Robustness of PFC networks under inter and intra-hemispheric patterned microstimulation perturbations

A central problem in neuroscience is to understand how the dynamics of neural populations support computation. These dynamics are created both by the local neural circuitry as well as the inputs to that circuit from other brain regions. Hence, characterizing how inputs drive different population responses is key to understanding how brain circuits are modulated to flexibly produce different behaviors. To this end, we developed a patterned microstimulation (uStim) protocol that allowed us to finely manipulate the activity of a neural population in a given area by directly stimulating the same population, or alternatively, by stimulating the activity of a second population in a different area. To achieve this, we implanted two macaques with either one or two 96-channel Utah arrays and electrically stimulated combinations of electrodes. We used our protocol to study how population dynamics in the prefrontal cortex (PFC), a high-order cognitive area that displays persistent memory encoding, is influenced by both inter and intra-hemispheric inputs. For this, we assessed the impact that different uStim patterns applied to the right-hemisphere PFC (PFCR) had on the contralateral PFC (PFCL) during a memory task. We also measured the effect of direct stimulation in the PFCL. We found that both inter and intra-hemispheric stimulation patterns pushed the PFCL activity in different directions and triggered transient responses lasting hundreds of milliseconds with diverse relaxation dynamics. The relaxation brought back the responses to an attractor state which encoded the different memory conditions. Thus, PFC activity displayed robustness to a wide range of inter and intra-hemispheric input perturbations. Patterned uStim might be an important causal tool to probe the dynamics of neural populations to understand brain computation.