RETRACTED ARTICLE: MicroRNA-3194-3p inhibits metastasis and epithelial-mesenchymal transition of hepatocellular carcinoma by decreasing Wnt/β-catenin signaling through targeting BCL9

We, the Editors and Publisher of the journal Artificial Cells, Nanomedicine, and Biotechnology, have retracted the following article: Bowen Yao, Yazhao Li, Liang Wang, Tianxiang Chen, Yongshen Niu, Qingguang Liu & Zhikui Liu (2019) MicroRNA-3194-3p inhibits metastasis and epithelial-mesenchymal transition of hepatocellular carcinoma by decreasing Wnt/β-catenin signaling through targeting BCL9. Artificial Cells, Nanomedicine, and Biotechnology, 47:1, 3885–3895, DOI: 10.1080/21691401.2019.1670190 Since publication, concerns have been raised about the integrity of the data in the article. In particular, Figures 3A, 5E, 6E and 6F appear to have been previously published in other journals with different author lists, specifically: Figure 3A appears to have been duplicated with Figure 3E, Yang et al., Cell Death & Disease (2017), DOI: 10.1038/cddis.2017.358 Figure 5E appears to have been duplicated with Figure 10C, Liu et al., Journal of Experimental & Clinical Cancer Research (2019), DOI: 10.1186/s13046-019-1188-x Figure 6E and 6F appear to have been duplicated with Figure 10G and 10F, Liu et al., Journal of Experimental & Clinical Cancer Research (2018), DOI: 10.1186/s13046-018-0717-3 When approached for an explanation, the authors checked their data and confirmed there are fundamental errors present and that the images had been “manipulated”. Therefore, they have agreed to the retraction of this article. We have been informed in our decision-making by our policy on publishing ethics and integrity and the COPE guidelines on retractions. The retracted article will remain online to maintain the scholarly record, but it will be digitally watermarked on each page as “Retracted.”


Background
Hepatocellular carcinoma (HCC) is one of the most common malignancies and the second leading cause of cancer-related mortality worldwide [1,2]. For the high prevalence of hepatitis B strongly predisposes individuals into developing chronic liver disease, and subsequently, HCC, the incidence of HCC is particularly high in China [3]. Although superior improvements in therapeutic strategy for HCC have undergone, the long-term survival of HCC patients remains poor due to high rates of recurrence and metastasis [4,5]. Therefore, understanding the molecular mechanisms underlying cancer progression may contribute to the discovery of more effective intervention for HCC treatment.
MicroRNAs (miRNAs), a group of evolutionarily conserved small non-coding RNA, inhibit translation or induce mRNA degradation by binding to the 3 0 -untranslated regions (3 0 -UTR) of target genes [6,7]. As post-transcriptional regulators, miRNAs play critical roles in physiological and pathological processes, including cell differentiation, proliferation, apoptosis, migration, and invasion [8]. Accumulating evidence confirmed that miRNAs play a role as promising therapeutic and prognostic biomarkers in HCC diagnosis and treatment [9]. Nevertheless, underlying molecular mechanisms and function of miR-3194-3p in HCC remain unknown.
Epithelial-to-mesenchymal transition (EMT) contribute to cancer invasion and metastasis by transformation of adherent and polarized epithelial cells, disruption of cell-cell adhesion, into an invasive mesenchymal cell phenotype [10]. EMT is marked of losing epithelial adhesion and cytoskeletal markers, such as E-cadherin, meanwhile, the obtaining of migratory mesenchymal symbols, such as N-cadherin and Vimentin [11,12]. Increasing studies reveal that EMT is a main cause for HCC invasion and metastasis [13]. However, the association between miR-3194-3p and EMT in HCC is poorly investigated.
In the present study, we demonstrated that downregulated miR-3194-3p was associated with adverse prognostic features of HCC patients. miR-3194-3p inhibited migration, invasion and EMT progression of HCC cells in vitro and in vivo. Notably, B-cell lymphoma 9 (BCL9) was a direct target of miR-3194-3p and mediated the function of miR-3194-3p in HCC cells. Besides, miR-3194-3p, BCL9 and their combination were valuable predictors for the prognosis of HCC patients.

Clinical samples
Patients' tumor tissues and corresponding adjacent nontumor tissues were obtained from 129 patients in our department between January 2008 and December 2011 after the informed consent was obtained from all patients. They did not receive any therapy before surgery. Xi'an Jiaotong University Ethics Committee approved the research on the basis of Declaration of Helsinki.

Cell culture
The normal immortalized human hepatocyte LO2 and a panel of HCC cells (Hep3B, HepG2, Huh7, MHCC-97H, and SMMC-7721) (Chinese Academy of Sciences, Shanghai, China) were maintained in DMEM (Invitrogen, Carlsbad, USA) containing 10% FBS (Gibco, Grand Island, USA) in 37˚C with 5% CO2. Cells were classified into several groups under different treatments. The transfection was performed according to the instructions of lipofectamine 2000 (#11668019, Invitrogen,CA, USA). All the sequence details were shown in Supporting Figure 1.

Immunohistochemical staining (IHC)
The sections were dewaxed, dehydrated, and rehydrated. BCL9 (1:300, Abcam, USA) was applied and covered the sections, which were incubated at 4 C overnight. Then, applyied the biotinylated secondary antibodies (Goldenbridge, Zhongshan, China) according to SP-IHC assays. The detailed experiment was conducted similar to the previously reported [16].

Luciferase reporter assays
The 3 0 -UTR sequence of BCL9, together with a corresponding mutated sequence within the predicted target sites, were synthesized and inserted into the pmiR-GLO dual-luciferase miRNA target expression vector (Promega, Madison, WI, USA). The assays were carried out as described previously [14,20].

Cell migration and invasion analyses
To evaluate cell migration and invasion, we used 2 Â 10 4 transfected cells suspended in 150 lL serum-free DMEM medium into the upper chamber of Matrigel-uncoated and -coated transwell inserts (8 mm pore size; Millipore), and 550 lL 20% FBS DMEM medium was placed in the lower chamber. After 24 h incubation,fixed in 4% paraformaldehyde for 20 min and stained with 0.1% crystal violet dye for 15 min, cells on the inner layer were softly removed with a cotton swab and counted at five randomly selected views and the average cell number per view was calculated.

In vivo experiments
Four-week-old male BALB/c nude mice (Centre of Laboratory Animals, The Medical College of Xi'an Jiaotong University, Xi'an, China) were randomized into two groups (n ¼ 6) and either MHCC-97H-miR-3194-3p or MHCC-97H-miR-control cells (1 Â 10 6 ); Hep3B-anti-miR-3194-3p or Hep3B-anti-miR-NC were injected into the tail veins for the establishment of pulmonary metastatic model. Mice were sacrificed 10 weeks' post-injection and examined microscopically by H&E staining for the development of lung metastatic foci. Accounting of lung-metastasis nodules was finished in 100-fold high-power field of microscope in sections after stain. Animals were housed in cages under standard conditions. All in vivo protocols were approved by the Institutional Animal Care and Use Committee of Xi'an Jiaotong University.

Statistical analysis
Performing at least three independent replicates, we collected the presented data as the mean ± SD. To evaluate the statistical significance, we operated SPSS software, 16.0 (SPSS, Inc, Chicago, IL, USA) and Graphpad Prism 6.0 (CA, USA) by two-tailed Student t-test, Pearson's correlation analysis, Kaplan-Meier method, and the log-rank test. Differences were defined as p < .05.

Reduced miR-3194-3p expression confers metastasis and recurrence of HCC
To determine the expression level of miR-3194-3p in HCC tissues, we performed qRT-PCR and found that miR-3194-3p expression was significantly lower in human HCC tissues than in adjacent non-tumor tissues (p < .05, Figure 1(A)), which was clarified in TCGA database and shown in Supporting Figure 2A. We selected the HCC tissues with intrahepatic metastasis and venous infiltration as aggressive HCC tissues and found that miR-3194-3p was obviously decreased in aggressive HCC tissues compared to non-aggressive tissues (p < .05, Figure 1(B)). Moreover, miR-3194-3p was down-regulated in tumor tissues from patients with recurrence compared to tissues without recurrence (p < .05, Figure 1(C)). In addition, miR-3194-3p was significantly down-regulated in several HCC cell lines as compared with hepatic normal cell line LO2 (p < .05, Figure 1(D)). The significant down-regulation of miR-3194-3p in HCC tissues and cell lines supports its role as a tumor suppressor in HCC and confers metastasis and recurrence of HCC.
miR-3194-3p inhibits epithelial-to-mesenchymal transition of HCC cells EMT has been identified as a key modulator in the initiation of metastasis progression of HCC. To elucidate the potential role of miR-3194-3p in HCC metastasis, we performed Western blotting and IF to confirm that miR-3194-3p overexpression increased the EMT marker in MHCC-97H cells (p < .05, respectively, Figure 3(A,C). By contrast, miR-3194-3p knockdown decreased E-cadherin and increased N-cadherin and Vimentin expression in Hep3B cells (p < .05, respectively, Figure 3(B,D). Furthermore, we also investigated the correlation between miR-3194-3p and EMT markers in HCC tissues. We demonstrated that E-cadherin expression in high miR-3194-3p group HCC tissues was higher than that in low expressing cases. Conversely, the expression of Vimentin in the high miR-3194-3p group was markedly lower than that in low expression group (p < .05, respectively, Figure 3(E)).

R E T R A C T E D
Therefore, our results suggest that miR-3194-3p inhibits EMT progress of HCC cells.

BCL9 is up-regulated in HCC tissues and promotes migration and invasion of HCC cells
To determine the potential role of BCL9 in HCC, we performed qRT-PCR and WB to show that the BCL9 mRNA and protein expression was higher in HCC than adjacent nontumor (p < .05, Figure 5(A,B) and Supporting Figure 2(B)). To confirm the functional effects on HCC, we established stable BCL9 overexpression in or knockdown cells p < .05, Figure  5(C)) by lentiviral transduction. Transwell assays showed that BCL9 over-expression significantly promoted cell migration and invasion of Hep3B cells (p < .05, Figure 5(D)). In contrary, BCL9 knockdown inhibited cell migration and invasion in MHCC-97H cells (p < .05, Figure 5(E)). In addition, WB confirmed that BCL9 alteration obviously regulated EMT process (p < .05, Figure 5(F)). Therefore, we demonstrated that alteration of BCL9 could mimic miR-3194-3p-induced function on HCC cells.

miR-3194-3p suppressed the metastatic capacity of HCC cells in vivo
To confirm the effects of miR-3194-3p in vivo, we used lateral veins injection model to construct a lung metastasis model. We demonstrated that miR-3194-3p over-expression group showed fewer and smaller foci in the lungs of nude mice via microscopic evaluation (p < .05, Figure 7(A)). Moreover, we confirmed that the lung sections of miR-3194-3p over-expression group showed decreased expression of BCL9 ( Figure  7(B)). In contrary, miR-3194-3p knockdown in Hep3B cells led to a significant increase of lung metastasis nudes and BCL9 expression (p < .05, Figure 7(C,D)). In addition, to confirm if miR-3194-3p regulates EMT of HCC cells in vivo, we found that miR-3194-3p over-expression increased E-cadherin expression and inhibited Vimentin expression (Figure 7(B)).

R E T R A C T E D
However, miR-3194-3p knockdown showed opposite effects (Figure 7(D)). Altogether, these results suggest that miR-3194-3p restrain metastatic behaviors and EMT phenotype of HCC by regulating BCL9 in vivo.  Figure  8(A-D)). With combination analysis, the data showed that patients with low miR-3194-3p and high BCL9 expression had the worst OS and DFS (p < .05, respectively, Figure 8(E,F)). These data suggest that combination of

R E T R A C T E D
miR-3194-3p and BCL9 is a potential biomarker for the clinical outcome of HCC patient.
Hypoxia condition induced miR-3194-3p downregulation and mediates the effects of hypoxia on HCC metastasis Hypoxia, an important feature of tumor microenvironments, have been recognized as a critical regulator of cancer metastasis [23]. Previous study confirmed that BCL9, which was a

R E T R A C T E D
downstream target of miR-3194-3p in this research, could be regulated by hypoxia [24]. Therefore, we try to explore whether miR-3194-3p could be regulated by hypoxia and its role in hypoxia. Hypoxia condition significantly increased HIF-1a expression in Hep3B cells (p < .05, Figure 9(B)) and led to decrease of miR-3194-3p expression (p < .05, Figure 9(C)). Furthermore, miR-3194-3p over-expression abolished the promoting effects of hypoxia on migration and invasion of Hep3B cells (p < .05, Figure 9(D)). Similarly, the positive effects of hypoxia on EMT process were reversed by miR-3194-3p restoration in Hep3B cells (p < .05, Figure 9(E)). Altogether, these results indicated that hypoxia-induced miR-3194-3p loss play a key role in migration, invasion, and EMT events of HCC cells.

Discussion
Accumulating evidence reported that aberrant miRNAs act as oncogenes or tumor suppressors in the cancer initiation, development, and progression [15,25]. New studies have established the potential usefulness of miRNAs as therapeutic molecules against cancers [26]. In this study, we propose a novel role for miR-3194-3p in coordinating the BCL9 expression in HCC cells and tissues. Being the first report, we demonstrated that miR-3194-3p was significantly down-regulated in HCC tissues and cell lines. The tissues with metastatic and recurrent phenotype showed reduced miR-3194-3p expression. These results strongly suggest that miR-3194-3p is a tumor suppressor in HCC and plays a pivotal role in the progression of HCC. Metastasis, a major reason that leads to the dismal prognosis of HCC [27], has been reported by numerous studies, which confirm that miRNAs are identified as regulators in metastasis of HCC [28]. To this end, we explored whether miR-3194-3p was involved in the progression of HCC by regulating the metastasis of HCC. In this study, gain-and loss-of-function experiment demonstrated that miR-3194-3p over-expression inhibited the migration and invasion of HCC cells while miR-3194-3p knockdown increased these metastatic behaviors in vitro and in vivo. EMT has been recognized as an important process in the invasion and metastasis of HCC [29]. We confirmed that miR-3194-3p inhibited EMT process of HCC cells. In patients' tissues, we also showed that miR-3194-3p was inversely associated with EMT markers. These findings indicated that miR-3194-3p inhibits metastasis of HCC by regulating EMT and potentially serve as a therapeutic target of HCC metastasis. B cell lymphoma 9 (BCL9), a component of aberrantly activated Wnt signaling, is an important contributing factor to tumor progression [30]. Despite overwhelming evidence highlighted the clinical implication of BCL9 in cancers [31,32], here we confirmed that BCL9 was a direct downstream target of miR-3194-3p and mediated the biological effects of miR-3194-3p on HCC. miR-3194-3p negatively regulated BCL9 mRNA and protein expression level in HCC, which we confirmed that was up-regulated in HCC tissues and promoted the migration and invasion of HCC cells. Moreover, miR-3194-3p was inversely correlated with the expression of BCL9 in tissues. Previous studies reported that dysregulated BCL9 expression is an activator of Wnt pathway. In our study, we discovered that miR-3194-3p restrained the activation of Wnt/b-catenin signaling.
In clinical roles, we confirm that miR-3194-3p and BCL9 could serve as valuable biomarkers for prognostic prediction. We found that low miR-3194-3p expression and high BCL9 were significantly associated with poor prognostic features of HCC patients. Nevertheless, we confirmed that miR-3194-3p down-expression and BCL9 over-expression, as well as their combination, were obviously correlated with poor prognosis of HCC patients. These results suggest that miR-3194-3p and BCL9 may be promising predictors for the prognosis of HCC. Finally, we elucidated the underlying mechanism for reduced expression of miR-3194-3p in HCC. Previous researches have

R E T R A C T E D
shown hypoxia tumor microenvironment play important roles in metastasis of HCC [33]. Moreover, BCL9, which was the target of miR-3194-3p, was also regulated by hypoxia condition.
To confirm the reason, we showed that miR-3194-3p level was decreased in hypoxia environment. Importantly, we also found that overexpression of miR-3194-3p level abrogated the prompting effects of hypoxia on migration, invasion, and EMT progress. These results suggest that hypoxia-induced miR-3194-3p loss promotes the metastasis and EMT of HCC. Altogether, our observations suggested that miR-3194-3p inhibits HCC migration, invasion, and EMT via decreasing Wnt/b-catenin signaling through targeting BCL9 and might be a therapeutic target for HCC.

Conclusion
In conclusion, we demonstrate miR-3194-3p is down-regulated in HCC tissues and cell lines, and its reduced expression is associated with adverse clinicopathological features. We confirm that miR-3194-3p inhibits migration, invasion, and EMT process of HCC cells by directly targeting BCL9-mediated Wnt/b-catenin pathway. Notably, a combination of downregulated miR-3194-3p and over-expressed BCL9 is potential prognostic predictors for the survival of HCC patients. Moreover, we confirmed that hypoxia was the underlying reason for miR-3194-3p down-regulation. In summary, the down-regulation of miR-3194-3p was involved in tumor metastasis and maybe a novel prognostic factor and potential therapeutic target for HCC.

Ethical approval
Ethics approval and consent to participate. All procedures performed in studies involving human participants were in accordance with the ethical standards of the Research Ethics Committee of The First Affiliated Hospital of Xi'an Jiaotong