Exposing the memory engine: the story of PKMzeta

View Images

[This is the second of three intertwined posts on PKMzeta – the molecule that keeps our memories intact]

You’ve got the phone number of a hot date – a vital piece of information that you need to keep in a safe place. You write it in a notepad, you save it on a file in your computer and you try to commit it to memory. This third method – the one involving your brain – is very different to the others.

In the other formats, stability is the norm. The ink on the paper won’t vanish (at least not for centuries). The magnetic information on the hard drive won’t spontaneously rearrange itself. Unless either material suffers physical damage, the information recorded within them will stand the test of time. In your brain, the fate of information is much less certain.

In the last decade, scientists have found that it takes active and unrelenting effort to keep our memories intact. Even long-term memories are constantly on the verge of being erased. To keep them stable, we need to continually recreate a protein called PKMzeta. This molecule is the engine of memory, constantly whirring to store information in our brains. Give the engine a boost, and old memories gain a new lease on life. Switch it off, and we forget things…. permanently.

When we learn new things, PKMzeta shows up at the gaps between neurons (synapses) and boosts the signals that go across them. This strengthens the connections between the neurons on either side, and this network of bolstered connections is the physical embodiment of our memories. “A synapse with PKMzeta is twice as strong as it otherwise would be,” says Todd Sacktor from SUNY Downstate Medical Center, who has led the way in uncovering the protein’s secrets. (See “Todd Sacktor talks about the memory engine)


The idea that proteins could help to store memories is an old one. Francis Crick, one of three scientists who discovered the structure of DNA, suggested it way back in 1984. But by the start of the 21st century, it had fallen out of favour. Scientists had found many molecules that were important for creating long-term memories, but not for storing them. Once memories were stable, you could knock out these molecules to little effect.

At the time, the prevailing idea was that memories were stored through the growth of new synapses. Networks of neurons became more strongly connected by increasing the number of links between them, like towns that build more roads between them.

But in the last five years, Sacktor’s research has challenged this view. In 2006, he showed that PKMzeta is vital for storing long-term spatial memories in the hippocampus. The team managed to erase recently stored memories in the brains of rats by injecting them with ZIP, a chemical that neutralises PKMzeta. A year later, together with Yadin Dudai and Reut Shema at Israel’s Weizmann Institute of Science, Sacktor used ZIP to erase a rats’s dislike for an artificial sweetener, by injecting a different part of the brain involved in taste. It even erased very strong memories that had been stored months before, and it did so irreversibly.

These were amazing results. Now, the team have sealed their case by doing the opposite – boosting old memories by loading rats with extra PKMzeta. They injected the rodents with viruses carrying the protein, before teaching them to avoid the taste of sweetener. With extra copies of PKMzeta in their brains, the rats were more likely to remember their distaste. Even if the team injected the viruses a week after the rats’ training, when their aversion to the sweetener had started to fade, it enhanced their dulled memories.

“Multiple old memories were robustly enhanced. These results have no precedent,” says Sacktor. Other groups have found ways of strengthening new memories, from sleeping to repeating the same information. Soon, we may have drugs that do the same thing. But once our memories shift to long-term storage, PKMzeta is one of a few molecules that can reinforce them, and the only one that works in healthy individuals. (See “Single protein can strengthen old faded memories“)

If you can erase or boost old memories by tweaking a single protein, then memories can’t be stored just by growing new synapses. Instead, Sacktor thinks that we store new memories by increasing the strength of synapses rather than their number, like towns that add extra lanes to the roads between them, rather than building new ones. “It’s a real revolutionary change in how neuroscientists have thought about memory,” says Sacktor.


As memories are formed, synapses are busy places. The gene that provides instructions for making PKMzeta is always on, but in most neurons, there’s a blockade that prevents these instructions from being followed. It takes a large committee of signalling molecules to lift the blockade, start the production of PKMzeta, and ignite the memory engine.

Once it’s made, PKMzeta probably only lasts for a matter of days. So our synapses need to constantly replenish their supply of this protein, if we’re to keep stable long-term memories. Fortunately, a series of looping chemical reactions ensures that once neurons continue producing PKMzeta, they don’t stop. Once the memory engine starts whirring, it can carry on indefinitely. If the experiments with rats are a guide to what’s happening in people, “PKMzeta is staying in the same synapse probably for decades,” says Sacktor. “It’s not the same molecule but the population is being maintained at a high level for maybe a hundred years.”

PKMzeta works by increasing the levels of AMPAR, a protein that sits at synapses and allows fast signals to travel across them. Normally, AMPAR is caught in a tug-of-war between proteins that try to drag it towards the synapse and others that drag it away. PKMzeta swings the battle in favour of the former group. When it’s around, AMPAR moves towards the synapse in great numbers. Each arrival strengthens the synapse.

This is a constant battle. Other proteins are always trying to drive AMPAR away from the synapse, so PKMzeta has to fight to keep it there. This is why it’s so easy to erase memories with ZIP. If you get rid of PKMzeta, the tide of battle turns, AMPAR is driven away, the synapse weakens, and memories are forgotten.

The implications of this are staggering. It means that your brain is constantly on the verge of erasing itself. It’s the perpetual drive of the memory engine that prevents it from being rewritten back to a blank state. “Our default state is no knowledge!” says Sacktor. “Empty nirvana is our preference!”

This isn’t just a human thing either. David Glanzman and Wayne Sossin have shown the California sea slug has its own version of PKMzeta and if you block it, lo and behold, you erase the animal’s long-term memories. The same goes for the fruit fly. The memory engine is an ancient evolutionary invention.

It’s also universal to different types of memory. By using ZIP, Sacktor and others have managed to erase all sorts of long-term memories, from fears to locations to physical skills. “It applies to all parts of the brain that store different types of memory like the hippocampus that stores place information, or the amydgala that stores fear memories, or motor memories in the motor strip,” says Sacktor. “They’re all using PKMzeta.”

At first glance, this seems like an odd state of affairs. Memories are incredibly important, so why are they always teetering on the edge of disappearance? It probably has something to do with flexibility. The vulnerable nature of our memories allows us to easily update our entire network with new information. Without this flexibility, we’d be incapable of learning new things – a flaw that’s just as dangerous as the threat of memory loss.

The simple nature of the PKMzeta system might makes it even easier to continuously update our memories. “Information storage systems, no matter what you look at, tend to be quite simple,” says Sacktor. A hard disk is using just one type of storage – magnetic charge on a spinning disc. Of course, the information that’s encoded within the pattern of zeroes and ones is incredibly complicated.

“By analogy, it’s the structure of the entire brain that’s incredibly complicated and that gives you the type of memory that’s being stored by PKMzeta. For example, the amygdala is important for fear, the motor cortex is important for movement and the visual cortex is important for seeing. Each of those is going to have its own complex language, but ultimately, the long-term store is still going to be a simple pattern of zeroes and ones – PKMzeta or not. The analogy for DNA is that you’ve got four bases – it could have been hundreds, but it’s a small number.”


This story is far from complete. We know a lot more about the memory engine than we did even a decade ago, but the list of unanswered questions is vast. When we remember something, our memories once again enter a fragile state when they can be edited or overwritten – is this because PKMzeta must be destroyed and created afresh? How does newly formed PKMzeta manage to find synapses that are already tagged with this protein? What stops all the neurons in our brain from becoming saturated with PKMzeta? How does sleep affect the levels of PKMzeta at synapses? And does the memory engine start having problems as we grow older?

Most of all, can we tweak the engine to either get rid of unwanted memories? Can we manipulate it to boost old, faded ones, in people with dementia? And does playing with memory open an ethical Pandora’s box? I was initially worried too if it somehow fell into the wrong hands,” says Sacktor. “[But] it’s not just about the dystopian fantasies of making zombies or toying with people’s memories. I think the actual good is going to far outweigh the potential for bad.

The benefits go beyond the obvious candidates like traumas and addictions. “There’s a condition called central neuropathic pain syndrome,” says Sacktor “where people catch their finger in the car door and even after the injury heals, a memory for the pain is set up in the central nervous system. ZIP could erase that too.” Indeed, Min Zhuo from the University of Toronto managed to alleviate this type of pain in mice by injecting them with ZIP.

But for Sacktor, the true value of his work in PKMzeta is in revealing how our brains work. The practical benefits follow on from that. “We have the first inkling of how long-term memories and information are stored in the brain. That’s an important thing for understanding how we’re humans. We are our memories – our mental states are based upon everything we’ve learned. You can’t do the technological side without doing the basic research first. You can’t hope to treat addiction or post-traumatic stress disorder in a fundamental way until you really know how these processes work.”



  • Pastalkova, E. (2006). Storage of Spatial Information by the Maintenance Mechanism of LTP Science, 313 (5790), 1141-1144 DOI: 10.1126/science.1128657
  • Shema, R., Sacktor, T., & Dudai, Y. (2007). Rapid Erasure of Long-Term Memory Associations in the Cortex by an Inhibitor of PKM Science, 317 (5840), 951-953 DOI: 10.1126/science.1144334
  • Sacktor, T. (2010). How does PKMζ maintain long-term memory? Nature Reviews Neuroscience, 12 (1), 9-15 DOI: 10.1038/nrn2949
  • Migues, P., Hardt, O., Wu, D., Gamache, K., Sacktor, T., Wang, Y., & Nader, K. (2010). PKMζ maintains memories by regulating GluR2-dependent AMPA receptor trafficking Nature Neuroscience, 13 (5), 630-634 DOI: 10.1038/nn.2531

More on PKMzeta

More on memory: