Identification of disease-associated RNAs in extracellular vesicles from saliva and plasma

Abstract

While the biopsy of specimens can offer direct, accurate information on diseases, it causes severe pain and inconvenience to patients. In contrast, the collection of body fluids is relatively comfortable to patients, and they contain a number of potential diagnostic factors including extracellular vesicles (EVs). EVs are bi-layered structures that contain molecules from cells and carry their contents to other locations by travelling through body fluids. Accordingly, the analysis of EVs can provide useful information regarding the state of cells, tissues, or organs. Among the types of molecules in EVs, this thesis focuses on RNAs with an aim of diagnostic application. As other body fluids, saliva is known to reflect the overall bodily health, and its role is not confined to oral health. For example, the levels of steroid hormones, viral antigens, and some disease-associated proteins in saliva were revealed to correlate with systemic conditions. However, in the case of RNA of systemic origin, its presence in saliva has been questioned due to the relative instability of RNA. Since EVs provide protection to their RNA contents by membranous structures, the early part of this thesis investigates EVs from saliva to search for melan-A mRNA, a systemic disease-derived mRNA, based on the assumption that RNA can be transferred from blood to saliva via EVs. In order to identify melan-A mRNA, human melanoma cells are implanted into nine immunodeficient nude mice. Six immunodeficient mice which are not injected with human melanoma cells are regarded as controls. Five to six weeks after the injection, saliva and plasma are collected from the mice, and RNAs in EVs from saliva and plasma are analyzed for melan-A RNA by RT-qPCR and the relative standard curve method. The result shows that there are melanoma mice with higher normalized melan-A value than control mice at fifth week or sixth week. Since the human melan-A sequence is not present in mouse species, and its expression is not identified in the control mice, this result suggests that human melan-A RNA, which is related to melanoma on skin and systemically originated, can be detected in saliva. Micro-RNAs, 15 to 25 nucleotides in length, are known to relate to the progress of various diseases. In the later part of this thesis, the levels of prostate cancer (PCa)-related micro-RNAs, miR-21, miR-141, and miR-221, in EVs from plasma and saliva are studied. Plasma is collected from total 46 patients: 8 patients with benign prostatic hyperplasia (BPH), 15 patients with localized PCa and 16 patients with local advanced PCa, and 7 patients with metastasized PCa. Saliva is collected from total 33 patients: 10 patients with BPH, 16 patients with localized PCa, and 7 patients with local advanced PCa. Samples from these two body fluids are not obtained from the same patients and independent of each other. According to the results, for miR-221 in EVs from plasma, the localized PCa group shows 17.08-fold higher expression than BPH with significance (p < 0.01). In the case of miR-141, even though the metastasized PCa group illustrates 13.58-fold higher expression than localized PCa (p< 0.05) and 7.60-fold higher expression than local advanced PCa group (p< 0.05) in plasma, its level is not significantly different between the BPH group and the localized PCa group. For the level of prostate specific antigen (PSA) in serum, the metastasized PCa group shows a higher expression than BPH group (p <0.001), localized PCa group (p <0.001), and local advanced PCa group (p <0.001), while there is no significant difference between the BPH group and the localized PCa group. Additionally, the AUC value of miR-221 is 0.98, which is higher than those of PSA and miR-141. Since it is crucial to discover PCa during its early stage of pathogenesis, this result suggests that miR-21 in EVs from plasma has its potential in the diagnostic application. In saliva, for miR-21, only the local advanced PCa group shows a significantly different expression from the localized PCa group (p > 0.05), but for miR-221 and miR-141, there is no group that shows a significant change in the expression level each other. The AUC values of miR-21, miR-141, and miR-221 in saliva are lower than the AUC value of PSA. Taken together, it is concluded that EVs from saliva can include RNA of systemic origin, but selected RNAs in this study do not reveal their diagnostic values in EVs from saliva with respect to a systemic disease, PCa. On the other hand, RNA in EVs from plasma shows a better potential in the diagnostic application: a type of RNA in EVs, miR-221, reveals its value in distinguishing the early stage of PCa, which could not be achieved by the level of PSA in serum.