Extracellular vesicles (EVs) are small nanovesicles secreted from the plasma membrane of cells and are found ubiquitously in bodily fluids. Recently, EVs have attracted much attention as a potential nanotherapeutic due to their ability to modulate cell and biological processes. These functional properties are a result of the synergistic roles of the EV membrane, which contain important signaling proteins and lipids, and the bioactive cargo, which include proteins and nucleic acids. However, the translational potential of EVs is met with several challenges due to difficulties in EV isolation and functional heterogeneity. Presently, bioengineering approaches have been developed to overcome these obstacles by identifying the therapeutically relevant components of EVs and recapitulating these mechanisms in a synthetically engineered bioinspired nanovesicle. This dissertation focuses on identifying biomaterials to mimic the EV membrane, wherein we hypothesized that cell-derived plasma membrane materials will be compositionally similar to EV membranes due to the known biogenesis pathways of EVs. We further leveraged the use of these membranes to synthesize two different classes of bioinspired nanovesicles that can be clinically applied for tissue repair and regenerative medicine.

The first study explored the use of lipid rafts, a subdomain of the plasma membrane, as a membrane biomaterial for exosomes, which are a subclass of EVs specifically deriving from the lipid raft regions. Lipid rafts and EVs were isolated from placental mesenchymal stromal cells (PMSCs) and were found to possess similar lipidomic and proteomic profiles. Lipid rafts were then synthesized into lipid raft nanovesicles (LRNVs) and found to promote neuroregenerative and angiogenic processes. LRNVs could also deliver loaded cargo in a proof-of-concept study, indicating their potential use as tailored biological nanovesicles for disease-specific purposes.

While lipid rafts represent a small section of the plasma membrane and are more relevant for exosomes, other types of membrane-bound EVs (such as microvesicles and apoptotic bodies) are derived from the general regions of the plasma membrane. Thus, in the next study, we characterized the whole plasma membrane isolated from endothelial progenitor cells (EPCs) and synthesized plasma membrane nanovesicles (PMNVs). Proteomic profiling and functional assessment indicated cell-specific biological activity, as EPC membranes showed the ability to mediate hemostasis and angiogenesis due to functional membrane proteins. PMNVs were also able to target different vascular cells, though at different efficiencies, and can potentially be leveraged for future use as targeted nanovesicles for the treatment of various vascular pathologies.

This work further demonstrates that membrane biomaterials can be used in conjunction with more traditional synthetic polymer materials. This allows the development of hybrid nanoparticles that can leverage both the scalability and tunability of polymer materials and the bioactivity of biological membrane materials. In this last study, we developed EPC membrane-coated PLGA nanoparticles (ENPs) for the clinical application of vascular repair following percutaneous coronary interventions (PCI). PCI is commonly used to remove occlusions in the coronary arteries and restore blood flow. However, in doing so, the endothelium that lines the inner surface of the blood vessel is damaged and exposes the underlying collagen. This initiates a cascade of unwanted cellular and inflammatory processes that can ultimately lead to thrombosis, neointimal hyperplasia, and restenosis. ENPs were designed specifically to target collagen, inhibit platelet activity, and promote re-endothelialization. ENPs were conjugated with a collagen-targeting ligand for more specific binding to type I collagen while the EPC plasma membrane coating facilitated decreased platelet adhesion and binding. Proangiogenic microRNA cargo loaded into the PLGA cores facilitated slight increase in re-endothelization, though it is also thought to have contributed to neointimal hyperplasia. These data confirm the ability to engineer a hybrid nanovesicle with promising therapeutic functions, but more work is needed to fully evaluate all potential adverse outcomes and optimize therapeutic effects.

Overall, this dissertation establishes the use of cell-derived membranes as a functionally active biomaterial for the synthesis of EV-inspired nanovesicles. By isolating different types of membranes from various cell sources, disease-specific nanovesicles could be engineered where biological activity was often parent cell-dependent. Furthermore, membranes can be further engineered with targeting ligands or used as drug delivery carriers to enhance therapeutic efficacy. Overall, cell membranes present a new class of biologically active biomaterials and hold high therapeutic potential for future bioengineering and translational strategies.

Mesenchymal stem/stromal cells (MSCs) and MSC-derived extracellular vesicles (MSC-EVs) have been shown to be promising therapeutic agents in regenerative medicine and tissue engineering due to their high potential capacity for repair in tissue injury, as well as for their neuroprotective, proangiogenic, and immunomodulatory abilities. New evidence suggests that MSC-EVs have greater regenerative and clinical application potential compared to the MSCs themselves, due to advantages of the MSC-EVs’ small size and easy homing. However, one major challenge in MSC-EV clinical application is the relatively low production yield of MSC-EVs. The purpose of this study was to determine if regulating fatty acid oxidation (FAO) could boost EV production yield in MSCs. Results have demonstrated that treatment of MSCs with FAO booster chemical, L-carnitine, robustly increased EV production by around two-fold. In addition, the FAO boosted MSC-EVs did not alter their immunomodulatory function regarding the polarization of macrophages to the anti-inflammatory phenotype. Therefore, FAO booster-mediated MSC-EV yield enhancement is a potential approach to meet the need of producing MSC-EVs in high yield for regenerative medicine applications.

Multiple sclerosis (MS) is characterized by an immunological attack of the myelin sheath which leads to demyelination and axonal degradation. This debilitating disorder typically manifests as a progressive loss of motor function, with most cases involving relapsing flare-ups throughout disease progression. The underlying pathophysiology of adult-onset MS is similar to pediatric MS, however pediatric disease typically displays more acute inflammatory responses leading to degeneration. According to the MS society, in the US alone, there are 8,000-10,000 children diagnosed with MS that suffer from frequent relapses. Current therapies for MS reduce the incidence of flare-ups, but do not prevent progressive axonal and neurological degeneration. Cell-based therapies using mesenchymal stem/stromal cells (MSCs) have been under investigation in clinical trials for neurogenerative diseases including adult MS. The placenta has been suggested to be a unique source of MSCs that possess robust immunomodulatory properties and have been reported to be beneficial in graft verse host diseases mouse models. MSCs derived from the placenta (PMSCs) may be a more appropriate cell source for pediatric diseases, because during pregnancy the placenta demonstrates “fetomaternal tolerance”, which is attributed to the expression of human leukocyte antigen-G (HLA-G), a non-classical major histocompatibility complex (MHC) class I molecule that inhibits natural killer cell (NK) killing. Unlike bone marrow derived MSCs (BM-MSCs), PMSCs express HLA-G on their surface in response to interferon gamma (IFN), which is a key inflammatory mediator involved with the onset of MS. Therefore, the expression of HLA-G on PMSCs would make them a unique therapeutic cell source for the treatment of autoimmune diseases like pediatric MS. Currently, a clinical trial is underway using term placenta-derived PMSCs for adult MS and no paradoxical worsening of MS lesion counts was noted.

Notably cell-based therapies are limited by potential immune rejection of donor cells and other safety concerns. Increasingly, studies have shown that MSC survival and integration within the host after transplantation are usually poor and that MSCs exert their therapeutic functions mainly via paracrine signaling mechanisms. Conditioned media of BM-MSCs is shown to protect neurons from apoptosis, activate macrophages and be pro-angiogenic. In particular, hepatocyte growth factor (HGF) secreted by MSCs into conditioned medium mediates recovery in a murine model of MS. Currently, the use of MSC conditioned media is limited in that the secreted protein factors are unstable, which creates technical difficulties for “off-the-shelf” clinical use. MSC derived extracellular vesicles (EVs) alternatively, are stable under long term storage conditions compared to freely secreted proteins and may serve as a superior source for cell-free therapy.

MSC-derived EVs readily cross the blood brain barrier (BBB) and deliver therapeutic cargo to reduce the effects of neuropathologic diseases, such as MS. It has been demonstrated that MSC-derived EVs exhibit systemic immunomodulatory effects and can facilitate neurological recovery in vitro. PMSC-EVs contain numerous proangiogenic, immunomodulatory, and neuroprotective proteins, including HGF. HGF is a pleiotropic factor shown to have neuronal and oligodendrocyte protective properties. In an experimental autoimmune encephalomyelitis (EAE) rodent model of MS, overexpression of HGF by neurons conferred neuroprotection by reducing inflammation in the CNS and activation of Tregs. HGF is secreted both in soluble form from MSCs and is also contained in exosomes; however, the effects of each form could lead to different cellular responses. It has been suggested that MSCs secrete multiple categories of exosomes, which are involved in differing cellular processes. Therefore, deciphering the molecular mechanism by which MSC-derived EVs alleviate the effects of neurodegenerative diseases is warranted. The goal of the current research project is to demonstrate the therapeutic potential of PMSC-EVs utilizing translational models of MS as well as to and to examine the role of HGF. These studies will be crucial as pre- investigational new drug data moving towards novel and bioengineered cell-free therapeutic treatments for MS patients.

Demyelinating disorders, including multiple sclerosis and Canavan disease, afflict millions worldwide, yet their underlying mechanisms remain largely elusive. Our study on Canavan disease utilized single-cell spatial-temporal transcriptomics and mass spectrometry imaging. This approach enabled us to profile 37,083 genes at 500 nm resolution and 85 lipids at 10 μm resolution, offering detailed insights into the dysregulated genes and metabolites affecting signal transduction, osmotic homeostasis, vision, auditory function, and circadian rhythm. The Canavan brain lacks aspartoacylase; therefore, N-acetylaspartate (NAA) cannot be utilized to generate acetyl-CoA, an essential precursor for lipid synthesis. Our data demonstrate that the alternate route for acetyl-CoA production, mediated by ATP:citrate lyase in citrate cleavage, is also impaired, thus contributing to the deficit in myelin lipid synthesis. Additionally, this impairment results in a broad attenuation of gene expression due to histone hypoacetylation and disturbances in the tricarboxylic acid (TCA) cycle. Such disruptions could manifest as increased oxidative stress, widespread reduction in protein synthesis, and decreased ATP production. Notably, our research also revealed that oligodendrocytes are more abundant but less communicative with neurons in Canavan disease. To restore their function at an early stage, we delivered adeno-associated viruses (AAVs) intracerebroventricularly to fetal mice. These AAVs were engineered to express the ASPA gene in oligodendrocytes, significantly alleviating disease symptoms. Additionally, we introduced two new methods, BrainSegmenter and TransMetaSegmentation (TMS), to address challenges in brain cell segmentation and metabolite allocation, respectively. We anticipate that this study will advance the understanding and treatment of neurological and genetic disorders.

Spinal cord injury (SCI) results in a wide range of clinical manifestations varying in severity from sensory deficits to irreversible paralysis, with approximately 12,500 new cases occurring every year in the US1. Recently, stem cell-based therapy, specifically with placenta-derived mesenchymal stem/stromal cells (PMSCs), has been suggested as a promising approach for the treatment of SCI, as PMSCs can secrete multifunctional therapeutic factors for remyelination, neuroprotection, and immunomodulation2–4. However, many studies have also shown poor PMSC survival and integration within the host after transplantation, suggesting that PMSCs exert their therapeutic functions mainly via a paracrine mechanism5,6. PMSCs secrete numerous factors that orchestrate key biological functions and cell behaviors including a significant number of extracellular vesicles (EVs)7. EVs are small membranous vesicles, ranging from 50-150nm in diameter, derived from the cell’s plasma membrane or endosomes and recognized to play an important role in long range cell-cell communication transporting various functional molecules, including proteins, lipids, microRNAs, and mRNAs8. Compared to cell-based therapy, EVs display long term storage stability, reduced immune rejection, and natural targeting surface markers that allow them to cross the blood brain barrier (BBB), a major obstacle for most cell-based and drug therapies for central nervous system (CNS) disease and injury9. Although EVs pose a promising alternative to cell-based therapy, targeted delivery in vivo – especially to the CNS - is lacking10,11. To correct for this limitation, researchers have modified their surface to endow them with active targeting molecules to enable specific cell uptake and tailor EV biodistribution12–16. A dominant paradigm has been to evaluate the EV surface functionalization using bulk analysis assays, such as western blotting and bead-based flow cytometry17,18. However, to our knowledge, there are no standardized practices to confirm the successful surface modification of EVs or calculate the degree of conjugation on EV surfaces (conjugation efficiency). Considering the many single-nanovesicular analysis technologies developed, the potential to inform the optimization of conjugation and to correlate the degree of surface conjugation required to see therapeutic effects in vivo is endless. Thus, the objectives of this dissertation were to 1) find the proper channels through which to confirm and calculate EV conjugation efficiency after surface modification, 2) use the identified methods to optimize conjugation of EVs for the improved targeting ability, and 3) test the optimized surface-modified EVs in relevant in vitro and ex vivo assays within the context of developing a targeted SCI therapy. First, we conducted a literature review to explore the bulk and single nanoparticle analysis techniques used to confirm that EVs or nanoparticles were successfully conjugated with a desired targeting molecule or fluorescent tag19. Then, we identified the reported conjugation efficiency calculation methods to figure out either 1) the % EV population modified with said molecules or 2) the estimated number of molecules modified on each EV. Then, we engineered the EV surface at the single-vesicle level. We applied orthogonal platforms with single vesicle resolution to determine and optimize the efficiency of conjugating the myelin-targeting aptamer LJM-3064 to single EVs (Apt-EVs). The aptamers were conjugated using either lipid insertion or covalent protein modification, followed by an assessment of single-EV integrity and stability. We observed unique aptamer conjugation to single EVs that depend on EV size. Finally, we tested the collective myelin-targeting, oligodendrocyte (OL) protection, and regenerative properties of the optimized Apt-EVs using relevant in vitro and ex vivo models. We used optimized lipid insertion-aided conjugation strategies to modify PMSC-EVs and observed 1) improved mature OL targeting, 2) improved rescue of Ols under oxidative stress, and 3) minimally altered intrinsic PMSC-EV functions (neuroprotection and immunomodulation) post-modification through various well-established in vitro assays. We also observed improved uptake in myelin-expressing oligodendrocytes compared to neurons in an ex vivo rat brain tissue model. This work study underscores the importance of use of single vesicle analysis technology for surface engineering EVs and provides a novel single-EV-based framework for informing optimized EV surface modification. This work is also essential in the study to effectively deliver therapeutic EVs to CNS injury sites to decrease off-target effects and increase treatment efficacy.

Placenta mesenchymal stem cell (PMSC) treatment has been investigated to repair the spina bifida defect in various animal models, including in an English Bulldog model and in the gold-standard fetal lamb model of spina bifida. Furthermore, these cells are currently undergoing a Phase 1/2a clinical trial to assess their safety and preliminary efficacy in humans. For these cells to move forward into Phase 2 and Phase 3 clinical trials, Biologics License Applications (BLA) and commercialization as a biological drug, they must be manufactured and characterized according to FDA guidelines and must be validated through potency assay(s). An industrial standard potency assay must contain replicates of the reference standard, samples, and positive and/or negative controls. The replicates must be in triplicate and serially diluted into 8 or 12 concentrations that would fit down the columns or across the rows of a 96-well plate or a 394-well plate. The range of concentration is determined by the linear range of the assay, which is part of the sample acceptance criteria. The assay must also have a validated assay (system) acceptance criteria and an appropriate analysis method. Currently, an in vitro neuroprotection assay in the 12-well format developed by Kumar et al. has been validated as a screening method for identifying PMSC lines. My thesis is aimed to further refine the parameters and develop a high-throughput, industrial standard potency assay for screening PMSCs. During development, the cell seeding densities for SH-SY5Y and ENStem-A cell line were optimized for the 96-well format. Extracellular matrix coating was also examined for ENStem-A optimal culturing conditions. ImageXpress PICO Automated Cell Imaging System/Neurite Tracing Image Analysis was compared and determined as a better system than the system used for the 12-well format, as this system was easier to use, more accurate, took less time to complete and was automated for high throughput assay. Two possible mechanisms of PMSC protection were elucidated, neurogenesis preservation and neuroprotection. Therefore, two assays were developed and tested with undifferentiated and differentiated SH-SY5Y cells. However, the results were inconclusive, lacking dosage responses. There were differences between the control and the PMSC-rescued samples. There were also differences between the PMSC dosages, but the differences did not produce a linear trend. The discrepancy could be due to the biology of undifferentiated SH-SY5Y as a neuroblastoma line. Plate setting, acquisition time, intensity and image analysis parameters could also be factors accounting for the inconsistency. Further optimization of differentiated SH-SY5Y is needed to achieve 100% mature neurons and specificity for the neuroprotection mechanism. Replacing the live-cell imaging procedure with fixed cells followed by immunocytochemistry (ICC) would preserve cell morphology, prevent cells from lifting off and capture fragile neurite processes by formaldehyde cross-linking, allow multiple acquisitions and analysis with the same sample/plate. Fixed cells/ICC would also aid in establishing the assay acceptance criteria such as plate setting, acquisition time, intensity, and image analysis parameters.