Engineering Human Neural Circuits with Single-Cell Precision

Biological neuronal networks (NNs) can be studied <em>in vivo</em> or <em>in vitro</em>. Although <em>in vivo</em> studies allow for the observation of processes in a natural setting, such as a freely behaving animal, readout and interaction with the neuronal system are typically limited due to experimental constraints. <em>in vitro</em> approaches on the other hand offer large-scale readout and interaction capabilities as well as better accessibility. These networks can be engineered to meet experimental needs; however, achieving precise, reproducible control of their architecture at the single-cell level has so far been out of reach. <br> In this work, I show for the first time how this limitation can be overcome and how <em>in vitro</em> networks, recorded with multielectrode arrays (MEAs), can serve as versatile models to study neuronal function, physiology, and development under varied conditions. First, I studied NN activity under altered gravity. While experimental platforms for altered gravity conditions inherently pose challenges for biological experiments, we were able to integrate and maintain our networks. We demonstrated that altered gravity influences human NN activity, resulting in altered firing and bursting in response to gravitational stimuli, as well as adaptation processes. Second, I used high-density recording technology to study the longitudinal morphological and physiological development of NNs. We observed the formation of neuronal clusters, morphological changes, and the increase of GABAergic neurons over time. These observations have also been reported <em>in vivo</em>. Third, I used <em>in vitro</em> networks and tried to teach them a classification task through continued electrical stimulation. Although I observed more coherent activity, I did not observe signs of learning or plasticity. Finally, I developed a method to construct NNs with single-cell precision in a reproducible and controlled manner, enabling the measurement and validation of ephaptic coupling effects, as predicted by computational modeling, which was previously difficult to achieve. I could confirm significant changes in action potential velocity, synchronicity and other measures as a result of ephaptic coupling. <br> Overall, I demonstrated that <em>in vitro</em> neuronal systems are well-suited to answering specific questions about human NN function and development. Our novel method of engineering NNs from the bottom up with single-cell precision has the potential to be applied to disease modeling, drug development, the study of optogenetic tools, and the validation of computational model predictions, as well as fundamental neuroscience research and the study of neuronal computation.