Ultracentrifugation and Ultrafiltration Differentially Alter the Composition and Functionality of the Biomolecular Corona of Extracellular Vesicles

ABSTRACT While extracellular vesicles (EVs) are increasingly recognized for their diagnostic and therapeutic potential, many unknowns remain. An emerging and important concept in EV biology is the EV biomolecular corona, a dynamic layer of absorbed biomolecules to the EV membrane that influences EV behaviour, cellular uptake, biodistribution and function. In de‐coronisation studies, a key challenge involves identifying effective purification methods to selectively and reliably remove enzymatically digested surface proteins. This study assesses the effectiveness of ultracentrifugation (UC) and ultrafiltration (UF) purification techniques following proteinase K (PK) treatment of pre‐isolated natural killer cell‐derived EVs (NK‐EVs). The efficiency and specificity/selectivity of these purification methods in removing cleaved proteins while preserving EV physical properties (size), biomolecular profile (surface immunophenotyping, cytokine and proteomic profiling) and cytotoxicity against cancer cells were evaluated. While both methods achieve comparable NK‐EV particle recovery, only UF purification selectively removed PK‐cleaved proteins, as depicted in the protein‐to‐particle ratio, biomolecular profiling and NK‐EV cytotoxicity after PK treatment. In contrast, UC purification caused non‐specific corona disruption, dramatically reducing the cytokine payloads and abolishing NK‐EV cytotoxicity. Interestingly, NTA and transmission electron microscopy (TEM) (positive staining) analyses showed that de‐coronised EVs were smaller than their untreated counterpart. This finding suggests that PK treatment impacts surface‐associated proteins with different sensitivity to proteolysis, an effect that remained consistent independent of the subsequent purification methods. Collectively, these results highlight UF as the purification method of choice for controlled de‐coronisation studies, potentially supporting the advancements of EV research across multi‐omics, investigatory, and preclinical applications.