Bottom-up assembly of synthetic neural networks from biological matter

Introduction

The long-term objective of my lab is the construction of “synthetic” neural networks from biological matter. In this way, we not only will understand how to build sustainable computing architectures but also provide a novel bottom-up approach towards understanding biological neural networks.

Research Question

To make progress in this direction, I ask: What is the minimal assembly of biological components that allows for bio-realistic electrical information processing?

Importance of Ion Channels

Clearly, voltage-sensitive ion channels form the molecular basis for electrical spiking activity in neuronal networks. However, spatial propagation of spikes is not a property inherent to individual ion channels but rather emerges from the arrangement of ion channels along a tubular lipid membrane, the axon.

Current Challenges

While the reconstitution of functional ion channels outside of living cells has been well established, the propagation of an action potential along a lipid bilayer nanotube has not yet been shown. Similarly, the physical realization of larger, non-living spiking networks using biological matter remains elusive.

Proposed Approach

Here, I propose to move forward from the state-of-the-art by:

1. Studying action potentials propagating along lipid nanotubes and demonstrating their electrical cable and spiking characteristics.
2. Understanding the coupling between membrane elasticity and electrical characteristics and how electromechanical coupling remodels and reshapes membranes and nanotube networks.
3. Exploiting these remodeling, reshaping, and self-healing abilities for biomimetic molecular mechanisms of information processing.

Complexity and Methodological Advances

This research is a challenge of enormous complexity. In this proposal, I argue that this challenge can be overcome using recent methodological advances.

Conclusion

My preliminary data and my research experience combining electrical engineering, biophysics, and synthetic biology will enable this leap in our understanding and design of biological neural networks.