PREFRONTAL MECHANISMS OF TRAINING IN WORKING MEMORY

Abstract

Working memory is the ability to maintain and manipulate information in the conscious mind over a timescale of seconds. This is considered to be centered on the prefrontal cortex—as evidenced by neurophysiological recordings in non-human primates—and further studies have demonstrated that training in related tasks may cause a variety of changes in prefrontal activity. For example, neurons in the prefrontal cortex are typically activated by different stimuli and abstract variables. A single neuron’s selectivity for a given stimulus dimension often changes depending on its context in a task, and this phenomenon is known as nonlinear mixed selectivity, which has previously been hypothesized to emerge as a result of training to perform tasks in different contexts. We therefore tested this hypothesis directly by examining the neuronal responses of different prefrontal areas before and after male monkeys had been trained to perform different working memory tasks involving visual stimulus locations and/or shapes. This revealed a modest increase in nonlinear mixed selectivity over the course of training in spatial—but not shape—working memory tasks. Moreover, to explore more deeply into the field of prefrontal training effects, we also investigated how prefrontal activity predicts rule learning across parallel spatial and object memory tasks. There is a significant gap regarding how related mechanisms may reflect behavioral improvements in different contexts, especially when examining across different tasks and modalities. This was confronted by recording single units from chronic electrode arrays implanted in the prefrontal cortex of four monkeys as they were trained to perform parallel spatial and object memory tasks with the goal of assessing the resulting activity changes that would be induced by rule learning. Progression of training thus allowed behavioral improvements to be correlated with a variety of neural effects. Some of these were unique for spatial and object learning tasks, including firing rate changes decoding information about match and nonmatch stimuli. Others, however, were shared between tasks, including a decrease in the variance explained by the reward and separation of task variables in principal component space as training progressed. Such generalizable mechanisms thus reveal that the object-spatial divide is not necessarily absolute. Altogether, our studies present both novel expansions and limitations on how changes in prefrontal activity may underlie training in working memory, highlighting specific mechanisms that may be targeted for the treatment of related impairments, the enhancement of healthy populations, and an improved understanding of human cognition.