A Synthetic Biology Approach to the Evolutionary Transformation of Visual Cortex Architecture
1 Campus Institute for Dynamics of Biological Networks; University medical center Göttingen, Institute for Neuropathology; Max-Planck-Institute for Experimental Medicine, Göttingen
2 Campus Institute for Dynamics of Biological Networks
3 Princeton Neuroscience Institute
4 University medical center Göttingen, Institute for Neuropathology
5 NYU Grossman School of Medicine, Neuroscience Institute
6 Max-Planck-Institute for Experimental Medicine, Göttingen
The arrangement of orientation selective neurons in the visual cortex into functional domains is a common feature over a wide range of mammalian taxa. This architecture presumably evolved from a primordial rodent-like salt-and-pepper layout. Here, we present a synthetic biology approach to investigate transition scenarios between the two types of functional architecture and assess evolutionary benefits that might have driven the emergence of functional orientation domains.
We used a bidirectional neuronal interface to connect a computational model of the retino-thalamic pathway to living neural tissue that served as a model for cortical input layer. The latter was either an in-vitro culture of dissociated cortical neurons or an acute brain slice. The in-silico components were connected to the neural tissue via optogenetic holographic stimulation. Neural activity was registered either by multielectrode array recording or optically using a genetically-encoded calcium indicator.
In this system, we implemented the Hubel&Wiesel feed-forward model of orientation tuning and explored the consequences of scaling orientation domains down to the size of a single neuron. We found that the fraction of orientation selective neurons only weakly decreases with shrinking domain size. Intriguingly, even in the absence of orientation tuned input a considerable level of orientation selectivity was retained. In this limit of infinitesimally small domains, the arrangement of orientation selective cells resembles a sparse salt-and-pepper layout.
Our results demonstrate that neural interface technology can be used to study evolutionary transitions in neuronal computation by a synthetic biology approach.