Researchers at the Institute for Basic Science (IBS) in South Korea have identified a previously unknown gut-brain signalling system that detects protein deficiency and selectively redirects feeding behaviour towards essential amino acids. The study, published in Science, reveals a dual-pathway mechanism coordinating both rapid neural and slower hormonal responses to nutritional state.
The research team used fruit flies as an initial model, employing neural imaging, genetic tools, and behavioural experiments to map the circuit. When flies were deprived of dietary protein, specialised intestinal cells produced a peptide hormone called CNMa. This signal operated through two coordinated pathways: a fast neural route activating gut-associated enteric neurons that relayed information about amino acid deficiency directly to the brain, and a slower hormonal route in which circulating CNMa reinforced and sustained protein-seeking behaviour over time.
Critically, the system did not simply increase overall appetite. CNMa signalling inhibited activity in sugar-sensing neurons, selectively shifting feeding preference away from carbohydrates and towards protein-rich nutrients. The researchers also found that gut microbiota influence the circuit: flies lacking commensal gut bacteria showed stronger activation of amino acid-seeking brain neurons, suggesting that microbial regulation of nutrient availability feeds into this signalling pathway.
The mechanism was found to be evolutionarily conserved in mammals. Protein-deprived mice similarly developed a strong preference for essential amino acids, and this response remained intact even in animals lacking FGF21, a hormone previously considered central to protein appetite regulation, indicating the existence of additional nutrient-sensing systems yet to be characterised.
The authors suggest the findings may inform future therapeutic approaches targeting obesity, metabolic disease, and eating disorders, particularly given the limited understanding of how endogenous gut signals influence brain-directed feeding behaviour.
Source: Kim B et al. Complex interplay of neuronal and hormonal gut-brain responses to essential amino acid deficit. Science (2026). DOI: 10.1126/science.adv3355.
The research team used fruit flies as an initial model, employing neural imaging, genetic tools, and behavioural experiments to map the circuit. When flies were deprived of dietary protein, specialised intestinal cells produced a peptide hormone called CNMa. This signal operated through two coordinated pathways: a fast neural route activating gut-associated enteric neurons that relayed information about amino acid deficiency directly to the brain, and a slower hormonal route in which circulating CNMa reinforced and sustained protein-seeking behaviour over time.
Critically, the system did not simply increase overall appetite. CNMa signalling inhibited activity in sugar-sensing neurons, selectively shifting feeding preference away from carbohydrates and towards protein-rich nutrients. The researchers also found that gut microbiota influence the circuit: flies lacking commensal gut bacteria showed stronger activation of amino acid-seeking brain neurons, suggesting that microbial regulation of nutrient availability feeds into this signalling pathway.
The mechanism was found to be evolutionarily conserved in mammals. Protein-deprived mice similarly developed a strong preference for essential amino acids, and this response remained intact even in animals lacking FGF21, a hormone previously considered central to protein appetite regulation, indicating the existence of additional nutrient-sensing systems yet to be characterised.
The authors suggest the findings may inform future therapeutic approaches targeting obesity, metabolic disease, and eating disorders, particularly given the limited understanding of how endogenous gut signals influence brain-directed feeding behaviour.
Source: Kim B et al. Complex interplay of neuronal and hormonal gut-brain responses to essential amino acid deficit. Science (2026). DOI: 10.1126/science.adv3355.