Research has turned up many molecules that contribute to the formation of memories, but a single peptide, protein kinase M-ζ (PKM-ζ) has been hailed as the key player in maintaining them. Now, new research casts doubt on that starring role. In the January 2 Nature, back-to-back papers report that PKM-ζ knockout mice learn and remember just as wild-types do. They also have normal synaptic plasticity. The two groups, one led by Richard Huganir at Johns Hopkins University, Baltimore, Maryland, and the other by Robert Messing at the University of California, San Francisco, made different knockouts but obtained similar results. Researchers in the field agree that this means the kinase may not be essential for memory maintenance, but they differ on the exact interpretation. Several scientists Alzforum contacted believe that PKM-ζ may still be the main player in normal mice, but when the gene is deleted, cells recruit redundant pathways to compensate for its loss. Other researchers are more inclined to sideline the kinase, suggesting that alternate memory pathways may take center stage. This is an active area of research in many labs, and future studies will likely clarify the issue.

“The message of these papers is not that PKM-ζ is not important, but rather that it is one molecule in the complex system that plays a role in memory maintenance,” Paul Frankland at the Hospital for Sick Children, Toronto, Canada, told Alzforum. Together with colleague Sheena Josselyn, he wrote an accompanying Nature commentary on the findings.

The discovery in 2006 that an inhibitor of PKM-ζ could erase memories “turned the field on its head,” according to Kurt Haas at the University of British Columbia, Vancouver (see ARF related news story on Pastalkova et al., 2006). The constitutively active kinase was the first and still the only molecule shown to be crucial for sustaining long-term, consolidated memories. Science magazine named it one of its 10 “Breakthroughs of the Year 2006.” Later work by one of the authors, Todd Sacktor at SUNY Downstate Medical Center, Brooklyn, New York, demonstrated that overexpression of PKM-ζ can strengthen previously formed memories in rats (see ARF related news story).

Huganir and colleagues wanted to further investigate PKM-ζ’s effect on synapses. Co-first authors Lenora Volk and Julia Bachman knocked out the gene in mice. To their surprise, hippocampal slices from the animals exhibited normal long-term potentiation (LTP). The knockouts also performed normally in fear conditioning and spatial memory tasks. Also unexpectedly, the authors found that the PKM-ζ inhibitor (ζ inhibitory peptide, or ZIP) quenched LTP in hippocampal slices from PKM-ζ knockouts just as well as it did in wild-types, showing that ZIP can act through pathways other than PKM-ζ to disrupt synaptic plasticity. “This indicates that PKM-ζ is not the normal target for ZIP and is not mediating the effect on memory,” Huganir told Alzforum. “Now it’s up to the field to figure out what ZIP is doing.”

In the second paper, Messing and colleagues describe the generation of a different PKM-ζ knockout, which also behaved normally in fear conditioning, object recognition, motor learning, and drug reward memory tests. The only difference was that these knockouts appeared less anxious than wild-type mice. Similar to the Huganir mice, ZIP still affected these animals. First author Anna Lee injected the inhibitor into the nucleus accumbens of the knockouts and found that it erased cocaine reward memory just as well as in wild-types.

What do these findings mean for PKM-ζ? The data are confusing, Haas told Alzforum. Previous work from more than a dozen labs worldwide suggests that PKM-ζ can play an important role in memory maintenance. In addition to the overexpression experiments and inhibition by ZIP, blocking PKM-ζ function with a dominant negative molecule or using a different inhibitor also wipes out long-term memories. However, overexpression and dominant negative experiments can produce non-physiological results, Haas noted. “My opinion is that it’s possible that PKM-ζ still plays a role under normal conditions, but I think that has to be reevaluated in light of these studies,” he said.

André Fenton at New York University, who coauthored the original PKM-ζ paper, believes that the previous work and the new findings do not necessarily contradict each other. “The logical possibility is that when you get rid of something fundamental like PKM-ζ, it’s compensated for,” he said. Other researchers agree. “It would make sense that the cell would have redundant mechanisms to maintain memories,” Daniel Alkon and Thomas Nelson at West Virginia University, Morgantown, wrote to Alzforum. Future work should focus on finding out what these backup mechanisms are, Sacktor suggested, adding, “The key thing is we have three different inhibitors of PKM-ζ and they all do what nothing else has ever done, which is to erase memory. I think the evidence is pretty strong that for normal animals, PKM-ζ is the main molecule [that maintains memory].”

The fact that ZIP still erases memory in the PKM-ζ knockouts intrigues researchers. What might the inhibitor be acting on? PKM-ζ belongs to the protein kinase C (PKC) family, and has a closely related isozyme, PKC-ι/λ. Using in-vitro assays, Messing and colleagues found that ZIP quieted PKC-ι/λ just as effectively as PKM-ζ. They then looked to see if PKC-ι/λ levels went up in the knockout mice, but saw no evidence of this by Western blot. Huganir and colleagues also found no sign of changes in PKC-ι/λ in their knockout mice.

Despite these negative findings, many commentators suggested that PKC-ι/λ may be compensating for the loss of PKM-ζ in the knockouts. PKC-ι/λ differs from its cousin in that it contains a regulatory sequence, but this portion can be cleaved to form constitutively active PKM-ι/λ, which is almost identical to PKM-ζ. PKM-ι/λ functionally compensates for missing PKM-ζ in other scenarios (see Kishikawa et al., 2008; Rodriguez et al., 2009), and like PKM-ζ, it becomes activated during LTP (see Kelly et al., 2007), noted Oliver Hardt at McGill University, Montréal, Canada. Fenton pointed out that levels of PKM-ι/λ are very low in normal brain, and suggested that the methods in the two Nature papers may not have been sensitive enough to detect the kinase. He is currently looking for changes in PKM-ι/λ in the same knockout mouse used by Messing.

To test these ideas, Huganir has made a PKC-ι/λ knockout and a double knockout of the ι/λ and ζ isoforms, which he will examine for memory defects and sensitivity to ZIP. He also plans to perform a biochemical screen to find other molecules that can bind the inhibitor. The results of these and other ongoing studies should help pin down the pathways that sustain our long-term memories.––Madolyn Bowman Rogers


Make a Comment

To make a comment you must login or register.

Comments on News and Primary Papers

  1. Before experiments with the PKM-ζ inhibitor ZIP, there was no evidence that long-term memories were maintained by a specific molecular mechanism. The effect of ZIP on memory is quite specific and unique—memories from one day to three months old are erased, whereas there is no effect on short-term memory, initial learning, or baseline synaptic transmission measured in the living animal (1). Until ZIP, there was no agent known to erase memory long after its consolidation without hampering the ability to acquire new memories later on. So the action of ZIP is not anything like an anesthetic. PKM-ζ increases during memory storage and enhances synaptic transmission (2). PKM-ζ inhibitors also reverse the maintenance of long-term potentiation, again the only agents known to do so (2). Moreover, molecularly jamming the mechanism by which PKM-ζ-mediated synaptic enhancement can be reversed prevents ZIP from erasing memories, strongly supporting that ZIP is working on PKM-ζ (3). All these studies, along with others using additional inhibitors and dominant negative versions of PKM-ζ (4), indicate that PKM-ζ maintains long-term memory in normal animals, from mice to invertebrates such as flies (5) and the mollusk Aplysia (6).

    What is new in these two studies is that they use genetically modified mice in which the gene that makes PKM-ζ had been completely deleted at the animals’ conception in order to study learning and memory. They find that the mice’s memory is intact. This is not too surprising because compensation for one gene by another, which might have taken place here but may have escaped detection, is a routine observation in knockout mice. If so, the compensatory gene product would subserve a similar function and exhibit learning-induced upregulation.

    So what is the backup mechanism for maintaining memory? The authors do not say, but a clue can be obtained by their observations that ZIP still works to erase memory and reverse long-term potentiation. ZIP is a pseudo-substrate inhibitor of the catalytic domain of atypical PKCs, and effectively inhibits PKM-ζ and the other atypical PKC found in forebrain, PKC-ι/λ. PKC-ι and PKC-λ are different names for the same gene—the human gene is termed PKC-ι and the mouse gene PKC-λ. Before vertebrates, there was only one atypical PKC gene, and a PKM form is made from it (7,8). But with the earliest vertebrates, this function was split into two very closely related atypical PKC genes. In the forebrain, the ζ gene became dedicated to making only PKM-ζ (7), but PKC-ι/λ can also make a PKM form similar to the way it is made in invertebrates—by shortening the PKC-ι/λ protein by proteolysis (8). Genetic studies have shown that these two isoforms can compensate for each other’s functions (9), and both ζ and ι/λ can do the same thing in neurons to enhance excitability (10). We have previously shown that PKC-ι/λ is also activated in LTP (11), and that there is normally a small amount of PKM-ι/λ present in the brain as well (12). The methods in the new papers may not have been optimized to detect synaptic PKM-ι/λ in the complete knockout after learning or LTP. So it is indeed possible that PKC/PKM-ι/λ is the backup mechanism for memory that is normally maintained by PKM-ζ. In such a case, the ability of ZIP to block LTP and memory, reported in the papers, is actually expected and would be strong support for the crucial role of atypical PKCs in maintaining long-term memories.

    One of the experiments in one of the papers used a drug that binds to the estrogen receptor in a technique to reduce gene expression of PKM-ζ after a mouse’s initial development, and this is a particularly interesting experiment because they still saw synaptic long-term potentiation. They didn’t look at the mouse’s behavior. However, there was still PKM-ζ around in the brain of the mouse and, more importantly, we don’t know if any new PKM-ζ was made during the LTP from the PKM-ζ mRNA that was still around. Moreover, the estrogen receptor binds to and activates atypical PKC (13), so it is not clear how chronic exposure to the drug they used would affect PKM-ζ or PKC-ι/λ activity or turnover.

    So I think future research from these studies will be to try to find the backup mechanisms for memory and to determine what role they play, along with PKM-ζ, in memory in normal animals. I am gratified that our research has stimulated the experiments of others to identify the molecular mechanisms that store long-term memories—one of the fundamental questions in neuroscience.


    . How does PKMζ maintain long-term memory?. Nat Rev Neurosci. 2011 Jan;12(1):9-15. PubMed.

    . Protein kinase Mzeta is necessary and sufficient for LTP maintenance. Nat Neurosci. 2002 Apr;5(4):295-6. PubMed.

    . PKMzeta maintains memories by regulating GluR2-dependent AMPA receptor trafficking. Nat Neurosci. 2010 May;13(5):630-4. PubMed.

    . Enhancement of consolidated long-term memory by overexpression of protein kinase Mzeta in the neocortex. Science. 2011 Mar 4;331(6021):1207-10. PubMed.

    . Memory enhancement and formation by atypical PKM activity in Drosophila melanogaster. Nat Neurosci. 2002 Apr;5(4):316-24. PubMed.

    . Protein kinase M maintains long-term sensitization and long-term facilitation in aplysia. J Neurosci. 2011 Apr 27;31(17):6421-31. PubMed.

    . Memory maintenance by PKMζ - an evolutionary perspective. Mol Brain. 2012;5:31. PubMed.

    . Serotonin-induced cleavage of the atypical protein kinase C Apl III in Aplysia. J Neurosci. 2012 Oct 17;32(42):14630-40. PubMed.

    . aPKC enables development of zonula adherens by antagonizing centripetal contraction of the circumferential actomyosin cables. J Cell Sci. 2008 Aug 1;121(Pt 15):2481-92. PubMed.

    . Nerve growth factor enhances the excitability of rat sensory neurons through activation of the atypical protein kinase C isoform, PKMζ. J Neurophysiol. 2012 Jan;107(1):315-35. PubMed.

    . Regulation of protein kinase Mzeta synthesis by multiple kinases in long-term potentiation. J Neurosci. 2007 Mar 28;27(13):3439-44. PubMed.

    . Distribution of protein kinase Mzeta and the complete protein kinase C isoform family in rat brain. J Comp Neurol. 2000 Oct 16;426(2):243-58. PubMed.

    . Role of atypical protein kinase C in estradiol-triggered G1/S progression of MCF-7 cells. Mol Cell Biol. 2004 Sep;24(17):7643-53. PubMed.

    View all comments by Todd Sacktor


News Citations

  1. Immune Receptor Controls Synaptic Plasticity; LTP Makes Memories
  2. The Guardians of Forever: Forming and Keeping Memories

Paper Citations

  1. . Storage of spatial information by the maintenance mechanism of LTP. Science. 2006 Aug 25;313(5790):1141-4. PubMed.
  2. . aPKC enables development of zonula adherens by antagonizing centripetal contraction of the circumferential actomyosin cables. J Cell Sci. 2008 Aug 1;121(Pt 15):2481-92. PubMed.
  3. . Atypical protein kinase C activity is required for extracellular matrix degradation and invasion by Src-transformed cells. J Cell Physiol. 2009 Oct;221(1):171-82. PubMed.
  4. . Regulation of protein kinase Mzeta synthesis by multiple kinases in long-term potentiation. J Neurosci. 2007 Mar 28;27(13):3439-44. PubMed.

Further Reading

Primary Papers

  1. . PKM-ζ is not required for hippocampal synaptic plasticity, learning and memory. Nature. 2013 Jan 17;493(7432):420-3. PubMed.
  2. . Prkcz null mice show normal learning and memory. Nature. 2013 Jan 2; PubMed.
  3. . Neuroscience: Memory and the single molecule. Nature. 2013 Jan 17;493(7432):312-3. PubMed.