- Homepage
- Programme
- Registration
- Practical Guide
- About Neuronus
Flavio Donato
Biozentrum, University of Basel, Switzerland
Abstract: The entorhinal-hippocampal network contributes to the formation of episodic memories by creating an internal representation of the environment where experience unfolds. Such internal representation, or cognitive map, is instantiated in the activity of several functionally-specific cell types whose activity is modulated by space. Among these cell types, we distinguish neurons that are active at one or more specific locations in the environment (place and grid cells), or next to borders (border cells), or when the animal faces specific directions (head-direction cells). In rodents, while the firing properties of head-direction and border cells are adult-like at the onset of spatial exploration, spatial tuning in grid and place cells emerges and is refined progressively during the first months of life. This maturation process might depend on the establishment of specific connectivity motifs between the entorhinal cortex and the hippocampus. In fact, we previously showed that the functional maturation of such cell types is accompanied by the structural maturation of the entorhinal-hippocampal circuit, which is driven by an activity-dependent instructive signal that instructs the stepwise maturation of excitatory and inhibitory neurons at each stage of the network. Here, we will discuss recent studies whose aim is to understand how the emergence of spatial tuning in the developing entorhinal-hippocampal network shapes learning and memory processes during early postnatal life. Furthermore, we propose that studying the functional ontogenesis of the brain’s representation of space offers a unique opportunity to understand the contribution of individual cell types to hippocampal computations, and to dissect the contribution of such computations to learning and memory processes at multiple stages of an animal’s life
Bartosz Zglinicki1, Urszula Włodkowska2, Edyta Balcerek2, Rafał Czajkowski2, Michał Ślęzak1
1Biology of Astrocytes Group, Life Sciences & Biotechnology Center, Lukasiewicz Research Network – PORT Polish Center for Technology Development, Wroclaw, Poland
2Laboratory of Spatial Memory, Nencki Institute of Experimental Biology, Warsaw,Poland
Abstract: Behavioral experiments are a desirable tool in neuroscience but they can suffer from various factors, which influence reliability and reproducibility. Automatization of the process can overcome limitations such as experimentator input or data collection, and can help to reduce bias. With an emergence of machine learning and deep neural networks in recent years, it is possible to standardize behavioral studies and produce high quality data sets.
In this study we present automated t-maze system, in which mice learn to acquire a reward, based on two separate visual cues. System uses Bonsai software and Arduino with mechanical components, to record and to train animals with predefined protocol. After reaching appropriate threshold of correct choices, mice were subsequently tested in the same protocol with only one set of visual cues. DeepLabCut software was used for pose estimation and further analysis of mice trajectories. Furthermore, successful pose estimation in T-maze jumpstarted analysis of animal behavior in the open field and adaptation to known concept of social boxes.
Results from T-maze showed significant differences between pre-trained and trained animals, both in performance-oriented and trajectory-oriented manner. Automatization and deep learning allowed for a degree of analysis that was hard to achieve before and highlighted small details that would pass unnoticed otherwise. These tools are becoming mandatory for our future studies for efficient parametrization of social behavior.
Magdalena Sabat1,2,3, Quentin Gaucher1,2, Sam Norman-Haignere4,5,6,7,8, Yves Boubenec1,2
1Département d’Études Cognitives, École Normale Supérieure, PSL University, CNRS, Paris, France 2Laboratoire des systèmes perceptifs, Département d’études cognitives, École normale supérieure,
PSL University, Paris, France
3Laboratoire de Neurosciences Cognitives et Computationelles, INSERM, Ecole Normale Supérieure,
Université PSL, Paris, France
4Department of Neuroscience, University of Rochester Medical Center, Rochester, NY, USA 5Department of Biostatistics and Computational Biology, University of Rochester Medical Center, Rochester, NY, USA
6Department of Biomedical Engineering, University of Rochester, Rochester, NY, USA
7Department of Brain and Cognitive Sciences, University of Rochester, Rochester, NY, USA 8Zuckerman Mind, Brain, Behavior Institute, Columbia University, New York, NY, USA
Abstract: Retrieving meaning from complex auditory stimuli requires flexible binding in time of auditory features at different timescales. Recent reports suggest that auditory perception relies on a hierarchy of auditory timescales and progressive processing of information through a hierarchy of cortical areas. Although temporal dynamics of auditory perception have been thoroughly studied, little is known about the plasticity of temporal integration windows (or timescales) in the auditory cortex. To address this question we study the neural responses of the auditory cortex of the ferret to complex continuous auditory stimuli across a range of behavioural states. First, we assess the similarity of the gradient of sensory timescales in the auditory cortex of the ferret to that of the human. Subsequently, we address the question of plasticity by comparing the structure of this gradient across different behavioural states. Finally, we explore other sources of variability such as the internal state of the animal by relating the structure of the gradient to the physiological state of the animal. This work provides insights into the cortical mechanisms underlying flexible auditory performance at multiple scales.
Maciej Kania1, Basile Confavreux2, Tim Vogels2
1Vrije Universiteit Amsterdam, Amsterdam, Netherlands
2Institute of Science and Technology Austria, Klosterneuburg, Austria
Abstract: Information in the brain is usually thought to be transmitted mainly via excitatory neurons, while local inhibitory circuits are considered to merely stabilize excitation. As such, in most network models, more attention is given to excitatory population dynamics, relegating inhibition to a support role. However, numerous studies highlight the functional importance of inhibitory neurons beyond stabilization, such as regulation of information transfer or memory processes.
Here, we explore the role of inhibitory-to-inhibitory plasticity as a mechanism to produce network models with plausible inhibitory dynamics. To study the interaction between the classical role of inhibition - stabilizing excitation - and other possible inhibition-specific roles, we analyze the behavior of a recurrent spiking network simulated with a plasticity rule similar to Vogels et al., 2011 on the inhibitory-to-inhibitory and inhibitory-to-excitatory connections.
We find that, depending on the firing rate of the inhibitory population, the network activity can be silenced, highly synchronized, or bi-stable. We propose conditions on network and plasticity parameters to enable inhibition to have its own rich dynamics while still stabilizing excitation. Overall, our study shifts the focus from excitatory to inhibitory dynamics in plastic network models to better account for the profuse inhibitory dynamics observed experimentally.
Marek Brodzki1, Otmane Bouchatta1, Jaquette Liljencrantz2, Eleni Frangos2, Dimah Sade3, Gabriela Carballo1, Houria Manouze1, Carsten Bönneman3, Alexander Chesler2,3, Håkan Olausson1, Saad Nagi1, Marcin Szczot1
1Center for Social and Affective Neuroscience, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden
2National Center for Complementary and Integrative Health, National Institutes of Health, Bethesda, MD, USA
3National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA
Abstract: Hair follicle innervation forms an elaborate sensory apparatus that provides information on even the most delicate hair movements. As a measure of protecting the pelage, hair pulling evokes sharp pain sensation. We decided to investigate the neural encoding and molecular mechanisms underpinning this unique painful modality. We first examined mechanotransduction mechanism and found that patients with PIEZO2 deficiency syndrome do not feel hair pull pain, indicating for the first time the role of PIEZO2 in acute pain. Subsequently, we established that a PIEZO2-positive class of sensory neurons present in primates is homologous to murine hair pull nociceptors. Therefore, to study neuronal coding, we used in vivo imaging in mice to show that a subset of hair-pull responding neurons shows specificity and selective tuning. Consistently, we observed diminished responses to low pull forces in PIEZO2-KO neurons which was in line with measured hair nociceptors activation thresholds. To test functional conservation of hair pull nociceptors we performed human microneurography and demonstrated that hair pull pain is indeed coded by a dedicated novel group of Aβ fibers. In conclusion, we show that hair pull pain is a PIEZO2-dependent sensation conveyed by an evolutionarily conserved labelled line of fast conducting low-threshold nociceptors.
