EBV-miR-BART12 accelerates cancer cell migration and invasion through targeting TPPP1 in EBV associated cancer

Background: Epstein-Barr virus (EBV) is suggested to actively utilize its EBV microRNAs (miRNAs) to manipulate viral and cellular functions during neoplasia transformation. But its role in tumor development and maintenance remains unclear. The objective of this study is to examine EBV-miRNA-BART12 implicated in EBV associated carcinogenesis. Methods :Expression of EBV-miRNA-BART12 and TPPP1 (Tubulin Polymerization Promoting Protein 1) was confirmed by GEO datasets, RT-PCR and in situ hybridization. Luciferase analyses, RT-PCR and western blotting experiments examined and confirmed that EBV-miRNA-BART12 targeted the 3'-UTR of TPPP1 mRNA. Migration and invasion ability were measured by wound healing and transwell assay. And the mechanism was assessed by western blotting and immunofluorescence. Results : Our results showed that EBV-miRNA-BART12 was highly expressed and TPPP1 was lowly expressed in NPC and/or GC and tightly associated with these patients' prognosis. Additionally, EBV-miR-BART12 advanced the HDAC6 (Histone Deacetylase 6) activity in cells by inhibiting TPPP1, and stimulated acetylation of α-tubulin and β-catenin, thereby increasing the microtubule instability and activating the epithelial-mesenchymal transition (EMT) process. Conclusion : EBV-encoded BART12 promotes migration and invasion of EBV associated NPC and gastric cancer by inhibiting TPPP1 mRNA and promoting the microtubule instability and the cellular EMT process. ability co-transfection EBV-miR-BART12 mimics TPPP1 examined the expression of α-tubulin and acetylated α-tubulin in cells with paclitaxel western blotting.

mRNA. Migration and invasion ability were measured by wound healing and transwell assay. And the mechanism was assessed by western blotting and immunofluorescence.
Results : Our results showed that EBV-miRNA-BART12 was highly expressed and TPPP1 was lowly expressed in NPC and/or GC and tightly associated with these patients' prognosis. Additionally, EBV-miR-BART12 advanced the HDAC6 (Histone Deacetylase 6) activity in cells by inhibiting TPPP1, and stimulated acetylation of αtubulin and β-catenin, thereby increasing the microtubule instability and activating the epithelial-mesenchymal transition (EMT) process.
Conclusion : EBV-encoded BART12 promotes migration and invasion of EBV associated NPC and gastric cancer by inhibiting TPPP1 mRNA and promoting the microtubule instability and the cellular EMT process.
We previously profiled all 44 EBV-encoded mature miRNAs in NPC biopsies and noncancerous nasopharyngeal tissues and found that EBV miRNAs located in the BART region were highly expressed in NPC biopsies. EBV-miR-BART12 was highly expressed in NPC samples compared with non-tumor nasopharyngeal epitheliums (NPE). But its function in NPC is still unclear.
Tubulin polymerization promoting protein 1 (TPPP1) is a microtubule (MT) regulatory protein that drives MT polymerization and stabilization. It encodes 25-kDa protein and belongs to the tubulin polymerization promoting protein family that includes TPPP2 (p20) and TPPP3 (p18) (20,21). Overexpression of TPPP1 in cells promotes MT polymerization and an increase in MT acetylation, whereas introduction of TPPP1 RNAi reduces MT acetylation. It binds to tubulin heterodimers and facilitates their incorporation into the growing microtubule filaments. It also binds to histone deacetylase 6 (HDAC6), a major MT deacetylase, and inhibits its activity leading to increased MT acetylation, a polymer stabilizing modification (22).
In this study, we tried to explore the function of EBV-miR-BART12 through TPPP1 5 gene on the microtubule network and the EMT process in three different cell lines.
Our results suggested that TPPP1 was a direct target of EBV-miR-BART12. EBV-miR-BART12 decreased acetylation of the microtubule network and affected the stability of the microtubule network through TPPP1, as well as decreased acetylation of βcatenin and facilitated the EMT process, therefore resulting in increased cell migration and invasion. Synthetic EBV-miR-BART12 mimics or inhibitors were products of Ruibo Company (RiboBioCo., Guangdong, China). EBV-miR-BART12 inhibitors were chemically synthesized, single-stranded, modified RNA molecules that can specifically inhibit endogenous target EBV-miR-BART12 miRNA. Both were worked at the concentration of 50 nM. Full-length cDNA of TPPP1 gene was amplified by PCR and constructed by inserting a PCR product into IRESneo3 vector. TPPP1 siRNAs were obtained from Genepharma (Genepharma, Shanghai, China) and worked at 100 nM concentration.

Clinical samples
The luciferase reporter was established by inserting synthetic oligonucleotides containing either the wild-type EBV-miR-BART12 binding site in TPPP1 3'-UTR (TPPP1-WT) or the one with mutant-binding sites (TPPP1-MT) into the pMIR-Report luciferase vector (Ambion, Austin, TX,). The sequences of the synthetic oligonucleotides are listed in the Additional file 1. Transfection of plasmids and miRNAs was performed with Lipofectamine 3000 (Invitrogen, the Nederlands) or Hipefect (Qiagen, Hilden, Germany) as recommended.

GEO data
Multiple gene chips (GSE12452, GSE36682, GSE32960, GSE13195, GSE38749, and GSE26253) were downloaded from NCBI's GEO datasets for data analysis (24)(25)(26)(27). We performed differential analysis of the data by the SAM software to select differentially expressed genes (28), and analyze the mRNA expression, the correlation between the expression level and the survival time by the GraphPad 7 Prism software.

Wound healing and transwell assay
Wound healing: When the cells were grown to 90% confluence after transfection, a 10 μl pipet tip was used to create a scratch in the cell monolayer. Images of the Wuhan, China). After washing, the blots were incubated with horseradish peroxidase-conjugated anti-rabbit secondary antibodies (Santa Cruz, Heidelberg, Germany) at a dilution of 1:2000 for 1 h at room temperature. Blots were visualized by exposure to X-ray film, through an enhanced chemiluminescence detection system (Millipore, Darmstadt, Germany), and quantified by densitometry using the Image J software (http://rsb.info.nih.gov/ij). GAPDH served as an endogenous control for equal loading and Histone H3 served as a nucleus control.

In situ hybridization (ISH) and immunohistochemistry (IHC)
ISH and IHC staining were performed as previously described (29). Paraffin sections were first dewaxed, then paraffin was removed from the turpentine, and the turpentine was washed with gradient alcohol and the slices was hydrated. For ISH, tissue sections were first inactivated by H 2 O 2 to inactivate endogenous peroxidase, fixed by 4% paraformaldehyde, digested by pepsin diluted with 3% citric acid, and then pre-hybridized and hybridized, followed by blocking, and biotinylated rat antidigoxin, SABC(strept avidin-biotin complex) and biotinylated peroxidase were added dropwise to tissues, final DAB (Diaminobenzidine) color development, hematoxylin counterstaining, gradient alcohol dehydration, neutral resin sealing. For IHC, the tissue sections were firstly subjected to antigen retrieval, and then 3% H 2 O 2 was used to inactivate endogenous peroxidase. After blocking for 30 min, the TPPP1 primary antibody was incubated overnight. After rewarming at room temperature, the secondary antibody was incubated for 30 min, DAB was colored, and hematoxylin was counterstained. Finally, neutral resin seals the tissues. Sequences of EBV-miR-BART12 probe could be seen in Additional file 1.

Animal experiments
Four-week-old male BALB/c nude mice with weight 20±2g were purchased from and raised at the Laboratory Animal Services Centre of Central South University with specific pathogen free. Animal handling and experimental procedures were approved by the Animal Experimental Ethics Committee of Central South University. Statistics analyses 11 Statistical analysis was performed using software of GraphPad Prism 5 (GraphPad, LaJolla, CA). Student's t-tests were used to evaluate significant differences between any two groups of data. One-way ANOVA was used when there are more than two groups. Survival estimates overtime were calculated using the Kaplan-Meier method, and the differences were compared using the log-rank test. The results of the analysis were considered significant in a log-rank test if P < 0.05.

Results
The expression of EBV-miR-BART12 and TPPP1 was associated with the poor prognosis of NPC or GC patients.
First, the NPC online GEO data (GSE36682 and GSE32960) were used to analyze the expression level of EBV-miR-BART12, which was highly expressed in the NPC tissues NPCs and 13 normal nasopharyngeal epitheliums by qRT-PCR. EBV-miR-BART12 was highly expressed in these NPC clinical samples ( Figure 1B). The expression of EBV-miR-BART12 was also associated with the poor prognosis in these NPC patients, in which high EBV-miR-BART12 expression patients had a lower survival rate in GSE36682 ( Figure 1C).
The bioinformatics software was used and predicted that TPPP1 gene was the potential target of EBV-miR-BART12 (Additional file 2). The GEO data GSE12452 (NPC) and GSE13195 (GC) showed that the expression of TPPP1 in cancers was lower than the non-tumor tissues ( Figure 1D). The qRT-PCR data also showed TPPP1 was lowly expressed in 34 NPC tissues compared with 24 NPE tissues ( Figure 1E). The expression of TPPP1 was tightly associated with the survival time of GC patients in GSE26253, the low TPPP1 expression patients with the short survival rate ( Figure   1F). The ISH and IHC results also show that EBV-miR-BART12 was highly expressed in NPC, while TPPP1 is just found in NPE part ( Figure 1G). Then we predicted through the bioinformatics software that TPPP1 was one of the targets of EBV-miR-BART12 and found that TPPP1 was also down-regulated in NPC of GSE12452 ( Figure 1H, We concluded that the expression of EBV-miR-BART12 and TPPP1 might be correlated with the survival rate of NPC or GC patients. Metastasis is one of the reasons for the death of cancer patients. Therefore, we speculate that EBV-miR-BART12 may affect the migration and invasion of EBV associated tumors by targeting the TPPP1, which plays an important role in the development of malignant tumors. EBV-miR-BART12 inhibits TPPP1 expression through directly targeting TPPP1 3'-UTR According to the above results, we tried to confirm if EBV-miR-BART12 targets the 3'-UTR of TPPP1 mRNA. Firstly, the EBV-miR-BART12 chemical synthesis analogues (EBV-miR-BART12 mimics) were transfected into EBV negative NPC 5-8F and GC AGS cell lines. EBV-miR-BART12 was overexpressed in the experimental group, while it was not detected in the control group (Figure 2A). Meanwhile, EBV-miR-BART12 inhibitors were also transfected into EBV positive NPC C666-1 cell lines and the qRT-PCR results showed that the inhibitors reduce the expression of EBV-miR-BART12.
Then the expression of TPPP1 in 5-8F, AGS, C666-1 was also detected by qRT-PCR after transfection of EBV-miR-BART12 mimics or inhibitors. The TPPP1 mRNA was reduced after EBV-miR-BART12 mimics or induced after the inhibitor's transfection ( Figure 2B). Western Blotting also confirmed that the expression of TPPP1 was inhibited by EBV-miR-BART12 mimics in 5-8F and AGS cell lines and promoted by EBV-miR-BART12 inhibitors in C666-1 cell lines ( Figure 2C). In order to detect whether EBV-miR-BART12 directs to target TPPP1, the bioinformatics software was unchanged. This showed that EBV-miR-BART12 mimics play its inhibitory effect by directly targeting the 3'-UTR of TPPP1 ( Figure 2E). EBV-miR-BART12 inhibitors increased the luciferase activities of TPPP1-WT but no effect in TPPP1-MT ( Figure   2E). These experimental results indicated that EBV-miR-BART12 directly combined with the 3'-UTR of TPPP1.

EBV-miR-BART12 promotes migration and invasion of cancer cells by inhibiting TPPP1
Metastasis is an important feature of tumor cells. We speculate that EBV-miR- Similar results were also seen in AGS ( Figure 3A and B). The TPPP1 overexpression vector reversed the ability of EBV-miR-BART12 mimics after co-transfection of EBV-miR-BART12 mimics and the TPPP1 overexpression vector. These results suggested that EBV-miR-BART12 promoted tumor cells migration and invasion through downregulating TPPP1. In C666-1, EBV-miR-BART12 inhibitors (BART12 In), the TPPP1 overexpression vector (TPPP1), TPPP1 siRNA (siTPPP1), or EBV-miR-BART12 inhibitors combined with TPPP1 siRNA (BART12 In + siTPPP1) were also transfected.
The results of scratch test and the transwell assay showed that EBV-miR-BART12 inhibitors reduced the ability of migration and invasion. Overexpression of TPPP1 also inhibited and siTPPP1 boosted the migration and invasion of C666-1( Figure 3A and 3B). We conclude that EBV-miR-BART12 promotes the migration and invasion of cancer cells by inhibiting TPPP1.
EBV-miR-BART12 impacts the microtubules dynamic instability by affecting TPPP1. TPPP1 is a microtubule related protein which affects the dynamics and stability of the microtubule network through interacting with microtubules. On the one hand, TPPP1 promotes tubulin polymerization and microtubule bundles by its 15 polymerization activity, on the other hand, TPPP1 influences microtubules acetylation level through binding with microtubule deacetylase. In order to explore whether EBV-miR-BART12 affects the microtubule dynamic stability through targeting TPPP1, the microtubule network morphology was detected in five groups, including control, BART12, TPPP1, siTPPP1, BART12 + TPPP1 group ( Figure 4A). The results showed that TPPP1 mainly distributed in cytoplasm. The green fluorescence representing the TPPP1 expression was significantly lower than the control group in the cells after EBV-miR-BART12 mimics transfection, indicating that TPPP1 expression was significantly decreased, and the content of tubulin represented by the red fluorescence was also reduced. Tubulin gathered around the nucleus and cytoplasm was scattered after EBV-miR-BART12 mimics transfection and was no longer the rules of radial distribution which was in the control group. After over expression of TPPP1, the green fluorescence value of TPPP1 was also significantly increased. The content of TPPP1 on the edge of cells increased, and the expression of microtubule protein increased correspondingly. The morphological changes of the cells suggested that EBV-miR-BART12 might be associated with apoptosis and metastasis. The cells had shrunken to a smaller circle. The morphological changes and infection of the group siTPPP1 were like that of the group BART12 ( Figure 4A).
In addition, western blotting also showed that the acetylated α-tubulin was inhibited by EBV-miR-BART12 mimics and siTPPP1 and induced by TPPP1 ( Figure 4B), which hinted that EBV-miR-BART12 reduced the acetylation through targeting TPPP1.
Immunofluorescence showed that acetylated α-tubulin and TPPP1 were colocalized with each other in the control group. After transfection of EBV-miR-BART12 mimics, the expression of TPPP1 significantly decreased compared to that of the control group. And the expression and distribution of acetylated α-tubulin was significantly different from the control group. In the EBV-miR-BART12 group, the distribution of acetylated α-tubulin decreased, while the density increased, and the cell shrinkage became smaller and round. After overexpression of TPPP1, the expression of TPPP1 in cells was significantly higher than that in control cells, with higher density and full in the cytoplasm, while the morphology of acetylated α-tubulin was more complete. The results for the EBV-miR-BART12 + siTPPP1 group were like those in the TPPP1 group, while the results of the siTPPP1 group were like those in the EBV-miR-BART12 group ( Figure 4C). From these results, we concluded that EBV-miR-BART12 affected the distribution, expression and acetylation degree of α-tubulin by downregulating the expression of TPPP1. This affects the microtubule network dynamic stability.
EBV-miR-BART12 influences the stability of microtubules by downregulating of TPPP1.
Acetylation of microtubules is an important indicator for evaluating microtubule stability. The stable microtubules usually have a higher degree of acetylation. In order to observe the distribution and expression of acetylated α-tubulin in cells more clearly, immunofluorescence was used to observe the expression and distribution of α-tubulin and acetylated α-tubulin in cells to explore the effect of EBV-miR-BART12 on the stability and dynamics of microtubule. In these experiments, the HDAC6 inhibitor TSA was used to treat cells. In the control group, α-tubulin and the acetylated α-tubulin were partly co-localized in the perinuclear region. When the 5-8F cells were treated with EBV-miR-BART12 mimics, the luminance of acetylated α-tubulin was lower than that of the control group. After the entry of TSA, the brightness of acetylated α-tubulin increased, and cold treated TSA cells resistant this increasing ( Figure 5A).
Similarly, western blotting experiments were performed to observe the changes of acetylated α-tubulin after treated with or without TSA on 5-8F and AGS cell lines transfected with EBV-miR-BART12 mimics (BART12) or C666-1 with EBV-miR-BART12 inhibitors (BART12 In), the TPPP1 overexpression vector (TPPP1) and TPPP1 siRNA(siTPPP1). It was found that TSA increased the acetylation of α-tubulin to a higher level in negative control cells, and the expression of acetylated α-tubulin protein in the control group was higher than that of TSA-treated BART12 group and lower than that of the TSA-treated BART12 In group. The expression of acetylated αtubulin protein in the siTPPP1 group was consistent with the BART12 group. In the TPPP1 group, the addition of TSA further encouraged acetylation of α-tubulin ( Figure   5B). In summary, TSA, an inhibitor of HDAC6, is resistant to microtubule depolymerization by cold treatment, which is produced by acetylated α-tubulin.
EBV-miR-BART12 reduced the content of acetylated α-tubulin. Therefore, EBV-miR-BART12 promoted the increase of HDAC6 content in cells through down-regulating the expression of TPPP1, leading to the down-regulation of microtubule acetylation in cells and the decrease of microtubule network stability and anti-depolymerization ability, which causes the migration and invasion of tumor cells. Then HDAC6 specific inhibitor Tubacin (Tub) was also used to treat cells and the results was consistent with TSA ( Figure 5C). Paclitaxel (PTX) is a famous antitumor chemotherapeutic drug, which promotes the assembly of free tubulin subunits into microtubules and facilitates microtubule stabilization by binding to microtubules (30). To test whether paclitaxel influences EBV-miR-BART12-associated NPC and GC or not, we examined the expression of αtubulin and acetylated α-tubulin in cells treated with paclitaxel by western blotting. 18 The results showed that paclitaxel did cause an increase in the protein expression levels of α-tubulin and acetylated α-tubulin. However, EBV-miR-BART12 group cells, with paclitaxel treatment, had little effect on the expression level of α-tubulin, but it greatly increased the expression level of acetylated α-tubulin ( Figure 5C). The results indicate that paclitaxel does not have a particularly good effect on the reduction of microtubules caused by EBV-miR-BART12 but can greatly enhance the stability of the remaining microtubule network. It is suggested that clinical treatment with paclitaxel will have a certain effect on the progression of NPC and GC. In summary, EBV-miR-BART12 promotes deacetylation of microtubules and increases microtubule instability by binding and down-regulating the expression of TPPP1.
EBV-miR-BART12 induces the EMT process by restraining the inhibitory effect of TPPP1 on β-catenin TPPP1 promotes the degradation of β-catenin by inhibiting the activity of HDAC6 (31).So we examined if the expression of acetylated β-catenin was regulated by EBV-miR-BART12 through TPPP1 gene in 5-8F, AGS, and C666-1 cell lines. The results showed that EBV-miR-BART12 reduced acetylation of β-catenin through TPPP1 gene ( Figure 6A). So, 5-8F, AGS, and C666-1 cell lines were treated with the HDAC6 inhibitor TSA or Tub after EBV-miR-BART12 mimics or inhibitors transfected into cells to detect whether EBV-miR-BART12 regulate the expression of acetylated β-catenin through interacting with TPPP1 and HDAC6. The western blotting results showed that TSA increased the expression of total β-catenin and reduced the protein level of acetylated β-catenin. EBV-miR-BART12 decreased the expression of acetylated β-catenin. After adding TSA or Tub to medium, the expression of acetylated β-catenin increased in the EBV-miR-BART12 mimics group, but it was still lower than that of the control group ( Figure 6B -6E). The results showed that the activity of HDAC6 in the cells was inhibited by TSA, and the effect of EBV-miR-BART12 on β-catenin and acetylated β-catenin was inhibited. On the other hand, it was indicated that EBV-miR-BART12 decreased the expression of TPPP1 and inhibited the function of TPPP1 and HDAC6, thereby promoting the deacetylation of β-catenin and making the β-catenin free from the fate of proteasome degradation.
Also, β-catenin is a key marker of the EMT signaling pathway. We speculate that EBV-miR-BART12 participates in the EMT signaling pathway by protecting β-catenin from degradation, promoting the activation of β-catenin and its downstream molecules. To this end, we detected the expression of some EMT related molecules in 5-8F and AGS cell lines with BART12, the TPPP1, or siTPPP1 groups. The results showed that E-cadherin, Claudin-1, and ZO-1 were down-regulated, transcription factors 4 (TCF4), N-cadherin, Vimentin, Slug, and Snail were upregulated in the EBV-miR-BART12 group. They are opposite to the TPPP1 group ( Figure 6E). The qRT-PCR results also showed the same trend of these EMT molecules ( Figure 6F). These results suggest that EBV-miR-BART12 stimulates the EMT process through downregulating the expression of TPPP1, and reducing the binding of TPPP1 and HDAC6, then resulting in the activation of β-catenin.
Finally, we detected the distribution of β-catenin expression in 5-8F, AGS, and C666-1 cells in order to prove if EBV-miR-BART12 promotes the EMT process through regulating β-catenin entering the nucleus. The histone H3 is a nuclear reference and α-tubulin works as a cytoplasmic reference. The results showed that EBV-miR-BART12 increased the expression of β-catenin in the nucleus and cytoplasm, compared with the control group ( Figure 6G). The results above indicate that EBV-20 miR-BART12 enhances the activity of HDAC6 by inhibiting the expression of TPPP1, thereby promoting the deacetylation of β-catenin, accelerating the enucleation of βcatenin, and finally initiating the downstream transcription factor to facilitate the EMT process.

Discussion
In this study, we found that EBV-miR-BART12 was highly expressed in the NPC clinical samples and positively correlated with the prognosis of the NPC patients. It deacetylated α-tubulin and β-catenin, resulting in microtubules polymerization and β-catenin entering the nucleus and activation. Thereby promoting the EMT process through directly targeting the 3'-UTR of TPPP1 mRNA.
TPPP1 is a microtubule associated protein (MAPs), which promote tubulin proteins aggregating into microtubules and changing the ultra-structure of microtubule (32).
The present studies showed that TPPP1 affected cell cycle and migration through regulating the microtubule, although there is still very few reports in cancer research (33)(34)(35)(36)(37)(38). It is reported that TPPP1 was low-expressed in oligodendrocyte gliomas (39), non-small cell lung cancer (40), and hepatocellular carcinoma(41) and associated with poor prognosis of the patients (31,41,42), suggesting that TPPP1 be used as a potential molecular cancer marker and potential target. In this study, we preliminarily study the mechanism of TPPP1 in tumorigenesis. EBV-miR-BART12 regulates the microtubule dynamic stability and EMT process in cancer cells through targeting TPPP1.
As one of the cytoskeleton component, the microtubule stability affects cancer cells invasion and metastasis (37,43). The co-localization of TPPP1 and α-tubulin proteins increases the microtubule stability (29). Many factors affect the microtubules 21 stability including the posttranslational modification and some microtubule proteins.
Microtubules in a stable state will gather more posttranslational modifications, such as microtubule acetylation (44,45). The regulation of acetylation is an important function of TPPP1. TPPP1 stabilized microtubules and promoted tubulin acetylation through binding and inhibiting HDAC6 activity (37,46). The interaction between HDAC6 and TPPP1 led to the inhibition of HDAC6 deacetylation activity, and thus increased the microtubule acetylation. In this study, we found that TPPP1 and αtubulin were co-localized in the cytoplasm. EBV-miR-BART12 affects acetylation of microtubules by binding TPPP1, causing the change of the microtubule structure.
It had been reported that HDAC6 also be involved in the formation of tumors by deacetylating other molecules, such as β-catenin (47) and then promoting the nuclear localization of β-catenin (47,48), the EMT process (49) and tumor invasion and metastasis (50). As an inhibitor of HDAC6, TPPP1 was decreased by EBV-miR-BART12, causing an increase of HDAC6 activity. When HDAC6 activity is elevated, the acetylation level of β-catenin is decreased, which reduces the amount of βcatenin that should be degraded. Then the level of β-catenin in the cytoplasm is improved, which promotes the entry of β-catenin into the nucleus, activates downstream transcription factors, and initiates EMT, thereby facilitating tumor cell invasion and metastasis.
In this study, we combined with clinical chemotherapeutic drugs to analyze the prospects of EBV-miR-BART12 in tumor therapy. Paclitaxel is a chemotherapeutic agent (30,32,51), which has strong anti-tumor ability in clinic mainly through promoting the assembly of microtubules and stabilizing the polymerized microtubules, resulting in stagnating cells in G0/G1 and G2/M stages and inducing apoptosis (52,53). It has been widely used to treat solid tumors, including breast cancer, ovarian cancer, and non-small cell lung cancer, because of its curative effect and small side effects (54)(55)(56). In this study, EBV-miR-BART12 had an antagonistic effect on the microtubule with paclitaxel in cancer cells. EBV-miR-BART12 might promote microtubule depolymerization by inhibiting the expression of TPPP1, but this depolymerization is reversed with the action of paclitaxel. It is suggested that paclitaxel may be effective in the treatment of patients with high expression of EBV-miR-BART12. And the development of a combination of EBV-miR-BART12 inhibitors and paclitaxel has a potential effect on the treatment of EBV related NPC and gastric cancer.

Conclusion
We have found the high expression of EBV-miR-BART12 is related to the poor prognosis of the NPC patients for the first time. EBV-miR-BART12 promoted NPC and GC cells through targeting the 3'-UTR of TPPP1 mRNA, resulting in an increasing the activity of HDAC6, and a deacetylation of α-tubulin and β-catenin, thereby changing the microtubules dynamic stability and promoting the EMT process, ultimately cancer migration and invasion ( Supplementary Fig. 1). We also confirmed our results in the mouse experiment, too ( Supplementary Fig. 2B). EBV-miR-BART12 and TPPP1 might be potential markers for the diagnosis and prognosis of NPC and GC and provided potential targets in NPC and GC treatment.