New Insights Into Exosome Biogenesis

Extracellular vesicles (EVs) are small membrane bound vesicles that bud from the plasma and endosomal membranes of nearly all cell types. EVs do not have a defined morphology but have the same topology as the cells they bud from. They have gained prominence over the years because of their involvement in intercellular communication; cells can transmit signals, nuclei acids, proteins, carbohydrates and lipids to other cells in a non-viral pathway via extracellular vesicles. EVs have been implicated in numerous physiological processes such as protein quality control, differentiation, and cell polarity. They have also been shown to play important roles in pathological processes (e.g. cancer, viral infections and neurodegenerative disorders). The size and composition of extracellular vesicles vary significantly, and this has led to a proliferation of names for EVs, including exosomes, microvesicles, oncosomes, ectosomes, membrane particles, etc. Of these, the two most commonly used terms are exosomes and microvesicles. Some have proposed that exosomes be defined as vesicles that are released when multivesicular bodies (MVB) fuse with the plasma membrane, while microvesicles be defined as vesicles that bud directly from the plasma membrane. However, there is little empirical support for these distinctions, and the extent to which differences in EV size and composition are caused by stochastic variations or mechanistic differences is not yet known. Our understanding of organelle biogenesis has been shaped by detailed studies of individual cargo proteins. My thesis work advanced our mechanistic understanding of EV biogenesis using the same type of cargo-based approach. Specifically, in chapter 1, I tested the endosomal hypothesis of the current model of exosome biogenesis in HEK293 cells using the classical exosomal cargoes (CD63, CD81 and CD9). I observed that (i) majority of these proteins are localized on the plasma membrane, (ii) cargo proteins localized on the plasma membrane bud better than those in the endosomes, and lastly (iii) redirecting cargo proteins from the endosome to the plasma membrane enhances their budding significantly. The most parsimonious explanation for these observations is that the majority of protein budding occurs at the plasma membrane of HEK293 and NIH3T3 cells, two of the most commonly used cell lines in biomedical research. In Chapter 2, I characterized the role of certain trans-acting factors that are required for the budding of classical exosomal markers. I showed that loss of tetraspanins, or other putative exosome factors such as alix and syntenin, do not cause a defect in the vesicular secretion of exosomal markers in HEK293 and 293T. Despite no defect in budding, I however noticed variation in the budding of exosome cargo proteins in some mutant clones. To identify the basis of this clonal heterogeneity in exosome biogenesis, I generated dozens of single cell clones (SCCs) from the parental HEK293 cell line and found that they varied by as much as 10-fold in their production of exosomes, demonstrating that the variation we observed in mutant cell lines was not due to the nature of the mutations we introduced or the other Cas9-induced mutations in our SCC lines. We next tested whether clonal heterogeneity in exosome biogenesis was a unique feature of HEK293 cells by generating multiple SCCs from several other human cell lines, again in the absence of any mutagenesis. These experiments revealed that SCCs derived 293T, K562, and Hap1 cells also displayed pronounced clonal heterogeneity. Furthermore, in collaboration with Louise Laurent, we found that SCCs derived from the human stem cell line WA09 also displayed pronounced clonal heterogeneity. We conclude from these results that the extent of exosome biogenesis in human cultured cells is under a complex regulatory pathway that can switch from low to high activity and back as cells are subjected to single cell cloning. Lastly, I described my observations in the differences in exosome composition and production in HEK293 and HEK293T cells. I report an increase in budding of CD63, CD81 and CD9 in HEK293T compared to HEK293, and propose that the increase in budding is a result of the elevated nSMase2 levels observed in HEK293T. The mechanisms for how nSMase2 affects budding of proteins is however not known