A fruit fly deprived of essential amino acids will, within minutes, lose interest in a sugar feast and start hunting for protein instead. That behavioral flip is not a vague instinct. According to research published in Science in 2025, it is driven by a specific peptide released from gut cells that travels to the brain through two separate channels, one fast and one slow, and actively suppresses sugar cravings while amplifying the drive to find protein. The peptide is called CNMamide, and its discovery is filling in one of the biggest blank spots in nutrition science: how the body knows, in real time, that it needs more protein and what it does about it.
A peptide with two delivery routes
The story begins in the gut lining. When a fruit fly’s diet falls short on essential amino acids, specialized intestinal cells called enterocytes increase production of CNMamide (CNMa). Earlier work, published in Nature in 2021, first identified this peptide and showed that the CNMa/CNMaR signaling axis links amino acid scarcity in the gut to changed feeding behavior at the whole-organism level.
The 2025 Science paper advanced that finding substantially. It revealed that CNMa does not rely on a single communication line. Instead, a fast neural pathway carries the deficit signal to the brain almost immediately, while a slower hormonal route reinforces the message over a longer window. The result is a layered alert system: the brain gets an urgent notification and then a sustained update confirming that protein is still needed.
In choice experiments, flies given access to both sugar and yeast-based protein shifted their intake toward protein almost as soon as the CNMa pathway was activated. When researchers silenced the enterocytes that produce CNMa, or disrupted CNMa receptors in the brain, the shift collapsed. Those flies kept gorging on sugar even while protein-starved. That kind of loss-of-function evidence, where removing a single molecular player erases a specific behavior, is among the strongest forms of causal proof in neuroscience.
Mammals run a parallel circuit
Fruit flies and humans diverged hundreds of millions of years ago, but the logic of protein hunger appears to be conserved even if the molecules differ. In mice, a 2025 study reported that vagal sensory neurons relay internal protein status from the gut to the brainstem, shaping subsequent food choices. Separately, the liver-derived hormone FGF21 rises during protein restriction and acts through a receptor called beta-Klotho in the brain to increase motivation for protein-rich foods. Published rodent work has also indicated that FGF21 can suppress carbohydrate intake, though the precise downstream neuronal targets in the hypothalamus are still being mapped in detail.
These mammalian findings matter because they establish that the “skip sugar, find protein” pattern is not a quirk of insect biology. Under normal conditions, gut-derived signals activate vagal and brainstem pathways that drive sugar preference. The CNMa system in flies and the FGF21 system in mammals both appear to override that default when the body’s protein needs become urgent, reprioritizing nutrient choices without necessarily changing total calorie intake.
An open question about sugar-sensing neurons
Not everything in the pathway is settled. One of the more contested points involves a group of brain neurons in flies called DH44 cells. The Science paper reports that CNMa suppresses these neurons to bias feeding away from carbohydrates, framing them as sugar sensors. But earlier research painted a more complicated picture. A 2018 study in Cell Research characterized DH44 neurons as gut-brain amino acid sensors, and a separate paper in Neuron found that DH44 cells are also regulated by Piezo-mediated stomach stretch and glucose levels.
Whether DH44 neurons primarily track sugar, amino acids, or some combined nutritional signal remains unresolved. The answer matters because it determines whether CNMa cleanly “switches off” sugar cravings or instead recalibrates a broader nutrient-sensing hub that weighs multiple inputs at once. Future imaging studies that record DH44 activity in real time across different dietary conditions should help clarify the picture.
The gap between flies and people
No study has yet identified a direct CNMa equivalent in mammals. The vagal and FGF21 pathways share functional similarities with the fly circuit, but the specific molecule that mammalian gut cells might release in response to amino acid shortage has not been pinpointed. It is also unclear whether mammalian enterocytes themselves act as the primary amino acid sensors, or whether enteroendocrine cells or liver-based sensors take the lead.
Human data are even thinner. As of June 2026, no clinical trial has connected amino acid deficit in people to measurable, real-time shifts in FGF21 or any CNMa-like signal. The rodent FGF21 work is well established, but translating those findings into dietary or pharmacological guidance for humans will require controlled intervention studies that have not yet been reported. Until they exist, claims that specific diets or supplements can “hack” protein hunger circuits in people remain speculative.
That gap is worth keeping in mind the next time a headline promises a protein-craving breakthrough. The biology is real and increasingly detailed, but the bridge from laboratory flies to human nutrition advice is still under construction.
What the fly data actually tell us about protein hunger
The Drosophila experiments provide some of the cleanest evidence in modern neuroscience that nutrient-specific hunger is not just a feeling but a molecularly defined process. Genetic manipulation of individual cell types, real-time neural imaging, and precise behavioral measurements all converge on the same conclusion: when essential amino acids run low, the gut dispatches a chemical messenger that rewires the brain’s feeding priorities within minutes.
The mammalian evidence, while distributed across multiple studies and molecules, points in the same direction. When protein is scarce, the body actively steers animals away from sugar and toward protein. That insight helps explain a pattern familiar to anyone who has ever craved eggs or meat after days of carb-heavy eating: the preference may not be random but biologically enforced.
As researchers work to clarify DH44 neuron function, identify mammalian counterparts to CNMa, and test these pathways in human volunteers, the picture of how the gut teaches the brain to seek protein will sharpen. For now, the fly data anchor the mechanism, the rodent studies sketch a parallel architecture, and the human story remains a compelling but still unproven extension of that biology. The next chapter will likely be written in clinical labs, where the question shifts from “does this circuit exist?” to “can we use it?”
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
