Date:November 2nd, 2006

Speaker's Name and Affiliation:
Mark Friedman, Ph.D.
Monell Chemical Senses Center

Title: "Overeating in Obesity: A Behavioral Response to a Hungry Liver"

Presiding Chair: Harry R. Kissileff, Ph.D.

Rapporteur: Kathleen L. Keller, Ph.D.

Attendees and their Affiliation:

Kathleen Keller	Columbia/Obesity Research Center
Harry Kissileff	Columbia/Obesity Research Center
Carol Maggio	New York Obesity Research Center
Michael Lewis	Hunter College
Boman Irani	VA Medical Center
Dennis Vanderweele	Independent/retired
Joe Vasselli	Obesity Research Center
Susan Ettinger	Obesity Research Center
Amy Samuelson	GlaxoSmithKline
Christelle Le Foll	VA Medical Center
Christa Patterson	UMDNJ
Suki Scott	Hunter College
Gary Schwartz	AECOM
Barry Levin	VA Medical Center
Anne Klein	New York Obesity Research Center
Chris Ochner	Obesity Research Center
Tatiana Ungredda	St. Luke's
Mary Yeotsas	Obesity Research Center
Joel Grinker	University of Michigan
Bart Hoebel	Princeton University
John G. Kral	SUNY Downstate
Tony Sclafani	Brooklyn CUNY
Karen Ackroff	Brooklyn CUNY
Lisa Neff	Rockefeller
Rhoda Gruen	Columbia University
Jennifer Nasser	Obesity Research Center
Rhoda Gruen	Columbia University
Allan Geliebter	Obesity Research Center
Chris Ochner	Columbia/ Obesity Research Center
Karen Ackroff	Brooklyn college

Summary (Provided by Speaker):

Several experiments with human subjects have shown an inverse correlation between body mass index (BMI), a measure of fatness, and liver ATP concentration or degree of recovery from ATP depletion: the higher the BMI, the lower the level of liver ATP or the smaller the fractional recovery of ATP after experimentally-induced depletion. These observations are consistent with the hypothesis based on studies with rats that a reduction in liver ATP production is a stimulus that controls eating behavior. In this talk, Dr. Friedman reviewed evidence for this hypothesis and presented newer studies showing that rats susceptible to diet-induced obesity have a preexisting deficit in hepatic fatty acid oxidation, which, when they are fed a energy-dense, high-fat diet, results in a decreased capacity for liver ATP production, hyperphagia and obesity.

A variety of evidence points to a role of changes in liver ATP production as a stimulus controlling feeding behavior in rats. Studies using various metabolic inhibitors alone or in combination demonstrate that reduction of liver ATP stimulates food intake. Fasting decreases liver ATP production and the compensatory hyperphagia during refeeding follows a time course that parallels restoration of hepatic energy production. Rats with experimental diabetes mellitus that are fed a low-fat diet are hyperphagic and show low levels of liver ATP production. Both the hyperphagia and reduced liver energy status are reversed by feeding diabetic rats fat, a metabolic fuel that these animals lacking insulin can readily oxidized.

Considering the shifts in fuel partitioning among different metabolic pathways that are seen with the development of obesity, Dr. Friedman previously suggested that increased storage and sequestration of fuels as fat in adipose tissue decreased the fuel supply to liver, resulting in lowered hepatic energy production and, consequently, increased food intake. More recent studies suggest that a deficit in the capacity to oxidize fatty acids in liver may also play a part in overeating and the development of obesity.

This role for reduced hepatic fatty acid oxidation appears to underlie susceptibility to diet-induced obesity. The capacity for fatty acid oxidation is predictive of body weight gain after rats are fed an energy-dense, high-fat diet; outbred obesity-prone rats oxidize dietary fatty acids to a lesser extent than do obesity-resistant rats. The same preexisting deficit in fatty acid oxidation is seen in inbred Levin DIO rats relative to DR rats, and in addition, DIO rats also evidence a lower rate of hepatic oxidation of endogenous fatty acids during fasting. Relative to DR rats, expression of enzymes involved in fatty acid oxidation (CD36 and long-chain acyl-CoA dehydrogenase) in liver is lower in DIO rats before they are fed a high-fat diet. DIO rats that had been eating a high-fat diet and gaining excess weight also show reduced expression of these enzymes in liver and, in addition, show reduced expression of liver carnitine palmitoyl transferase 1a (CPT1a), a rate-limiting enzyme for fatty acid oxidation. Expression of liver enzymes involved in other aspects of fat metabolism is similar in DIO and DR rats as is expression of these enzymes in muscle.

Additional evidence that a deficit in hepatic fatty acid oxidation underlies diet-induced overeating and obesity stems from experiments showing that treatment with fenofibrate, which stimulates hepatic fatty acid oxidation, reduces food intake and body weight and increases expression of CPT1a in obese obesity-prone rats, but not in lean obesity-resistant rats. These results suggest that fenofibrate treatment reverses a critical deficit in rats that are susceptible to diet-induced obesity. Inhibition of hepatic fatty acid oxidation stimulates food intake in rats eating a high-fat diet, an effect which is associated with a reduction in liver energy production.

These findings raise the possibility that a low capacity for or a greater vulnerability to decrements in liver energy production may underlie diet-induced overeating and obesity. Several observations supports this hypothesis: (i) Obesity-prone rats, which eat more than do obesity-resistant rats during refeeding after a fast, also restore liver energy status more slowly under these conditions, and (ii) Injection of 2,-5-anhydro-D-mannitol, reduces liver ATP more and stimulates feeding behavior to a greater extent in obesity-prone rats as compared with obesity-resistant rats. The role of reduced liver ATP production in the hyperphagia that accompanies the development of obesity may extend beyond diet-induced obesity as pilot studies show lower liver ATP levels in obese Zucker rats and rats with lesions of the ventromedial hypothalamus.

In conclusion, obesity appears associated with a shift in the partitioning of fuels into adipose tissue that is independent of overeating. An inherited propensity for diet-induced obesity and overeating is associated with a preexisting deficit in hepatic fatty acid oxidation. The loss of metabolic fuels into adipose tissue, combined with a limited capacity to oxidize fatty acids in liver, diminishes hepatic energy production, which in turn drives overeating that further contributes to excessive fat deposition.

Discussion:

Q. Is (the recovery of liver ATP production after a period of food deprivation) because the liver is making glycogen?
A. Yes, I think you are right.

Q. In diabetic rats, are you restricting carbohydrates?
A. It doesn't matter, you can dilute the diet with carbohydrates if you want.

Q. Are you suggesting that liver ATP acts as a signal to stimulate food intake?
A. It's a transient signal that comes and goes, but yes, that is what I'm suggesting.

Q. If you had a food deprived rat and you infused ATP, would the animal eat?
A. You would get serious changes in circulation. In order for ATP to serve as a signal for intake, it first needs to get into the cell.

Q. Where does the vagus message go in the brain?
A. The paraventricular nucleus, the nucleus of the solitary tract, and a variety of other places along vagal afferents.

Q. What is the correlation between liver ATP and liver glycogen?
A. We do not always see a correlation, or at least not a constant correlation.

Q. What happens if you give phosphate without giving 2,5 AM?
A. Nothing happens.

Q. Haven't conflicting results been found when the liver is denervated?
A. Yes, that's true. I don't like doing vagotomies, as they grow back, and the autonomic nervous system has quite a bit of plasticity.

Q. Does the short term composition of the fuel make any difference, for example, would a higher % fat fuel result in a difference in phosphate trapping?
A. Not sure, but I think it could. Medium chain fatty acids would, of course, be absorbed more quickly, and would produce a lot of ketone bodies.

Q. Why does 2,5 AM work in the liver if the liver isn't oxidizing much glucose?
A. Because of phosphate trapping, and this can occur regardless of the fuel source.

Q. Does fructose produce satiety?
A. Fructose is oxidized in the liver, and it is considered a calorie for the liver similar to other sugars.

Q. Do we have any idea what controls the switch from oxidation to storage?
A. Not really. AMP-Kinase might be one possibility at the cellular level.

Q. It seems to me that if you buy that model, obese would continue to overeat. What in your model suggests that eventually, an obese person will reach a point of homeostatic balance?
A. Well, as fat cells get larger, they become less sensitive to insulin. Storage eventually reaches a maximum, and food intake will reduce.

Q. Have you looked at enzyme catalytic activity for any of these targets (CD36, CPTIA, AC ADL)?
A. No, not yet.

Q. There was a study at SSIBS that said that if you give fenofibrates, you will get a decrease in food intake through leptin signaling.
A. It would argue that leptin is not a signal for decreasing food intake at all.

Q. Is there any evidence that fibrates decrease body weight in humans?
A. You tend to gain weight no them, but the weight is redistributed.

Q. What if the signal for food intake doesn't have to do with ATP in the liver at all, but rather is ATP in another tissue, such as the muscle?
A. It's possible, but I have no reason to believe that muscle ATP is involved in food intake.

Q. Are there leptin receptors in the liver?
A. I'm not sure.

Q. Is there a triglyceride aspect to this story?
A. As a fuel, they might be important as they are not oxidized as regularly. However, we don't have data on that yet.

Q. If you remove part of the animal's liver, what would you predict?
A. Animals are a mess when you do this type of surgery. When they are finally recovered, the liver has grown back, so we wouldn't be able to see much.