Dynamic self-organization of microwell-aggregated cellular mixtures

Abstract

Cells with different cohesive properties self-assemble in a spatiotemporal and context-dependent manner. Previous studies on cell self-organization mainly focused on the spontaneous structural development within a short period of time during which the cell numbers remained constant. However the effect of cell proliferation over time on the self-organization of cells is largely unexplored. Here, we studied the spatiotemporal dynamics of self-organization of a co-culture of MDA-MB-231 and MCF10A cells seeded in a well defined space (i.e. non-adherent microfabricated wells). When cell-growth was chemically inhibited, high cohesive MCF10A cells formed a core surrounded by low cohesive MDA-MB-231 cells on the periphery, consistent with the differential adhesion hypothesis (DAH). Interestingly, this aggregate morphology was completely inverted when the cells were free to grow. At an initial seeding ratio of 1 : 1 (MDA-MB-231 : MCF10A), the fast growing MCF10A cells segregated in the periphery while the slow growing MDA-MB-231 cells stayed in the core. Another morphology developed at an inequal seeding ratio (4 : 1), that is, the cell mixtures developed a side-by-side aggregate morphology. We conclude that the cell self-organization depends not only on the cell cohesive properties but also on the cell seeding ratio and proliferation. Furthermore, by taking advantage of the cell self-organization, we purified human embryonic stem cells-derived pancreatic progenitors (hESCs-PPs) from co-cultured feeder cells without using any additional tools or labels.

Small-scale demixing in confluent biological tissues

“…There is an emerging consensus [9][10][11][12][13] that adhesive molecules help to regulate cortical acto-myosin, which can strongly impact cell sorting. However, it remains controversial whether differential adhesion or differential cortical tension alone is sufficient to generate the level of cell sorting and compartmentalization observed in embryos and cell culture systems [14][15][16][17][18][19][20]. Several experiments have suggested that additional processes such as specialized cell-cell signaling [16] or cellular jamming [17] enhance or disrupt sorting in living tissues.…”

Computational 4D-OCM for label-free imaging of collective cell invasion and force-mediated deformations in collagen

“…The results of the previous section demonstrate several advantages provided by our computational 4D-OCM imaging methods. Micrometer-scale details of both spheroid and collagen structure were observed over a large mm 3 scale volumetric field-of-view using only native scattering contrast. Rapid volumetric image acquisition (on the order of 10-15 seconds per volume, as detailed in the Methods) enabled the measurement of temporal fluctuations in the backscattered optical signals from cells, and these fluctuations were successfully exploited to enable the segmentation of cells from the surrounding collagen.…”

“…All OCT image reconstruction and subsequent image processing was performed in MATLAB R2018a. the MATLAB function imresize3 to yield a final voxel size of (1.88 μm) 3 and (2.34 μm) 3 for images of monoculture spheroids and co-culture spheroids, respectively. This down-sampling was used in order to perform smoothing and reduce computational load.…”

All-in-one 3D printed microscopy chamber for multidimensional imaging, the UniverSlide

“…Another category of samples for which observation chambers need to be designed and optimized are 3D multicellular aggregates. Proposed as in vitro tumour models 30 years ago 21 , they are now the subject of renewed interest because of the advances in optical sectioning microscopy 22 , the development of microfluidics strategies to produce them in a controlled fashion 23 24 25 , and the necessity for pharmaceutical and cosmetic companies to reduce animal testing for toxicity assays 26 . Initially restricted to the field of cancer biology 21 , the 3D multicellular aggregates have now numerous variants, such as embryoid bodies, organoids, neurospheres 27 28 29 30 31 .…”