Learning and aging affect neuronal excitability and learning

Highlights

• Intrinsic neuronal excitability is enhanced via reduced AHP by successful learning.

• Intrinsic neuronal excitability is reduced via enlarged AHP with normal aging.

• Hippocampal neurons in learning-impaired aged subjects have enlarged AHP.

• Hippocampal neurons in learning-unimpaired aged subjects have reduced AHP.

• Reducing hippocampal neuron AHP ameliorates learning deficits in aged subjects.

Abstract

The first study that demonstrated a change in intrinsic neuronal excitability after learning in ex vivo brain tissue slices from a mammal was published over thirty years ago. Numerous other manuscripts describing similar learning-related changes have followed over the years since the original paper demonstrating the postburst afterhyperpolarization (AHP) reduction in CA1 pyramidal neurons from rabbits that learned delay eyeblink conditioning was published. In addition to the learning-related changes, aging-related enlargement of the postburst AHP in CA1 pyramidal neurons have been reported. Extensive work has been done relating slow afterhyperpolarization enhancement in CA1 hippocampus to slowed learning in some aging animals. These reproducible findings strongly implicate modulation of the postburst AHP as an essential cellular mechanism necessary for successful learning, at least in learning tasks that engage CA1 hippocampal pyramidal neurons.

Introduction

In the early 1980′s there were several lines of research that were intersecting in the learning and memory field. First, the hippocampus was (and still remains) an actively researched brain region thought to be essential for associative learning since the ground-breaking publication by Scoville and Milner (1957) described the profound memory deficit in Henry Molaison (H.M.) following his bilateral hippocampectomy to alleviate his seizures. Second, investigators working in invertebrate model systems, including Eric Kandel and Daniel Alkon and their colleagues, had been training invertebrates in a variety of tasks, removing portions of the central nervous system presumed to be importantly involved in mediating the plastic changes, and studying cellular mechanisms for learning and plasticity in ex vivo preparations. Kandel and his colleagues focused on presynaptic changes in transmitter release, especially those mediated by cyclic AMP, as involved in both sensitization and their model of classical conditioning (Abrams, Castellucci, Camardo, Kandel, & Lloyd, 1984, Cedar, Kandel, & Schwartz, 1972, Lee, Bailey, Kandel, & Kaang, 2008). In contrast, Alkon and his group focused on nonsynaptic changes, conditioning-specific reductions in the A-type potassium current and a calcium-dependent potassium current that were observed in isolated type B photoreceptor cells from Hermissenda crassicornis (an invertebrate sea slug) after classical conditioning (Alkon, 1984). Third, the demonstration of long term potentiation in conscious rabbits by Bliss and Lomo (Bliss and Gardner-Medwin, 1973, Bliss and Lomo, 1973, Lømo, 2003) had been extended into ex vivo brain slices, where long term potentiation was being induced and mechanisms were being explored (Andersen, Sundberg, Sveen, & Wigström, 1977, McNaughton, Douglas, & Goddard, 1978, Wigström & Gustafsson, 1983). Fourth, Woody and colleagues had done an important series of studies in conscious, head fixed cats examining the brainstem and cortical changes involved in eyeblink and nose-twitch classically conditioned responses. Importantly, they observed that a smaller electrical stimulus was needed to evoke an action potential from pericruciate cortical neurons in vivo after delay eyeblink conditioning in cats (Brons & Woody, 1980, Woody & Black-Cleworth, 1973). Fifth, in vivo single-unit recordings in rabbits during delay eyeblink conditioning revealed that the firing rate of hippocampal principal cells increased to the presentation of the conditioned stimulus and mirrored the conditioned response, closure of the nictitating membrane (Berger and Thompson, 1978, Berger et al., 1976).

These and other reports strongly suggested that successful learning would lead to an excitability change within the principal neurons of the hippocampus that could potentially be observed and studied in detail from brain tissue slices. Thus, over thirty years ago, John Disterhoft, Douglas Coulter, and Daniel Alkon tested this hypothesis in a series of experiments that they carried out at the Marine Biological Laboratories in Woods Hole, MA, where Alkon was located at that time. Very importantly, they used the same approach that had been used extensively in invertebrates. They trained rabbits in the eyeblink associative learning task, removed a portion of the brain known to be dramatically changed during the task – the hippocampus – and studied it ex vivo. In 1986 they reported that the postburst afterhyperpolarization (AHP) is reduced in hippocampal CA1 pyramidal neurons from rabbits that successfully learned the Pavlovian delay eyeblink conditioning task (Fig. 1; Disterhoft, Coulter, & Alkon, 1986). Prior to their 1986 publication, there had not been a study designed to examine learning-related changes in intrinsic properties of neurons in mammalian brain slices. The field was and still remains focused on long-term potentiation of synaptic transmission in the hippocampus as the mechanism for learning and memory, based upon the in vivo discoveries made by Lomo, Bliss and Gardner-Medwin (Bliss and Gardner-Medwin, 1973, Bliss and Lomo, 1973) in mammals as well as the invertebrate work initiated by Kandel and colleagues (Castellucci, Pinsker, Kupfermann, & Kandel, 1970, Kandel & Tauc, 1965, Kandel & Tauc, 1965). But the learning-related AHP reduction reported by Disterhoft and colleagues was evoked using a direct current injection into the cell, and was not due to synaptic stimulation. Hence, this paper was the first to demonstrate that a ‘memory trace’ caused by associative learning can be stored in mammalian neurons ex vivo and opened the field to examine the cellular mechanisms that cause changes in intrinsic neuronal excitability, not synaptic plasticity, as successful learning takes place.