Food taste found to rein in eating within seconds, before gut stretch

New research has found that signals linked to how food tastes are sent from our mouths to our brains almost immediately to slow our eating pace, with the well-known stretch signals from the gut coming later. The findings provide insights into how different parts of the body control appetite and how weight loss drugs like Ozempic may work.

That eating causes stretch signals from nerves in the gut to be sent to the brain to tell us that we’re full is well-established. Indeed, newer weight-loss drugs like Ozempic and Wegovy capitalize on this signaling pathway. Now, UCSF researchers have delved deeper into the brain to explore other mechanisms that control eating, making the surprising discovery that signals sent by the mouth to the brain and linked to the taste of food are the first to control our food intake.

“We’ve uncovered a logic the brainstem uses to control how fast and how much we eat, using two different kinds of signals, one coming from the mouth, and one coming much later from the gut,” said Zachary Knight, the study’s corresponding author. “This discovery gives us a new framework for understanding how we control our eating.”

The caudal nucleus of the solitary tract (cNTS), a series of sensory nuclei deep within the brainstem, is the first site in the brain where many meal-related signals are sensed and integrated. Animal studies have consistently shown that the cNTS contains neurons activated by meal-related signals, but these studies weren't able to record the sensory and motor feedback that occurs during natural ingestion in awake, active animals.

The researchers explored the action of prolactin-released hormone (PRLH) neurons, a cNTS cell type whose stimulation inhibits feeding. Using Prlh knock-in mice, they showed that optogenetic stimulation of PRLH neurons inhibited food but not water intake, confirming these cells specifically regulated feeding. Optogenetics uses light and genetic engineering to control a neuron’s activity. When the researchers infused liquid food directly into the stomachs of the mice, they observed that PRLH neurons progressively activated over tens of minutes after the infusion.

They next fasted mice overnight and measured responses to self-paced feeding. In contrast to intragastric infusion of the same substances, PRLH neurons were activated within seconds by oral consumption. Similar rapid activation was seen during the consumption of chow or a high-fat diet but not water or saline. Notably, PRLH neuron activation during oral ingestion didn’t further increase and didn’t track cumulative food intake, unlike what occurred with intragastric feeding. This discrepancy persisted when the researchers matched food delivery to the mouth and stomach.

Examining whether taste perception affected PRLH neuron activation, the researchers found that a high percentage of cells were activated when the mice consumed sweet solutions (74% of all neurons) and fatty solutions (80% of all neurons), implying that most PRLH neurons are not specialized to respond to a single taste. Further experiments suggested that PRLH neurons were activated by food tastes, which in turn rapidly modulated the palatability of food, thereby restraining the pace of ingestion. The researchers propose that this response may be important for preventing the gastrointestinal distress that occurs when food is eaten too quickly.

“It was a total surprise that these cells were activated by the perception of taste,” said Truong Ly, lead author of the study. “It shows that there are other components of the appetite-control system that we should be thinking about.”

The researchers then compared the activity of PRLH neurons to another distinct cNTS cell type involved in regulating satiety: GCG neurons. GCG neurons are activated by ingestion and express the appetite-reducing glucagon-like peptide 1 (GLP-1), the peptide that’s mimicked by the class of drugs that includes Ozempic. They found GCG neurons tracked cumulative food intake and were inherently more sensitive to gastrointestinal feedback, responding preferentially to stretch in the gut. They also noted that, compared to PRLH neurons, GCG neurons produced a longer-lasting satiety.

The study’s findings reveal that sequential negative feedback signals from the mouth and the gut engage distinct circuits in the brainstem, which, in turn, controls feeding behavior in the short and long term.

“Together, these two sets of neurons create a feed-forward, feed-back loop,” Knight said. “One is using taste to slow things down and anticipate what’s coming. The other is using a gut signal to say, ‘This is how much I really ate. Ok, I’m full now!’”

The study has also provided a better understanding of how drugs like Ozempic work, opening the door to weight loss regimens that address the individual ways people eat by optimizing how the signals from these two sets of neurons interact.

“Now we have a way of teasing apart what’s happening in the brain that makes these drugs work,” Ly said.

The study was published in the journal Nature.

Source: UCSF