Super-multiplex imaging with Raman and fluorescent labels proves 20 times faster than conventional models

Together with a collaborative research group that includes researchers from Columbia University, Tsinghua University, and the University of Hawaii, Professor Yasuyuki Ozeki and graduate student Jingwen Shou of the Graduate School of Engineering, the University of Tokyo, as well as Specially Appointed Assistant Professor Robert Oda, Associate Professor Mutsuo Nuriya, and Professor Masato Yasui of Keio University School of Medicine have developed a technology to analyze complex and diverse cells in detail by integrating stimulated Raman scattering (SRS) microspectroscopy, which detects intracellular biomolecules by SRS, and fluorescence microscopy, which detects fluorescence from fluorescent molecules.

The developed SRS/fluorescence-integrated imaging system can detect the molecular vibration frequency, fluorescence excitation frequency, and fluorescence wavelength at high speed simultaneously while acquiring 30 frames of Raman image/fluorescence image per second based on the molecular vibration in each frame. As the frequency, fluorescence excitation wavelength, and fluorescence detection wavelength can be set, the time required for supermultiplex imaging of a biological sample can be substantially reduced. This means that imaging of eight types of live cells labeled with Raman and fluorescence reporters can be performed within 30 seconds (more than 20 times faster than conventional methods). Using this technique, the researchers succeeded in examining in detail how organelles in living cells interact with each other by moving around in a complicated manner as well as the spatial distribution of organelles in a large number of cells. These findings were published in the online edition of iScience.

Raman imaging, which optically detects molecular vibrations, provides visualization of label-free biomolecules and biomolecules using Raman labels, such as alkynes, isotopes, and Raman probes, and allows supermultiplex imaging using the narrow-bandness of molecular vibration spectra. It has been recently shown that a diverse variety of biomolecules can be visualized by simultaneously performing Raman and fluorescence imaging. Among them, SRS imaging is advantageous for simultaneous Raman and fluorescence imaging because there is almost no effect of fluorescence emission.

However, conventional SRS and fluorescence imaging have low time resolution owing to factors such as the switching time of SRS and fluorescence, the switching time of the laser wavelength for acquiring signals of multiple molecular vibration frequencies, and the switching time of the optical filter for detecting fluorescence of multiple wavelengths. For instance, in ultra-multicolor imaging, it takes 10 minutes or more to obtain one image, and it is difficult to analyze intracellular interactions in time series as well as to measure a large number of cells. The research group developed the SRS and fluorescence-integrated imaging system for clarifying complicated interactions in the cell and for exploring cellular diversity.

In this newly developed system, molecular vibration frequency, fluorescence excitation, and fluorescence signal, which are detected by SRS for every frame, are controlled and SRS and fluorescence images are simultaneously acquired in 30 frames per second. In particular, two-color near-infrared picosecond pulses (pump light and Stokes light) for SRS imaging and four-color excitation light for fluorescence imaging are combined and fed into a microscope. The SRS signal is acquired by lock-in detection of the pump light wavelength from among the sample-transmitted light. At the same time, the fluorescence emitted from the sample is detected by a photodetector through confocal pinholes and wavelength filters.

The Stokes and transmission wavelengths of the fluorescent filter are controlled at high speed by changing the angle of the galvanometric scanner (GS1 and GS2). Finally, the excitation wavelength is controlled by turning the fluorescence excitation light on and off. The SRS and fluorescence switching and the time of laser wavelength switching is reduced by using this mechanism, thereby speeding up SRS and fluorescence imaging. With this SRS/fluorescence-integrated imaging system, a total of 8 colors (4 SRS colors and 4 fluorescent colors) can be imaged in just 0.14 seconds.

In this experiment, four types of fluorescently labeled polymers (PA, PEMA, PMMA, and PS) and four types of fluorophores were observed. Images with different contrast were obtained according to the molecular vibration frequency, fluorescence excitation, and fluorescence detection wavelength detected using SRS. In addition, an image of each component was obtained by linear separation processing, which enabled a colored display. Furthermore, supermultiplex imaging of biological samples was carried out.

First, mitochondria, lysosomes, fat droplets, and vesicles of living cells were stained using a Raman probe, and cell nuclei, cell membranes, tubulin, and actin were stained using a fluorescent probe, and 8 micrograph images were acquired in 30 seconds. The researchers also succeeded in performing 8-color imaging within 2 seconds through unlabeled SRS and fluorescence imaging of biomolecules, and in performing 7 color depth-resolved imaging in 16 second fluorescence imaging of the mouse brain tissue. In addition, isotope-labeled saturated fatty acid and unsaturated fatty acid molecules were administered to cells, and SRS imaging and color fluorescence imaging was performed. This allowed them to investigate the state of intracellular organelles when two types of fatty acids are taken up by cells and metabolized.

This system allows researchers to achieve detailed observation of organelle movement in cells and to investigate differences in organelle localization on a cell-by-cell basis. This was not possible with conventional SRS/fluorescence imaging systems. Thus, this SRS/fluorescence-integrated imaging system has greatly expanded the applicability of supermultiplex imaging.