Date: May 6th, 2010

Speaker's Name and Affiliation:

• Khalid Touzani,
• Brooklyn College, CUNY

Seminar Title: "Neurobiology of Learned Food Preferences"

Chair: Harry R. Kissileff, Ph.D.

Rapporteur: Kathleen L. Keller, Ph.D.

Attendees and their Affiliation:

Kathleen Keller	Columbia/Obesity Research Center
Harry Kissileff	Obesity Research Center
Tony Sclafani	Brooklyn College
Hstang Ju Pan	St. Luke's
Charlisa Gibson	NYORC
Annemarie Olsen	NYORC
Joe Vasselli	NYORC
JA Grinker	University of Michigan
Susan Carnell	NYORC
Roni Friedman	NYORC
Allan Geliebter	NYORC
Karen Ackroff	Brooklyn College
Rich Bodnar	Queen's College
George Collier	Rutgers
Gerry Smith	Weill-Cornell
John Glendinning	Barnard College, CU
Emmanuel Pothos	Tufts University
Steven Zukerman	Brooklyn College
Chris Ochner	NYORC
Katherine Halmi	Weill Cornell MC
Rhoda Gruen	Columbia

Summary: (Prepared by the Rapporteur)

Obesity is raging at a higher pace in the United States in spite of the dietary guidelines that recommend the consumption of low-fat and low-sugar diets. Craving sugar- and fat-rich foods promotes overeating and obesity not only in humans but in animals as well including laboratory rats. What makes fat- and sugar-rich foods so attractive? Certainly, there are some innate preferences for certain flavors (taste, odor, texture) of foods such as the preference for the sweet taste of sugars. However, most food preferences are learned through experience. Among the different types of learning that contribute to food preference learning, we distinguished two major forms called flavor-flavor and flavor-nutrient learning. Flavor-flavor learning is based on associations among the different components of the flavor, e.g., taste-odor association. Flavornutrient learning is based on associations between the oro-sensory information and the post-oral reinforcing properties of food nutrients. These types of association can be studied separately in the laboratory and are forms of Pavlovian learning in which the flavor of food is the conditioned stimulus (CS) and the positive oral and post-oral reinforcing properties of food are the unconditioned stimuli (US).

Flavor-flavor and flavor-nutrient preference learning require the neural integration of oro-sensory and viscero-sensory information and the formation of long-term flavor memories. The brain mechanisms underlying these processes are incompletely understood. Our initial lesions studies have focused on brain structures implicated in flavor aversion learning. The results of these studies indicate that the medial parabrachial nucleus, lateral hypothalamus and amygdala play a crucial role in flavor preference learning. That is, lesions of the medial parabrachial nucleus prevent animals from learning to prefer a CS taste (sour, bitter) paired with intragastric nutrient infusions. The same lesions only attenuate preference conditioning when the CS is a complex flavor that has a strong odor component. Flavor preference learning conditioning is completely blocked by large lesions that include all amygdala nuclei. The same lesions impair, but do not prevent, taste preference conditioning by intragastric nutrients. Thus, the amygdala and medial parabrachial nucleus seem to function in a complementary way. Lateral hypothalamus lesions totally block flavor preference learning when the intragastric infusions of nutrient occur after a certain delay following the ingestion of the CS flavor, suggesting that that this learning deficit reflects a lesion-induced decay in the CS memory trace. Where and how this trace is maintained and where the memory of the learned preferences is stored are unknown.

Food appetite and preferences are mediated by potent brain reward systems that are also involved in drug addiction. There is extensive literature on the role of opioids in food reward. To date, both systemic and intra-nucleus accumbens administrations of the opioid receptor antagonist, naltrexone, has not interfered with the acquisition or expression of flavor-flavor and flavor-nutrient preference learning, providing no evidence for the involvement of the opioids in food preference learning. There is also extensive work on the role of dopamine in food reward and initial microdialysis and pharmacological studies suggested a role of dopamine signaling in food preference learning. Subsequent pharmacological studies revealed that the dopamine D1 receptor antagonist, SCH23390 blocked both the acquisition and expression of fructose-conditioned flavor preference (flavor-flavor learning), where as it blocked only the acquisition of IG sucrose-CFP (flavor-nutrient learning). The D2 receptor antagonist, raclopride, blocked the acquisition of fructose-CFP but had no effect on IG sucrose-CFP. These findings indicate differential involvement of dopamine receptors in flavor-flavor and flavor-nutrient conditioning. However, they do not identify the sites where dopamine is acting to promote flavor preference learning. Dopamine neurons in the ventral tegmental area project to cortical and limbic structures including the nucleus accumbens, amygdala and prefrontal cortex and dopamine transmission within these structures is involved in reward-related learning. Our recent studies, using intra-cerebral administrations of SCH23390 and raclopride, revealed that activation of D1 receptors in the nucleus accumbens, amygdala or medial prefrontal cortex is critical for the acquisition but not the expression of a flavor-nutrient preference conditioned by intragastric glucose infusions. Activation of D1 and D2 receptors in the in the nucleus accumbens or amygdala had minimal effect on the expression of a flavor-flavor preference conditioned by the sweet taste of fructose, did not block its acquisition but accelerated its extinction. The striking resemblance in the effects of the antagonists within different dopamine targets suggests that dopamine transmission within different discrete components of a distributed network is involved in flavor preference learning in a complementary way. This does not mean that dopamine transmission in each of these structures promotes the same learning process. It is possible that dopamine transmission in the amygdala is involved in the association of flavor cues with the affective significance of the rewarding properties of nutrients, dopamine transmission in the medial prefrontal cortex allows updating the representation of flavor cues following their association with food reward, and dopamine transmission in the nucleus accumbens promotes association between food flavors and the consequences of its ingestion as well as the execution of actions upon the presentation of food rewardassociated flavors.

The appetite for high-fat and high-sugar foods and the learning processes through which these foods become more attractive and preferred are presumably mediated at least in part by the brain dopamine systems that are also linked to drug addiction. The link between dopamine and obesity has been a hot topic of recent research. One hypothesis is that, initially, dopamine acts as a reward signal that promotes both the incentive value of foods and reward-related learning. This enhanced reward value may promote overeating and seeking foods rich in sugars and fats and the repeated stimulation of the brain dopamine system by these foods may lead, in turn, to uncontrolled dopamine release that leads to a down regulation of the dopamine receptors in the synapse cliff. Consequently, the reward level mediated by dopamine signal becomes low and requires the over-consumption of highly palatable foods usually rich in fats and sugars to reach a certain level of reward. A better understanding of the basic cellular and molecular mechanisms involved in appetite and learned food preferences may provide insights into the clinical treatment of overeating and obesity.

Question & Answers:

Q. Is the lesion that wipes out conditioned taste aversions preferential to the parabrachial nucleus?
A. Yes, this was a physiologically directed lesion. We just destroyed the part responsible for gustatory perception.

Q. Do either of these lesions (parabrachial nucleus [(PBN) or amygdala (AMY)] block taste preference of sucrose?
A. These rats prefer saccharin over water and lesions in either the PBN or AMY do not change that.

Q. Do either of these lesions affect both taste and flavor?
A. The AMY lesion only slightly attenuates saccharin taste preferences, but does not influence flavor lesions. I'm not aware if either of these lesions affect both taste and flavor.

Q. So the conclusion is that flavor preference learning does not request the primary cortex? Is this correct?
A. Yes, that is correct. Comment: With gustatory cortex lesions, animals will still prefer saccharin over water. There is some recovery of function in these animals.

Q. What happens to taste when you lesion the Insular Cortex?
A. We have not done that yet.

Q. Is maltodextrin a glucose derivative?
A. It is a polymer of glucose, but it's effects on the plasma are similar to that of glucose.

Q. Did insular cortex lesions block flavor aversion learning completely?
A. It's not completely clear. Several studies in the literature have not seen an effect using different methodologies.

Q. What do you think would be the result of hippocampal lesions?
A. You may see some affect on flavor aversion learning, but it's not clear what.

Q. Have there been any previous studies on flavor aversion learning and hippocampus lesions?
A. There is only one study that I know of and they only found an effect on flavor aversion learning after a time delay.

Q. Is it true that if the electrode is placed in the nucleus accumbens, then animals will not exhibit self-stimulation behavior (as observed originally by Olds and Milner, 1954)?
A. Yes, it's true. The Amygdala is a poor place to get brain stimulation.

Q. If you block dopamine receptors, do you also block self-stimulation?
A. Yes, that was Roy Wise has shown.

Q. In your flavor-flavor learning paradigm, why did you switch to fructose?
A. We did several pilot studies and started with sucrose, but we found that fructose did not elicit a post-ingestive learning effect, so we stuck with that.

Q. Why doesn't fructose cause any post-ingestive learning?
A. It's not clear. The data are open to interpretation.

Q. Is the intake effect you see when you administer naltrexone the same whether you administer the drug systemically or directly in the accumbens?
A. Yes. The effects are the same.

Q. Does naltrexone decrease wanting?
A. We aren't sure.

Q. Regarding Gerry Smith's data, how do you know that the finding that raclopride reduced the reward potency of sucrose is not really a conditioned taste aversion?
A. It is not possible to demonstrate a conditioned taste aversion with a sweet taste.

Q. Is your summary model suggestion that there is a preferred level of dopamine in an animal and that animals that are below this will overeat to achieve this preferred level?
A. Under normal situations, an animal would eat a palatable food and dopamine and dopamine receptor synthesis would increase. In an obese animal, the amount of dopamine released and dopamine receptor synthesis is thought to be downregulated. It is possible that animals are overeating to increase the amount of dopamine due to this "inefficiency of the dopamine system."