Our lab's goal is to understand cell signaling and information processing in biological cellular neural networks in the brain and retina (which is an extension of the brain), and how the break down of these processes contribute to neurological disorders. A broad goal is understanding how the complex dynamics of the brain at a systems level emerge from (often stereotyped and ubiquitous) molecular and cellular foundational processes.
In particular, we are interested in the mechanisms that underlie signal and information propagation in biological cellular neural networks, and the computational potential of such networks in the brain. There are some key things a network of cells must be able to do: Store different pieces of information, morph or modify stored information, and manipulate stored information. These must occur in some logical and physically constrained way that results in a meaningful outcome for the organism. A broad objective of this work aims to understand how given a network of a certain size with a specific set of operational parameters (i.e. a defined amount of degrees of freedom), one can tell what kinds of structures (i.e. classes of objects) such a network can represent, how much information it can store (i.e. the size of the objects it can hold), and how a network can manipulate such objects. The goal is to identify the minimum set of properties a network must have to do each of these, so that given a task or objective (i.e. something the organism needs to accomplish, whether it be sensory recognition, learning something new, or responding to an input with a specific motor or cognitive output) the kinds of circuits and networks needed and how they necessarily must interact to accomplish the task can be predicted. The objective is not to do this from a purely theoretical perspective, but by developing theories that can be directly experimentally validated and put into a practical neurobiological context.
Another specific interest in our lab are glial neurobiology and neuron-astrocyte interactions. Astrocyte neural glial cells engage in bi-directional chemical signaling with neurons and have the ability to modulate and directly participate in information processing in the brain, which necessitates more than just interactions between neurons and almost certainly involves astrocytes somehow. At present though, despite beginning to uncover the molecular and cellular details, we do not know or understand what astrocytes are doing in the brain in the context of information processing, the extent of their functional roles, and if such functions differ from one brain region to another. Neuron-astrocyte interactions likely result in emergent dynamic properties at the cellular network level that ultimately determine behavioral and cognitive processes in the organism, but we do not understand how.
In certain parts of the brain astrocytes outnumber neurons 10 to 1, resulting in a network engaged in information processing ten times the size of the neuronal network in terms of the number of nodes and several orders of magnitude greater in the number of links or functional connections between nodes. In fact, in some parts of the brain one astrocyte usually connects to six neurons but forms between 100,000 to 140,000 synaptic connections with those neurons. As one moves up the phylogenetic tree, species have more and more astrocytes in their brains as a function of brain complexity culminating with humans, which have the greatest number. Indeed, anecdotally several decades ago neuroscientists made the observation that Albert Einstein's brain had a larger number of astrocytes than an average brain after parts of Einsteinâ€™s brain were sectioned and histologically stained following his death. More empirically, recent work published earlier this year has shown that human astrocytes are quite different than rodent astrocytes: they are structurally more complex and display more complex signaling. Astrocytes might not just be participating in information processing in the brain, they might actually be making us smarter.
We approach these questions by developing and using experimental and computational methods in order to reverse engineer how the nervous system is built, so that we can understand how it functions. Our lab operates at the interface of experiment, theory, and computation, integrating experimental neurobiology and physiology with engineering, physics and chemistry. Experimentally we rely on optical imaging methods, as well as traditional molecular and cellular neurobiology methods. More technologically intensive, we are very engaged in the development of nanotechnologies as biosensors for neural cells in order to study both individual cells and neural circuits and networks. Theoretically and computationally we are developing mathematical and physical models for identifying and mapping functional signaling and information propagation in biological neural networks, and neurophysiological and biophysical models that provide mechanistic insights. Computer science and engineering comes into all of this by providing the theory and tools necessary to computationally implement the models, and our lab is engaged in using and pushing the limits of graphics processing unit (GPU) computing in neurophysiological and systems neuroscience. Finally, we are also interested in neural engineering applications, in particular retinal neural prosthesis and technologies for interfacing with the nervous system.
D Yu, M Buibas, Z Singer, I Lee, and GA Silva (2009) Characterization of calcium mediated intracellular and intercellular signaling in the rMC-1 glial cell line. Cellular and Molecular Bioengineering 2:144-155.
M Buibas, D Yu, and GA Silva (2009) A framework for simulating and estimating the state and functional topology of complex dynamic geometric networks. ArXiv e-print (article ID 0908.3934v1).
C MacDonald, D Yu, Buibas, M. and GA Silva (2008) Diffusion modeling of ATP signaling suggests a partially regenerative mechanism underlies astrocyte intercellular calcium waves. Frontiers in Neuroengineering 1:1-13.
GA Silva (2006) Neuroscience nanotechnology: Progress, challenges, and opportunities. Nature Reviews Neuroscience 7:65-74.
GA Silva, C Czeisler, KL Niece, E Beniash, D Harrington, JA Kessler, and SI Stupp (2004) Selective differentiation of neural progenitor cells by high-density epitope nanofibers. Science 303:1352-1355.