| dc.description.abstract | Space is a fundamental substrate for animal interactions with the external world. The hippocampus, a region deep within the temporal lobe of mammalian brains, is a multisensory structure that has been heavily implicated in spatial cognition. Systematic intra- and extra-cellular recordings from the hippocampus have revealed the prevalence of temporal patterns of neural activity that embody spatial encoding and memory storage. Among these patterns, hippocampal theta oscillations (4–10 Hz) are implicated in the encoding of spatial information, and are prominently observed when the animal is exploring external space or during rapid eye movement sleep. In contrast, ripple-frequency oscillations (120–250 Hz) are associated with memory consolidation and are largely observed during slow wave sleep or wakeful immobility. The questions addressed in this thesis focus on the implications for, the origins of and the co-existence of these two temporal patterns of neural activity.
A high-resolution phase code of space within a hippocampal neuron’s place field manifests as a monotonic advancement of spike phase with respect to the extracellular theta rhythm, and is referred to as phase precession. In the first part of the thesis, we built a heterogeneous population of biophysically detailed neuronal models and posed phase-coding efficiency as an information maximization problem. Our analyses unveiled a permissive role for intrinsic neuronal properties in efficient phase coding, and suggested parametric degeneracy as a framework for achieving the twin goals of efficient encoding and homeostasis. Importantly, we suggest a synergistic counterbalance between neural gain and synaptic strength as a functional constraint for efficient phase coding. Our analyses also advocated the intricate balance between excitatory inputs, inhibitory inputs and neuronal intrinsic excitability (E-I-IE balance) as a generic mechanism that could govern stable and efficient encoding.
In the second part of thesis, we quantitatively address the origins of ripple frequency oscillations in the hippocampus. We specifically focused on the disparity in the firing rates (30–40 Hz) of individual interneurons during sharp wave ripple complexes in vivo and ripple-frequency oscillations (120–250 Hz) that have been postulated to emerge from networks of such interneurons. We built a network model that incorporated different degrees of two distinct forms of biological heterogeneities, with reference to the amplitude and duration of afferent stimuli onto individual neurons in the network. Our simulation outcomes accounted the experimental disparity in individual neuronal firing rates and ripple frequency oscillations when the mean excitation amplitudes were low and the degrees of heterogeneity were high. Thus, our emergent model for ripple-frequency inhibition proposes biological heterogeneity as a feature of the network that aids spike dyssyncrony, which collectively yields oscillations in a frequency range that is higher than the firing frequency of individual neurons.
The third part of the thesis challenges the prevalent behavioral and physiological dichotomy between the theta- and ripple-frequency oscillations. Based on several existing experimental observations, we hypothesized that theta-frequency oscillations and ripples should coexist. We tested this hypothesis using in vivo polytrode recordings from multiple strata of the hippocampus of foraging rats, and assessed the coexistence of theta oscillations and ripples using cross-electrode analyses. Strikingly, we found ~15–20% of the total detected ripples to occur within theta epochs (theta ripples) at frequencies higher than that of non-theta ripples. Behaviorally, although ripples are traditionally considered to occur during immobile non-theta periods, we found a considerable fraction of ripples in other categories (20–35%). Our observations point to potential encoding roles for ripple events that occur during theta oscillations. | en_US |
