Journal Article Research Support, Non-U.S. Gov't Ppublish 3 citations
PubMed: 38844049 DOI: 10.1016/j.jlr.2024.100573
Adipose-derived mesenchymal stem cells' adipogenesis chemistry analyzed by FTIR and Raman metrics.
J Lipid Res 65(7):100573
Authors: Karolina Augustyniak, Monika Lesniak, Hubert Latka, Maciej P Golan, Jacek Z Kubiak, Robert Zdanowski, Kamilla Malek
Affiliations
- Department of Chemical Physics, Faculty of Chemistry, Jagiellonian University in Krakow, Krakow, Poland; Doctoral School of Exact and Natural Sciences, Jagiellonian University in Krakow, Krakow, Poland.
- Laboratory of Molecular Oncology and Innovative Therapies, Military Institute of Medicine - National Research Institute, Warszawa, Poland.
- Department of Chemical Physics, Faculty of Chemistry, Jagiellonian University in Krakow, Krakow, Poland.
- Laboratory of Molecular Oncology and Innovative Therapies, Military Institute of Medicine - National Research Institute, Warszawa, Poland; Institute of Psychology, The Maria Grzegorzewska University, Warsaw, Poland.
- Laboratory of Molecular Oncology and Innovative Therapies, Military Institute of Medicine - National Research Institute, Warszawa, Poland; Dynamics and Mechanics of Epithelia Group, Institute of Genetics and Development of Rennes (IGDR), Faculty of Medicine, University of Rennes, CNRS, UMR 6290, Rennes, France.
- Laboratory of Molecular Oncology and Innovative Therapies, Military Institute of Medicine - National Research Institute, Warszawa, Poland.
- Department of Chemical Physics, Faculty of Chemistry, Jagiellonian University in Krakow, Krakow, Poland.
Abstract
The full understanding of molecular mechanisms of cell differentiation requires a holistic view. Here we combine label-free FTIR and Raman hyperspectral imaging with data mining to detect the molecular cell composition enabling noninvasive monitoring of cell differentiation and identifying biochemical heterogeneity. Mouse adipose-derived mesenchymal stem cells (AD-MSCs) undergoing adipogenesis were followed by Raman and FT-IR imaging, Oil Red, and immunofluorescence. A workflow of the data analysis (IRRSmetrics4stem) was designed to identify spectral predictors of adipogenesis and test machine-learning (ML) methods (hierarchical clustering, PCA, PLSR) for the control of the AD-MSCs differentiation degree. IRRSmetrics4stem provided insights into the chemism of adipogenesis. With single-cell tracking, we established IRRS metrics for lipids, proteins, and DNA variations during AD-MSCs differentiation. The over 90% predictive efficiency of the selected ML methods proved the high sensitivity of the IRRS metrics. Importantly, the IRRS metrics unequivocally recognize a switch from proliferation to differentiation. This study introduced a new bioassay identifying molecular markers indicating molecular transformations and delivering rapid and machine learning-based monitoring of adipogenesis that can be relevant to other differentiation processes. Thus, we introduce a novel, rapid, machine learning-based bioassay to identify molecular markers of adipogenesis. It can be relevant to identification of differentiation-related molecular processes in other cell types, and beyond the cell differentiation including progression of different cellular pathophysiologies reconstituted in vitro.
MeSH
Cells, Adipogenesis, Cell Differentiation, Microscopy, Electron, Scanning Transmission, Regenerative Medicine, Stem Cells, Mesenchymal Stem Cells, Machine Learning, Learning, Biological Assay