Aerial respiration is mediated by a 3-neuron CPG whose sufficiency and necessity has been demonstrated. However, the output of this ‘hard-wired’ CPG is ‘plastic’. For example, aerial respiratory behaviour can be operantly conditioned, a form of associative learning, and it exhibits long-term memory. With the passage of time the memory weakens. We hypothesize that this forgetting, or memory transience, is also a form of learning.

To test this hypothesis we have performed a number of different experimental procedures following the establishment of memory. (1) Cold shock (intact snails are maintained at 4 °C) blocks new protein synthesis, and snails subjected to cold shock have their memories significantly extended. (2) If we remove the soma of RPeD1, one of the CPG neurons, after memory has been formed, forgetting does not occur. We remove the soma in such a way as to not damage the primary neurite (where all synaptic activity occurs) and the primary neurite remains functional for up to 3 weeks. However, since there is no longer any nucleus, altered gene activity, which we hypothesize to be necessary for forgetting, cannot occur. (3) If a new association is prevented (snails are not allowed to perform aerial respiration and thus the association of opening the pneumostome and no reinforcing stimulus) memory is also extended. All the results are consistent with the hypothesis that altered gene activity and new protein synthesis have to occur in order to forget. Forgetting of a learned behaviour thus appears to be a memory of a new learned behaviour that resembles the na•ve state.

Research Symposium Ð Control and Modulation of Respiratory SystemsResearch Symposium Ð Control and Modulation of Respiratory Systemsnt with the hypothesis that altered gene activity and new protein synthesis have to occur in order to forget. Forgetting of a learned behaviour thus appears to be a memory of a new learned behaviour that resembles the na•ve state.