Computational modeling of peptide-mediated charge transport and membrane-associated phenomena

Meigooni, Moeen Shamseddin

Permalink

https://hdl.handle.net/2142/127339 Copy

Description

Title

Computational modeling of peptide-mediated charge transport and membrane-associated phenomena

Author(s)

Meigooni, Moeen Shamseddin

Issue Date

2024-11-06

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

Tajkhorshid, Emad

Doctoral Committee Chair(s)

Tajkhorshid, Emad

Committee Member(s)

• Schroeder, Charles M
• Shukla, Diwakar
• Pogorelov, Taras

Department of Study

School of Molecular & Cell Bio

Discipline

Biophysics & Quant Biology

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• molecular dynamics
• membrane curvature
• peptide conductance
• electron transport

Abstract

Proteins and membranes play a crucial role in fueling cellular activity and maintaining cellular homeostasis. Proteins act as energy conduits within cells by shuttling electrons through a series of reactions to fuel the synthesis of adenosine triphosphate (ATP). In mitochondria, membranes provide an ideal structural framework to efficiently carry out the electron transport that forms the basis of biological energy production processes. By relying on the principle of compartmentalization, chemical redox processes within the cell can reliably proceed forward and avoid back reactions. Disturbances to both membrane and protein structure are therefore crucially linked to disease, and as the role of membrane and protein biology is uncovered, the need to understand the molecular and atomic bases of protein-, peptide-, and membrane-mediated charge transport remains a paramount challenge. Despite recent progress, theoretical and computational modeling of the complex interplay between chemical composition, structure, and function of these biomolecules remains challenging. To address these challenges, I have pursued the characterization of peptide-mediated charge transport and membrane-associate phenomena, particularly through the use of molecular dynamics (MD) simulations. MD simulations set a standard for revealing the underlying molecular mechanisms of biophysical and biochemical experiments. It is a particularly effective technique for uncovering atomic-level interactions that contribute to experimental observables. First, I explore the structure-activity relationships governing transmembrane ion conductance mediated by Aβ(1-42) peptides. Enabled by precise NMR structures of putative neurodegenerative structures of Aβ(1-42) oligomers, our simulations revealed a mechanism of membrane disruption in which permeation occurred through lipid-stabilized pores mediated by hydrophilic residues located on the core β-sheet edges of the Aβ(1-42) oligomers. Such behavior resembles the toroidal pore-type behavior shown by many antimicrobial peptides and would also be consistent with the reported antimicrobial activity for Aβ, and could form the basis for presumed ion transport across neuronal membranes that results in disruption of homeostasis and neurodegeneration. Moreover, the role of hydrophilic residues in membrane disruption is consistent with previous studies that show small penalties to expose charged side chains, such as lysine and arginine, to lipids due to the stabilizing influence of membrane deformations for the protonated form. Next, I sought to explore the impact of peptide sequence and structure on bioenergetic electron transport. While the process of electron transport is often mediated by major cofactors like heme, amino acids and peptides themselves can contribute to molecular conductance. During redox-mediated electron transfer events, intervening residues between redox cofactors are thought to provide a conductive matrix for electron transport, but the structure-function relationship governing their electronic properties is not fully understood. Single-molecule conductance measurements using a scanning-tunneling microscope break junction show an unexpected bimodal distribution for oligopeptide conductance across diverse amino acid sequences. To address this, I develop a series of computational strategies for interrogating the conformational basis of experimentally observed differential peptide conductance, and show that this two-state molecular conductance arose from the conformational flexibility of peptide backbones, with high conductance for defined secondary structures (e.g., a beta turn) and low conductance for extended peptide structures. Classical peptide forcefields were supplemented with custom rationally-guided potentials representing interactions between peptides and gold electrodes. Extensive molecular dynamics (MD) simulations within this scheme allowed for direct structural characterization of the full conformational ensembles of various peptide molecular junctions. Molecular conformations from MD were used to calculate charge transport properties in non-equilibrium Green’s function density functional theory (NEGF-DFT) calculations, confirming our predictions regarding conformational dependence of conductance. Overall, our results provide new avenues for understanding electron transport in longer peptides or proteins with defined secondary structures. Next, I investigate the role played by cofactors like heme on electron transport. Metal-binding proteins have the exceptional ability to facilitate long-range electron transport in biological systems. However, the sequence-structure-function relationships governing electron transport in heme-binding peptides and protein assemblies are not yet fully understood. In this work, the electronic properties of a series of heme-binding peptides inspired by human cytochrome c oxidase are studied using a combination of molecular electronics experiments and molecular modeling. Self-assembled monolayers (SAMs) are prepared by our collaborators using sequence-defined heme-binding peptides capable of forming secondary structures and assembling into coiled-coil helical bundles. Our results show a staggering 1000-fold increase in current density across SAM junctions upon addition of heme compared to identical peptide sequences in the absence of heme or in the presence of metal-free porphyrin, while maintaining a constant junction thickness. Our results show that amino acid composition and sequence directly control enhancements in electron transport in heme-binding peptides. Overall, this work demonstrates the potential of using sequence-defined synthetic peptides inspired by nature as functional bioelectronic materials. Next, I examine the interplay between lipid colocalization and membrane structure that drives the formation of membrane curvature, and create a new tool for interrogating these properties in MD simulations. Curvature is an important feature of biological membranes that plays a diverse role in regulating vital cellular processes including oxidative phosphorylation, vesicle trafficking, and membrane remodeling. Lipid composition can modulate membrane curvature, particularly in the inner mitochondrial membrane, where locally enriched nanodomains of cardiolipin have been shown to spontaneously generate local negative mean curvature. MD simulations of highly curved membranes can give key insights into the interplay between curvature and biology, but there is currently a lack of reliable, high-performance methods for calculating curvature properties of large membranes with many thousands of lipids. More specifically, there is a lack of such methods that calculate both the mean and Gaussian curvatures. We present a grid-free method for calculation of membrane curvature properties that draws upon the generalizability and efficiency of the kernelized support vector machine (SVM). An SVM is trained on a single snapshot of the membrane, with the kernel of the headgroup positions as the training input, and binary leaflet labels as the training labels. Leaflet labels are generated by agglomerative clustering, allowing for automation of the training procedure. The calculated membrane midplane is thus the implicit surface formed by the decision boundary of the SVM: the maximum-margin hypersurface between the headgroups of the two membrane leaflets. This smooth, analytically-differentiable function is then used to calculate mean and Gaussian curvatures by the traditional methods of differential geometry. This representation of the bilayer midplane is naturally extended to provide other information such as membrane thickness and area-per-lipid for curved membranes. We apply our self-supervised learning approach to cardiolipin-containing membranes and demonstrate the Gaussian curvature sensitivity of cardiolipin in highly curved periodic membranes.

Graduation Semester

2024-12

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/127339

Copyright and License Information

© 2024 by Moeen Shamseddin Meigooni. All rights reserved.

Owning Collections