Cancer-Associated Fibroblasts Exosomal miR-106a Promotes Breast Cancer Invasion and Metastasis by Down -regulation of TCEAL7

Studies have shown that cancer-associated broblasts (CAFs) play an irreplaceable role in the occurrence and development of tumors. Therefore, exploring the action and mechanism of CAFs on tumor cells is particularly important for designing new and effective treatments and improving prognosis of tumors. For exosomes have been shown to play vital roles in intercellular communication, in this study, we compared the effects of CAFs-derived exosomes and NFs-derived exosomes on breast cancer cell proliferation, migration, and metastasis. The results showed that exosomes from both CAFs and NFs could enter into breast cancer cells and CAFs-derived exosomes had a more enhancing effect on breast cancer cell proliferation and invasion than NFs-derived exosomes. Furthermore, it was found that the expression levels of miR-106a in exosomes derived from CAFs were signicantly up-regulated than that of NFs-derived exosomes and what’s more, in vitro and in vivo studies have shown that miR-106a can promote breast cancer cell proliferation, migration and metastasis by specically binding to the 3'UTR of TCEAL7. It is inspiring to nd that the miR-106a-TCEAL7 pathway promotes Snail nuclear ectopic activation by activating NF-κB, thereby inducing epithelial-mesenchymal transition and promoting cell proliferation and metastasis. Moreover, a mouse xenograft model conrmed that CAFs-derived exosomes miR-106a could promote tumor metastasis. The above data shows that CAFs-derived exosomes miR-106a promote Snail nuclear ectopic by targeting TCEAL7 to activate the NF-κB pathway, thereby inducing EMT, invasion and metastasis of breast cancer. Targeting CAFs-derived exosome miR-106a may be a potential treatment option to overcome breast cancer progression. targeting miR-106a can to inhibit of inclusion cohort should be used future to prove the denite diagnostic miR-106a.


Introduction
Breast cancer is one of the common malignant tumors threatening women's health and recurrence and metastasis are the main causes of death in breast cancer patients [1]. In breast tumor tissues, epithelial tumor cells co-exist with different stromal cell types which create a microenvironment for tumor cells progression. In the past, most studies managed to overcome the initiation and deterioration of tumors by focusing on the cancer cells themselves, but in recent years, more and more evidences proved that the tumor microenvironment has a more signi cant impact on the development and metastasis of tumors.
As the main components of tumor stroma, CAFs account for 80% of the broblasts in breast tumors and similar to myo broblasts, CAFs are active fusiform or polygonal broblasts with highly heterogeneous and contractile characteristics [2]. CAF may affect not only cancer cells but also other stromal cells by releasing high levels of factors such as growth factors, cytokines, chemokines, and metalloproteinases [3][4][5] and thus CAFs can promote tumor progression and metastasis by providing mechanical support or by using paracrine, direct interaction and immune response regulation, as well as by enhancing the chemical resistance of tumor cells [6].
Exosomes are vesicle corpuscles with lipid bilayer structure secreted by various somatic cells and tumor cells [7]. With a diameter of about 40 ~ 100nm, exosomes contain miRNAs, RNAs, proteins and other small molecular substances and have attracted extensive attention as an important medium of intercellular communication. Studies have shown that bioactive molecules can be transferred from donor cells to recipient cells through exosomes, thus affecting the biological activity of cells including promoting the occurrence, development and metastasis of tumors [8][9][10].
MiRNAs are single-stranded non-coding RNAs of about 20 nucleotides in length. They can speci cally bind to post-transcriptional mRNA to regulate gene expression leading to degradation of target genes or inhibition of protein expression [11]. A large number of studies have shown that miRNAs can in uence the occurrence and development of tumors by regulating oncogenes or anti-oncogenes [12][13][14][15].
As a transcriptional regulator on X chromosome, Transcription elongation factor A-like 7 (TCEAL7) was rst discovered by Chien et al in 2005 through transcriptional cloning [16]. TCEAL7 is a member of transcription elongation factor A (SII) -like genes family that contain TFA domains and can act as nuclear phosphoproteins that regulate transcription in a promoter context-dependent manner [17]. The TCEAL7 gene encodes a cell death regulator protein containing 100 amino acids and is homologous in sequence with TCEAL1, TCEAL6 and brain-expressed (Bex) [18].
In this study, we found that miR-106a in CAF-derived exosomes promotes the invasion and metastasis of  horseradish peroxidase was incubated for 1 hour and the electrochemical luminescence was observed by chemiluminescence apparatus. Results WB quantization was performed by ImageJ software.

Electron Microscopy (EM)
The separated exosomes were xed with 4% paraformaldehyde and then dripped into the formaldehydecoated electron microscope grid and xed with 1% glutaraldehyde for 10 minutes. The sample was stained with 1% uranyl-oxalate solution for 5 minutes. Removing excess liquid and images were obtained using JEOL 1011 transmission electron microscope at 60kV.

Cell transfection
A lentiviral plasmid encoding miR-106a/mimics or inhibitor, TCEAL7 and its siRNA and negative control was designed and produced by Genechem (Shanghai, China). Lipofectamine 2000 reagent (Invitrogen, California, USA) was used to transfect breast cancer cells according to the instructions. GW4869 (Sigma, California, USA) was used to inhibit exosomal release.
RNA extraction and reverse transcriptase quantitative real-time PCR (qRT-PCR) All operations were performed according to the manufacturer's instructions. TRIzol reagent (Invitrogen; Thermo Fisher Scienti c, Inc) isolated total RNA from cells and exosomes, and the RevertAid First Strand cDNA synthesis kit (Thermo Fisher Scienti c, Inc) was used for cDNA synthesis. The expression of mature miRNAs was detected by real-time quantitative PCR (qPCR) using SYBR Premix (Perfect Real Time) kit. Transcripts of U6 act as internal parameters to standardize RNA input. Determining analysis and fold changes using the comparative threshold cycle (Ct) method.

Bioinformatics analysis
Meta-analysis of global gene expression data in the Oncomine database (Compendia Bioscience, Ann Arbor, USA) was performed using primary lters for "breast cancer" and "cancer vs normal analysis", data type lter to use "mRNA" data sets and sample lter to use "clinical specimens" (10 datasets representing 2941 patients). Patients include patients of all ages, genders, disease stage or treatment. Data were acquired in an unbiased manner by compiling all the Oncomine studies with signi cantly altered TCEAL7 expression at the threshold settings (P-value = 0.05, foldchange = 2, and gene rank = all). Signi cant studies, in which at least one analysed group was comprised of three patients or less were excluded. In the Oncomine database, all data are reported as log2 median-centered intensity. Export the dataset from Oncomine and analyze it in GraphPad Prism V7 software.

Migration and invasion
In vitro invasion experiments were performed by 6.5mm Transwell 24-well plates. Add 600 µL of RPMI-1640 medium containing 10% FBS in the lower chamber, and coat the upper chamber with 25 µL of Matrigel Basement Membrane Matrix (BD) and serum-free RPMI-1640 mixed solution (con gure scale bit 1: 6), and dry at 37°C at least 4 h. Then implant 120,000 cells in the upper chamber and incubate for 16 h in the incubator. Wash three times with PBS, x with 4% paraformaldehyde, and stain Giemsa for 15 minutes. Rinse off the staining solution with slow running water and leave to dry. Scanned under a microscope after sealing with resin.

DiD Labeling of Exosomes
Vybrant DiD (Life Technologies) was used to label exosomes according to the manufacturer's instructions [20]. In brief, DiD and PBS were diluted at 1:1000 and incubated at room temperature for 10 minutes. Exosomes were then re-puri ed using a qEV Size Exclusion Columns (Izon Science Ltd.).

Immuno uorescence analysis
The cells were xed with 4% paraformaldehyde, 0.5% TritonX-100 was in ltrated for 15 min, and washed three times with PBS. Blocking with BSA for 1 hour, and incubate at 4°C with the primary antibody for 1 hour in the dark, then incubate with Alexa Fluorescence 568 (1: 1000, Invitrogen) at 37°C for 1 hour in the dark with the secondary antibody, stained with DAPI (1: 5000, Sigma). Finally, the cells were observed with a confocal uorescence microscope.

Luciferase assays
The 3′UTR region (5′-GCUUUCUGUUUGUCUGCACUUUC-3′) of TCEAL7 might be targeted by miR-106a was screened by TargetScan software. 3′UTR terminal mutation reporting vector was designed and constructed by Shanghai GenePharma Co.,Ltd. According to the instructions of Lipofectamine2000, the luciferase reporter vector was used to co-transfected MDA-231 cells with miR-106a, miR-106a inhibitor, miR-106a mutant and control sequence, and then the cells were put into an incubator at 37°C and 5%CO 2 for 48 h. Luciferase activity was tested by detector.

Immunohistochemistry (IHC)
The tissue was embedded in para n xed with formalin, dewaxed with xylene and dehydrated with ethanol, and then extracted with 0.01mm citrate buffer (pH6.0). The Rabbit anti-SMA monoclonal antibody was incubated overnight at 4℃, and anti-rabbit secondary antibody was incubated at 37℃ for 30 minutes. Finally, it was developed with DAB and counterstained with hematoxylin. Taking photos under an optical microscope.

Immuno-precipitation (IP)
Immuno-precipitation using anti-ZRANB2 antibody was performed at 48 h after treatment. Brie y, cells were collected and lysed by the lysis buffer. Then Lysates were centrifuged at 14,000×g for 15 min. The supernatant was mixed with ZRANB2 antibody (A nity) overnight at 4°C, and then co-cultured with beads (santa cruz) for incubated at 4°C for 4 h. The beads were washed ve times in lysis buffer and then subjected to Western blot (WB) analysis.

In vivo tumour growth and metastasis experiment
The prepared 10 mice were divided into two groups. One group of MDA-231 cells were oating in 150 mL of PBS and injected into the mice through the tail vein, and the other group was injected with cancer cells co-cultured with exosomes miR-106a. Thirty days later, the tumor growth was monitored using bio uorescence imaging technology, and then the mice were sacri ced, and the lung tissue was removed for observation and photographing. All experimental procedures involving mice were performed in accordance with the Guidelines for the Care and Use of Experimental Animals and were approved by the Animal Care and Use Research Committee of Weifang Medica University.

Statistical analysis
All data were analyzed by SPSS 17.0 statistical software. Each experiment was repeated for 3 times, and the comparison of measurement data was conducted by independent sample t test and paired sample t test. Mean ± SD was used for quantitative data, and P < 0.05 was considered statistically signi cant.

CAFs promotes the migration and invasion of breast cancer cells
We rst isolated CAFs and NFs (Normal Fibroblasts) from BC tissues and adjacent normal tissues.
Western blot experiments were used to observe the expression of α-SMA, FAP and PDGFR-β in NFs and CAFs and the results showed that the expression levels of α-SMA and FAP in CAFs were higher than those in NFs, but there was no signi cant difference in the expression of PDGFR-β between the two groups ( Fig. 1a). Immuno uorescence further veri ed the higher expression of α-SMA and FAP in CAFs (Fig. 1b) and consistently, immunohistochemical staining showed that the expression of α-SMA and FAP in tumor tissues of breast cancer patients was much higher than that in adjacent normal tissues (Fig. 1c). To evaluate the effect of isolated CAFs on cancer cells, we treated MDA-231 breast cancer cells with NF-CM and CAF-CM of NFs and CAFs. The scratch test showed no signi cant difference between NF-CM groups and control groups whereas cells co-cultured with CAF-CM showed higher proliferation capacity (Fig. 1d). The invasion test also con rmed that the CAF-CM signi cantly improved the invasion ability of cancer cells than NF-CM (P < 0.05) (Fig. 1e). These ndings indicated that CAFs can promote the migration and invasion of breast cancer cells.

CAFs derived exosomes enhance migration and invasion of breast cancer cells
Exosomes, nanometric membrane vesicles secreted by almost all kinds of cell types, play an important role in intercellular communication. Since CAF-CM can promote the migration and invasion of breast cancer cells, we considered whether exosomes from CAFs contribute to this effect. We cultured MDA-231 cells in CAFs-CM with exosome depletion and CAFs-CM containing exosomes respectively and found that the conditioned medium containing exosomes signi cantly promote the migration and invasion of cells by transwell assay (P < 0.05) (Fig. 2a). Then exosomes were isolated from the CM of NFs and CAFs and electron microscopy showed that the isolated exosomes displaying typical lipid bilayer morphology ranging in size from 30 to 150nm (Fig. 2b-c). In addition, the exosome expressed exosome protein markers CD63 and Hsp70 (Fig. 2d) which con rmed that we got puri ed exosomes from both CAFs-CM and NFs-CM. CM-DiD labeled CAF-derived exosomes were co-cultured with MDA-231 cells, and distribution of CM-DiD in cells was analyzed by immuno uorescence which showed that exosomes could be uptaken into cancer cells (Fig. 2e). To evaluate the effect of exosomes from NFs and CAFs on cancer cells, MDA-231 cells were co-cultured with derived exosomes from NFs and CAFs respectively. Transwell assay showed that CAFs-derived exosomes could promote migration and invasion of MDA-231 cells more obviously than that of NFs-derived exosomes (P < 0.05) (Fig. 2f). Collectively, these data indicated that exosomes released by CAFs promote the migration and invasion of breast cancer cells.

The up-regulated expression of miR-106a in exosomes enhanced migration and invasion of breast cancer cells
Studies have shown that exosomes contain a variety of bioactive molecules such as miRNAs, and the expression pro le of miRNAs is similar to those of the parent cells [21]. The transfer of exosomes containing miRNAs between cells is considered to be a new important mechanism of genetic exchange. We applied miRNA microarray technology to identify differentially expressed miRNAs in exosomes derived from CAFs and NFs, of which 20 functional miRNAs in 196 exosomes miRNAs changed (13 upregulated and 7 down-regulated) ( Table 1). MiR-106a has been reported to be abnormally expressed in many tumors [22][23][24], therefore, it could be an important target for cancer treatment. Then we veri ed the expression level of miR-106a in NFs and CAFs derived exosomes by qRT-PCR, the results showed that the miR-106a expression in CAFs-derived exosomes is higher than that in NF-derived exosomes (P < 0.05) (Fig. 3a). As exosomes have been shown to be able to transport miRNAs to affect cell function, we examined the expression of miR-106a in cancer cells co-cultured with NFs or CAFs-derived exosomes and the results suggested that the expression of miR-106a in cancer cells co-cultured with CAFs-derived exosomes was higher than that of NF-derived exosomes(P < 0.05) (Fig. 3b). We further transfected MDA-231 cells with miR-106a mimics and inhibitors respectively and found that the miR-106a mimics signi cantly enhanced migration and invasion capacity of MDA-231 cells(P < 0.05) whereas the miR-106a inhibitors have the contrary effects(P < 0.05) (Fig. 3c-d). Interestingly, miR-106a has no effect on cell proliferation. These data emphasized the importance of miR-106a in CAFs derived exosomes in migration and invasion of breast cancer cells.

The packaging of miR-106a into exosomes is mediated by ZRANB2
In order to explore the mechanism of miR-106a being speci cally packaged into exosomes, we analyzed the speci c interaction between miR-106a sequence and RNA binding protein (RBPs) motifs through the RBP speci c database (RBPDB, http://rbpdb.ccbr.utoronto.ca/; threshold 0.7). The results showed that zinc nger ran-binding domain protein 2 (ZRANB2), aconitase 1 (ACO1) and muscleblind-like 1 (MBNL1) motifs have speci c miR-196a binding sites (Fig. 3e). Further research revealed that knockdown of ZRANB2 in CAFs by speci c siRNAs signi cantly reduced exosome miR-106a levels (P < 0.05) while intracellular miR-106a levels remained almost unchanged (P > 0.05) (Fig. 3f-g). This indicated that ZRANB2 plays a regulatory role of miR-106a in exosomes. The miRNA pull-down analysis suggested that the interaction between ZRANB2 and miR-106a can be observed in the cytoplasm and exosomes, but cannot be detected in the nucleus. However, ZRANB2 binding ability was impaired when the AGUGCU sequence of miR-106a was mutated (Fig. 3h). In addition, subsequent RNA immunoprecipitation (RIP) analysis of CAFs cells and exosome lysates revealed that miR-106a was enriched in the ZRANB2 antibody group compared to the IgG group (P < 0.05) (Fig. 3i). The above data demonstrated that ZRANB2 might play an important role in packaging miR-106a into exosomes by binding a speci c motif (AGUGCU) of miR-106a, thus ZRANB2-mediated miR-106a enrichment in exosomes might play an active role in breast cancer progression.

miR-106a reduce the expression of TCEAL7 by binding to its 3'UTR
MiRNA can regulate the expression of transcribed genes by binding to the 3'-UTR of target mRNAs, leading to degradation of mRNA or translational inhibition. To further disclose the mechanism of exosomes miR-106a promoting tumor cells invasion and metastasis, we predicted TCEAL7 as its candidate target gene through Targetscan (Fig. 4a). In addition, luciferase assay demonstrated that miR-106a was complementary to the 3'UTR region of TCEAL7(P < 0.05) (Fig. 4b-c) and RIP analysis MDA-231 transfected with miR-106a showed that mRNAs of TCEAL7 could be speci cally adsorbed by miRNP complex isolated by anti-AgoI antibody (P < 0.05) (Fig. 4d). ( RIP analysis showed that mRNAs of TCEAL7 from MDA-231 cells transfected with miR-106a could be speci cally adsorbed by miRNP complex isolated by anti-AgoI antibody (P < 0.05) (Fig. 4d)) PCR and Western blot results suggested that CAFderived exosomes suppressed the MDA-231 cells and MCF-7 cells expression of TCEAL7 (P < 0.05) ( Fig. 4e-g) after these cells were co-cultured with exosomes from NFs and CAFs, for 48hs. Then we cotransfected isolated exosomes into cancer cells with miR-106a mimics or inhibitors respectively, interestingly, the more reduced TCEAL7 expression was observed in the mimics group, and the reduced TCEAL7 expression was rescued in the inhibitor group (Fig. 4h). In accordance with above mentioned ndings, the expression of TCEAL7 in breast cancer tissues was lower than that in normal tissues, and decreased expression level of TCEAL7 was negatively correlated with pathological grade (Sample clinical information is shown in Table 2). Thus, miR-106a e ciently restrained TCEAL7 expression by directly binding to its 3'UTR in BC cells. To further verify the low expression of TCEAL7 in human breast cancers, a meta-analysis of publicly available gene expression data was performed using the Oncomine database. We compared TCEAL7 expression in BC with normal adjacent BC samples from ten datasets and found underexpressed TCEAL7 in BC samples (gene median rank: 445.5, P = 7.73e-9) in eight datasets included in the meta-analysis (Fig. 5a). Conformably, reduced TCEAL7 mRNA levels were also found in BC samples including invasive breast carcinoma, invasive ductal breast carcinoma, Ductal carcinoma in situ, invasive lobular breast carcinoma as compared with the corresponding normal breast tissues (P < 0.05) (Fig. 5b-e). In addition, Kaplan-Meier analysis on breast cancer patients strati ed by TCEAL7mRNA levels showed that low TCEAL7 mRNA level (probe 227705_at) is correlated with lower overall (Fig. 5f) survivals than patients with high TCEAL7 mRNA levels. Thus these in silico data suggest the reduced TCEAL7, level in human BC which is related to a worse prognosis. Collectively, we proved that miR-106a e caciously inhibited the expression of TCEAL7 in human BC through binding to its 3'UTR region. 3.6 TCEAL7 reverse EMT by suppressing the NF-κB pathway EMT refers to the transformation of epithelial cells into mesenchymal cells morphologically to obtain cytoskeletal reconstruction and enhanced cell mobility and EMT play key roles in mediating tumor cell invasion [25]. To determine whether CAFs is related to EMT of breast cancer cells, western blot was used to detect the expression of epithelial marker E-cadherin, mesenchymal markers N-cadherin and Vimentin in tumor cells after co-cultured with CAFs. The results showed that MDA-231 cells co-cultured with CAFs had the lower expression of E-cadherin and higher expression of N-cadherin and Vimentin than that of MDA-231 cells co-cultured with NFs and we got similar results from MCF-7 cells (Fig. 6a, right). We further veri ed that tumor cells co-cultured with CAFs-derived exosomes expressed lower E-cadherin and higher N-cadherin and Vimentin than that of cells co-cultured with NFs-derived exosomes (Fig. 6b). In addition, the results of immuno uorescence staining were consistent with those of western blot (Fig. 6c). As mentioned before, our results showed that CAF-derived exosomes decreased the expression of TCEAL7 in BC cells, to explore the effect of TCEAL7 on EMT, we up-regulated the expression of TCEAL7 in BC cells and found that the expression of E-cadherin was increased, whereas the expression of N-cadherin and Vimentin was decreased (Fig. 6a, left). Expression of E-cadherin, N-cadherin, and Vimentin was also validated using real-time PCR to show regulation at the transcriptional level (Supplementary Figure S1-2).
To further explore the upstream molecules that affected EMT, we measured the expression levels of several transcription factors in MDA-231 cells. It was found that the Snail protein but not the Snail mRNA was signi cantly up-regulated after co-cultured with CAFs-derived exosomes and the expression levels of other transcription factors (Zeb 1, Zeb 2 and Slug) were stable (Fig. 6d-e). Then we try to determine whether CAFs-derived exosomes was related to the Snail nuclear translocation event. As shown in Fig. 6F, the expression of Snail protein in the nucleus of MDA-231 cells co-cultured with CAFs-derived exosomes was signi cantly higher than that of control MDA-231 cells. Recent evidence revealed that TCEAL7 negatively regulates NF-kB pathway [26], as NF-κB could induce snail promoter activity [27][28][29], then MDA-231 and MDA-231/siTCEAL7 cells were treated with speci c inhibitors to determine which signaling pathway is involved in Snail stabilization. The results suggested that only NF-κB inhibitor (BAY11-7082) could signi cantly inhibit the stability of Snail (Fig. 6g). To analyze the effect of TCEAL7 on NF-κB activation, we transfected MDA-231, MDA-231/TCEAL7 and MDA-231/siTCEAL7 cells with NF-κB luciferase reporter gene plasmid and found that the luciferase activity of MDA-231/TCEAL7 cells was signi cantly lower than that of MDA-231 cells while that of MDA-231/siTCEAL7 cells was apparently higher than that of MDA-231 cells(P < 0.05) (Fig. 6h). Studies have found that activating the NF-kB pathway in the loss of TCEAL7 may be one of the mechanisms by which normal cells obtain proliferation and survival advantages. In order to study the mechanism of CAFs-exo regulation of TCEAL7 activity on invasion, and metastasis, we evaluated the expression levels of MMP9, and ICAM-1. The results showed that, compared with the control group, CAFs-exo up-regulated the expression of MMP9, and ICAM-1, and this effect was mimicked by the loss of TCEAL7. Overexpression of TCEAL7 reduces the expression of MMP9, and ICAM-1. In addition, overexpression of TCEAL7 can rescue the effects of CAFs-exo on MMP9, and ICAM-1. Expression of MMP9,ICAM were also validated using real-time PCR to show regulation at the transcriptional level. Taken together, these results indicate that the TCEAL7 involved in invasion, and metastasis progression by regulating the transcriptional modulation of NF-kB target genes involved in invasion, and metastasis and CAFs-exo can inhibit the role of TCEAL7 in this process. Together, these in vitro data indicate that TCEAL7 could reverse EMT by suppressing NF-κB pathway to suppress breast cancer cell invasion and metastasis.

Exosome miR-106a induces tumor invasion and metastasis in mice
Finally, we did experiments in mice to further evaluate the roles of exosome miR-106a in breast cancer metastasis and survival situation. MDA-231 cells, simultaneously with exosomes miR-106a or alone, were injected intravenously into the mice through tail vein respectively. After six weeks, we found that mice of exosomes miR-106a group had more metastases in lungs than that of control group. Colonization in lungs was examined through H&E staining. As showed in Fig. 7a, the results showed that the number of metastatic tumor nodules in the lungs of exosomes miR-106a group was more than that of the control group. Meanwhile, the survival rate of exosomes miR-106a group was signi cantly lower than that of the control group as shown in Fig. 7b(P < 0.05). Then we separated cancer cells from the metastases of mice lung tissues and detected miR-106a and TCEAL7 expression in lung metastatic cells using qRT-PCR, and western blot separately. The results showed that in exosomes miR-106a group the content of miR-106a was higher whereas the expression of TCEAL7 was signi cantly lower than that of the control group (P < 0.05) (Fig. 7c-d). Taken together, these ndings suggest that miR-106a-loaded exosomes had a cancerpromoting effect in the xenograft model. Finally, we provide a schematic diagram to reveal the biological role of transcellular signaling pathway, comprising exo-miR-106a, ZRANB2, TCEAL7, in promotes BC cells invasion and metastasis (Fig. 7e).

Discussion
Breast cancer is one of the malignant diseases that seriously threaten the life and health of women all over the world [30]. Tumor microenvironment is usually composed of tumor cells, extracellular matrix (ECM), stromal cells, immune cells and cells from the blood and lymphatic system [31]. The composition of TME and the intercellular communication affect tumor progression, prognosis and therapeutic effect [32]. Meanwhile, tumor cells can reprogram TME stromal cells to form the optimal tumor phenotype [33,34]. The difference between tumor stroma and normal stroma has been widely studied.
Studies have shown that NFs can inhibit tumor progression [35]. Normal broblasts can be transformed into CAFs after co-cultured with cancer cells and promote tumor progression through speci c communication with cancer cells [36,37]. CAFs can increase the proliferation of cancer cells by secreting stromal cell-derived factor 1 (SDF1) or produce ECM-degrading protease to affect cell invasion and movement [3,38,39]. This study compared the effects of CAFs and NFs on tumor cells and showed that CAFs promoted cell proliferation, migration and metastasis.
Exosomes are extracellular vesicles enclosed by lipid bilayers that carry molecules such as miRNAs and proteins and they play an important role in the communication between the tumor microenvironment and cancer cells. In this study, we isolated exosomes from the culture medium of NFs and CAFs to explore their role and potential mechanisms in the development of breast cancer. By co-culturing of exosomes with breast cancer cells, we found that CAFs-derived exosomes had stronger effects on cancer cell proliferation, migration and metastasis.
MiRNAs can perform biological functions by binding to 3'UTR of target genes. Volinia et al have found that miR-106a is one of the differentially expressed miRNAs in multiple tumor types [40]. You et al demonstrated that miR-106a promotes the proliferation, invasion and chemoresistance of breast cancer cells [41]. However, so far, the expression pattern and biological function of miR-106a in breast cancer progression and its underlying mechanism have not been fully elucidated. The results of this study indicate that miR-106a is signi cantly up-regulated in CAFs-derived exosomes and miR-106a overexpression can promote cell proliferation and motility. Further experiments demonstrated that exosomes miR-106a signi cantly promote tumor metastasis in vivo.

TCEAL7 (Transcription Elongation Factor A Like 7) is a member of the transcription elongation factor A
(SII)-like gene family and relatively poorly reported. TCEAL7 was cloned as a pro-apoptotic nucleoprotein for the rst time, and has been proved to have the function as an anti-oncogene, and a cell death regulator. In addition, previous studies found that TCEAL7 was down-regulated in a variety of human tumors including ovarian cancer [16], but its speci c roles in tumors have been rarely reported [42]. Consistent with that TCEAL7 is associated with a reduced risk of invasive serous ovarian cancer, we have illuminated that TCEAL7 is frequently downregulated in breast cancer malignancy and functions as a tumor suppressor.
NF-κB signaling pathway is widely activated in diverse human cancers and plays an important role in tumorigenesis and development [43]. Snail is recognized as a potent E-cadherin repressor and is a downstream target for NF-κB. The expression of Snail can be regulated by NF-κB through both transcriptional and post-translational mechanisms [44]. A previous study showed that, miR-210-3p maintains the continuous activation of NF-κB signaling by targeting the negative regulators of NF-κB signaling, TNIP1 and SOCS1, leading to EMT, invasion, migration and bone metastasis of prostate cancer cells [45]. Here, our data veri ed that TCEAL7 was negatively correlated with the expression of miR-106a in breast cancer tissues and was identi ed as the direct target gene of miR-106a in breast tumor cell lines. It is widely accepted that NF-κB induces a signi cant increase of snail expression which leads to a remarkable decrease of E-cadherin-mediated intracellular adhesion, EMT and metastasis/invasion of cancer cells. Consistently, our results revealed the role of NF-κB in the downstream pathway of miR-106a-TCEAL7 axis in breast cancer by demonstrating that TCEAL7 can inhibit NF-κB activation, thereby stabilizing Snail and ultimately inhibiting EMT. Interestingly, we clari ed that miR-106a may play a procancer role by inhibiting the expression of TCEAL7 protein and promoting the activation of the downstream NF-κB pathway.
In summary, our research indicates that the overexpressed miR-106a in CAFs-derived exosomes may promote the development and metastasis of breast cancer cells. MiR-106a can be transported through exosomes to facilitate cell-to-cell communication, leading to breast cancer invasion and metastasis.
Further study suggests that CAFs-derived metastatic exosomes miR-106a could activate NF-κB pathway by down-regulating TCEAL7 expression in breast cancer cells to promote malignant progression of tumors. The results of this study suggest that targeting miR-106a can be used as a new strategy to inhibit the development of breast cancer. Nevertheless, stricter inclusion criteria should be developed and a larger cohort should be used in future studies to prove the de nite diagnostic e ciency of miR-106a.

Declarations
Financial support: