The pair found that the level of DNA-methylating enzymes (DNMTs) increased in the hippocampus when rats learned a conditioned response. For conditioning, rats were placed in an unfamiliar chamber, left to explore for a few minutes, and then given an electric shock and removed from the chamber. When the rats were put back in the chamber a day later, they froze in anticipation of the shock.
This associative learning was forgotten, however, when the rats were given a DNMT inhibitor immediately after their conditioning. The normal increase in DNMTs upon learning was coupled with an increase in DNA methylation and reduced transcription at a memory suppressor gene. In the presence of the DNMT inhibitor, the suppressor gene remained active.
Despite the increase in DNMT activity during conditioning, a gene that promotes neuronal plasticity was demethylated and thus exhibited increased transcription. This finding came as a bit of a surprise, since DNA methylation was considered to be a stable epigenetic modification in adult tissue.
In the presence of the DNMT inhibitor, the plasticity gene was further demethylated and thus further activated. Since memory was nonetheless impaired, the authors suggest that repressing the memory suppressor by methylation might be the more important cellular event during learning.
Both the methylation of the memory suppressor and the demethylation of the plasticity promoter were apparent just one hour after conditioning and returned to baseline levels by the next day. The lack of permanent change to the hippocampus, explains Sweatt, might be related to the fact that this part of the brain is associated with memory consolidation rather than memory storage. The team now plans to look at the cortex, where memories are stored, to see whether methylation states there become as fixed as our memories.