Circulating exosomes contain protein biomarkers of metastatic non-small-cell lung cancer

2.1 Patients and clinical samples

A total of 183 patients with NSCLC (113 males and 70 females; mean age, 58.7 years; range 34 to 88) were enrolled in the Shandong Cancer Hospital. Written informed consent was obtained from all subjects. Based on the tumor and metastasis classification system, 94 patients with non-metastatic NSCLC and 89 patients with metastatic NSCLC were analyzed. None of the patients had received any treatment. Controls were recruited from outpatients who had undergone a general medical examination. Blood samples of included patients were collected and centrifuged at 3000 g, at 4°C for 15 min. The supernatant was stored at −80°C.

2.2 Isolation of exosomal proteins from human serum

Serum samples were thawed on ice,27 and the serum was centrifuged at 10 000 g at 4°C for 30 min. Then, the samples were ultracentrifuged (Beckman Coulter Class H, R, and S preparative ultracentrifuges, Type 50.4 Ti Rotor; Beckman Coulter, Brea, CA, USA) at 100 000 g at 4°C for 120 min. Next, the exosome pellets were washed with PBS, which was filtered through a 0.22-μm pore filter, followed by a second step of ultracentrifugation at 100 000 g at 4°C for 120 min. Measurement of nanoparticle size was based on tunable resistive pulse sensing (TRPS) and conducted using the qNano (Izon Science Ltd, Christchurch, New Zealand), combining tunable nanopores with proprietary data capture and analysis software, Izon Control Suite v.3.3.2.2000 (Izon Science Ltd). The supernatant was discarded, and pelleted exosomes were resuspended in 100 μL lysis buffer (8 mol/L urea, 1% protease inhibitor, 3 μM Trichostatin A (TSA), 50 mmol/L nicotinamide (NAM), and 2 mmol/L EDTA) for proteomic analysis.

2.3 Sample preparation, MS, and interpretation of mass spectra

Protein concentration in the supernatant was determined using a 2D Quant kit (GE Healthcare, Little Chalfont, UK) according to the manufacturer's instructions. For digestion, the protein solution was reduced with 5 mmol/L dithiothreito (lDTT) (Sigma Chemical Co., St Louis, MO, USA) for 30 min at 56°C and alkylated with 11 mmol/L iodoacetamide (IAA) for 15 min at room temperature in the dark; the protein sample was then diluted for trypsin digestion. Then, the peptides were combined into 18 fractions and dried by vacuum centrifuging. The peptides were subjected to an nitrogen solubility index (NSI) source followed by tandem mass spectrometry (MS/MS) in Q Exactive (Thermo Fisher Scientific, San Jose, CA, USA) coupled online to an ultra-performance liquid chromatography (UPLC) system. The resulting MS/MS data were processed using MaxQuant with integrated Andromeda search engine (v.1.5.2.8). Gene Ontology, or GO, is a major bioinformatics initiative to unify the representation of gene and gene product attributes across all species. The GO annotation proteome was derived from the UniProt-GOA database (www. http://www.ebi.ac.uk/GOA/). The Kyoto Encyclopedia of Genes and Genomes (KEGG) connects known information on molecular interaction networks. These pathways were classified into hierarchical categories according to the KEGG website.

2.4 Immunoblotting

Lysis buffer (100 μL; 1:10 dilution: 20 mmol/L Tris [pH 7.5], 150 mmol/L NaCl, 1% Triton X-100, sodium pyrophosphate, β-glycerophosphate, EDTA, Na₃VO₄, leupeptin) was added to the samples, which was subsequently mixed and incubated on ice for 30 min. The denatured protein was stored in a refrigerator at −20°C. Proteins (10 μL) from EV were loaded on an 8%-15% SDS gel for electrophoresis. Primary antibodies used were rabbit polyclonal anti-HSP70 (clone W70) and rabbit polyclonal anti-LBP (both from Proteintech, Chicago, IL, USA). Peroxidase-conjugated AffiniPure goat anti-mouse (H + L) and goat anti-rabbit (H + L) (both from Jackson ImmunoResearch, West Grove, PA, USA) were used as secondary antibodies.

2.5 Enzyme-linked immunosorbent assay

Blood plasma samples of patients with metastatic NSCLC, patients with early NSCLC (stage I to IIIa), and healthy donors were analyzed for expression of LBP using a Human LBP DuoSet ELISA (R&D Systems, Minneapolis, MN, USA). For ELISA testing, all serum samples were diluted with reagent diluent (1:1000) before ELISA analysis. The samples were analyzed as recommended by the manufacturer. Protein content of exosomes was determined using a fluorescent microplate reader (Molecular Devices, Biberach an der Riß, Germany), following the manual instructions.

2.6 Statistical analysis

Data were expressed as the median with interquartile range. The data between two groups were compared using the Mann-Whitney U-test and t test. Values of P < .05 were considered to be statistically significant. All statistical analyses were carried out using SPSS 22.0 statistical software.

3.1 Isolation and identification of exosomes

Extracellular vesicles from the serum of patients with NSCLC and healthy donors were isolated by ultracentrifugation, and were called exosomes based on the following observations. Scanning electron microscopy (SEM) showed oval-shaped extracellular vesicles of 50-150 nm diameter (Figure 1A). Moreover, the qNano system also showed that the diameter of the exosomes was 50-150 nm (Figure 1B). Quantitative results of protein concentration showed that the expression of total proteins from exosomes in patients with NSCLC was higher compared with that of healthy donors (P < .0001; Figure 1C). Median expression of proteins in patients with NSCLC was 0.416 mg/mL (range, 0.164-1.377 mg/mL) and the median expression of exosomal LBP in healthy donors was 0.234 mg/mL (range, 0.089-0.269 mg/mL). Total proteins from exosomes in patients with metastatic NSCLC were significantly higher compared with those in patients with non-metastatic NSCLC (P < .0001). Median expression of proteins was 0.996 mg/mL (range, 0.435-2.462 mg/mL) and 0.204 mg/mL (range, 0.056-0.400 mg/mL), respectively. However, no significant differences were found between healthy controls and non-metastatic NSCLC (P = .308, Figure 1C).

3.2 Profile of proteome in NSCLC cell-derived exosomes

Non-small-cell lung cancer exosomes are enriched in proteins and are implicated in the tumor microenvironment. Studies have shown that they can regulate tumor metastasis.6 In the present study, patients with distant metastatic disease (bone, brain etc., n = 5), non-metastatic NSCLC (stage I-IIIa, n = 5), and healthy donors (n = 5) were selected as a discovery cohort for proteomic analysis. In total, 628 proteins from exosomes were identified, among which 552 proteins were quantified. A quantitative ratio of more than 1.5 was considered upregulation, whereas that less than 0.667 was considered downregulation. All identified quantifiable proteins with increased (≥1.5 fold) and decreased (≤0.667) expression levels are listed in Tables S1–S4 (P < .05).

Compared with metastatic NSCLC, non-metastatic NSCLC induced 61 differentially expressed proteins: 15 upregulated proteins, such as serum amyloid A-2 protein; and 46 downregulated proteins, such as apolipoprotein M. Compared with healthy donors, NSCLC induced 43 differentially expressed proteins: 34 upregulated proteins and nine downregulated proteins.

3.3 Functional enrichment of differentially quantified proteins and clustering for protein group

Based on the differentially regulated proteins in exosomes derived from patients with NSCLC and healthy donors, the GO analysis tool was used to screen the classification of affected biological functions, including biological process, molecular function, and cellular component. The affected biological functions and the number of related proteins in patients with NSCLC and healthy donors as well as in patients with metastatic NSCLC and non-metastatic NSCLC were analyzed and compared. The top 5 affected biological functions and the amount of differentially expressed proteins are summarized in Tables 1 and 2 (Figures S1 and S2).

The enrichment test was then carried out to identify the GO terms (molecular function, cellular component, and biological process) that were significantly enriched to indicate the nature of the differentially expressed proteins between patients with NSCLC and healthy donors. The biological process was first investigated (Figure 2A); it was found that vesicle-mediated transport [-log10 (P-value) = 3.01] was enriched with upregulated proteins. Meanwhile, the downregulated proteins related to complement activation [-log10 (P-value) = 2.39] were significantly enriched in patients with NSCLC; and on the ontology of molecular function (Figure 2B), the protein heterodimerization activity [-log10 (P-value) = 1.35] was enriched with upregulated proteins. Endopeptidase activity [-log10 (P-value) = 2.4] in downregulated proteins was enriched in patients with NSCLC. Moreover, on the ontology of cellular component (Figure 2C), the upregulated component including membrane-enclosed lumen [-log10 (P-value) = 2.43] and the downregulated proteins related to blood microparticle [-log10 (P-value) = 1.47] were significantly enriched in patients with NSCLC. The results thus indicated that biological function might be important in regulating tumorigenesis.

Next, enrichment tests were also carried out to identify the GO terms to indicate the nature of the differentially expressed proteins between metastatic and non-metastatic NSCLC (stage I-IIIa). On the ontology of biological process (Figure 2A), the upregulated proteins related to the acute-phase response [-log10 (P-value) = 2.7] were significantly enriched in metastatic NSCLC; meanwhile, the downregulated proteins related to protein carboxylation [-log10 (P-value) = 3.89] were enriched in biological process terms. On the ontology of molecular function (Figure 2B), protease binding [-log10 (P-value) = 2.23] was enriched with upregulated proteins and enzyme inhibitor activity [-log10 (P-value) = 2.25] was enriched with downregulated proteins. On the ontology of cellular component (Figure 2C), the upregulated component including extracellular space [-log10 (P-value) = 3.45] was significantly enriched in patients with metastatic NSCLC, and the Golgi lumen [-log10 (P-value) = 1.74] was significantly enriched with downregulated proteins. The results thus indicated that the biological function of regulating tumorigenesis and metastasis was different, and that proteome changes could distinguish between patients with metastatic and non-metastatic NSCLC. Clustering analysis based on the KEGG pathway was carried out to identify cellular pathways and the protein complex related to NSCLC (Figure 2D).

Next, the cellular pathways that related to metastatic NSCLC (Figure 3A) were identified. Results showed that the extracellular matrix-receptor interaction pathway [-log10 (P-value) = 1.81] was the dominant pathway enriched in upregulated proteins in metastatic NSCLC compared with the non-metastatic NSCLC. Meanwhile, the results showed that the metabolic pathway [-log10 (P-value) = 1.56] was the dominant pathway enriched in downregulated proteins in the same compared group.

3.4 Validation of LBP in serum exosomes could serve as a metastatic biomarker

Based on the quantitative results, the list of proteins of interest was narrowed down, and functional studies were carried out for target candidates, paying more attention to the proteins with the most significant expression changes, such LBP, the fold change of which between patients with NSCLC and healthy donors was 2.574. Meanwhile, the fold change was 1.845 in metastatic NSCLC compared with the non-metastatic NSCLC.

Samples including metastatic and non-metastatic NSCLC and healthy donors were analyzed using western blotting in an independent patient cohort comprising patients with metastatic NSCLC (n = 51), patients with non-metastatic NSCLC (n = 51), and healthy donors (n = 51). Presence of exosomal marker HSP70 was confirmed by western blotting (Figure 3A). Quantitative results showed that patients with metastatic NSCLC showed high expression levels of LBP compared with patients with non-metastatic NSCLC and healthy donors. However, the non-metastatic NSCLC group showed similar exosomal LBP levels to healthy donors (Figure 3A).

Concentration of LBP was measured using ELISA in a symptomatic set with 273 subjects to determine whether LBP proteins entered the bloodstream at levels that could be measured and used as cancer biomarkers (Table3). Of the 273 individuals, 183 patients with NSCLC (89 metastatic NSCLC and 94 non-metastatic NSCLC) and 90 healthy donors were identified. Median expression level of exosomal LBP in patients with NSCLC and healthy donors was 1.895 ng/mL (range 0.588-3.887 ng/mL) and 0.577 ng/mL (range 0.143-1.138 ng/mL), respectively, with a significant difference (P < .0001; Figure 3D). Area under the curve (AUC) was 0.713 with a sensitivity of 65% and a specificity of 75.6% (Figure 3E). Based on this result, expression level of exosomal LBP from patients with metastatic and non-metastatic NSCLC was further analyzed. Analysis of the validation cohort showed that median expression level of exosomal LBP from patients with metastatic NSCLC and patients with non-metastatic NSCLC was 3.23 ng/mL (range 1.876-5.330 ng/mL) and 0.881 ng/mL (range 0.204-1.994 ng/mL), respectively. Expression level of exosomal LBP in patients with metastatic NSCLC was significantly higher than the level in those with non-metastatic NSCLC (P < .0001; Figure 3D). Receiver operating characteristic (ROC) curves indicated that exosomal LBP proteins had an AUC of 0.803 with a sensitivity of 83.1% and a specificity of 67% (Figure 3F). However, no statistically significant difference was found between patients with non-metastatic NSCLC and healthy donors (P = .09; Figure 3D).

Expression level of selected exosomal LBP of patients with AC and SCC was analyzed using ELISA. Results showed that the median expression level of exosomal LBP from patients with AC and patients with SCC was 2.609 ng/mL (range 1.188-5.260 ng/mL) and 0.918 ng/mL (range 0.361-2.924 ng/mL), respectively, showing an AUC value of 0.659 with a sensitivity of 85.5% and a specificity of 47.4% in distinguishing patients with AC from patients with SCC (P = .003; Figure 3B,C). Next, serum samples of the validation cohort (40 healthy donors, 36 non-metastatic NSCLC, and 44 metastatic NSCLC) were analyzed for circulating LBP levels. Results showed that the median expression level of circulating LBP was 25.334 ng/mL (range 20.503-29.020) in patients with NSCLC and 18.340 ng/mL (range 15.643-23.569) in healthy donors, and the AUC was 0.676 with a sensitivity of 68.8% and a specificity of 67.5%. This indicated that patients with NSCLC had a high expression level of circulating LBP compared with the healthy donors (P = .002; Figure 3G,H). Moreover, high levels of circulating LBP were associated with advanced stages and presence of metastasis in metastatic NSCLC. Expression level of circulating LBP was higher in patients with metastatic NSCLC (95% confidence interval [CI], 27.964 [25.456-30.472]) compared with patients with non-metastatic NSCLC (95% CI, 22.551 [20.638-24.464]; P = .005; Figure 3G). ROC curves showed an AUC of 0.683 with a sensitivity of 79.5% and a specificity of 47.2% (Figure 3I). Moreover, the circulating LBP levels in patients with non-metastatic NSCLC were similar to those in healthy donors (P = .061; Figure 3G).