Pablo Sáez (LaCàN, IMTech)
Cells engage in continuous interactions with the extracellular matrix (ECM). This connection between cells and the ECM is responsible for various cellular and tissue responses, including tissue morphogenesis, repair, and organ development [1]. However, when unregulated, this cell-ECM connection can lead to pathological conditions. Among the factors influencing cell-ECM behavior, the composition and mechanical properties of the ECM are crucial. Central to these interactions are cell membrane receptors called integrins, which act as molecular sensors recognizing specific ECM components [2]. Once bound to the ECM, integrins facilitate intracellular tension transmission, influencing the localization and activity of key transcriptional regulators (genes involved in the regulation of gene expression) like YAP, which is closely linked to cancer initiation [3]. Cells with an overexpression of YAP are highly invasive. Moreover, integrins dynamically respond to changes in ECM mechanical properties and composition, thereby modulating signaling pathways and gene expression.
While extensive research has focused on ECM components like fibronectin and collagen, the precise role of laminin remains unclear. Laminin plays a pivotal role in guiding cellular processes, from maintaining healthy epithelial homeostasis to promoting cancer metastasis. A recent publication led by Jenny Kechagia and Pere Roca-Cusachs from IBEC investigated the role of laminin in cell adhesion [4]. Using MCF10A breast epithelial cells, they observed that cells seeded on laminin showed reduced traction forces and YAP nuclear localization compared to cells on Collagen I and fibronectin substrates. To unravel the mechanisms underlying laminin-mediated effects on nuclear mechanoresponses, they turned their attention to the cell cytoskeleton—a network of protein filaments providing structural support within the cell, and its connections to the ECM. Comprising actin filaments, intermediate filaments (e.g., keratin), and microtubules, the cytoskeleton can link from one side to specific integrins and from the other to the cell nucleus. This chain of intracellular components transmits internal or external forces to the ECM and the nucleus. When it reaches the nucleus, it induces nuclear expression of YAP through several mechanosensitive events. For instance, the keratin network links to the ECM through α6β4 integrins and to the nucleus through a protein called nesprin-3.
To understand such mechanotransduction chain, and elucidate how force tramission influences YAP expression, they blocked integrin subunits involved in laminin interactions, showing that α6β4 integrins play a crucial role in modulating the mechanoresponse of the cells. Contrary to expectations, blocking these integrins, which should reduce force transmission to the nucleus, led to increased YAP nuclear localization. However, focal adhesion size or traction forces did not change, which indicated that force transmission neither changed. This raises a puzzling question: if cell adhesion is pivotal for force transmission from the ECM to the nucleus, how can YAP expression change without corresponding changes in traction forces? To further analyze this mechanotransduction chain, the researchers knocked down nesprin-3, a protein connecting the keratin cytoskeleton to the nucleus, reducing YAP nuclear-to-cytoplasmic ratios. This was expected as tension transmission to the nucleus was inhibited. These opposing effects implied that the keratin network alone was not responsible for the changes in YAP expression.
Given the intricate interplay between actin and intermediate filament cytoskeletal networks, it is plausible that the keratin cytoskeleton indirectly affects the nucleus by regulating how actin-mediated force generation reaches it. To explore this hypothesis, Marino Arroyo and I, from LaCàN, developed a computational model of the interaction between these networks and the ECM. To study this cellular system, we described the balance of forces in the actin network as
and in the keratin network as
The model treated the actomyosin cytoskeleton as an active and viscous gel undergoing turnover, while the keratin cytoskeleton was modeled as a passive viscoelastic gel [5, 6]. Therefore, the constitutive relations for the actin and the keratin network were described as 
and 
, respectively. μa and μIF are the viscosity of the actin and keratin network, respectively. ζ represents the contractility of the myosin motors exerted on the actin filaments, the main source of intracellular force generation, and ρ is the density of the actomyosin network, which was also modeled by a transport equation (see [4]). G is the elastic stiffness of the keratin network and λ is its stretch, which we also model separately. Integrin-mediated adhesions between these networks and the substrate were incorporated with cytoskeletal-substrate friction coefficients. The friction of the actin and keratin network with the ECM is given by ηa and ηIF, respectively. Moreover, there is friction between these two networks modeled as 
. This coupled system of equations results in the velocity of the actin and keratin network, va and vIF respectively. As a result, we can compute the tension on the nucleus and estimate the increase or decrease in YAP expression.
The model predictions revealed that inhibiting the cell-ECM connection through α6β4 integrins, as experimentally done, results in the actomyosin network dragging the keratin network inward. This manipulation reduced the connection of the keratin network with the ECM and, therefore, its friction coefficient. Consequently, keratin accumulates around the cell nucleus, while actin flow and organization remain largely unaffected. Experimental validation confirmed disrupted keratin organization in mutant integrin-expressing cells compared to controls, while the actin network remained unaffected. However, computation of the traction force on the nucleus showed that these changes in intracellular forces were insufficient to explain the nuclear morphological shape changes measured experimentally, which served as an indicator of tension exerted on them and, consequently, on YAP expression.
The model was successful in recapitulating actin and keratin velocity and distribution but fell short in predicting the actual tension on the nucleus. The missing piece of the puzzle lay in understanding how the mechanical properties of the keratin network change when the connection through α6β4 integrins is inhibited. Previous publications demonstrated that when the keratin network has lower cross-linking with the ECM, its stiffness reduces. To incorporate this idea, we adjusted the model to make the stiffness of the keratin network, represented by G, proportional to the density of keratin and the adhesion strength, which was also confirmed experimentally. This modification in the model predicted a more pronounced increase in nuclear deformation in mutant cells, as quantified by lower sphericity in the experimental results and, effectively, in an increase in YAP expression.
In summary, through close and interactive collaboration between experimentalists and modelers, we were able to elucidate how cells behave when attached to substrates of different compositions and mechanical properties. We demonstrated how the intracellular transmission of forces between different intracellular structures induces changes in nuclear tension, thereby affecting YAP expression. These findings provide insights into cellular behavior in various contexts where laminin and keratin play crucial roles, such as cancer progression or early developmental stages.
References
[1] Hynes, Richard O. Extracellular matrix dynamics in development and regenerative medicine. Journal of Cell Science 125, 24 (2012): 5597–5608.
[2] Geiger, Benjamin and Bershadsky, Alexander. Integrin signaling: specificity and control of cell survival and cell cycle progression. Current Opinion in Cell Biology 13, no. 5 (2001): 563–573.
[3] Piccolo, S., Panciera, T., Contessotto, P. et al. YAP/TAZ as master regulators in cancer: modulation, function and therapeutic approaches. Nature Cancer 4 (2023): 9–26.
[4] Kechagia, Z., Sáez, P., Gómez-González, M., Canales, B., Viswanadha, S., Zamarbide, M., Andreu, I., Koorman, T., Beedle, A. E. M., Elosegui-Artola, A., Derksen, P. W. B., Trepat, X., Arroyo, M., & Roca-Cusachs, P. (2023). The laminin–keratin link shields the nucleus from mechanical deformation and signalling. Nature Materials, 22(11), 1409–1420.
[5] J. Prost, F. Jülicher, J. F. Joanny, Active gel physics, Nature Physics 11 (2) (2015) 111–117.
[6] Betorz, J., Bokil, G. R., Deshpande, S. M., Kulkarni, S., Araya, D. R., Venturini, C., & Sáez, P. (2023). A computational model for early cell spreading, migration, and competing taxis. Journal of the Mechanics and Physics of Solids, 179, 10539