Supplemental Data
Files in this Data Supplement:
- supplamental figure 1 - Supplemental Figure 1. Summary of the SOMAscanTM array platform method. SOMAmers® are modified aptamers, composed of a unique 40-mer sequence of single stranded DNA, which exhibit a selective protein binding domain. The SOMAmers® are configured with a fluorescent reporter, a photo-cleavable linker and a biotin tag at the 5’ end (A). Samples containing mixtures of soluble proteins are added to SOMAmer® mixtures that are immobilised through biotin to a streptavidin coated surface, and unbound proteins are washed away (B). Bound proteins are subjected to a biotinylation step (C), and SOMAmers® are thereafter liberated from the surface by exposure to ultraviolet light (D). Protein-SOMAmer® complexes are immobilised for a second time, through the biotinylated proteins to a streptavidin coated surface (E). Following denaturation, SOMAmers® are released from associated proteins (F) and hybridised to custom micro-array chips, containing complementary probes to each individual SOMAmer® (G). Cy3 fluorescence intensity is measured, and is proportional to the quantity of specific protein in the original sample.
- supplamental figure 2 - Supplemental Figure 2. Selecting sample buffer conditions for SOMAscan analysis. Exosomes were purified from Du145 cell conditioned media using the sucrose cushion method (A), and diluted in standard buffer conditions (SB17), or with other reagents added as indicated (B). The samples were assayed using an earlier version of the SOMAscanTM array, analysing 300 analytes. The RFU values from each of the 300 individual analytes are shown (dots), together with the mean (vertical bar) (B). Heat-map depiction of the same data emphasising similarities between NP40/DCO and MPER conditions and the effect of adding DTT (C). Similar data were generated using exosomes purified from the PC3 and LNcAP cell lines (not shown). (D-F) Some randomly selected examples comparing SB17 with SB17+NP40/DCO conditions, from the full v3.0 version of the array platform (covering 1129 analytes) depicting identifications where the extra detergent elevated the signal (D), or had little effect (E) or had decreased the signal (F) for the specified proteins.
- supplamental figure 3 - Supplemental Figure 3. Comparing SOMAscanTM identifications with the Vesiclepedia database Using the search terms “human”, “exosomes” and “prostate” we identified 11 Vesiclepedia database entries with MS-generated protein lists, and the identifier numbers for these entries are specified (A, table). Gene name lists extracted from these were merged, and replicates removed. A second gene name list was generated from the array, taking the gene names reporting an RFU >200. A third list containing all the filtered SOMAscanTM proteins irrespective of RFU levels were also compared (A, Venn diagram). There were 91 common identifications between the Vesiclepedia database and SOMAscanTM positive identifications (B, list). Some database entries, also covered by the array, were below the 200RFU threshold, and hence considered not expressed according to these criterion (C, list), suggesting that ~2.6% of the total SOMAscanTM coverage should have been detected as positive, but are reported as negatives, based on the conservative 200 RFU-cut off. A total of 392 proteins were found exclusive to the SOMAscanTM dataset (D, list), and were therefore not identified by MS-based studies of prostate-related exosomes. Of the 532 Vesiclepedia identifications, 415 (78%) were not covered by the SOMAscanTM assay, and hence were not identifiable by this approach. Those names in bold and underlined have been verified by us using other methods. Given the diversity in methods across the Vesiclepedia studies, we cannot comment on the level of exosome purity or stringency of the MS-reported identifications in the database entries.
- supplemantal figure 4 - Supplemental Figure 4. Comparing SOMAscanTM identifications with the Vesiclepedia database Analysis performed as for Fig S3, except following the removal of all ambiguous identifications (those listed in Table S3). The relevant identifications were removed from all lists (Vesiclepedia and SOMAmer-derived), and compared by Venn diagram.
- supplemental table 1 - Supplemental Table 1. Summary of the SOMAscanTM SB17+NP40/DCO conditions. Data generated from the SB17+NP40/DCO conditions, were subjected to filtering to remove high abundance replicates (as explained in the methods), and the remaining gene names with their corresponding log-transformed RFU values presented for both exosomes and cells. The averaged coefficient of variation and the Benjamini-Hochberg corrected p value is included. The fold change where + relates to elevation, and – to decrease in levels in exosomes compared to cells. The table is ranked in order of exosomes versus cells fold change, and uses red-text to identify those of >2 fold elevation, grey shading depicting proteins <2 fold difference, orange shading indicating those of comparable levels and green shading highlighting proteins >2 fold elevated in cells.
- Table s2 - Supplemental Table 2. Summary of the SOMAscanTM standard conditions. Data generated from the standard assay buffer (SB17+tween-20) conditions, were subjected to filtering to remove high abundance replicates (as explained in the methods), and the remaining gene names with their corresponding log-transformed RFU values presented for both exosomes and cells. The averaged coefficient of variation and the Benjamini-Hochberg corrected p value is included. The fold change where (+) relates to elevation, and (–) to decrease in protein levels in exosomes compared to cells. The table is ranked in order of exosomes versus cells fold change, and uses red-text to identify those of >1.5 fold elevation, grey shading depicting proteins <1.5 fold difference, orange shading indicating those of comparable levels and green shading highlighting proteins >1.5 fold elevated in cells.
- supplemental table 3 - Supplemental Table 3. Protein to Gene name conversion anomalies. SOMAmer-based protein identifications do not always map to a single gene name, in situations where the SOMAmer detects a complex of polypeptides such as the  and  chains of an integrin receptor. Thus in tables S1 and S2 some gene names, arising from a single SOMAmer exhibit an identical RFU output. The SOMAmers and corresponding gene names affected are summarised in the table. Comparisons with the Vesiclepedia database are affected by these gene names and such analysis was performed with these included and removed. These identifications were removed for the DAVID-based bioinformatics analyses to avoid their unwanted influence on p values.
- supplemental table s4 - Supplemental Table 4. DAVID biological themes. Table of biology themes resulting from DAVID analyses of the 50 genes with the largest fold change in exosomes for SB17 and SB17 + NP40/DCO. Biological terms are clustered according to the similarity of their gene content
- legend - legends for supplemental materials