A paper published two weeks ago in Cell promises an exciting new therapy for PTSD, at least in mice. And it’s a little more like something out of “Eternal Sunshine” than you might think. The funny thing is, the reason for that comes down to how memory — especially learned fears such as those that cause PTSD — is written, and rewritten, in your brain. Like Amy said in a great post a while ago, memory is fluid, constantly rewriting itself. There are a two interesting gems here, for me: first, it underscores the fluidity of memory and the constant reprogramming that goes on in the brain. Second, it highlights how that fluidity and reprogramming are actually essential parts of brain function — a brain of steel is unhealthy compared to one of rubber. Finally, it shows how so many promising drugs are discovered through an odd mixture of happenstance and curiosity, fueled by basic research.
Write Your Own History
So how does memory work, anyway?
Or, in other words, how is cellular memory (epigenetics) leveraged in the brain to create organismal memory (either episodic memory like people have or learned fears as can be seen in mice)?
Many of the details are still being studied, and we’re not exactly to a point where we can target specific cells as being responsible for specific memories other than asking a research subject to retell the story of a memory and measure which cells get more blood flow. Which is a rather blunt instrument in general. But we have a bit of a start, especially in thinking about which chemical reactions are necessary to successfully remember something.
A large number of epigenetic changes take place in neurons in response to a stimulus and in the process of encoding that stimulus and response into memory. It comes down to two groups which are added to either DNA or histones: an acetyl group (two carbons, three hydrogens, and an oxygen) and a methyl group (one carbon and three hydrogens). Short term, histones get acetylated in response to a neuron firing. The presence of these acetyl groups triggers other enzymes to add in methyl groups, either directly to DNA or to histones. The acetyl groups are short-lived; taken off by specific enzymes called HDACs, but the methyl groups stay longer. More specifically, in the first hour of response to a stimulus, histones in certain neurons are probably modified by adding an acetyl group, and this new pattern is transcribed into more long-lived marks such as DNA methylation and histone methylation over the next day or so. (For a little bit more about what these modifications are and how they work in the cell, try wikipedia for histones or DNA methylation.)
These methyl marks can change expression of nearby genes, and they stay on the genome for a while, so if the same stimulus comes up, and the same neurons are firing, they’ll be primed to react in the same way. Hence, a memory.
Most interestingly (to me) is that many of these same pathways are activated when we remember things, not just when we learn new things. Which means that, practically, every time you remember something, you’re rewriting the epigenetic marks that cause you to have that memory in the first place.
This has two effects:
First, small changes can come in (neural plasticity). Because no copier is perfect, and your memories are basically copies of copies of copies.
Second, each time you recall something, it gets engraved a little bit deeper. It gets harder to write over. It might not stay exactly the same (no copier is perfect), but it’ll stick in there better.
That engraving has a lot to do with the HDAC enzymes I talked about above. The crucial bit of information in this recent study, and the central breakthrough for it, is that over time, as we rehash and revisit memories and fears, proteins called HDACs (Histone De-ACetylases) are turned on. The more times we rehash that memory, the more HDACs the neurons involved in it will have. With lots of HDACs in a cell, that acetyl group will get taken right back off. So there’s no time to convert those acetyl groups into more lasting marks, and the new response doesn’t get remembered.
As a side note, this provides a wonderful biochemical demonstration of confirmation bias: you’re more likely to disregard conflicting evidence for things that you “know” — things that you rehash all the time — because those memories are associated with cells with higher levels of HDACs.
And, since biology is an empirical science, we know this because when we treat mice or rats with HDAC inhibitors, they tend to be better at learning new things. Just like we know that DNA methylation and histone methylation are the marks that get added as acetylation gets removed because when we treat with inhibitors of those processes, the acetylation signal doesn’t go anywhere, and a memory can’t be formed.
HDAC inhibitors have been suggested as treatment for memory disorders like Alzheimer’s and Huntington’s since 2008, and they show some promise. But histone acetylation and deacetylation doesn’t just effect the brain, it’s used in every part of the body to control gene regulation. So drugs like this need a lot of refinement before being used continually and regularly to boost memory formation.
Relearning is ReLearning
But, once again, the process of even thinking about a memory, much less being reminded of it by an outside stimulus, triggers many of the same neurons (and epigenetic pathways) that the initial stimulus and response did. In fact, if anything, it’s irrelevant what the initial stimulus and response did, since in disorders like PTSD a patient is responding to the memory of the event more than the event itself.
Which gives clinicians and researchers a golden opportunity: HDAC inhibitors need a lot of testing before anyone would feel comfortable suggesting taking them every day for the rest of your life. They may not be the best option for generally making it easier to learn, because they effect so many other things. But what if we just want to change one thing?
What if we just want to relearn a specific memory?
Which is, in essence, the coolest part about this study, and what the researchers did. They used HDAC inhibitors — which we’ve known for a few years boost plasticity and learning — in a very short time-span, and used that time to bring up a conditioned fear response. And the mice forgot their fear, with few or no side effects.
Because the HDAC inhibitor treatment didn’t last very long, it was unlikely to mess up other pathways. But because researchers were able to use that time specifically to rewrite a certain memory — and not others — the now inaccurate fear response was overwritten.
And when you think about it, in a certain way, this is exactly the same idea as the plot device in Eternal Sunshine of the Spotless Mind: take a pill before bed, and that night remember everything you want to forget; for the last time.