pic

pic

Assuming optimal wiring and applying our morphological modelling technique, we study the relation between morphology and connectivity in small networks of neurons. These methods open up a realm of approaches sketched below when used in conjunction with various experimental data; see our collaboration partners in the links section.
I. Space filling and network structure emergence

If neurons are packed in an optimal manner in the restricted space given by the brain it becomes feasible to seek a generative model which describes this particular space-filling behaviour. Not only the branching structure of the single cell must be intrinsically optimal but also the location of the cells themselves, their organization in layers and ganglia, and their electric isolation and maintenance by glia. We therefore believe that from a precise input-output topography of a network follows a precise optimal geometric layout and the full morphology of all neurons and glia in the network must follow this layout. We study neurons and glial cells in networks of different brain regions in particular the dentate gyrus, the retina (see picture) and the cerebellum.


pic
II. Structural plasticity

What happens during wiring up of the brain undergoing structural plasticity, e.g. during development, a cortical lesion or adult neurogenesis? How does a neuron integrate into an existing circuitry? What better way to learn how neurons connect than to watch them do it when they are involved in structural remodelling! We study these questions using time-lapse reconstructions of developing neurons in the fly PNS (see picture), of maturing newborn neurons in the adult olfactory bulb and of newborn neurons in the dentate gyrus which integrate into the existing functioning circuitry (or when axonal inputs were depleted after entorhinal cortex lesion).
Active collaboration with Gaia Tavosanis

pic



III. Connectomics

In order to understand how an entire circuitry is connected, it is important to study the architecture of dendrites and axons in the context of a whole network of cells. We study full connection maps, the connectome, of all cells in a network such as the stereotypic dendritic and axonal forests of Purkinje cells in the cerebellum using mice with selective brainbow expression and by exploiting their underlying growth principles for a model-based automated reconstruction.

pic

IV. Selective wiring and function

The technique of morphological modelling enables us to produce large anatomically realistic neural networks that are based on specific wiring schemes. We study the optic flow processing system in the lobula plate of the fly for which both the precise wiring of its interneurons as well as its function are well known and electrophysiological data is available for many of the neurons involved and their connections. We are also involved in an ambitious project with the aim of replacing reconstructions of real in vivo cortical neurons by synthetic ones to generate a synthetic anatomical model of the cortical column. The resulting anatomical components will in turn be used in a corresponding network model to investigate the propagation of neural signals there.
pic
V. Neuroinformatics: the TREES toolbox

Our theories and algorithms allow us to better formalise neuronal branching and have led to the development of a vast software package, the TREES toolbox, with which one can analyze, edit, simplify, visualize and database neuroanatomical data. These tools are ready to be used in large-scale models with accurate anatomical morphology and connectivity, integrating well in the newly emerging field of Neuroinformatics.
pic

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 License
last updated on 21 March 2024, Hermann Cuntz