- New AI-based tool can automatically search and focus on individual molecules within living cells.
- It is 10 times faster than other techniques and can be implemented with high-magnification microscopy.
Analyzing cell signaling and other molecular phenomena in living cells has been made possible by single-molecule imaging techniques. They allow scientists to directly monitor the behaviors of biomolecules, detect their movements, locations, and complex formation with high sensitivity, deepening current medical and biological knowledge.
However, existing single-molecule imaging methods are not much efficient: they still require significant expertise to search for cells suitable for observation, focus on nanometer precision, and statistically investigate each molecule. Also, the quality and speed of data acquisition and analysis suffer due to lack of skills.
Now, researchers from Osaka University and RIKEN have come up with a tool that can automatically search and focus on individual molecules within living cells. Within a short period of time, it can examine hundreds of thousands of individual molecules in hundreds of cells, offering reliable information on the dynamics of target molecules.
Fully Automated In-Cell Single-Molecule Imaging System (AiSIS)
The tool, named AiSIS, relies on an AI-powered total internal reflection fluorescence microscope. It features neural networks trained to precisely focus on a sample, search for cells, and then track single fluorescently labeled molecules.
This technique is nearly 10 times faster than other single-molecule measurement methods. To automate traditional workflow, it can be implemented with high-magnification microscopy, which enhances the current scenarios of imaging and analysis in biomolecules, while requiring fewer efforts.
Reference: Nature Communications | doi:10.1038/s41467-018-05524-7 | Osaka University
Testing and Verification
The tool is tested on an Epidermal growth factor receptor (EGFR) – a transmembrane protein that is almost free to move along the plasma membrane in which it’s expressed according to a specific alteration. The results show that the tool can distinguish between altering and non-altering scenarios by tracing the paths of individual receptors in membranes.
Researchers used these outcomes to measure numerous dynamics and pharmacological parameters like the diffusing speed of molecules (from their initial position) and those affecting the drugs’ efficacy.
The results matched the data obtained in previous researches using conventional labor-intensive approaches. One of the major advantages of this technique is that the impact of inhibitors and ligands on target(s) can be measured at the scale of a single molecule. Overall, the automation offered by this technique allows researchers to accurately characterize a larger number of molecules at low cost.
Read: New Calcium-Depended MRI Sensor Provides More Detailed Brain Imaging
Researchers plan to use this tool to monitor individual molecules elsewhere in the cells (like organelles and nucleus) using different optical microscopes, including super-resolution microscopy, oblique illumination, and correlation. They hope that their tool will be soon clinically available for genome-wide screening and testing.
