Search

How are long-term memories stored in the brain? The formation of memories is believed to depend on the strengthening of connections between neurons. During learning, neurons produce an enzyme called PKMzeta (or PKMζ), which is thought to be responsible for maintaining the newly strengthened connections. Inhibitors of PKMzeta, such as a drug called ZIP, disrupt long-term memories. This suggests that the brain may be like a computer hard disc in that its stored information — its memories — could be erased. However, recent experiments on genetically engineered mice have thrown the role of PKMzeta into question. Knockout mice that lack the gene for PKMzeta can still strengthen connections between neurons and can still learn and remember. Moreover, ZIP still works to reverse the strengthening and to erase long-term memories. This indicates that ZIP can act on something other than the PKMzeta enzyme. These results have led many neuroscientists to doubt that PKMzeta has anything to do with memory. Yet there are two possible explanations for the normal memory in PKMzeta knockout mice. First, PKMzeta is not required for memory, so getting rid of it has no effect. Second, PKMzeta is essential for long-term memory in normal mice. However, knockout mice recruit a back-up mechanism for long-term memory storage, which is also sensitive to the effects of ZIP. To test these possibilities, Tsokas et al. used a modified piece of DNA that prevents neurons with the gene for PKMzeta from producing the enzyme. The DNA blocked memory formation in normal mice, consistent with a role for PKMzeta in memory. However, it had no effect in knockout mice — the DNA had nothing to work on. This suggests that another molecule does indeed act as a back-up for PKMzeta in these animals. Further experiments revealed that an enzyme closely related to PKMzeta, called PKCiota/lambda (PKCι/λ), substitutes for PKMzeta during memory storage in the knockout mice. These findings restore PKMzeta to its early promise. They show that PKMzeta is crucial for long-term memory in normal mice, but that something as important as memory storage has a back-up mechanism should PKMzeta fail. Future work may reveal when and how this back-up becomes engaged.