Circulating Extracellular-Vesicle-Incorporated MicroRNAs as Potential Biomarkers for Ischemic Stroke in Patients With Cancer

Abstract

Background and purpose: This study aimed to evaluate whether extracellular-vesicle-incorporated microRNAs (miRNAs) are potential biomarkers for cancer-related stroke.

Methods: This cohort study compared patients with active cancer who had embolic stroke of unknown sources (cancer-stroke group) with patients with only cancer, patients with only stroke, and healthy individuals (control groups). The expression profiles of miRNAs encapsulated in plasma exosomes and microvesicles were evaluated using microarray and validated using quantitative real-time polymerase chain reaction. The XENO-QTM miRNA assay technology was used to determine the absolute copy numbers of individual miRNAs in an external validation cohort.

Results: This study recruited 220 patients, of which 45 had cancer-stroke, 76 were healthy controls, 39 were cancer controls, and 60 were stroke controls. Three miRNAs (miR-205-5p, miR-645, and miR-646) were specifically incorporated into microvesicles in patients with cancer-related stroke, cancer controls, and stroke controls. The area under the receiver operating characteristic curves of these three miRNAs were 0.7692-0.8510 for the differentiation of patients with cancer-stroke from cancer-controls and 0.8077-0.8846 for the differentiation of patients with cancer-stroke from stroke controls. The levels of several miRNAs were elevated in the plasma exosomes of patients with cancer, but were lower than those in plasma microvesicles. An in vivo study showed that systemic injection of miR-205-5p promoted the development of arterial thrombosis and elevation of D-dimer levels.

Conclusion: Stroke due to cancer-related coagulopathy was associated with deregulated expression of miRNAs, particularly microvesicle-incorporated miR-205-5p, miR-645, and miR-646. Further prospective studies of extracellular-vesicle-incorporated miRNAs are required to confirm the diagnostic role of miRNAs in patients with stroke and to screen the roles of miRNAs in patients with cancer.

Keywords: Biomarker; Cancer; Coagulopathy; Extracellular vesicle; MicroRNA; Stroke.

Figure 1.

Study design. miRNA, microRNA; AUC, area under the receiver operating characteristic curve; KEGG, Kyoto Encyclopedia of Genes and Genomes; GO, Gene Ontology; NETosis, neutrophil extracellular traposis.

Figure 2.

Characteristics of plasma extracellular vesicles. (A) Process for the isolation of exosomes and microvesicles. (B) Size distribution of microvesicles and exosomes, as determined by NanoSight Tracking Analysis. (C) Size and lipid double layers of purified exosomes and microvesicles, as determined by electron microscopy. (D) Western blot analysis of extracellular vesicle markers of microvesicles and exosomes. PBS, phosphate buffered saline.

Figure 3.

Analysis of plasma microvesicle-microRNAs in pooled samples of patients with cancer and stroke and patients with cancer and without stroke. (A) Microarray data of pooled samples of patients with cancer only (cancer-control [CC]) and patients with cancer-stroke (CS). (B) Venn diagrams of increased levels of microRNAs identified in exosomes and microvesicles of patients with CC and those with CS. The numbers represent the number of microRNAs that were increased in microvesicles (more than 5 folds) and exosomes (3 folds) compared with health individuals at Ct (cycle threshold) less than 35.

Figure 4.

Analysis of plasma microvesicle microRNAs in individual samples of patients with cancer. (A) Quantitative real-time polymerase chain reaction (qRTPCR) results of relative abundance of microvesicle-microRNAs in patients with cancer-stroke (CS) compared with healthy controls and cancer controls (CC). (B) Receiver operating characteristics curve for cancer-stroke in patients with cancer. AUC, area under the receiver operating characteristics curve.

Figure 5.

Analysis of plasma microvesicle microRNAs (miRNAs) in individual samples of patients with stroke. (A) Quantitative real-time polymerase chain reaction (qRT-PCR) results of relative abundance of microvesicle-miRNAs in patients with cancer-stroke compared with stroke-controls. (B) Receiver operating characteristics curve for cancer stroke in patients with stroke. The cutoff score was 8.2-fold for miR-205-5p, 5.6-fold for miR-645, and 129.2-fold for miR- 646 (as normalized with miR-39, a spike-in control). AUC, area under the receiver operating characteristics curve.

Figure 6.

Microvesicle-incorporated microRNA (miRNAs) profiles of a validation cohort determined by xeno-sensors. (A) The absolute copy numbers of the three miRNAs within microvesicles were determined by performing XENO-ONT direct sequencing. (B) The expression of the three miRNAs within microvesicles was validated by performing the XENO-Q miRNA detection assay. (C) Receiver operating characteristic curve analysis of three miRNAs for the presence of cancer-related stroke vs. cancer control or stroke control. *P<0.05; **P<0.01; ***P<0.001. AUC, area under the receiver operating characteristics curve.

Figure 7.

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) functional analysis and target gene validation. (A and B) GO of biological processes and KEGG pathways presented using the 10 most relevant terms for target genes of microRNAs enriched in microvesicles. (C) Human umbilical vein endothelial cells (HUVEC) were transfected with control microRNA (con-miR) or three microRNA mimics (for microRNA overexpression) respectively. The level of vascular endothelial growth factor A (VEGF-A) mRNA was assessed through quantitative real-time polymerase chain reaction (qRT-PCR) analysis. Data were presented with mean ± standard deviation. Statistical analyses were performed using Student’s t-test in three independent experiments (**P<0.01). (D) Western blot analysis was performed to evaluate the expression level of VEGF-A in HUVEC. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a loading control. Quantification of image were performed with ImageJ program and statistical analysis were presented by Student’s t-test (**P<0.01).

Figure 8.

The effect of miR-205-5p in a mouse model of carotid artery thrombosis. (A) The scheme of in vivo common carotid artery (CCA) thrombosis. (B) Occlusion time of the common carotid arteries and the plasma levels of D-dimer in the control agomir- or hsa-miR-205-5p agomir-treated mice (n=5 per group). Data are presented as mean±standard deviation. **P<0.01, ***P<0.001 vs. control agomir. (C) Representative blood flow curve and laser speckle image in control agomir- or hsa-miR-205-5p agomir-treated mice. The laser was exposed on the common carotid artery (a) followed by common carotid artery occlusion (b). (D) Hematoxylin and eosin (H&E) staining of transverse section of common carotid artery in agomir-treated mice.