Cognitive neuroscientists need not abandon their understanding of an evolutionary continuum between humans and all other creatures when faced with the reality of our many unique abilities. A complete explanation of our singular intelligence may yet lay beyond our grasp, but we know that the answer will likely be revealed as we unravel the twisted circuits of the prefrontal cortex. No other part of our brains evolved more radically as we became Homo Sapiens, and a superficial reading of everything from Freudian repression to moral reasoning will appeal to the “involvement of the prefrontal cortex.”
Although we have accumulated only a small bit of knowledge about this part of our brains, we are beginning to appreciate how its function underlies higher cognitive abilities such as executive control. A fundamental aspect of the prefrontal cortex (PFC) is its connectivity: neurons in PFC integrate highly-processed streams of information from a diverse range of cortical areas into representations that can guide voluntary behavioral strategies. One unanswered question is how information about rewards influences the working-memory traces in the PFC. On laboratory tasks that involve several possible rewards, animals will tend to perform better on trials when they know they will be receiving a reward of greater subjective value. Understanding how the PFC brings together information about what an animal is supposed to do and what it will get if it does it will help us understand this effect.
In this week’s Journal of Neuroscience, Steven Kennerley and Jonathan Wallis report a study investigating the interaction of reward and spatial working memory in the lateral prefrontal cortex of the rhesus monkey. They trained their monkeys on two tasks that required them to remember both a spatial instruction and a reward cue over several seconds before performing an eye movement to the instructed location so as to receive the reward (a squirt of fruit juice, with one of five different volumes being available for each trial). On one version of the task, the monkeys first saw a picture indicating how much juice they would be working for; after that cue disappeared, a spot was illuminated at one of 24 potential locations in the monkey’s visual field. After another delay, the monkeys were cued to look at the remembered location of the spatial instruction and, if correct, were rewarded with the juice. In the second version of the task, the order of the instructions was reversed.
Kennerley and Wallis recorded from neurons in the dorsolateral (dlPFC) and ventrolateral (vlPFC) regions of the prefrontal cortex as the monkeys performed these tasks. Traditionally, dlPFC has been thought to be responsible for representing spatial information in working memory, while vlPFC is thought to encode reward information. The results of this experiment, however, revealed that the ventral region produced a greater response both to reward and to spatial instructions. Kennerley and Wallis reached this conclusion based on three main variables: which neurons responded to the cues earlier, which responded more strongly, and which sustained their activation better over the delays preceding the behavioral response. Across all three measurements, vlPFC outperformed dlPFC. Furthermore, individual vlPFC neurons appeared to integrate the two modalities of information into a single signal in the period directly preceding the eye-movement.
This single signal, which encodes a stronger representation of spatial information when making the right eye-movement will yield a larger reward, could underlie the monkey’s improved performance on those high-value tasks. Kennerley and Wallis thus suggest that the ventral prefrontal cortex might be more involved in directing attention than previously thought. In general, though, this experiment provides further support for the hypothesis that dorsal and ventral regions of the PFC perform distinct roles as they guide cognitive tasks. Going forward, elucidating this distinction will allow us to understand how the prefrontal cortex generates ever more abstract and complex behaviors.
Kennerley, S., & Wallis, J. (2009). Reward-Dependent Modulation of Working Memory in Lateral Prefrontal Cortex Journal of Neuroscience, 29 (10), 3259-3270 DOI: 10.1523/JNEUROSCI.5353-08.2009
