I just got back from the swimming pool. When I was in the pool, I very vividly recalled my memories from my childhood when I used to swim competitively. Why did this distant, abstract experience feel so powerfully familiar? As I finished pondering this bizarre feeling, it occurred that I had been (thankfully) swimming without thinking about it the whole time. How was I able to stay afloat as if it were an unconscious habit? What motivated my decision to swim over the practically endless possible combinations of other actions I could perform? What triggered my memory to write for the Ex Aula competition this evening? Could the chance of winning a cash prize have anything to do with it?

These questions are specific examples of survival-orientated problems which the brain must constantly solve. Fundamentally, our brains exist to encode and store representations of the world and the outcomes of our actions in order to select future behavioural responses that maximise survival chances. Understanding how the brain encodes and retrieves these informationally-rich experiences to guide behaviour is the key focus of my lab. The environments we encounter are comprised of more than just spatial dimensions; different brain regions show specialised responses to rewards, emotions and decisions to name a few examples. These discretely encoded signals in the brain are married to form a neural interface for the translation of environmental inputs into learnt behavioural responses.

The focus of my PhD project is to understand how the orchestration of these information signals is hijacked during the acquisition and recall of unwanted memories: namely drug-taking experiences. Due to the powerfully rewarding properties of addictive drugs, the sequence of actions (behaviours) which lead to these pleasurable experiences become over-learnt and ultimately compulsive. The schema by which actions are reinforced based on the rewards they lead to is known as instrumental learning, which enforces the future selection of these conditioned responses. Despite being so powerfully learnt, drug-taking behaviours are not constantly engaged by addicts. Depending on the context, competing conditioned responses will be implemented when deemed more appropriate. If for example, the individual finds themselves in a swimming pool, swimming behaviour will be selected over a drug-taking one. This selection process is guided by recognition of cues in the environment (conditioned stimuli), which predict the onset of reward. It was exactly this learning regime that caused Pavlov’s dogs to salivate upon hearing the bell which Ivan Pavlov repeatedly rang prior to their meal. This motivating anticipatory response is a healthy form of that which is felt much more powerfully by drug addicts. Heroin users report immense craving sensations upon presentation of a foil chocolate bar wrapper [this being used for administration of the drug]. The acquisition of these conditioned stimulus-reward associations is hence known as Pavlovian learning.

During Pavlovian and instrumental learning, the conditioned stimulus and response that are mapped onto a mutual reward become linked via a process known as Pavlovian-instrumental transfer. The result of this inferential computation is that the conditioned response is triggered in subsequent encountering episodes of the conditioned stimulus. This permits that the most contextually-appropriate behaviour is selected. When I was swimming, I was able to consciously disengage from my actions, whilst still performing this complex behaviour. In much the same way do drug-taking behaviours manifest. Repeated epochs of drug-associated Pavlovian-instrumental transfer lead to drug-taking behaviours becoming more automatic – if the rewarding outcome is so certain, there is less of a need for conscious processes to intervene and consider alternative behaviours. Whilst habit development is useful for swimming and all other everyday behaviours, habitual drug-taking repertoires can be very dangerous.

Dopamine is released in the brain when we experience reward and is delivered to many different brain circuits which represent different types of information. These signals are thought to both facilitate instrumental and Pavlovian learning programmes and motivate the selection of future reward-promoting behaviours. Drugs of abuse pharmacologically increase dopamine and other pleasure chemicals, and/or mimic their actions. Drug-taking in addicts raises dopamine levels to a much lower extent. Dopamine response is instead tuned more to the conditioned stimulus. This observation is reminiscent of the fact that in these individuals, drug-taking behaviours have become over-learnt to the habitual level. The reduction of a dopamine response to the drug itself means that drug use is no longer motivated by the rewarding effects of the drug itself, but by drug-predicting cues (conditioned stimuli). Repeated drug use over time also leads to a desensitisation to its pleasurable effects. As a result of these two consequences of repeated drug-taking, drug use can escalate to a socially-disruptive level and to more perilous dosages.

Drug-paired memories which are encoded, cemented and retrieved by the brain form the underlying framework for the development of these pathological compulsions. Unveiling a signature of these pathological memories will uncover potential treatment targets for their perturbation; and hopefully the extinction of the damaging behaviours they encourage.

Want some coke?

I’ll stick to swimming…