Journal: Scientific Reports

Article Title: An automated real-time microfluidic platform to probe single NK cell heterogeneity and cytotoxicity on-chip

doi: 10.1038/s41598-021-96609-9

Figure Lengend Snippet: Experimental setup of high-throughput droplet-based cytotoxicity platform. ( A ) Experimental schematics showing cytotoxicity platform that combines (i) droplet generation and cell pairing using microfluidics, (ii) droplet immobilization for real-time microscopy, (iii) automated image analysis using custom-made MATLAB script to allow unbiased and high throughput detection of cytotoxic events. Stained NK cells and K562 cells were loaded into the chip using 200 µL pipette tips and encapsulated into droplets using a 3-inlet microfluidic chip. The viability dyes were included within the cell medium. The immobilized droplets were incubated in a stage top incubator set at 5% CO 2 and 37 °C. Image acquisition was performed at every hour interval for 10 h. ( B ) The three-inlet microfluidic device with flow-focusing junction to generate droplets. ( C ) A qualitative test of the observation chamber was performed by monitoring droplets movement in the chamber under the microscope for 10 h.

Article Snippet: Automated Image analysis was performed using custom-made in-build MATLAB script (Mathworks), DMALAB (available on request).

Techniques: High Throughput Screening Assay, Microscopy, Staining, Transferring, Incubation

Journal: Scientific Reports

Article Title: An automated real-time microfluidic platform to probe single NK cell heterogeneity and cytotoxicity on-chip

doi: 10.1038/s41598-021-96609-9

Figure Lengend Snippet: Image processing and analyzing steps in MATLAB script. ( A ) Image processing by stacking different channels, droplet tracing, and detection of cells over each channel. ( B ) Droplet movement over one hour. An adjustable amount of movement is allowed by the script, but if this allowed movement is more than the droplet radius, the risk exists of detecting a different droplet at the next time point. Droplet detection over time is done by comparing the coordinates of the center of the droplet between two consecutive time points and selecting the closest center within the allowed movement range. ( C ) Evaluation of the script for cells recognition over different channels, here two cells labelled with cell tracker blue are identified in DAPI channel). ( D ) Validation of the script by comparing differences in cell distribution, cell pairing, and dead cell identification at three different time points (t = 3 h, t = 4 h, and t = 10 h) in-between script generated data with manually counted data.

Article Snippet: Automated Image analysis was performed using custom-made in-build MATLAB script (Mathworks), DMALAB (available on request).

Techniques: Biomarker Discovery, Generated

Journal: Molecular Therapy. Methods & Clinical Development

Article Title: Vector Copy Distribution at a Single-Cell Level Enhances Analytical Characterization of Gene-Modified Cell Therapies

doi: 10.1016/j.omtm.2020.04.016

Figure Lengend Snippet: Single-Cell Vector Copy Number Assay on Two Cell Samples (A) Diagram of the scVCN workflow. The steps in the light blue box are performed within the closed Fluidigm C1 system. ddPCR analysis is performed on each single-cell preamplified material on the QX200 Bio-Rad system. (B and C) Single-cell VCN values from each duplex assay combination are shown for the low (B) or the high (C) pVCN samples, whereas the values from bulk gDNA analysis (gray) include all six combinations and are shown as reference. Combinations originating from RG1 are in light blue, whereas the combinations with RG2 are in dark blue. Each boxplot is represented with mean and standard deviation. (D and E) The mean of the six scVCN combinations is shown for each single cell in the low (D) and high (E) pVCN samples and represented with standard error and 95% confidence interval. Single cells are ordered by increasing mean values. (F and G) Proportion of single cells with predicted vector copy units determined by Bayesian analysis in the low (F) and high (G) pVCN samples. The percentage of single cells with five or more vector copies is grouped. pVCN from bulk population analysis is indicated at the bottom with the standard deviation, alongside the mean of the single-cell VCN predictions. See also Figure S3 .

Article Snippet: A custom-made thermal protocol was designed with the C1 Script Builder (Fluidigm) to perform cell capturing, staining, and processing.

Techniques: Plasmid Preparation, Standard Deviation