Taste Aversion and Preference Learning in Animals

views updated


Historically taste aversion learning arose as a problem in evolutionary biology. The English naturalist Charles Darwin was puzzled by an incongruity: Some tender caterpillars were brightly colored and exposed themselves so that they caught the eye of every passing bird. Such behavior appeared maladaptive. Years later, the English anthropologist and naturalist Alfred Russell Wallace suggested that brightly colored butterfly larva probably tasted bitter and might be poisonous; therefore the colors served to deter birds and other predators. Subsequent research supported Wallace's hypothesis. Consumption of the colorful insects causes gastric nausea and emesis, and after one or two trials birds and other predators learn to avoid them. As larva, these insects feed on plants that evolved the bitter toxins as a defense against herbivores; the insects turned that defense to their own advantage.

Taste aversion learning proved to be widespread in phylogeny and ontogeny. Taste-toxin conditioned aversions have been observed in snails, insects, fish, frogs, salamanders, lizards, snakes, domestic and wild birds, and in mammals, ranging from fetal and neonate rats, to young children and adult humans. Even protozoans reject bitter, the natural taste of plant poisons. The ubiquity of the phenomenon indicates that this mechanism to protect the gut must have evolved many millions of years ago.

In the mid-1960s taste aversion learning caught the attention of experimental psychologists. They observed that when an animal drinks a tasty solution marked by a bright-noisy signal, and is later injected with a mild toxin, the animal will develop an aversion to the taste but not to the bright-noisy signal. Conversely, if the animal is mildly shocked on the feet, it will avoid the bright-noisy signal but not the tasty solution. Second, and more important, the shock must be applied immediately after the signal for effective learning, but the toxin injection can be delayed for several hours after the consumption of the tasty solution. Moreover, a one trial situation, that is, a single pairing of the taste with the toxin injection, is sufficient to elicit a strong aversion to that particular taste. These factors, known as (1) selective association and (2) long-delay learning, are the major behavioral characteristics of taste aversion learning. This type of behavioral paradigm, that came to be known later as conditioned taste aversion (CTA), can tolerate an interstimulus interval of up to six to eight hours between the taste (the conditioned stimulus, CS) and the malaise inducing agent (the unconditioned stimulus, US) during the training session. After consumption, the internal representation of the taste would probably be encoded in an "on-hold" position over hours before a decision is being made of whether the food is safe or not. This long interstimulus interval enables the dissociation in time of neuronal events that generate the memory of the sensory stimulus from those that subserve the association of the memory of the CS with the US, that is, the negative reinforcer.

Since the 1980s, the ecological paradigm of CTA, by virtue of its aforementioned experimental advantages, has been adopted by several research groups in the study of the behavioral, pharmacological, cellular, and molecular aspects of learning and memory in mammals. In the laboratory, a routine CTA protocol is composed of the following steps: (1) the preconditioning session (three to four days), in which rats learn to drink water from the liquid container (usually the liquid is supplied in glass pipettes); (2) the conditioning session (a single day), where rats sample the taste (usually a solution of saccharin, but many other tastes can be used) and around thirty to sixty minutes later administered with the transient malaise-inducing agent (an intraperitoneal injection of a solution of lithium chloride is used as the standard US, but the US can range from rotation and irradiation to drugs and poisons); and (3) the testing session, which can vary between one and six days. Rejection of the conditioned taste can be easily monitored by measuring the amount of the taste consumed by the animal on the day of the test (a single-bottle test), and comparing this volume to that consumed on the day of conditioning. Another way of quantifying aversion to a specific taste is to calculate a so-called "aversion index," based on the total amount of water consumed compared to the taste in a multiple-choice situation.

The neurobiological mechanisms of the CS-US association in CTA are known only in broad outline. In the rat, the processing of gustatory information begins with transduction of chemical stimuli which reach the oral cavity. The taste receptors send afferents via the facial and glossopharyngeal nerves to the nucleus of the solitary tract in the brainstem. The receptors in the viscera also send vagal fibers converging to the same nucleus. The blood carries absorbed food products to the area postrema, where blood monitors report to the solitari nucleus. The CS-US routes then proceed to the parabrachial nucleus (PBN) located in the pons, the more anterior region of the brainstem. Neurophysiological experiments indicate that a complex series of looping circuits interconnect these nuclei in the brainstem with higher brain areas such as the gustatory cortex, located in the insular cortex (IC). Although the IC is unnecessary for simple reflex responses to gustatory stimulus, anatomical and metabolic lesion experiments have shown that the IC is required for the retention of learned taste aversion. The amygdala, another region interconnected with the PBN and the IC, is believed to play an important role is assessing the hedonic value of the consumed taste (i.e., the emotional aspect of the taste learning experience). All in all, the behavioral, anatomical, and pharmacological data accumulated to date suggest that the IC is the area of the brain involved in the encoding of the memory of the taste and in the processing the detection of taste unfamiliarity (see below); the amygdala as the region responsible for the evaluation of the hedonic value of the taste as well as the expression of CTA; and the PBN as the locus of the CS-US association.

An important element in CTA is the novelty, or unfamiliarity, of the taste stimulus. Sampling of any kind of a novel tastant by the rat, even in the absence of the negative reinforcer, will result in the formation of a memory to that specific taste. However, if these same animals are then subjected to CTA training (now in the presence of the malaise-inducing compound), they will show a poorer aversion to the taste compared to animals that were not pre-exposed to it. The CS in CTA is most effective in rendering a strong aversion response if it is unfamiliar to the organism at the time of conditioning, A major question is, How does the brain "know" when a taste is familiar or unfamiliar? Taste novelty detection is expected to require some type of fast internal comparator that matches the on-line (sensory) information with off-line (memory) information. A potential candidate for this comparator is a corticothalamo-brainstem system. The thalamus may compare the on-line sensory information coming from the brainstem with previous taste memory representations retained in the IC, and when a mismatch is identified (if the on-line taste information is novel), it triggers the behavioral response on the one hand, and initiates memory encoding in the IC on the other.

By using the advantage of a single conditioning trial, investigators have examined the role of several biochemical and molecular processes involved in the discrete phases of acquisition, consolidation, and extinction of CTA memory. For example, the microinfusion of a protein synthesis inhibitor into the IC blocks CTA learning when administered before the exposure to the taste, but not before the testing trial, indicating that the synthesis of new proteins in the cortex is a critical step for the formation, but not for the retrieval, of the taste memory. Along this line of experimentation, researchers have found that the cellular and molecular mechanisms that subserve CTA are similar to those that subserve other forms of learning. These specific molecular devices ("switches") are turned on/off in different brain regions during CTA (e.g., activation of cholinergic receptors and phosphorylation of glutamate NMDA receptors in the IC, and activation of specific intracellular signaling cascades together with modulation of gene expression in the IC and in the central nucleus of the amygdala). These essential molecular entities are also differentially activated in the brain according to the stimulus dimension and context. For example, muscarinic receptors and activation of members of the mitogen-activated protein kinase (MAPK) signaling cascade are necessary for the acquisition of CTA to a novel taste but not to a familiar one.

As mentioned, CTA results in a robust learning, but yet, this behavior is very plastic. The aversive memory can last for several months without significant decay. However, if after conditioning animals are subsequently exposed to the taste in the absence of the negative reinforcer (as in a test situation), and provided that they sample even a tiny amount of the taste, the aversive memory will commence to decay. This phenomenon, first described by the Russian physiologist Ivan Pavlov as experimental extinction, does not result in the erasure of the original aversive memory, but rather reflects a relearning process in which now the new CS-NoUS association comes to control behavior.



Berman, D. E., and Dudai, Y. (2001). Memory extinction, learning anew, and learning the new: Dissociations in the molecular machinery of learning in cortex. Science 291, 2,417-2,419.

Berman, D. E., Hazvi, S., Neduva, V., and Dudai, Y. (2000). The role of identified neurotransmitter systems in the response of insular cortex to unfamiliar taste: Activation of ERK1-2 and formation of a memory trace. Journal of Neuroscience 20, 7,017-7,023.

Berman, D. E., Hazvi, S., Rosenblum, K., Seger, R., and Dudai, Y. (1998). Specific and differential activation of mitogen-activated protein kinase cascades by unfamiliar taste in the insular cortex of the behaving rat. Journal of Neuroscience 18, 10,037-10,044.

Bures, J., Bermudez-Rattoni, F., and Yamamoto, T. (1998). Conditioned taste aversion: Memory of a special kind. New York: Oxford University Press.

Bures, J., Buresova, O., and Krivanek, J. (1988). Brain and behavior: Paradigms for research in neural mechanisms. New York: Wiley.

Garcia, J., Ervin, F. R., and Koeling, R. A. (1966). Learning with prolonged delay of reinforcement. Psychonomic Science 5, 121-122.

Garcia, J., Kimmeldorf, D. J., and Koelling, R. A. (1955). Conditioned aversion to saccharin resulting from exposure to gamma radiation. Science 122, 157-158.

Garcia, J., McGowan, B. K., Ervin, F. R., and Koelling, R. A. (1968). Cues: Their relative effectiveness as a function of the reinforcer. Science 160, 794-795.

Lamprecht, R., and Dudai, Y. (2000). The amygdala in conditioned taste aversion: It's there, but where. In J. Aggleton, ed., The amygdala: A functional analysis, 2nd edition. Oxford: Oxford University Press.

Rosenblum, K., Berman, D. E., Hazvi, S., Lamprecht, R., and Dudai, Y. (1997). NMDA receptors and the tyrosine phosphorylation of its 2B subunit in taste learning in the rat insular cortex. The Journal of Neuroscience 17, 5,129-5,135.

Schafe, G. E., Sollars, S. I., and Bernstein, I. L. (1995). The CS-US interval and taste aversion learning: A brief look. Behavioral Neuroscience 109, 799-802.

Yamamoto, Y., et al. (1994). Neural substrates for conditioned taste aversion in the rat. Behavioral Brain Research 65, 123-137.

Diego E.Berman