The results were delivered through the 4 objectives:
Objective 1. Design and fabrication of gradient cell engineering substrates.
Key results: The development of nanopatterned libraries applicable for high-content screening. A number of groups have developed different platforms for exploring/screening cell-topography interactions. The vast majority of them do not have the different patterned areas physically separated which can lead to cross-chemokine interaction and thus skewing the results. To that end, we developed a platform centered on 96-well plates, one of the most common cell culture consumables, where each well represents a different pattern or functional cue. It is even possible to incorporate variance in mechanical properties in the different wells. This was published in Biofabrication 2020 (DOI: 10.1088/1758-5090/ab5d3f).
Objective 2 – Model cell systems and FA proteins
Key results: We establish model cell lines to explore the dynamics at the single cell level. Here transfection of cell lines allowed us to express reporters for either single molecule expression used for super resolution microscopy (currently on BioRxiv, doi.org/10.1101/2020.07.23.191858) or the force-sensing receptors. The result of the latter is currently being prepared for publication but also led to a start-up (www.forcebiology.com).
Objective 3-Cell imaging and analysis
Key results: Super resolution microscopy was developed to quantify the direct interactions with nanostructures (currently on BioRxiv, doi.org/10.1101/2020.07.23.191858). The unique designs of our nanostructures allow us to align super resolution imaging to topography with a resolution of better than 20 nm. For the first time, this has enabled us to directly image the dynamic interaction of single molecules and the nanotopography but also visualise the formation of focal adhesions in relation to the topography.
Objective 4 – Linking morphometric parameters with FA structure and dynamics
Key results: We applied machine learning to correlate cells, materials and function (Sci Rep 2017 doi.org/10.1038/s41598-017-03780-z PLOS One 2020 doi.org/10.1371/journal.pone.0237972 and Nature Comms 2020 doi.org/10.1038/s41467-020-15114-1). The AI development led to cell phenotype identification greatly reducing the workflow in for example immune cell work. Central to our work in the identification of cells, is the free software package CellProfiler. However, the interface is technical and is a barrier for initiating cell segmentation. To that end, we developed an AI-driven plug in removing all parameters from the user and placed them in an AI optimised engine. Finally, we have demonstrated that quantitative gene expression levels are directly encoded in the cell morpheme (shape, texture, intensity, etc). This was achieved using AI optimized algorithms to link fluorescent images with gene expression level. As a consequence, it is now possible to identify expression levels at the single cell level and spatially.
Dissemination: Numerous invited presentation were given during the project. During the pandemic, the injection moulding technology was repurposed and 10.000 visors were manufactured and distributed to the local health authorities. One start-up (www.ForceBiology.com) based on technology developed in the project.
Unfortunately, COVID struck at the most critical part of the project as we were performing the concluding experiments. Lock-down and restrictions has a significant impact on the final stage of the project. However, remaining work is continued through other funded project to complete the vision the project.
