Please note, this is recovered content from the former website of the New York Obesity Research Center website.


Interaction Between Gut and Brain in the Control of Feeding

Date: February 20th, 2003
Title: “Gut-Brain Communication in the Control of Feeding in Mice”
Speaker: Anthony V. Azzara, Bourne Laboratory-Cornell Medical Center, White Plains, NY.

Gut-Brain-CommunicationThe gut and the brain interact to control food intake. The nature of gut-brain communication has been extensively studied in a variety of species, yet little of this work has been done in mice.

This presentation provides background information on how gut-brain communication has been studied, and reviews novel data we have collected on the controls of feeding in mice.

The study of food intake is the study of meals, as the meal is the functional unit of eating. It has been hypothesized that meal size is controlled in part by positive and negative sensory feedback systems related to food. For example, when an animal consumes a palatable sugar solution, the taste of the sugar provides a positive feedback of a given intensity.

As the sugar solution accumulates in the gut, however, it generates negative feedback signals that increase in intensity as the amount of nutrient consumed increases. When the stimulus intensity of positive and negative feedback are judged to be equal by a central comparator mechanism, the meal ends. Food stimuli in the digestive tract (post-oral) contributes to the negative-feedback control of ingestion.

This is demonstrated in 2 types of experiments: 1) Those that increase negative feedback from the gut, and 2) Those that reduce negative feedback from the gut.

Increasing negative feedback from the gut is accomplished by administering gastric or duodenal preloads before a meal. Both types of infusions have been shown to reduce meal size. In rats, however, it has been shown that the stomach is only sensitive to an infusion’s volume, while the duodenum is sensitive to both volume and nutrient concentration.

Another way of increasing negative feedback is to administer cholecystokinin (CCK). CCK is released from the gut in response to the presence of food, and converging lines of experimental evidence suggest that CCK is an endogenous mediator of the negative feedback control of meal size. CCK reduces meal size in a variety of species.

Decreasing the negative feedback from the gut can be accomplished by sham feeding: a preparation in which anything an animal consumes drains from the stomach via a surgically implanted gastric fistula.

Animals can be tested with the cannula closed, with postingestive feedback presumably normal, or with the cannula open ­ greatly reducing the presence of food in the gut and greatly reducing negative feedback from the gut. Sham feeding animals dramatically increase their intake when the cannula is open, as opposed to when the cannula is closed.

While the controls of meal size have been studied extensively in rats, little work has been done in mice. Establishing that these controls are functional in mice is critical, as most genetic manipulations that influence feeding behavior exist only in mice.

Additionally, classic mouse models of obesity, such as ob/ob and db/db mice, are hyperphagic, and this hyperphagia is primarily expressed by increased meal size. These changes suggest that the function of feedback systems which mediate meal size is compromised.

Finally, as genetic techniques become both more common, and more sophisticated, careful behavioral analysis of the feeding phenotypes of the animals generated becomes more critical. In order to assess these phenotypes in the mutant animals, we must first describe the controls of meal size and feeding behavior in normal mice.

We have begun to assess the negative feedback controls of meal size in mice by adapting techniques previously used in rat models. Our first step was to determine if stimuli shown to reduce meal size in rats would reduce meal size in mice. Mice drinking 12.5% glucose in a one-hour test were given a gastric preload of saline, or 12.5% glucose, or an injection of CCK at 4 or 10 micrograms.

The saline infusion did not reduce intake, but the glucose infusion, and both doses of CCK reduced subsequent glucose intake. These same stimuli which reduced intake (12.5% glucose infusion, 4 and 10 micrograms of CCK) also produced C-fos expression in the AP, NTS, and PBN; hindbrain sites which are known to be involved in the regulation of food intake in rats. This provides convincing evidence that the functional and anatomical pathways that control meal size in rats are functional in mice.

In the next set of experiments, we implanted mice with duodenal catheters and administered preloads before a one-hour glucose-drinking test. The preloads were either carbohydrate (glucose), protein (peptone) or fat (Intralipid), administered in concentrations of 0.25, 0.5 and 1.0 kcal/ml. The infusions were a volume of .5 ml, administered over 10 minutes. As soon as an infusion ended, mice were allowed to consume 12.5% glucose for 1 hour. All 3 classes of macronutrients suppressed intake in a concentration-dependant manner; however, peptone appeared to be somewhat less effective in reducing meal size than the other nutrients.

We have also begun to study sham feeding in mice. Preliminary tests in four mice revealed that they will consume more than twice the amount of a liquid diet, during a one-hour test, when they are sham feeding (i.e., with a gastric cannula open) than when they are ‘real’ feeding (i.e., with a gastric cannula closed).

We interpret these findings as strong evidence that the gastrointestinal negative feedback controls of ingestion in mice are similar to those in rats and primates. Increasing the negative feedback from the gut via preloads decreases meal size, while reducing feedback from the gut, via sham feeding, increases meal size. The results with CCK, as well as the c-fos activation in the hindbrain suggest that gut-brain vagal afferent communication is involved in mediating these controls.

These results encourage the analysis of gastrointestinal controls of feeding and meal size in genetic mouse models of altered food intake and energy homeostasis.

An example of this is our ongoing analysis of NSE-RB mice. These mice have the db/db mutation, which eliminates the long form (Rb) of the leptin receptor and results in an obese, hyperphagic phenotype. NSE-RB mice undergo a transgenic ‘rescue’ of this mutation by selectively replacing the Rb receptor in the brain.

Preliminary studies have demonstrated that the correction of hyperphagia is largely due to a reduction in meal size­ NSER-B animals eat meals that are similar in size to controls, and are much smaller than those eaten by db/db animals. Future research will focus on identifying deficits in the negative feedback controls of intake in genetically modified mice with disordered food-intake phenotypes.


Q. So is there evidence that specific volumes stop the mice’s meals?
A. From these data we can’t say for certain.

Q. Although ‘sham-feeding’ paradigms intend to measure ingestive behaviors in the absence of post-absorptive/ingestive consequences, hasn’t Dr. Sclafani reported elevated blood glucose among rats sham feeding on a glucose solution? Does this suggest the paradigm may not remove all post-ingestive consequences?
A. While Sclafani’s work did demonstrate a rise in blood glucose during sham feeding, this effect was minimal compared with levels observed in ‘real’ feeding paradigms.

Q. Why did you choose to work with maltose?
A. I was expanding on a preparation employed by Kevin Meyers, who had used maltose.

Q. Do intra-duodenal fat infusions cause latency to bottle-feed in the mice?
A. Surprisingly, they did not.

Q. Is the fatty-acid composition of Intralipid atypical of a mouse’s daily diet?
A. Possibly; I’m not aware of any studies that address this subject.

Q. What are typical meal durations for mice, and how do you know when a meal is over?
A. They typically consume one tenth of a milliliter per minute. If they stop licking for five minutes, or pause all activity for three minutes, we consider a feeding bout to be finished. In the preload experiments, the mice will have meals of 45 minutes to an hour (with a saline preload), but they are drinking something very palatable, and they are mildly deprived.

Q. When the mice are given intra-duodenal glucose preloads, do they keep eating?
A. Yes, but these meals are very small.

Q. Have you examined licking patterns associated with these meals?
A. We are currently analyzing these data.

Q. You indicated that the mice lacking the leptin receptor begin gaining a lot more weight at approximately six weeks of age; what do you think is happening at this juncture to drive such weight gain?
A. For mice, six weeks of age represents a developmental stage similar to puberty.

Q. Is this a tonic or phasic difference in leptin activity?
A. Since these studies were conducted under acute conditions, we cannot distinguish between such patterns. However, we are addressing this question in our current experiments.

Q. If meal number changed in the NSE-RB mice, can you conclude that leptin-receptor effects on food intake are primarily mediated by leptin’s effect on meal size?
A. It’s true that the transgenic animals also have increased meal number­ but something has to give­ if the meals are much smaller, the animal will have to take more of them to meet some minimum caloric need. We think that this is a meal size effect, as exogenous leptin reduces meal size in normal rats, and replacing leptin receptors in db/db mice reduces meal size.

Q. Don’t meal sizes change depending on the food offered?
A. Yes- meal size merely represents a unit by which one can measure manipulations on the controls of food intake.

Q. Why don’t you see influences of circadian rhythms on meal size?
A. In our mice these patterns were abnormal as a result of our laboratory paradigm. One contributing factor could have been the palatable liquid diet used in these studies.