Designing in vitro patterned neuronal networks

Advances in neuroscience and bioengineering have led to a nascent discipline known as neural engineering, whose activity ranges from neurally inspired computer algorithms to brain and peripheral nerve prostheses to imaging of functional brain activity. This paper describes progress in a neural engineering research area in which the goal is to learn how to construct neuronal networks de novo with individual living nerve cells as the fundamental circuit elements. The effort is in part microlithography, applicable because the sizes of neurons and transistors are the same. In part it is surface chemistry, as it is necessary to couple diverse chemistries and molecules to surfaces in order to control neuron attachment and growth. It is in part electrical engineering, as the principal I/O devices are microelectrode arrays capable of recording or stimulating at dozens and in the future hundreds of sites. Modern data acquisition and real-time analysis are also necessary to store, recognize, and analyze the multichannel data. Overwhelmingly, however, the problems are ones of neuroscience and cell biology, of controlling the many factors that influence cell attachment, growth, and interactions with other cells. Despite this seemingly unending list of prerequisites, substantial progress is being made to the point where we can seriously begin to ask a variety of scientifically relevant questions. One of these is to ask whether or not these biological neuronal networks behave at all like their computational namesakes.