Spinning arrayed disk (SpAD) for large-scale image-based live cell assays

Abstract

Being intrinsically linked to cell function, changes in cell morphology can reflect how cells respond to various stimuli or environmental changes, are thus the crucial indicators of numerous biological processes and disease states. For instance, alterations in cell shape and structure can signal cancer progression and metastatic potential, differentiation processes in stem cells, and responses to drug treatments. Understanding these morphological changes provides valuable insights into cellular behavior and mechanisms, aiding the development of targeted therapies and diagnostic tools.

While image-based assays can illustrate cell morphology, they are limited by constraints such as laborious fluorescence labeling, imaging speed, resolution, and field-of-view. To address these challenges, I developed a spinning arrayed disk (SpAD) integrated with an ultrafast imaging system (mATOM) to enable ultrafast, large-scale, label-free, single-cell, high-resolution imaging of live adherent cells.

SpAD is a custom-designed circular disk equipped with over 96-chamber for large-scale monitoring of live adherent cells. Its chamber design facilitates easy fluid handling preventing fluid leakage during spinning. The platform (18.8cm²) can collect real-time data (at 10GBytes/sec) from all chambers by coordinating it with the imaging system (mATOM) without compromising resolution (~1μm). Utilizing an ultrafast quantitative phase imaging (QPI) system with a ~10 MHz laser-scan rate, it can achieve multiple label-free image contrasts (brightfield, QPI contrasts) of single cells at high spinning speed (~1300RPM). Over 100 optical and physical (optophysical) features were extracted from the single-cell images, encompassing overall shape, global and local textures of optical and mass density. This platform can screen more than 100,000 cells under nearly a hundred conditions within 12 minutes, facilitating large-scale live adherent cell imaging.

To validate SpAD platform, plasma membrane integrity and apoptosis assays on H1975 and H2170 cell lines under continuous spinning showed negligible impact on cell viability compared to static controls. The optophysical features and single-cell images obtained at different time points during prolonged monitoring on SpAD indicated that the spinning platform introduced minimal effect on cell morphology and maintained consistent imaging performance during continuous operation.

To demonstrate the scalability and efficacy of using label-free optophysical features, we performed two perturbation studies. In a drug screening assay involving over 40 conditions, different drugs at various concentrations on two lung cancer cell lines, drug concentration-dependent label-free morphological profile shifts were observed, highlighting the sensitivity of these morphological features to drug treatment. Additionally, a CRISPR-based cell-cell fusion assay involving ACE2 receptor-expressing human cells and SARS-CoV-2 spike protein-expressing cells was performed on SpAD. The morphological differences in cell images enabled the identification of subpopulations that distinguished cells with different gene knockouts. Furthermore, morphological profiling could predict the proportion of cell-fusion events in the chamber.

In summary, this thesis introduced a large-scale screening platform for single-adherent cell studies. This platform offered transformative potential for morphological profiling, providing novel insights into disease- and gene-related phenotypes and accelerating therapeutic discovery. Significantly, our findings demonstrated that label-free morphological features are sensitive and specific to gene perturbations, marking an important advancement in the field and underscoring the value of label-free technologies in biomedical research.