Uncovering the Molecular Basis of Glycosylated Peptide Aggregation

Abstract

Peptide therapeutics often lose their structural stability and biological activity due to uncontrolled self-assembly, which can lead to toxic aggregates. Therefore, improving the stability and solubility of peptides remains a major challenge in peptide engineering. Among the various strategies developed to overcome this limitation, glycosylation has emerged as an effective post-translational modification that modulates peptide structure, intermolecular interactions, and self-assembly. Glycosylation has been one of the most effective methods for reducing the aggregation propensity.[1] However, the phenomenon largely depends on several factors, such as the attachment site and the length of the saccharides.[2]

In this work, we performed extensive molecular dynamics simulations using additive and polarizable force fields to investigate the self-aggregation phenomenon of glycosylated peptides derived from the A-chain of human insulin. The mono and di-glycosylated peptides were constructed by glycosylating at the possible glycosylation site, and compared with the unglycosylated peptide. Aggregation propensity analysis revealed that increasing the degree of glycosylation from mono- to di- significantly reduced peptide aggregation, with di-glycosylation exhibiting the lowest tendency to self-assemble; this holds irrespective of the force field. Although both additive and polarizable simulations suggest a similar glycosylation effect, the morphology of the resulting aggregates differs for the unglycosylated form, which is an interesting observation and likely reflects that the inclusion of electronic polarization alters the aggregate architecture. Our study further reveals that the unglycosylated peptides exhibit frequent aggregation-disaggregation events, indicating the formation of relatively dynamic and reversible aggregates, whereas the event becomes unidirectional upon formation of stable aggregates, with inclusion of polarization. The transition network analysis [3,4] demonstrated that the unglycosylated system exhibited a more interconnected aggregation landscape, with dominant transitions toward highly aggregated intermediates of high population density, indicating a greater propensity for stable oligomer formation. Through this study, we identified key intermediate conformational states formed during aggregation that can be further nullified by selective glycosylation.[5]