Alexander D. Reyes - US grants

We know that the brain is a highly efficient computational machine but little is known about how neurons in the central nervous system process information. Deciphering the neural code is made difficult by the enormous and complex set of variables at both the cellular and systems level that determine the behavior of a neuron and of the network in general. Neurons in a network communicate via small electrical signals termed postsynaptic potentials (PSPs). A neuron integrates the PSPs from thousands of other neurons; when the composite signal is sufficiently large, a large all-or-none electrical event, termed the action potential (AP; which is the equivalent of a binary digit), is generated. The AP is subsequently transmitted to other parts of the nervous system where it results in a PSP in other neurons, thereby repeating the process. Information is encoded in the timing and frequency of the APs. The cortex contains a vast array of neurons, each with different biophysical properties. Designing experiments that control for these variables is challenging.
While biophysical studies permit detailed characterization of subcellular ion channels in single neurons, they do not directly address how these properties translate to encoding in a network of neurons. Conversely, behavioral electrophysiology describes the behavior of neurons under natural conditions but provides little data about the underlying cellular mechanisms. In the current study, the strategy is to interface a computer with a live neuron. By using computer simulations to generate realistic inputs into the neuron, the manner in which a neuron processes signals within a neural network can be examined systematically and be correlated directly with its biophysical properties. The data obtained from these experiments will provide a solid foundation for constructing realistic models of networks, developing better computing algorithms and hardware, and designing neural prostheses.