Contenidos

Fluctuation-based super-resolution (FBSR) imaging is a relatively novel approach for generating super-resolved microscopy images. The FBSR methods take advantage of pixel to pixel fluctuations present in camera-based microscopy images (Gustafsson, Culley et al. 2016, Zhao, W., Liu et al. 2018). Multiple frames of the same position are imaged and processed with FBSR algorithm to generate super-resolution images. We took an approach where we implemented FBSR imaging as a part of the traction force microscopy pipeline to gain more spatial resolution to our displacement and traction force maps (III, Fig 1A).

First, we decided to streamline the protocol for making SR capable PAA gels. The protocol for embedding the 40 nm fluorescent beads only on the top layer was has been described in detail (Colin-York, H., Shrestha et al. 2016, Colin-York, H., Eggeling et al. 2017). We took inspiration from this previous work and simplified the process by removing some steps we felt redundant. Making the gels with a new streamlined protocol does not take longer than the classical TFM gel preparation.

Furthermore, our protocol is suitable for glass-bottomed culture dishes (III, Fig. 1B, S1A).

Accuracy and sensitivity of TFM depend on trackable fluorescent beads (Style, Boltyanskiy et al. 2014). Moreover, the spatial resolution of TFM can be improved by enhancing bead detection. We decided to illustrate the enhanced bead recognition capability of FBSR by imaging a sample from the same position using spinning confocal and widefield microscopy (III, Fig 1C, 1D). Subsequently, the images were processed with two different FBSR algorithms, liveSRRF and SACD. Both algorithms were capable of significantly improving the bead images (III, Fig 1C, 1D). However, from the full field of view images it is apparent that SACD is more sensitive for having exactly right focus plane. Surprisingly, widefield microscopy performed especially well and could easily be implemented as part of TFM pipeline if there is no need for detecting planar structures like FAs. Widefield microscopy yielded more homogenous full-field images of the gels (III, Fig 1D). To associate how much the implementation of FBSR improved the bead density we compared the densities obtained in the previous publication of our group (III, Fig 1F). Classical TFM in combination with spinning disk confocal microscopy yielded bead densities close to 0.5 beads per square micron. The number coincides with other published work (Colin-York, H., Shrestha et al. 2016). With FBSR we were able to detect 1.2 beads per square micron, a considerable improvement to the classical method, but under the 2.2 beads per square micron reported using STED microscope (Colin-York, H., Shrestha et al. 2016). However, utilisation of FBSR circumvents some of the drawbacks which STED microscopy suffers from. FBSR have reduced phototoxicity and is cable of imaging large field of view, which are not features that STED microscopy is known (Schermelleh, Ferrand et al. 2019).

Super-resolution microscopy is known for artefacts caused by heavy in silico processing of the images (Wegel, Göhler et al. 2016). We wanted to make sure we are not detecting artefacts, so we decided to implement NanoJSQUIRREL a software method for detecting SR artefacts (Culley, Albrecht et al. 2018), as part of our experimental pipeline (III, Fig 1E, S1B, S1C). Furthermore, apart from FBSR processing and quality control steps, FBSR TFM pipeline is similar to the classical TFM experimental pipeline (III, Fig 2A). To compare the performance of FBSR TFM

and classical TFM we performed experiments with both methods from the same position and observed remarkable improvement in spatial resolution when using FBSR TFM (III, Fig 2b-2F). Both FBSR algorithms performed better compared to the classical TFM method. The improvement was probably caused by the ability to track more beads over more pixel units since FBSR methods artificially expand the images over a higher number of pixels. As part of FBSR TFM experimental pipeline, we use standard open-access methods for the analysis (Han, Oak et al. 2015, Tseng, Duchemin-Pelletier et al. 2012) and mathematical framework called Fourier Transformation Traction Cytometry. In principle, we should be able to improve the spatial resolution of our method even further if more mathematically heavy algorithms would be implemented (Soiné, Brand et al. 2015, Schwarz, Balaban et al.

2002).

Finally, we test FBSR in cell biological experiments to provide further insight into the versatility of the method (III, Fig 4A-4E, S4B). FBSR can be implemented to standard experiments where TFM is used to study the function of protein or effect of drug treatment with high accuracy. The reduced phototoxicity also makes FBSR TFM suitable for long-term live measurements of cellular forces (Culley, Tosheva et al. 2018). These experiments highlight the potential utility of FBSR TFM for the field of basic cell biological research and mechanobiology.

In the context of FA in hPSCs, FBSR TFM could be used for detecting the force changes during cell differentiation in live-imaging setup since the mechanobiological properties are known to influence the stem cell faith (Engler, Sen et al. 2006). It would also be intriguing to study if the structural differences between the FA edges and centres would translate to differences in forces exerted by individual FAs in hPSCs. However, to achieve this, the protocol would need additional optimisation to improve the spatial resolution further. The improvement could be attained by utilising two different coloured beads (Plotnikov, Sabass et al. 2014) to increase the density of detectable units and more mathematically heavy analysis pipeline (Han, Oak et al. 2015).

7 Conclusions

The purpose of the work in this thesis was to increase the knowledge of basic cell biological features of hPSCs. We think that there are significant gaps in the research when it comes to the adhesion mediated signalling and FA biology. To our knowledge this is the first time that somebody characterises the structural features of FAs and actin cytoskeleton in hPSCs We hope that the data derived from carefully done experiments and in detail analysis provided by this book will prove out to be useful in the future and offer additional perspectives for understanding the complicated signalling cascades regulating the pluripotency of hPSCs. Two publications included in this thesis book showcase the peculiar FAs in human induced pluripotent stem cells. The last original publication provides a method for high accuracy force measurement at the cellular level.