A single-cell Raman-based platform to identify developmental stages of human pluripotent stem cell-derived neurons

Area of Science:

• Biomedical Engineering
• Stem Cell Biology
• Spectroscopy

Background:

• Stem cells offer therapeutic potential for neural diseases but lack standardized noninvasive characterization methods.
• Clinical applications are limited by the absence of comprehensive and noninvasive techniques to analyze cell molecular profiles.

Purpose of the Study:

• To develop and validate a label-free, noninvasive single-cell Raman microspectroscopy (SCRM) platform for identifying neural cell lineages.
• To establish a method for distinguishing human induced pluripotent stem cells (hiPSCs) from their neural derivatives using intrinsic biochemical profiles.

Main Methods:

• Utilized single-cell Raman microspectroscopy (SCRM) to analyze biochemical profiles of 8,774 individual cells.
• Identified a Raman biomarker (glycogen) for differentiating hiPSCs from neural cells.
• Employed a machine learning classification model with t-distributed stochastic neighbor embedding (t-SNE)-enhanced ensemble stacking for cell categorization.

Main Results:

• SCRM successfully distinguished hiPSCs from iPSC-derived neural cells based on unique Raman spectra.
• A glycogen-based Raman biomarker was identified and validated using conventional assays.
• The machine learning model achieved 97.5% accuracy in categorizing hiPSCs across different developmental stages.

Conclusions:

• The SCRM platform provides a high-content, noninvasive, and label-free approach for monitoring stem cell development.
• The identified glycogen biomarker and SCRM platform are potentially extensible to other cell types, impacting neural stem cell therapy.