Comments by Harvey Grill (December 2014)

When reporting that an experimental manipulation reduces rodent food intake, scientists are obliged to consider alternate interpretations for the observed behavioral changes.  The development of taste reactivity arose from Ralph Norgren’s and my desire to develop a methodology that would provide explanations for feeding effects that had previously been refractory to analysis because [1] treatment-induced deficits in appetitive behavior (e.g. rodents did not approach or sample food) interfered with food intake measurement and [2] assessments of rodent food hedonics were unavailable.  The experimenter-initiated oral infusion of nutrients used in taste reactivity tests and later in the intra-oral intake test eliminates an appetitive behavior requirement and measures only ingestive consummatory behavior.  These approaches also provided a behavioral metric for assessing hedonic responses to food in rodents as discussed in detail below.

Prior to the development of the taste reactivity test, the terms avoidance and aversion were offered as potential explanations for treatment-induced changes in food intake.  Unfortunately, however, a rationale for selecting one over the other lacked an empirical basis.  The source of the term “aversion” in rodent models of energy balance research derives, for most of us, from the phenomena termed conditioned taste aversion and made popular by John Garcia’s work.  Despite the fact that the term “aversion” is part of the name of the paradigm, the test applied to assess aversion simply measures the reduced intake of the taste, food or flavor associated with the negative consequences of a LiCl injection. A reduction in intake however provides no justification for selecting an aversion explanation over the more appropriate avoidance explanation.  The work of Pelchat and Rozin (1982) provides a thoughtful analysis of this distinction by drawing particular attention to the anatomical loci of the different negative consequences (symptoms) arising from consuming food under different natural or experimental conditions.  These authors reported that when nausea and vomiting followed food consumption (described as an upper GI locus of effect), people both reduced their intake of that food as well as described a negative change in their hedonic response to that food.  By contrast, when abdominal cramping, gas or diarrhea symptoms (i.e. symptoms associated with the lower GI tract such as those observed with lactose malabsorption) followed food consumption, humans ate less but did not alter their hedonic response to the food and continued to describe the symptom-associated food as pleasant, tasty or attractive.  In an attempt to help distinguish these two explanations for reduced intake responding Pelchat and Rozin used the term “danger” to explain the reduced feeding without hedonic change as that arising from adjusting intake to reduce the magnitude of negative lower GI symptoms and “distaste” for the reduced feeding related to a negative hedonic response to the food associated with nausea/vomiting. 

Assessing food pleasantness and distaste in humans is commonly accomplished by verbal metrics such as the questionnaire approach employed by Pelchat and Rozin (1982).  Non-verbal, orofacial motor reactions had also been used to assess human hedonic responses to food as described by Darwin (1872, The Expression of the Emotions in Man and Animals) for disgust reactions and by others, such as Steiner (1973) for distinguishing reactions to different tastes.  In the analysis of animal behavior, the biologist Wallace Craig (1918) defined an aversive reaction as an innately determined reaction adapted to getting rid of a disturbing stimulus.  This concept directly applies to assessing negative hedonic reactions to food in rodents.  From this perspective, an aversive response to food would be one that resulted in removing the “offending” food from the mouth.  As a postdoc working with Ralph Norgren at the Rockefeller University in the mid-70s, I set out to develop an assessment of behavioral reactions to food stimuli that could be used as an alternative to amount consumed to examine ingestive consummatory responses of rats that had been subjected to extensive brain ablation and disconnection as those rats would no longer seek out or approach food.  By implanting a tube into the front of the oral cavity and delivering tiny volumes of stimuli such as sucrose, NaCl, HCl and quinine, the elicited orofacial motor reactions could then be videotaped and analyzed frame-by-frame.  In intact control rats, this taste reactivity analysis revealed two highly stereotyped, basic patterns of taste-stimulated responses, each made up of distinct oral-motor reactions.   Quinine concentrations above a threshold elicited a response that was radically different than the responses rats emitted in response to sugar concentrations (and also to a range of salt and acid solutions).  Quinine taste reactivity fit Craig’s definition of an “aversive response” as all of the elicited response elements (defined and described in the accompanying paper) resulted in the rejection of taste stimulus from the rat’s mouth.  Adding a interesting twist to these taste stimulus-driven behavior patterns just described we found (abstract presented at the second annual SFN meeting, 1975), that a sucrose stimulus would trigger an aversive reaction to in intact rats that had previously associated sucrose with LiCl injection.  The oral-motor reactions triggered by LiCl-paired sugar no longer resembled the rhythmic oral lingual reactions (emulating licking) and swallowing seen prior to pairing, but now the reaction to the CS+ was identical to that evoked by novel quinine indicating that the taste-LiCl pairing produced an alteration in neural processing resulting in a dramatic change in stimulus-response coupling (sucrose-rejection).

To probe the aversion-avoidance, danger-distaste distinctions in rats, I collaborated with Pelchat and Rozin.  In Pechat et al. (1983) we used taste reactivity testing to provide a hedonic assessment and an empirical basis for a applying the term food aversion, and a two-bottle intake tests to assess food avoidance.  In separate groups of rats, novel sugar taste was associated with one of three different negative consequences – lactose gavage to trigger malabsorption, food shock for peripheral discomfort, or LiCl injection to induce upper GI symptoms of malaise.  Irrespective of the negative consequence condition, all rats avoided the sugar relative to control rats, and the magnitude of the sugar avoidance was comparable across conditions.  By contrast, only rats that received sugar-LiCl pairing showed aversive taste reactivity responses.  The rats who received sugar paired with either lactose or foot shock continued to express sugar-elicited taste reactivity associated with licking, swallowing and ingestion observed in naïve rats.  This outcome in rats mirrors the results in humans described by Pelchat and Rozin (1982) and establishes an approach to assessing the distinction between avoidance and aversion in rodents.  Work with my former trainees Kent Berridge, Paul Breslin, and Alan Spector exploited the taste reactivity metric to probe the development and persistence of taste aversion in rats following taste-LiCl association, and to make inferences about the utility of different models of palatability processing (one- vs two- dimensional models – see Kent Berridge’s Ingestive Classic Commentary on PT Young). 

Having first defined the taste reactivity methodology in intact rats, Norgren and I next addressed whether the taste reactivity responses of chronic decerebrate rats that lacked forebrain neural processing and hindbrain-forebrain communications were similar to intact rats.  Despite the absence of appetitive behavior and other impairments related to the forebrain surgical disconnection, we found that the taste reactivity responses of these rats mirrored that of intact control rats.  These data support the conclusion that the hindbrain neural circuits are sufficient for mediating taste-driven oral-motor response production.  The taste reactivity of thalamic rats, which lack all structures dorsal and anterior to the thalamus, was also investigated.  Thalamic rats, like decerebrate rats, are aphagic and adipsic, however the taste reactivity reactions of the two neurologic preparations differed dramatically.  All taste stimuli elicited a quinine-like aversive reaction/rejection sequence in the thalamic rats.  Decerebrate rats with isolated caudal brain stems display both acceptance and rejection responses while  all tastes provoke aversive reactions in thalamic rats. This suggests that control of response valance can be influenced by transmission from rostral brain structures to hindbrain response output controllers. Which neurons participate in this descending control and what  guides their modulatory output remains to be explored further.  An example is instructive. Rats with bilateral lesions of the lateral hypothalamus examined during the post lesion aphagic phase display rejection of sugar stimuli if tested after gavage  feeding maintenance but normal hedonic response to sugar if examined when food deprived.(Fluharty and Grill 1981).

Despite its applicability for dissociating aversion vs. avoidance as explanations for reduced feeding, taste reactivity is not a widely used methodology.  Unfortunately the term food or taste aversion still periodically appears in the literature as an explanation for a treatment that reduces intake without also employing taste reactivity to substantiate this assertion.  As research on food intake has expanded from a focus on satiation and “homeostatic feeding” to investigate the neural basis of processes such as reward and motivation, it will be useful to examine how assessments of food hedonics in animal models are applied to the neural analysis of appetitive aspects of ingestive behavior.

References

Berridge K.C.  Grill H.J. Isohedonic tastes support a two-dimensional hypothesis of palatability. Appetite. 1984 Sep;5(3):221-31.
Craig, W.  Appetites and aversions as constituents of instincts.  Biological Bulletin. 1918, 34(2):91-107
Fluharty, S.J.,  Grill H.J. Taste reactivity in lateral hypothalamic lesion end rats: effects of deprivation and tube feeding. 1981 Neuroscience Abstracts 6:28.
Grill H.J., Berridge K.C.  Taste reactivity as a measure of the neural control of palatability, Progress in Psychobiology and Physiological Psychology. 1985 11:1-64
Grill H.J., Norgren R. The taste reactivity test. I. Mimetic responses to gustatory stimuli in neurologically normal rats. Brain Res. 1978 Mar 24;143(2):263-79.
Grill H.J., Norgren R. The taste reactivity test. II. Mimetic responses to gustatory stimuli in chronic thalamic and chronic decerebrate rats. Brain Res. 1978 Mar 24;143(2):281-97
Pelchat M.L., Grill H.J., Rozin P., Jacobs J.  Quality of acquired responses to tastes by Rattus norvegicus depends on type of associated discomfort. J Comp Psychol. 1983 Jun;97(2):140-53.
Pelchat M.L., Rozin P. The special role of nausea in the acquisition of food dislikes by humans. Appetite. 1982 Dec;3(4):341-51.
Spector A.C., Breslin P., Grill H.J. Taste reactivity as a dependent measure of the rapid formation of conditioned taste aversion: a tool for the neural analysis of taste-visceral associations. Behav Neurosci. 1988 Dec;102(6):942-52.
Steiner, J.E. The gustofacial response: observation on normal and anencephalic newborn infants. Symp Oral Sens Percept. 1973;(4):254-78.