RT Journal Article
SR Electronic
T1 Effects of Altered Excitation-Inhibition Balance on Decision Making in a Cortical Circuit Model
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 100347
DO 10.1101/100347
A1 Norman H. Lam
A1 Thiago Borduqui
A1 Jaime Hallak
A1 Antonio C. Roque
A1 Alan Anticevic
A1 John H. Krystal
A1 Xiao-Jing Wang
A1 John D. Murray
YR 2017
UL http://biorxiv.org/content/early/2017/01/16/100347.1.abstract
AB Background Disruption of the synaptic balance between excitation and inhibition (E/I balance) in cortical circuits is a leading hypothesis for pathophysiologies of neuropsychiatric disorders, such as schizophrenia. However, it is poorly understood how synaptic E/I disruptions propagate upward to induce cognitive deficits, including impaired decision making (DM).Methods We investigated how E/I perturbations may impair temporal integration of evidence during perceptual DM in a biophysically-based model of association cortical microcircuits. Using multiple psychophysical task paradigms, we characterized effects of NMDA receptor hypofunction at two key synaptic sites: inhibitory interneurons (elevating E/I ratio, via disinhibition), versus excitatory pyramidal neurons (lowering E/I ratio).Results Disruption of E/I balance in either direction can similarly impair DM as assessed by psychometric performance, following inverted-U dependence. Nonetheless, these regimes make dissociable predictions for task paradigms that characterize the time course of evidence accumulation. Under elevated E/I ratio, DM is impulsive: evidence early in time is weighted much more than late evidence. In contrast, under lowered E/I ratio, DM is indecisive: evidence integration and winner-take-all competition between options are weakened. These effects are well captured by an extended drift-diffusion model with self-coupling.Conclusions Our findings characterize critical roles of cortical E/I balance in cognitive functions, the utility of timing-sensitive psychophysical paradigms, and relationships between circuit and psychological models. The model makes specific predictions for behavior and neural activity that are testable in humans or animals under causal manipulations of E/I balance and in disease states.