What are the microcircuit properties that enable the prefrontal cortex to subserve cognitive functions, in contrast to early processing in primary sensory areas? In this presentation, I will discuss biophysically realistic microcircuit modeling and related experimental data that have begun to shed insights into this key issue. Recent work has led to a basic circuit mechanism that depends on slow, NMDA receptor mediated, recurrent synaptic excitation balanced by fast feedback inhibition. When recurrent connections exceed a threshold level, the network undergoes a sudden transition and becomes capable of self-sustained persistent activity for maintaining working memory. The same circuit is also well suited for decision-making computations, which typically proceed from slow accumulation of information (by long transient signals in the form of stochastic ramping neural activity) toward categorical choice formation (through attractor network dynamics). Moreover, such a circuit endowed with reward-dependent synaptic plasticity is able to account for adaptive learning in value-based economic choice behavior. Finally, I will show how this framework may be extended to describe “rule cells” and flexible rule-based tasks. Taken together, these results identify a set of core circuit properties, and suggest a common theory for understanding the prefrontal cortex and other association areas underlying cognitive functions.

 

Video registration (1h37min)

Podcast interview (43min)

 

References:

1.    Wang X-J (2001) Synaptic reverberation underlying mnemonic persistent activity. Trends in Neurosci 24: 455-463.

2.    Wang X-J (2008) Decision making in recurrent neural circuits. Neuron 60: 215-234.