Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article


Master of Science




Lagugné-Labarthet, François


Extracellular vesicles (EVs), which are nanoscale vesicles secreted by cells into biofluids, are of research interest due to their roles in intercellular communication. EVs released from mesenchymal stromal cells (MSCs) have tremendous potential in cell-free regenerative medicine, while EVs released from diseased cells are being studied as biomarkers for minimally invasive and early disease detection. Presented in this thesis are gold nanohole arrays for the capture and sensitive detection of EVs by surface-enhanced Raman spectroscopy (SERS), a plasmonic technique capable of single molecule detection. Herein, we have characterized EVs released from MSCs and ovarian cancer cells, with a focus on cell lines that have been underexplored by SERS in literature. Using a hybrid principal component analysis-machine learning approach, we have demonstrated the platform’s potential in classifying EV groups with high (~ 99 %) accuracy, sensitivity, and specificity, which we hope will one day translate to point-of-care detection for disease diagnosis.

Summary for Lay Audience

The fields of rehabilitative and diagnostic medicine are constantly evolving, where the former is seeking safer and more effective ways to repair tissue and organ damage, and the latter is developing methods for rapid, non-invasive, and early disease detection. One area of research with applications in both these fields is extracellular vesicle (EV)-based technology. EVs are a complex group of membrane-bound vesicles released from cells into biofluids including blood, saliva, and urine. EVs are traditionally separated into three subclasses based on attributes such as size, biomolecular cargo, and mechanisms of formation and release. The most interesting subclasses of EVs consist of exosomes and microvesicles since they play roles in intercellular communication. Released from mesenchymal stromal cells (MSCs), EVs have shown immense promise in cell-free regenerative and restorative applications, while EVs released from diseased cells (e.g., cancer cells) are studied for their applications as disease biomarkers. However, the nanoscale size and molecular heterogeneity of EVs pose a significant research problem since sensitive methods are required for their detection.

Surface-enhanced Raman spectroscopy (SERS) is a sensitive, non-destructive plasmonic technique that has shown potential in biosensing applications, including EV detection. The highly sensitive nature of SERS is based on the collective oscillation of free electrons at a nanoscale, metallic surface. Consequently, large electromagnetic fields are produced at the edges of the nanoscale features, and analytes that are confined to these regions experience significant enhancement with respect to their Raman signal intensity. Proposed in this work are gold nanohole arrays for the capture and SERS characterization of EVs. The SERS spectra gathered provide insight into the biochemical composition of EVs, and EVs from both MSCs and ovarian cancer cells are explored. SERS characterization of EVs from the specific cell sources investigated in this thesis have been largely underexplored in literature, and so the work presented here is novel. Statistical analysis is utilized to find patterns in the complex spectra acquired, and machine learning is further implemented to classify EVs from various cell sources. The high (~ 99 %) accuracies, sensitivities, and specificities reported in this thesis demonstrate great promise for translation to clinical testing.

Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License