Impact of L- and D-Peptide Isomer Mixtures on Self-Assembly in Tissue Engineering: A Molecular Dynamics Study

Tissue regeneration is a critical challenge that necessitates the design of dynamic materials. Such design requires the material to mimic the extracellular matrix (ECM), by understanding features such as proteolytic remodeling, cell adhesion, and viscoelastic properties. Recent studies have focused on the design of ECM-mimetic, water-swollen hydrogels, due to the flexibility of their mechanical and biochemical properties. Specifically, our group has previously developed Rapidly Assembling Pentapeptides for Injectable Delivery (RAPID) hydrogels, composed of the peptide sequence KYFIL (lysine–K, tyrosine–Y, phenylalanine–F, isoleucine–I, and leucine–L), and has used turbidity, oscillatory rheology, and X-ray diffraction to examine the effects of amino acid stereochemistry (i.e., ʟ- and ᴅ-amino acids) on peptide self-assembly and hydrogel mechanical properties. We used atomistic molecular dynamic simulations to probe the molecular mechanism governing the stereocomplexation-directed self-assembly of KYFIL. Five peptide systems with varying ʟ:ᴅ ratios, solvated with physiological ion concentrations, were conducted under constant pressure and temperature using NAMD and GROMACS software. Simulations trajectory files revealed that pure ʟ- and ᴅ-KYFIL systems formed larger but fewer clusters of self-assembled peptides, while stereoisomer mixtures formed a larger number of smaller clusters. Hydrogen bonding and pi-pi stacking had similar results with the ʟ- and ᴅ- mixtures and pure peptide systems. Lastly, secondary structure analysis indicated the formation of β sheets and random coils in all systems. Overall, varying stereochemical ratios significantly impacted the assembly mechanism of KYFIL systems, influencing the size of the cluster, secondary structure, and hydrogen bonding, in qualitative agreement with previous experimental observations. Future investigations will include coarse-grained simulations utilizing GROMACS and MARTINI, to simulate hydrogels at larger length scales and longer timescales.