Christian Tetzlaff’s research while affiliated with University of Göttingen and other places

Intrinsic within-type neuronal heterogeneity is a ubiquitous feature of biological systems, with well-documented computational advantages. Recent works in machine learning have incorporated such diversities by optimizing neuronal parameters alongside synaptic connections and demonstrated state-of-the-art performance across common benchmarks. However, this performance gain comes at the cost of significantly higher computational costs, imposed by a larger parameter space. Furthermore, it is unclear how the neuronal parameters, constrained by the biophysics of their surroundings, are globally orchestrated to minimize top-down errors. To address these challenges, we postulate that neurons are intrinsically diverse, and investigate the computational capabilities of such heterogeneous neuronal parameters. Our results show that intrinsic heterogeneity, viewed as a fixed quenched disorder, often substantially improves performance across hundreds of temporal tasks. Notably, smaller but heterogeneous networks outperform larger homogeneous networks, despite consuming less data. We elucidate the underlying mechanisms driving this performance boost and illustrate its applicability to both rate and spiking dynamics. Moreover, our findings demonstrate that heterogeneous networks are highly resilient to severe alterations in their recurrent synaptic hyperparameters, and even recurrent connections removal does not compromise performance. The remarkable effectiveness of heterogeneous networks with small sizes and relaxed connectivity is particularly relevant for the neuromorphic community, which faces challenges due to device-to-device variability. Furthermore, understanding the mechanism of robust computation with heterogeneity also benefits neuroscientists and machine learners.

Recent technological developments in molecular biology led to large data sets providing new insights into the molecular organisation of cells. To fully exploit their potential, these developments have to be complemented by computer simulations that allow to gain in-depth understanding on molecular principles. We developed the Python-based, reaction-diffusion simulator PyRID, integrating many features for the efficient simulation of molecular biological systems. Amongst others, PyRID is capable of simulating unimolecular and bimolecular reactions as well as pairinteractions to assess the dynamics resulting from individual interacting proteins to polydisperse substrates consisting of many different molecules. Furthermore, PyRID supports mesh-based compartments and surface diffusion of particles, enabling analyses of the interaction between (trans-)membrane proteins with intra- and extracellular proteins. PyRID is written entirely in Python, which is a programming language being known for its readability and easy accessibility, such that the scientific community can easily extend PyRID to its current and future needs.

A bstract Stimulus-triggered synaptic long-term plasticity is the foundation of learning and other cognitive abilities of the brain. In general, long-term synaptic plasticity is subdivided into two different forms: homosynaptic plasticity describes synaptic changes at stimulated synapses, while heterosynaptic plasticity summarizes synaptic changes at non-stimulated synapses. For homosynaptic plasticity, the Ca ²⁺ -hypothesis pinpoints the calcium concentration within a stimulated dendritic spine as key mediator or controller of underlying biochemical and -physical processes. On the other hand, for heterosynaptic plasticity, although theoretical studies attribute important functional roles to it, such as synaptic competition and cooperation, experimental results remain ambiguous regarding its manifestation and biological basis. By integrating insights from Ca ²⁺ -dependent homosynaptic plasticity with experimental data of dendritic Ca ²⁺ -dynamics, we developed a mathematical model that describes the complex temporal and spatial dynamics of calcium in dendrites.We show that the influx of calcium into a stimulated spine can lead to its diffusion to neighboring spines, triggering heterosynaptic effects such as synaptic competition or cooperation. By considering different input strengths, our model explains the ambiguity of reported experimental results of heterosynaptic plasticity, suggesting that the Ca ²⁺ -hypothesis of homosynaptic plasticity can be extended to also model heterosynaptic plasticity. Furthermore, our model predicts thata synapse can modulate the expression of homosynaptic plasticity at a neighboring synapse in an input-timing-dependent manner, without the need of postsynaptic spiking. The resulting sensitivity of synaptic plasticity on input-spike-timing can be influenced by the distance between involved spines as well as the local diffusion properties of the connecting dendritic shaft, providing a new way of dendritic computation.

Synaptic proteins need to be replaced regularly, to maintain function and to prevent damage. It is unclear whether this process, known as protein turnover, relates to synaptic morphology. To test this, we relied on nanoscale secondary ion mass spectrometry, to detect newly synthesized synaptic components in the brains of young adult (6 mo old) and aged mice (24 mo old), and on transmission electron microscopy, to reveal synapse morphology. Several parameters correlated to turnover, including pre- and postsynaptic size, the number of synaptic vesicles and the presence of a postsynaptic nascent zone. In aged mice, the turnover of all brain compartments was reduced by ∼20%. The turnover rates of the pre- and postsynapses correlated well in aged mice, suggesting that they are subject to common regulatory mechanisms. This correlation was poorer in young adult mice, in line with their higher synaptic dynamics. We conclude that synapse turnover is reflected by synaptic morphology.

Ripples are a typical form of neural activity in hippocampal neural networks associated with the replay of episodic memories during sleep as well as sleep-related plasticity and memory consolidation. The emergence of ripples has been observed both dependent as well as independent of input from other brain areas and often coincides with dendritic spikes. Yet, it is unclear how input-evoked and spontaneous ripples as well as dendritic excitability affect plasticity and consolidation. Here, we use mathematical modeling to compare these cases. We find that consolidation as well as the emergence of spontaneous ripples depends on a reliable propagation of activity in feed-forward structures which constitute memory representations. This propagation is facilitated by excitable dendrites, which entail that a few strong synapses are sufficient to trigger neuronal firing. In this situation, stimulation-evoked ripples lead to the potentiation of weak synapses within the feed-forward structure and, thus, to a consolidation of a more general sequence memory. However, spontaneous ripples that occur without stimulation, only consolidate a sparse backbone of the existing strong feed-forward structure. Based on this, we test a recently hypothesized scenario in which the excitability of dendrites is transiently enhanced after learning, and show that such a transient increase can strengthen, restructure and consolidate even weak hippocampal memories, which would be forgotten otherwise. Hence, a transient increase in dendritic excitability would indeed provide a mechanism for stabilizing memories.

Different real-world cognitive tasks evolve on different relevant timescales. Processing these tasks requires memory mechanisms able to match their specific time constants. In particular, the working memory utilizes mechanisms that span orders of magnitudes of timescales, from milliseconds to seconds or even minutes. This plentitude of timescales is an essential ingredient of working memory tasks like visual or language processing. This degree of flexibility is challenging in analog computing hardware because it requires the integration of several reconfigurable capacitors of different size. Emerging volatile memristive devices present a compact and appealing solution to reproduce reconfigurable temporal dynamics in a neuromorphic network. We present a demonstration of working memory using a silver-based memristive device whose key parameters, retention time and switching probability, can be electrically tuned and adapted to the task at hand. First, we demonstrate the principles of working memory in a small scale hardware to execute an associative memory task. Then, we use the experimental data in two larger scale simulations, the first featuring working memory in a biological environment, the second demonstrating associative symbolic working memory.

Calcium (Ca ²⁺ ) is a well-known second messenger in all cells, and is especially relevant for neuronal activity. Neuronal Ca ²⁺ is found in different forms, with a minority being freely soluble in the cell and more than 99% being bound to proteins. Free Ca ²⁺ has received much attention over the last few decades, but protein-bound Ca ²⁺ has been difficult to analyze. Here, we introduce correlative fluorescence and nanoscale secondary ion mass spectrometry imaging as a tool to describe bound Ca ²⁺ . As expected, bound Ca ²⁺ is ubiquitous. It does not correlate to free Ca ²⁺ dynamics at the whole-neuron level, but does correlate significantly to the intensity of markers for GABAergic pre-synapse and glutamatergic post-synapses. In contrast, a negative correlation to pre-synaptic activity was observed, with lower levels of bound Ca ²⁺ observed in the more active synapses. We conclude that bound Ca ²⁺ may regulate neuronal activity and should receive more attention in the future.

Different real-world cognitive tasks evolve on different relevant timescales. Processing these tasks requires memory mechanisms able to match their specific time constants. In particular, the working memory utilizes mechanisms that span orders of magnitudes of timescales, from milliseconds to seconds or even minutes. This plentitude of timescales is an essential ingredient of working memory tasks like visual or language processing. This degree of flexibility is challenging in analog computing hardware because it requires the integration of several reconfigurable capacitors of different size. Emerging volatile memristive devices present a compact and appealing solution to reproduce reconfigurable temporal dynamics in a neuromorphic network. We present a demonstration of working memory using a silver-based memristive device whose key parameters, retention time and switching probability, can be electrically tuned and adapted to the task at hand. First, we demonstrate the principles of working memory in a small scale hardware to execute an associative memory task. Then, we use the experimental data in two larger scale simulations, the first featuring working memory in a biological environment, the second demonstrating associative symbolic working memory.

Developing intelligent neuromorphic solutions remains a challenging endeavor. It requires a solid conceptual understanding of the hardware's fundamental building blocks. Beyond this, accessible and user-friendly prototyping is crucial to speed up the design pipeline. We developed an open source Loihi emulator based on the neural network simulator Brian that can easily be incorporated into existing simulation workflows. We demonstrate errorless Loihi emulation in software for a single neuron and for a recurrently connected spiking neural network. On-chip learning is also reviewed and implemented, with reasonable discrepancy due to stochastic rounding. This work provides a coherent presentation of Loihi's computational unit and introduces a new, easy-to-use Loihi prototyping package with the aim to help streamline conceptualization and deployment of new algorithms.

The necessity of having an electronic device working in relevant biological time scales with a small footprint boosted the research of a new class of emerging memories. Ag-based volatile resistive switching memories (RRAMs) feature a spontaneous change of device conductance with a similarity to biological mechanisms. They rely on the formation and self-disruption of a metallic conductive filament through an oxide layer, with a retention time ranging from a few milliseconds to several seconds, greatly tunable according to the maximum current which is flowing through the device. Here we prove a neuromorphic system based on volatile-RRAMs able to mimic the principles of biological decision-making behavior and tackle the Two-Alternative Forced Choice problem, where a subject is asked to make a choice between two possible alternatives not relying on a precise knowledge of the problem, rather on noisy perceptions.