Sensory-specific appetition: Postingestive detection of glucose rapidly promotes continued consumption of a recently encountered flavor

Highlights

• Intragastric glucose infused during a meal increased rats’ licking and intake.

• This effect was produced by glucose but not fructose nor maltodextrin infusion.

• Intake stimulation was abolished if the flavor was switched after glucose infusion.

• Within-meal feedback from gut nutrients can alter meal size via flavor evaluation.

Abstract

It is generally thought that macronutrients stimulate intake when sensed in the mouth (e.g., sweet taste) but as food enters the GI tract its effects become inhibitory, triggering satiation processes leading to meal termination. Here we report experiments extending recent work (see Zukerman et al., 2011 [1]) showing that under some circumstances nutrients sensed in the gut produce a positive feedback effect, immediately promoting continued intake. In one experiment, rats with intragastric (IG) catheters were accustomed to consuming novel flavors in saccharin daily while receiving water infused IG (5 ml/15 min). The very first time glucose (16% w/w) was infused IG instead of water, intake accelerated within 6 min of infusion onset and total intake increased 29% over baseline. Experiment 2 replicated this stimulatory effect with glucose infusion but not fructose nor maltodextrin. Experiment 3 showed that the immediate intake stimulation is specific to the flavor accompanying the glucose infusion. Rats were accustomed to flavored saccharin being removed and replaced with the same or a different flavor. When glucose infusion accompanied the first bottle, intake from the second bottle was stimulated only when it contained the same flavor, not when the flavor switched. Thus we confirm not only that glucose sensed postingestively can have a rapid, positive feedback effect (‘appetition’ as opposed to ‘satiation’) but that it is sensory-specific, promoting continued intake of a recently encountered flavor. This sensory-specific motivation may represent an additional psychobiological influence on meal size, and further, has implications for the mechanisms of learned flavor-nutrient associations.

Introduction

A basic heuristic principle that has emerged from a century of behavioral neuroscience research on appetite is that the meal – the “fundamental behavioral unit of eating” [2] – is largely determined by opposition between two main influences. Food palatability stimulates eating rate, whereas satiation factors that arise during eating suppress it and bring the meal to an end. Of course this basic algebra is extended and elaborated by layers of cognitive, environmental, and social influence, but at its core it is generally thought to be a system whereby nutrients sensed in the mouth promote intake until nutrients sensed in the gut inhibit intake through negative feedback. In this paper we present experimental evidence from our lab, which supports and extends recent work from others [1], showing that this model is incomplete in including only negative-feedback immediate effects of gut nutrients. This work describes a rapidly-onsetting, positive feedback response to nutrients sensed in the gut within a meal, which stimulates ongoing intake instead of inhibiting it.

Regarding the fundamental influence of oral hedonics as a driver of intake, humans, rodents, and other omnivorous mammals clearly like the tastes of many macronutrients. The sweetness of sugars, for some species a starchy taste of complex carbohydrates, and the taste/texture of fats, are strongly hedonically positive (reviews [3], [4], [5], [6], [7]) The umami taste of glutamate and some ribonucleotides, which serves as a marker of protein content, can enhance the palatability of other food constituents, and may be perceived by some animals as palatable on its own [8], [9], [10]. These responses to a narrow category of biologically relevant stimuli serve to make nutrient-dense foods attractive and to promote intake in proportion to macronutrient content.

Were it only for these positive responses to nutrient taste, more calorically dense foods would typically be eaten in larger amounts. But this is usually not the case, especially on initial exposures, since nutrient-dense foods are also more “filling.” That is, once nutrients are swallowed they begin to activate a number of satiation processes. Although the palatability of nutrient taste does passively wane within a meal due to habituation independently of postingestive feedback [11], [12], to a considerable extent ongoing eating becomes actively inhibited and eventually ceases due to the afferent detection of food accumulating in the gut and the products of its digestion and absorption giving rise to negative feedback signals [2], [13].

The entire post-oral extent of the gastrointestinal tract is rich in receptors that transduce physical and chemical properties of ingested food. In some cases these are structurally similar to taste receptors in the mouth, possessing some of the same transduction and signaling pathways [14], [15]. Mechano- and chemoreception of food in the gut initiates a cascade of local and systemic responses, influencing not only a variety of functions including gut motility, efficient digestion, and nutrient partitioning, but also psychological effects on the enjoyment of eating and motivation to continue. Although they are distributed and complex, the direct effects arising from nutrient stimulation of post-oral sensory pathways are thought to be exclusively inhibitory [16], [17], [18], [19]. In fact the only known endogenous gut peptide that stimulates intake is secreted during fasting and is inhibited by ingested nutrients [20].

An important extension to this simple negative feedback model involves the well-documented effects of postingestive nutrient detection on future meal size, mediated by learning. That is, even if gut nutrient action is entirely inhibitory during a meal, animals learn about a food's postingestive nutritive actions and may preferentially select that food and eat more the next time around [21], [22]. This learning is sensory-specific, in that the oral/cephalic sensations (the particular tastes, odors, flavors, and perhaps for some species visual appearance) are remembered in association with the postingestive nutritional consequences detected in the gut. Through Pavlovian conditioning, these oral/cephalic sensations then serve as cues that elicit conditioned increases in preference and intake on subsequent encounters, so we refer to this learning as “flavor-nutrient” conditioning.

In this way animals need not rely only on the inborn repertoire of basic hedonic responses to macronutrients themselves, and can more adeptly forage in a complex environment containing a diversity of foods that could not be anticipated by genetic endowment alone, and which constantly vary in quality and availability. Learning to respond to flavors, odors and other food properties that are reliable cues for nutrient type and density becomes a predominant influence on an individual's food preferences and meal patterns. Thus the postingestive effects of nutrients come to exert a positive influence on meal size, in contrast to their immediate satiating influences. Sclafani has coined the term “appetition” (in contrast to satiation) for this category of intake-promoting effects of macronutrients sensed post-orally [23].

Much work has attempted to identify the physiological identity of the signal(s) involved in these rewarding, appetition effects, since it is still undetermined what specific receptor sites or afferent pathways are involved. One approach has been to experimentally manipulate different pathways through deafferentation, selective nutrient infusion sites, or with receptor knockout models (for overviews, see [24], [25]). In sum, this work has highlighted the input of pre-absorptive nutrient detection in the proximal intestines, but another take is that it is rather difficult to completely block flavor-nutrient conditioning, suggesting that perhaps multiple, redundant afferent pathways are involved.

In our lab we have been attempting a different strategy, using behavioral methodology to investigate when during or after a meal rats show evidence of detecting nutrients and experiencing different effects on motivation. By identifying the time course of the different psychological effects of post-oral nutrient detection, we may gain some clues of the underlying physiological events.

One informative experimental paradigm we've employed (e.g., [26], [27]) involves modifying the standard flavor-nutrient conditioning protocol. The standard protocol has rats consume non-nutritive, flavored solutions in different sessions, with distinct flavors either paired or unpaired with intragastric (IG) nutrient infusion, to study how intake is altered by the flavor-nutrient association. But in our modified protocol, instead of each training session (i.e., ‘meal’) consisting of one flavor, there are two flavors in consecutive sequence, with particular flavors consumed only in the beginning half of a meal and others only in the latter half. For a meal that provides appetitive postingestive stimulation with glucose, which flavor would the rats learn to prefer? If the relevant postingestive signal accrues relatively slowly and is experienced late in the meal or after, the last flavor should become most strongly associated, since Pavlovian temporal contiguity and retroactive interference effects should minimize or prevent learning about the first flavor. But for meals with IG glucose, we found this was not the case. Instead rats did learn a strong preference for the flavor encountered in the first several minutes of the meal, and in subsequent tests they only expressed that preference when hungry, like they were when they encountered it in training [26].

These experiments have led us to argue that there is an unconditioned (in the Pavlovian sense) effect of glucose detected post-orally that arises within minutes of meal initiation to support appetitive learning. This evidence for a rapid-onset US signal is consistent with a variety of evidence that the US supporting flavor-nutrient conditioning is psychologically and physiologically dissociable from the nutrient's satiating effects that terminate the meal [28], instead acting as a separate, positive influence on intake. Based on this principle, we sought to investigate the possibility that appetition does not only act to increase subsequent intake the next time the cue flavor is encountered, but may also act in an immediate, unconditioned fashion to stimulate intake within the meal as soon as beneficial nutritive effects are first detected. This proposal differs from the common view that the immediate, direct effects of nutrients in the gut are only satiating. This possibility is further suggested by the fact that strong preference for a flavor can be acquired in only a single flavor-nutrient pairing [29], [30].

Such an unconditioned appetition effect has recently been reported in mouse behavior by Sclafani and co-workers [1]. In their model, IG water self-infusion was yoked to licking of flavored saccharin in an electronic esophagus preparation. Then in three consecutive 1-h sessions, a flavor was accompanied by IG glucose instead of water. As would be expected in flavor-nutrient conditioning, licking rate and total intake increased across the three flavor + glucose pairings. But the critical finding was that IG glucose increased intake within the first session, consistent with a rapid, unconditioned appetition response to gut glucose sensing.

Our experiments sought first to demonstrate this immediate appetition effect in rats. If rats rapidly detect the unconditioned appetitive effects of glucose in the gut within minutes of meal initiation, would they, upon first encounter of a novel flavor, accelerate licking and consume more of a flavor accompanied by IG glucose than when the flavor is accompanied by IG water?

But a second goal was to extend this finding by investigating its specificity. The positive effect occurring during the first flavor-nutrient pairing could reflect a nonspecific activation or arousal stemming from a rapid rise in blood glucose. But if it genuinely reflects the initial formation of a flavor-nutrient association, we could expect it to be flavor-specific, as if rapid detection of a food's nutritive consequences feeds back onto flavor evaluation systems to promote continued eating of that food, not just eating in general.

Our rationale for this question came from a curious observation in experiments with the consecutive flavor paradigm described previously. When the first flavor in a meal paired with glucose infusion was removed and replaced with the second flavor, rats unexpectedly tended to suppress intake of the second flavor [26]. This occurred even though the second flavor was also accompanied by IG glucose and rats could continue to self-infuse more glucose simply by continuing to consume. This pattern was not seen for a flavor switch in IG water infusion sessions, nor did this suppression indicate learned rejection of the second flavor. Instead it appeared as if the rapid-onset postingestive signal was already detected and ‘attributed’ to (i.e., associated with) the first flavor. That remains speculative since the experiment was not designed to examine this unexpected behavior and therefore didn't include all the control conditions needed for conclusive interpretation. Nonetheless, it did call our attention to the potential motivational significance of rapid glucose detection. Therefore we investigate in the current experiments, first, whether detecting postingestive glucose rapidly stimulates intake within a meal, and second, if that effect is specific to the flavor that accompanied onset of the glucose.