The Neurons That Would Starve You

Hunger is complicated. It's not merely a single drive, though this is mostly how may experience it consciously: a single dimension of hunger magnitude. We are more or less hungry, sometimes not at all. But there's something else lurking in the brain: anti-hunger. We can be hungry and not hungry simultaneously, in a sense.

In more concrete terms, we can imagine that there is in the brain a certain subset of "hunger neurons." When we feel hungry—as during periods of fasting—it means that these neurons are active. Otherwise, the hunger neurons are silent. Hunger neurons are quite real: neuroscientists have demonstrated their function by stimulating hunger neurons artificially, causing mice to eat at weird times and gain weight.

But something interesting happens as we start cranking hunger neurons (agouti-related protein, or AgRP, neurons) up. There's a limit. Mice won't just eat themselves to death. This indicates that there's something else to hunger, a moderating factor. This factor is described for the first time this week in Nature Neuroscience by researchers at Harvard Medical School: a new population of neurons that intermingle with AgRP neurons and basically have the opposite effect. Anti-hunger.

Anti-hunger is in itself not a brand new idea. For a long time, neuroscientists looked to pro-opiomelanocortin (POMC) neurons, which are likewise intermingled with the AgRP hunger neurons, for filling this role. This is reasonable: genetic mutations and manipulations to the POMC neurons have been observed to lead to obesity in mice.

Further research, however, hasn't borne out the POMC anti-hunger hypothesis. "Chemogenetic and optogenetic activation of POMC neurons has little effect on short-term feeding behavior, suggesting that other neurons are involved in meal termination," Richard Palmiter, a biochemist at the University of Washington, explains in a perspective accompanying the new paper. "Hence, POMC neurons appear to be more important for long-term body-weight homeostasis."

Why should we even expect that short-term hunger-quenching neurons exist? For one thing, the AgRP neurons do something kind of unexpected when the body is presented with food. Rather than wait for the stomach to feel full or for the body to start releasing metabolism-related hormones, these neurons flip off even before the first bite of food. Yet, we keep eating until we are full, or at least not hungry.

Instead, feeding eventually stops as these various feedback mechanisms trigger the new neural circuit described in the current paper, which is located in a part of the brain called the solitary nucleus, or NTS. These neurons project into multiple parts of the brain that also happen to be connected to POMC and AgRP neurons.

"Activation of ARC glutamatergic neurons rapidly reduced food intake by 50 percent during the time when mice normally eat," Palmiter writes. "Conversely, inactivation of these neurons during the day stimulated food intake. Consistent with these observations, these glutamatergic neurons are more active during the fed state and are inhibited by neighboring AgRP neurons."

So, what can we actually do with this information? Not much at the moment. We're still at the very early stages of understanding this stuff. The interplay between how the body actually senses hunger and satiety ("fullness") is finally coming into focus, but neuroscientists have a long ways to go. For now you'll have to diet the old fashioned way.