Proteomic Profiling of Two Distinct Populations of Extracellular Vesicles Isolated from Human Seminal Plasma

Abstract

Body fluids contain many populations of extracellular vesicles (EV) that differ in size, cellular origin, molecular composition, and biological activities. EV in seminal plasma are in majority originating from prostate epithelial cells, and hence are also referred to as prostasomes. Nevertheless, EV are also contributed by other accessory sex glands, as well as by the testis and epididymis. In a previous study, we isolated EV from seminal plasma of vasectomized men, thereby excluding contributions from the testis and epididymis, and identified two distinct EV populations with diameters of 50 and 100 nm, respectively. In the current study, we comprehensively analyzed the protein composition of these two EV populations using quantitative Liquid Chromatography-Mass Spectrometry (LC-MS/MS). In total 1558 proteins were identified. Of these, ≈45% was found only in the isolated 100 nm EV, 1% only in the isolated 50 nm EV, and 54% in both 100 nm and 50 nm EV. Gene ontology (GO) enrichment analysis suggest that both originate from the prostate, but with distinct biogenesis pathways. Finally, nine proteins, including KLK3, KLK2, MSMB, NEFH, PSCA, PABPC1, TGM4, ALOX15B, and ANO7, with known prostate specific expression and alternate expression levels in prostate cancer tissue were identified. These data have potential for the discovery of EV associated prostate cancer biomarkers in blood.

Keywords: LC-MS/MS; extracellular vesicles; prostasomes; proteomic analysis; seminal plasma; subtypes.

Figure 1

Isolation and characterization of LEV and SEV. Total EV in seminal plasma of vasectomized men were collected by UC at the interface of a sucrose block gradient. LEV and SEV were separated by their distinct velocities during upward displacement into a continuous sucrose density gradient. (A) Gradient fractions were analyzed by SDS-PAGE followed by Sypro ruby staining for total protein. (B) Gradient fractions were analyzed by immunoblotting for the presence of EV associated proteins, including CD9, CD81, PSCA, Galectin-3, CD63, HSP70, Annexin A1, and GLIPR2. Molecular weight markers are indicated on the left in kDa. (C) Particles in the LEV and SEV containing fractions were analyzed by TEM. Scale bar, 500 nm. Arrows exemplify incidental SEV in the LEV isolate. (D) Size distribution of the particles in LEV and SEV isolates as determined by NTA.

Figure 2

MS/MS identification and relative quantification of proteins in LEV and SEV. (A) Venn diagram indicating the total number and overlap of proteins in LEV and SEV detected in experiment 1 (red) and experiment 2 (blue). (B) Venn diagram comparing of proteins detected in LEV (yellow) and SEV (beige) in experiment 1 (red numbers) and experiment 2 (blue numbers). (C) Venn diagram comparing proteins that are shared by LEV and SEV in experiment 1 (red) with those in experiment 2 (blue). (D) iBAQ ratios (LEV/SEV) for all 684 proteins that shared within both experiment 1 and experiment 2, as indicated in C. Proteins are plotted on x-axes by decreasing iBAQ ratio in experiment 1 (red line). The same proteins are plotted in the same order along the x-axes for experiment 2 (blue dots).

Figure 3

(A) IBAQ heatmap for all 1557 proteins that were detected either in LEV and/or SEV in both experiments (see Figure 1A). Color code key on the right indicates iBAQ values. Red color is corresponding to relatively high abundance of proteins, darkest blue indicates absence of protein. (B) IBAQ values of six selected proteins in LEV and SEV that were also compared by immunoblotting (see Figure S3B).

Figure 4

Gene ontology (GO) annotations of proteins detected in both experiments in LEV only (539 proteins), SEV only (seven proteins), or shared by LEV and SEV (684 proteins). Ranking is according to the database in FunRich and ordered by the percentage of genes per GO term. (A) Cellular component annotations. (B) Molecular function annotations. (C) Biological process annotations.

Figure 5

IBAQ heatmaps for specific protein groups potentially related to EV biogenesis. (A) ESCRT and associated proteins. (B) Small GTPases. (C) Heat shock proteins. (D) RNA binding proteins. (E) Tetraspanin proteins. Color code key on the right indicates iBAQ values. Red color is corresponding to relatively high abundance of proteins, darkest blue indicates absence of protein.

Figure 6

Protein–protein interaction network as determined by STRING for all proteins shown in Figure 5. Red encircled proteins were detected in LEV only. All other proteins were detected in both LEV and SEV. Dotted lines indicate the different protein groups in Figure 5A–E. Note that these protein groups are maintained by this presentation. Additionally, note the linkage between “RNA binding proteins” and “ESCRT and associated proteins” by RPS27A, and that both “tetraspanins” and “small GTPases” are linked to “RNA binding proteins” via “heat shock proteins”.