In the study, our results showed that miR-21 was up-regulated in plasma samples of human glioma tumors compared to healthy controls, whereas miR-128 and miR-342-3p were down-regulated. ROC analysis demonstrated the sensitivity and specificity of miR-21, miR-128 and miR-342-3p for GBM diagnosis. In order to further indentify the relationship between plasma level of the three miRNAs and classification and treatment effect of glioma, we next performed statistical analysis of our miRNAs expression data. There was a significant difference in plasma levels of miR-128 between the earlier stages (grade II) and the later subgroups (grade III and IV). Plasma level of miR-342-3p was notably decreased in glioma with ascending tumor grades. Expression levels of three miRNAs in plasma samples of patients treated reached levels comparable with control subjects. Additionally, the three miRNAs can specifically discriminate glioma from other brain tumor such as pituitary adenoma and meningioma.
MiRNAs were firstly discovered in 1993 when Lee et al. studied regulation of developmental timing in Caenorhabditis and reported a small RNA, lineage- definicient-4 (lin-4) . To date, more than 1 000 miRNAs in human have been discovered according to miRBase sequence Database Release 14 (http://www.mirbase.org/). MiRNAs represent approximately 1% of the eukaryotic transcriptome. They play key regulatory roles in a diverse range of pathway, including tumorigenesis and progression of cancer. Furthermore, variation of specific miRNAs in plasma offers the potential for detection, subtype and prognosis determination.
MiR-21 level is markedly elevated in human GBM tumor tissues [11–13]. It targets multiple components and plays an anti-apoptotic function in GBM. We found that miR-21 is significant higher in plasma of GBM patients than in controls, which is consistent with the finding of miR-21 with significant levels in CSF sample and tissue from patients with glioma [9, 11]. Furthermore, although circulating miR-21 is reduced in postoperation compared to preoperation, no significant difference existed. MiR-21 is observably decreased after further treatment with chemo-radiaton. Thus, these data suggest a possible association between miR-21 and treatment effect.
The expression level of brain-enriched miRNA-128 in glioma tissues is inversely correlated with tumor grade and function as a tumor suppressor . Similarly, we found that expression level of miR-128 in plasma of GBM patients was also decreased and negatively relevant to high and low grade glioma, just same as the tendency reflected in the test results of glioma tissues. But another research reported that miR-128 was up-regulated in peripheral blood of GBM patients . The reason may be that miRNAs contained blood cells cause the difference. Our data also revealed that miR-128 is up-regulated after glioma patients were treated, so miR-128 may be associated with curative effect.
To date, little is known whether miR-342-3p is dysregulated in glioma tissues and has an effect on glioma development. Roth et al. reported that miR-342-3p was down-regulated in peripheral blood of GBM patients . In the present study, our results also showed that the expression level of miR-342-3p is reduced in the plasma of glioma patients and also inversely correlated with glioma grade. In addition, we assessed the expression of miR-342-3p by real-time PCR in the group of patients who had been treated by operation and chemo-radiation. miR-342-3p is significantly increased and there are no differences between normal, control plasma and plasma sampling received therapies. All these results reveal that plasma-derived miR-342-3p may be a suitable biomarker which can function as diagnosis, classification and therapeutic effect.
The mechanism of origin of extracellular miRNAs remains to be fully elucidated. Some researchers have demonstrated that miRNAs in plasma are released from cells in membrane-bound vesicles which are named microvesicles (exosomes). These exosomes come from multivesicular bodies and are released by exocytosis and also can be shed by outward budding of the plasma membrane [18–21]. These early reports are confirmed by which cultured cells release exosomes containing miRNAs [22–24]. Similarly, one study has also demonstrated that microvesicles (exosomes) containing miRNAs are released from glioblastoma cells and the size of them is from 50 to 500 nm . The other two researches have revealed most of the extracellular miRNAs is bound to protein and form complexes rather than vesicles [26, 27]. Comprehensive previous researches, we preliminarily speculate that miRNAs in the plasma of patients with glioma derive from glioma cells because (1) blood brain barrier (BBB) is partly destroyed in patients with glioma; (2) exosomes or complexes may be through the BBB by unknown mechanisms. It is necessary to further investigate if microvesicles encapsulation is the only mechanism for miRNAs in plasma with glioma or if other potentially more predominant mechanisms exist. One interesting point we observed in our study and other studies is that the expression level of some miRNAs is different in different body fluids. For example, our results found that miR-15b in plasma doesn’t dysregulate, but another study has indicated that it is significantly increased in CSF from patients with glioma compared to samples from control patients . Because BBB exists, it is necessary to systematically explore the origin of plasma miRNAs of glioma patients and find the relationship between miRNAs of tumor cells and that of plasma.
In summary, our results demonstrate cell-free miR-21, miR-128 and miR-342-3p of plasma are specificity and sensitivity for diagnosis of GBM, suggesting that these miRNAs may be used as non-invasive biomarkers in GBM. Moreover, our data also find that particular miRNAs have a strong correlation with classification and clinical course and aid in therapeutic decisions for glioma patients through detecting plasma.