Extracellular vesicles: the next generation of biomarkers for liquid biopsy-based prostate cancer diagnosis

Abstract

Prostate cancer (PCa) is a leading cause of cancer death for males in western countries. The current gold standard for PCa diagnosis - template needle biopsies - often does not convey a true representation of the molecular profile given sampling error and complex tumour heterogeneity. Presently available biomarker blood tests have limited accuracy. There is a growing demand for novel diagnostic approaches to reduce both the number of men with an abnormal PSA/ DRE who undergo invasive biopsy and the number of cores collected per biopsy. 'Liquid biopsy' is a minimally invasive biofluid-based approach that has the potential to provide information and improve the accuracy of diagnosis for patients' treatment selection, prognostic counselling and development of risk-adjusted follow-up protocols. Extracellular vesicles (EVs) are lipid bilayer-delimited particles released by tumour cells which may provide a real-time snapshot of the entire tumour in a non-invasive way. EVs can regulate physiological processes and mediate systemic dissemination of various types of cancers. Emerging evidence suggests that EVs have crucial roles in PCa development and metastasis. Most importantly, EVs are directly derived from their parent cells with their information. EVs contain components including proteins, mRNAs, DNA fragments, non-coding RNAs and lipids, and play a critical role in intercellular communication. Therefore, EVs hold promise for the discovery of liquid biopsy-based biomarkers for PCa diagnosis. Here, we review the current approaches for EV isolation and analysis, summarise the recent advances in EV protein biomarkers in PCa and focus on liquid biopsy-based EV biomarkers in PCa diagnosis for personalised medicine.

Keywords: biomarker; diagnosis; extracellular vesicles; liquid biopsy; prostate cancer.

Figure 1

Liquid biopsy as a non-invasive novel technique for PCa diagnosis.

Figure 2

Principles of common EVs isolating methods. (A) UC separates EVs based on sedimentation speed. Large non-EV particles (yellow) with a higher density are collected earlier at the bottom. EVs (red) are collected with a higher centrifuge force with a buffer (blue) replacement to remove soluble contaminates. (B) Precipitation reagent is added to change the solubility of EVs and soluble proteins, so that EVs can be separated with lower speed centrifugation. (C) UF separates EVs with specific cut-off membranes. Soluble proteins and particles smaller than EVs are pushed through the membrane, whereas EVs are collected on the membrane. (D) In SEC, particles with different sizes pass the column with different speeds by compressing through porous matrix (brown ball), so that EVs and particles with different sizes come out in different fractions. (E) Immunoaffinity magnetic beads are added to the EVs mixtures and capture EVs. The coated monoclonal antibody on the bead's surface binds to an antigen exposed on the targeted (red) EVs only. Beads with captured EVs are separated by the magnet, whereas non-EVs particles are washed out. (F). Field-based microfluidic device separates EVs thorough microchannels along the flow direction using external forces (such as surface acoustic wave) or physical obstacles [such as nanoscale deterministic lateral displacement (nanoDLD)]. Large particles are deviated towards one side, while small EVs are deviated towards another side.

Figure 3

Integrated nanoDLD arrays for PCa EVs isolation and analysis. (A) Layout of integrated nanoDLD chip. (B) The workflow of sample injection with zigzag and enriched, bump-particle fluid isolation. Collection in common reservoirs on the back side of the chip from the different via sets. (C) Heatmap of 50 literature-curated PCa ncRNA and mRNA markers expression levels in EVs isolated from serum by nanoDLD and UC. Adapted with permission from , copyright 2018 Royal Society of Chemistry.

Figure 4

Exodisc for PCa EVs isolation. (A) Layout and working principle of Exodisc-B with centrifugal double-filtration system. (B) Heat map of protein types of human-plasma-driven EVs extracted from PCa patients and healthy donors. Measurements sorting along HSP90 expression level for three EV markers (CD81, CD63 and CD9) and five cancer markers (EGFR1, PSMA, PSA, EpCAM and HSP90) measured from lysates of EVs using direct ELISA. Adapted with permission from , copyright 2019 Ivyspring International Publisher.

Figure 5

TAS for PCa EVs isolation. (A) EVs bound to aptamers detected by TAS. (B) The unweighted sum of the expression level of 7 markers (SUM signature) from EVs in different cohorts (means ± s.d.). (C) ROC curves showing a superior of SUM signature with higher accuracy (AUC = 0.9436) than serum PSA (AUC = 0.6842) in PCa versus benign disease discrimination. (D) Comparison of the SUM signature above 0.48 with serum PSA above 4 for the detection of PCa and benign disease. The vertical line indicates the threshold value of serum PSA (4 ng/mL) and the horizontal line indicates the threshold value of the SUM signature of EVs (0.48). Statistical differences were determined by two-tailed Mann-Whitney U-test. P values are indicated on the chart. Adapted with permission from . Copyright 2019 Springer Nature.