Nerve growth factor facilitates conditioned taste aversion learning in normal rats.

Chronic intracerebroventricular (i.c.v.) infusion of 3.2 micrograms/day of nerve growth factor (NGF) in normal rats elevated choline acetyltransferase (ChAT) activity of the striatum, medial septum, and basal forebrain and improved performance of a conditioned taste aversion (CTA) task. Relative to bovine serum albumin (BSA) or Cytochrome C treatments, NGF treatment facilitated acquisition and prolonged extinction of a lithium chloride (LiCl)-induced saccharin aversion. This facilitation was evident at saccharin/LiCl intervals ranging up to 1 h. Also, NGF treatment did not increase reactivity to LiCl-induced illness and neither shifted detection thresholds nor altered hedonic reactions to taste stimuli, indicating that NGF did not produce simple changes in sensory function. NGF treatments that elevate ChAT also facilitate memory of CTA in normal, adult rats.

Aversive conditioning properties of caffeine in rats.

Four experiments tested the conditioning effects of caffeine. Flavor and place cues were paired with IP caffeine injections and followed by tests for cue preference. In Experiment 1A, saccharin was paired with 1.25, 5 or 20 mg/kg of caffeine. In Experiment 1B, caffeine was delivered 30 min before, 5 min before, or 30 min after saccharin. Dose- and time-dependent conditioned taste aversions were produced. In Experiment 2, a place and taste cue were paired simultaneously with 5 or 20 mg/kg of caffeine. Conditioned place and taste aversions developed at 20, but not at 5 mg/kg. In Experiment 3, a place cue alone was paired with 0, 5, or 20 mg/kg of caffeine; dose-dependent conditioned place aversions developed. In Experiment 4, place and taste cues were paired with control treatments: pH-buffered caffeine, purine or vehicle. Caffeine produced taste aversions whereas the purine and vehicle did not. These aversive conditioning effects of caffeine across a variety of situations, doses and temporal arrangements stand in contrast to results obtained with other psychoactive drugs, such as amphetamine and alcohol.

Coyote control and taste aversion.

Studies in which conditioned taste aversion was used as a non-lethal method to suppress coyote predation are reviewed in light of the controversy that surrounds such research. It is concluded that the negative results obtained to date may have been due to theoretical and methodological problems in the studies. Uncritical acceptance of those results has slowed progress on an effective and inexpensive method of coyote management.

Taste potentiation of poisoned odor by temporal contiguity.

The synergistic interaction between odor and taste in flavor-toxicosis conditioning was tested in two experiments. The temporal interval between a 2-min odor and a 2-min taste was varied for thirsty rats licking at a water spout. In the first experiment, taste was presented at time zero, and odor was presented at -10, -1, 0, 1 and 10 min to independent groups in a simple compartment. In the second experiment, taste was presented at 0, and odor was presented at -5, -2, and 0 min in a "wind tunnel" apparatus. The results indicated that odor alone is an ineffective conditioned stimulus for a toxic unconditioned stimulus under our conditions, simultaneous (0-min) presentation of odor with taste potentiates the odor component so that it becomes more effective than the taste component, a 2-min interval between odor and taste attenuates potentiation, and a 5-min interval disrupts the effect, and the interaction in asymmetrical, that is, odor has no such systematic effect on the conditioning of taste.

Alcohol-ingestive habits. The role of flavor and effect.

We propose that alcohol is a food and that drinking alcohol is fundamentally a feeding habit. Many feeding habits are acquired by Pavlovian conditioning in which the flavor of food [conditioned stimulus (CS)] is adjusted by the aftereffects of the ingested meal [unconditioned stimulus (US)]. Thus, repletion of nutritive deficits or recuperation from illness enhances flavor palatability, whereas toxicosis or nausea reduces it. After reviewing the major functional and neurophysiological features of this homeostatic conditioning process, we relate them to the flavor (CS) and effect (US) of alcohol. We present evidence suggesting that the caloric effects of low-dose alcoholic drinks may help to establish and maintain alcohol habits through conditioned palatability changes. Included is a general treatment of the antagonistic interactions between such homeostatic conditioning and the behavioral mechanisms used to defend again peripheral insults such as electric shock. In closing, we discuss the implications for modifying alcoholic habits.

Developmental flavor experience affects utilization of odor, not taste in toxiphobic conditioning.

Feeding experiences were varied in developing rats and the effects upon flavor neophobia and lithium chloride-induced flavor aversions were observed. In Experiment 1, nursing experience of neonate rats was reduced by artificial feeding via intragastric cannula; the rats then were tested with apple juice paired with lithium chloride injection at weaning or maturity. Conditioned aversions were not affected, but neophobia to novel apple juice was attenuated in artificially-reared rats tested at maturity. In Experiment 2, rats received enriched feeding experience after weaning, which consisted of (a) obtaining many complex flavors, a few of which were paired with poisoning, effortlessly in the home cage, or (b) foraging for various foods on an elevated maze. No dramatic effects on neophobia or conditioned taste aversion for saccharin water were apparent. In Experiment 3, rats were given experience after weaning with vanilla-scented water either paired or unpaired with quinine water, and then tested with the odor of almond or that odor compounded with saccharin water for neophobia and lithium-induced aversions. Flavor-experienced rats exhibited more pronounced odor conditioning and more resistance to extinction of the odor aversion after both simple and compound conditioning. In contrast, saccharin taste aversions were relatively unchanged. Apparently, enriched feeding and drinking experience facilitates the utilization of odor more than taste cues.

Flavor-illness aversions: potentiation of odor by taste is disrupted by application of novocaine into amygdala.

Taste and odor have different properties in toxiphobic conditioning. When each is used alone, taste becomes aversive when followed by immediate or delayed poison, while odor becomes aversive only if followed by immediate poison. However, if odor and taste are presented as a compound and followed by delayed poison, then odor does become aversive when tested alone. It is as if taste has potentiated the odor signal. Several experiments assessed the role of the amygdala in this potentiation effect by anesthetizing the amygdala with 10% novocaine. Novocaine applied 30 min before presentation (Pre-CS) of an odor-taste compound disrupted the potentiated odor aversion but not the taste aversion. In contrast, novocaine applied 1 min after the compound odor-taste or 1 min prior to LiCl poison did not dissociate odor and taste aversions; both odor and taste aversions were facilitated. Novocaine applied 30 min before an odor alone also disrupted an odor aversion induced by immediate LiCl. But identical treatment did not disrupt odor avoidance conditioned by immediate foot-shock, suggesting that amygdala anesthesia does not simply produce anosmia. Pre-CS novocaine treatment also disrupted flavor neophobia prior to conditioning. The results suggest that novocaine applied to the amygdala disrupts the integration of odor with taste and illness during toxiphobic conditioning.

Potentiation of odor by taste in rats: tests of some nonassociative factors.

The contribution of nonassociative neophobia and sensitization to the potentiation of odor by taste in rats was tested in three experiments. In Experiment 1, neophobia for almond odor (O), saccharin taste (T), and odor-taste compound (OT) cues was tested before and after noncontingent lithium chloride poisoning and compared with conditioned aversions produced by OT-LiCl temporal pairing. The OT compound potentiated unconditioned neophobia, but there was no evidence of poison-enhanced neophobia, disinhibition of neophobia, or sensitization by noncontingent LiCl; temporal pairing produced aversions for the compound and its elements. In Experiment 2, generalization to a novel odor was tested after O-LiCl or compound OT-LiCl pairing. The potentiated odor aversion did not generalize to the novel odor; it was specific to the odor paired with taste and LiCl. In Experiment 3, potentiation of the odor component by a discriminant or nondiscriminant taste component was tested. Potentiation was evident only when a novel discriminant taste was in compound with odor prior to LiCl poisoning. These studies support an associative "indexing" hypothesis of the potentiation effect in rats.

Flavor-illness aversions: potentiation of odor by taste with toxin but not shock in rats.

Potentiation of odor by taste in rats was tested in a variety of situations. In three experiments, almond odor and saccharin taste were presented either as a single conditioned stimulus (CS) or as a compound CS and followed by either toxic lithium chloride or footshock. Extinction tests with the almond and saccharin components were then given. In single CS-toxin experiments, taste was more effective than odor, and after compound conditioning, the taste component potentiated the odor component. Conversely, in single CS-shock experiments, odor was more effective than taste, and after compound conditioning, no potentiation was observed. Rather, interference effects were observed. In Experiments 1 and 2, the addition of taste disrupted odor CS-shock conditioning, and in Experiment 3, odor interfered with taste CS-shock conditioning. Visceral feedback is apparently a necessary unconditioned stimulus for the potentiation of odor by taste. These data support the neural convergence and gating hypothesis of flavor aversion conditioning.

Flavor-illness aversions: gustatory neocortex ablations disrupt taste but not taste-potentiated odor cues.

Two experiments evaluated the contribution of the gustatory neocortex (GN) to the potentiation of odor by taste during illness-induced aversions in rats. In Experiment 1, rats lacking GN and control rats were given an odor, a taste, or an odor-taste compound cue followed by intragastric gavage of lithium chloride. Prior to conditioning, neophobia for flavored solutions was absent in rats with GN lesions. After pairing with LiCl, GN rats developed normal conditioned odor aversions (Experiment 1B), whereas conditioned taste aversions were attenuated (Experiment 1A) or totally blocked (Experiment 1B). Potentiation of odor by taste after compound conditioning was evident in both control and GN rats, although GN lesions attenuated the effect slightly in Experiment 1B. In Experiment 2, normal rats were given compound conditioning to induce potentiated odor aversions and then given GN lesions prior to tests with the odor and taste components. Taste aversion retention was disrupted totally by GN ablation; potentiated odor aversions were retained by both groups, although the GN group extinguished faster. Gustatory neocortex ablations produced differential effects on odor and taste, disrupting taste memorial and associative processes but leaving odor conditioning and the potentiation of odor by taste processes relatively unaffected. Integrity of the GN apparently is not necessary for the acquisition or retention of potentiation odor aversions.

Vagotomy facilitates extinction of conditioned taste aversions in rats.

Results from three experiments indicate that severing the subdiaphragmatic vagus in rats increased the rate of extinction of learned taste aversions. In Experiment 1, although vagotomized rats acquired a saccharin aversion equivalent to that of controls when the illness-inducing agent was the blood-borne toxin apomorphine, vagotomized rats tended to consume more saccharin than controls during repeated extinction tests. In Experiment 2, vagotomy disrupted retention and increased extinction of a preoperatively acquired saccharin aversion. Disruptions were found when the taste aversion was induced by copper sulfate, a local gastric irritant (Experiment 2A), or apomorphine, a systemic toxin (Experiment 2B); in each experiment vagotomized rats consumed more saccharin than controls on the first retention test and extinguished the prior to surgery. Experiment 3 demonstrated that vagotomy did not affect retention or extinction of a shock-induced conditioned emotional response (lick to suppression) to noise. It is concluded that integrity of the vagus is not necessary for acquisition of a learned taste aversion when a blood-borne toxin is used as the ill-inducing agent. However, the vagus apparently mediates an integral portion of the conditioned response following taste-illness acquisition regardless of whether the illness agent is a local gastric irritant or a systemic toxin.

Flavor-illness aversions: the peculiar roles of odor and taste in memory for poison.

When either taste or odor alone was followed by poison, rats acquired a strong aversion for the taste but not for odor, especially if poison was delayed. When odor-taste combinations were poisoned, however, odor aversions were potentiated, as if odor could gain the enduring memorial property of taste by associative contiguity.

Bait shyness: avoidance of the taste without escape from the illness in rats.

Thirsty rats habituated to drinking .12 M sodium chloride accepted .12 M lithium chloride for 5 min on the first trial but stopped short of their sodium baseline. With repeated trials they reduced consumption of the toxin by either (a) detecting subtle oral (conditioned stimulus, CS) differences, thus avoiding toxicosis (unconditioned stimulus, US) or (b) detecting earlier signs of malaise (US), thus escaping further distress. When both solutions were masked with saccharin, the discrimination was more difficult but still possible. When both solutions were mixed in a solution masking all four taste qualities, the discrimination was severely disrupted. When oral sensors were bypassed with nasopharyngeal tubes, intragastric pumping rats were unable to use postingestional cues to escape, even though such cues were proximal to the ultimate malaise. Oral cues at the distal end of the consummatory chain were extremely effective.

Behavioral regulation of the milieu interne in man and rat.

In regulating the internal homeostatic environment mammals, by necessity, employ behavioral strategies that differ from the tactics used in coping with contingencies in the external environment. When an animal consumes a meal, the palatability of that meal is automatically adjusted in accordance with the ultimate internal effects of that meal. If the meal causes toxicosis, the animal acquires an aversion for the taste of the meal; conversely, if recuperation follows ingestion of the meal, the taste of that meal is enhanced. Unlike the learning that occurs when externally referred visual and auditory signals are followed by punishment in the form of peripheral pain or reward in the form of food in the mouth, conditioning to the homeostatic effects of food can occur in a single trial and rarely requires more than three to five trials, even though the ultimate effects of the meal are delayed for hours. Paradoxically, the animal need not be aware of the ultimate internal effect in the same sense that it is aware of external contingencies. For example, an aversion can be acquired even if the animal is unconscious when the agent of illness is administered. Thus, the way in which food-effects are stored in memory may be fundamentally different from the way in which memories of specific time-space strategies devised for external contingencies are stored. This separation of function is indicated by limbic lesions which disrupt conditioning to a buzzer that is followed by shock and facilitate conditioning to a taste that is followed by illness. Operationally speaking, one can describe both aversion conditioning and buzzer-shock conditioning in the spacetime associationistic terms of classical conditioning. However, psychologically speaking, one must realize that in aversion conditioning the animal does not act as if it were acquiring an "if-then" strategy. It acts as if a hedonic shift, or a change in the incentive value of the flavor were taking place. Such hedonic shifts are critical in regulation of the internal milieu. When an animal is in need of calories, food tends to be more palatable; as the caloric deficit is restored, food becomes less palatable. If the animal's body temperature is below optimum, a warm stimulus applied to the skin is pleasant. When body temperature is too high, the converse is true. In this way, homeostatic states monitored by internal receptors produce changes in the incentive values of external stimuli sensed by the peripheral receptors, and guide feeding behavior. In mammals at least, the gustatory system, which provides sensory control of feeding, sends fibers to the nucleus solitarius. This brainstem relay station also receives fibers from the viscera and the internal monitors of the area postrema. Ascending fibers bifurcate at the level of the pons and project toward the feeding areas of the hypothalamus and the cortex. The olfactory system which primarily projects to the limbic system does not play a primary role in adjusting food incentives. Rather, it plays a secondary role in the activation of feeding, as do other external sensory systems. This specialized conditioning mechanism, which specifically adjusts gustatory hedonic values through delayed visceral feedback, is widespread among animals, including man and rat. These two species are remarkably similar in their thresholds and preferences for gustatory stimuli. The behavioral similarities are based on the animals' having similar gustatory systems, similar convergence of gustatory and internal afferents to the nucleus solitarius, and similar midbrain regulatory mechanisms. Thus, it is not surprising that the feeding of obese rats with internal hypothalamic damage resembles the feeding of obese human beings insensitive to the internal signs of this caloric state. Obviously, man has a highly specialized form of symbolic communication and the rat does not, yet man's cognitive specialization does not prevent him from developing aversions to food consumed before illness even when he knows that his illness was not caused by food (43).

Coyote predation control by aversive conditioning.

Conditioned aversions were induced in coyotes by producing lithium chloride illness in them following a meal, and the effects upon eating and attack behavior were observed. One trial with a given meat and lithium is sufficient to establish a strong aversion which inhibits eating the flesh of that prey. One or two trials with a given flesh (lamb or rabbit) specifically suppresses the attack upon the averted prey but leaves the coyote free to attack the alternative prey. A method of saving both prey and predator is discussed.