Discrimination of functional hepatocytes derived from mesenchymal stem cells using FTIR microspectroscopy

摘要

Functional hepatocytes differentiated in vitro from mesenchymal stem cells (MSCs) need to be fully characterized before they could be applied as a therapy to treat liver disease. Here, we employed Fourier Transform Infrared (FTIR) microspectroscopy to investigate the characteristics of hepatocyte-like cells derived from rat bone marrow mesenchymal stem cells (rBM-MSCs) by detecting changes in macromolecular composition occurring during the hepatogenesis process. Partial Least Squares Discriminant Analysis (PLS-DA) enabled us to discriminate undifferentiated rBM-MSCs, and early, mid-stage and late stage rBM-MSCs derived hepatocytes by their characteristic FTIR "spectroscopic signatures". The predominant spectroscopic changes responsible for this discrimination were changes in FTIR absorbance bands at: 3012 cm ~(-1) (cis CC stretch from unsaturated lipids), 2952 cm ~(-1) (ν _(as)CH _3 from lipids), 2854 cm ~(-1) (ν _sCH _2 from lipids) and 1722 cm ~(-1) (CO stretching from lipids), which were associated with triglyceride and unsaturated fatty acid accumulation in the hepatocyte-like cells occurring during differentiation. Based on these findings, rBM-MSCs derived hepatocytes are characterized by high lipid content which facilitates a means of identifying hepatocytes from their stem cells progenitors by using FTIR microspectroscopy. Other complex changes in spectral bands assigned to proteins and nucleic acids were observed during hepatocyte differentiation indicating that mRNA translation was taking place producing proteins related to the formation of the new hepatocyte-like phenotype, which was corroborated by immunohistochemistry. The results show FTIR microspectroscopy combined with bioinformatic modeling constitutes a powerful new phenotypic-based methodology for monitoring and characterization of the process of stem cell differentiation leading to the formation of hepatocytes, providing complementary information to existing methodologies such as immunohistochemistry and gene analysis, but having advantages of being reagent-free and non-destructive of the sample.