Biology subjects

Matsumoto, J.

Publications and source records attributed to Matsumoto, J..

3 recordsLinked to original sources

bioRxiv2018

Monkey Features Location Identification Using Convolutional Neural Networks

Understanding animal behavior in its natural habitat is a challenging task. One of the primary step for analyzing animal behavior is feature detection. In this study, we propose the use of deep convolutional neural network (CNN) to locate monkey features from raw RGB images of monkey in its natural environment. We train the model to identify features such as the nose and shoulders of the monkey at about 0.01 model loss.

animal behavior and cognition↗

bioRxiv2018

mitoNEET Regulates Mitochondrial Iron Homeostasis Interacting with Transferrin Receptor

Iron is an essential trace element for regulation of redox and mitochondrial function, and then mitochondrial iron content is tightly regulated in mammals. We focused on a novel protein localized at the outer mitochondrial membrane. Immunoelectron microscopy revealed transferrin receptor (TfR) displayed an intimate relationship with the mitochondria, and mass spectrometry analysis also revealed mitoNEET interacted with TfR in vitro. Moreover, mitoNEET was endogenously coprecipitated with TfR in the heart, which indicates that mitoNEET also interacts with TfR in vivo. We generated mice with cardiac-specific deletion of mitoNEET (mitoNEET-knockout). Iron contents in isolated mitochondria were significantly increased in mitoNEET-knockout mice compared to control mice. Mitochondrial reactive oxygen species (ROS) were higher, and mitochondrial maximal capacity and reserve capacity were significantly decreased in mitoNEET-knockout mice, which was consistent with cardiac dysfunction evaluated by echocardiography. The complex formation of mitoNEET with TfR may regulate mitochondrial iron contents via an influx of iron. A disruption of mitoNEET could thus be involved in mitochondrial ROS production by iron overload in the heart.

cell biology↗

bioRxiv2018

Automated Adherent Cell Elimination by a High-Speed Laser Mediated by a Light-Responsive Polymer

Conventional cell handling and sorting methods require manual dissociation, which decreases cell quality and quantity and is not well suited for monitoring applications. To purify adherent cultured cells, in situ cell purification technologies that are high throughput and can be utilized in an on-demand manner are expected. Previous demonstrations using direct laser-mediated cell elimination revealed only limited success in terms of their usability and throughput. Here, we developed a Laser-Induced, Light-responsive-polymer-Activated, Cell Killing (LILACK) system that enables high-speed and on-demand adherent cell sectioning and purification. This system employs a visible laser beam, which does not kill cells directly, but induces local heat production through the trans-cis-trans photo-isomerization of azobenzene moieties in only the irradiated area of a light-responsive thin layer. Using this system in each passage for sectioning, human induced pluripotent stem cells (hiPSCs) were maintained their pluripotency and self-renewal during long-term culture. Furthermore, combined with deep machine-learning analysis on fluorescent and phase contrast images, a label-free and automatic cell processing system was developed by eliminating unwanted spontaneously differentiated cells in undifferentiated hiPSC culture conditions.

bioengineering↗