The goal of this work is to elucidate evolving multimodal interactions between structural, dynamical, and functional network properties based on the interdependencies observed between neuronal and astrocytic networks acting on diverse spatial and temporal scales. Specifically we will investigate how dynamical and structural network characteristics interact on these different time scales to form evolving, functional neural ensembles. To achieve this goal we plan to combine computational modeling with experimental approaches that include calcium optical imaging, multi-electrode recordings, and structural labeling studies in primary neuronal cultures. This will allow for monitoring of multi-scale, simultaneous dynamical and structural changes in networks under different conditions. In particular we want to address the following questions: What properties of spatio-temporal patterning are mediated through fast and/or slow network interactions? How does network connectivity influence multimodal network activity? Whether and how do local network changes modify local patterns of fast and slow dynamics? Finally, we want to understand how the functions of these dynamical modes evolve during network development.
Networks of interacting elements are ubiquitous in nature and man-made systems. The interactions in these networks happen on different spatial and temporal scales. Neural networks are a prime example of such a complex system. While interconnected neurons establish fast modes of communication, the slower astrocytes can modulate overall activity in the network that in turn will affect its connectivity structure and function. It becomes then pertinent to understand the interactions between these two modes of transmission and what their differential roles in network function are. This interdisciplinary project connects neurobiology with dynamical systems approaches to tackle these problems and can potentially have important impact on diverse fields. Identifying dynamical mechanisms underlying these interactions will allow for a better understanding of the network processes that underlie both cognitive tasks and brain pathologies: for example, on one hand structural network reconfiguration is crucial for memory formation but at the same time may drive formation of epileptic seizures. At the same time, the use of these dynamical modes for controlled network development may also lead to formulation of new strategies for network repair after injury. Finally, the insight obtained from this work could be generalized and applied to man-made networks that employ multi-scale temporal dynamics and could undergo functional reorganization online.
This project will also provide opportunity for graduate and undergraduate students to experience interdisciplinary research connecting biology, physics and dynamical systems theory.
The brain is an extremely complex network that works on multiple spatial and temporal scales. Interactions on these scales underlie what we commonly call cognition. To understand them we need to apply advances from biology, engineering, math and physics. The goal of this cross-disciplinary project was to investigate the role that such multi-scalar (or multimodal) interactions play in formation of spatial and temporal activity patterns in the neural systems, and consequently their role in information processing in the brain. The project pursued these questions both experimentally and through computational modeling. In experiments we focused on possible role of astrocytes in shaping network dynamics. Here we have a clear separation of time scales as neurons emit electrical signals (spikes) on fast timescales of about ~1ms, while astrocytic dynamics is dominated by slow calcium release and propagation (~hundreds of milliseconds timescale). We recorded neuronal activity using multi-electrode arrays (MEA) and at the same time we performed optical imaging to monitor astrocytic calcium activity. We studied these interactions in cultured hippocampal networks. We specifically have investigated what happens if we disrupt connectivity of astroctytic network (formed via gap junctions). We showed, that when we disrupt this connectivity the patterning of astrocytic activity changes significantly consistent with existence of extracellular coupling mechanism based on diffusive transport. In terms of computations we pursued couple of avenues of research: We constructed a reduced model of multimodal interactions in the network to grossly simulate astrocytic interactions in cultures. The modeled cells were communicating actively through network connections (fast transport) and through extracellular diffusion (slow transport). We show that these two facets of communications sometimes cooperate, while at other times compete, depending on network topology. We investigated how slow, neuromodulatory effects of acetylholine (Ach) change spatio-temporal dynamics of a network of cortical neurons. We also investigated the role this mechanism can have in anatomic network reorganization during sleep. We showed that this dynamic mechanism could underlie process known as synaptic renormalization (overall downscaling of synapses during slow wave sleep). We proposed a novel mechanism underlying information processing in the brain based on slow, voltage dependent changes in neuronal resonance frequency. Namely we argue that small, sub-threshold synaptic input can lead to overall membrane depolarization, which in turn causes the resonance shift. We showed that this effect, coupled with global oscillatory rhythms (ubiquitous in the brain) can underlie robust feature binding and separation mechanism needed for example for storage and retrieval of memories. In all we showed that multi-scalar interactions can dramatically alter network dynamics and thus control brain function. More importantly they cannot be considered independently. The results obtained as a part of this project were published in 7 articles and 3 additional articles will be published shortly. The results were also presented on numerous conferences.
