Towards biologically constrained attractor models of schizophrenia

Abstract

Highlights• People with schizophrenia have quantitative working memory (WM) deficits • WM deficits cannot be readily mapped on the stability of continuous attractors • Different attractor models yield inconsistent behavioral predictions upon changes in E/I ratio • Grouping different perturbations under E/I ratio can be inadequate for some deficits • Insufficient experimental and computational understanding precludes model selection • We need data showing NMDAR-dependence of rate tuning and correlated noise in WM circuits

Cholinergic neuromodulation of prefrontal attractor dynamics controls performance in spatial working memory

“…Weaker tuning in our continuous attractor network framework can lead to better performance when it is concomitant with a reduction of the noisy diffusion of the memory traces (Pouget et al, 1999; Zhang and Sejnowski, 1999b; Butts and Goldman, 2006; Ma et al, 2006b; Stein et al, 2021). How these two components are integrated into attractor dynamics depends on specific network implementations.…”

“…The marked impact that connection heterogeneities have on memory diffusion made us consider how more realistic networks might respond to changes in excitability. In these networks, the conditions to observe lower memory diffusion following increased excitability may depend on the specific biophysical mechanisms that help maintain a quasi-continuous attractor dynamical regime (Stein et al, 2021). We investigated these mechanisms in a biologically realistic model of spiking neurons.…”

“…Furthermore, WM is affected in most neuropsychiatric disorders (Forbes et al, 2009). However, the causal chain from cellular neuromodulation to changes in network dynamics and behavioral performance in WM remains largely unknown, even for simple behavioral readouts, such as spatial memory precision in delayed-response tasks (Stein et al, 2021).…”

Myelin dystrophy in the aging prefrontal cortex leads to impaired signal transmission and working memory decline: a multiscale computational study

“…In addition to age-related changes in memory duration and precision, our network model predicts an age-related increase in systematic errors due to an increased drift of the activity bump ( Supplementary Figure 8 ). Delay-dependent systematic working memory errors have been observed in behavioral experiments ( Bae, 2021; Panichello et al, 2019; Stein et al, 2021 ) and it would be interesting to test whether those biases also change with aging. While the prefrontal cortex is certainly a key player in working memory, it is not the only brain area involved in this function; working memory is likely sustained by the interaction of several fronto-parietal brain areas ( Leavitt et al, 2017 ; Christophel et al, 2017 ).…”

“…Moreover, if demyelination is spatially localized in a part of the network, the model predicts a repulsive bias away from the memories encoded in the affected zone ( Supplementary Figure 7 ). Delay-dependent systematic working memory errors have been observed in behavioral experiments ( Panichello et al, 2019 ; Bae, 2021 ; Stein et al, 2021 ) and it would be interesting to test whether those biases also change with aging. In addition to the prefrontal cortex, working memory is likely sustained by interactions between several fronto-parietal brain areas ( Leavitt et al, 2017 ; Christophel et al, 2017 ).…”