Study Of Serum MiRNAs As Biomarkers For Non-small Cell Lung Cancer

Lung cancer is the leading cause of cancer-related mortality, and the majority of cases are diagnosed at an advanced stage. New biomarkers are needed to assist diagnosis and early detection. miRNAs have been reported to be potential biomarker candidates for non-small cell lung cancer (NSCLC). In this study, we sought to identify novel serum miRNAs that could help distinguish NSCLC from healthy individuals and one miRNA, miR-660 is selected for the further investigation.Serum samples from 44 patients with NSCLC and 22 healthy individuals were collected for discovering the altered miRNAs, and another 300 serum samples (178 patients with NSCLC and 122 healthy individuals) were enrolled and divided into training cohort and test cohort for the validation. A total of 380 miRNAs were isolated using TaqMan(?) miRNA ABC Purification Kit, and the expression levels of these miRNAs were detected by RT-qPCR. We found miR-194, miR-652 and miR-660 overexpressed in the serum of NSCLC patients, and one of the miRNAs, miR-660 was also overexpressed in lung tumors compared with adjacent normal tissues, which may suggesting a potential functional role in lung tumorigenesis.In discovering cohort, three miRNAs (miR-194, miR-652 and miR-660) were selected from 380 miRNAs by two normalization methods. Then the levels of miR-652 and miR-660 were validated to be significantly upregulated in serums of patients with ADC and SCC compared with healthy controls both in training cohort and test cohort (p< 0.01). Next, the AUC of miR-652 (training cohort:0.819, test cohort:0.819) and miR-660 (training cohort:0.735; test cohort:0.714) were assessed. The combination of miR-652 and miR-660 showed significantly higher AUC than miR-660 for ADC and SCC diagnosis in training cohort and test cohort (p< 0.05). While compared with miR-652, the miR-652+miR-660 panel showed significantly higher AUC for NSCLC, ADC and SCC diagnosis only in training cohort (p< 0.05). Moreover, when compared with the existing protein biomarkers, the miR-652+miR-660 panel showed significantly higher AUC for ADC and SCC diagnosis than CEA and CA125 in both training and test cohort (p< 0.05). However, the Cyfra21-1 showed the comparable diagnostic ability with the miR-652+miR-660 panel. Furthermore, the diagnosis ability of combination of miR-652+miR-660 with Cyfra21-1 were assessed. The results showed that the combination of miR-652+miR-660 with Cyfra21-1 significantly improved diagnostic ability for discriminating patients with ADC from healthy controls (training cohort:AUC, 0.941; test cohort:AUC,0.942). However, for SCC diagnosis, the combination of miR-652+miR-660 and Cyfra21-1 showed comparable ability with Cyfra21-1.As for miR-660, transient inhibition of miR-660 using miRNA inhibitor reduced migration, invasion and proliferation properties in the Anip973 cell line. The inhibition of miR-660 icduced cell cycle G2/M arrest which leaded to the slow proliferation. We used bioinformatics method and identified a gene named PCDH9 to be the target of miR-660 and validated the results in the Anip973 cell line. PCDH9 is a member of the protocadherin superfamily which is frequently lost in many different cancer types. Our result showed that with the low level of miR-660, a high amount of PCDH9 protein was observed. And there was a signigicant inverse correlation between miR-660 and PCDH9 protein levels. The results suggested that high expression of miR-660 may promote cell proliferation, invasion and migration by modulating expression of PCDH9.Taken together, the combination of miR-652+miR-660 and cyfra21-1 was a novel biomarker for NSCLC diagnosis, especially for ADC. And miR-660 may inhibit the translation of PCDH9 mRNA by targeting its 3’UTR, which suggest that miR-660 may play an important role in lung cancer.