skip to content

The Fly’s Brain: DFG Research Unit Investigates Circuits and Memory in the Fruit Fly

Subproject at the University of Cologne aims to implement memories as plastic networks / scientists hope for insights into how to control the behaviour of robots

The neuronal processes underlying memory formation and their structure are the focus of the Research Unit ‘Dissection of a Brain Circuit: Structure, Plasticity and Behavioral Function of the Drosophila Mushroom Body’ (FOR 2705), established by the German Research Foundation (DFG).

Professor Martin Paul Nawrot from the Institute of Zoology at the University of Cologne heads a subproject of the Research Unit. Using the model organism of the fruit fly Drosophila melanogaster, he explores how simultaneous and temporally distributed memories are formed in the flexible network of the fly brain. ‘The aim of my subproject is to be able to model the highly complex process of memory performance as a biologically realistic and at the same time functional neural network’, explains Nawrot. With the help of the model, individual memory traces in the fly brain can be precisely localized.

How memory traces unfold in the complex architecture of the brain and how they develop over time in the plastic network depends on individual experiences and determines individuality of behaviour. Nawrot has already investigated the relationship between learning and behaviour in another project using the model organism of the honeybee. ‘The bee is indeed the best established model organism for learning, especially in behavioural studies. Bees are extremely good at learning, which means that they can also carry out complex learning tasks that are normally only found in vertebrates’, says Nawrot.

The fly’s memory is not comparable to that of honeybees, also because a fly’s brain is ten times smaller than that of a honeybee. However, in contrast to bees, Drosophila has the great advantage that various biological techniques have already been developed to investigate and manipulate its brain. Thanks to these tools, scientists know much more about the anatomy and function of the fly’s brain. Nawrot’s subproject at the University of Cologne incorporates a large amount of anatomical and functional data on Drosophila into the modelling of the memory network. These data come partly from the research group of Professor Dr. Kei Ito, who came to the University of Cologne from the University of Tokyo in 2016. As an associated member of the research group, he provides his expertise, methods and genetically modified fly lines.

‘Whether bee or fly, despite their relatively small brains, insects show complex and adaptive behaviours. The limited neuronal resources in particular make it necessary to decipher stimuli arriving in the brain with very little effort and to transform them into actions. We can learn a lot from these net-like circuits in order to find out basic characteristics of the functioning of the nervous system’, Nawrot reflects.

In Nawrot’s project, for example, the behaviour of an autonomous robot with approximately 4,000 artificial ‘nerve cells’ will be controlled in real time. The zoologist explains: ‘Our robot can see through a camera and collect visual stimuli such as light or colour. It learns that the perception of a certain colour requires a certain action to gain a ‘reward’. The more frequently it performs the action correctly, the more stable becomes the pathway from sensory input to motor output  in the connections – or synapses – between the artificial nerve cells.’ Nawrot already came to similar conclusions in his honeybee study: The more often a stimulus was rewarded, the more stable became the memory trace, which literally ‘burned itself into the memory’. This plasticity rule, as Nawrot calls it, inspired the research group to implement a neuronal spiking network – a biologically realistic network with neurons that induce behaviour – on an analogous microchip for the robots.

FOR 2705 is managed by Georg-August-Universität Göttingen (Professor Dr. André Fiala) and has six subprojects in Cologne, Bonn, Göttingen, Munich, Magdeburg and Berlin. It will be funded for an initial period of three years. The research group plans to organize a large annual conference. The first one is planned to take place in May 2019.

Media Enquiries:     
Professor Dr. Martin Paul Nawrot
Institute of Zoology at the University of Cologne
+49 221 470-7307
mnawrotSpamProtectionuni-koeln.de

Press and Communications Team:
Frieda Berg
+49 221 470-1704
f.bergSpamProtectionuni-koeln.de