Parsing the early cytoskeletal and nuclear organizational cues that demarcate stem cell lineages

• Published in: Cell cycle (Georgetown, Tex.) Vol. 9; no. 11; pp. 2108 - 2117
• Main Authors: Liu, Er, Gordonov, Simon, Treiser, Matthew D., Moghe, Prabhas V.
• Format: Journal Article
• Language: English
• Published: United States Taylor & Francis 01.06.2010
• Subjects: Actin Cytoskeleton - metabolism; Binding; Biology; Bioscience; Calcium; Cancer; Cell; Cell Differentiation; Cell Lineage; Cluster Analysis; Cycle; Gene Expression Profiling; Humans; Landes; Mesenchymal Stromal Cells - cytology; Mesenchymal Stromal Cells - metabolism; Nuclear Proteins - genetics; Nuclear Proteins - metabolism; Organogenesis; Principal Component Analysis; Proteins; rho-Associated Kinases - metabolism
• ISSN: 1538-4101; 1551-4005
• DOI: 10.4161/cc.9.11.11864

Our recent report suggests that subtle changes in early cytoskeletal protein-level organization correlate with long-term stem cell lineage commitment1. In this extra-view, we dissect changes in the expression of both cytoskeletal and nuclear-regulating genes that may precede and, possibly, govern the formative lineage-specific organizational cues. Human mesenchymal stem cells cultured on glass under basal, osteogenic, and adipogenic induction media were analyzed for gene expression profiles within the first 24 hours. Several key actin organization regulating genes and nuclear and cell cycle regulatory genes were found to be upregulated in osteogenic media compared to adipogenic and basal conditions. Given the role of both cytoskeletal and nuclear genes, we examined the possibility of classifying stem cell subpopulations using high content imaging approaches based on the organization of both actin, as previously proposed, as well as nuclear organization and distribution of a nuclear organizational protein, the nuclear mitotic apparatus (NuMA). A pool of combined cytoskeletal and nuclear descriptors were merged into a composite feature space via dimensionality reduction, data fusion, and classification methodologies. This composite approach enabled feature-based identification of specific lineage committed as well as non-differentiating cell populations. Using the improved classification of this high-content imaging-based profiling tool, we demonstrate that MSCs induced to differentiate to either osteogenic or adipogenic lineages are discernable within the first 24 hours from each other and from non-differentiating cells.