The analysis of specific cell types is only the starting point for our quest to understand small neuronal networks, their physiology and their successive integration into larger networks in the brain. As an example, we are currently studying the feedback circuit of the optic tectum with the nuclei of the isthmic area. The latter nuclei are well-defined, compact structures in the midbrain with an almost exclusive in- and output relation to the optic tectum. This system has a comparable layout in most vertebrates, and has been studied in detail in various avian species. The the anatomical layout in birds is intriguing and suggests a function as a winner-takes-all circuit, possibly subserving the role of disambiguating the tectal saliency map which is fundamental to bottom-up models of attention. Since parts of the tecto-isthmo-tectal networks remain intact in a slice preparation of the midbrain, we have the unique opportunity to study the physiology and the temporal dynamics of a feedback circuit in a slice preparation, with all the advantagees of the in-vitro approach (ease of whole-cell patch recording, targeting of neurons under visual control, selective electrostimulation and pharmacological and microsurgical manipulation options).
The development of cellular interactions in the nervous system rely heavily of molecular gradients and cell-cell interactions through a plethora of membrane-bound molecules. The molecular cues that define parts of the tecto-isthmo-tectal networks (e.g., R-Cadherin) have been analysed in combined tracing/immunocytochemical studies. This knowledge offers up the possibility of manipulating the expression of specific factors during development (e.g., through RNAi) that could result in an altered layout of the feedback network. This in turn would allow us to investigate our concepts of the role of a given nucleus in the complex feedback loop directly, either in the slice or even in the intact behaving animal.

The multimodal map of space in the optic tectum is assembled during development and is dominated by the visual system which acts as a master map for the other sensory afferents. Thus, manipulations in the visual map results in plastic events leading to the alignment of other (e.g., the auditory map) with the new visual topography. This process has been worked out in a series of experiments in barn owls by Eric Knudsen, and can probably also be extended to the other spatial sensory  afferents such as somatonsensory, lateral line, or infrared vision. How the discrepancy in the sensory maps is read out and how this information leads to the plasticity in the afferent pathways, is still under debate. With intracellular methods, we have characterized the development of the auditory input to the optic tectum, and have identified a cell type in the tectum that might be the source of the error signal fed back into the auditory pathway. At the moment, we are investigating the multimodal midbrain network with a combination of intracellular labeling, patch physiology, tracing and immunocytochemistry, to get a deeper insight into the cellular mechanisms of multimodal alignment.

Neuronal networks have to be wired during embryonic development. This is especially pressing for precocial animals such as the chick, since the hatchling does not have a nesting phase where the fine-tuning of sensory input and cellular computation can occur. Thus, the determination of the neuronal networks largely has to be coded genetically and has to interact in a complex manner with molecular cues and electrical patterns of excitation prior to hatching. Over the years, we have analysed the dendritic development of several cell types in the optic tectum of chicken with and without embryonal manipulation (such as removal of the eye anlage early in development). To carry this analysis further, we have established a cell culture of chicken neurons from both the optic tectum and the auditory brainstem, and developed a method to sort neurons through magnet-based immunocoupling. We have studied the development of cellular characteristics under culture conditions, and delineated factors that control these steps. As a side aspect, we have collaborated with electrical engineers to eventually build a biohybrid system where we want to study the interplay of electrical stimulation and intracellular processes. This line of work also aims at reconstructing a given neuronal circuit in a biohybrid approach