Memory is a complicated process, one that involves further experiences, changes in gene expression, and the structure of nerve connections. Forming a memory will clearly involve a variety of processes and proteins, but over the last several years, a variety of evidence has pointed to one protein appearing to be key to the formation of memories: protein kinase M ζ.
Drugs that inhibit this protein could wipe out long-term memories. Expressing more of the protein stabilized memories when they'd normally start to fade. PKM-ζ appeared to alter a process called long-term potentiation, in which repeated patterns in nerve activity get translated into long-term changes in brain behavior.
But now, researchers have thrown a monkey wrench into the gears of this otherwise smooth-running model. When the gene for PKM-ζ is knocked out, the resulting mice have no problems forming memories. But when these mice are exposed to the drug that is supposed to target the (now missing) protein, they still showed deficits in memory. All of this suggests the drug that got everyone excited about PKM-ζ isn't all that specific after all. Oops.
It's a shame. PKM-ζ made an appealing candidate as a key mediator of memory. It's actually an alternate, shortened form of the Protein Kinase C ζ gene. The full-length PKC form includes sections that allow its activity to be regulated by what's going on in the cell. The shortened, PKM form lacks all of this, meaning it's always active. So, it's easy to envision a model where a nerve cell that only has PKC-ζ is able to flexibly respond to transient neural activity. In contrast, the cells expressing PKM-ζ have their activity state locked in, allowing them to retain information (like a memory) indefinitely.
Preliminary results also looked good. Expressing more PKM-ζ seemed to allow cells to hang on to memories for longer. A drug called ZIP, which was designed to look like a stretch of protein that is normally targeted by PKM-ζ (which adds phosphates to other proteins), could inhibit memory formation. It even hastened the loss of older memories. Everything seemed to be pointing in one direction.
That's why the knockout results are a bit of a shock. Since the two genes (PKC-ζ and PKM-ζ) overlap, it's impossible to eliminate only one of them. But the mice that lacked both ended up looking and acting perfectly normal. One of the two papers describing them failed to detect any differences between the knockout and normal mice, while the second only noted a handful of subtle differences on a handful of anxiety tests out of a large collection of behavioral assays they performed.
Both teams of researchers made sure a related kinase wasn't compensating for the loss of PKM-ζ. One checked the expression of all the related proteins in the brain and found they were largely unchanged. The other team made a conditional PKM-ζ deletion so the gene would only be eliminated after administration of a drug. As far as memory tests were concerned, these mice were also normal.
But the most surprising result came when both teams tested the drug ZIP in their knockout mice. ZIP is a short section of protein designed to look like the normal target of PKM-ζ; it has been thought to alter memory by simply swamping PKC-ζ, preventing it from performing its normal role. Yet, even when PKC-ζ was completely missing, ZIP continued to interfere with the formation and maintenance of memories. That strongly suggests ZIP is acting through a different target entirely, and it raises further questions about PKM-ζ's role in memory.
It's probably too early to write off PKM-ζ entirely. It's still possible the balance between PKC-ζ and PKM-ζ is important—since this knockout eliminated both, it's impossible to tell. But it should be possible to test this by modifying the gene so that it can express only PKC-ζ.
However, the results do suggest that PKM-ζ simply can't be the essential mediator of memory formation that people once suspected. It may still have a role, but more as a part of large network of activities that help consolidate a memory—a perspective more similar to the one that prevailed before the earlier PKM-ζ results started coming in.
But the papers, in addition to showing it is possible to publish negative results, do indicate that ZIP is doing something to memory, which almost certainly involves targets other than PKM-ζ. The labs involved in this research are undoubtedly already doing their best to find out what.
Nature, 2012. DOI: 10.1038/nature11802, 10.1038/nature11803 (About DOIs).
reader comments
34