Presiding Chair: Harry R. Kissileff, Ph.D.
Rapporteur: Kathleen L. Keller, Ph.D.
Date: September 8, 2005

Title: "Obesity, eating disorders and alcohol intake: What goes wrong in the brain?"
Speaker: Sarah Leibowitz, Rockefeller University

Attendees and their Affiliation:

Kathleen Keller	Columbia/Obesity Research Center
Harry Kissileff	Columbia/Obesity Research Center
Khalid Touzani	Brooklyn College
Mike Lewis	Hunter College
Rebecca Gordon	St. Luke's Obesity Research Center
Tony Sclafani	Brooklyn College
Rich Bodnar	Queen's College
Amy Samuelson	GlaxoSmithKline
Judith Gibbs	Columbia University
Erica Goldstein	Columbia University
Mary O'Brien	Columbia University Health Sciences
Carol Boozer	ORC/Columbia
Allan Geliebter	St. Luke's Roosevelt Hospital
Janet Guss	St. Luke's Roosevelt Hospital/Obesity Research Center
Karen Acroff	Brooklyn College
B. Timothy Walsh	Columbia/NY Psych Institute
Bart Hoebel	Princeton University
Doug Mook	Astoria, NY
Allegra Broft	Columbia
Diane Klein	Columbia
Alisa Schwartz	Columbia Counseling Center

Summary:
Why are we so obsessed with dieting? Why do we overeat fat-rich foods? Why do good foods and good wine go so well together? These are some of the questions discussed by renowned neurobiologist Dr. Sarah Leibowitz in her thought-provoking title, "The neurobiology of obesity, eating disorders, and alcohol intake." In this seminar, Dr. Leibowitz provided data to support the role of a non-homeostatic positive feedback loop involved in the persistence of high-fat and/or alcohol intake. The critical mediator in this feed-forward circuit is circulating lipids, in particular triglycerides, which rise in response to fat and/or alcohol intake. Furthermore, an increase in circulating lipids stimulates the mesolimbic dopamine system, thus pairing a reward stimulus with the ingestion of fats and/or alcohol. The following summary will detail the various lines of evidence that have lead Dr. Leibowitz and her colleagues to this conclusion.

Dr. Leibowitz began by giving a historically oriented overview of hypothalamic neurochemicals involved in the regulation of eating and body weight. In the 1970s, much of the interest was on monoamines, including norepinephrine, dopamine, and serotonin. Gut-brain peptides occupied much of the interest in the 1980s, with the list including galanin, NPY, CCK, and Peptide YY. Advances in molecular biology made it possible to discover and investigate the function of a new group of neurochemicals in the 90s, including agouti-related peptide and the family of orexins. In addition, there are a number of circulating hormones that are known to mediate action of the hypothalamic neurochemicals. These include insulin, leptin, estrogen, progesterone, testosterone, and corticosterone. Moreover, circulating nutrients also mediate the action of hypothalamic neurochemicals, and these include triglycerides, glucose and alcohol.

Homeostatic negative controls of consummatory behavior are mechanisms that promote energy balance by inhibiting further consumption in the fed state. These mechanisms evolved in a time of limited supply of natural foods. However, these mechanisms do not always regulate energy balance in the face of palatable, high-fat foods, such as those seen in current US society. The best example of this can be seen in the cafeteria diet: when animals are given access to highly palatable foods, they overconsume and gain weight. These abundant palatable foods are rich in fat and sugar, readily available, and they increase consummatory behavior beyond physiological need. It is probably that they do this through a non-homeostatic positive feedback mechanism, where instead of resulting in decreased intake, palatable foods stimulate mechanisms that increase intake.

The mechanisms by which palatable foods facilitate the non-homeostatic positive feedback mechanism are through nutrients, hypothalamic peptides, and forebrain neurotransmitters. Palatable, high-fat foods and alcohol stimulate the release of triglycerides, fatty acids, and glucose into the circulation. These nutrients serve to increase production and release of hypothalamic peptides which moreover encourage further consumption, and consequently, override negative feedback inhibition. The cycle is mediated by neurotransmitters in the forebrain, including dopamine, which are stimulated by both nutrients and peptides, and provide a reward to the animal (or human) for consumption of the palatable food. The result is a cycle that allows fat and alcohol consumption to override the negative feedback inhibition that would normally function to decrease intake in the fed state.

Dr. Leibowitz reviewed several experiments to support the notion that nutrients such as fat and alcohol encourage a cycle of sustained and increasing intake. For example, animals consume more chow 90 minutes after a high-fat meal (15 kcals) than they do after a low-fat meal (15 kcals). High-fat meals increase levels of triglycerides, relative to a lower fat meal, but do not increase circulating levels of insulin and leptin. Furthermore, high-fat meals increase levels of galanin mRNA in the PVN, but not the ARC (Leibowitz et al, Brain Research 2004). In addition, galanin in the PVN is stimulated by peripheral intralipid injection. Double labeled neurons suggest that there is a direct effect of fatty acids on stimulation of hypothalamic peptides such as galanin and enkephalin. Central oleic acid injections have a similar effect on these neuropeptides in the PVN, but not the ARC. The overall result of fat intake, as these data suggest, is that peptides that stimulate intake are increased, while hormones that would counteract this effect are not.

What role does nucleus accumbens dopamine (DA) play in this non-homeostatic feedback mechanism? Hypothalamic peptides, such as galanin, stimulate dopamine. In addition, lipids may have a direct effect on increasing levels of dopamine. As a consequence of dopamine release, both hypothalamic peptides and nutrients directly result in a reward response that allows feedback signals to be overridden.

It is well known that sugar increases the palatability of many foods. Sugar also increases daily caloric consumption in animals and has well-studied addictive properties, particularly through its ability to increase levels of dopamine. When animals are injected with 10% glucose, the initial effect after 30 min is a suppression of NPY and AgRP mRNA in the arcuate nucleus. After this effect, however, there is a rebound of peptide expression and a significant increase above baseline levels at 90 min after glucose administration. These data suggest that sugar, like fat, also results in non-homeostatic positive feedback by increasing levels of intake-enhancing neuropeptides.

Finally, how does alcohol consumption mimic the effects seen with both sugar and fat? First, alcohol intake stimulates production of circulating triglycerides, without concomitant increases in leptin and insulin. Alcohol also stimulates expression of peptides in the PVN, most notably, galanin, opioids and orexins, while having little effect on or suppressing NPY. Thus, alcohol exerts influences on consumption that are similar to those seen with fat.

In sum, fat, sugar, and alcohol result in stimulation of neuropeptides that instead of decreasing intake, actually appear to increase and encourage intake. The current nature of the food supply in the US is abundant in all of these dietary components. Because these effects appear to be more sensitive to female sex hormones, these data may in part explain why women are more prone to disordered eating than men.

Discussion:
Q. Are you going to address how homeostatic systems change in response to the supply or limits of natural foods?
A. Yes, but I am focusing my talk on what happens with non-homeostatic systems in relation to palatable foods.

Q. What is the rest of the diet? If animals get a high-fat meal, how does this affect their intake of chow after the meal?
A. We have shown that after a preload and a 90 minute delay, animals given the high-fat preload eat more chow compared to those given the low-fat preload.

Q. Are the animals food restricted?
A. No

Q. Do animals eat all the preload?
A. Yes

Q. When you reduce fat in the preload, how do you still make it the same calories as the low-fat preload? What do you replace the fat with?
A. Sucrose

Q. Is it possible that the effects you are seeing (differences between intake following low and high-fat preloads) are due to the sugar, and not the fat?
A. Yes, it's possible, but we do have specific paradigms where the peptides are found to shift in response to a rise in fat rather than a fall in carbohydrate.

Q. How much would the animals eat if they were not given a preload?
A. That's a good question. I'm not sure.

Q. When did you see a differential effect with leptin?
A. The difference in leptin (high-fat vs. low-fat preload) was not significant.

Q. Do you have any data on ghrelin?
A. No, but there are published data to suggest that nutrient infusions directly into the stomach do affect ghrelin.

Q. Do fats and fatty acids cross the blood-brain barrier?
A. Yes, fatty acids can cross the blood-brain barrier, and although triglycerides are generally thought to remain outside the blood-brain barrier, the evidence is not totally clear on this. Circulating triglycerides certainly affect fatty acids in the brain.

Q. Have you also shown that fat increases orexins?
A. Yes, I mentioned it briefly.

Q. Is apolipoprotein B related to these peptide mechanisms?
A. It's an important point, but I don't know the relationship.

Q. Is there a galanin knock-out?
A. Yes, and some studies have been performed by others in relation to fat ingestion, although the results have yet to be published.

Q. The alternative view to the one that you are taking is that fat exhibits less of a negative feedback, rather than directly stimulating intake?
A. Perhaps both feedback processes, a reduction in satiety and stimulation of food intake, may be involved.

Q. Does fat also increase testosterone?
A. Yes, but the pattern of peptide change in the male is different. Testosterone exerts effects on the ARC, whereas the female steroids affect peptides in the PVN and medial preoptic area.

Q. Couldn't you also say the same for fat as for sugar? Does fat have addictive properties?
A. This is possible, but there is no clear evidence yet to demonstrate this.

References:

Leibowitz, S.F., Dourmashkin, J.T., Chang, G.Q., Hill, J.O., Gayles, E.C., Fried, S.K., Wang, J., 2004. Acute high-fat diet paradigms link galanin to triglycerides and their transport and metabolism in muscle. Brain Res, 1008, 168-178.