Mechanobiological Insights and Unfolding Molecular Dynamics of Cell-Matrix Bond Mutant

Introduction:: Cell-matrix bonds allow cell communication with its outside environment, serving as a regulator of proliferation, differentiation, and apoptosis1,2. One such cell-matrix bond, ɑ5ꞵ1 integrin and fibronectin (ɑ5ꞵ1-FN), has been implicated as a mediator of cell-tissue homeostasis and as a potential therapeutic target for diseases such as cancer and lung fibrosis. Notably, a synergy site double mutation of ɑ5ꞵ1-FN, namely R1374/9A, significantly reduces cell adhesion strength compared to wildtype (WT)3. Describing the molecular mechanisms that dictate these adhesion dynamics could illustrate how this disruption in cell adhesion strength occurs at the nanoscale, with potential broader implications for tailoring cell-matrix bonds for engineered tissues. Here, we present a computational study that compares the unique adhesion dynamics and unfolding behavior of WT and mutant (MT) ɑ5ꞵ1-FN under an applied displacement rate.

Materials and Methods:: All-atom steered molecular dynamics (MD) simulations of the ectoplasmic ɑ5ꞵ1-FN complex were run in GROMACS 2020.44. The structure file was downloaded from the protein data bank5. The integrin heads and 7-10 FN segments were isolated in PyMOL6 (Fig. 1A). We used MODELLER 10.47 to impose the virtual R1374/9A double mutation (Fig. 1A). The wildtype and mutated structures were solvated in a TIP3P water box (18nm x 45nm x 19nm) with 0.15 mM NaCl. Energy was minimized for 15k steps with the steepest gradient descent algorithm, then equilibrated with a 1ns NVT at 310K followed by a 10ns NPT at 1 bar and 310K. After verifying equilibration, K559 and E36 at the proximal ends of the integrin headpieces were restrained. P1142 at the distal end of the FN chain was pulled at 10nm/ns using a 50kJ/mol/nm spring with an umbrella potential for 3ns. The trajectories were analyzed using Time-Resolved Force Distribution Analysis8 to measure the punctual stresses, akin to molecular forces, at each residue across all timesteps.

Results, Conclusions, and Discussions:: During extension, the FN-9 segment of the WT ɑ5ꞵ1-FN unfolded, in contrast to the primary unfolding of FN-10 in the MT ɑ5ꞵ1-FN (Fig. 1B). This preferential unfolding pattern did not lead to any alterations in the overall force transmitted through the structure, with forces peaking at 729 pN and 704 pN for the WT and MT cases, respectively (Fig. 1C, top). The force required to unfold FN may be the predominant limiter in terms of force transmission through the complex, i.e., more force is required to unfold FN than it is to rupture the ɑ5ꞵ1-FN bond. However, the reduced interaction between R1374/9, which resides in FN9, and the integrin head could explain why FN9 separates first in the MT case, rather than holding near the ɑ5 head as observed in the WT complex. Although the salt bridge between R1379 and D154 was broken after 1.5ns, it did not lead to immediate separation of FN9 from the ɑ5 head (Fig. 1C, middle). This could be due to heightened punctual stress transmission between E1405 in FN9 and ɑ5. This increase in interaction was not observed in the MT variant (Fig. 1C, bottom). The unfolding sequence of ɑ5ꞵ1-FN changed due to the R1374/9A double mutation disrupting the link between FN9 and ɑ5. However, the overall “spring-like” behavior of ɑ5ꞵ1-FN was maintained. Therefore, an alternative mechanism, such as a reduction in binding energy, may lead to the reduction in cell adhesion strength. Ongoing work focuses on calculating WT and MT binding energies to later feed into a multiscale whole-cell model of cell adhesion. These multiscale dynamics could provide insight on how cell-matrix bonds behave within larger cell and tissue engineered constructs.

Acknowledgements (Optional): :

References (Optional): : 1Ingber DE. Proc. Natl. Acad. Sci. 2003, 100:1472-1474. 2Jahed ZH et al. Internat. Rev. Cell Mol. Biol. 2014, 110: 2475-2483. 3Friedland et al. Science. 2009. 4Abraham MJ et al. SoftwareX, 1-2, 19-25, 2015. 5Available at: rcsb.org/structure/7nwl 6Schrodinger L & DeLano W. PyMOL. 2020. Available at: http://www.pymol.org/pymol. 7Webb B and Sali A. Current Protocols in Bioinformatics. 2016. 8Costescu and Grater. BMC Biophysics. 2013.