Peptides self-assemble into β-sheet nanostructures that are central to both amyloid disease and advanced biomaterial design. The canonical pleated β-sheet, with its characteristic cross-strand eclipsed side chain geometry, dominates natural systems and most synthetic assemblies. Yet in 1953, Pauling and Corey predicted a second fold: the rippled β-sheet. This structure requires alternating enantiomeric l- and d-peptide strands, producing a staggered side chain conformation that creates a zigzag or rippled pattern. Despite being theoretically sound, the rippled β-sheet has remained largely unexplored for over seventy years. Recent crystallographic evidence and assembly studies have begun validating the Pauling and Corey prediction, but the sequence scope and thermodynamic drivers of rippled versus pleated β-sheet formation remain poorly understood.
Researchers in the Nilsson Group at the University of Rochester, published in Biochemistry, investigated how amino acid sequence order influences rippled β-sheet assembly by studying enantiomeric mixtures of five isomeric amphipathic peptides: Ac-(FKFE)2-NH2, L1, Ac-(FK)2(FE)2-NH2, L2, Ac-KE(F)4KE-NH2, L3, Ac-(KFFE)2-NH2, L4, and Ac-FF(KE)2FF-NH2, L5. These sequences share identical amino acid composition but differ in the arrangement of hydrophobic phenylalanine and hydrophilic lysine and glutamate residues. The team had previously demonstrated that these sequence variations profoundly affect pleated β-sheet self-assembly propensity, with L1 and L2 assembling strongly, L3 showing intermediate behavior, and L4 and L5 assembling weakly. The researchers prepared equimolar mixtures of each l-peptide with its corresponding d-enantiomer and compared assembly thermodynamics to single enantiomer systems using Fourier transform infrared spectroscopy, circular dichroism, transmission electron microscopy, and molecular dynamics simulations.
Critical aggregation concentration measurements revealed that all five enantiomeric mixtures coassemble into putative rippled β-sheets at significantly lower concentrations than their single enantiomer counterparts self-assemble into pleated β-sheets. The L1/D1 mixture exhibited a critical concentration of 0.032 millimolar compared to 0.100 millimolar for L1 alone, corresponding to a thermodynamically favorable free energy difference of 0.65 kilocalories per mole favoring rippled β-sheet formation. This trend held across all sequences tested. L2/D2 showed a critical concentration of 0.026 millimolar versus 0.067 millimolar for L2, with a free energy advantage of 0.55 kilocalories per mole. Even sequences with weak pleated β-sheet assembly propensity demonstrated enhanced coassembly in enantiomeric mixtures. L3/D3 assembled at 0.228 millimolar compared to 1.267 millimolar for L3 alone, representing a substantial 1.01 kilocalorie per mole favorable energy change. Most striking were L4/D4 and L5/D5, which formed rippled β-sheets at 0.394 and 0.339 millimolar respectively, while L4 and L5 showed minimal pleated β-sheet formation even at four millimolar concentration. Molecular dynamics simulations at 350 Kelvin confirmed these experimental observations, revealing that enantiomeric mixtures rapidly organize into cross-β structures with alternating l- and d-strand architecture characteristic of rippled β-sheets.
These findings establish that rippled β-sheet formation from enantiomeric peptide mixtures is thermodynamically favored over pleated β-sheet formation across diverse sequence contexts, not just for ideally alternating amphipathic sequences. The universal enhancement of assembly propensity in racemic mixtures suggests that stereochemical diversity can be exploited to tune biomaterial properties including mechanical strength, proteolytic stability, and gelation kinetics. This work significantly expands the design space for peptide nanomaterials by demonstrating that sequence patterns governing pleated β-sheet assembly do not strictly predict rippled β-sheet behavior, opening new avenues for engineering supramolecular architectures with tailored function.
