Recurrent activity propagates through labile ensembles in macaque dorsolateral prefrontal microcircuits.

Human and non-human primate studies clearly implicate the dorsolateral prefrontal cortex (dlPFC) as critical for advanced cognitive functions.(1)(,)(2) It is thought that intracortical synaptic architectures within the dlPFC are the integral neurobiological substrate that gives rise to these processes.(3)(,)(4)(,)(5)(,)(6)(,)(7) In the prevailing model, each cortical column makes up one fundamental processing unit composed of dense intrinsic connectivity, conceptualized as the "canonical" cortical microcircuit.(3)(,)(8) Each cortical microcircuit receives sensory and cognitive information from upstream sources, which are represented by sustained activity within the microcircuit, referred to as persistent or recurrent activity.(4)(,)(9) Via recurrent connections within the microcircuit, activity propagates for a variable length of time, thereby allowing temporary storage and computations to occur locally before ultimately passing a transformed representation to a downstream output.(4)(,)(5)(,)(10) Competing theories regarding how microcircuit activity is coordinated have proven difficult to reconcile in vivo, where intercortical and intracortical computations cannot be fully dissociated.(5)(,)(9)(,)(11)(,)(12) Here, using high-density calcium imaging of macaque dlPFC, we isolated intracortical computations by interrogating microcircuit networks ex vivo. Using peri-sulcal stimulation to evoke recurrent activity in deep layers, we found that activity propagates through stochastically assembled intracortical networks wherein orderly, predictable, low-dimensional collective dynamics arise from ensembles with highly labile cellular memberships. Microcircuit excitability covaried with individual cognitive performance, thus anchoring heuristic models of abstract cortical functions within quantifiable constraints imposed by the underlying synaptic architecture. Our findings argue against engram or localist architectures, together demonstrating that generation of high-fidelity population-level signals from distributed, labile networks is an intrinsic feature of dlPFC microcircuitry.