Caveola mechanotransduction reinforces the cortical cytoskeleton to promote epithelial resilience

Introduction

Epithelia form large sheets to create protective barriers and interfaces to external surfaces¹. These tissues often face disruptive forces that can compromise their integrity and cause diseases to develop^2,3. To withstand physical stress, epithelial cells have developed mechanisms to sense and respond to forces^4–6, which rely on conformational changes due to increased force – in proteins like adherens junctions (AJs) protein α-catenin⁷, or in membrane structures like caveolae⁸.

Caveolae are cholesterol- and sphingolipid-rich lipid rafts that form Ω-shaped invaginations of the plasma membrane^9,10 and play a role in endocytosis, lipid metabolism and mechanosensing¹¹. Caveolae consist of two protein families – caveolins and cavins^12,13. The bulb-like morphology of caveolae provides a membrane reservoir which can be flattened by stretch and, thus, can passively buffer tension⁸. Previous work of R. Parton’s and A. Yap’s groups¹⁴ found caveolae to modulate mechanical tension within epithelial monolayers at the tissue level by affecting the pool and organisation of cortical F-actin, resulting in perturbation of oncogenic cell extrusion.

The following preprint focuses on the effects of acute, mechanically-induced caveolae disassembly on tissue stabilization and mechanical resilience aiming to uncover additional functions of caveolae beyond passive tension buffering.

Key findings

Mechanical stress increases tension at adherens junctions in a caveola-dependent fashion

To induce acute disassembly of caveolae, the authors used two types of mechanical stressors – hypoosmotic stimulation and external stretch – and observed the cell-cell junctional and cytoskeletal response to the stimulation. Hypoosmotic stimulation of MCF-10 mammary epithelial cell monolayers for 5 minutes caused cells to swell, leading to release of cavins from caveolae and, thus, to disassembly of the structure. As a response, F-actin got reinforced at cell-cell junctions and myosin II increasingly localized to the cell-cell contacts. However, F-actin enrichment did not occur in caveolin-1 knock-down (cav-1 KD) cells suggesting caveolae’s importance for this short-term response.

To determine whether the reinforcement of F-actin at cell-cell junctions correlated with increased tension and altered junctional mechanics, the authors measured the recoil of bicellular junctions upon laser ablation and quantified the amount of conformationally unfolded α-catenin at AJs. After hypoosmotic stimulation, the baseline recoil of junctions was increased significantly and recovered to baseline after addition of iso-osmotic medium, suggesting short-term increase of AJs tension, which was confirmed by increased levels of open-conformation α-catenin. In cav-1 KD cells, hypoosmotic stimulation did not increase AJs tension, as measured by both methods, suggesting that hypo-osmotic stimulation increases mechanical tension at AJs in a caveola-dependent manner.

Caveolae mechano-activation increases mechanical tension at cell-cell junctions via F-actin reinforcement by the PtdIns(4, 5)P₂-FMNL2 pathway

To uncover the source of increased AJs tension – the membrane itself or cellular contractility – the authors inhibited myosin II. Para-aminoblebbistatin treatment prior to hypoosmotic shock abolished the short-term increase in AJs tension, leading to the conclusion that increased tension at AJs results primarily from F-actin accumulation and is facilitated by altered cytoskeletal contractility.

To characterise the molecular pathway of junctional actin reinforcement upon caveolae stimulation by hypoosmotic treatment, the authors examined levels of phosphoinositide‑4, 5‑bisphosphate (PtdIns(4, 5)P₂) at AJ membranes motivated by previous work of the group¹⁴. Transient expression of a location sensor was used to visualize PtdIns(4, 5)P₂, which localized primarily to the cell membrane. Upon hypoosmotic treatment, PtdIns(4, 5)P₂ levels increased by ~20% within 5 minutes – a response which did not happen in cav-1 KD. The phenotype was, however, rescued by expression of RNAi-resistant cav-1 transgene. To examine the pathway further downstream, the authors expressed FMNL2-EGFP – a formin that promotes actin assembly at the junctional cortex¹⁴, which increasingly accumulated at AJs upon hypoosmotic stimulation, in line with the PtdIns(4, 5)P₂ increase. This increase was again abolished in the cav-1 KD cells, suggesting a role of caveolae for the effect. Further, pre-treatment with PtdIns(4, 5)P₂ antagonist neomycin abolished FMNL2-EGFP accumulation at cellular junctions upon hypoosmotic shock, confirming the requirement of PtdIns(4, 5)P₂ for FMNL2 recruitment. Together, these results highlight that acute caveolae mechanoactivation leads to increase of PtdIns(4, 5)P₂ levels and to downstream activation of FMNL2 activity to reinforce junctional F-actin.

Molecular stabilisation of caveolae antagonises protective cortical reinforcement

To test if PtdIns(4, 5)P₂-FMNL2 pathway action is a direct consequence of caveolae disassembly upon mechanical stimulation, the authors employed a molecular strategy to prevent caveolae disassembly. To increase resistance to disassembly, the authors expressed zebrafish cavin1b (DrCavin1b), which contains five undecad cavin1 (UC1) repeat domains that can strongly interact with phosphatydylserines, compared to the only two UC1 domains mammalian cavins contain. The authors expressed either mouse cavin-1 (MmCavin1-EGFP), the zebrafish cavin1 (DrCavin1b-EGF) or a control zebrafish cavin1 with deletion of four UC1 domains (∆4UC1-EGF) in cavin-1 KD MCF-10 cells. As expected, cavin1-EGFP fluorescence upon hypoosmotic stimulation decreased in the MmCavin1-EGFP mutants, signifying dissociation of caveolae, while the dissociation was much less pronounced in the DrCavin1b-EGF mutant. With this successful caveolae stabilization strategy, the authors assessed the effects of caveolae stabilization on the F-actin reinforcement phenotype observed under hypoosmotic stimulation. Reconstitution of cavin1 KD with MmCavin1-EGFP or with ∆4UC1-EGFP restored the ability of cells to reinforce the cortex, while cortical reinforcement was effectively abolished by the DrCavin1b-EGFP, where caveolae were less prone to disassemble. This implied that the process of caveola disassembly is necessary for MCF-10 cells to reinforce the cortex in response to mechanical stress.

Lastly, the authors focused on the role of caveolae in the protection of epithelial integrity against disruptive contractile mechanical stress. To test this, they increased cellular contractility from within the epithelium by using calyculin A, which increases non-muscle myosin II contractility and ultimately leads to monolayer tears. The authors measured the time to initial appearance of sheet fractures upon calyculin A treatment in cavin1 KD cells reconstituted with either MmCavin1-EGFP, ∆4UC1-EGFP or with DrCavin1b-EGFP. MmCavin1-EGFP and ∆4UC1-EGFP cavin1 KD monolayers showed similar times as WT MCF-10 cells, while molecular stabilization of caveolae by DrCavin1b-EGFP expression accelerated the rupture process by ~20%, similar to the group in which PtdIns(4, 5)P₂ signalling was blocked with neomycin. This result suggests that caveolae disassembly forms part of the physiological protective reaction of the epithelium to increased contractile stress.

What I like about this preprint

Traditional mechanobiology research has primarily focused on well-established players such as cell-cell junctional proteins, integrin receptors, and ion channels. However, lesser-known membrane structures, too, deserve attention if not only for their intriguing contributions to mechanosensation. This is particularly relevant in epithelial tissues, where membrane processes and transport mechanisms play crucial roles in maintaining tissue polarization and function. I chose this preprint especially because it highlights the importance of a relatively unexplored membrane-associated structure for mechanosensing and epithelial resilience – the caveolae.

Further, I was particularly intrigued by the molecular stabilization approach employed by the authors to prevent caveolae disassembly and I liked the multi-sided approach to narrow down particular molecular contributors. Lastly, this paper is of personal interest to me, and the insights gained from it offer valuable perspectives for my own research.

Questions for the authors

How do you explain the independence of myosin II accumulation at cell-cell junctions from caveolae? Further, is there a general difference in F-actin turnover or in F-actin polymerization activity between wild type and cav-1 KD cells?

How is the long-term mechanoadaptation different in cav-1 KD cells compared to wild type? Is there some permanent adaptation in terms of cellular cytoskeleton structure or other junctional complexes (e.g., desmosomes) and their dynamics as means for compensation?

Is there any preferential disassembly of caveolae containing a certain cavin isotype given that different isoforms may trigger different downstream responses? Did you further explore the intrinsic signalling the released cavins may trigger in the explored conditions?

References

1. Ganz, T. Epithelia: Not just physical barriers. Proc National Acad Sci 99, 3357–3358 (2002).
2. Leoni, G., Neumann, P.-A., Sumagin, R., Denning, T. L. & Nusrat, A. Wound repair: role of immune–epithelial interactions. Mucosal Immunol 8, 959–968 (2015).
3. Kılıç, A. et al. Mechanical forces induce an asthma gene signature in healthy airway epithelial cells. Sci Rep-uk 10, 966 (2020).
4. Yap, A. S., Duszyc, K. & Viasnoff, V. Mechanosensing and Mechanotransduction at Cell–Cell Junctions. Csh Perspect Biol 10, a028761 (2018).
5. Jo, J., Nansa, S. A. & Kim, D.-H. Molecular Regulators of Cellular Mechanoadaptation at Cell–Material Interfaces. Frontiers Bioeng Biotechnology 8, 608569 (2020).
6. Chen, Y., Ju, L., Rushdi, M., Ge, C. & Zhu, C. Receptor-mediated cell mechanosensing. Mol Biol Cell 28, 3134–3155 (2017).
7. Yao, M. et al. Force-dependent conformational switch of α-catenin controls vinculin binding. Nat Commun 5, 4525 (2014).
8. Pozo, M. A. D., Lolo, F.-N. & Echarri, A. Caveolae: Mechanosensing and mechanotransduction devices linking membrane trafficking to mechanoadaptation. Curr Opin Cell Biol 68, 113–123 (2021).
9. Parton, R. G. & Collins, B. M. The structure of caveolin finally takes shape. Sci Adv 8, eabq6985 (2022).
10. Predescu, D., Predescu, S., McQuistan, T. & Palade, G. E. Transcytosis of alpha1-acidic glycoprotein in the continuous microvascular endothelium. P Natl Acad Sci Usa 95, 6175–80 (1998).
11. Lo, H. P. et al. The caveolin–cavin system plays a conserved and critical role in mechanoprotection of skeletal muscle. J Cell Biology 210, 833–849 (2015).
12. Drab, M. et al. Loss of Caveolae, Vascular Dysfunction, and Pulmonary Defects in Caveolin-1 Gene-Disrupted Mice. Science 293, 2449–2452 (2001).
13. Wei, Z. et al. The N-terminal leucine-zipper motif in PTRF/cavin-1 is essential and sufficient for its caveolae-association. Biochem Bioph Res Co 456, 750–756 (2015).
14. Teo, J. L. et al. Caveolae Control Contractile Tension for Epithelia to Eliminate Tumor Cells. Dev Cell 54, 75-91.e7 (2020).

 

Tags: adherens junctions, caveolae, epithelia, lipid rafts, mechanobiology, mechanosensing, membrane tension, plasma membrane, tissue mechanics

Posted on: 15 June 2023 , updated on: 17 March 2024

doi: https://doi.org/10.1242/prelights.34871

Read preprint (1 votes)