Analysis of traction forces and intracellular mechanics in aggregates of endothelial and stromal cells

FINAL DEGREE PROJECT Degree in Materials Engineering and Nanotechnology ANALYSIS OF TRACTION FORCES AND INTRACELLULAR MECHANICS IN AGGREGATES OF ENDOTHELIAL AND STROMAL CELLS Author: Alice Savoia Director: Andrea Malandrino Codirector: José Muñoz Announcement: February 2022 i ii Acknowledgements I would like to express my sincere gratitude to my tutors Andrea Malandrino and Jose Muñoz, for their insightful comments and suggestion and their assistance at every stage of the research project. iii ABSTRACT Animal cells sense and transduce physical forces. In everyday physiological activities, tissues composed by groups of cells interacting together can stretch significantly, for instance when heart pumps, or lung inflate during breathing. Evidence also show that altered cell mechanosensing and mechanotransduction occurs in pathologies, such as cancer. These cell mechanics events can be studied in simplified 2D in vitro systems, where forces exerted by cells are measured with accuracy. One commonly used technique is 2D traction force microscopy (TFM), which makes use of high-resolution microscopy and fluorescently decorated elastic culture substrates to measure the substrate displacement caused by cell forces. In a previous study, we have built on analysis pipelines that convert displacement into tractions, to further obtain computationally the intracellular (cytoskeletal) structural organization that could give rise to such forces, starting from TFM data on single cells. For that, we have solved an elastic minimization problem assuming that intracellular stresses are propagated in cable-like elements and seeking a mechanical equivalence leading to the experimentally measured tractions. However, our previous work and other research in experimental literature have not focused on more realistic situation, such as intracellular organization when multiple cell types are interacting. Our experiments show that distinct cell types, endothelial and fibroblasts exert different level of forces on the substrate. Moreover, by varying the stiffness of the substrate, we measured an increase of the spreading area for both endothelial cells and fibroblasts, as stiffness increases, up to a plateau. Finally, as a proof of concept, we obtained computationally the intracellular organization for a case of traction field produced by iv several endothelial cells interacting and observe a possible mechanical communication between cells by transmission of intracellular forces. 1 SOMMARIO 1 INTRODUCTION ..................................................................................................................... 3 1.1 Cell mechanics ........................................................................................................ 3 1.2 Previous results on endothelial and stromal cell cluster mechanics. ...................... 9 2 MATERIALS AND METHODS ............................................................................................. 13 2.1 Materials ................................................................................................................ 13 2.2 Glass surface activation ......................................................................................... 14 2.3 PAG gels preparation ............................................................................................ 15 2.4 PAG stiffness measurement .................................................................................. 17 2.5 PAA functionalization ........................................................................................... 18 2.6 Cell culture ............................................................................................................ 19 2.7 Seeding Cells on PAG substrate ........................................................................... 20 2.8 Acquisition of cell and fluorescent bead images ................................................... 21 2.9 Estimating the displacement field and computing the traction field ..................... 22 2.10 Measuring median tractions values ....................................................................... 23 2.11 Measuring of the spreading area ........................................................................... 24 2.12 Computing Dipoles from the Traction Field ......................................................... 25 3 RESULTS ................................................................................................................................ 27 3.1 PAG stiffness measurement .................................................................................. 27 3.2 Acquisition of cell and fluorescent bead images ................................................... 31 3.3 Estimating the displacement field and computing the traction field ..................... 34 3.4 Measuring median tractions values ....................................................................... 38 3.5 Measuring of the spreading area ........................................................................... 39 3.6 Computing Dipoles from the Traction Field ......................................................... 40 2 4 CONCLUSIONS AND DISCUSSION .................................................................................... 43 5 BIBLIOGRAPHY .................................................................................................................... 45 Appendix a. Cells and Displacements .............................................................................................. 48 Appendix b . Cells and Tractions ...................................................................................................... 55 3 1 INTRODUCTION 1.1 Cell mechanics The cell represents the fundamental unit of living organisms. The shape, movements and organization of the cells and tissues are physically maintained through structural intracellular elements that allow them to adapt to the mechanical stresses to which they are subjected. Cells might be regarded as active materials which can detect the activation of biochemical signalling pathways, some of these pathways being sensitive to physical forces. Cells react to physical signals by reorganizing the cytoskeleton (CSK) and by further generating forces. As such, cells are subjected to mechanics, both as individual organism and as part of complex systems. The mechanical load is fundamental for the growth and regeneration of all the tissues like bones, ligaments, tendon, skeletal muscle, intervertebral disc and meniscus. Many organs systems are influenced by mechanical forces: respiratory, cardiovascular, nervous and integumentary system. For instance, during exhalation and inspiration, the bronchial mucosa and alveoli are subjected to a volume change and so to a compressive and tensile load, generating significant deformations. Also, the blood flowrelated forces influence the entire shape and function of the cardiovascular system. As further example, epithelial barriers, such as the skin, have interesting mechanical functions. Most skin wound healing experiments to study epithelial barriers have so far focused on the effect of biological and chemical stimuli but one of the most interesting aspects is the response to the application of mechanical loads. The skin consists of the epidermis and the underlying layer, the dermis, which contains cells and extracellular matrix. This represents the substrate for adhesion, growth and cell differentiation providing the necessary 10 cadherin provides a signalling system through cadherin-integrin pathway or actin cytoskeleton regulation, so, the impact of E-cadherin on tensional homeostasis is due to its role on cellular architecture, but also to its relevance as a signalling-transduction platform. Other studies [4] have developed finite element-based traction force microscopy (TFM) to accurately calculate and visualize the traction maps resulting from multiple clusters of cells. Biological considerations have been derived from this research: the cluster can behave as a single contractile unit and peripheral cells behave as anchorage sites; the force is transmitted between the cells as if the peripheral cells were pulling the internal cells outwards. It has been suggested that this occurs through cellular junctions and the cytoskeleton. Moreover, cases have been identified [4] where traction is limited to small regions within clusters that may have locally balanced traction (forming dipoles), leaving the rest of the clusters almost unrestrained and weakly adherent to the substrate. Regions without traction tend to minimize the surface being circular and is also likely that the cells inside are under compression due to the surface tension of peripheral cells. In any case, the traction map detected [4] has shown to be greatly dynamic: as well as the cell clusters coalesce, the map can swap modifying the disposition and the scale of net force also by an order of magnitude in a short time. Since cells are known to generate contractile forces, what is expected from cells in a 2D cluster is that they are under intercellular tension. Evidence to the contrary was found [4]. When cells are on soft substrates on which they do not spread strongly but adhere to the substrate, some of the cells in the cluster may be subject to compression in which neighbouring cells exert repulsive forces while adhering to the substrate. This could be due to the growth of neighbouring cells but their adhesion sites were steady. 11 To better interpret traction maps in cell mechanics studies, one often missing fundamental correlation with traction forces is the intracellular (cytoskeletal) tension. Intracellular tensions are challenging to measure. Even derived computationally, the knowledge of intracellular tension alongside traction forces would be important in cell mechanics. In an attempt of deriving the intracellular mechanics as a function of measured traction forces, we have started to build analysis pipelines that (i) convert displacement into tractions, and (ii) further obtain computationally the intracellular (cytoskeletal) structural organization that could give rise to such forces. This algorithm deduces a set of mechanically equilibrated contractile dipoles that best matches the measured traction forces. The resulting network of dipoles, which is obtained through an iterative filtering process, provides information on the plausible cytoskeletal structure compatible with the traction field. However, our previous work and other research in experimental literature have not focused on more realistic situation, such as intracellular organization when multiple cell types are interacting. Our tissues and organs are indeed multicellular systems with different cells. Endothelial and stromal cells are fundamental cellular building blocks of the vascular systems. Our study aims to observe more specifically the behaviour of small groups of these two cell types on substrates of different stiffness. Tractions exerted by clusters of endothelial cells and fibroblasts have thus been measured on PAGs and analysed to understand if there are and what are the differences present in the mechanical interaction between cells of the same type and the different type. Additional measures, such the spreading area of the cells have been performed to see if there is a correlation with the stiffness of the PAG, both for the endothelial cells and the fibroblasts. Finally, and as a proof-of-concept, the computation of intracellular systems of dipoles associated with such 12 tractions has been performed using the current version of an algorithm developed at UPC [13] in the laboratory of mathematical and computational modelling (LaCàn). 13 2 MATERIALS AND METHODS 2.1 Materials The following materials have been used for the elaboration of the polyacrylamide gels (PAGs) ▪ 3-aminopropyltrimethoxysilane (silane). ▪ Acetic acid. ▪ Acrylamide. ▪ Ammonium persulfate (APS) solution (Sigma). ▪ Collagen type 1 ▪ Confocal Microscope (Carl ZEISS LSM 800). ▪ Cover glass, Ø18 mm. ▪ Cytochalasin D. ▪ Fluorescent beads solution (Sigma), 0.5 μm, red fluorescent (580/605), 2% solids. ▪ Fume hood. ▪ Glass-bottomed Petri dish, Ø35mm. ▪ Milli-Q water ▪ N, N’-methylenebisacrylamide (bis-acrylamide). ▪ N,N,N’,N’-Tetramethylethylenediamine (TEMED, Sigma). ▪ Normal Human Lung Fibroblasts (NHLF, Lonza). ▪ Phosphate Buffered Saline (PSB) solution. ▪ Pure ethanol. ▪ Rain-X ▪ Red Fluorescent Protein (RFP)- expressing Human Umbilical Vein Endothelial Cells (HUVEC, from AngioProteomie). 14 2.2 Glass surface activation To allow the PAG to bind during the polymerization. The activation of the glass petri dish surface is performed with a link of silane that generate a monolayer of free chemical groups that covalently bind the PAG we are going to polymerise on top. The solution is prepared in the laboratory hood with pure ethanol (95%), acetic acid (4.7%) and silane (0.3%). The hood must be used since the silane is a toxic element. The coating is left to act on the petri dish, which is covered in order to prevent evaporation of the solution, for 10 minutes. After that, a pipette is used to remove the excess of solution and then the surface is rinsed with ethanol 80% . 15 2.3 PAG gels preparation PAG hydrogels have been used in this work as cell culture substrate due to its deformability and tuneable stiffness (table 1). Polyacrylamide gels result from the free radical polymerization of vinyl groups among acrylamide 40% and the crosslinker bisacrylamide 2%. In the mixture is also added a saline solution as buffer, the common phosphate-buffered saline (PBS) and fluorescent microbeads that will allow us to visualise the deformation of the gel. PAGs Soft Intermediate Stiffer Total volume [ul] 500 500 500 Bis-acrylamide 2% 11 40 162.5 Acrylamide 40% 93.25 94 150 Beads 8 8 8 APS 2.5 2.5 2.5 TEMED 0.25 0.25 0.25 PBS 385 355.25 176.75 Table 1. The reaction is initiated by tetramethylethylenediamine (TEMED) and catalysed by ammonium persulfate (APS). These two are added immediately before the placement of the prepolymer on the Petri dish. 16 Figure 1. Prepolymer drop on the Petri dish Then, a liquid drop of 22 uL of prepolymer is deposited on the bottom coverslip (Figure 1) and the top one is placed on to define the hydrogel surface topography while surface tensions set the shape of the polymerized hydrogel and hence its thickness. The top coverslip surface is treated with Rain-X, a hydrophobic synthetic product to ease coverslip removal after polymerization. The dish is flipped with the coverslip facing down during polymerization in order to allow the beads to aggregate in what will be the upper layer of the gel, improving surface displacement resolution (see 2.9). After 30 minutes PBS is added to prevent the gel from drying out and after others 30 minutes the top coverslip is removed very carefully with the aid of forceps (Figure 2). Figure 2. Removal of the coverslip 17 2.4 PAG stiffness measurement Gel stiffness was measured by indenting the gel with a tungsten sphere (diameter, 979.455 μm) of known mass (6mg). The indentation caused by the weight of the sphere was determined using confocal microscopy. From the measured indentation and sphere mass, we obtained Young’s modulus by applying Hertz theory, corrected for the finite thickness of the gel [14]. Both for the indentation depth and gel thickness the measures have been taken using Fiji [17]. 18 2.5 PAA functionalization Different strategies can be used to functionalize the gels: functional co-polymers or surface functionalization. We used a simple method which involves surface activation with ultraviolet light in the presence of the photoactivable heterobifunctional reagent sulfoccinimidyl-6-(4’-azido-2’-nitro phenylamino) hexanoate (sulfo-SANPAH). The photoactivatable region of sulfo-SANPAH binds polyacrylamide and the sulfosuccinimidyl group at the other end reacts with protein primary amine groups. After defrosting 20 ul of sulfo-SAMPAH, this was added to 480uL of PBS. 80uL are deposited on the gel after eliminating all the excess of PBS present with the help of a pipette. The gels were then placed under a UV lamp for 3 minutes. Then three rinses, two with milli-Q water (purified and deionized water with appropriate purification systems) and one with PBS, of a duration of 5 minutes each were performed. To adhere the cells to the gel, collagen coating was used (4.1 mg/ml). Collagen coating presents some problems due to collagen rapid polymerization after an increase in temperature or pH. For this reason, we prepared a collagen solution (2%) and acetic acid (98%) which is deposited on the gel and left for 24h to 4 C to coat the gel surface. 19 2.6 Cell culture Normal Human Lung Fibroblasts (NHLF) are transfected with LentiBrite GFP Control Lentivial Sensor to express a green fluorescent protein in their cell body. This procedure is not performed in this work, but the fibroblasts have been transfected and donated (laboratory of Kristina Haase, EMBL Barcelona) for our experiment. The medium used for cell culture of these NHLF is DMEM, supplemented with 2mM glucose and 10% Fetal Bovine Serum (FBS) . RFP Expressing Human Umbilical Vein Endothelial Cells (RFP-HUVECs)[18] have also been used. HUVEC were isolated from normal human umbilical vein and transfected with RFP-Lentiviral particles to express a red fluorescent protein in their cell body. These cells are commercially available. The medium used for RFP-HUVECs culture is the Endothelial Cell Growth Medium, a low-serum cell culture medium by PromoCell, (97.6%) with supplement (2.4%). 26 Traction forces are decomposed into Tx and Ty and then further decomposed in the new axes eij , of the D matrix, to deduce the dipoles. The number of dipoles created is a compromise between a minimisation of the equilibrium error and a reduction of the computational cost. A filtering process removes dipoles with a minimal contribution to the total mechanical error. Some criteria are imposed which aim at reducing the number of dipoles created at each increment, and removing those that would give unrealistic result (compression). Moreover, tolerances have been calibrated to minimise the equilibrium error while not generating and exceeding number of dipoles. 27 3 RESULTS 3.1 PAG stiffness measurement In this section we show the images obtained with the confocal microscopy and used to measure the PAG thickness and stiffness (Figure 4). Figure 4. Multifocal z-stack image of the indentation depth of sample _3. a)XY frontal view. b) XZ orthogonal view. c) YZ orthogonal view. We have used the orthogonal views to measure the gel thickness (table 2) and the depth of the indentation (table 3) caused by a tungsten sphere to calculate the real stiffness of all our PAG gels. a) b) c) 28 Table 2. PAGs thickness measures Table 3. PAGs indentation depth measures sample Intended stiffness Gel thickness [μm] 1st measure 2nd measure 3rd measure Mean value 1 Soft 118,55 123,54 116,05 119,3800 2 81,73 79,86 81,11 80,9000 3 104,82 106,07 108.56 105,4450 4 Intermediate 116,68 110,44 112,94 113,3533 5 84,85 79,86 76,12 80,2767 6 86,1 84,23 81,74 84,0233 7 Stiffer 99,7 97,96 101,7 99,7867 8 84,23 80,49 80,48 81,7333 sample Intended stiffness Indentation depth [μm] 1st measure 2nd measure 3rd measure Mean value 1 Soft 36,19 38,68 42,43 39,1000 2 21,21 26,20 23,70 23,7033 3 23,71 24,33 25,58 24,5400 4 Intermediate 14,97 14,35 14,35 14,5567 5 24,96 17,47 18.72 21,2150 6 21,84 19,34 18,72 19,9667 7 Stiffer 15,60 13,73 14,98 14,7700 8 9,980 8,74 7,49 8,7367 29 The Young’s modulus of PAG is determined (Table 4) according to the published method [14]. According to the different percentage of acrylamide and bis-acrylamide used (2.3), the Young’s modulus should vary from 5.37 kPa to 48kPa. sample Intended stiffness Gel thickness [μm] -mean valueIndentation depth [μm] -mean valueReal stiffness [kPa] 1 Soft 119,3800 39,1000 0,8400 2 80,9000 23,7033 1,3482 3 105,4450 24,5400 2,0553 4 Intermediate 113,3533 14,5567 7,0511 5 80,2767 21,2150 1,9300 6 84,0233 19,9667 2,2200 7 Stiffer 99,7867 14,7700 5,8318 8 81,7333 8,7367 13,721 Table 4. Real stiffness obtained for each gel The soft gels have an average measured real stiffness of 1,4145 kPa, the intermediate of 3,7337 kPa and the stiffer gels of 9,7764 kPa (Figure 5). 30 Figure 5. Average measured real stiffness The ratios of acrylamide and bis-acrylamide used in this experiment gives a gel that is not too stiff. This can be due to old reagents or polymerization issues. The real stiffness and thickness will be introduced in the MATLAB code for the computation of the tractions to obtain more reliable results. 0 2 4 6 8 10 12 14 16 1 Average measured real stiffness [kPa] SOFT INTERMEDIATE STIFFER 31 3.2 Acquisition of cell and fluorescent bead images In this section we show the results obtained from the images acquisition of the cells placed on the PAG. Two images have been extracted from each position: the "force-loaded image” and the "null-force image" after Cytochalasin D is applied on top of the gel (see section 2.8). For each image we have three channels (Figure6 and Figure7), corresponding to the three fluorophore used: one channel of the beads (far red fluorescence), one of the endothelial cells (red fluorescence) and the third of the fibroblasts (green fluorescence). To compute the displacement and traction fields we used the first channel. 32 Figure 6. Confocal image[159,73 x 159,73 microns] sample_3 pos_2 a) endothelial cells channel b)fibroblasts channel c) the channel of endothelils and fibroblasts merged and d)beads channel. a) c) b) d) 33 Figure 7. Confocal image[159,73 x 159,73 microns] sample_3 pos_3 a) endothelial cells channel b)fibroblasts channel c) the channel of endothelils and fibroblasts merged and d)beads channel. a) b) d) c) 34 3.3 Estimating the displacement field and computing the traction field The images of the fluorescent beads, before, i.e. the “force-loaded image” (Figure 8.a and Figure 9.a), and after the adding of Cytochalasin-D, i.e. the “null-force image” (Figure 8.b and Figure 9.b), are used in this. A comparison is made between them for each selected positions by means of an algorithm that uses crosscorrelation[15] to calculate the displacement in pixels of one image with respect to the other, using these reference points (2.9). Figure 8. Beads channel sample_2 pos_2 a) pre Cytochalasin-D and b) post Cytochalasin-D a) b) 35 Figure 9. Beads channel sample_3 pos_3 a) pre Cytochalasin-D and b) post Cytochalasin-D Representative results of the image processing are attached below (Figure 10 and Figure 11). The procedure has been executed for all the positions of the PAG with the soft and intermediate stiffness (all the results in Appendix a). The soft gels show larger displacement with respect to the intermediate ones so the images are less affected by experimental noise and the corresponding displacement fields are more reliable. For the same reason, the stiffer gels have been excluded from the results since the displacements of the beads were too low. As it can be seen, the magnitude of the displacement is presented in a gradient of colours. In dark blue are the pixels that show no displacement and the colour is ascending through the light blue, green, yellow, orange until it reaches red which represents the largest displacement. White arrows indicate the vector of the displacement at each analysed point of the image. a) b) 42 Figure 17. Number of dipoles as a function of algorithm iteration (blue is for dipole size=1, red is for size=2, green is size=3, cyan is size=4, and magenta is size=5). 43 4 CONCLUSIONS AND DISCUSSION First, the results from this report indicate that the calculated real PAG stiffness are quite low. This can be due to the old reagents used and to the learning process in making gels. There may also be experimental errors in measuring it with the confocal microscope, in measuring the gel thickness and/or the indentation depth, especially when indentations are small (for stiffer gels). But nonetheless, our measurements are indicative of their true bulk stiffness. Moreover, by using this method (see 2.5) we can estimate a value for every gel we have prepared. Second, the traction fields and the mean traction values obtained that fibroblasts’ detected tractions are very small, while we have measured larger tractions force patterns in endothelial cells. This could have two different and opposite explanations, (i) fibroblasts have lower adhesive forces than endothelial cells or (ii) since they are attached with stronger adhesive forces, the Cytochalasin-D affect them more slowly than endothelial cell. By running the computation of the median value of the traction forces we noticed that there were positions in which the fluorescence of the fibroblasts was so weak that the algorithm did not detect them. In particular this occurs in the position 1 of the sample 3 where we can clearly see (Appendix 2) that the fibroblasts are present and that they exert tractions. We can hypothesize that the low fluorescence of those fibroblasts was an indication of their health, so the Cythochalasin-D was able to destroy their actin filaments and we could detect the displacements of the fluorescent beads. This observation can be of support of the second proposed explanation. This aspect should be more investigated in the future, for instance, by running experiments leaving more time to the Cytochalasin-D to act. 44 Third, further observations can be done over the results of the calculation of the median spreading area. The graph shows an initial tendency to increase with the stiffness until approaching a plateau, namely a maximal footprint area. This concurs with the results of earlier studies [20]. Furthermore, the spread is greatest for fibroblast with respect to endothelial cell. Because adhesion must occur for cells to test the elasticity of the ECM and continue spreading [20] this results supports the hypothesis that fibroblasts actually adhere strongly to PAG gels as compared with endothelial cells. Again, further experiments should be performed to discern among conflicting hypotheses. Last, our computational reconstruction of force dipoles shows, as proof of concepts, the existence of communication dipoles between cells exerting traction, which could indicate that there is transmission of forces between them. This is very promising and it is a substantial difference compared to the single cell case. However, results are quite preliminary in this sense. In conclusion, in this thesis, an experimental protocol for fabricating call culture and measure both substrate stiffness and traction forces was combined with intracellular dipole calculation related to these traction forces. We found that the forces exerted on the substrate from the cell depends on cell type and that the substrate characteristics, specifically the stiffness, tune the behaviour of the cells enhancing their spreading until a cell-characteristic threshold. We also had proof of the possible existence of a mechanical communication system between cells. 45 5 BIBLIOGRAPHY [1] Moeendarbary E, Harris AR. Cell mechanics: principles, practices, and prospects. Wiley Interdiscip Rev Syst Biol Med. 2014 Sep-Oct;6(5):371-88. doi: 10.1002/wsbm.1275. PMID: 25269160; PMCID: PMC4309479. [2] https://spiegato.com/che-cose-la-biomeccanica-cellulare [3] Zollinger AJ, Xu H, Figueiredo J, Paredes J, Seruca R, Stamenović D, Smith ML. Dependence of Tensional Homeostasis on Cell Type and on Cell-Cell Interactions. Cell Mol Bioeng. 2018 May 8;11(3):175-184. doi: 10.1007/s12195-018-0527-x. PMID: 31719884; PMCID: PMC6816663. [4] Tang X, Tofangchi A, Anand SV, Saif TA. A novel cell traction force microscopy to study multi-cellular system. PLoS Comput Biol. 2014 Jun 5;10(6):e1003631. doi: 10.1371/journal.pcbi.1003631. PMID: 24901766; PMCID: PMC4046928. [5] Ribeiro AJ, Denisin AK, Wilson RE, Pruitt BL. For whom the cells pull: Hydrogel and micropost devices for measuring traction forces. Methods. 2016 Feb 1;94:51-64. doi: 10.1016/j.ymeth.2015.08.005. Epub 2015 Aug 8. PMID: 26265073; PMCID: PMC4746112. [6] Ulrich S. Schwarz, Jérôme R.D. Soiné,Traction force microscopy on soft elastic substrates: A guide to recent computational advances,Biochimica et Biophysica Acta (BBA) - Molecular Cell Research,Vol 1853, Issue 11, Part B,2015,Pages 3095-3104,ISSN 01674889,https://doi.org/10.1016/j.bbamcr.2015.05.028. [7] Benjamin M, Hillen B. Mechanical influences on cells, tissues and organs - 'Mechanical Morphogenesis'. Eur J Morphol. 2003 Feb;41(1):3-7. doi: 10.1076/ejom.41.1.3.28102. PMID: 15121543. 46 [8] Sgarlata C, Torti S., Borroni G, Ciocca O. (2012). Meccanobiologia della riparazione tissutale: dalla fisiopatologia alle implicazioni terapeutiche. Bollettino della Società Medico Chirurgica di Pavia 125(3):507-511. Clinica Dermatologica, Università degli Studi di Pavia, Fondazione IRCCS Policlinico San Matteo, Pavia, Italia [9] Efremov YM, Lomakina ME, Bagrov DV, Makhnovskiy PI, Alexandrova AY, Kirpichnikov MP, Shaitan KV. Mechanical properties of fibroblasts depend on level of cancer transformation. Biochim Biophys Acta. 2014 May;1843(5):1013-9. doi: 10.1016/j.bbamcr.2014.01.032. Epub 2014 Feb 11. PMID: 24530505. [10] Panzetta V, Musella I, Rapa I, Volante M, Netti PA, Fusco S. Mechanical phenotyping of cells and extracellular matrix as grade and stage markers of lung tumor tissues. Acta Biomater. 2017 Jul 15;57:334-341. doi: 10.1016/j.actbio.2017.05.002. Epub 2017 May 5. PMID: 28483699. [11] Lehoux S., Redox signalling in vascular responses to shear and stretch. Cardiovascular Research 71 (2006) 269 – 279 [12] Kolahi KS, Mofrad MR. Mechanotransduction: a major regulator of homeostasis and development. Wiley Interdiscip Rev Syst Biol Med. 2010 Nov-Dec;2(6):625-39. doi: 10.1002/wsbm.79. PMID: 20890961. [13] Laia Ortega Pérez, Measuring Mechanical Contractility In Single Cells, master thesis in biomedical engineering at the UPC (2021) [14] Latorre E., Kale S., Casares L., Gómez-González M., Uroz M., Valon L., Nair R.V., Garreta E , Montserrat N., del Campo A., Ladoux B, Arroyo M, Nature X.T. Active superelasticity in three-dimensional epithelia of controlled shape. 2019 May 16. doi: 10.1038/s41586-018-0671-4 47 [15] Landauer, A.K., Patel, M., Henann, D.L. et al. A q-Factor-Based Digital Image Correlation Algorithm (qDIC) for Resolving Finite Deformations with Degenerate Speckle Patterns. Exp Mech 58, 815–830 (2018). https://doi.org/10.1007/s11340-0180377-4 [16] Bar-Kochba, E., Toyjanova, J., Andrews, E. et al. A Fast Iterative Digital Volume Correlation Algorithm for Large Deformations. Exp Mech 55, 261–274 (2015). https://doi.org/10.1007/s11340-014-9874-2 [17] Trepat, X., Wasserman, M., Angelini, T. et al. Physical forces during collective cell migration. Nature Phys 5, 426–430 (2009). https://doi.org/10.1038/nphys1269 [18] https://www.angioproteomie.com/commerce/ccp3099-rfp-expressing-human-umbilicalvein-endothelial-c-cap-0001rfp.htm [19] https://imagej.net/sftware/fiji [20] Bell S, Redmann AL, Terentjev EM. Universal Kinetics of the Onset of Cell Spreading on Substrates of Different Stiffness. Biophys J. 2019 Feb 5;116(3):551-559. doi: 10.1016/j.bpj.2018.12.020. Epub 2019 Jan 5. PMID: 30665696; PMCID: PMC6369430. [21] Malandrino A, Trepat X, Kamm RD, Mak M (2019) Dynamic filopodial forces induce accumulation, damage, and plastic remodeling of 3D extracellular matrices. PLOS Computational Biology 15(4): e1006684. https://doi.org/10.1371/journal.pcbi.1006684 [22] Hazlett, L., Landauer, A.K., Patel, M. et al. Epifluorescence-based three-dimensional traction force microscopy. Sci Rep 10, 16599 (2020). https://doi.org/10.1038/s41598-02072931-6 48 Appendix a. Cells and Displacements Softer PAG gels Sample1_pos1 Sample1_pos2 49 Sample1_pos3 Sample2_pos2 50 Sample3_pos1 Sample3_pos2 51 Sample3_pos3 58 Sample3_pos3 59 5.1.1 Intermediate PAG gels Sample4_pos1 Sample4_pos2 60 Sample4_pos3 Sample6_pos1 61 Sample6_pos2 Sample6_pos3 62