The chemical enzyme is sent by fat cells as a sensor to indicate enough food has been eaten.
This stops any more eating – preventing us from putting on too much weight.
The breakthrough could could also lead to treatments for eating disorders such as anorexia nervosa.
Drugs that target the protein known as BH4 by boosting or lowering levels could suppress the appetite of those hooked on food or increase it for others addicted to dieting.
Experiments on flies found it plays a key role in reducing eating in flies – offering hope of a similar chemical pathway in humans.
Professor Walton Jones, of the Korea Advanced Institute of Science and Technology in Daejeon, said: “Our study indicates fat tissue sends a molecular signal to the fly brain to regulate feeding behaviour.
“Further studies will be needed to determine if a similar system acts in mammals, and if so, whether it can be safely manipulated to help achieve weight loss, or gain, in people.”
The study published in PLOS Biology said fat is the primary long-term energy storage chemical in animals and controlling levels is critical for survival.
In mammals the hormone leptin induces eating in response to fat loss but so far no corresponding signal has been identified stopping eating in response to fat gain.
Fruit flies replicate many of the feeding-related regulatory mechanisms and genes known to operate in humans.
Professor Jones, a neurobiologist, explained this is why they make a good model for the search for such an inhibitory signal.
His team focused on bits of DNA called microRNAs as they are well-known inhibitors of gene expression.
This identified the chemical miR-iab-4 in fat tissue which increased feeding by more than 27 per cent and a gene called purple which was expressed in fat bodies.
Blocking the gene increased feeding – suggesting its normal function was to inhibit it.
Purple is known to be one of two enzymes that build a molecule called PTP which is released by fat bodies and circulates in the fly brain.
There, a third enzyme converts PTP into BH4 (tetrahydrobiopterin) where neurons produce the protein NPF that regulates feeding.
Loss of purple in the fat body – or of BH4 in brain cells – led to increased release of NPF and increased feeding.
Conversely, increasing BH4 in neurons reduced NPF release and decreased feeding.
Finally, the researchers showed feeding flies a low-calorie diet reduced expression of the fat body enzymes that control BH4 production – and led to increased feeding.
They said the results suggest BH4 plays a key role in suppressing appetite in flies – and that PTP released from fat bodies delivers a signal to the brain indicating energy stores are sufficient and feeding can stop.
Professor Jones said: “Because the enzymes that produce BH4 and the enzymes that require BH4 are all highly conserved, it will be interesting to see whether our results in flies also apply to mice or humans.”