Exploring Membrane Binding Targets of Disordered Human Tau Aggregates on Lipid Rafts Using Multiscale Molecular Dynamics Simulations

Abstract

The self-aggregation of tau, a microtubule-binding protein, has been linked to the onset of Alzheimer’s Disease. Recent studies indicate that the disordered tau aggregates, or oligomers, are more toxic than the ordered fibrils found in the intracellular neurofibrillary tangles of tau. At present, details of tau oligomer interactions with lipid rafts, a model of neuronal membranes, are not known. Using molecular dynamics simulations, the lipid-binding events, membrane-damage, and protein folding of tau oligomers on various lipid raft surfaces were investigated. Tau oligomers preferred to bind to the boundary domains (Lod) created by the coexisting liquid-ordered (Lo) and liquid-disordered (Ld) domains in the lipid rafts. Additionally, stronger binding of tau oligomers to the ganglioside (GM1) and phosphatidylserine (PS) domains, and subsequent protein-induced lipid chain order disruption and beta-sheet formation were detected. Our results suggest that GM1 and PS domains, located exclusively in the outer and inner leaflets, respectively, of the neuronal membranes, are specific membrane domain targets, whereas the Lod domains are non-specific targets, of tau oligomers binding to neurons. The molecular details of these specific and non-specific tau bindings to lipid rafts may provide new insights into understanding membrane-associated tauopathies leading to Alzheimer’s Disease.

Recent Advances in Molecular Dynamics Simulations of Tau Fibrils and Oligomers

“…Cheng et. al investigated the lipid-binding events, membrane damage, and protein-folding of tau oligomers using the K-18 construct on different lipid-raft surfaces using molecular dynamics simulations [ 38 ]. CG monomers were created from the corresponding AA structures of the peptide using the AA to CG resolution transformation program martinize.py based on the Martini 2.20 CG force fields [ 35 ].…”

“…CG monomers were created from the corresponding AA structures of the peptide using the AA to CG resolution transformation program martinize.py based on the Martini 2.20 CG force fields [ 35 ]. The CG monomers were solvated in 0.1 M NaCl at 310 K and 1 atm pressure and subjected to energy minimization and positional constraint to reduce the high-energy local structure formed during the solvation steps [ 38 ]. The generation of a 130-residue-long all-atom tau K18 monomer was based on a tau fibril structure generated with Cryo-EM (PDB ID:5O3L) [ 3 , 20 ].…”

“…The 73-residue-long peptide Tau 308-372 was first extracted from chain A of the pentamer structure ( Figure 8 a). Then, a 57-residue-long random coil Tau 243-307 was attached to the N-terminus of Tau 308-372 using a homology modeling algorithm to generate the final 130-residue-long Tau 243-372 or wild-type Tau-K18 (WT-K18) monomer ( Figure 8 b) [ 38 ]. Three strongly hydrophobic residues at V287, I308, and V318 were replaced by three negative residues at E287, E308, and E318 to generate the mutant tau-K18 membrane-binding-deficient (MBD) ( Figure 8 c).…”

“…Three strongly hydrophobic residues at V287, I308, and V318 were replaced by three negative residues at E287, E308, and E318 to generate the mutant tau-K18 membrane-binding-deficient (MBD) ( Figure 8 c). The primary and AA structures of WT-K18 and MBD-K18 both show mutations at V287E, I308E, and V318E [ 38 ]. After the CG simulations of the binding between the oligomer and lipid raft, each replicate of the CG tetramer-raft system was converted to the AA structure using backward.py [ 39 ], a program for converting the resolution of CG to AA.…”

“…After the CG simulations of the binding between the oligomer and lipid raft, each replicate of the CG tetramer-raft system was converted to the AA structure using backward.py [ 39 ], a program for converting the resolution of CG to AA. The same equilibration was performed with similar energy minimization and position constraint procedures as in the simulations of the CG oligomer-force complexes [ 38 ]. The atomistic force fields AMBER99SB [ 40 ] for proteins and SLIPID [ 41 ] for lipids were used for the AA simulations.…”