A three-step neural circuit lets C. elegans forget one odour but keep another

Researchers at Kyushu University have mapped the molecular and neural pathway that makes the nematode Caenorhabditis elegans actively forget an odour it had learned. The circuit is selective: when worms memorised two odours at once, it removed only one of them. The work, led by Professor Takeshi Ishihara with first author and PhD candidate Yuuki Onishi, is published in Current Biology.

Why forgetting needs explaining

Most memory research has concentrated on how memories are formed and maintained. Loss of memory has received less attention, and is often treated as simple failure. Ishihara argues that forgetting is also needed for adaptation and survival, which raises the question of how a nervous system chooses what to discard.

C. elegans is well suited to the problem. It has only 302 neurons, each of them identified, so a memory and its erasure can be traced to individual cells. The animal is blind but has a highly developed sense of smell. In 2013, Ishihara's group identified neurons involved in active forgetting of an odour memory held in a different neuron, but the molecules driving that pathway remained unknown.

The experiment

The team trained worms to learn the odour diacetyl. Worms normally forget it after about four hours. When diacetyl was presented again at that point, movement towards it indicated that the memory had gone, while a lack of approach indicated that it had been retained.

Earlier work had shown that AWA neurons store the odour memory and that signals from AWC neurons erase it. The researchers set out to find the forgetting signal released by AWC. They identified a neuropeptide, NLP-47, and its receptor, GNRR-1. Tracing where these act led them to the AIZ neurons, which sit between AWC and AWA in the circuit.

A three-step relay

The pathway proposed by the authors runs as follows:

• AWC neurons release NLP-47, which activates GNRR-1 receptors on AIZ neurons.
• AIZ neurons then release the neurotransmitter glutamate.
• Glutamate acts on GLR-4 receptors on AWA neurons, the cells holding the odour memory, and the memory is erased.

Using optogenetics to suppress activity in AWC and AIZ neurons made the worms more resistant to forgetting. Worms lacking functional NLP-47, GNRR-1 or glutamate signalling also held on to the odour memory for longer.

Selectivity and links to mammals

The most notable result concerned specificity. Worms trained on both diacetyl and isoamyl alcohol lost only the diacetyl memory through the NLP-47/GNRR-1 pathway. The system therefore does not act as a general wipe of stored associations, and can single out one memory. Onishi said he had not expected to find such a concrete mechanism with this degree of selectivity.

The authors note that receptors similar to GNRR-1 occur in other species, including humans. In rodents they are found in the hippocampus, a region central to memory, although their function there is not yet known. The authors suggest the findings could help explain how mammals forget particular memories, but this remains a hypothesis.

What comes next

The study is in a nematode, and the source does not report any tests in mammals, so the relevance to human memory disorders is speculative at this stage. The immediate value lies in a fully specified circuit, from peptide to receptor to relay neuron to memory-holding neuron, which can be manipulated and probed. Open questions include what determines which memory is selected for removal, and whether the hippocampal receptors in rodents play a comparable role. Answering the second would be an early step towards judging whether this form of selective forgetting is conserved beyond worms.

The paper is titled "Neuropeptidergic-glutamatergic circuit regulates selective forgetting of olfactory learning" (Current Biology, 2026; DOI: 10.1016/j.cub.2026.10.015).

Source: No neuralyzer needed: The molecular pathways and circuit that trigger forgetting in C. elegans (Medical Xpress, 3 November 2026).

Image: AI-generated illustration.