Memory erasure by very high concentrations of ZIP may not be due to PKM‐zeta

The molecular basis of memory at synapses is one of the major outstanding problems in neuroscience. The article of Pastalkova et al. (2006), and subsequent articles based on their methods, makes the strong claim that memory maintenance (i.e., storage) is due to the protein kinase, PKM-zeta. In support of this, Pastalkova et al. performed the key tests of any model of memory maintenance: whether inhibition of the putative memory molecule can erase previously induced long-term potentiation (LTP) (at the cellular level) or previously learned information (at the behavioral level). Their claim that PKM-zeta is the memory molecule stems from success on both tests. Close examination of their article, however, reveals problems that were not discussed and which raise doubts about their claims. To erase LTP and memory, the authors used ZIP, a membrane permeable peptide inhibitor of PKM-zeta. In previous work on hippocampal slices their laboratory showed that 1 lM bath-applied ZIP could erase LTP, i.e., return the excitatory postsynaptic potential (EPSP) to the preLTP baseline (Ling et al., 2002). Their in vitro biochemistry showed that at higher concentrations, ZIP can also inhibit other kinases (IC-50 PKM-zeta 5 0.1 lM; IC-50 CaMKII 5 3 lM). It is therefore important to know what concentration Pastalkova et al. used in their in vivo experiments. This was not stated, but can be estimated from the information provided. They injected 1 lL of a solution containing 10 nmol of ZIP into the hippocampus. This solution was thus 10 mM ZIP. As shown in their Figure 1F (reproduced below), ZIP diffused widely in the hippocampus, but because of a barrier at the edge of the hippocampus, did not diffuse out of the hippocampus. By 2 h, the concentration of ZIP was nearly homogeneous across the four electrodes that span the hippocampal cross-section. Consistent with this, the physiological effect was similar at the four electrodes and was near steady-state (their Fig. 1E reproduced below). One can therefore assume that ZIP had spread fairly uniformly through the hippocampus (the very high concentration near the cannula is due to continuous leakage from the cannula). The 1 lL (51 mm) injected would have been diluted as it spread throughout the hippocampus (total volume <50 mm). Assuming that half the injected solution escapes the hippocampus through the hole in the hippocampus made by the cannula (their Fig. 1F), the concentration in the hippocampus is estimated to be 100 lM. This is three orders of magnitude greater than the IC-50 of PKM-zeta and two orders of magnitude higher than the concentration required to reverse LTP in hippocampal slices. At this concentration, other enzymes may be well affected.