Formation and maintenance of robust long-term information storage in the presence of synaptic turnover

Abstract

A long-standing problem is how memories can be stored for very long times despite the volatility of the underlying neural substrate, most notably the high turnover of dendritic spines and synapses. To address this problem, here we are using a generic and simple probabilistic model for the creation and removal of synapses. We show that information can be stored for several months when utilizing the intrinsic dynamics of multi-synapse connections. In such systems, single synapses can still show high turnover, which enables fast learning of new information, but this will not perturb prior stored information (slow forgetting), which is represented by the compound state of the connections. The model matches the time course of recent experimental spine data during learning and memory in mice supporting the assumption of multi-synapse connections as the basis for long-term storage. Author SummaryIt is widely believed that information is stored in the connectivity, i.e. the synapses of neural networks. Yet, the morphological correlates of excitatory synapses, the dendritic spines, have been found to undergo a remarkable turnover on daily basis. This poses the question, how information can be retained on such a variable substrate.In this study, using connections with multiple synapses, we show that connections which follow the experimentally measured bimodal distribution in the number of synapses can store information orders of magnitude longer than the lifetime of a single synapse. This is a consequence of the underlying bistable collective dynamic of multiple synapses: Single synapses can appear and disappear without disturbing the memory as a whole.Furthermore, increasing or decreasing neural activity changes the distribution of the number of synapses of multi-synaptic connections such that only one of the peaks remains. This leads to a desirable property: information about these altered activities can be stored much faster than it is forgotten. Remarkably, the resulting model dynamics match recent experimental data investigating the long-term effect of learning on the dynamics of dendritic spines.

Silent Synapses in Cocaine-Associated Memory and Beyond

“…Within a relatively homogeneous microenvironment, synapses in a compound connection may share similar spatial and temporal activation profiles. In this case, although activation of individual synapses is stochastic, the integrated activity of these synapses, referred to as the collective dynamics, is predicted to be strikingly stable, resilient to the probabilistic failure or reasonably paced synaptic turnover (Fauth et al, 2015). As a theoretical extrapolation, compound connections may serve as the basic functional modules for a synaptic engram to support stable circuit dynamics for a memory.…”

Intrinsic spine dynamics are critical for recurrent network learning in models with and without autism spectrum disorder

“…Another study explored the consequence of volatile synaptic strengths in recurrently connected networks without studying how such volatility affects activity-dependent plasticity (Mongillo et al 2018) . Other studies model stochastic changes of synaptic strength (Loewenstein et al 2011) or connectivity (Deger et al 2012;Fauth et al 2015) without distinguishing activity-dependent and -independent parts. Hence, these works do not distinguish their separate roles in memory and synaptic normalization.…”

A neuromimetic approach to the serial acquisition, long-term storage, and selective utilization of overlapping memory engrams

“…Our biomimetic strategy for permanent storage is permanent stabilization of the pertinent synapses; accordingly, in both our network and biologically-embedded ones, memory acquisition correlates with accretion of permanent synapses. Our confidence in this strategy is reinforced by the contrast between models that employ slowly-changing synapses (5,6,32) and models whose synapses have state descriptor variables that explicitly tend to stay constant after initial learning (33,34). A second challenge concerns a different type of interference, anterograde, in which a network's previously-acquired memories hinder the subsequent acquisition of memories that a tabula rasa network would have no trouble acquiring.…”