The mysteries of remote memory

Abstract

Long-lasting memories form the basis of our identity as individuals and lie central in shaping future behaviours that guide survival. Surprisingly, however, our current knowledge of how such memories are stored in the brain and retrieved, as well as the dynamics of the circuits involved, remains scarce despite seminal technical and experimental breakthroughs in recent years. Traditionally, it has been proposed that, over time, information initially learnt in the hippocampus is stored in distributed cortical networks. This process-the standard theory of memory consolidation-would stabilize the newly encoded information into a lasting memory, become independent of the hippocampus, and remain essentially unmodifiable throughout the lifetime of the individual. In recent years, several pieces of evidence have started to challenge this view and indicate that long-lasting memories might already ab ovo be encoded, and subsequently stored in distributed cortical networks, akin to the multiple trace theory of memory consolidation. In this review, we summarize these recent findings and attempt to identify the biologically plausible mechanisms based on which a contextual memory becomes remote by integrating different levels of analysis: from neural circuits to cell ensembles across synaptic remodelling and epigenetic modifications. From these studies, remote memory formation and maintenance appear to occur through a multi-trace, dynamic and integrative cellular process ranging from the synapse to the nucleus, and represent an exciting field of research primed to change quickly as new experimental evidence emerges.This article is part of a discussion meeting issue 'Of mice and mental health: facilitating dialogue between basic and clinical neuroscientists'.

Keywords: ACC; consolidation; epigenetics; hippocampus; memory; remote memory.

Figure 1.

Current theories on remote memory formation and retrieval. (a) The standard consolidation theory states a linear relationship of decay in the hippocampus (HIP) and strengthening in the cortex, such as the ACC, over time, with the hippocampus unilaterally driving the mnemonic information transfer from earlier stages until the memory is completely transferred to cortical sites for its long-term storage. (b) The multiple trace theory postulates hippocampal–neocortical bidirectional interaction as early as the time of encoding as conjoint neuronal ensembles. Accordingly, the mnemonic trace is stored at multiple sites across the network, and for contextual or episodic memories, the influence of the hippocampus never decays. (c) According to the synaptic tagging hypothesis proposed here, an early distinctive synaptic or molecular signal occurs at the encoding in cortical sites and influences through as of yet unknown mechanisms the hippocampus for encoding. Such signal is critical for the formation of remote memories to persist over time. For references, please refer to the text.

Figure 2.

The relationship between structural synaptic plasticity and current theories on remote memory consolidation. (a) Findings of structural plasticity changes in alignment with the standard model of system consolidation. Recent memory elicits the formation of basal dendritic spines in hippocampal area CA1 while remote memory is associated with both apical and basal spine changes in the anterior cingulate cortex (aCC) ([, fig. 4]). (b) Findings of structural plasticity changes in alignment with the multiple trace theory of memory consolidation. Spine density changes in anterior cingulate cortex occur within hours after contextual fear conditioning (FC) ([, fig. 3]). (c) Local structural plasticity changes may not be needed for a memory to be accessible. DG engram-specific spine density at day 15 (remote memory) was significantly reduced compared with that on day 5 (recent memory), but on both days, optogenetic activation of DG engram cells induced behavioural freezing ([, fig. 3]). All figures are reproduced with permission. (Online version in colour.)