Gregory A. Clark

Operant Conditioning of Gill Withdrawal in Aplysia

A basic question in neuroscience is how different forms of learning are related. To further address that question, we examined whether gill withdrawal in Aplysia, which has already been studied extensively for neuronal mechanisms contributing to habituation, sensitization, and classical conditioning, also undergoes operant conditioning. Animals were run in pairs. During the initial training period, the contingent (experimental) animal received a siphon shock each time its gill relaxed below a criterion level, and the yoked control animal received a shock whenever the experimental animal did, regardless of its own gill position. This was followed by an extinction period when there was no shock, a retraining period when both animals were contingent, and another extinction period. The experimental animals spent more time with their gills contracted above the criterion level than did the control animals during each period, demonstrating operant conditioning. The type of gill behavior modified by learning shifted over time: the experimental animals had a larger increase in the frequency and duration of spontaneous contractions than did the control animals during the first but not the last extinction period and a larger increase in the level of tonic contraction during the last but not the first extinction period. Because many of the neurons controlling spontaneous and tonic gill withdrawal have already been identified, it should now be possible to examine the cellular locus and mechanism of operant conditioning and compare them with those for other forms of learning of the same behavior.

Modeling Hermissenda: I. Differential contributions of IA and IC to type-B cell plasticity.

We developed a multicompartmental Hodgkin-Huxley model of the Hermissenda type-B photoreceptor and used it to address the relative contributions of reductions of two K+ currents, Ia and IC, to changes in cellular excitability and synaptic strength that occur in these cells after associative learning. We found that reductions of gC, the peak conductance of IC, substantially increased the firing frequency of the type-B cell during the plateau phase of a simulated light response, whereas reductions of gA had only a modest contribution to the plateau frequency. This can be understood at least in part by the contributions of these currents to the light-induced (nonspiking) generator potential, the plateau of which was enhanced by gC reductions, but not by gA reductions. In contrast, however, reductions of gA broadened the type-B cell action potential, increased Ca2+ influx, and increased the size of the postsynaptic potential produced in a type-A cell, whereas similar reductions of gC had only negligible contributions to these measures. These results suggest that reductions of IA and IC play important but different roles in type-B cell plasticity.

GABA-Induced Synaptic Facilitation at Type B to A Photoreceptor Connections in Hermissenda

Gamma-aminobutyric acid (GABA) is a prevalent neurotransmitter in both vertebrate and invertebrate systems. Here we report that, in addition to its usual inhibitory actions, GABA induced synaptic facilitation at type B to A photoreceptor connections of the marine mollusk Hermissenda when applied transiently to the isolated nervous system. Synaptic facilitation also occurred in response to mechanical stimulation of the GABAergic hair cells, which are normally activated by rotational unconditioned stimuli during behavioral training of the intact animal. This synaptic facilitation represents a novel form of GABA-induced neuromodulation which may contribute to learning-dependent suppression of phototaxis in Hermissenda.

Modeling Hermissenda: II. Effects of variations in type-B cell excitability, synaptic strength, and network architecture.

Because the Hermissenda eye is relatively simple and its cells well characterized, it provides an attractive preparation for detailed computational analysis. To examine the neural mechanisms of learning in this system, we developed multicompartmental models of the type-A and type-B photoreceptors, simulated the eye, and asked three questions: First, how do conductance changes affect cells in a network as compared with those in isolation; second, what are the relative contributions of increases in B-cell excitability and synaptic strength to network output; and third, how do these contributions vary as a function of network architecture? We found that reductions in the type-B cells of two K+ currents, IA and IC, differentially affected the type-B cells themselves, with IC reductions increasing firing rate (excitability) in response to light, and IA reductions increasing quantal output (synaptic strength) onto postsynaptic targets. Increases in either type-B cell excitability or synaptic strength, induced directly or indirectly, each suppressed A-cell photoresponses, and the combined effect of both changes occurring together was greater than either alone. To examine the effects of network architecture, we compared the full network with a simple feedforward B-A pair and intermediate configurations. Compared with a feedforward pair, the complete network exhibited greater A-cell sensitivity to B-cell changes. This was due to many factors, including an increased number of B-cells (which increased B-cell impact on A-cells), A-B feedback inhibition (which slowed both cell types and altered spike timing relationships), and B-B lateral inhibition (which reduced B-cell sensitivity to intrinsic biophysical modifications). These results suggest that an emergent property of the network is an increase both in the rate of information acquisition (“learning”) and in the amount of information that can be stored (“memory”).

Emotional Learning: Fear and loathing in the amygdala

The amygdala in the brain plays a critical role in learning emotional components of experience, such as conditioned fear; these processes in turn affect many other aspects of memory and cognition.

How neural are neural networks? a comparison of information processing and storage in artificial and real neural systems

Abstract In this paper we compare the operations of biological and artificial neural networks and highlight some of the key strategies that real brains use to acquire, process, and store information. Artificial neural networks do indeed employ some of the same fundamental processes used in biological networks, such as changes in connection strengths, thus supporting the proposed similarity between the two systems. However, biological networks also exhibit several additional forms of plasticity which have received less attention in network models, including changes in neuronal excitability; changes in the fidelity (as well as strength) of synaptic transmission; changes in signal-to-noise ratios; changes in type of neurotransmitter synthesized and released; and changes in neuron number. The richness of plastic mechanisms found in biological neurons suggests there may be a number of effective computational tricks used by real nervous systems that could be advantageously incorporated into artificial neural networks.