Engineering cell membrane vesicles for regenerative medicine

• Main Author: Syed Abdullah Bin Syed Ahmad Alkaff
• Other Authors: Czarny Bertrand Marcel Stanislas
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Materials::Biomaterials
• Online Access: https://hdl.handle.net/10356/170788
• Institution: Nanyang Technological University

Extracellular vesicles (EVs) are intensely investigated at the frontiers of medicine, for use in therapeutics and diagnostics, because of the endogenous origin and secretion by virtually all cells. Particularly, mesenchymal stem cells (MSCs) were reported to produce the most EVs, and these were reported to be safe and efficacious in regenerative therapy. Obstacles remain in generally low yield and high heterogeneity despite developments in the isolation techniques. Biomimetic solutions exist to overcome this problem. Specifically, this thesis proposed cell-derived nanovesicles (CDNs) as suitable substitutes because of similarities in the physical description and chemical composition; and set out to prove this via analysis of biological responses. CDNs were generated by centrifugation of viable cells in spin cups. Cell-based experiments confirmed cellular uptake, migration, proliferation, and expression of markers for growth and development. Tissue-based experiments verified permeation and angiogenesis. Animal trials demonstrated that CDNs were therapeutic in full-thickness skin wounds and that the outcomes of CDNs were equivalent to EVs and cell therapy, but only EVs exhibited signs of systemic influence. Beforehand, proteomic screening had revealed repeatability in the production of CDNs ~78%, with >70% overlap with EVs. Despite the difference in the abundance of proteins related to the nucleus, the proteins found common in biological pathways explained the comparable outcomes in vitro and in vivo, such as in vesicle transport and uptake, and cell survival and growth. Thus, because of the similarities in structure and content to descriptions of EVs, as well as concentration-dependent cellular responses and observations of tissue regeneration, CDNs were concluded to have mimicked corresponding EVs, but not in all the modes of action. The mechanism of biomimicry was tested further as EVs were engineered to acquire a unique second function that CDNs were expected to inherit. Intermediate adipogenic differentiation of stem cells, with subsequent priming using either a Gram-positive or -negative bacterial strain, had generated EVs that were antibacterial against the Gram-negative strain yet therapeutic for re-epithelialization. When stem cells were not differentiated but primed using either one of the bacterial strains, the CDNs produced had developed antibacterial activity without re-epithelialization effects. All other conditions did not show promise of antibacterial effect and were not followed up. The findings indicated an inability to recapitulate the changes caused by stimuli, attributed to the randomness in the production process of CDNs. Yet, the same process allowed the incorporation of compounds within the core of CDNs, which remains a challenge for EVs. The first strategy, remote loading, used a pH gradient to draw in compounds and the second used concurrent extrusion with cells to mix the compound with CDNs. Both remote loading of small drugs and extrusion with nanoparticles were successful as neither caused the final structure of CDNs to deviate from EVs, but studies are needed to test functional similarity.