Andres Silva, Ignacio Fuentes, Jessenia C Melio, Francisco Parra, Diana Majluf, Diego Rojas, Felipe M Llancalahuen, Maria J Barros, Katina Schinnerling, Fernando Gil, Iván L Calderón, Juan A Fuentes
Outer membrane vesicles (OMVs) are bacterial membrane nanoparticles with applications in host-microbe interaction studies, vaccines, molecular delivery, and microbial biotechnology. However, OMV workflows often rely on ultracentrifugation, limiting accessibility and process integration. Here, we evaluated a laboratory-scale workflow for Salmonella enterica serovar Typhi that sequentially combined culture optimization, TFF-based preconcentration, diafiltration performed in the same TFF system, and lyophilization. We first assessed how culture volume and agitation affected apparent OMV-associated recovery, flagellin carryover, vesicle morphology, and epithelial cell-associated fluorescence. Higher agitation increased protein- and particle-associated recovery readouts but also promoted co-isolation of flagellin-containing filaments, whereas low agitation yielded morphologically better-defined preparations with reduced flagellin carryover and higher cell-associated and trypan blue-resistant fluorescence. TFF and normal-flow filtration were then compared as preconcentration steps before ultracentrifugation and were comparable across morphology, particle size, global protein profiles, and recovery readouts, supporting selection of TFF for workflow integration. During refrigerated, frozen, and lyophilized storage, bulk protein- and lipid-associated signals remained detectable, whereas size distribution, clustering, and surface charge changed over time. Lyophilized preparations retained detectable β-lactamase activity, mCherry fluorescence, and epithelial cell-associated fluorescence after reconstitution. Finally, an integrated ultracentrifugation-free workflow, in which TFF preconcentration was followed by diafiltration in the same TFF system and subsequent lyophilization, preserved multiple OMV-associated structural and functional readouts without ultracentrifugation, including recognizable vesicular morphology, nanoscale particle size, global protein profile, membrane-protected mCherry-FLAG cargo, and detectable fluorescent reporter signal. Compared with ultracentrifugation-based controls, the integrated workflow showed a lower total-protein readout, whereas lipid- and particle-associated readouts were not significantly reduced. These findings define a sequential, ultracentrifugation-free processing framework for research-grade S. Typhi OMVs and support further development of scale-compatible workflows while identifying colloidal stability, cargo-specific retention, and compositional definition as priorities for subsequent validation.