Fibronectin III9-10 adsorption to self-assembled monolayers and interdomain orientation in the context of material-driven fibrillogenesis studied with molecular dynamics simulations

Student thesis: Doctoral ThesisDoctor of Philosophy


Fibronectin (Fn) is a structural glycoprotein that plays a major role in the extracellular matrix (ECM) and communication with the intracellular environment. Fibronectin brillogenesis is triggered after its quaternary structure transitions from a compact to an extended conformation. In vivo, this takes place via the FnIII9-10 - integrin α5β1 binding after which the intracellular skeleton exerts force on the structure. However, brillogenesis can also take place on certain materials, such as ethyl acrylate (EA) functionalised self-assembled monolayers (SAMs). Surprisingly, bronectin brillogenesis does not take place on a chemically similar surface, methyl acrylate (MA) SAMs. Classical molecular dynamic simulations are used to investigate the adsorption of FnIII9-10 and its relation to brillogenesis. I show that surface water hydration explains why Fn adsorbs to EA SAMs but not to MA SAMs, which di ers only by one extra methylene bridge. The FnIII9-10 domains adsorb with the 9th domain to EA SAMs with the CHARMM36 forcefield, which leads to exposure of the RGD and PHSRN motifs for potential binding. I also show, however, that the 10th domain by itself adsorbs well to the surface. The two domains adsorb well to methyl SAMs in a non-specific way, always burying the two motifs in the surface. I reproduce the simulations with the CHARMM36m forcefield and show that, in contrast to CHARMM36, FnIII9-10 adsorbs with the 10th domain to EA SAMs, which also buries the motifs in the surface. Moreover, with CHARMM36m, adsorption of FnIII9-10 to methyl SAMs converges to the same adsorption state, which always makes the RGD and PHSRN motifs unavailable for binding. Furthermore, I show that the force eld CHARMM36 does not reproduce the correct behaviour of the FnIII9-10 in bulk water. Unfolding of the tertiary structure of the 9th domain takes place, which is inconsistent with the experimentally determined stability and structure of the domain. I further show that this problem disappears when the CHARMM36m forcefield is used, which I then use to investigate the different FnIII9-10 interdomain orientations. I show that the two domains have a major preference for the interdomain conformation found in the crystal structure with PDB 1FNF. However, I also show that there are two less-common interdomain orientations which appear to be stable, and one of those has an acute angle between the two domains. Our investigation of FnIII9-10 is discussed in the light of the available experimental data and any extrapolations are discussed with respect to their potential biological role.
Date of Award1 Apr 2020
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorChris Lorenz (Supervisor)

Cite this