Compartmentalized PKA signaling events are required for synaptic tagging and capture during hippocampal late-phase long-term potentiation

Abstract

Synaptic plasticity, the activity-dependent change in the strength of neuronal connections, is a proposed cellular mechanism of memory storage that is critically regulated by protein kinases such as cAMP-dependent protein kinase (PKA). Despite the fact that a neuron contains thousands of synapses, the expression of synaptic plasticity can be specific to subsets of synapses. This is surprising because signal transduction pathways underlying synaptic plasticity involve diffusible second messenger molecules such as cAMP and diffusible proteins such as the catalytic subunit of PKA. One way in which this specificity can be achieved is by the localization of signal transduction molecules to specific subcellular domains. Spatial compartmentalization of PKA signaling is achieved via binding to A kinase-anchoring proteins (AKAPs). We report here that pharmacological inhibition of PKA anchoring impairs synaptically activated late-phase long-term potentiation (L-LTP) in hippocampal slices. In contrast, potentiation that is induced by the pharmacological activation of the cAMP/PKA pathway, which can potentially affect all synapses within the neuron, is not impaired by inhibition of PKA anchoring. These results suggest that PKA anchoring may be particularly important for events that occur at the synapse during the induction of L-LTP, such as synaptic tagging and capture. Indeed, our results demonstrate that blocking PKA anchoring impairs synaptic tagging and capture. Thus our data highlight the idea that PKA anchoring may allow for specific populations of synapses to change in synaptic strength in the face of plasticity-related transcription that is cell-wide.

Introduction

The cellular mechanisms underlying memory storage likely involve activity-dependent changes in the strength of neuronal connections (Martin et al., 2000). Long-lasting modifications in synaptic strength require products of transcription (Nguyen et al., 1994) and translation (Frey et al., 1988). Even though these newly generated molecules may be distributed throughout the cell, only a subset of these synapses remain changed over time (Nguyen et al., 1994). Out of thousands of synapses, how does a neuron identify those synapses that will selectively undergo long-term change? Furthermore, how is this degree of specificity possible when signal transduction pathways underlying synaptic plasticity involve diffusible molecules such as cAMP? One way in which specificity can be achieved is by localizing signal transduction molecules to specific subcellular domains. Indeed, PKA is concentrated in certain subcellular regions through interaction with a family of functionally distinct but structurally related proteins called A kinase-anchoring proteins (AKAPs), a protein family consisting of more than 50 members (reviewed in Wong and Scott, 2004). Spatial compartmentalization of PKA may contribute to the specificity of the cAMP/PKA signaling pathway in affecting downstream proteins (for example, in studies by Fink et al., 2001, inhibition of PKA anchoring resulted in redistribution of RII and decreased compartmentalization of PKA).

A well known and widely studied form of synaptic plasticity is hippocampal long-term potentiation (LTP) (Bliss and Collingridge, 1993; Bliss and Lømo, 1973; Malenka and Nicoll, 1999). In slice preparations of the hippocampus, brief patterns of high-frequency stimulation increase the amplitude of subsequent synaptic potentials. At hippocampal Schaffer collateral-CA1 synapses, L-LTP requires NMDA receptor activation (Collingridge et al., 1983), protein synthesis (Frey et al., 1988), transcription (Nguyen et al., 1994), and PKA (Abel et al., 1997; Frey et al., 1993; Huang and Kandel, 1994; Matthies and Reymann, 1993; Woo et al., 2000, Woo et al., 2002, Woo et al., 2003). In area CA1, cAMP levels are increased 1 min after tetanic stimulation that induced L-LTP (Frey et al., 1993), with a corresponding increase in PKA activity briefly after stimulation (Roberson and Sweatt, 1996). Treatment of hippocampal slices with cAMP analogs induces a potentiation that resembles L-LTP, whereas treatment with PKA inhibitors blocks L-LTP (Frey et al., 1993). Transgenic mice expressing a dominant negative regulatory subunit of PKA have reduced L-LTP in area CA1 and exhibit impaired performance in hippocampal-dependent memory and place cell stability (Abel et al., 1997; Rotenberg et al., 2000; Woo et al., 2000, Woo et al., 2002, Woo et al., 2003). Therefore, PKA activity is crucial in L-LTP and long-term memory.

An interesting property of hippocampal L-LTP is that of pathway specificity. A two-pathway experimental setup, where two independent sets of presynaptic inputs converge on a common set of postsynaptic neurons, can be used to monitor synaptic activity in two separate populations of synapses on the same synaptic neurons. In this two-pathway setup, only synapses that received L-LTP stimulation remain persistently potentiated; synapses that did not receive L-LTP stimulation do not undergo long-term functional change (Nguyen et al., 1994). This pathway specificity with which subsets of synapses become stably potentiated over time suggests that plasticity-related gene products are selectively used by those synapses that have received L-LTP stimulation to increase synaptic strength. Frey and Morris tested the idea that plasticity-related proteins, widely distributed throughout the cell, can be captured at specific “tagged” synapses (Frey and Morris, 1997). They demonstrated using two-pathway experiments that transient synaptic potentiation induced by weak stimulation in one pathway can be converted to stable synaptic potentiation if it is paired with strong stimulation in the other pathway. Thus synaptic activity seems to “tag” active synapses. It has been proposed that strong stimulation of synapses not only tags the synapses but also induces transcriptional and translational activity; weak stimulation of synapses only tags the synapses, but these tags have the ability to “capture” the products of gene expression produced by the neuron in response to strong stimulation in other synapses (Barco et al., 2002, Barco et al., 2005; Frey and Morris, 1997, Frey and Morris, 1998; Sajikumar and Frey 2004; Sajikumar et al., 2005; Young and Nguyen, 2005).

Because selectively modifying subsets of synapses likely requires highly compartmentalized signal transduction pathways, we explore here the hypothesis that PKA anchoring is essential for hippocampal L-LTP. To inhibit PKA–AKAP interactions, we use a cell-permeable form of the Ht31 peptide, a truncated form of an AKAP that competitively binds the regulatory subunits of PKA, blocking the interaction of the regulatory subunit with most AKAPs (Carr et al., 1992). We demonstrate that synaptically activated L-LTP is impaired by this pharmacological inhibitor of PKA anchoring in a dose-dependent manner. We provide further evidence for the compartmentalized nature of PKA signaling in L-LTP by showing that synaptic potentiation caused by the global elevation of cAMP does not require PKA anchoring. Because tagging synapses for long term changes is specific to subsets of synapses, we investigate whether PKA anchoring is required to maintain the spatial specificity that is critical to synaptic tagging. Using a two-pathway paradigm, we demonstrate that the process of synaptic tagging and capture is impaired by inhibition of PKA anchoring. Thus the spatial specificity of the PKA signaling pathway, mediated by AKAPs, is critical to long-term synaptic changes in the hippocampus.