miR‐3184‐3p enriched in cerebrospinal fluid exosomes contributes to progression of glioma and promotes M2‐like macrophage polarization

Abstract Liquid biopsy is a novel strategy for tumour diagnosis. The contents of cerebrospinal fluid (CSF) exosomes could reflect glioma status, hence sampling exosomes from CSF is a means of liquid biopsy for glioma. However, few studies have focused on the function of microRNAs in CSF exosomes. In this study, we found that miR‐3184‐3p was enriched in CSF exosomes in glioma patients and was downregulated after tumour resection. We found that miR‐3184 facilitates glioma progression in two ways. On the one hand, miR‐3184 directly promotes proliferation, migration, and invasion while inhibiting apoptosis in glioma. On the other hand, miR‐3184 in glioma‐derived exosomes polarizes macrophages to an M2‐like phenotype, which further aggravates tumour progression. Overall, the current findings uncovered a new mechanism and highlighted the significant role of miR‐3184 in glioma progression. Furthermore, exosomal miR‐3184 could be a considerable factor with potential applications in glioma diagnosis and treatment in the future.


| INTRODUC TI ON
Glioma is the deadliest malignant brain tumor in adults. 1 Glioblastoma (GBM), which is the most common type of glioma, 2 progresses rapidly, and the median survival for GBM patients is only 14 months.
According to the Stupp protocol, tumor resection followed by radiotherapy and temozolomide chemotherapy is the first-line treatment for GBM patients. 3 However, the effectiveness of this approach is still poor, partially due to a poor understanding of the mechanism of glioma progression.
Tumor-associated macrophages (TAMs) account for 30% of cells in glioma tissues and display an immunosuppressive effect in glioma. 4 TAMs enhance tumor progression via the secretion of protumor cytokines, exosomes, or anti-inflammatory factors. [5][6][7] We previously reported that glioma cells can produce periostin and facilitate the transition of macrophages to the TAM phenotype. 5 This immunosuppressive phenotype resembles that of alternately activated M2 macrophages, therefore we described this process of M2-like polarization in macrophages. 5 Whether there is another mechanism driving M2-like polarization in macrophages remains to be studied.
Exosomes are nanoparticles (with diameters between 50 and 150 nm) secreted by a variety of cells, including tumor cells and immune cells. 8 Large quantities of bioactive factors could be transferred cells to cells by exosomes, changing the gene expression in recipient cells. 9 It is believed that exosomes play a key role in mediating tumor progression and the exosomes in plasma and cerebrospinal fluid (CSF) could be used as diagnostic markers for glioma. 10 Aberrant expression of microRNAs (miRNAs) is a hallmark of neoplasms, including glioma. 11 We discovered a series of miRNAs related to glioma proliferation, invasion, mesenchymal transition, and TAM M2-like polarization. 6,7,12,13 Since exosomal RNA profiles in CSF reflect the disease stages of glioma, 14 we focused on exosomal miRNAs enriched in the CSF of preoperative glioma patients.
We proposed that the enriched miRNAs whose expression levels decreased after tumor resection were potential tumor-related miR-NAs. In the current study, we discovered that among these miRNAs, miR-3184 significantly promoted glioma progression and M2 macrophage polarization, indicating that CSF exosomal miR-3184 is a potential prognostic factor for glioma.

| Cell lines and cell culture
Human glioma cell lines (LN229 and U251) and THP-1 cells were obtained from the Culture Collection of the Chinese Academy of Sciences and cultured in Dulbecco's modified Eagle medium (DMEM; Thermo Fisher Scientific) or Roswell Park Memorial Institute 1640 (RPMI 1640; Thermo Fisher Scientific) with 10% fetal bovine serum (FBS; Gibco). The human primary GBM cell line P3 was a kind gift from Dr Rolf Bjerkvig, University of Bergen, and was cultured in neurobasal medium (Gibco) supplemented with GlutaMAX (2 mM), B-27 supplement (1×), and basic fibroblast growth factor and epidermal growth factor (20 ng/ml). These cell lines were maintained in a humidified chamber containing 5% CO 2 at 37 °C.

| Peripheral blood mononuclear cell isolation and induction of macrophages
Peripheral blood mononuclear cells (PBMC) were isolated from the healthy blood of volunteers as previously described. 5 The collected monocytes were cultured in RPMI 1640 complete medium supplemented with 100 ng/ml macrophage colony-stimulating factor (M-CSF, Peprotech) for 7 days to induce the M0 macrophage phenotype.

| Exosomes labelling
PKH67 (Sigma-Aldrich) was used to label exosomes as previously described. 15 To eliminate excess dye, PKH67-labelled exosomes were centrifuged at 100,000 g for 1 h, and the supernatants were discarded. The exosome pellet was diluted in 100 μl of PBS and then used for subsequent uptake experiments.

| Flow cytometry
To detect CD11b+ CD163+ macrophages, anti-CD163-PE (BD Biosciences) and anti-CD11b-APC (eBioscience) antibodies were used to stain cells according to the manufacturer's instructions. Cells were harvested, washed, and incubated in blocking buffer; isotype controls were run in parallel. Flow cytometry was performed using a BD Accuri C6 flow cytometer (BD Biosciences).  The exosomes from glioma cells were isolated from conditioned medium from glioma cell lines, and isolation was performed as previously described. 16 The exosomes were stored at −80 °C and verified by electron microscopy and nanoparticle tracking technology (Nanosight).

| Luciferase reporter assay
The reporter genes containing pGL3-PDCD4/RSAD2 and pGL3-mut PDCD4/RSAD2 were synthesized by BioAsia. LN229 cells or THP-1derived macrophages were cotransfected with luciferase reporters and miR-3184-3p mimics/inhibitors. Forty-eight hours later, the activity of the reporter protein was measured using a luciferase assay kit (Promega) according to the manufacturer's instructions.

| Western blotting
The harvested cells were lysed using heat denaturation in RIPA cell lysis buffer. The protein lysates were loaded and separated using SDS-PAGE and then transferred to PVDF membranes. The blots were incubated with primary antibodies targeting the following proteins which can see in Table S1. To visualize the protein bands, enhanced chemiluminescence (ECL; Millipore) was used. The intensity of the protein bands was analysed using ImageJ software and normalized to that of the GAPDH bands.

| Quantitative real-time PCR
Total RNA was isolated from glioma cells or macrophages using  Table S2. The mRNA expression was normalized to that of GAPDH to obtain the relative expression level.

| Statistical analysis
Data analysis was performed and data were visualized using GraphPad Prism. Each experiment was carried out at least in triplicate, and all results are presented as the mean ± SD. One-way ANOVA was used to assess statistical significance. Kaplan-Meier survival curves were also constructed, and log-rank tests in GraphPad Prism software were used to assess survival. Differences with the following p values were considered significant: P value <0.05, denoted by "*"; P value <0.01, denoted by "**"; P value <0.001, denoted by "***"; and P value <0.0001, denoted by "****." Difference with P values >0.05 were considered nonsignificant and are denoted by "ns".

| miR-3184-3p is enriched in glioma CSF-derived exosomes and promotes malignancy in glioma cells
First, we examined the exosomal miRNA expression profile in the CSF of glioma patients pre-(pre-CSF) and post-(post-CSF) tumor resection. We found that 123 miRNAs were downregulated after surgery ( Figure 1A, pre-CSF vs. post-CSF), suggesting that these miRNAs are probably important for glioma maintenance and diagnosis. In addition to fold change, the absolute expression abundance of miRNAs in exosomes should be considered. We therefore ranked the top 20 most highly expressed miRNAs in pre-CSF exosomes ( Figure 1A, exosomal miR) and found that four of these 20 miRNAs showed decreased levels after tumor resection. We further performed miRNA sequencing for glioma tissues and normal brain tissues. The 279 miRNAs upregulated in glioma tissues were selected ( Figure 1A, upregulated miR). The intersection of the abovementioned three groups of miRNAs is shown in a Venn diagram, and we finally chose the novel miRNA miR-3184-3p for further study ( Figure 1A). We further performed nuclear-cytoplasmic fractionation as well as FISH experiments in four glioma tissues and demonstrated that mir-3184-3p is localized in the cytoplasm ( Figure S1A,B). We then performed CCK8 and EDU assays and confirmed that proliferation was increased after miR-3184 overexpression in glioma cells ( Figure 1B-D). Overexpression of miR-3184 in glioma cells also decreased the apoptosis rate ( Figure 1E However, silencing miR-3184 induced the opposite effects on these factors ( Figure 1P). In summary, these findings suggest that miR-3184-3p is enriched in pre-CSF exosomes and promotes progression in gliomas.

| PDCD4 is a direct target and mediates the protumor effect of miR-3184
To identify the direct target of miR-3184, we screened the downregulated mRNAs after miR-3184 overexpression. We intersected these downregulated mRNAs with predicted miR-3184 targets from two widely accepted prediction websites, TargetScan and miRDB (Figure 2A), and the drastically downregulated mRNA PDCD4 was chosen for further study ( Figure 2B). We performed PCR and western blot assays to detect the changes in PDCD4 after miR-3184 overexpression and inhibition. We found that both the mRNA and protein levels of PDCD4 decreased after miR-3184 overexpression ( Figure 2C,D). Additionally, we performed IHC staining for PDCD4 in xenografted glioma samples. PDCD4 expression was restored in the miR-3184 silencing group, which confirmed our prediction were similar to those after miR-3184 overexpression ( Figure 3J).
These results demonstrate that PDCD4 is directly targeted by miR-3184 and mediates the protumor function of miR-3184.

| Exosomal miR-3184 induces macrophage M2like polarization
To confirm tumor cells secret exosomes containing miR-3184-3p, we collected exosomes from the supernatants of cultured GBM cells, which exhibited similar typical cup-shaped morphology, size, and number ( Figures 4F and S3B), and further confirmed their identity by detection of the exosome markers TSG101 and CD9 ( Figure 4G), indicating that we successfully isolated exosomes from GBM cells.
Furthermore, we collected exosomes from glioma cells transfected with miR-NC, miR-3184, sh-miR-NC, and sh-miR-3184. The miR-3184 expression level was examined by PCR. We confirmed that the expression level of miR-3184 in exosomes was concomitantly changed with cell transfection ( Figure S3D). We also found that the expression of miR-3184-3p in exosomes was downregulated in cells treated by neutral sphingomyelinase-2 (nSMase) GW4869, which blocks exosome formation ( Figure S3E), thus confirming the existence of mir-3184-3p in exosomes. We wondered whether exosomal miR-3184 could alter the immune microenvironment in glioma. We have previously reported that glioma cell-derived exosomal miR-1246 promotes macrophage M2-like polarization. 6 We therefore examined the role of glioma cell exosomal miR-3184 on macrophages.
First, we tested the effects of miR-3184 on macrophages. We found In addition, other factors related to M2-like polarization were regulated by exosomal miR-3184 ( Figure 4J). We further confirmed the abovementioned findings in vivo. According to our previous study, 6 we overexpressed miR-3184 in macrophages and coinjected these cells with glioma cell lines into nude mice. We found that overexpression of miR-3184 in macrophages significantly promoted tumor growth in vivo ( Figure 4K,L) and shortened the survival times of mice ( Figure 4M). The results of IHC staining also showed that miR-3184 overexpression in macrophages increased the cell proliferation rate in xenografted glioma tissue ( Figure 4N). Furthermore, we studied the role of macrophage transfected with miR 3184-3p in the proliferation, invasion, and migration of glioma cells using the in vitro coculture system. The CCK8 and EDU assay results showed that glioma cells cocultured with miR-3184-3p overexpressing macrophages have better proliferation ability ( Figure S4A,B). The invasion and migration assays demonstrated that macrophages transfected

| miR-3184 promotes macrophage M2-like polarization via RSAD2
To identify the downstream target of miR-3184, we performed mRNA sequencing for macrophages overexpressing miR-3184. We then intersected the downregulated genes with predicted targets for miR-3184 in the TargetScan website ( Figure S4E). The gene RSAD2 was finally found ( Figure 5A), of which the 3′ UTR contained potential binding sites for miR-3184 ( Figure 5B). We cotransfected RSAD2 and miR-3184 in macrophages. Although RSAD2 transfection significantly increased the expression of RSAD2, cotransfection of RSAD2 and miR-3184 attenuated this increase ( Figure 5C).
Then we performed a luciferase reporter assay to further confirm the direct binding of miR-3184 with RSAD2 ( Figure 5D). We confirmed that RSAD2 overexpression attenuated miR-3184-induced M2-like polarization in macrophages via flow cytometry, ELISA, and PCR assays ( Figure 5E-H). In addition, knockdown of RSAD2 in macrophages promoted the M2-like phenotype, further suggesting that the role of miR-3184 is mediated by RSAD2 ( Figures 5I-L and S4F).
It is widely accepted that the NF-κB pathway plays an important role in the M2-like polarization of macrophages, 17 thus we examined the protein levels of p65 and p-p65 in miR-3184-altered cells. The results showed that overexpression of miR-3184 or knockdown of RSAD2 inhibited the activation of p65, and p-p65 expression was rescued when miR-3184 and RSAD2 were co-overexpressed in glioma cells ( Figure 5M). Moreover, miR-3184-overexpressing gliomaderived exosomes decreased the protein expression of RSAD2 and inhibited the phosphorylation of p65 ( Figure 5N). In summary, we confirmed that glioma exosome miR-3184 promotes macrophage polarization via RSAD2.

| DISCUSS ION
In the current study, we revealed the glioma-promoting role of miR- MiR-3184 is a recently defined miRNA regulating tumor malignancy. 22 In the current study, we first found that miR-3184 promotes the progression of glioma cells. These protumor effects were mediated by the targeting of PDCD4. Furthermore, we proved that miR-3184 could be transferred to macrophages via exosomes and induce M2-like polarization of macrophages by targeting RSAD2.
Programmed cell death 4 (PDCD4) is a tumor suppressor whose expression is frequently downregulated in various types of cancers.
PDCD4 functions as a protein translation inhibitor in tumors and thereby blocks pathways involving cell proliferation, survival, and invasion. 23 In the current study, we confirmed the antitumor role of PDCD4 in glioma. We found that knockdown of PDCD4 induced the activation of factors in the Wnt-b-catenin pathway and EMT process. These results were in accordance with previous findings that PDCD4 is a suppressor of EMT in tumors. [24][25][26] In Figure 5A, when overexpressing mir-3184-3p, we found the changes of many genes related to angiogenesis and immunosuppression. 27 We determined that Radical S-adenosylmethionine (SAM) domain-containing protein 2 (RSAD2) downregulated is one of the most significant genes and may be the directly downstream targets of miR-3184-3p. RSAD2 is a key enzyme in innate immune responses that is highly expressed in response to viral infection and inflammatory stimuli in many cell types. 28 Among human THP-1-induced macrophages, RSAD2 is upregulated in M1 subtype macrophages, indicating a correlation between RSAD2 and M1-like polarization. 29 Our study found that inhibition of RSAD2 by miR-3184 in macrophages induced M2-like polarization.
In summary, we uncovered a novel interaction between glioma cells and TAMs that enhances tumor progression. miR-3184, which functions as a key regulator during this process, is probably a predictive factor for glioma progression. However, there are some limitations to this study. The diagnostic ability of miR-3184 should be further assessed, and the detailed downstream pathway by which miR-3184-PDCD4 or miR-3184-RSAD2 induces tumor progression needs further study.

ACK N OWLED G M ENTS
We thank the surgeons and patients who participated in these studies, and Novogene Co. Ltd and Tianjin Novogene Bioinformatics Technology Co. Ltd for miRNA sequencing and mRNA sequencing technical development and support.

D I SCLOS U R E
The authors declare that they have no competing interests. were responsible for clinical sample collection and subsequent sample delivery. X.G., L.D., and H.X. helped to revise the manuscript. All authors read and approved the final manuscript.

E TH I C S A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
The research protocol was reviewed and approved by the Ethical

Committee on Scientific Research of Shandong University Qilu
Hospital (approval number: KYLL-2018-324), and written informed consent was obtained from each patient included in the study. The patient data were acquired from publicly available datasets, which contained complete informed consent information for the patients.

A N I M A L E XPER I M ENTS
All procedures that involved mice were approved by the Animal Care and Use Committee of the Qilu Hospital of Shandong University.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data used in this work can be acquired from the TCGA database