Behavioral Classification of Sequential Neural Activity Using Time Varying Recurrent Neural Networks.

Shifts in data distribution across time can strongly affect early classification of time-series data. When decoding behavior from neural activity, early detection of behavior may help in devising corrective neural stimulation before the onset of behavior. Recurrent Neural Networks (RNNs) are common models for sequence data. However, standard RNNs are not able to handle data with temporal distributional shifts to guarantee robust classification across time. To enable the network to utilize all temporal features of the neural input data, and to enhance the memory of an RNN, we propose a novel approach: RNNs with time-varying weights, here termed Time-Varying RNNs (TV-RNNs). These models are able to not only predict the class of the time-sequence correctly but also lead to accurate classification earlier in the sequence than standard RNNs. In this work, we focus on early sequential classification of brain-wide neural activity across time using TV-RNNs applied to a variety of neural data from mice and humans, as subjects perform motor tasks. Finally, we explore the contribution of different brain regions on behavior classification using SHapley Additive exPlanation (SHAP) value, and find that the somatosensory and premotor regions play a large role in behavioral classification.

Mesoscale cortex-wide neural dynamics predict self-initiated actions in mice several seconds prior to movement.

Volition - the sense of control or agency over one's voluntary actions - is widely recognized as the basis of both human subjective experience and natural behavior in nonhuman animals. Several human studies have found peaks in neural activity preceding voluntary actions, for example the readiness potential (RP), and some have shown upcoming actions could be decoded even before awareness. Others propose that random processes underlie and explain pre-movement neural activity. Here, we seek to address these issues by evaluating whether pre-movement neural activity in mice contains structure beyond that present in random neural activity. Implementing a self-initiated water-rewarded lever-pull paradigm in mice while recording widefield [Ca++] neural activity we find that cortical activity changes in variance seconds prior to movement and that upcoming lever pulls could be predicted between 3 and 5 s (or more in some cases) prior to movement. We found inhibition of motor cortex starting at approximately 5 s prior to lever pulls and activation of motor cortex starting at approximately 2 s prior to a random unrewarded left limb movement. We show that mice, like humans, are biased toward commencing self-initiated actions during specific phases of neural activity but that the pre-movement neural code changes over time in some mice and is widely distributed as behavior prediction improved when using all vs. single cortical areas. These findings support the presence of structured multi-second neural dynamics preceding self-initiated action beyond that expected from random processes. Our results also suggest that neural mechanisms underlying self-initiated action could be preserved between mice and humans.

The phosphorylation of serine55 in enamelin is essential for murine amelogenesis.

Amelogenesis imperfecta (AI) is an inherited developmental enamel defect affecting tooth masticatory function, esthetic appearance, and the well-being of patients. As one of the major enamel matrix proteins (EMPs), enamelin (ENAM) has three serines located in Ser-x-Glu (S-x-E) motifs, which are potential phosphorylation sites for the Golgi casein kinase FAM20C. Defects in FAM20C have similarly been associated with AI. In our previous study of EnamRgsc514 mice, the Glu57 in the S55-X56-E57 motif was mutated into Gly, which was expected to cause a phosphorylation failure of Ser55 because Ser55 cannot be recognized by FAM20C. The severe enamel defects in ENAMRgsc514 mice reminiscent of Enam-knockout mouse enamel suggested a potentially important role of Ser55 phosphorylation in ENAM function. However, the enamel defects and ENAM dysfunction may also be attributed to distinct physicochemical differences between Glu57 and Gly57. To clarify the significance of Ser55 phosphorylation to ENAM function, we generated two lines of Enam knock-in mice using CRISPR-Cas9 method to eliminate or mimic the phosphorylation state of Ser55 by substituting it with Ala55 or Asp55 (designated as S55A or S55D), respectively. The teeth of 6-day or 4-week-old mice were subjected to histology, micro-CT, SEM, TEM, immunohistochemistry, and mass spectrometry analyses to characterize the morphological, microstructural and proteomic changes in ameloblasts, enamel matrix and enamel rods. Our results showed that the enamel formation and EMP expression in S55D heterozygotes (Het) were less disturbed than those in S55A heterozygotes, while both homozygotes (Homo) had no mature enamel formation. Proteomic analysis revealed alterations of enamel matrix biosynthetic and mineralization processes in S55A Hets. Our present findings indicate that Asp55 substitution partially mimics the phosphorylation state of Ser55 in ENAM. Ser55 phosphorylation is essential for ENAM function during amelogenesis.

Detection and Quantitation of Label-Retaining Cells in Mouse Incisors using a 3D Reconstruction Approach after Tissue Clearing.

The murine incisor is an organ that grows continuously throughout the lifespan of the mouse. The epithelial and mesenchymal stem cells residing in the proximal tissues of incisors give rise to progeny that will differentiate into ameloblasts, odontoblasts, and pulp fibroblasts. These cells are crucial in supporting the sustained turnover of incisor tissues, making the murine incisor an excellent model for studying the homeostasis of adult stem cells. Stem cells are believed to contain long-living quiescent cells that can be labeled by nucleotide analogs such as 5-ethynyl-2´-deoxyuridine (EdU). The cells retain this label over time and are accordingly named label-retaining cells (LRCs). Approaches for visualizing LRCs in vivo provide a robust tool for monitoring stem cell homeostasis. In this study, we described a method for visualizing and analyzing LRCs. Our innovative approach features LRCs in mouse incisors after tissue clearing and whole-mount EdU staining followed by confocal microscopy and a 3-dimensional (3D) reconstruction with the imaging software. This method enables 3D imaging acquisition and non-biased quantitation compared to traditional LRCs analysis on sectioned slides.

Uncovering the effect of different brain regions on behavioral classification using recurrent neural networks.

As our ability to record neural activity from a larger number of brain areas increases, we need to develop tools to understand how this activity is related to ongoing behavior. Recurrent neural networks (RNNs) have been shown to perform successful classification for sequence data. However, they are black box models: once trained, it is difficult to uncover the mechanisms that they are using to classify. In this study, we analyze the effect of RNNs on classifying behavior using a simulated dataset and a widefield neural activity dataset as mice perform a self-initiated behavior. We show that RNNs are comparable to, or outperform, traditional classification methods such as Support Vector Machine (SVM), and can also lead to accurate prediction of behavior. Using dimensionality reduction, we visualize the activity of the RNNs to better understand the classification mechanisms of the RNNs. Finally, we are able to accurately pinpoint the effect of different regions on behavioral classification. This study highlights the utility and interpretability of RNNs while classifying behavior using neural activity from different regions.

Transparent and Hard Siloxane Based Hybrid UV-Curable Coating Materials with Amphiphobic Properties.

In this study, highly transparent siloxane-based hybrid UV-curable coating materials were prepared using (acryloxypropyl)methylsiloxane monomer (APMS), a thiol-ene monomer, with benzoin ethyl ether. For the thiol-ene monomer, either pentaerythritol tetrakis(3-mercaptopropionate) (PETTMP) or trimethylolpropane tris(3-mercaptopropionate) (TMPTMP) was used. The siloxane-based hybrid coating materials were highly transparent and hard (pencil hardness of 6-7H). The materials were also amphiphobic, with a water static contact angle of 92-100° and an oil contact angle of 46-63°, when prepared with a high siloxane-monomer-to-PETTMP/TMPTMP ratio. In general, both hybrid coating materials exhibited improved oleophobicity, high hardness, and surface smoothness with increasing siloxane content, although the TMPTMP-based hybrid coating films exhibited slightly higher oleophobicity (lower hydrophobicity) and a smoother surface than the PETTMP-based hybrid coating films.