Synaptic tagging during memory allocation

Abstract

There is now compelling evidence that the allocation of memory to specific neurons (neuronal allocation) and synapses (synaptic allocation) in a neurocircuit is not random and that instead specific mechanisms, such as increases in neuronal excitability and synaptic tagging and capture, determine the exact sites where memories are stored. We propose an integrated view of these processes, such that neuronal allocation, synaptic tagging and capture, spine clustering and metaplasticity reflect related aspects of memory allocation mechanisms. Importantly, the properties of these mechanisms suggest a set of rules that profoundly affect how memories are stored and recalled.

Figure 1

a) Neurons with increased cyclic AMP-responsive element-binding protein (CREB) levels (green 'halo' in ensemble 2) are more excitable and therefore more likely to fire postsynaptic action potentials in response to presynaptic action potentials and are more likely to be involved in synaptic changes underlying memory. b) Thus, these these neurons are more likely to be over-represented in the memory trace ('memory-positive' neuron). c) Synapse-specific potentiation results in the local diffusion of plasticity-related proteins (for example, RAS and RHOA) from an activated synapse and a local enhancement in excitability. d) This increased excitability promotes plasticity in nearby synapses for a brief period of time. e) This results in clustering of potentiated synapses in close proximity to previously activated synapses. L-LTP, late long-term potentiation.

Figure 2

a) During acquisition, neurons in a neural circuit (grey circles) are recruited into encoding episode A (blue). This increases their excitability so that shortly thereafter, they are also very likely to be involved in encoding episode B (purple). b) With time, the increase in excitability wanes and sequent episodes are no longer stored in the same neurons. A consequence of this pattern of storage is that recall of episode B will also result in the recall of episode A (and vice versa), whereas recall of subsequent episodes will be unaffected. Figure is adapted, with permission, from Ref. © (2009) American Association for the Advancement of Science.

Figure 3

a) The synaptic tagging and capture hypothesis proposes that the synapses activated during early long-term potentiation (E-LTP) induction (as depicted by the presence of presynaptic action potentials) become tagged in a protein synthesis-independent manner that involves calcium/calmodulin-dependent protein kinase II (CaMKII) and actin (not shown). b) These tagged synapses then capture plasticity-related proteins (PRPs) downstream of the CaMKII–CaMKIV–cyclic AMP-responsive element-binding protein (CREB) pathway, which is needed for the maintenance of LTP and, by extrapolation, long-term memory linked to ensemble 1. c) The formation of the strong memory A (indicated by turquoise shading) induces late LTP (L-LTP) in a subset of synapses in neuronal ensemble 1 (depicted as a single neuron for clarity) but not in neuronal ensemble 2 (not shown). For at least 1 hour after strong training, neuronal ensemble 1 is able to share plasticity-related proteins that can convert a weak memory into a strong one ('synapse-specific potentiation'). d) An arriving action potential (memory B) at the top right synapse of ensembles 1 and 2 sets a new synaptic tag. e) Subsequently, the weak memory B (depicted in part d) is able to elicit L-LTP in neuronal ensemble 1 (owing to the presence of plasticity-related proteins) but not in neuronal ensemble 2. Ensemble 1 therefore becomes positive for memory A and memory B (indicated by dual blue and purple shading). This is an example of how synaptic tagging and capture can determine which neurons (not just synapses) would encode a given memory.

Figure 4

Neuron 1 is strongly activated (depicted by multiple red action potential traces), which leads to the formation of synaptic tags involving calcium/calmodulin-dependent protein kinase II (CaMKII) and actin in stimulated synapses. Diffusion of RAS and RHOA (indicated by green shading) from the activated synapses promotes plasticity in nearby synapses (~10 µm) for a brief period of time (<10 minutes). Production of plasticity-related proteins (PRPs) in the postsynaptic neuron is needed for stable synaptic strengthening, and these PRPs are shared by tagged synapses. In response to a single stimulation bout (single red trace), presynaptic neuron 2 fires weakly within 10 minutes, resulting in the formation of synaptic tags. Synapses closer to those that are strongly stimulated in neuron 1 will be more likely to be potentiated, resulting in synaptic clustering.