The atypical protein kinase C (aPKC), PKMζ (PKMzeta), has garnered a great deal of attention in neurobiology due to its potential role as a molecular determinant of long-term synaptic plasticity (Sacktor, 2011). PKMζ has been implicated in the maintenance of late long-term potentiation (LTP), the storage of long-term memories, and in the maintenance of chronic pain states in multiple CNS regions (Sacktor, 2012). Evidence for this has depended, at least in part, on the use of a pseudo-substrate inhibitor of PKMζ called ZIP. Two very recent papers have demonstrated that knockout of PKCζ and PKMζ fails to disrupt the maintenance of late LTP and similarly fails to disrupt the maintenance of several forms of long-term memory (Lee et al., 2013; Volk et al., 2013). Moreover, these papers have demonstrated that ZIP reverses late LTP and long-term memory in knockout mice, indicating that ZIP has effects that are PKMζ independent (Lee et al., 2013; Volk et al., 2013). Hence, in our opinion, the PKMζ area has suddenly become much more interesting! Below we will briefly review what this means for work on PKMζ and pain, and we will offer some potential explanations for these findings in the context of work on PKMζ and pain.
First, what do these papers mean for work on PKMζ and pain? Here, it is important to distinguish between findings related to PKMζ itself versus those that are dependent on ZIP. Several groups, including ours, have shown that PKMζ phosphorylation and (presumably) translation are increased in important CNS areas for pain in response to injury. In the anterior cingulate cortex, phosphorylation of PKMζ increases, persistently, after peripheral nerve injury in mice and rats (Li et al., 2010; King et al., 2012). Hence, ACC PKMζ phosphorylation is correlated with the presence of neuropathic pain. In the spinal dorsal horn, both formalin and capsaicin injections into the hind paw increase total and phosphorylated PKMζ (Laferriere et al., 2011; Marchand et al., 2011). Consistent with this, we have shown that virally mediated overexpression of constitutively active PKCζ (PKMζ mimetic) leads to the development of a chronic pain state (Asiedu et al., 2011) that closely parallels hyperalgesic priming models pioneered by Jon Levine’s group (Reichling and Levine, 2009). Importantly, Todd Sacktor and colleagues have also shown that simple infusion of PKMζ into hippocampal neurons is sufficient to induce enduring LTP, and that virally mediated overexpression of PKMζ in cortex leads to an enhancement of memory (Shema et al., 2011). Therefore, although these recent papers have called into question whether PKMζ is necessary for the maintenance of long-term memory, there is strong evidence suggesting that PKMζ is sufficient for induction of late LTP, enhancement of memory, and expression of an enduring pain state. Thus, we propose that it is not possible to discount the role of PKMζ in pain and/or learning and memory based on these knockout studies.
Where things become much more problematic is in the interpretation of studies using the PKMζ inhibitor ZIP, as it clearly has non-specific effects as elucidated in these knockout mice. First, in PKC/Mζ knockout mice, late LTP can be induced, and ZIP still leads to a decay of late LTP. Second, several forms of long-term memory are normal in these mice, and ZIP continues to lead to a reversal of these forms of memory where its effects were actually tested (Lee et al., 2013; Volk et al., 2013). Studies exploring the role of PKMζ in pain have relied heavily on the use of ZIP as an inhibitor of PKMζ. These include studies in the ACC and in the spinal cord (Li et al., 2010; Asiedu et al., 2011; Laferriere et al., 2011; Marchand et al., 2011; Kwapis et al., 2012). Importantly, Sandkuhler’s group has shown that ZIP fails to reverse LTP in outer lamina dorsal horn neurons receiving C-fiber input (Drdla-Schutting et al., 2012); therefore, it is possible that the effects of ZIP in the spinal cord differ mechanistically from effects in the cortex or hippocampus related to learning and memory, where the use of this compound is clearly linked to a reversal of late LTP. Thus, it is not outside the realm of possibilities that ZIP has a specific action on PKMζ in the pain pathway, but that this mechanism is distinct from a reversal of late LTP. We propose that our findings in hyperalgesic priming models (Asiedu et al., 2011), along with those in DRG neurons where ZIP blocks nerve growth factor-dependent excitability (Zhang et al., 2012), are consistent with this notion. Therefore, a potentially exciting area of work in relation to PKMζ and pain would be to elucidate mechanisms wherein PKMζ leads to sensitization without affecting an LTP-like process.
There are, however, several other potential explanations for these recently elucidated discrepancies. One, which cannot be ruled out, is that ZIP has non-specific effects on other non-aPKC targets. Clearly, a screen of potential ZIP targets is warranted. Having said that, we hypothesize that such a target may have already been identified. The aPKC family consists of three kinases: PKMζ, PKCζ, and PKCλ (in rodents; PKCι in humans) (Moscat and Diaz-Meco, 2000). While PKMζ has been extensively investigated in CNS plasticity, PKCλ, which is expressed in neurons (Suzuki et al., 2003), has largely been ignored. Moreover, ZIP, which is identical to the pseudo-substrate sequence in PKCζ, is predicted to have 100 percent homology with the same region of PKCλ, suggesting that this inhibitor may also block PKCλ action. In fact, recent evidence shows that ZIP blocks PKCλ in in-vitro kinase assays (Lee et al., 2013). With genetic removal of PKC/Mζ, it is possible that PKCλ plays a compensatory role, allowing for the maintenance of CNS plasticity in the absence of PKC/Mζ. Under such a scenario, knockout of PKC/Mζ would fail to reveal a loss of late LTP and long-term memory, whereas ZIP would still be capable of reversing these processes. This is precisely what has been observed in the two recent Nature papers (Lee et al., 2013; Volk et al., 2013). Further exploration of the role of PKCλ in CNS plasticity is, therefore, highly likely to lead to novel insight into mechanisms of LTP, memory, and chronic pain maintenance.
Without question, these recent revelations have fundamentally altered the PKMζ narrative. While an initial reading may lead the casual observer to discount some fundamental findings in the field, we would argue that the reaction should be exactly the opposite. The field has been blown wide open, and this is now the perfect time for novel discoveries concerning the role of aPKCs in CNS plasticity. It also reveals an important possibility to identify differences in the role of aPKCs in learning and memory and chronic pain that may lead to the development of therapeutics that selectively affect the pain pathway without disrupting the fundamental mechanisms of learning and memory. There has never been a more exciting time to study aPKCs and their evolving role in nervous system plasticity.
Correction: An earlier version of this article referred to the three aPKC isoforms as PKMζ, PKCζ, and PKCγ. The third should have been PKCλ (PKC lamda). The text has been corrected.