Brains are made up of complex connections between many different types of neurons. Ultimately, these connections allow to process incoming (e.g., visual) signals and respond to them appropriately. Most of the neuronal wiring takes place during development. However, up to now, it has been largely unclear which processes ensure that the adult brain is correctly wired (and thus functional).

An interdisciplinary research team involving neurobiologists and mathematicians at the Freie Universität Berlin set out to investigate, both biologically and mathematically, the fundamental mechanisms that determine neuronal wiring. As a model system, they investigated the so-called “visual map formation” in fruit flies: The fly’s compound eyes consist of approximately 800 small eyes (ommatidia), each containing eight light-sensitive receptor neurons, six of which are tuned to detect motion and form the primary visual representation of the outside world in the brain.

Due to the optics of the compound eye, these six motion-detecting neurons in the same ommatidium receive visual input from different points in space (i.e., they have different visual axes). Hence, if these six visual inputs were processed by the same post-synaptic neuron, the fly would perceive a blurred image. However, during a process known as “neural superposition wiring,” the outputs of the 800×6 motion-detecting receptors are re-sorted so that six distinct receptor neurons that originate from six neighboring ommatidia but with identical visual axis are bundled together. The major question the researchers investigated is: how is this sorting process achieved during development?

In their research, recently published in Science, the team used intravital microscopy and derived mathematical rules from this spatially and temporally resolved data. They convincingly showed that “neural superposition wiring” is determined through self-organization.

Importantly, self-organization only requires a minimal set of local interaction rules and is robust to external perturbations and intrinsic errors. These attributes, along with evolutionary considerations, argue that self-organization may be a general principle applicable to other aspects of brain development and possibly development in general.

The work has emerged as a result of the MATH+ project EF3-2 and the ECMATH project CH20.

Original publication in Science:

Axonal self-sorting without target guidance in Drosophila visual map formation

Egemen Agi, Eric T. Reifenstein, Charlotte Wit, Teresa Schneider, Monika Kauer,

Melinda Kehribar, Abhishek Kulkarni, Max von Kleist, P. Robin Hiesinger

https://www.science.org/doi/10.1126/science.adk3043

LINKS

• Research Group Max von Kleist
• Research Group Peter Robin Hiesinger
• Research Unit „Robust Circuit“
• MATH+ Project EF3-2: Model-Based 4D Reconstruction of Subcellular Structures