Enabling Metabolomics Based Biomarker Discovery Studies Using Molecular Phenotyping of Exosome-Like Vesicles

Abstract

Identification of sensitive and specific biomarkers with clinical and translational utility will require smart experimental strategies that would augment expanding the breadth and depth of molecular measurements within the constraints of currently available technologies. Exosomes represent an information rich matrix to discern novel disease mechanisms that are thought to contribute to pathologies such as dementia and cancer. Although proteomics and transcriptomic studies have been reported using Exosomes-Like Vesicles (ELVs) from different sources, exosomal metabolome characterization and its modulation in health and disease remains to be elucidated. Here we describe methodologies for UPLC-ESI-MS based small molecule profiling of ELVs from human plasma and cell culture media. In this study, we present evidence that indeed ELVs carry a rich metabolome that could not only augment the discovery of low abundance biomarkers but may also help explain the molecular basis of disease progression. This approach could be easily translated to other studies seeking to develop predictive biomarkers that can subsequently be used with simplified targeted approaches.

Fig 1. Nanoparticle Tracking Analysis (NTA).

Determination of the concentration and particle size of human from 500, 1000 and 2000 μL of human plasma (Panel A). Size distribution of the nanovesicles was around 150 nm. Western blot analysis to confirm the expression of exosome specific markers (Panel B). Human plasma ELVs enrichment confirmed in exosomes isolated from different volumes of human plasma (2000, 1000 and 500 μL) and no presence of exosomal specific markers when loading total plasma (non-enriched). Lane A through C represent ELVs isolated from 2000, 1000 and 500 μL of plasma respectively; Lane D represents non-enriched total plasma.

Fig 2. Total Ion Chromatograms (TIC) of human plasma Exosome-Like Vesicles (EVLs) in the negative electrospray ionization mode.

ELVs were isolated from different volumes of human plasma (500, 1000 and 2000 μL) and subjected to TOF-MS analysis. The X-axis represents the chromatographic retention time while the Y-axis represents the relative intensity.

Fig 3. Metabolomic ELVs composition.

Characterization of the metabolite content of human plasma Exosome-Like Vesicles (P ELVs) (Panel A) or PANC 1 cell culture media (CCM ELVs) (Panel B).

Fig 4. Total Ion Chromatograms (TIC) of PANC1 pancreatic cancer cell culture media Exosome-Like Vesicles (ELVs) in the positive electrospray ionization mode.

ELVs were isolated from pancreatic cell line PANC 1 media and subjected to TOF-MS analysis. The X-axis represents the chromatographic retention time while the Y-axis represents the relative intensity.

Fig 5

Panel A. TGF-beta induces epithelial to mesenchymal transition (EMT) in PANC1 cells. Real Time PCR analysis showed a decrease of the epithelial marker E-cadherin (p-v<0.05) and an increase of the mesenchymal markers N-Cadherin and Vimentin (p-v<0.05) expression after TGF-β treatment in PANC1 cells. Panel B. Principal Component Analysis (PCA) of exosomal metabolome for MS negative ionization mode showing the separation between the two study groups. TGF-β treated (T) and control (C) PANC1 cells. We analyzed three replicates per condition. The x-axis shows interclass separation while y-axis illustrates the intra-class variability on Y-axis.

Fig 6. Multivariate analysis reveals distinct metabolic changes in plasma derived Exosome-Like Vesicles (ELVs) isolated from endometrial cancer (EC) patients compared to the control subjects.

Panel A. Heat map visualization of ion rankings of volcano plot based m/z, corresponding to their relative levels (intensity) in plasma ELVs isolated from EC patients and control subjects for MS positive ionization mode. Each row on the heat map represents a unique feature with a characteristic mass to charge ratio and retention time while each column represents one subject. Panel B. Principal Component Analysis (PCA) plot for MS positive ionization mode showing the separation between EC and control plasma ELVs. The x-axis shows interclass separation while y-axis illustrates the intra-class variability.

Fig 7. Comparison of the number of features detected in total plasma (non-purified) or plasma ELVs.

The number of common and unique features for each type of sample is represented for the negative (Panel A) and positive (Panel B) electrospray ionization mode.

Fig 8. Workflow.

Schematic showing experimental design in order to analyze the composition of exosome-like vesicles (ELVs) derived from plasma samples and cell culture media. (WB = western blot, NTA = nanoparticle tracking analysis).