Peptide glasses are an emerging class of biofunctional amorphous materials, but their atomic-level structure has remained elusive. Here, we resolve the three-dimensional (3D) organization of peptide glasses using a cyclic dipeptide model, combining molecular dynamics simulations with 2D solid-state NMR fingerprinting. Our analysis quantifies three defining hallmarks of the glassy state: (i) pronounced conformational heterogeneity that distinguishes it from the crystalline state; (ii) reorganization of diverse H-bonding types producing annealing-rate-dependent spectral fingerprints; and (iii) dominance of non-hydrogen-bonded contacts that generate annealing - temperature-dependent spectroscopic signatures. These features give rise to molecular clusters with a branched H-bonding topology that quantitatively reproduce bulk spectroscopic properties, establishing a representative structural unit analogous to the crystalline unit cell. This framework enables the identification of structural organization in amorphous peptide glasses with different thermal histories, paving the way for rational design of functional small-molecule glassy materials.
The formation of fibrous architectures via peptide self-assembly underpins numerous biological functions and biomaterial applications; however, the thermodynamic origins of multistep assembly pathways remain elusive. Here, we map the complete free-energy landscape governing the liquid-liquid phase separation (LLPS)-mediated self-assembly of an amphiphilic peptide by exploiting temperature as a tunable parameter. We discover an unexpected thermodynamic mechanism: the initial LLPS-like clustering is enthalpy-driven but limited by a positive enthalpic barrier (+121 kJ mol-1), arising from the endothermic disruption of intramolecular hydrogen bonds before interpeptide contacts can form. Subsequent nucleation and fibril growth are governed by negative entropic barriers (-56 and -39 kJ mol-1, respectively), reflecting the reorganization cost of partially ordered oligomers. The energy landscape identifies LLPS as the rate-limiting step with the highest Gibbs free-energy barrier (+26 kJ mol-1). Our findings establish a generalizable framework for decoding multistep biomolecular self-organization, with implications for designing adaptive biomaterials and understanding aberrant phase transitions in diseases.
Yufan Yang, Haoning Gong, Peng Zhou et al.· Journal of Physical Chemistr...· 0 citations