MIR29A Impedes Metastatic Behaviors in Hepatocellular Carcinoma via Targeting LOX, LOXL2, and VEGFA

Primary liver cancer accounts for the third most deadly type of malignant tumor globally, and approximately 80% of the cases are hepatocellular carcinoma (HCC), which highly relies on the activity of hypoxia responsive pathways to bolster its metastatic behaviors. MicroRNA-29a (MIR29A) has been shown to exert a hepatoprotective effect on hepatocellular damage and liver fibrosis induced by cholestasis and diet stress, while its clinical and biological role on the activity hypoxia responsive genes including LOX, LOXL2, and VEGFA remains unclear. TCGA datasets were retrieved to confirm the differential expression and prognostic significance of all genes in the HCC and normal tissue. The Gene Expression Omnibus (GEO) dataset was used to corroborate the differential expression and diagnostic value of MIR29A. The bioinformatic identification were conducted to examine the interaction of MIR29A with LOX, LOXL2, and VEGFA. The suppressive activity of MIR29A on LOX, LOXL2, and VEGF was verified by qPCR, immunoblotting, and luciferase. The effect of overexpression of MIR29A-3p mimics in vitro on apoptosis markers (caspase-9, -3, and poly (ADP-ribose) polymerase (PARP)); cell viability and wound healing performance were examined using immunoblot and a WST-1 assay and a wound healing assay, respectively. The HCC tissue presented low expression of MIR29A, yet high expression of LOX, LOXL2, and VEGFA as compared to normal control. Serum MIR29A of HCC patients showed decreased levels as compared to that of normal control, with an area under curve (AUC) of 0.751 of a receiver operating characteristic (ROC) curve. Low expression of MIR29A and high expression of LOX, LOXL2, and VEGFA indicated poor overall survival (OS). MIR29A-3p was shown to target the 3′UTR of LOX, LOXL2, and VEGFA. Overexpression of MIR29A-3p mimic in HepG2 cells led to downregulated gene and protein expression levels of LOX, LOXL2, and VEGFA, wherein luciferase reporter assay confirmed that MIR29A-3p exerts the inhibitory activity via directly binding to the 3′UTR of LOX and VEGFA. Furthermore, overexpression of MIR29A-3p mimic induced the activity of caspase-9 and -3 and PARP, while it inhibited the cell viability and wound healing performance. Collectively, this study provides novel insight into a clinical-applicable panel consisting of MIR29, LOX, LOXL2, and VEGFA and demonstrates an anti-HCC effect of MIR29A via comprehensively suppressing the expression of LOX, LOXL2, and VEGFA, paving the way to a prospective theragnostic approach for HCC.


Introduction
Primary liver cancer accounts for the third most deadly type of malignant tumor globally and approximately 80% of the cases are hepatocellular carcinoma (HCC) [1,2]. To date, despite the availability of many different therapeutic interventions, such as surgical resection, liver transplantation, radiofrequency ablation, chemoembolization, and target therapy [3], the prognosis of HCC remains poor, with a five-year overall survival (OS) rate below 20% [1]. One of the key factors is its propensity for metastatic progression and poor response to pharmacological approaches [3]. As such, the identification of promising theragnostic targets to improve the prognosis of the HCC patients is of vital importance.
The lysyl oxidase family members are secreted copper-dependent oxidases, including five paralogues: LOX, LOX-like 1-4 (LOXL1-4), which act to exert catalytic activity to remodel the cross-linking of the extracellular matrix (ECM) of fibrotic liver and that of a corrupted tumor microenvironment (TME) [4]. To date, the roles of LOX and LOXL2 in the clinical significance and therapeutic implication of HCC are mostly studied as compared to the other family members [5]. The fact that LOX and LOXL2 serving as critical factors in mediating the formation of a corrupted TME and promoting the progression of metastasis of HCC through activating pathways involved in hypoxia responsive signaling and angiogenesis, and epithelial mesenchymal transition (EMT) has been highlighted [5]. For example, LOX acts to mediate angiogenesis via increasing the secreted vascular endothelial growth factor A (VEGFA) from HCC cells [6] and to augment metastatic behaviors by activating EMT program [7]. On the other hand, LOXL2 was reported to activate the process of angiogenesis through upregulating the HIF-1α/VEGF pathway [8] and promote HCC metastasis to distant organ via an AKT/fibronectin-dependent pathway [9]. Given the biological importance of LOX, LOXL2, and VEGFA in the cancer progression, a variety of targeting drugs has currently been evaluated in clinical trials [5,[10][11][12]. Nonetheless, a treatment approach that can comprehensively targeting these factors to render therapeutic potential for the treatment of HCC is yet to be identified.
MicroRNAs (miRNAs), a large family of small non-coding RNAs, was reported to feature diverse and crucial effect on the development and metastasis of HCC by targeting the 3 untranslational region (3 UTR) of a variety of genes [13]. MIR29 family members consist of MIR29A, MIR29B, and MIR29C. The UCU sequence of MIR29A results in its high stability as compared to the rapid decay of MIR29B and MIR29C caused by the presence of UUU tri-uracil sequence [14]. The MIR29A expression features clinical significance in a variety of disease scenarios, including Alzheimer's disease [15], Parkinson's disease [16], ankylosing spondylitis [17], atherosclerosis [18], atrial fibrillation [19], active pulmonary tuberculosis [20], thoracic aneurysms [21], tendon disease [22], diabetes [23], scleroderma [24], cholestatic pediatric liver disease [25], and non-alcoholic fatty liver disease (NAFLD) [26]. Importantly, our previous studies have reported the hepatoprotective role of MIR29A in the scenario of cholestasis-or diet-induced steatohepatitis and liver fibrosis [27][28][29][30][31][32][33][34][35][36][37][38], and growing evidence has revealed its regulatory role on HCC metastasis [13,39]. In terms of biological activity, MIR29A was reported to directly target VEGFA to perturb the wound healing process [40], and to directly target LOXL2 to impede tumor progression [41]. In this study, we demonstrated that MIR29A, along with LOX, LOXL2, and VEGFA, exhibits diagnostic and prognostic significance and that overexpression of MIR29A notably contributes to suppression of cellular metastatic behaviors by comprehensively targeting the 3 UTR of the aforementioned genes, offering novel insights into the MIR29A-involved signaling in the development of a practical diagnostic/prognostic panel and a therapeutic strategy.

The Aberrent Expression and the Prognostic Value of MIR29A, LOX, LOXL2, and VEGFA in HCC Patients
We firstly examined the MIR29A expression levels in HCC and normal tissue by accessing the TCGA datasets in the OncoMir Cancer Database (OMCD). Using MIR16 as a normalization control as previously described [42], HCC tumor tissue (375 TCGA datasets) presented significantly lower expression levels of MIR29A than that of normal tissue (51 TCGA datasets) (p < 0.01; Figure 1A). In addition, serum MIR29A expression normalized to MIR16 in HCC patients and non-cancer individuals from the Gene Expression Omnibus (GEO) dataset presented a similar trending (p < 0.001; Figure 1B). The number (%) of HCC patients presenting a lower value than normal individuals at the cut-off of maximum, mean + 1SD, mean, mean-1SD, and minimum value of normal serum was 315/317 (99.4%), 311/317 (98.1%), 229/317 (72.2%), 165/317 (52.1), and 7/317 (2.2%), respectively ( Figure 1C). We then asked whether MIR29A gene alteration in HCC plays a factor in its expression change. In this regard, we surveyed the MIR29A data with regard to structural variant, mutation, and copy number alterations (CNA) from TCGA PanCancer Atlas Studies using the cBioPortal web platform, where 10,953 patients/10,967 samples are included as of 26 May 2021. In HCC, MIR29A presented a 1.084% alteration frequency in amplification without the presence of deletion and mutation (Supplementary Figure S1), indicating that copy number alteration, deletion, and mutation scarcely contribute to an expression change of MIR29A. In addition, MIR29A expression between male and female HCC tissue showed no difference (Supplementary Figure S2), suggesting no potential gender bias acting on its expression in HCC tissue. Additionally, we determined the diagnostic value of serum MIR29A using a receiver operating characteristic (ROC) curve and noted that MIR29A presented good diagnostic accuracy for the discrimination of disease, with an area under curve (AUC) of 0.751 (95% CI: 0.711-0.792, p < 0.0001) ( Figure 1D). Next, UALCAN web server-accessed TCGA datasets revealed that the mRNA of LOX, LOXL2, and VEGFA presented elevated expression levels in HCC tissue as compared to that in normal tissue (p < 0.001; Figure 1E-G).
The correlation between the expression of MIR29A, LOX, LOXL2, and VEGFA and corresponding clinical follow-up information was analyzed by Kaplan-Meier curves and the log-rank test. High MIR29A expression was found to be associated with increased overall survival (OS) (HR = 0.43; p = 0.0032) ( Figure 2D) was associated with decreased OS. These results indicate that the downregulated MIR29A and upregulated LOX, LOXL2, and VEGFA predicts poor prognosis, serve as an independent prognostic panel, and may contribute to HCC progression.

MIR29A Act as A Common Supressor of LOX, LOXL2, and VEGFA
The contrary expression pattern between MIR29A and the aforementioned mRNAs in HCC tissue prompted us to ask whether MIR29A presents a pathway-centric manner on modulating the genes responsible for hypoxic adaptation, matrix remodeling, and angiogenesis. We conducted gene functional enrichment analysis to examine the microRNAinteracting network of LOX, LOXL2, and VEGFA using GSCALite and TargetScan Release 7.2. As shown in Figure 3, MIR29A-3p acts as a central hub in regulating LOX, LOXL2, and VEGFA ( Figure 3A,B) via putative binding at the 3 UTR ( Figure 3C). We then verified the bioinformatic prediction by an in vitro experimental setting, where overexpression of MIR29A-3p in human HCC cell line HepG2 significantly inhibited the gene and protein expression of LOX ( Figure 4A,B), LOXL2 ( Figure 4C,D), and VEGFA ( Figure 4E,F) (all p < 0.05). Additionally, the molecular interaction at posttranscriptional level was corroborated by luciferase reporter assay. As shown in Figure 5A,B, the pMIR-REPORTER TM plasmid comprises a strong protomer CMV to drive the expression of luciferase that is fused with the 3 UTR sequence of LOX or VEGFA ( Figure 5A,B). Binding of MIR29A-3p to wild type 3 UTR leads to reduced luciferase signal, whereas the 3 UTR mutant serves as a negative control. Overexpression of MIR29A-3p suppressed the reported luciferase activity of wild type but not that of mutant, as compared to the negative control (NC)-treated group (all p < 0.05) ( Figure 5C,D), revealing that the direct binding to LOX 3 UTR and VEGFA 3 UTR accounts for the suppression activity of MIR29A-3p.

MIR29A Act as A Common Supressor of LOX, LOXL2, and VEGFA
The contrary expression pattern between MIR29A and the aforementioned mRNAs in HCC tissue prompted us to ask whether MIR29A presents a pathway-centric manner on modulating the genes responsible for hypoxic adaptation, matrix remodeling, and angiogenesis. We conducted gene functional enrichment analysis to examine the microRNAinteracting network of LOX, LOXL2, and VEGFA using GSCALite and TargetScan Release

MIR29A Act as A Common Supressor of LOX, LOXL2, and VEGFA
The contrary expression pattern between MIR29A and the aforementioned mRNAs in HCC tissue prompted us to ask whether MIR29A presents a pathway-centric manner on modulating the genes responsible for hypoxic adaptation, matrix remodeling, and angiogenesis. We conducted gene functional enrichment analysis to examine the microRNAinteracting network of LOX, LOXL2, and VEGFA using GSCALite and TargetScan Release mid comprises a strong protomer CMV to drive the expression of luciferase that is fused with the 3′UTR sequence of LOX or VEGFA ( Figure 5A,B). Binding of MIR29A-3p to wild type 3′UTR leads to reduced luciferase signal, whereas the 3′UTR mutant serves as a negative control. Overexpression of MIR29A-3p suppressed the reported luciferase activity of wild type but not that of mutant, as compared to the negative control (NC)-treated group (all p < 0.05) ( Figure 5C,D), revealing that the direct binding to LOX 3′UTR and VEGFA 3′UTR accounts for the suppression activity of MIR29A-3p.

Overexpression of MIR29A Might Induce Apoptosis Signaling and Inhibits Wound Healing Performance and Viability in HCC Cells
We then evaluated the effect of overexpression of MIR29A-3p mimics on the metastatic behaviors of HCC cells. As shown in Figure 6, overexpression of MIR29A-3p mimics induced a significant increase in the markers representing activated apoptosis signaling, including the increased ratio of cleaved/non-cleaved caspase-9 ( Figure 6A), cleaved caspase-3 ( Figure 6B), and cleaved poly (ADP-ribose) polymerase (PARP) ( Figure 6C). Additionally, the wound healing assay showed that overexpression of MIR29A-3p mimics features a suppressive effect on cellular wound healing performance ( Figure 6D,E). WST-1 assay revealed an overexpression of MIR29A-3p mimics saps cell viability ( Figure 6F). Together, these results indicate that MIR29A acts toward the impeded metastatic behaviors of HCC cells.   VEGFA (B,D). Nucleotides in red denotes mismatched mutation as compared to that of wild type. After transfection of the pMIR-REPORTER TM plasmid, HepG2 cells were treated with transfection reagent transfection reagent as a control (Ctrl), miR negative control sequence (NC), or MIR29A-3p mimic (mimic), followed by the detection of luciferase activity. * p < 0.05, *** p < 0.001 between the indicated groups.   including the increased ratio of cleaved/non-cleaved caspase-9 ( Figure 6A), cleaved caspase-3 ( Figure 6B), and cleaved poly (ADP-ribose) polymerase (PARP) ( Figure 6C). Additionally, the wound healing assay showed that overexpression of MIR29A-3p mimics features a suppressive effect on cellular wound healing performance ( Figure 6D,E). WST-1 assay revealed an overexpression of MIR29A-3p mimics saps cell viability ( Figure 6F). Together, these results indicate that MIR29A acts toward the impeded metastatic behaviors of HCC cells.

Discussion
In this study, we demonstrated that MIR29A, LOX, LOXL2, and VEGFA represent a panel of differential biomarkers and independent prognostic factors for HCC patients and

Discussion
In this study, we demonstrated that MIR29A, LOX, LOXL2, and VEGFA represent a panel of differential biomarkers and independent prognostic factors for HCC patients and that MIR29A acts as a common suppressor on these genes to impede the metastatic behaviors of HCC. The proposed model of the clinical value and the mechanism underlying MIR29A-centric pathways is illustrated in Figure 7.
There are a number of studies reporting that low expression of MIR29A represents a biomarker and a predictor for poor outcome in HCC patients [43][44][45][46], whereas a few reports showed an opposite expression manner [47,48]. In this study, a decreased expression of MIR29A is noted in HCC tissue as compared to that of normal tissue and serves as an independent factor for poor OS, in support of the aforementioned reports [43][44][45][46]. Zhu et al. and Xue et al. reported increased serum MIR29A as a diagnostic factor of HCC [42,49], while our study demonstrated that serum MIR29A expression was in parallel with that of tissue and exhibited a good diagnostic accuracy.
Although MIR29A-activated metastasis-suppression signaling pathways involving carcinogenesis, epigenetics, metabolic adaptation, and immunomodulation in HCC have been noted [39], its role in regulating factors modeling the structure of TME was yet to be clarified. In this regard, our study identified MIR29A as a common suppressor for LOX, LOXL2, and VEGFA, which notably present a panel of diagnostic and prognostic factors along with MIR29A. As such, this finding provides new insights into clinical decisionmaking and the development of predictive settings for HCC. that MIR29A acts as a common suppressor on these genes to impede the metastatic behaviors of HCC. The proposed model of the clinical value and the mechanism underlying MIR29A-centric pathways is illustrated in Figure 7. There are a number of studies reporting that low expression of MIR29A represents a biomarker and a predictor for poor outcome in HCC patients [43][44][45][46], whereas a few reports showed an opposite expression manner [47,48]. In this study, a decreased expression of MIR29A is noted in HCC tissue as compared to that of normal tissue and serves as an independent factor for poor OS, in support of the aforementioned reports [43][44][45][46]. Zhu et al. and Xue et al. reported increased serum MIR29A as a diagnostic factor of HCC [42,49], while our study demonstrated that serum MIR29A expression was in parallel with that of tissue and exhibited a good diagnostic accuracy.
Although MIR29A-activated metastasis-suppression signaling pathways involving carcinogenesis, epigenetics, metabolic adaptation, and immunomodulation in HCC have been noted [39], its role in regulating factors modeling the structure of TME was yet to be clarified. In this regard, our study identified MIR29A as a common suppressor for LOX, LOXL2, and VEGFA, which notably present a panel of diagnostic and prognostic factors along with MIR29A. As such, this finding provides new insights into clinical decisionmaking and the development of predictive settings for HCC.
There are increasing reports demonstrating the promising effect of targeting hypoxia-responsive genes, such as LOX, LOXL2, and VEGFA. For instance, a couple of drugs targeting LOX family members are in the early stage of clinical trials [5], including pancreatic and colorectal adenocarcinoma [10,11]. Sorafenib was shown to hit the first breakthrough systemic therapy for treating advanced HCC via disrupting VEGF signaling [50]. Three angiogenesis inhibitor, Trebananib, rebastanib, and MEDI3617, have been evaluated in phase I and II clinical trials [51]. Despite the present lack of HCC trial using approaches targeting the activity of LOX family members, further investigation into broadening the horizons of this issues and the therapeutic strategy widely targeting critical factors are promising. In this study, we demonstrated the biological role of MIR29A in comprehensively inhibiting LOX, LOXL2, and VEGFA and its effect on counteracting metastatic behaviors by inducing apoptosis and repressing cell viability and wound healing performance, indicating the potential of MIR29A as an adjuvant therapy that may further There are increasing reports demonstrating the promising effect of targeting hypoxiaresponsive genes, such as LOX, LOXL2, and VEGFA. For instance, a couple of drugs targeting LOX family members are in the early stage of clinical trials [5], including pancreatic and colorectal adenocarcinoma [10,11]. Sorafenib was shown to hit the first breakthrough systemic therapy for treating advanced HCC via disrupting VEGF signaling [50]. Three angiogenesis inhibitor, Trebananib, rebastanib, and MEDI3617, have been evaluated in phase I and II clinical trials [51]. Despite the present lack of HCC trial using approaches targeting the activity of LOX family members, further investigation into broadening the horizons of this issues and the therapeutic strategy widely targeting critical factors are promising. In this study, we demonstrated the biological role of MIR29A in comprehensively inhibiting LOX, LOXL2, and VEGFA and its effect on counteracting metastatic behaviors by inducing apoptosis and repressing cell viability and wound healing performance, indicating the potential of MIR29A as an adjuvant therapy that may further enhance the treatment effectiveness of the targeting therapy for HCC. Nevertheless, further studies mimicking the physiological circumstance such as in vivo evaluation is warranted to clarify the safety and efficacy in a preclinical setting.
In summary, this study provides novel insight into a clinically applicable panel consisting of MIR29, LOX, LOXL2, and VEGFA and demonstrates an anti-HCC effect of MIR29A via comprehensively suppressing the expression of LOX, LOXL2, and VEGFA, paving the way to a prospective theragnostic approach for HCC.

Analysis of Gene Differential Expression and Prognostic Significance
The MIR29A expression levels in HCC and normal liver tissue were analyzed using the OncoMiR Cancer Database (OMCD) web server, a repository enabling systemic comparative genome analysis of miR expression sequencing data derived from over 10,000 cancer patients with associated clinical information and organ-specific controls present in The Cancer Genome Atlas (TCGA) database [52]. The raw data of TCGA samples retrieved from OMCD are shown in Supplementary File 1. Serum MIR29A expression levels in HCC and noncancer cohorts were analyzed using National Center for Biotechnology Information Gene Expression Omnibus (GEO) database (available online: http://www.ncbi.nlm.nih.gov/geo (accessed on 25 April 2021)) [53], GSE113740. The comparison of the gene expression level in HCC and normal tissue was undertaken in UALCAN web server [54], which facilitates access to publicly available cancer OMICS data, such as TCGA and Clinical Proteomic Tumor Analysis Consortium (CPTAC). MIR16 served as normalization control as previously described [42].
Kaplan-Meier analysis, which was used to determine the prognostic value of biomarkers, was conducted in the Kaplan-Meier plotter web server (available online: https: //kmplot.com/analysis/ (accessed on 25 April 2021)), which includes data sources from GEO, European Genome-phenome Archive (EGA), and TCGA to enable the assessment of the effect of 54 k genes, including mRNA, miR, and protein on survival in 21 cancer types [55]. The prognostic value of miR and mRNA was evaluated by using the panel miRpower for liver cancer [56] and RNA-seq for liver cancer [57] of the KM plotter, respectively.

Luciferase Reporter Assay
To investigate the molecular basis of MIR29A-3p acting on LOX and VEGFA, HepG2 cells were respectively transfected with pMIR-LOX-3 UTR and pMIR-VEGFA-3 UTR followed by transfection of reagent only, MIR negative sequence, or MIR29A-3p mimic. Cells transfected with pMIR-LOX-3 UTR mutant and pMIR-VEGFA-3 UTR mutant served as the negative control. Specifically, the sequence of wild type 3 UTR or mutated 3 UTR were cloned into the multiple cloning site after the CMV-driven luciferase ( Figure 3B) of the pMIR-REPORTER TM plasmid (AM5795, Thermo Fisher Scientific, Rockford, IL, USA). Then, HepG2 cells were seeded at 1.8 × 10 6 cells in a 100 mm dish for 18 h. Next, 6 µg reporter plasmids were transfected into a 100 mm dish cultivating HepG2 cells in the presence of Turbofect transfection reagent (R0531, Thermo Fisher Scientific, Rockford, IL, USA) for 20 h. HepG2 were subsequently trypsinized and seeded at 9 × 10 5 cells in a 60 mm dish with fresh growth medium.

Quantitative Real-Time PCR (qPCR)
Total RNA of the cells was extracted by using TRIzol ® reagent (15596026, Invitrogen, Carlsbad, CA, USA) and then underwent reverse transcription to yield cDNA with an oligodeoxynucleotide primer (oligo dT15)-based method according to the manufacturer's protocol (M1701, Promega, Madison, WI, USA). The qPCR reaction was undertaken using 2 × SYBR Green PCR Master Mix (04887352001, Roche Molecular Systems, Inc., Pleasanton, CA, USA) on LightCycler480 ® (Roche). Each PCR reaction included 0.5 µM forward and reverse primers, 30 ng of cDNA, and 1 × SYBR Green PCR Master Mix in a total reaction volume of 10 µL. For qPCR program, an initial amplification was done with a denaturation step at 95 • C for 10 min, followed by 45 cycles of denaturation at 95 • C for 30 s, primer annealing at 62 • C for 15 s, and primer extension at 72 • C for 25 s, followed by melting curve analysis. The primer sequences for human genes were shown in Table 1. Table 1. Sequence of primers pairs.

Forward Primer
Reverse Primer

Wound Healing Assay
Wound healing assay was employed to detect the wound healing performance of the cells. In each culture-insert, 5 × 10 5 HepG2 cells were seeded (ibidi culture-insert

Cell Viability Assay
WST-1 assay was used to determine the cell viability. In each 96-well, 5 × 10 5 HepG2 cells were seeded at a volume of 100 µL. After attachment, cells were transfected with 25 nM MIR29A mimics and their corresponding MIR negative control for 24 h, followed by the addition of 10 µL WST-1 reagent (Roche Diagnostics, Laval, QC, Canada) for another 1 h at 37 • C. The absorbance signal was measured using a microplate reader (Hidex Sense microplate reader, Turku, Finland) at a test wavelength of 450 nm and reference wavelength of 630 nm.

Statistical Analysis
All values are expressed as mean ± standard error (SE). Quantitative data were analyzed using unpaired t-test for two-group comparison or one-way analysis of variance for three-or-more-group comparison when appropriate. Least significant difference (LSD) test was used for post-hoc analysis. Two-sided p-values less than 0.05 were considered to be statistically significant.