Revolutionary real-time microscopy reveals secrets of motor protein behavior

A team led by Professor Seo Dae-ha from the Department of Physics and Chemistry at DGIST (President Lee Kun-woo) has developed a modern real-time microscopy technology and successfully observed the behavior of “motor proteins” that may be the key to unraveling an capable strategy for transporting material through cells. The research team used a nanoparticle probe, high-resolution microscopy, and Fourier transform algorithm technologies to develop “Fourier transform-based dark-field plasmonic microscopy” (FT-pdf microscopy) with positional and angular accuracy comparable to electron microscopy, achieving the highest level of existing optical microscopy.

Cells transport materials efficiently through intracellular vesicles called endosomes. Materials are transported to their destination by motor proteins that move along a elaborate network of microtubules. Observing the movement and rotation of endosomes that occur during the transport process provides critical information for understanding how intracellular transport is efficiently regulated, which in turn helps to explain cellular function and disease.

To visualize this transport process, the research team developed FT-pdf microscopy, which allows analysis using the Fourier transform technique using nanoparticles that have “polar angle dependence”. Images of the scattering signals observed by rotating polarized lithe are continuously recorded over a long period of time and, in combination with existing single-molecule tracking technology, the movement and rotation of the molecules can be observed in real time.

Using a dark-field plasmonic microscope, the research team discovered temporal patterns (high-resolution time series characteristics) in the rotational movements of endosomes in cells, which they interpreted as similar to the reinforcement learning strategy of navigation robots or web search engines. The real-time endosome transport strategy can be analyzed and applied to disease cell models to elucidate and diagnose the cause of diseases.

As this study shows, ordinary cells that make up the human body appear to be equipped with robotic data learning technology that humans are actively developing.

Professor Seo Dae-ha, Department of Chemical Physics at DGIST

“This molecular strategy is key to precise material transport and is the next research topic. Our research results are expected to contribute to the understanding and diagnosis of diseases in the future by applying them to diseased cells,” added Prof. Seo.

The research was supported by the Ministry of Science and ICT and the Basic Research Program of the National Research Foundation of Korea and the Engineering Research Center (ERC), the DGIST Grand Challenge Research Innovation Project (D-GRIP) and the HRHR+ Program. The research results were published in the international journal Advanced Science (IF 15.1).