Technologies for and electrophysiological studies of structured, living, neuronal networks on microelectrode arrays

In this thesis, we have developed techniques for modifying and patterning surfaces for cellular growth and applied them to the recording of electrical activity of neurons grown in patterns. We found that silanes can attach to oxygen-plasma-treated polyimide and platinum and remain bound for at least 8 weeks. Attached silane also permits further chemical modification through cross-linkers. We also improved protein microstamping using a release layer and made it compatible with extended use and cell growth. Amphiphiles with various head groups were tested and shown to attract or repel proteins, suggesting the possibility of stamping supramolecular structures.; We also recorded neurons patterned by photoresist image transfer and found patterned neurons capable of firing action potentials. Their firing depended on neuronal communication as action potentials were suppressed by high magnesium media. In addition, we found that network activity depended on network morphology. Well-confined networks fired action potentials independent of cell density and varied minimally in activity level, while less-confined networks had both greater dependence and variation. We attributed this effect to the greater glia:neuron interaction as both glial and synaptic density were higher in patterned than in random cultures. This interaction may depend on cell density, as synaptic density was equivalent in both patterned and random cultures when cultures were initiated at a lower plating density, implying that neuronal growth was stunted when culture is sparse. Currently, we cannot distinguish the direct effects of patterning on synaptic efficacy from the indirect effects of glia because our experiments were not designed to distinguish them. Future experiments should be conducted to test the hypothesis that distributed connectivity weakens the contribution of each synapse to cellular depolarization.