Francis Crick and the long-term storage of the memory trace

Since my Ph.D. is related to memory consolidation, I was interested in a strange idea from Francis Crick. He asked the question of long-term storage of the memory trace 1. How is this memory trace stored in our brain? And, more importantly, how is it protected against molecular turnover? In his view, Crick suggested three hypothesis:

  1. Memory could be encoded in alterations of some part of the cell DNA. This will imply that each neuron synapse would be represented by a part of the neuron DNA since the actual paradigm states that memory is encoded in the strength of individual synapse. This first hypothesis seems unlikely.
  2. Memory could otherwise be stored in a local piece of DNA or RNA, at the synapse (a bit like the mitochondrion has its own DNA). This piece would be immune to the molecular turnover. Although more logical, this hypothesis seems unlikely too.
  3. Finally, Crick’s last hypothesis states that molecules at the synapse level would interact in such a way they could be replaced by new ones, one at a time, without altering the general status (strength). The figure below shows a working example of this hypothesis …

In this figure, two monomers (squares) forms a hypothetical protein
highly involved in a memory process at the synapse level. Each monomer can be in two states: active (plus sign) or inactive (minus sign). Activation of the monomer could be done by phosphorylation (in this example ; any other modification could be applied here). The hypothetical protein can either be active (plus plus) or inactive. The key point in Crick’s hypothesis is an enzyme which will phosphorylate a monomer if the protein is in state (plus minus), giving an active (plus plus) protein, but not if it is in state (minus minus). This will counteract the molecular turnover which transform an active (plus plus) protein into an inactive (plus minus) one.

Of course, Crick’s hypothesis can be extended to proteins that are trimers, tetramers, … other process than phosphorylation could be used (methylation, glycosylation, …) and more conditions could also be added (anchoring, maturity, …).

What do you think of this hypothesis?

1 Crick F. “Neurobiology: Memory and molecular turnover” in Nature 312:101 (1984) – read the PDF