There appears to be a paradox in brain research. On the one hand, our brain can store memories for decades. On the other hand, it is highly adaptive. How can the brain satisfy these seemingly contradictory requirements? Henning Sprekeler intends to examine these and many other questions in detail. He received the Bernstein Award 2011 from the German Federal Ministry for Education and Research (BMBF) for this project, and with this grant he will over the next five years establish an independent research group at the Humboldt-Universität zu Berlin (HU) and the Bernstein Center Berlin.

Henning Sprekeler. © Henning Sprekeler.

Memories can only be stored long-term in the neuronal networks of our brain if the structure and activity of these networks remain stable, regardless of time and external influences. But how does the brain ensure this stability? “We have known for many years that the activity of neuronal networks must be very well balanced in order to function properly,” explains Sprekeler. For the network to be balanced, it is essential that the activity of excitatory and inhibitory neurons is in balance: “If inhibitory neurons have too little influence, the excitatory cells synchronize sooner or later. However, if inhibitory cells have too much influence, the neurons of the network fall almost silent. Both can critically disturb brain activity,” describes the brain researcher. Disruptions of this balance seem to play an important role in neurological disorders such as epilepsy and schizophrenia. Also because of this, research into the “activity of balanced neuronal networks” has come more into the focus of neuroscience lasting recent years.

Synaptic plasticity often shows a characteristic, asymmetric dependence on the relative timing of pre- and postsynaptic activity (characteristic graph sketched as balance). The central goal of Sprekeler’s research agenda is to understand how such forms of synaptic plasticity interact with the observed balance between excitatory and inhibitory currents in cortical networks. © Henning Sprekeler.

In addition to its stability, the flexibility of the brain is extremely important for its function to learn new content. It has long been known that the process of learning causes changes in the strength of connections between neurons, as well as forming new connections or removing old ones. This connectivity plasticity has been extensively studied from the molecular level up to largescale network models. Therefore, nowadays we have a much better understanding of the changes which are set in motion. But still, many basic questions remain unanswered.

The plasticity of neuronal connections, for example, which is essential for the learning process, automatically changes the influence of the excitatory or inhibitory neurons in the network. What mechanism prevents the network activity getting out of hand or becoming silent at such moments? How does the brain avoid these instabilities, which are almost inevitable when neuronal plasticity is introduced into strongly recurrent - and therefore biologically realistic – model networks? So far, not many have investigated these questions. “Once certain neurons obtained more influence, the entire balance ran out of control and the network ’exploded’, so to speak,” describes Sprekeler on the situation up to now.

Past research focused primarily on the plasticity of excitatory neurons and paid less attention to changes in inhibitory neurons. Together with colleagues, Sprekeler has now developed a model that could explain how a continuous balance between activation and inhibition persists in complex networks when the balance of power changes by learning. To this end, the scientists paid more attention to the plasticity of inhibitory neurons than previous models. “In this way we can now systematically investigate the influence of learning processes in balanced networks.” In initial studies, the researchers already found that the readjustment of the balance in turn affects the learning process - how this happens exactly, is as yet unexplored. Therefore, over the next few years, Sprekeler would like to develop models that simulate the plasticity and self-organization of networks, in particular, taking the inhibitory elements into account.

Sprekeler belongs to the theoreticians among neuroscientists. He studied physics in Freiburg and Berlin. “Originally I was interested in mathematical signal analysis in neuroscience, a field whose mathematical methods were familiar to me from quantum mechanics,” he explains. “Over the years, my interest has steadily moved towards biology.” His PhD thesis on “Slowness Learning: Mathematical Approaches and Biological Mechanisms” in the lab of Laurenz Wiskott at the HU Berlin finally marked his arrival in Computational Neuroscience. A stay of over two years in the laboratory of Wulfram Gerstner at the Brain Mind Institute of the Ecole Polytechnique Federale de Lausanne in Switzerland, from 2008 onwards, brought him closer to experimental research. “I wanted to work with more experimentally oriented scientists. As a theoretician, one must get close to the experiment to understand which models make sense and which don’t.” He now wants to contribute this experience to the Bernstein Network.

But his theory-oriented background is still important to him. With his projects within the Bernstein Award, he also wants to get back to his academic roots. He seeks, among other things, for answers to fundamental questions in neuroscience. For example, on which principles do neuronal networks develop and organize themselves? He is also interested in how stimuli are encoded in the brain. By receiving external stimuli, we learn about our environment and develop an internal representation of the world. What are the effects of these stimuli and the associated learning processes on the development, structure and activity patterns of the brain? He is interested in whether simple principles are veiled behind these developmental and organizational processes and what they might look like.

When Sprekeler returned to the HU Berlin in early 2011, he was able to quickly develop and expand connections to other members of the Bernstein Network: besides Laurenz Wiskott, Richard Kempter and Susanne Schreiber, with whom he had been in contact since the times of his PhD thesis, collaborations are planned with Michael Brecht and Dietmar Schmitz, among others. There, Sprekeler will examine the data transfer from the hippocampus into the cortex and the development of episodic memory. “I want to contribute to a future understanding of how complex networks like the brain learn,” describes the Munster-born researcher his motivation in brain science. What does the Bernstein Prize mean to him? “It gives me the security to calmly pursue my research in the next few years. For that, I am very glad and grateful.” This calm will be needed while devoting the next few years to the paradoxes of brain research.