Development of the hyperspectral cellular imaging system to apply to regenerative medicine

Abstract

Regenerative medicine by the transplantation of differentiated cells or tissue stem cells has been clinically performed, particularly in the form of cell sheets. To ensure the safety and effectiveness of cell therapy, the efficient selection of desired cells with high quality is a critical issue, which requires the development of a new evaluation method to discriminate cells non-invasively with high throughput. There were many ways to characterize cells and their components, among which the optical spectral analysis has a powerful potential for this purpose. We developed a cellular hyperspectral imaging system, which captured both spatial and spectral information in a single pixel. Hyperspectral data are composed of continual spectral bands, whereas multispectral data are usually composed of about 5 to 10 discrete bands of large bandwidths. The hyperspectral imaging system which we developed was set up by a commonly-used inverted light microscope for cell culture experiments, and the time-lapse imaging system with automatic focus correction. Spectral line imaging device with EMCCD was employed for spectral imaging. The system finally enabled to acquire 5 dimensional (x, y, z, time, wavelength) data sets and cell-by-cell evaluation. In this study, we optimized the protocol for the creation of cellular spectral database under biological understanding. We enabled to confirm spectrum of autofluorescence of collagen, absorption of specific molecules in the cultural sample and increase of scattering signal due to cell components although detail spectral analyses have not been performed.