A Multimodal Imaging Framework to Advance Phenotyping of Living Label-free Breast Cancer Cells

Now our protocol, we present multimodal imaging using confocal Raman microspectroscopy and tomographic phase microscopy for rapid and unbiased morphochemical phenotyping of human breast cancer cells in their native physiological environment. Combining Raman microspectroscopy and tomography phase microscopy enables label-free imaging under physiological condition, providing biochemical and morphological information about cells and addressing the limitation of conventional perturbative methods, such as fluorescence labeling or chemical fixation. Our results demonstrate a high potential of implementing multimodal approach in providing morphochemical phenotyping of breast cancer cells. Moreover, its versatility, reproducibility, and non-invasiveness made a promising tool for a wide range of biomedical application, from fundamental cell biology studies to diagnostics.

We aim to create an atlas of optical biomarkers for non-invasive embryonic quality control, detecting cellular senescence, and drug screening in organoid models to advance regenerative medicine and disease research.

[Instructor] To begin, place the MDA-MB-231 human breast cancer cell sample into the onstage incubator. Add water for the water immersion objective. Switch on the automated water immersion feeder to supply water to the objective lens. Regularly check the water levels to avoid evaporation and ensure consistent imaging conditions throughout the measurements. Now, turn on the pump laser and adjust its output to attain a laser power of 75 milliwatts at the sample plane. Open the Micro-Manager microscope control software. In the configuration settings, select BF and click live on the left side of the software window. Choose the desired single cell, then click stop to end the bright-field visualization. Open the CCD camera control software. In the experiment setting, set the exposure time to 1.5 seconds. Click run on the experiment toolbar and turn on the laser beam. Next, adjust the Z position to maximize the signal counts in the fingerprint region from 600 to 1,800 inverse centimeters, then turn off the laser. Click stop on the experiment toolbar and exit the software. Start the data acquisition process using the MATLAB script, which automatically manages Raman channels in multidimensional measurements and saves the data. Set the field of view to 50 by 50 micrometer square with a resolution of 40 by 40 pixels, each pixel covering 1.25 by 1.25 micrometer square. Set the exposure time for each Raman spectrum to 1.5 seconds and run the MATLAB script. For the post-processing of Raman hyperspectral maps, launch the RamApp web-based tool. Import the hyperspectral data by clicking on new in the my analyses tab and upload the .mat file containing the spatial and spectral data variables, along with the X axis variable. To truncate the spectrum, in the crop and rotate menu, choose spectral crop. Define the inferior and superior borders of the Raman shift region as 600 and 1,800 inverse centimeters respectively. To correct for cosmic rays, go to denoising, click despike, select Z-score as method, and set threshold to eight. Check the options use first difference and correct spikes, then click run to apply the correction. To remove noise from the maps, click smooth map, choose the median filter, set the size of the square to three pixels, then click run. To smooth the spectrum of each pixel, choose the Savistky-Golay filter. Set filter window length to seven and polynomial order to two, then click run. To subtract the average background signal, open optical substrate removal and click identify substrate. Set k-means as the method with two clusters. then click on advanced options. Choose morphological cleaning to remove small blobs from the foreground, then click run. To remove the average background signal, click on remove substrate, then choose global average or mean of the central 95% of values method and click run. To correct for baseline fluorescence, open the baseline panel, click on correct baseline, select adaptive smoothness PLS, then set lambda to 5 million and click run. For the normalization of each Raman hyperspectral map, go to miscellaneous, select normalize, choose Frobenius method, and click run. To generate a false color image showing subcellular distribution, on images panel, click the open menu symbol under intensity image, then click edit image. On the editing panel, select single band and define the spectral range. Now, choose double color on the color map, then define the color, intensity threshold, and opacity before clicking confirm. For the download of a single point spectrum, choose a pixel of interest, then on the spectral legend panel, click download spectrum. Finally, to export foreground and substrate spectra, navigate to spectral legend, then click download spectrum for both foreground and substrate sections. Place a drop of distilled water onto the objective lens of the microscope. Position the sample on the microscope's translation stage, then adjust the sample position to align it with the objective lens. Turn on the microscope and launch the imaging software. Click the microscope icon on the toolbar. After initialization, click configuration, select live cell in the job panel and PBS as the medium. Next, access the calibration tab in the control panel of the imaging software. Set the axial positions of the objective and condenser lenses by selecting focus and surface respectively. Now, click scanning mode to manually adjust the lenses and ensure the illumination patterns are centered and nearly static. Choose normal mode, then adjust the translation stage to bring the cell into the field of view and focus on the sample. Adjust the translation stage to locate a region without cells. Choose calibrate to capture multiple 2D holograms at varying illumination angles. Adjust the translation stage to center the cell, then navigate to the acquisition tab and select 3D snapshot to capture the tomogram of the cell. To visualize the holographic tomograms, select the data on the data navigation panel, right click, and click open. On the data manager panel, click on RI tomogram. In the volume visualization panel, select RI and draw four rectangular color boxes within the RI canvas. Set the minimum and maximum values of the RI range for each color box, then associate an opacity and color with each box. Click save to store the defined RI ranges and color boxes. To attain quantitative descriptors of cell morphology, utilize the analysis interface. Select manual under the segmentation panel, then set 1.3450 as the RI threshold and click apply. Click save to save the calculated morphological indexes. Raman spectral mapping revealed the spatial distribution of cytoplasm, nucleus, phenylalanine, and lipids within individual MDA-MB-231 cells. Distinct Raman spectral peaks corresponding to cytoplasm, nucleus, phenylalanine, and lipids were confirmed in the fingerprint region between 600 and 1,800 inverse centimeters. Refractive index heat maps of single cells revealed the internal distribution of subcellular components in the axial XY plane, sagittal YZ plane, and coronal XZ planes. A 3D tomographic reconstruction identified regions of volumetric refractive index corresponding to internal substructures. Quantitative morphological measurements of a single MDA-MB-231 cell showed different descriptors of cell morphology, with a total volume of 4024.189 cubic micrometers and surface area of 2383.707 square micrometers, with a mean refractive index of 1.356.