Identification and characterization of exosomes in cell-free serum specimens.
Transmission electron microscopy (TEM), nanoparticle tracking analysis (NTA) and western blot analysis were performed to determine whether exosomes were extracted efficiently from serum samples collected from 72 GC patients and 20 healthy controls. The morphology of exosomes was directly observed by TEM. The exosomes showed spherical or oval vesicle shape and were found to have the lipid bilayer membrane structure with a diameter of 50–100 nm (Fig. 1a). Meanwhile, the results of NTA showed that the diameter of exosomes from the serum of GC patients was similar to that from healthy controls, and the average size of exosomes was 50–100 nm (Fig. 1b). The concentration of exosomes from the serum of GC patients was higher than that from healthy controls (P < 0.01) (Fig. 1c). Furthermore, we also confirmed that the isolated exosomes from the serum of GC patients and healthy controls expressed specific exosomal markers such as CD9 and CD81 by using western blot analysis (Fig. 1d). Taken together, these results showed that exosomes were successfully isolated from the serum samples, which provided a solid foundation for the further study of exosomal biomarkers.
Screening and evaluation of exosomal miR-17-92 cluster in the serum samples of gastric cancer patients.
The expression profiles of the serum exosomal miR-17-92 cluster, including miR-17, miR-18, miR-19a, miR-19b, miR-20 and miR-92, were detected by RT-qPCR analysis. As shown in Fig. 2, all of the six members of the exosomal miR-17-92 cluster could be clearly detected in the serum samples of human GC patients and healthy controls, albeit with different levels. Our data showed that the expression level of serum exosomal miR-17 in GC patients was significantly higher than that in healthy controls (P < 0.001) (Fig. 2a). Accordingly, the expressions of serum exosomal miR-18, miR-19a and miR-92 were also statistically much higher in GC patients compared with those of healthy controls (Fig. 2b, c, f). However, there were no significant differences in the expression of serum exosomal miR-19b and miR-20 between GC patients and healthy controls (P > 0.05) (Fig. 2d, e). Therefore, our results showed that the expression profile of the serum exosomal miR-17-92 cluster in GC patients was different from the healthy controls.
Evaluation of the diagnostic potential of serum exosomal miR-17-92 cluster for gastric cancer.
ROC (receiver-operating characteristic curve) analysis was used to explore the diagnostic utility of the serum exosomal miR-17-92 cluster for GC, with the sensitivity as the y axis and the 1-specificity as the x axis. The AUC (area under the curve) for serum exosomal miR-17 was 0.750 (95%CI, 0.626–0.874), the sensitivity was 84.7% and the specificity was 70.0% (Fig. 3a). Besides, the AUC for miR-18 was 0.736 (95%CI, 0.590–0.881) with sensitivity of 88.9% and specificity of 65.0% (Fig. 3b). Accordingly, the AUC for miR-19a and miR-92 was 0.700 (95%CI, 0.562–0.838) and 0.689 (95%CI, 0.567–0.811), respectively (Fig. 3c, d). These results showed that the AUC of serum exosomal miR-17 was the highest among the four miR-17-92 cluster members, indicating that miR-17 had the most powerful diagnostic efficacy for GC. In addition, we further explored whether combination of the exosomal miR-17-92 could be used to better distinguish GC patients from healthy controls. As shown in Fig. 3e, the combined detection of miR-17 and miR-18, which was the top two up-regulated miR-17-92 cluster members in GC patients, received the AUC of 0.774 (95%CI, 0.638–0.911), with sensitivity of 87.5% and specificity of 70.0%. It seemed that the combination of miR-17 and miR-18 showed much higher predictive value for the diagnosis of GC than each alone. Furthermore, we wondered whether the combination of all the four members could be the most powerful predictive marker for the diagnosis of GC. As expected, the combined panel consisting of miR-17, miR-18, miR-19a and miR-92 received the AUC of 0.808 (95%CI, 0.680–0.937), with sensitivity of 90.3% and specificity of 70.0% (Fig. 3f), which showed much higher clinical diagnostic value for GC than any of the four alone or any pair. Taken together, these results indicated that the serum exosomal miR-17-92 cluster could serve as potential predictive biomarker for the diagnosis of GC, and the exosomal miRNA combination panel could enhance the diagnostic power for GC.
Combining multiple biomarkers enhances the diagnostic power for gastric cancer.
To better evaluate the diagnostic potential of exosomal miR-17-92, we collected the clinical data from the 72 GC patients and 20 healthy controls for further study. The traditional tumor biomarkers such as CEA and CA19-9, which were commonly used in our clinical work, were calculated between the GC patients and healthy controls. As shown in Fig. 4a, the expression level of CEA was significantly upregulated in the serum of GC patients compared with that of healthy controls. Similarly, the expression of CA19-9 was also statistically upregulated in the GC patients than that of healthy controls (Fig. 4b). Then, ROC analysis was performed to explore the diagnostic efficiency of the two traditional tumor biomarkers. The AUC for CEA was 0.697 (95%CI, 0.569–0.825), the sensitivity was 56.9% and the specificity was 85.0% (Fig. 4c). The AUC for CA19-9 was 0.676 (95%CI, 0.547–0.805) with sensitivity of 70.8% and specificity of 65.0% (Fig. 4d). Besides, the combined panel of CEA and CA19-9 was built to further investigate the diagnostic efficiency for GC. As expected, the combination of CEA and CA19-9 showed much higher diagnostic value with the AUC of 0.738 (95%CI, 0.615–0.861) than any of the two alone (Fig. 4e). Furthermore, we also built the panel consisting of the two traditional tumor biomarkers and the four miR-17-92 cluster members, and the ROC of the newly developed diagnostic panel was calculated. Our results showed that the AUC for the newly developed panel was 0.881 (95%CI, 0.765–0.998) with sensitivity of 91.7% and specificity of 90.0% (Fig. 4f), which showed the greatest powerful clinical diagnostic value for GC. Taken together, these results suggested that the serum exosomal miR-17-92 cluster, together with traditional tumor biomarkers, could significantly enhance the diagnostic power for GC.
Correlation of serum exosomal miR-17-92 expression with clinicopathological parameters of gastric cancer patients.
In order to better understand the potential role of the serum exosomal miR-17-92 cluster in the development of GC, the association of the expression levels of each member of the miR-17-92 cluster with various clinicopathological features of GC patients was further analyzed. As shown in Table 2, the serum exosomal miR-17 expression was strongly related to the tumor size, tumor depth, distant metastasis and TNM stage of GC patients. However, there was no significant correlation between the expression of serum exosomal miR-17 and other clinical pathological characteristics such as age, gender, histological grade, lymph node metastasis and venous invasion of GC patients. Correspondingly, the serum exosomal miR-18 expression was significantly associated with tumor size, distant metastasis and TNM stage of GC patients. The serum exosomal miR-19a expression was markedly relevant to tumor depth, lymph node metastasis and TNM stage of GC patients. The serum exosomal miR-92 expression was greatly related to tumor depth, distant metastasis and TNM stage but not to the other clinicopathological features of GC patients. Taken together, these results suggested that exosomal miR-17-92 cluster might play an oncogenic role in the progression of GC.