Abstract
That moment when suddenly, everything makes sense. The first time the relationship between our behaviour and consequences comes into sharp relief. This moment of abrupt associative learning has been at the periphery of psychological and neurobiological studies of learning for decades. But no investigations have been able to definitively identify its neurobiological correlates. Historically, investigations have tended to focus on an aggregate measure of learning which obscures individual learning or are conducted in preparations that have little to do with the actual learning context (in-vitro physiology). These methods have made conclusions regarding mechanisms of associative learning difficult.
More recently, recording of neural ensembles in prefrontal areas has revealed neuronal activity that closely tracks, and in some cases predicts, subsequent behavioural performance in situations with rapidly-changing rules. While these studies indicate population-level dynamics that underlie rapid learning of changing rules, they do not shed light on the initial moment when a new association is acquired.
To shed light on this mechanism we used a classical conditioning protocol paired with a statistical algorithm that allowed us to quantitatively specify when a rat has learned the association between a conditioned stimulus (CS) and its outcome (delivery of a food pellet) also known as unconditioned stimulus (US). We quantified the phosphorylation of ERK, a signaling molecule related to learning and synaptic plasticity, in the central amygdala, a region of the brain involved in behavior driven by the value of a stimulus during normal associative learning and during aberrant behavior towards addictive stimuli. We found that rats that had experienced associative learning expressed significantly more p-ERK than those that did not or those that were overtrained following acquisition. Thus, learning in our paradigm is accompanied by cellular changes that transiently peak when the association between CS and US has been established and then subside following it.
We then used Gradient-index (GRIN) lenses paired with the UCLA’s miniscope technology to try to image changes in the influx of calcium ions (Ca2+) in the moments leading to the activation of ERK and the formation of associative learning in rats. Our hypothesis for this experiment was that we could capture the instant when populations of neurons responding to the food pellet started to respond to the predictive sound, signaling that the association between CS and US had formed in the brain of these animals.
Although this last experiment has not produced the results we were hoping for, due to COVID disruptions and a long process of troubleshooting, it has rewarded us with a new knowledge about an interesting and promising new technology and an invaluable set of new skills. Indeed, we can now say that in future research we will be able to image deep brain regions of freely behaving rats during associative learning tasks.