Biology subjects

Tatsuno, M.

Publications and source records attributed to Tatsuno, M..

6 recordsLinked to original sources

bioRxiv2025

Influence of STDP rule choice and network connectivity on polychronous groups and cell ensembles in spiking neural networks

Spike-timing dependent plasticity plays an important role in how biological neural networks modify themselves with experience. However, the relationship between STDP and memory is not fully understood. Previously, an important advancement in understanding the relationship between spike-timing dependent plasticity (STDP) and memory was made through cortical simulations. The proposed memory items, polychronous groups (PGs), combined the network anatomy with precise spike-timing relationships between the connected cells of the network. However, there are some challenges with this previous work. It is unclear how different STDP rules would impact the PG results and if the PG results are complementary with purely spike-pattern defined memory items called cell ensembles (CEs). Lastly, it is unclear how these results are affected by changes in network connectivity. We address these challenges by comparing the PGs and CEs detected in spiking neural network simulations of cortical and hippocampal CA1 networks with two different STDP rule implementations. We show that the PG and CE results differ greatly for the two different STDP rules and for the cortical and CA1 networks. Our results show an important disconnect between anatomically defined and spike-pattern defined memories in spiking neural network simulations illustrating that care must be taken when drawing conclusions on the relationship between STDP and memory in simulation studies.

neuroscience↗

bioRxiv2025

Behaviorally relevant cell ensembles in rat motor cortex are replayed during sleep and implicate hippocampal involvement in motor skill learning

Motor memory is essential for our daily activities. It involves complex neural processes during learning and sleep. However, unlike explicit memory where its neural activation and role in memory consolidation are well-studied, the properties of cell ensembles for motor memory are less understood. In this study, we re-examined rats behavior and neural activity in the primary motor cortex (M1) and hippocampus while the animals were trained daily on a single-pellet reaching task. Recordings included both the training and 3 hr rest epochs before and after training. Behaviorally, the animals were classified into two learning types: rapid and gradual learners. Unsupervised cell ensemble detection on M1 neurons revealed that about 60% of the ensembles were modulated during reaching behavior. Those reach-related ensembles were further categorized into four types, and their replay was detected during both slow-wave sleep (SWS) and REM sleep. In SWS, replay preferentially occurred during spindles, especially slow-oscillation coupled spindles (SO-spindles). In addition, about 30% of the reach-related cell ensembles were modulated during the hippocampal sharp-wave ripples (SWRs). The direction of modulation and the temporal coupling between SWRs and SO-spindles depended on the training phase and the animals learning types. Our results demonstrate the replay of rats skilled-reaching memory during SWS and REM sleep and the possible involvement of the hippocampus through the modulation of M1 activations during SWRs. This study will advance our understanding of how neural activity patterns evolve during skilled-reaching learning and sleep, and help develop medical applications that leverage sleeps memory functions. Significance StatementMost evidence for memory replay comes from hippocampal studies in which neural activity during rest is compared to a template constructed from activity recorded during a behavioral task. Here, we used an unsupervised ensemble detection on recordings from rat primary motor cortex (M1) that included both rest and a skilled reaching task. We discovered a variety of ensembles with different sizes and time scales. Many were related to reaching behavior and were replayed during sleep. Some M1 ensembles were also modulated by the hippocampal activity, suggesting their involvement in motor skill learning. This study will advance our understanding of motor memory and sleep and help develop medical applications that leverage sleeps memory functions.

neuroscience↗

bioRxiv2025

Using Information Geometry to Characterize Higher-Order Interactions in EEG

In neuroscience, methods from information geometry (IG) have been successfully applied in the modelling of binary vectors from spike train data, using the orthogonal decomposition of the Kullback-Leibler divergence and mutual information to isolate different orders of interaction between neurons. While spike train data is well-approximated with a binary model, here we apply these IG methods to data from electroencephalography (EEG), a continuous signal requiring appropriate discretization strategies. We developed and compared three different binarization methods and used them to identify third-order interactions in an experiment involving imagined motor movements. The statistical significance of these interactions was assessed using phase-randomized surrogate data that eliminated higher-order dependencies while preserving the spectral characteristics of the original signals. We validated our approach by implementing known second- and third-order dependencies in a forward model and quantified information attenuation at different steps of the analysis. This revealed that the greatest loss in information occurred when going from the idealized binary case to enforcing these dependencies using oscillatory signals. When applied to the real EEG dataset, our analysis detected statistically significant third-order interactions during the task condition despite the relatively sparse data (45 trials per condition). This work demonstrates that IG methods can successfully extract genuine higher-order dependencies from continuous neural recordings when paired with appropriate binarization schemes.

neuroscience↗

bioRxiv2024

Long-Term Reactivation of Multiple Sub-Assemblies in the Hippocampus and Prefrontal Cortex

In resting or sleep periods following a task, neurons in various parts of the brain become reactivated, with firing patterns often similar to that during the task. This reactivation plays an essential role in memory consolidation. However, detecting these reactivating episodes has been challenging because not all neurons recorded may be directly relevant to the memory being consolidated. Here, we propose a novel spike train clustering (STC) method for detecting groups of neurons (clusters) with partial synchronous firing. From a mechanistic standpoint, the correlated activity of an ensemble can arise from neuronal interactions within the recorded local ensemble, common external inputs, or both. To quantify these contributions, we propose to use information geometry (IG) by taking advantage of its capacity for orthogonal decomposition of neural interactions. We analyzed simultaneous single unit activity from rat medial prefrontal cortex (mPFC) and area CA1 of the hippocampus when animals explored novel objects and when they slept before and after the exploration. We demonstrate that multiple reactivations by different subsets of neurons (clusters) occurred. Those reactivations could extend over 11 hours, the entire recording duration of post-task rest after the exploration epoch. Long-lasting reactivation was not detected when all neurons were included as a single cluster in the analysis. We also showed that pairwise interactions in reactivating clusters tended to be more strongly modified than in non-reactivating ones. In addition, the pairwise interactions of the reactivating clusters in CA1 were strongly modulated by the task experience but not in the mPFC. These results indicate that hippocampal reactivation following novel experience is likely to be induced within the hippocampal circuits. In contrast, mPFC reactivation is likely to be driven by external inputs, possibly in part from the hippocampus.

neuroscience↗

bioRxiv2022

Regional variation in cholinergic terminal activity determines the non-uniform occurrence of cortical slow-wave activity during REM sleep

Sleep consists of two basic stages: non-rapid eye movement (NREM) and rapid eye movement (REM) sleep. NREM sleep is characterized by slow high-amplitude cortical EEG signals, while REM sleep is characterized desynchronized cortical rhythms. While, until recently, it has been widely believed that cortical activity during REM sleep is globally desynchronized, recent electrophysiological studies showed slow waves (SW) in some cortical areas during REM sleep. Electrophysiological techniques, however, have been unable to resolve the regional structure of these activities, due to relatively sparse sampling. We mapped functional gradients in cortical activity during REM sleep using mesoscale imaging in mice, and observed local SW patterns occurring mainly in somatomotor and auditory cortical regions, with minimum presence within the default mode network. The role of the cholinergic system in local desynchronization during REM sleep was also explored by calcium imaging of cholinergic terminal activity within the mouse cortex. Terminal activity was weaker in regions exhibiting SW activity more frequently during REM sleep. We also analyzed Allen Mouse Brain Connectivity dataset and found that these regions have weaker cholinergic projections from the basal forebrain.

neuroscience↗

bioRxiv2017

Unsupervised detection of cell-assembly sequences with edit similarity score

Cell assembly is a hypothetical functional unit of information processing in the brain. While technologies for recording large-scale neural activity have been advanced, mathematical methods to analyze sequential activity patterns of cell-assembly are severely limited. Here, we propose a method to extract cell-assembly sequences repeated at multiple time scales and various precisions from irregular neural population activity. The key technology is to combine \"edit similarity\" in computer science with machine-learning clustering algorithms, where the former defines a \"distance\" between two strings as the minimal number of operations required to transform one string to the other. Our method requires no external references for pattern detection, and is tolerant of spike timing jitters and length irregularity in assembly sequences. These virtues enabled simultaneous automatic detections of hippocampal place-cell sequences during locomotion and their time-compressed replays during resting states. Furthermore, our method revealed previously undetected cell-assembly structure in the rat prefrontal cortex during goal-directed behavior. Thus, our method expands the horizon of cell-assembly analysis.

neuroscience↗