59.32 ± 14.77 and ID Sucrose: 44.93 ± 14.77, t(25) = 2.03, P = 0.053], NaCl [ID NaCl: 65.44 ± 6.42 and ID Sucrose: 62.11 ± 7.70, t(25) = 1.20, P = 0.24], Polycose [ID NaCl: 58.09 ± 12.28 and ID Sucrose: 46.33 ± 16.25, t(25)=2.07, P = 0.05], and Intralipid [ID NaCl: 57.66 ± 11.06 and ID Sucrose: 49.37 ± 17.40, t(25)=1.43, P = 0.16].

Table 3.

Experiments 3A and 3B brief access taste testing statistical outcomes

ANOVAs on Lick Scores			t-Tests on Trials Initiated
ID Effect	Conc. Effect	ID × Conc.	ID Effect
Sucrose (0.03, 0.1, 0.3 M)
F(1,25) = 3.96, P = 0.058	F(2,50) = 1.23, P = 0.30	F(2,50) = 1.66, P = 0.20	t(25) = −0.44, P = 0.66
NaCl (0.03, 0.1, 0.3 M)
F(1,25) = 1.32 P = 0.26	F(2,50) = 11.12, P = 0.0001	F(2,50) = 0.04, P = 0.96	t(25) = −0.19, P = 0.85
Polycose (1, 3, 10 %)
F(1,25) = 7.13, P = 0.01	F(2,50) = 3.79, P = 0.03	F(2,50) = 0.17, P = 0.85	t(25) = 0.13, P = 0.90
Intralipid (0.41, 1.3, 4.1 %)
F(1,25) = 5.30, P = 0.03	F(2,50) = 0.32, P = 0.73	F(2,50) = 1.77, P = 0.18	t(25) = −1.97, P = 0.60
Quinine (0.03, 0.1, 0.3 mM)
F(1,5) = 0.07, P = 0.80	F(2,10) = 117.8, P < 0.00001	F(2,10) = 0.004, P = 0.99	t(5) = 0.21, P = 0.85

Significant effects are denoted by italics.

Experiment 3B: Effects of Intraduodenal Sucrose on Taste-Guided Behaviors to Quinine HCl

Taken together, the results of experiments 1–3A suggest that postoral sugar stimulation does not produce a general positive effect on ingestive behaviors. Rather, its influence appears to depend on the orosensory properties of the ingesta. Specifically, whereas ID sucrose enhanced licking for sucrose and two other stimuli thought to be inherent signals of nutritious outcomes (Polycose and Intralipid), it did not influence licking for a representative salt stimulus. Moreover, when the sucrose stimulus was rendered hedonically negative via CTA in experiments 1 and 2, post-oral sugar stimulation actually enhanced aversive responding and inhibited appetitive behavior.

Given that sucrose, Polycose, and Intralipid are all hedonically positive stimuli, and, by comparison, NaCl and aversively conditioned sucrose (i.e., CTA) presumably fall into the less hedonically positive and/or aversive domains, it is possible that ID sucrose simply reinforces the response pattern dictated by the taste input, in effect making positive substances more positive and negative substances more negative, respectively. The other possibility is that the interactions are somewhat sensory-specific, such that sugar stimulation has privileged access to interact with certain types of taste stimuli more than others. Accordingly, perhaps intraduodenal sucrose makes CTA sucrose more aversive because it engages a common central sensory representation as oral sucrose, but does not affect licking for NaCl because those sensory qualities are more segregated in high-order circuitries. Provided this lack of effect on NaCl, it is of interest to test the above-mentioned specificity hypothesis with another taste quality, “bitter.” In experiment 3B, taste-guided responses to a “bitter” ligand, quinine, were measured in the brief-access taste test, following intraduodenal infusions of sucrose vs. the NaCl control. This was done to determine whether the positive/rewarding effects of postoral sugar stimulation would mitigate the aversiveness of quinine.

MATERIALS AND METHODS

Subjects.

Subjects were a cohort of male rats that were previously used in other intraduodenal infusion experiments, including some from experiment 3A, but were naïve to quinine prior to the start of this experiment.

Brief-access testing for quinine.

For this experiment, a within-subjects design was used, such that all rats were tested for their responses to three concentrations of quinine hydrochloride (0.03, 0.1, 0.3 mM; Sigma Aldrich, St. Louis, MO; and dH₂O) in two separate brief-access test sessions, employing the same trial structure and order of randomized stimulus presentation described above for experiment 1. Half of the rats received an intraduodenal infusion of 0.15 M NaCl, beginning 2 min before the start of one test session (3 ml, 0.75 ml/min) and then received an equivalent rate and volume intraduodenal infusion 0.3 M sucrose, beginning 2 min before the start of the second test session. The order of test (intraduodenal NaCl vs. intraduodenal sucrose) was counterbalanced across rats. Rats were maintained on the same restricted food and water access schedule described for experiment 3A. Two sessions with dH₂O only in 10-s trials (with 1-s rinses) interceded the two test sessions.

Data analyses.

Lick scores for each quinine concentration were calculated by standardizing against mean lick responses for dH₂O in the same test (see above). Then, the lick scores for each of the three concentrations of quinine following an ID infusion of NaCl vs. an ID infusion of sucrose were analyzed in a repeated-measures ANOVA. A t-test was used to compare intraduodenal effects on trial initiation.

RESULTS

Clearly, lick scores for quinine decreased as the concentration increased, as would be expected, but lick scores did not further vary as a function of intraduodenal infusate type (Fig. 6, Table 3). Nor did ID sucrose influence the number of trials initiated. It is important to note that, in this case, all rats completed at least one full trial block, and, on average, rats completed five full-trial blocks.

Fig. 6.

A: means ± SE lick responses (standardized against licks to dH₂O) on the quinine brief-access taste tests, as a function of quinine concentration and ID infusate. Two minutes before the start of the tests, rats were infused with either ID sucrose or ID NaCl (test order counterbalanced) (n = 6). B: corresponding means ± SE trials initiated on each of these test sessions. For reference, means ± SD licks to dH₂O on the ID NaCl test was 64.48 ± 4.79 and on the ID sucrose test was 62.99 ± 7.30 [t(5) = 0.64, P = 0.55].

GENERAL DISCUSSION

Post-oral sugar sensing has been recently shown to rapidly stimulate ingestion (e.g., 3, 28, 62, 63). This early-phase positive feedback is thought to serve a disambiguating role, whereby the immediate detection of sugar in the GI tract reinforces continued consumption of that food. A critical question then was whether detection of sugars in the GI tract can promote consumption, irrespective of the sensory and/or hedonic information being concurrently derived from signals arising from the oral cavity. Here, we found that sucrose delivered directly to the intestine selectively enhanced licking for oral sucrose, Polycose, and Intralipid solutions, but had no effect on responding to a representative salt stimulus (NaCl) nor to an aversive bitter compound (quinine) in unpaired or naïve rats (experiments 1 and 3). Interestingly, when the sucrose was rendered aversive (via CTA), an equivalent intestinal sucrose stimulus had contrasting effects on behavior. Under these conditions, intraduodenal sucrose rapidly inhibited appetitive responding and ingestive oromotor reactions (experiments 1 and 2, respectively). Taken together, the results of the present experiments demonstrate that post-oral sugar feedback does not universally facilitate taste-guided behavior, but rather that the efficacy and directionality of its influence depend on the taste properties of the substance being consumed.

To date, the rapid-onset, early-phase stimulatory actions of post-oral nutrients have largely been studied in a single experimental context, in which the subject licks for a single flavored sweetened solution (i.e., with Na-saccharin or sucralose), while being intragastrically infused with water or nutrient (or nutrient substitute) (e.g., 3, 28, 62, 63). Considering the orosensory properties of the solution at the sipper spout are critical determinants of ingestive motivation, with “sweet” solutions being particularly potent drivers, it was important to determine whether the postoral stimulatory effects of nutrients were dependent on the type of taste input concurrently received. Using a modified procedure that allowed us to present various taste solutions to the animal and measure (with millisecond resolution) licking responses, we found that, indeed, the nature of the orosensory stimulus is an important factor in this phenomenon. With respect to the behaviors of just the unpaired rats, intraduodenal sucrose enhanced preferential licking for sucrose, as well as for two qualitatively distinct stimuli, Polycose and Intralipid, but did not impact responding for NaCl and quinine. The basis for this breakdown among taste stimuli is unclear. On the one hand, sucrose, Polycose, and Intralipid all contain macronutrients with hedonically positive taste properties, whereas quinine is a nonnutritive, hedonically negative stimulus (e.g., 19, 47, 52, 56, 61). The lack of an intraduodenal sugar effect on NaCl is difficult to reconcile in terms of simple hedonics though. While lower concentrations of NaCl are thought to have an inherently acceptable taste, high concentrations (e.g., 0.3 M) are not, except under conditions of sodium depletion (e.g., 5). Thus, it is reasonable to speculate that in the sodium-replete state, there may be sensory-affective mechanisms in place to prevent the overconsumption of NaCl, even if it is paired with positive intraduodenal outcomes. Taken together, these data argue that intraduodenal sugar is not universally effective at motivating ingestion. Taste plays an important role, but the underlying organization of these sensory/reward interactions is still unknown.

CTA was used in experiments 1 and 2 to explicitly dissociate the sensory qualities of sucrose from its inherent hedonic value for some rats. Under these conditions, ID sucrose abruptly inhibited trial initiation in the brief-access taste test, and, when a second group of rats was later forced to intraorally sample sucrose in the serial taste reactivity test, intraduodenal sucrose preloads rapidly reduced the number of ingestive reactions elicited. Although conditioned sucrose aversion and unconditioned quinine aversion differ on a number of dimensions (e.g., taste quality, intensity, training history), all of which may be contributing factors, the fact that intraduodenal sucrose influenced appetitive behavior on the sucrose test, but had no such effect on the quinine test favors the view that there is a sensory-specific aspect to early-phase oral-post-oral integration. Moreover, the fact that intraduodenal sucrose actually provoked avoidance and rejection behaviors to oral sucrose demonstrates that intraduodenal sugar is not a simple “go” signal early in the meal; rather, it appears as though intraduodenal sugar can either selectively enhance or suppress ingestion of sucrose, depending on the current value or biological significance of the stimulus. Yet, precisely how the intraduodenal sucrose signal accesses such sensory- or context-specific information is unknown; a few possibilities are discussed, in turn, below.

Perhaps the simplest account assumes that ID sucrose does not undergo any revision itself, but instead, places gain on the concurrently incoming taste signal from the tongue, in effect making the taste CS more readily detectable or salient. Note that this type of intraduodenal influence could just as well account for the patterns observed in the unconditioned or naive state, assuming that sugar signals arising from the gut have privileged access to certain types of sensory inputs arising from the oral cavity. There is some electrophysiological evidence to suggest that vagal/visceral afferent signals modulate and/or converge upon at least some gustatory neurons (e.g., 4, 22, 24). Furthermore, while some gustatory neurons may be broadly tuned to various classes of tastants, others may be more selectively responsive to sucrose or NaCl or quinine (e.g., 1, 8, 17, 23, 29, 30, 33, 58, 59). Thus, perhaps what makes quinine different from oral (aversive) sucrose in the present context is that the sucrose signals ascending from the gut are channeled into neurons (or even a subset of those neurons) that respond best to sucrose (or neurons that respond best to Intralipid or Polycose) and not to neurons that respond best to quinine. Moreover, if intraduodenal sucrose signals were to simply amplify activity in these specific subsets of neurons, then this would presumably be sufficient to render the incoming taste signal more intense and more palatable in the naïve (or unpaired) animal and more intense, but less palatable in the conditioned animal. Accordingly, it would be of interest now to see whether post-oral sugar stimulation affects the activity of taste-responsive neurons in such a chemospecific manner, both in the naive and conditioned animal.

While that certainly has the appeal of parsimony, it assumes that the intraduodenal sucrose signal remains the same under both of those states, when, in fact, there is reason to believe that postoral signals are also modulated by experience. For instance, Schier et al. (39, 40) showed that rats can associate the chemosensory properties of a stimulus (i.e., denatonium benzoate or sucrose) delivered directly into the duodenum with LiCl and can later use that post-oral stimulus to curb ongoing ingestion. Thus, perhaps, in addition, to learning about the orosensory cues associated with sucrose consumption, the rats in experiments 1 and 2 paired its post-oral sensory attributes with the LiCl consequences. Then, at testing, both intraduodenal sucrose and oral sucrose elicited those respective negatively charged associations, combining to make a more robust response to sucrose than either input alone would elicit. A second, but not unrelated possibility, is that intraduodenal sucrose and oral sucrose share some chemosensory features that permits both to access a common central representation and provoke a corresponding response. Tracy et al. (54, 55) discovered a phenomenon consistent with the latter alternative. In those studies, rats that had experience with a particular intragastric/intraduodenal nutrient (e.g., maltodextrin) paired with LiCl were later able to avoid that same nutrient in a two-bottle preference test. Future experiments that expand the range of oral and intraduodenal test stimuli used and that involve the conditioning of aversions to oral vs. post-oral stimulus features separately will be critical for distinguishing which of these putative integrative processes underlie these phenomena. For example, it would be revealing to determine following the development of a CTA to sucrose, whether ID sucrose affects licking for another aversive substance (e.g., quinine) or whether its effects are specific to stimuli that have common sensory features (i.e., oral sucrose, Na-saccharin, but not Intralipid, NaCl, or quinine).

We have focused on a single representative post-oral sugar stimulus in the present studies, but it should be noted that at least in other preparations, the post-oral reinforcing efficacy of sugar appears to be related to its glucose content (3, 42, 45, 48, 49, 63). Other sugars (e.g., fructose) fail to stimulate early-phase ingestion and condition weaker flavor preferences (42, 45, 63). Moreover, glucose-containing sugars (or analogs) are not the only nutrients that elicit the early-phase response; some fats do as well (e.g., 62). Thus, it would be of interest to examine whether those post-oral sensory events, which are presumably quite different from the ones elicited by sucrose, affect taste-specific behaviors as well. Electrophysiological evidence suggests that intraduodenal Intralipid directly modifies the sucrose-elicited responses in sucrose-best neurons of the parabrachial nucleus, but has only a weak effect, if any, on NaCl-elicited responses in sodium-best neurons (22). It is at least tempting to speculate then that intraduodenal Intralipid would enhance lick responses to sucrose, but would likewise not affect licking for NaCl.

Another factor that likely plays a role in these interactions is the deprivation state of the animal. To motivate at least moderate levels of ingestive behavior and help to ensure the GI tracts were comparably free of food and fluid at the time of testing, we maintained all of the rats on a partial food and water access schedule. Both food and water deprivation state are known to differentially affect the responsivity of the gustatory system and temper taste-guided behaviors, in some cases in a chemospecific manner (e.g., 6, 7, 16, 18, 22, 26, 38, 41, 64). In the brief-access taste tests conducted here, intraduodenal sucrose rapidly and selectively promoted or maintained high levels of licking for sucrose solutions over other stimuli in the array (e.g., dH₂O, 0.12 M NaCl). Thus, it is important to consider whether this was due to a post-oral sugar stimulation effect that enhanced the salience (hedonic or otherwise) of sucrose (relative to the other stimuli) or whether this was due to a reduction in the deprivation state that selectively diminished the motivational potency of the other stimuli. Importantly, in addition to being matched on volume, the intraduodenal infusates were specifically chosen to match for preabsorptive osmolarity; therefore, it seems unlikely that the rapid early-phase differences that we observed were due to differences in fluid repletion per se, although we cannot entirely rule that out. Moreover, if the small intraduodenal sucrose load was simply enhancing satiation, then that ought to manifest in a reduction in trial initiation and/or in consummatory responding on other tests as well (i.e., including to quinine, NaCl, or dH₂O) in the naïve or unpaired animals. Yet no such suppressive effects were observed in these other tests. Nor did the same preload reduce ingestive responding for 0.3 M sucrose at any point across the serial TR pretest. Thus, overall, the present results favor the view that these intraduodenal sucrose loads exerted a selective effect on some taste stimuli and not others. Whether the same chemospecific interactions would be observed under strict water or food deprivation or in an ad libitum state remains to be determined.

On the face of it, the dependency of the post-orally mediated early-phase response on the orosensory properties of the substance being consumed is somewhat surprising, considering a number of recent studies that suggest post-oral signals alone are “sufficient” to reinforce operant behaviors (including dry licking) and guide nutrient selection, even in the absence of taste input altogether (12, 15, 32, 37, 49, 53; though see Refs. 9 and 25). Moreover, it is well known that post-oral feedback plays a large role in amending appetites and flavor preferences in lasting ways (see Refs. 43 and 44 for reviews). Collectively, these phenomena have fostered the compelling view that foods and fluids engage dual sensory/reward pathways—one via taste input and the other via post-oral input—which, although can be integrated (as with conditioning), can also work independently to best accommodate adequate nutrient ingestion (and prevent illness). To be clear, the present data do not necessarily refute this framework, but are indicative of other types of processing, whereby the two systems engage a common pathway; this may be especially relevant to online, within-meal ingestive decisions.

Accordingly, it is important to consider that almost all of the ingestive decisions in a given meal are made before most of the food is even absorbed. Thus, the ability for an animal to predict the consequences of a particular food as fully and as early as possible is paramount for adequate nutrition and survival. The existence of oral, post-oral GI, and metabolic sensors that separately track foods and their consequences coupled with systems that can integrate information to enhance the predictive value of the earliest signal in that cascade (e.g., taste) would seem advantageous in this regard. Interestingly, there are hints in the literature that hedonic value for some tastes (e.g., sweet) are more readily reinforced by caloric outcomes than are others (e.g., bitter), suggesting that certain oral signals are more readily integrated with certain post-oral outcomes (e.g., 27, 35, 36). That said, when the orosensory signals are altogether equivocal, then it is important that the post-oral consequences can attach to other types of cues as well (e.g., odors, spatial locations). Arguably then, the underlying circuitry may be arranged in such a way to more readily permit certain sensory-specific links, while forestalling others, and there is some flux along these lines that depend on the context and/or stimuli available. For instance, it may be that the post-oral sensory signals that are elicited within the meal are more limited with respect to their capacity to influence taste-guided behaviors to the inherent or already established sensory and hedonic programs, while the longer-term cues can be integrated with taste and other cues in a broader manner.

Perspectives and Significance

Considering overweightness and obesity are associated with the overconsumption of foods and fluids high in refined sugars and fats, coupled with the fact that the most effective treatment strategies known to date involve major revisions of the post-oral GI tract (bariatric surgery), underscores the need to more fully understand how oral and post-oral sensory signals are functionally and centrally organized. Here, we used experimental paradigms that allowed us to extend the literature by examining the effects of intraduodenal sugar on immediate behavioral responding to taste stimuli that varied in hedonic value, nutritive value, and/or learned value. As such, the present experiments highlight a novel and arguably significant dimension of oral–post-oral sensory integration in the early phase of ingestion. Although some foods and fluids may be capable of eliciting particular behavioral repertoires from stimulating only an oral or post-oral site, there is clearly a level of sensory-specific cross-talk among these systems throughout the meal and later. While studying these as separate systems has some obvious advantages in terms of determining their distinctive contributions to the meal and underlying neurocircuitries, this approach in isolation runs the risk of key domains of central and functional integrative processing in the controls of food and fluid intake going unnoticed. Indeed, much is left to be interrogated along these lines.

GRANTS

This work was supported by grants from National Institute on Deafness and Other Communication Disorders to A. C. Spector (R01-DC-009821) and L. A. Schier (F32-DC-013494). Portions of this work were presented at the 25th Annual meeting for the Association of Chemoreception Sciences (Huntington Beach, CA, 2013) and the 21st Annual meeting for the Society for the Study of Ingestive Behavior (New Orleans, LA, 2013).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

L.A.S. and A.C.S. conception and design of research; L.A.S. performed experiments; L.A.S. analyzed data; L.A.S. and A.C.S. interpreted results of experiments; L.A.S. prepared figures; L.A.S. drafted manuscript; L.A.S. and A.C.S. edited and revised manuscript; L.A.S. and A.C.S. approved final version of manuscript.