Computational investigation of early events in plant growth signaling

Moffett, Alexander Scott

Permalink

https://hdl.handle.net/2142/105644 Copy

Description

Title

Computational investigation of early events in plant growth signaling

Author(s)

Moffett, Alexander Scott

Issue Date

2019-07-05

Director of Research (if dissertation) or Advisor (if thesis)

Shukla, Diwakar

Doctoral Committee Chair(s)

Shukla, Diwakar

Committee Member(s)

• Aksimentiev, Aleksei
• Hind, Sarah R
• Zhao, Huimin

Department of Study

School of Molecular & Cell Bio

Discipline

Biophysics & Computnl Biology

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• brassinosteroid
• molecular dynamics simulations
• BRI1
• BAK1
• computational biophysics

Abstract

Brassinosteroids are a class of plant hormones critical for control of growth and development. While the signaling pathway through which brassinosteroid signals are processed has been the subject of intense study for the past 50 years, relatively little is known about the precise nanoscopic events underlying brassinosteroid signal transduction in terms of protein conformational dynamics. Focusing on the two primary brassinosteroid coreceptors, BRI1 and BAK1, we use a number of computational methods grounded in molecular dynamics simulations with the goal of understanding the physical mechanisms of brassinosteroid-induced BRI1-BAK1 association and activation. First, we examine the brassinosteroid-induced association of the BRI1 and BAK1 extracellular domains. Using a replica exchange umbrella sampling scheme together with alchemical free energy calculations and conventional molecular dynamics simulations, we examine the mechanisms through which brassinosteroid binding to the BRI1 extracellular domain induces association with the BAK1 extracellular domain. We find that in addition to the stability provided by specific interactions between a brassinosteroid and both BRI1 and BAK1 in a manner depending on protonation of BAK1 residue H61, brassinosteroid binding stabilizes the BRI1 island domain while BRI1 can undergo a large conformational change which may further stabilize the BRI1-BAK1 complex. The remaining work is focused on the BRI1 and BAK1 kinase domains. We performed adaptive sampling simulations on the isolated BRI1 and BAK1 kinase domains in their fully phosphorylated forms and built Markov state models from those simulations in order to examine the behavior of features characteristic of active protein kinases. We find that the BAK1 kinase domain displays local unfolding of the αC helix, while the BRI1 αC helix also unfolds, though to a lesser extent, while also swinging out away from the rest of the protein. Both behaviors are indicative of deactivation. Circular dichroism experiments and bioinformatic analysis indicate that the BRI1 and BAK1 αC helices are to some extent disordered, and that this may be a common feature in Arabidopsis thaliana protein kinases. Next, we examine the mechanisms of phosphorylation-induced BAK1 activation, performing Gaussian accelerated molecular dynamics simulations on seven phosphorylation states likely to be important for BAK1 activation as well as one ATP-bound system. Again monitoring features known to be important for protein kinase activation, we find that phosphorylation of BAK1 T450 is critical for stabilization of the activation loop while T450 and T455 phosphorylation stabilize a more active αC helix conformation. Phosphorylation of T446 and T449 have much less of an effect, though both may have a role in interaction with BAK1 substrates. Finally, we use adaptive sampling simulations and Markov state models to gain insight into the experimental finding that BAK1 is deactivated in vitro by S-glutathionylation. Simulating BAK1 with each of its three solvent-accessible cysteine residues individually S-glutathionylated, we find that modification of C408 induces further αC helix unfolding due to direct interactions, while causing global shifts in the BAK1 kinase domain dihedral angles. S-glutathionylation of C353 and C374 appears to have much less of an effect. This suggests that S-glutationylation of C408 may be responsible for the observed deactivation of BAK1.

Graduation Semester

2019-08

Type of Resource

text

Permalink

http://hdl.handle.net/2142/105644

Copyright and License Information

Copyright 2019 Alexander Scott Moffett

Owning Collections