Learned expectations are not always reliable
The bus doesn’t always arrive on time and your favourite dish is not always on the menu at the local restaurant. To account for an ever-changing world, animals must possess neural mechanisms to constantly update their learned predictions so that their choices correctly incorporate the new, sometimes contradicting information.
A recent study from the Waddell lab in Nature reveals how the strength of memory-directed behavior is either suppressed or maintained depending on whether a new experience opposes or fits a previously learned expectation.
Different dopaminergic neurons in the Drosophila brain work in support or opposition to a previous memory of food reward and so allow the fly to accumulate information from all experiences to optimize behaviour.
If reward is not present when expected, initial reward memories are suppressed, or extinguished, by the formation of parallel competing aversive memories. Alternatively, the initial reward memories can be maintained following a period of reconsolidation, during which they are susceptible to disruption.
Pharmacological intervention during memory reconsolidation has potential application for the treatment of troublesome memories in humans.
The full paper can be viewed here.
Johannes Felsenberg discusses the work in a short film, entitled Malleable Memories
The work was highlighted in PNAS Journal Club