A computational study of the mammalian subiculum

Abstract

The subiculum occupies a central position as the mediator of information flow between the hippocampal memory system and both cortical and subcortical brain structures. To further the goal of understanding how the subiculum fulfils this critical role, we constructed detailed biophysically realistic computational models of subicular principal neurons. In chapter III we developed Hodgkin-Huxley and Markov models of a total of fourteen sodium, potassium, calcium, and mixed cation currents that are prominent in pyramidal cells of the hippocampal complex. In chapter IV active models of subicular principal cells were created by selectively incorporating subsets of these currents into models of passive subicular membranes. Mechanisms for an unprecedentedly wide range of electrophysiological features (EPFs) of subicular pyramidals were simulated. These included inward and outward membrane rectification, M- and H-resonance, burst and single spiking, frequency-dependent bursting, spike frequency adaptation, plateau potentials, delayed firing, afterhyperpolarizations and rebound firing. The number of electrophysiological events treated in this thesis necessarily constrains the level of detail at which many EPFs were investigated. However, given the evident importance of burst firing within the brain (Lisman, 1998), particular attention was paid to the mechanisms, ionic composition, and regulation of the subicular burst event. A novel theory of subicular burst firing, the adaptive two-component burst firing theory, was proposed. A central theme of this thesis is that the expression of subicular EPFs are subject to adaptive regulation. Evidence for this hypothesis was marshalled and major findings within the subicular electrophysiological literature were re-interpreted from the adaptive EPF perspective. In chapter V the evidence for the adaptive regulation of subicular frequency preferences, bursting, and spike frequency adaptation was discussed in more detail along with the possibility of integrated adaptive control mechanisms. Finally, the limitations of the prevailing electrophysiological analysis paradigm were noted and a novel empirical framework for investigating adaptive EPFs, feature-based analysis, was proposed.