Characterization of time-course morphological features for efficient prediction of osteogenic potential in human mesenchymal stem cells.
Bottom Line: In this report, we aimed to scale-down our proposed method into a more practical, efficient modeling scheme that can be more broadly implemented by physicians on the frontiers of clinical cell therapy.We investigated which morphological features are critical during the osteogenic differentiation period to assure the performance of prediction models with reduced burden on image acquisition.To our knowledge, this is the first detailed characterization that describes both the critical observation period and the critical number of time-points needed for morphological features to adequately model osteogenic potential.
Affiliation: Department of Biotechnology, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Aichi, 464-8603, Japan.Show MeSH
Related in: MedlinePlus
Mentions: We also identified limitations of the morphological features reflecting cellular potential. As shown in Figure 4, the “shortening type” analysis revealed that including changes in morphological features from the first day (16–32 h) undermines the prediction performance. These observations suggest morphological changes in the first day of differentiation culture are likely more random compared to subsequent changes in morphology that more directly contribute to prediction of osteogenic potential. Since cells had attached sufficiently during the expansion culture period in the same plate wells prior to day 0 in our experiment, these extraordinary morphological changes are likely a response to changing the culture environment to differentiation medium. We therefore further examined the morphological responses to medium changes (Fig. 5). When the differentiation period started, medium was changed at 64 h (day 3) and at 216 h (day 10; Fig. 5a). Surprisingly, many morphological features showed irregular changes after the medium was refreshed, which we interpret as clear morphological responses to the medium change. When viewed collectively these morphological responses to medium change were limited to particular morphological features. The morphological parameter, “fiber lengths,” was one parameter sensitive to medium changes (Fig. 5b). Even though these changes were small from a statistical point of view, this parameter, which indicates cell shrinkage, was most sensitive to medium change across all cell lots. Although most of these features were insensitive to the second medium change at time-point 26, fiber length still reflected this change in the environment. We found a drift in model performance in the “shortening type” and “window-shift type” analyses, which was due to parameter disturbances reflecting this response to medium change. Based on our observations we suggest correcting for these types of morphological parameters, which may be a source of noise when used in predictions of cellular ostogenic potential. This finding also strongly indicates the possibility that cellular morphology changes can also be used to detect changes in cellular microenvironment as a way to quality check animal cells during high-throughput assays.
Affiliation: Department of Biotechnology, Graduate School of Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Aichi, 464-8603, Japan.