PFC–Hippocampus Interactions during Sleep and Awake Ripples
Wenbo Tang, Justin D. Shin, Loren M. Frank, and Shantanu P. Jadhav
(see pages 11789–11805)
Interactions between the hippocampus and prefrontal cortex (PFC) are essential for the formation of episodic memories, incorporating these memories into models of the world (schema), and using these memories to guide ongoing decision making. Hippocampus–PFC interactions involve coordinated activity in which PFC neurons fire in phase with oscillatory activity in the hippocampus. During active exploration, PFC spiking is coupled to hippocampal theta-frequency oscillations, but during pauses in exploration and during slow-wave sleep, PFC activity occurs in conjunction with hippocampal sharp-wave ripples (SWRs). Ensembles of neurons activated during exploration are reactivated during SWRs, and this is thought to be essential for memory consolidation and schema formation.
Normalized cross-covariance of spiking between pairs of CA1 and PFC neurons shows a stronger clustering of peaks around 0 ms for awake SWRs than for sleep SWRs, indicating stronger synchronization during awake SWRs. See Tang et al. for details.
To identify differences between sleep and awake SWRs, Tang et al. recorded hippocampal CA1 and medial PFC activity continuously across sleep and awake periods in which rats learned a spatial alternation task. Hippocampus–PFC coordination during sleep and wake SWRs differed in several ways. First, synchronization between PFC and hippocampal neuron spiking was stronger during awake SWRs than during sleep SWRs. In addition, coactivation of hippocampal and PFC neurons that had overlapping place fields—suggesting reactivation of exploration-associated activity patterns—was stronger during awake SWRs than during sleep SWRs, particularly in the initial stages of learning. Furthermore, more PFC neurons were inhibited during awake SWRs than during sleep SWRs. The inhibited neurons tended to encode places where the rat was stationary, and there were more of these during awake periods than during sleep, in which rats mainly stayed in the nest box. Finally, whereas the power of PFC spindle and theta-frequency oscillations increased during sleep SWRs, no such coordination was seen during awake SWRs.
These differences suggest that sleep and awake SWRs have distinct roles. The authors suggest that the stronger reactivation of patterns representing recent experiences during awake SWRs might be essential for new learning and to guide decision making during the ongoing task, whereas hippocampus–PFC coordination during sleep SWRs is more important for integrating new learning with old memories to form schemas. Future experiments should test these hypotheses and examine reactivation of nonspatial representations during sleep and awake SWRs.
Contribution of CaV1.2 to Extinction of Cocaine Memories
Caitlin E. Burgdorf, Kathryn C. Schierberl, Anni S. Lee, Delaney K. Fischer, Tracey A. Van Kempen, et al.
(see pages 11894–11911)
A major challenge for recovering addicts is to resist relapse when returning to an environment associated with drug use. One way to overcome this challenge might be to enhance extinction of memories linking contextual cues to drug use. Previous work has shown that such extinction involves synaptic plasticity in the prefrontal cortex, nucleus accumbens, and hippocampus. Intriguingly, blocking CaV1.3 L-type calcium channels in the ventral tegmental area, which contains dopaminergic neurons involved in reward learning, facilitates extinction of cocaine-associated memories (Degoulet et al. 2016 Mol Psychiatry 21:394). In contrast, Burgdorf et al. show that activation of L-type CaV1.2 channels in the hippocampus is essential for such extinction.
When mice were trained to associate one chamber of a cage with cocaine delivery, they developed conditioned place preference (CPP), spending more time in that chamber than in a saline-paired chamber. After mice were exposed to the cage for several days without cocaine, CPP was extinguished. At that time, CaV1.2 levels increased in the hippocampus, but not in the nucleus accumbens or prefrontal cortex. Importantly, knocking out CaV1.2 selectively in dorsal hippocampal neurons that expressed D1 dopamine receptors impaired extinction, but not acquisition of CPP. In contrast, knocking out CaV1.2 in glutamatergic hippocampal neurons had no effect on acquisition or extinction of CPP. Additional experiments demonstrated that in addition to CaV1.2, extinction (but not acquisition) of cocaine CPP required activation of the Ca2+/calmodulin-dependent kinase CaMKII and phosphorylation of the AMPA receptor GluA1 subunit on serine 831.
D1 receptors in the dorsal hippocampus are present primarily in GABAergic interneurons (Puighermanal et al 2017 Brain Struct Funct 222:1897). Therefore, the current results suggest that extinction of cocaine CPP requires activation of CaV1.2 channels in GABAergic neurons, leading to activation of CaMKII and phosphorylation of AMPA receptors. Previous work indicates that AMPA receptor phosphorylation increases channel conductance and consequently enhances long-term potentiation. Thus, extinction training might strengthen inhibitory pathways that suppress cocaine CPP. The results thus reinforce the idea that extinction involves new learning rather than degradation of existing memories. They also suggest that any treatments based on inhibition of L-type calcium channels should selectively target CaV1.3 channels.
Footnotes
This Week in The Journal was written by Teresa Esch, Ph.D.
