Retinoic Acid Induced 16 Enhances Tumorigenesis and Serves as a Novel Tumor Marker for Hepatocellular Carcinoma

Our previous work identified downregulated miR-483-5p in hepatocellular carcinoma (HCC). This study aims to identify the target of miR-483-5p, evaluate the potential value of this target as a tumor marker for HCC and explore the role of this target in HCC tumorigenesis. Upregulated retinoic acid induced 16 (RAI16) (17/18 cases) was negatively correlated with downregulated miR-483-5p (14/18 cases) in HCC tissues. The dual-luciferase reporter assay showed that RAI16 is a target of miR-483-5p. Immunohistochemistry analysis showed RAI16 was moderate or strong staining in tumor tissues but negative or weak staining in adjacent non-tumor tissues. The sensitivity and specificity of RAI16 for HCC diagnosis were 70.6 and 93.6%, respectively, and increased to 80.9 and 92.0% when combined with glypican-3. Finally, overexpression or knock-down of RAI16 increased or decreased cell viability and colony formation in HCC cell lines and enhanced or inhibited tumor cell growth in xenograft nude mice. Mitogen-activated protein kinase/extracellular signal-regulated kinase (ERK) and transforming growth factor-β pathways were mostly affected by RAI16. RAI16 could activate the phosphorylation of ERK1/2 and SMAD2/3. In conclusion, RAI16 may serve as a useful therapeutic agent for HCC gene therapy and tumor marker for HCC diagnosis.


Introduction
Hepatocellular carcinoma (HCC) is common tumor worldwide and the leading cause of cancer-related death (1).HCC tumorigenesis is a multifactorial and multistep process that involves activating oncogenes and inactivating tumor suppressor genes in different stages of HCC progression (2).Clarifying and investigating the roles of the genes involved in HCC development will contribute to our understanding of the mechanisms of tumorigenesis and application of efficient new targets for the early diagnosis and effective treatment of HCC (3).
Our previous work reported the significant downregulation of miR-483-5p in HCC tumor tissues (4).By TargetScan, retinoic acid induced 16 (RAI16) was predicted as a target of miR-483-5p.RAI is a family of proteins induced by retinoic acid.RAI is conserved in human, rat, mouse, rabbit and zebrafish by homology analysis, which indicated that RAI might play an important role in basic cell function.RAIl was related to Smith-Magenis syndrome (5,6); RAI2 was reported to play a role in transmembrane signal transduction (7,8); RAI3 was identified as a therapeutic target for breast cancer (9,10); RAI14 was involved in NAD(P)H oxidase-dependent CD44 and BMP4Id pathway in restenosis and atherosclerosis (11,12); RAI17 regulated by alpha adducing (Add1) might contribute to endothelial rarefaction dysfunction in hypertension (13,14).To our knowledge, there is no functional study on RAI16.
In the present study, we identified RAI16 as a target of miR-483-5p, evaluated the potential value of RAI16 as a tumor marker for HCC and further explored the possible role of RAI16 in HCC tumorigenesis in vitro and in vivo.

Cell lines and tissue specimens
Human embryonic kidney cells (HEK293T) and human HCC cells lines HepG2 and Huh7 were cultured in Dulbecco's modified Eagle's medium plus 10% fetal bovine serum (Invitrogen, Carlsbad, CA) at 37°C in a humidified atmosphere containing 5% CO 2 .
Human HCC tumor and adjacent non-tumor tissues were obtained and prepared as described previously (4).Written consent for tissue donation (for research purposes) was obtained from the patients before tissue collection and the protocol was approved by the Institutional Review Board of Eastern Hepatobiliary Surgery Hospital and Second Military Medical University.Clinical and pathological information of patients were profiled previously (4).
Quantitative real-time PCR Synthesis of complementary DNA (cDNA) and quantitative real-time PCR analysis of miR-483-5p expression was carried out with TaqMan microRNA assay kits (Applied Biosystems) according to the manufacturer's protocol.Briefly, total RNA was extracted using TRIzol Reagent (Invitrogen) from HCC tissues or cell lines and were used to synthesize cDNAs with miR-483-5p specific primer.The reactions were incubated for 30 min at 16°C, 30 min at 42°C, 5 min at 85°C, and then held at 4°C.The cDNA product was used for the following quantitative real-time PCR analysis directly.Reactions were incubated at 95°C for 5 min, followed by 40 cycles at 95°C for 15 s and at 60°C for 1 min.PCR reactions were run on a StepOne Plus real time PCR machine (Applied Biosystems) and the data were analyzed by SDS v2.3 software.

Immunohistochemistry
The paraffin-embedded blocks of tumor and adjacent non-tumor tissue from 18 HCC specimens were fixed on slide glasses.In addition, tissue arrays (L01, L02, L03; SuperChip, Shanghai, China) were used for immunohistochemistry, which contained 126 paired HCC liver tissues (40 well differentiated, 65 moderately differentiated and 21 poorly differentiated HCC).Tissues were deparaffinized in xylene and rehydrated by reducing the concentration of ethanol (100 and 70%, 5 min each).Antigens were unmasked with microwave irradiation for 5 min in pH 6.0 citric buffer three times.The slides were incubated with RAI16 or glypican-3 antibodies (1:100 dilution) at 4°C overnight, followed by incubation with horseradish peroxidase-conjugated goat anti-rabbit or goat anti-mouse antibody (1:1000 dilution, KPL) at room temperature for 60 min.Finally, 3,3′-diaminobenzidine tetrahydrochloride was used for signal development, and 20% hematoxylin was used for counterstaining.The slides were dehydrated, cleared and evaluated by three pathologists.The expression level of RAI16 or glypican-3 was scored as 0 (negative), 1 (weak), 2 (moderate) or 3 (strong) by the staining intensity of the protein.Receiver operating characteristic curve between HCC tumor and adjacent non-tumor tissues was plotted and the sensitivity and the specificity of RAI16 were determined by SPSS 16.0 v.

Cell viability assay
Twenty-four hours after transfection, 1000 transfected HepG2 and Huh7 cells were placed in a fresh 96-well plate in triplicate and maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum for 5 days.Cells were tested for proliferation per 24 h using Cell Titer-Blue cell viability assay (Promega Corporation) according to the manufacturer's instructions and the fluorescence value was recorded by multi-plate reader (Synergy 2).

Colony formation assay
Twenty-four hours after transfection, 2000 transfected HepG2 and Huh7 cells were placed in a fresh 6-well plate in triplicate and maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum for 2 weeks.Cell colonies were fixed with 20% methanol and stained with 0.1% coomassie brilliant blue R250 at room temperature for 15 min.The colonies were counted by ELIspot Bioreader 5000 (BIO-SYS, Karben, GE).

Cell cycle analysis
Forty-eight hours after transfection, 1 × 10 5 transfected HepG2 and Huh7cells were harvested, washed once in PBS, and fixed in 70% ethanol at 4°C overnight.Staining for DNA content was performed with 50 mg/ml propidium iodide and 1 mg/ml RNase A at room temperature for 30 min.Populations in G 0 -G 1 , S and G 2 -M phase were measured by Cell Lab Quanta SC flow cytometry (Beckman Coulter, Fullerton, CA) and the data were analyzed by FlowJo v7.6 software.

Apoptosis assay
Forty-eight hours after transfection, 1 × 10 5 transfected HepG2 and Huh7cells were harvested, washed once in PBS, resuspended and incubated with

Retinoic acid induced 16 tumorigenesis and tumor marker
FITC-Annexin V (Promega Corporation) for 15 min at 4°C in the dark, according to the manufacturer's instructions.After staining, the cells were incubated with propidium iodide for 5 min at 4°C in the dark and then analyzed by Cell Lab Quanta SC flow cytometry and the data were analyzed by FlowJo v7.6 software.

Tumorigenicity assay in xenograft nude mice
Male BALB/c nude mice (5-6 weeks of age) were obtained from Shanghai Experimental Animal Center (Shanghai, China).Animal handling and experimental procedures were approved by the Animal Experiments Ethics Committee of Second Military Medical University.For in vivo tumorigenicity assay, all pyrimidine nucleotides in the RAI16 siRNA or NC siRNA were substituted by their 2′-O-methyl analogs to improve RNA stability.HepG2 cells (1 × 10 5 ) stably expressing RAI16 or transfected with RAI16 siRNA were suspended in 100 µl PBS and then injected subcutaneously into the left side of the posterior flank of six BALB/c nude mice, respectively.NC-siRNA-or pcDNA3.1-transfectedHepG2 cells (1 × 10 5 ) were injected subcutaneously into the right side of same 12 mice.Tumor growth was examined daily and the tumor volumes were calculated every week using the formula for hemi-ellipsoids: V = length (cm) × width (cm) ×height (cm) × 0.5236.After 5 weeks, the mice were killed and the tumors were dissected and photographed.

Cancer pathway reporter array
HepG2 cells stably expressing RAI16 or transfected with RAI16 siRNA were assayed by Cignal Finder 10 Cancer Pathway Reporter Array (Qiagen, Valencia, CA) according to the manufacturer's protocol.Relative firefly luciferase activity was calculated and normalized to the constitutively expressed Renilla luciferase.

Statistical analysis
All experiments were performed at least three times, and the data are presented as mean ± SD.Comparisons were made by using a two-tailed t-test or one-way analysis of variance for experiments with more than two subgroups.Correlation analysis was made by using Spearman correlation coefficient.P < 0.01 was considered statistically significant.

RAI16 is a target of miR-483-5p
MiR-483-5p was found to be downregulated in 77.8% (14/18) HCC tumor tissues compared with adjacent non-tumor tissues (Figure 1A).RAI16 was found to be upregulated in 94.4% (17/18) HCC tumor tissues compared with adjacent non-tumor tissues (Figure 1B).A negative correlation was found between the upregulated RAI16 protein and downregulated miR-483-5p (r = −0.726,P < 0.01) (Figure 1C).Furthermore, RAI16 is predicted as a potential target of miR-483-5p by TargetScan and miRanda but not by PicTar.The 3′-UTR of RAI16 mRNA contained a complementary site for the seed region of miR-483-5p (Figure 1D).These results indicated that RAI16 may be regulated by miR-483-5p and both of them may be involved in HCC tumorigenesis.
To validate whether RAI16 is a target of miR-483-5p, a human RAI16 3′-UTR fragment containing wild-type or mutant miR-483-5p binding sequence (Figure 1D) was cloned into the downstream of the firefly luciferase reporter gene of pGL3.In HEK293 cells cotransfected with the reporter plasmids and miR-483-5p mimic or NC duplex, the luciferase activity of the reporter that contained wild-type 3′-UTR was significantly suppressed by miR-483-5p mimic, but the luciferase activity of mutant reporter was unaffected (Figure 1E), indicating that miR-483-5p may suppress gene expression through miR-483-5p binding sequence at the 3′-UTR of RAI16.Furthermore, transfection of miR-483-5p mimic decreased RAI16 expression and transfection of miR-483-5p inhibitor increased RAI16 expression in HepG2 cells at protein (Figure 1F) but not mRNA level (data not shown), suggesting that RAI16 expression could be inhibited by miR-483-5p at posttranscriptional level.

Immunohistochemical analysis of RAI16 expression in HCC liver tissues
The expression of RAI16 was investigated in 18 paired HCC by immunohistochemistry analysis.As shown in Figure 2A, strong staining of RAI16 was observed in the cytoplasm (90%) and plasma membrane (10%) of tumor cells, but weak or negative staining in non-tumor tissues, which confirmed the upregulation of RAI16 in HCC tumor tissues and major cytoplasm localization in HCC liver cells.
To validate the usefulness of RAI16 staining for the diagnosis of HCC, commercial HCC tissue arrays were used, in which the degree of tumor differentiation had been proven (Table I).Most HCC tumor tissues showed moderate (score 2) or strong (score 3) RAI16 expression (89 of 126 cases).In contrast, most non-tumor tissues showed negative (score 0) or weak (score 1) RAI16 expression (118 of 126 cases) (Table IIA).The sensitivity and specificity of RAI16 expression for the diagnosis of HCC were 70.6 and 93.6%, respectively (Supplementary Figure S1, available at Carcinogenesis Online and Table IIB).Glypican-3 has been reported as a promising marker for HCC (15,16).The sensitivity and specificity of glypican-3 were 60.3 and 96.0%, respectively (Table IIB).Strikingly, the sensitivity and specificity increased to 80.9 and 92.0% when glypican-3 was used with RAI16 (Supplementary Figure S1, available at Carcinogenesis Online and Table IIB).These results indicate that RAI16 may be potential tumor marker for HCC diagnosis, especially combined with glypican-3.In addition, RAI16 expression level in tumor tissues correlates with tumor differentiation (r = 0.974, P < 0.01) (Figure 2B and 2C and Table IIA), but not other clinicopathological features (age, sex, stage and tumor size) (data not shown).These results indicate that RAI16 may be associated with cell differentiation and immunostaining of RAI16 could contribute to the pathological diagnosis of HCC.

RAI16 induces HCC tumorigenesis in vitro and in vivo
The significant upregulation of RAI16 expression in HCC tissues indicated possible biological significance in tumorgenesis.At first, the effect of RAI16 on cell growth was evaluated in HepG2 and Huh7 cells transfected with pcDNA3.1-RAI16or RAI16 siRNA.The expression of RAI16 was increased 2.5-fold (HepG2) and 2.2fold (Huh7) in cells transfected with 1 µg pcDNA3.1-RAI16,but deceased 3.0-fold (HepG2) and 4.3-fold (Huh7) in cells transfected with 20 nM RAI16 siRNA (Figure 3A).The overexpressed RAI16 is predominately located in cytoplasm of HepG2 cells (Supplementary Figure S2, available at Carcinogenesis Online).From 2 days (HepG2) or 3 days (Huh7) after subculture, the viability of cells transfected with pcDNA3.1-RAI16significantly increased compared with that of empty vector pcDNA3.1-transfectedor non-transfected cells, but the viability of cells transfected with RAI16 siRNA significantly decreased (Figure 3B).These results indicate RAI16 could enhance HCC cell growth.
To further confirm the above findings, an in vivo mouse model was used.For the duration of the treatment with HepG2 cells stably expressing RAI16 or transfected with RAI16 siRNA, tumor volume curves revealed a significant increase in growth rates at the third, fourth and fifth week after treatment with HepG2 cells stably expressing RAI16 and a significant decrease in growth rates at the fourth and fifth week after treatment with HepG2 cells transfected with RAI16 siRNA whereas no significant differences in tumor growth rates were observed between the NC-siRNA-transfected or non-transfected HepG2 cells groups (Figure 3F).These results indicate that introduction of RAI16 significantly promotes tumorigenesis in xenograft nude mouse model.
To identify whether miR-483-5p is involved in enhancement of cell growth by RAI16, the effects of miR-483-5p were also examined.The expression of miR-483-5p was increased 25-fold   shown).These results indicate that miR-483-5p could inhibit the cell growth, consistent with its downregulatory effect on RAI16.

RAI16 activates MAPK/ERK and TGF-β pathways in HCC
To investigate the mechanism of RAI16 enhancement of tumor growth in HCC, 10 major cancer-related pathways were analyzed in transfected HepG2 cells using a dual-luciferase reporter system (Promega).Figure 4A shows the relative reporter expression in HepG2 transfected with pcDNA3.1-RAI16normalized to that with pcDNA3.1.The most affected pathways were transforming growth factor-β (TGF-β), mitogen-activated protein kinase (MAPK)/ERK and Wnt. Figure 4B shows the relative reporter expression in HepG2 transfected with RAI16 siRNA normalized to that with NC siRNA.The most affected pathways were TGF-β, MAPK/ERK and nuclear factor κB.The upregulated MAPK/ERK and TGF-β pathways in RAI16 overexpressed cells are consistent with the downregulated MAPK/ERK and TGF-β pathways in RAI16 knockdown cells.These results indicate MAPK/ERK and TGF-β pathways may be involved in RAI16 induction of tumorigenesis.
To verify above findings, the key molecules in MAPK/ERK and TGF-β pathways were detected by immunoblotting.RAI16 overexpression increased the phosphorylation of MEK and ERK1/2 significantly (Figure 4C), even the total ERK (Figure 4C) but not total MEK expression (data not shown).Similar phenomena were observed in TGF-β pathway, and the phosphorylation of SMAD2/3 and total SMAD2/3, together with SMAD4, was increased by RAI16 overexpression (Figure 4D).RAI16 knockdown led to the reverse results.Interestingly, phosph-SMAD2/3 and SAMD4 in HCC tumor tissues were increased much more significantly than that in RAI16-transfected cells (Figure 4D), but phosphorylation of MEK and ERK1/2 seemed more significant in RAI16-transfected cells (Figure 4C).These results suggest it is possible that MAPK/ERK and TGF-β pathways play the role in different stages of RAI16 induction of tumorigenesis.

Discussion
Identifying the miRNAs and their targets that are essential for HCC progression may provide promising therapeutic opportunities.Previously, 20 deregulated miRNAs were identified by Taqman low density miRNA array and quantitative real-time PCR (4).The most significant miRNAs, including upregulated miR-520g, miR-217 and miR-518e and downregulated miR-483-5p, miR-138 and miR-199a-5p were selected for further study (4).MiR-483-5p was reported to be associated with histological subtypes in ovarian cancers (17), poorer adrenocortical carcinomas survival (18) and malignant pheochromocytoma (19).MiR-483-5p could act as an endogenous angiogenesis-inhibiting factor by targeting serum response factor (20). High expression of miR-483-5p appeared to be a marker for adrenocortical malignancies (21,22).MiR-483-5p suppresses the proliferation of glioma cells (23).All of these findings indicate miR-483-5p plays important role in tumorigenesis.We showed miR-483-5p regulation of RAI16 in HCC may provide new insights into the regulation of miRNA in tumorigenesis.

Retinoic acid induced 16 tumorigenesis and tumor marker
RAI16 was upregulated in majority of HCC tumor tissues (17 of 18 cases) compared with adjacent non-tumor tissues by immunoblotting that drove us to verify its expression by immunohistochemistry and evaluate the potential diagnostic value for HCC.Consistent with immunoblotting data, RAI16 also showed strong staining in HCC tumor tissue by immunohistochemistry analysis, whereas weak staining was observed in most non-tumor tissues.These results indicated that RAI16 may play a role in HCC tumorigenesis.Moreover, upregulated RAI16 expression in HCC was significantly associated with histological differentiation.Most of the moderately and poorly differentiated tumor samples were positive for RAI16 expression, but RAI16 expression was profoundly weaker in the welldifferentiated tumor samples.Thus, the expression of RAI16 might be associated with the differentiation state of tumor cells.RAI16 was further found to be useful for the pathological diagnosis of HCC.The sensitivity of RAI16 for HCC diagnosis is even much better than GPC-3 in our analyses.In addition, the sensitivity and specificity significantly increased when the combination of GPC-3 and RAI16 was used.Thus, RAI16 is diagnostically useful for screening of HCC.Accuracy of RAI16 application in HCC diagnosis needs validation in larger populations by multi-investigators.
Although some researchers mentioned it in their work on cell growth and differentiation or human cDNAs sequencing (24,25) (29,30).In the present study, RAI16 showed great potential as a tumor marker for HCC, so it is very important to investigate its role in HCC tumorigenesis.Plasmid expressing RAI16 and RAI16 specific siRNA were transfected into HepG2 and Huh7 cells.RAI16 overexpression significantly enhanced cell viability and colony formation but RAI16 knockdown inhibited cell viability and colony formation.In xenograft nude mouse, RAI16 overexpression also enhanced tumor growth.RAI16 overexpression also promoted cell cycle and inhibited apoptosis of HCC cells.These findings indicate that RAI16 enhances tumorigenesis by promotion of cell cycle and inhibition of apoptosis in vitro and in vivo, which strengthen the hypothesis that RAI16 acts as an oncogene.
In order to further examine the mechanism of RAI16 effect, 10 cancer-related pathways were assayed in RAI16 plasmid or siRNA-transfected HepG2 cells.MAPK/ERK and TGF-β pathways were found to be activated by RAI16.Immunoblotting analysis showed phosph-MEK, phosph-ERK and total ERK, as well as phosph-SMAD2/3, total SMAD2/3 and SMAD4, were upregulated both in RAI16-transfected HepG2 cells and in HCC tumor tissues.The MAPK/ERK signaling pathway is involved in diverse cellular processes such as cell survival, differentiation and proliferation (31), and the overexpression of members of this pathway was found to be correlated to HCC (32).The TGF-β pathway also regulates cell proliferation, differentiation and adhesion (33).These two pathways have been reported to play significant roles in HCC tumorigenesis (34).These results suggest MAPK/ ERK and TGF-β pathways are most probably involved in RAI16 induction of tumorigenesis.
In summary, the study found that: (i) RAI16 downregulated in the majority of the HCC tumor tissues was a potential tumor marker for HCC; (ii) RAI16 regulated by miR-483-5p enhance HCC tumorigenesis in vitro and in vivo possibly through MAPK/ERK and TGF-β pathways dependent mechanism (Figure 4E).Our findings provide evidence that RAI16 could function as an oncogene and a potential tumor marker in HCC, which is a useful therapeutic target for HCC early diagnosis and gene therapy.

W.
Wang et al.

Fig. 2 .Fig. 3 .
Fig. 2. Immunohistochemistry analysis of RAI16 expression in HCC tissues.(A) HCC tumor tissues (T) and adjacent non-tumor tissues (N) were stained with RAI16 antibody.HCC tumor tissues showed much stronger staining of RAI16 than adjacent non-tumor tissues.(B) RAI16 was stained in HCC tissue arrays.Strong or moderate staining of RAI16 in the cytoplasm of HCC tumor tissues but compared with negative or weak staining in non-tumor tissues was observed.Well: well-differentiated HCC; moderately: moderately differentiated HCC; poorly: poorly differentiated HCC.(C) RAI16 expression level correlates with tumor differentiation.r: correlation coefficient.

Fig. 4 .
Fig. 4. RAI16 activated MAPK/ERK and TGF-β pathways in HCC.(A and B) The relative reporter expression in HepG2 transfected with pcDNA3.1-RAI16(A) or RAI16 siRNA (B).(C and D) The effect of RAI16 on MAPK/ERK pathway (C) or TGF-β pathway (D).Glyceraldehyde 3-phosphate dehydrogenase was used as an internal control.(E) A schematic representation of the main findings in this study.

Table I .
Histology of HCC tissues on tissue array

Table II .
Immunohistochemical analysis from tissue array of HCC , RAI16 (also called FAM160B2) is almost a new one to us.The function study about RAI16 was from only one group: Deng et al. reported RAI16 interacted with tyrosine kinase express in hepatocellular carcinoma (Tec) by yeast two hybrid system (26-28); Luo et al. reported RAI16 suppressed the HepG2 cell proliferation and promoted cell differentiation