Physical interactions between epithelial cells organize the protective coverings that line the surface and cavities of many organs such as the colon and mammary gland. These complex interactions were thought to have largely emerged from the cooperation between the cortical actin cytoskeleton and adhesive binding between cells. In epithelia, this is evident from the molecular to the tissue level. From the molecular level, E-cadherin adhesion complexes couples the contractile cortical actin cytoskeleton of neighbouring cells together to support the apical Zonula Adherens (ZA). Resistance of cortical actomyosin contractility by E-cadherin adhesion then produces tension at the ZA that is transmitted across epithelial tissues. While much is known about the physical coupling between cadherin adhesion and the actomyosin cytoskeleton, the functional interplay between these components remains incompletely understood. A clearer understanding is essential as dysregulation of the cadherin-actomyosin system occurs in many diseases, including cancer. To achieve this, a variety of photonic methods were employed to probe the molecular and biophysical properties of intercellular contacts. This is combined with genetic, growth factor and pharmacological perturbations to characterize the molecular and cellular mechanisms of epithelial interactions during homeostasis and upon oncogenic transformation.
I began by studying the ZA to understand the role of E-cadherin in mechanotransduction. First, a recombinant lentivirus system that depletes endogenous E-cadherin and reconstitutes with fluorescently-tagged E-cadherin was generated, to transduce cells and enable real-time quantitative analysis of E-cadherin junctional movements. Further, both oscillatory and unilateral translational components of apical ZA movements were extracted and attributed to the Myosin II isoforms, thus providing evidence that Myosin II isoforms exert contractile forces on E-cadherin junctions. This capacity for Myosins to support the ZA also carries the implication that these motor proteins might also be targeted to perturb junctional integrity. Indeed, Hepatocyte Growth Factor (HGF), which is known to disrupt cell-cell interactions, acutely perturbs ZA integrity much more rapidly than generally appreciated and this entails displacement of Myosin IIA, IIB and VI from junctions. Of note, disruption of Myosin VI interactions with E-cadherin leads to a failure of actin filament anchorage to E-cadherin junctions ultimately resulting in the fragmentation of the ZA.
However, the ZA is not the only region where cadherins engage in adhesion. E-cadherin receptors were observed to distribute as adhesive clusters throughout the apicolateral axis of cell-cell junctions. These E-cadherin clusters engage with, and act to establish, the contractile actomyosin cortex throughout the apicolateral axis of cell-cell contacts. Interestingly, laser nanodissection of intercellular contacts revealed that cells can establish distinct regions of contractile activity within individual junctions, generating high tension at the apical ZA but substantially lower tension elsewhere. This contrasts with the prevailing notion that tension is a homogeneous property of junctions. Ultimately, I have characterized three interacting processes that establish the distinct dynamic properties of the lateral junctions, namely cadherin-dependent actin assembly, actin-based myosin recruitment (self-assembly), and myosin-induced local turnover of actin filaments.
Quantitative live-imaging of lateral E-cadherin clusters and the cortical actin cytoskeleton further revealed that cyclical condensation and disassembly of the actomyosin network is a distinctive property of the lateral junction, whereas F-actin networks are more stable at the ZA. Indeed, my data indicates that apical stabilization is mediated by N-WASP at the ZA, which is responsible for generating higher apical tension. Further, ectopic targeting of N-WASP to stabilize F-actin at lateral junctions was sufficient to increase tension at these sites. This implied that modulation of F-actin stability by N-WASP is an independent determinant of contractile tension, thus complementing the conventional focus on regulating tension through myosin II contractility.
Finally, the propensity for cancer cells to be retained in cohesive monolayers was perturbed when distinctive apicolateral patterning of junctional contractility was altered; either when N-WASP redistributes to lateral junctions in cells expressing oncogenic H-RasV12 or upon mosaic redistribution of active N-WASP itself. As studies investigating the impact of oncogenic transformation on junctional biomechanics are still relatively limited, this finding provides novel mechanistic insights into the emerging role of tension in the regulation of cancer cell metastasis. Taken together, I propose that local control of the organization and stability of junctional actin filaments regulates the landscape of cortical myosin contractility to determine whether, or not, cells integrate into epithelial populations.
In summary, my PhD work provides a quantitative description of core molecular processes that determine the biomechanics of epithelial interactions and its dysregulation upon oncogenic transformations.