Dynamic and social behaviors of human pluripotent stem cells

Human pluripotent stem cells (hPSCs) can self-renew or differentiate to diverse cell types, thus providing a platform for basic and clinical applications. However, pluripotent stem cell populations are heterogeneous and functional properties at the single cell level are poorly documented leading to inefficiencies in differentiation and concerns regarding reproducibility and safety. Here, we use non-invasive time-lapse imaging to continuously examine hPSC maintenance and differentiation and to predict cell viability and fate. We document dynamic behaviors and social interactions that prospectively distinguish hPSC survival, self-renewal, and differentiation. Results highlight the molecular role of E-cadherin not only for cell-cell contact but also for clonal propagation of hPSCs. Results indicate that use of continuous time-lapse imaging can distinguish cellular heterogeneity with respect to pluripotency as well as a subset of karyotypic abnormalities whose dynamic properties were monitored.


Supplementary
. Flow Diagram for Cell Moment Tracker (CMT) The overall purpose of CMT is to act as a user interface to analyze large time-lapse data sets of cells. CMT can save prior work or export for post processing, accurately track cells, and compute information about cell shape and migration patterns.
Currently, the program has been configured to study cluster formation and morphologies of cells to better understand social behaviors of cells. (Supplementary Movie 5)

Binary Mask
Using optimization methods developed by the Takeo Kanade image analysis group at Carnegie Mellon 1 , phase contrast images were analyzed to find where cell boundaries are located. From this, a binary mask, an image solely constructed out of black and white pixels, is created to differentiate regions where cells are located from the background of the image. This step is the most crucial step in CMT program because proper cell boundary detection is crucial for the collection of important data about cell morphology and migration patterns

Boundary Error Correction
When dead cells or artifacts are detected, the user can often edit these by adjusting parameters such as the minimum cell size in the image. If the cell boundaries are over-fitted (which may result in overly segmented cell detection), a Gaussian point spread function is applied across the entire image so that the boundaries are smoother and cells are properly detected. CMT also offers options to manually edit the boundaries of cells in case cell or cell-cluster morphology is highly valued in a given experiment.

Computing Tracks and Cell Feature Information
After cell boundaries are detected, a nearest temporal neighbor cell detection method is performed to generate cell migration tracks. Additionally, CMT computes cell shape information such as elliptic Fourier harmonics and spherical harmonics, which can be used to model cell behavior quantitatively. Error correction tools are provided for cell tracking errors or manually detecting mitotic events.

Saving and Post-processing Data
CMT provides a very user-friendly interface to save old data, so that edits can be made on a given data set over a period of several days. Additionally, after computing individual cell features, the program can export the data along with some important outputs about each cell's environment (such as the number of cell neighbors, eccentricity of the cells, and minimum number of neighbors). Using the mitotic detection feature, the program can construct a generation mapping based on mitotic events over the course of a time-lapse video. The ultimate goal will be to correlate the generation number and cell environment to the individual cell's behaviors. When the injection to withdrawal ratio is 1:7 or more and the probe is less than 100 um above the surface, hydrodynamic forces form a laminar flow, thus preventing the injected fluid from mixing with the media. Once the fluid injection is stopped it can be completely removed from the dish with minimal contamination. This allows the formation of a zone of influence next to the probe where the injected fluid is present (indicated by the green area). When the probe is placed above a target cell on the culture dish, it gets exposed to the injection fluid whereas the neighboring cells hardly sense the fluid. In this study, cells (imaged for 100 mins and analyzed by CMT) were exposed to accutase in a similar fashion and we were able to selectively detach tracing of the daughter cells shows spontaneous differentiation in the progeny of (ii) one daughter cell, and maintenance of pluripotency in the progeny of the (i) second daughter cell. (iii) Representative lineage plot of hESCs exposed to differentiation medium indicates lengthier cell cycle as compared to the pluripotent cells (i).
Supplementary movie 3a: hESCs seeded at low density (1,500 cells/cm 2 ). The probe is initially calibrated by first loading a blue food dye into the injection syringe. Then all the capillaries and tubes are primed with media. The probe is mounted on an inverted phase contrast microscope such that the two capillaries are centered with the microscope objective. A cell free medium loaded petri dish is placed on the microscope. The probe is lowered into the dish such that the capillary tips are approximately 30-100 m above the dish surface and fully submerged into media. The injection and withdrawal flow rates are adjusted until a clear area of influence is seen demarcated by the dye (see Supplementary Fig. S4b). The dye is removed from the syringe and accutase is loaded after the syringe is thoroughly washed with sterile di-H2O.
Then the test dish was replaced with the cultured cell's dish. The cells of interest are identified and positioned in the center of the field of view. The probe is lowered into the dish and positioned right above the cell of interest, such that the cell is right below the injection capillary. The injection and withdrawal flow are initiated and the cell is monitored for dislodgement. As soon as the cell is released from the dish surface the flow is stopped. Then the injection capillary used to withdraw the loose cell by reversing the flow in the corresponding syringe pump. As soon as the cell is removed the flow is stopped. The probe is removed form the pertri dish and placed into 5 L of 2x RT-qPCR mixture. 5 L is ejected from the injection capillary into the RT-qPCR mixture this would include the extracted cell. The RT-qPCR reaction can be run the corresponding primers and probes for gene expression analysis.