Malignant glioma, the most common cancer of the central nervous system, accounts for around 70% of malignant primary brain tumors [26]. Patients with malignant glioma have a poor prognosis, with a median survival time of 3 years [27]. Hence, it is necessary to identify promising biomarkers to help diagnose gliomas early and to develop new therapeutic modalities with favorable prognosis. Recent evidence suggests that exosomes are an important communication medium that can transfer information from one cell to another and reprogram recipient cells [28, 29]. Exosomal miRNAs have reported to exert important physiological functions in tumorigenesis and development [17, 30]. In the present study, we identified seven differentially expressed miRNAs in infiltrating macrophages based on dataset GSE51332 from GEO database. Through qRT-PCR detection, miR-15a and miR-92a were found to be under-expressed in M2 macrophages, consistent with the expression levels in GEO. In addition, we confirmed the down-regulation of miR-15a and miR-92a in M2 macrophage exosomes. Subsequently, we demonstrated that M2 macrophage-derived exosomes promoted glioma migration and invasion in T98 and U251 cell lines, whereas exosomal miR-15a and miR-92a inhibited the migration and invasion of glioma cells. We thus speculated that miR-15a and miR-92a might inhibit the invasion and migration of gliomas.
According to bioinformatics analysis, we identified four target genes for miR-15a and one for miR-92a. We then selected CCND1 as the key target gene for miR-15a since only CCND1 was down-regulated at the transcript and protein levels after miR-15a was over-expressed. Similarly, RT-qPCR and western blot assays confirmed that miR-92a over-expression was associated with significant reduction in levels of RAP1B transcript and protein. Through dual luciferase reporter gene assays, we verified that CCND1 and RAP1B are the target genes of miR-15a and miR-92a, respectively.
CCND1, namely G1/S-specific cyclin-D1, regulates the cell cycle process during the transition from G1 to S phase. A study by Ciznadija D et al revealed that the down-regulation of CCND1 is characterized by reduced activation of cyclin-dependent kinases cdk4, and CCND1 and cdk4 are necessary for glioma progression and the development of surrounding stroma [31]. Increasing studies show that many genes affect the viability and migration of glioma cells by regulating CCND1. For example, Alqudah MA elucidated that via activation of CCND1 and EGFR, NOTCH3 promotes glioma cell proliferation, migration and invasion [32]. Similar to our results, the reduction of miR-17 in glioma cells improves cell viability and migration ability by increasing the expression of CCND1, p-Akt and Akt [33].
RAP1B encodes a member of the RAS-like small GTP-binding protein superfamily, which regulates a variety of cellular processes, including cell adhesion, growth, and differentiation [34]. Accumulating data indicate that the deregulated activation of RAP1B is related to a series of malignant tumors, and that RAP1B has effects on cell proliferation, metastasis, angiogenesis, and treatment resistance [35, 36]. In addition, numerous studies have reported that CCND1/RAP1B is associated with glioma cell proliferation and invasion [37–40]. Consistent with our findings, studies have shown that abnormal miRNA expression can affect gliomas by targeting RAP1B. For example, She X et al demonstrated that overexpression of miR-181 can inhibit the aggressive proliferation of glioblastoma cells by targeting RAP1B-mediated cytoskeletal remodeling and related molecular changes [41]. The expression of miR-128 and miR-149 was down-regulated in glioblastomas, and their over-expression inhibited glioblastoma invasion by targeting RAP1B-mediated changes in the cytoskeleton [42].
Futhermore, we demonstrated that knockdown of CCND1 or RAP1B could inhibit the PI3K/AKT/mTOR signaling pathway. Moreover, in the rescue experiment, we found that miR-15a and miR-92a could block the PI3K/AKT/mTOR signaling pathway. Overexpression of CCND1/RAP1B in miR-15a/miR-92a group can reverse the blocking effect of miR-15a / miR-92a on the signaling pathway, and rescue the phosphorylated protein in a certain extent. It has been demonstrated that many microRNAs as biomarkers for glioblastoma [43–45]. However, recent evidence suggests that microRNAs could migrate between cells and mediate repression of target genes [46]. In addition, miRNAs play a negative regulation and confer characteristic changes in the expression levels of target genes, and may regulate a variety of signaling pathways, and will generate integral effects on recipient cells [47]. To some extent, the above research provides new insights into the mechanism research. p-AKT and p-mTOR were reported to be activated or over-expressed in human gliomas, and the labeling index of the PI3K/AKT/mTOR pathway increased with increasing grade of malignancy [48, 49]. In gliomas, the PI3K/AKT/mTOR pathway helps to induce invasion and angiogenesis in cells, and patients with activated PI3K/AKT/mTOR pathway have a worse prognosis than those without carcinogenic activation of the pathway [50, 51]. In addition, BRCA1-associated proteins inhibit glioma cell proliferation and migration through the TGF- AKT/ PI3K/ mTOR signaling pathway [52]. Besides in gliomas, this pathway was involved in the pathological mechanism of other cancer. For example, Xu J et.al reported that exosome MALAT1 promotes the malignant behavior of CRC cells by activating the PI3K/Akt/mTOR pathway [53]. Similarly, the other article reported that Dhw-208 inhibits the growth of human breast cancer cells by inhibiting the PI3K/AKT/ MTOR signaling pathway [54]. BRCA1-associated proteins inhibit glioma cell proliferation and migration through the TGF- AKT/ PI3K/ mTOR signaling pathway [55].
To better understanding our study, we deduced a mechanism diagram of miR-15a and miR-92a (Fig. 8): M2 macrophages secrete miR-15a and miR-92a to glioma cells through exosomes, and then miR-15a and miR-92a separately bind to CCND1 and RAP1B, thereby blocking the PI3K/AKT/mTOR signaling pathway to inhibit glioma invasion and migration.
As is well known to all, the main function of M2-type macrophages is to promote the progress of tumor cells. For the first time, our findings elucidated that M2 macrophage-derived miR-15a and miR-92a could be secreted through exosomes to inhibit the invasion and migration of glioma cells. Combining bioinformatics analysis and experiments, CCND1 and RAP1B were found to be the target genes of miR-15a and miR-92a, respectively. To sum up, we unearthed that M2 macrophage-derived miR-15a and miR-92a could target CCND1 and RAP1B, respectively, thereby inhibiting the invasion and migration via blocking PI3k/AKT/mTOR signaling pathway in gliomas. Despite these promising results, the limitations cannot be ignored. In this paper, there is a lack of animal experiments and some rescue experiments for verification. In addition, whether miR-15a and miR-92a can affect the differentiation of macrophages. M1 macrophages mainly play an anti-tumor and pro-inflammatory role, while M2 macrophages chiefly promote tumor progression. If miR-15a and miR-92a can affect the proportion of these two phenotypes of macrophages, they will also play a role in glioma invasion and migration, which will be our future research direction.