Learning, aging and intrinsic neuronal plasticity John F. Disterhoft 1, 2 and M. Matthew Oh 1 1 Department of Physiology, Feinberg School of Medicine, Northwestern University, Chicago, IL 60611-3008, USA 2 Institute for Neuroscience, Feinberg School of Medicine, Northwestern University, Chicago, IL 60611-3008, USA In vitro experiments indicate that intrinsic neuronal excitability, as evidenced by changes in the post-burst afterhyperpolarization (AHP) and spike-frequency accommodation, is altered during learning and normal aging in the brain. Here we review these studies, high- lighting two consistent findings: (i) that AHP and accom- modation are reduced in pyramidal neurons from animals that have learned a task; and (ii) that AHP and accommodation are enhanced in pyramidal neurons from aging subjects, a cellular change that might con- tribute to age-related learning impairments. Findings from in vivo single-neuron recording studies comple- ment the in vitro data. From these consistently repro- duced findings, we propose that the intrinsic AHP level might determine the degree of synaptic plasticity and learning. Furthermore, it seems that reductions in the AHP must occur before learning if young and aging subjects are to learn a task successfully. Introduction Synergistic cellular and subcellular mechanisms are pre- sumably involved in storing memories in the brain. The parable of the blind men and the elephant would suggest that how we study this fascinating problem might well define what we see. As we examine the interactions between molecules and pathways that lead to profound changes in neurons and neural circuits, there is not one single process but rather a series of parallel processes that combine to mediate information storage. There is no doubt that synaptic alterations occur, as has been documented in learning experiments. But we have been especially fasci- nated by changes in intrinsic neuronal excitability that occur reliably in animals that are trained and learn. These intrinsic alterations might be required for or at least contribute to the synaptic changes involved in information storage. This review will focus on summarizing our insights into the mechanisms of associative learning, gained over the past two decades from a series of studies that use both in vitro and in vivo recording techniques and that substantiate the importance of learning-related and aging-related alterations of intrinsic neuronal excitability in the hippocampus and neocortex. Post-burst afterhyperpolarization An important cellular measure of intrinsic excitability that has been observed to change after learning is the post-burst afterhyperpolarization (AHP). The term afterhyperpolarization was originally used by Eccles, Eccles and Lundberg [1] almost 50 years ago to describe the hyperpolarization that follows an action potential in motoneurons. They postulated that the AHP determined the firing frequency of these neurons. This hypothesis was further refined by Kandel and Schwartz [2] who proposed that ‘after-hyperpolarization of motoneurons limits firing frequency but also counteracts the inactivation process (of spike generators) and thereby allows the motoneuron to sustain a long repetitive train.’ Since then much work has been done to understand the role of the AHP in controlling neuronal firing capacity and to identify the channels that underlie it. To describe all these studies in depth is beyond the scope of this review (for some recent reviews, see Refs [3–8]). However, there are a few fundamental properties of the AHP that we will briefly describe. First and foremost, the AHP is mediated largely by Ca 2+ -dependent K + current(s). Evidence for Ca 2+ depen- dence was first illustrated in motoneurons and sympa- thetic neurons [9–13]. Shortly thereafter, Nicoll, Prince, Schwartzkroin and colleagues showed that the AHP of hippocampal pyramidal neurons is mediated by a Ca 2+ - dependent outward K + current [14–17]. During a sustained depolarization, hippocampal pyramidal neurons respond ‘with an initial rapid action potential discharge’ which gradually slows over time, a process referred to as spike- frequency adaptation (or accommodation) [18]. The AHP has a significant role in the accommodation of hippocampal pyramidal neurons. Accommodation is reduced (i.e. the neuron fires more action potentials to a sustained depolar- ization or a train of depolarizing stimuli) when the AHP is small and, conversely, accommodation is enhanced when the AHP is large. We should also note that recent studies in hippocampal and neocortical neurons highlight the con- tribution of a Na + -dependent K + current to the AHP [19], which will no doubt be further investigated within the context of learning in the near future. More detailed examination has revealed that currents flowing through several channels contribute to different phases of the AHP [4,5,20–22]. The fast AHP on the down- stroke of each action potential has a role in spike broad- ening and action potential repolarization, and is thought to comprise a mixture of I C , I M and I A currents [4,7,20]. The medium AHP, I AHP , lasts for the initial 50–200 ms after an action potential burst and has been a target of much research and debate in the past decade. Apamin, a neuro- toxin from bee venom, blocks the medium AHP, and Review TRENDS in Neurosciences Vol.29 No.10 Corresponding author: Disterhoft, J.F. (jdisterhoft@northwestern.edu). Available online 30 August 2006. www.sciencedirect.com 0166-2236/$ – see front matter ß 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.tins.2006.08.005