In this Chapter, we used purified DU145 exosomes as a basis to characterise vesicles for future experiments, to determine whether silencing putative regulators of exosome biogenesis and secretion will impact vesicle secretion.
Here, we showed that the prostate cancer cell line, DU145, was confirmed to have a cobble- stone like morphology (Stone et al., 1978), similarly shown in other studies with prostate cancer cells (Hayward et al., 2001, Alimirah et al., 2006). Exosomes from DU145 cells were isolated using the sucrose cushion method and revealed circular structures, demonstrating the presence of a lipid bilayer. The DU145 exosomes were a heterogeneous population, in which the size of vesicles ranged between 30 – 150nm in diameter, in agreement with NTA analysis. The NTA enabled size distribution measurements of nanoparticles in fluid phase, as opposed to dehydrated and fixed exosome samples used in traditional (non-cryo) EM. This observation may give a greater representation of exosomes, as they are in their native state. The NTA data indicated a broad nanoparticle size ranging up to 500nm for purified DU145 exosomes, despite passing through 220nm-pore filters. The purity of this method of isolating exosomes has been previously questioned as high molecular weight complexes and apoptotic blebs may co-sediment with these exosomes. Though, other studies have demonstrated the purity of these isolated exosomes contain minimal contaminating organelles (Mitchell et al., 2008, Webber and Clayton, 2013, Jeppesen et al., 2014). In contrast, we observed fewer size distribution peaks that are >300nm in size with nanoparticles in cell CM as detected by NTA. It could be possible that cell CM is less concentrated with EVs that may not co-isolate as many mixed EV populations. Both NTA and cryo-EM gave evidence that nanoparticles greater than >500nm were not present, suggesting the absence of apoptotic bodies or large organelles. Both these methods provide an indication of the heterogeneity present within samples that can characterise individual exosomes. NTA alone cannot distinguish EVs from co-isolated, non-membranous particles of similar size and should be compared with cryo-EM to acquire wide-field images of exosomes in question. Other methods to determine the presence of exosome-associated proteins must be considered to ascertain the isolated samples are in fact exosomes.
Typically, a general overview of the protein composition of each exosome preparation should be provided in a semi-quantitative manner, with exosomal-associated proteins expected to be present and components not necessarily expected. Although numerous proteomics studies have highlighted proteins commonly found in exosome preparations, it is becoming
90 clear that these do not represent exosome ‘specific’ markers, but rather exosome ‘enriched’ proteins (Mathivanan and Simpson, 2009, Kalra et al., 2012). Within the literature, the relative proportion of exosomal-associated proteins varies with different EVs or cell types. It has been proposed that several proteins (three or more) in a semi-quantitative manner should be reported to characterise EVs (Lotvall et al., 2014). The expression of these tetraspanins: CD9, CD63, CD81 were detected on our purified DU145 exosomes, commonly found on the surface of exosomes from various cell types (Raposo et al., 1996, Escola et al., 1998, Lamparski et al., 2002). These findings were similar to other studies with prostate and bladder cancer cells (Zoller, 2009, Welton et al., 2010). The microplate immune-phenotype assay and Western blot analysis revealed DU145 exosomes to be enriched with LAMP1 and LAMP2. The LAMP proteins are markers for lysosomes and could reflect the late endosomes containing ILVs, although the presence of LAMPs varies for most cell types. Some studies only demonstrate the presence of LAMP proteins in exosomes, as reported in tumour cells (Wolfers et al., 2001), but absent in DC-derived exosomes (Zitvogel et al., 1998). DU145 exosomes also showed an enrichment of HSP70 and HSP90. These proteins are involved in antigen presentation and participate in loading peptides to MHC molecules (Srivastava, 2002). Here, MHC-I was also present on exosomes and it is normally present in exosomes from most cell types. In general, the level of cellular contamination present within the exosome preparation was minimal, as the calnexin protein (ER marker) was absent in exosome samples. Within these DU145 exosomes, the presence of CD9, CD63, CD81, ALIX, TSG101, LAMP1, LAMP2, HSP70 and HSP90, in combination of the absence of calnexin, strongly suggests that the studied vesicles are exosomes.
One other approach to determine sample purity of exosome preparations is by determining the levels of particle to protein (P:P) ratio (Webber and Clayton, 2013). It has been shown that vesicle preparation that are considered pure exhibit a relatively high ratio of particles to protein and thus contaminating protein within samples should have a negative effect on this ratio. Typically, an arbitrary P:P ratio of <1 x 1010 would be considered an impure isolated
vesicle sample. From the data, isolated DU145 exosomes from the sucrose cushion method as one example, demonstrated a P:P ratio of 1.11 x 1010 compared to cell CM with a ratio of
1.78 x 108. This suggests the exosome preparation from the sucrose cushion would be
deemed purer. In contrast, analysing cell CM indicates an impure exosome preparation, possibly indicating a greater level of soluble proteins present. Applying the ratio method provides a relatively simple and quantitative manner to estimate and compare purity. Though, there are some caveats using the P:P ratio as a sole method to analyse exosomes.
91 There it is difficulty in discriminating vesicles from non-vesicular particular material, there is an assumption that all detected particles are vesicles and there could be an overestimation of particles present within the sample. Furthermore, there is an assumption that each vesicle has a comparable and stable quantity of protein; though different disease states may alter the protein content to some degree, but this remains poorly understood. Nevertheless, bearing issues in mind, the P:P ratio method does provide an additional method to determine the purity of the vesicle preparation for future experiments.
Other methods to isolate exosomes have claimed to isolate pure exosomes from cell CM, such the commercially available ExoQuick® (EQ) or Total Exosome Isolation Reagent (TEI). Though, the main caveat is that these reagents co-precipitates exosomes with non-vesicular proteins, leading to a low yield of impure exosomes (Van Deun et al., 2014). It has been shown that isolating vesicles based only on ultracentrifugation can retain a relatively moderate pure sample of vesicles, compared to density-based methods. One of the main advantages of ultracentrifugation method is it shows a greater protein yield, compared to other methods and remains enriched in exosomal-associated proteins. The ease of obtaining high overall yields of vesicles will be important for future experiments, as it will be utilised to not only be characterised by various methods, but also to perform functional experiments. In this Chapter, we have demonstrated different methods used to determine the phenotype of exosomes and the amount or concentration of exosomes present. All these methods act to analyse exosomes. By doing so, these methods act as a platform to evaluate whether exosomes are modulated by silencing putative regulators of exosome biogenesis and secretion in future Chapters.
92