Primary liver cancer, specifically hepatocellular carcinoma (HCC), is known as the third cause of tumor-associated death worldwide. It is estimated that the world incidence of liver cancer is 10.1 cases per 100,000 person-years.¹ There are numerous risk factors contributing to the initiation of HCC, the most common of which includes alcohol abuse, infection of hepatitis B and C viruses, nonalcoholic fatty liver disease, and obesity.^2,3 Up to now, the most common and effective therapeutic strategy of HCC is the combination treatment of surgical resection and chemotherapy. However, HCC patients are usually diagnosed at advanced clinical stages due to the lack of efficient and accurate diagnosis measures, resulting in a surgical cure of less than 20%.⁴ The limitation of therapeutic options in HCC treatment causes a poor prognosis with 5 years of survival remains around 10% only.⁵ Thus, it is essential to dig the potential mechanisms associated with HCC initiation and progression, which must be contributed to the development of novel therapeutic measures for HCC patients.

Long noncoding RNAs (lncRNAs) are a group of highly conserved transcripts without protein-encoding capacity. It consists of more than 200 of nucleotides and widely exists in the mammal cells. The implication of lncRNAs in human cancer has well been documented by an amount of evidences.⁶ lncRNA cytoskeleton regulator RNA (CYTOR), also named LINC00152, is a novel lncRNA that firstly identified to be dysregulated in gastric cancer by Cao et al. in 2013.⁷ In the following years, lncRNA CYTOR has been reported to play a role in the process of tumorigenesis of various human tumors, such as ovarian cancer, colorectal cancer, and non-small lung cancer.^8-10 Although the role of CYTOR in HCC carcinogenesis has been reported, its potential mechanisms remain largely unclear.

HS1-associated protein X-1 (HAX-1) was firstly reported as a binding partner of HS1 protein, which involves in T cell maturation.¹¹ Since it is structurally similar with Bcl-2, a well-known anti-apoptosis protein, HAX-1 was proposed to represent a new protein that modulates apoptosis. This speculation has subsequently been demonstrated by numerous studies.^12,13 Recently, HAX-1 was found to be upregulated in HCC and demonstrated to promote HCC cell proliferation.¹⁴

miR-125a-5p is a member of microRNA family, recently, that has been reported to be downregulated in HCC and to repress HCC cell proliferation.¹⁵ In the present study, we revealed a miR-125a-5p binding site in CYTOR. Meanwhile, HAX-1 was predicted to be a target gene of miR-125a-5p. Therefore, we aimed to investigate whether CYTOR could affect the HCC cell apoptosis in vitro through miR-125a-5p/HAX-1 axis.

1. Blood samples

The blood samples were collected from 20 HCC patients and 20 healthy volunteers. This study was approved by the Ethics Committee of Aerospace Center Hospital, Peking University Aerospace School of Clinical Medicine (approval number: 2021-HDPY-004). Written informed consent was obtained by all participants. The samples were stored at –80°C. All blood samples were derived from examination on admission to hospital, and remaining samples after examination were tested.

2. Cell culture

Normal human liver cell line (LO-2) and HCC cell lines (Bel-7402, Bel-100, HepG2, Huh-7, and SMMC-7721) were obtained from Cell bank of Chinese Academy of Sciences (Shanghai, China). They were routinely cultured in the fetal bovine serum (10%) included Dulbecco Modified Eagle Medium (HyClone, Logan, UT, USA) at 37°C in an incubator filled with 5% CO₂ and 95% O₂.

3. Quantitative real-time polymerase chain reaction (qRT-PCR) analysis

TRIzol reagent (Invitrogen, Waltham, MA, USA) was employed to extract total RNAs of HCC cells according to the procedures of manufacturers. The quality of extracted RNAs were verified by examining the ratio of A260/A280, then, 2 μg RNA was utilized as template to synthesis cDNA by the use of Gibco BRL kit (Life Technologies, Carlsbad, CA, USA). The process of qRT-PCR was carried out on an ABI Prism system (7700; PE Applied Biosystems, Foster, CA, USA) using indicated primers. The relative RNA expression levels were calculated with the 2^−ΔΔCt method and normalized to that of GAPDH (internal control).

4. RNA transfection

Specific siRNA that targets CYTOR (siCYTOR) and its negative control (siNC), miR-125a-5p mimics, inhibitor and its negative control (miRNA NC), as well as specific siRNA targets HAX-1 (si-HAX-1) were all designed and purchased from Gene Pharma (Shanghai, China). Indicated RNAs were transfected into HCC cells using Lipofectamine 3000^TM reagent (Invitrogen).

5. Western blot assay

HCC cells were lysed with RIPA buffer (Beyotime, Shanghai, China) and total proteins were isolated using high-speed centrifugation at 10,000 g for 20 minutes. After the concentration was determined, 50 μg protein samples were loaded into and isolated by SDS-PAGE. Next, isolated proteins were transferred into nitrocellulose membrane. The membranes were incubated with primary antibodies that against HAX-1 (1:1,000, Abcam), Bcl-2 (1:1,000, Abcam), Bax (1:2,000, Abcam), cleaved caspase-9 (1:2,000, Abcam), and total caspase-9 (1:2,000, Abcam) overnight at 4°C. GAPDH (1:2,000, Abcam) was used internal control. After washed with phosphate buffered saline with three times, membranes were incubated with indicated secondary antibodies for 2 hours. Bands were visualized using enhanced chemiluminescence reagent (EMD Millipore, Burlington, MA, USA), and the difference was analyzed through Image-Pro software version 6.0 (Media Cybernetics, Inc., Rockville, MD, USA).

6. Cell Counting Kit-8 assay

After 24 hours of transfection of indicated RNAs, HCC cells were collected and seeded into 96-well plates containing in 200 μL Dulbecco’s Modified Eagle’s Medium at a concentration of 3,000 cells/well. Cell Counting Kit-8 solution (10 μL) was added into each well, and incubated at 37°C for another 2 hours. The optical density of each well was determined through a Microplate Reader (Bio Rad, Hercules, CA, USA) at 450 nm.

7. Flow cytometry analysis

Treated HCC cells were collected and fixed using 75% ethanol for 1 hour. Next, HCC cells were subjected to propidium iodide (5 μg/mL) and Annexin V-FITC (0.5 μg/mL) double staining for 10 minutes (protected from light). Cell apoptosis was detected through a FACSCalibur flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA).

8. Caspase-9 activity assessment

Protein samples (10 µg) were employed to measure caspase-9 activity using casepase-9 activity kit (C1158, Beyotime) following the procedure of manufacturers. Cells were harvested and incubated with the reaction buffer and Asp-Glu-Val-Asp-pNA substrate for 1 hour at 37°C followed by the measurement of absorbance (405 nm).

9. Terminal deoxynucleotidyl transferase dUTP nick-end labeling

An in situ cell death assessment kit (Roche Diagnostics, Basel, Switzerland) was adopted in terminal deoxynucleotidyl transferase dUTP nick-end labeling staining assay to estimate cell apoptosis according to the manufacturer's instruction. After washed three times with phosphate buffered saline, treated HCC cells (2×10⁶) were fixed with 4% paraformaldehyde. Apoptosis cells were labeled by the kit, and nuclei were labeled with DAPI (ab104139, Abcam).

10. Dual-luciferase reporter assay

The wild type (WT) and mutant (MUT) reporter plasmids of CYTOR were established by inserting the WT and MUT miR-125a-5p binding fragments into pGL3 vector (Promega, Madison, WI, USA), named as CYTOR-WT and CYTOR-MUT. Establishment of HAX-1-WT and HAX-1-MUT reporter plasmids were carried out same as above. After 24 of culture in 96-well plates, HCC cells (1×10⁷ cells/well) were co-transfected with CYTOR-WT or CYTOR-MUT and miR-125a-5p mimics or miRNA NC using Lipofectamine 3000 (Invitrogen). The alteration of luciferase activity was examined by Dual-Luciferase Reporter Assay System (Promega). Validation of the interaction between miR-125a-5p and HAX-1 was completed as same as above.

11. Statistical analysis

Data in the present study were expressed as mean± standard deviation. SPSS version 19.0 (IBM Corp., Armonk, NY, USA) was employed to analyze the difference between groups using the Student t-test or one-way analysis of variance. A p-value less than 0.05 was considered statistically significant.

1. lncRNA CYTOR was highly expressed while miR-125a-5p was lowly expressed in HCC

To determine whether CYTOR and miR-125a-5p play a role during HCC carcinogenesis, we firstly examined their levels in the blood samples collected from HCC patients and healthy controls. Results indicated that CYTOR was elevated while miR-125a-5p was decreased in the HCC samples compared to normal samples (Fig. 1A and B). We also analyzed the correlation between the expression of CYTOR and miR-125a-5p and found a negative correlation between them in HCC samples (Fig. 1C). We also tested their expression in normal liver LO-2 cell line and HCC cell lines (Bel-7402, Bel-100, HepG2, Huh-7, and SMMC-7721) by qRT-PCR. Results indicated that the expression level of CYTOR was dramatically upregulated in HCC cell lines compared to normal liver LO-2 cell line (Fig. 1D). On the contrary, the expression level of miR-125a-5p was found to be significantly downregulated in HCC cell lines compared to normal liver LO-2 cell line (Fig. 1E). Moreover, we found the expression levels of CYTOR and miR-125a-5p were most dramatically changed in HepG2 and SMMC-7721 HCC cell lines. These findings suggested that CYTOR was highly expressed while miR-125a-5p was lowly expressed in HCC, implying they might play a role in HCC tumorigenesis.

Figure 1. Long noncoding RNA CYTOR was highly expressed while miR-125a-5p was lowly expressed in HCC. (A) CYTOR and (B) miR-125a-5p expression in the blood samples of HCC patients and healthy volunteers were detected by qRT-PCR. (C) A correlation was found between the expression of CYTOR and miR-125a-5p in HCC samples. (D) CYTOR and (E) miR-125a-5p expression in five HCC cell lines (Bel-7402, Bel-100, HepG2, Huh-7, and SMMC-7721) was analyzed via qRT-PCR, and LO-2 was used as a control.
CYTOR, cytoskeleton regulator RNA; HCC, hepatocellular carcinoma; qRT-PCR, quantitative real-time polymerase chain reaction. *p<0.05, ^†p<0.01, ^‡p<0.001.

2. Silencing of lncRNA CYTOR facilitated HCC cell apoptosis.

To further determine the functions of CYTOR in HCC, we estimated the impacts of CYTOR silencing on HCC cell proliferation and apoptosis in HepG2 and SMMC-7721 cells. After 24 hours of siCYTOR transfection, a significant downregulation of CYTOR was observed in both HepG2 and SMMC-7721 cells (Fig. 2A), indicating that siCYTOR could dramatically knockdown the expression of CYTOR in HCC. In Cell Counting Kit-8 assay, a dramatic repression was observed in siCYTOR transfected HepG2 and SMMC-7721 cells compared to siNC transfected cells (Fig. 2B). Results from flow cytometry analysis suggested that CYTOR knockdown resulted in a remarkable upregulation of cell apoptosis in both HepG2 and SMMC-7721 cells (Fig. 2C). Consistently, the intensity of transferase dUTP nick-end labeling staining was stronger in siCYTOR transfected group than siNC group (Fig. 2D). The activity of caspase-9 was found to be significantly increased in CYTOR blocked HepG2 and SMMC-7721 cells (Fig. 2E). Moreover, by using Western blot, we revealed that HAX-1 and Bcl-2 were significantly downregulated while Bax and cleaved caspase-9 were significantly upregulated in siCYTOR transfected HepG2 and SMMC-7721 cells (Fig. 2F). Taken together, these results indicated that silencing of CYTOR repressed HCC cell proliferation and facilitated HCC cell apoptosis.

Figure 2. Silencing of long noncoding RNA CYTOR facilitated hepatocellular carcinoma cell apoptosis. (A) Knockdown efficiency of siCYTOR was tested by qRT-PCR in HepG2 and SMMC-7721 cells. (B) Cell viability of HepG2 and SMMC-7721 cells was detected by Cell Counting Kit-8 assay after 1, 2, 3, and 4 days of siCYTOR or siNC transfection. Cell apoptosis in HepG2 and SMMC-7721 cells was analyzed using (C) flow cytometry following propidium iodide and Annexin V double staining and (D) TUNEL staining (scale bar=200 μm). (E) Activity of caspase-9 of HepG2 and SMMC-7721 was evaluated in cells transfected with siNC or siCYTOR. (F) Western blot analysis of cleaved caspase-9, total caspase-9, HAX-1, Bcl-2, and Bax was performed in HepG2 and SMMC-7721 cells transfected with siNC or siCYTOR.
CYTOR, cytoskeleton regulator RNA; OD, optical density; siCYTOR, siRNA of CYTOR; siNC, siRNA of negative control; qRT-PCR, quantitative real-time polymerase chain reaction; TUNEL, terminal deoxynucleotidyl transferase dUTP nick-end labeling. *p<0.05, ^†p<0.01, ^‡p<0.001.

3. lncRNA CYTOR acted as a sponge of miR-125a-5p

We then investigated whether CYTOR could interact with miR-125a-5p in HCC cells. According to the bioinformatics prediction (starBase, http://starbase.sysu.edu.cn/), we found a miR-125a-5p binding site in CYTOR (Fig. 3A). To validate the interaction between CYTOR and miR-125a-5p, dual-luciferase reporter assay was carried out in HepG2 and SMMC-7721 cells. The luciferase intensities of HepG2 and SMMC-7721 cells driven by CYTOR-WT were significantly attenuated by the transfection of miR-125a-5p mimics, but not mimics NC (Fig. 3B). Moreover, in HepG2 and SMMC-7721 cells, CYTOR overexpression was found to result in a significant downregulation of miR-125a-5p, and its knockdown caused a remarkable upregulation of miR-125a-5p (Fig. 3C). These findings indicated that CYTOR bind to and negatively regulated miR-125a-5p in HCC cells.

Figure 3. Long noncoding RNA CYTOR acted as a sponge of miR-125a-5p. (A) Complementary sequence between miR-125a-5p and wild type (WT) CYTOR; the putative miR-125a-5p binding site was mutated (MUT) in CYTOR. (B) HepG2 and SMMC-7721 cells cotransfected with miR-125a-5p and CYTOR-WT or CYTOR-MUT plasmid were detected via luciferase intensity. (C) Relative miR-125a-5p expression of HepG2 and SMMC-7721 cells were examined after overexpression or knockdown of CYTOR.
CYTOR, cytoskeleton regulator RNA; siCYTOR, siRNA of CYTOR; siNC, siRNA of negative control. *p<0.05, ^†p<0.01.

4. HAX-1 was targeted by miR-125a-5p

Furthermore, bioinformatics analysis (starBase, http://starbase.sysu.edu.cn/) revealed an miR-125a-5p binding site in the 3’-UTR of HAX-1 mRNA (Fig. 4A), and in dual-luciferase reporter assay, we demonstrated that miR-125a-5p, but not mimics NC, transfection could reduce the luciferase activities of HepG2 and SMMC-7721 cells driven by HAX-1-WT (Fig. 4B). To further evaluate the correlation between the expression levels of miR-125a-5p and HAX-1 in HCC cells, we measured the expression of HAX-1 in miR-125a-5p overexpressed and silenced HepG2 and SMMC-7721 cells by qRT-PCR and Western blot. The overexpression and knockdown efficiency of miR-125a-5p mimics and inhibitor was verified in HepG2 and SMMC-7721 cells, respectively (Fig. 4C). Overexpression of miR-125a-5p led to a significant downregulation of HAX-1 mRNA and protein in both HepG2 and SMMC-7721 cells, on the contrary, knockdown of miR-125a-5p caused an upregulation of HAX-1 (Fig. 4D and E). These results indicated that HAX-1 was targeted by miR-125a-5p in HCC cells.

Figure 4. HAX-1 was targeted by miR-125a-5p. (A) Complementary sequence between miR-125a-5p and wild type (WT) HAX-1; the putative miR-125a-5p binding site was mutated (MUT) in HAX-1. (B) HepG2 and SMMC-7721 cells were subjected to luciferase examination after 24 hours of co-transfection with miR-125a-5p and HAX-1-WT or HAX-1-MUT vector. Relative (C) miR-125a-5p and (D) HAX-1 mRNA expression of HepG2 and SMMC-7721 cells were estimated after miR-125a-5p mimics or inhibitor treatment. (E) HAX-1 protein expression of HepG2 and SMMC-7721 cells was assessed by Western blot after miR-125a-5p mimics or inhibitor treatment.
HAX-1, HS1-associated protein X-1; NC, negative control. *p<0.05, ^†p<0.01.

5. lncRNA CYTOR repressed HCC cell apoptosis via miR-125a-5p/HAX-1 cascade

As results above showed that CYTOR acted as a sponge of miR-125a-5p, and miR-125a-5p directly targets HAX-1 mRNA, we wondered whether CYTOR/miR-125a-5p/HAX-1 axis is involved in the events of HCC tumorigenesis. Transfection of siCYTOR or miR-125a-5p alone resulted in a significant upregulation of miR-125a-5p and a significant downregulation of HAX-1 in HepG2 and SMMC-7721 cells, and this phenomenon could be enhanced by the co-transfection of siCYTOR and miR-125a-5p (Fig. 5A and B). By using Western blot, we demonstrated that HAX-1 and Bcl-2 were downregulated, while Bax and cleaved caspase-9 were upregulated in siCYTOR treated cells compared to siNC cells, and co-transfection of siCYTOR and miR-125a-5p enhanced the siCYTOR induced dysregulations of HAX-1, Bcl-2, Bax and caspase-9 (Fig. 5C). Moreover, results from flow cytometry analysis indicated that miR-125a-5p transfection could promote the siCYTOR transfection induced cell apoptosis in HepG2 and SMMC-7721 cells (Fig. 5D). Transfection of miR-125a-5p mimics or siHAX-1 alone caused a remarkable downregulation of HAX-1 mRNA, and co-transfection of miR-125a-5p mimics and siHAX-1 facilitated this phenomenon in HepG2 and SMMC-7721 cells (Fig. 5E). The activity of caspase-9 was found to be increased in miR-125a-5p or siHAX-1 transfected HepG2 and SMMC-7721 cells, and co-transfection of miR-125a-5p and siHAX-1 further increased its activity (Fig. 5F). In Western blot assay, we found that transfection of miR-125a-5p or siHAX-1 alone resulted in a downregulation of HAX-1 and Bcl-2, as well as an upregulation of Bax and cleaved caspase-9 in HepG2 and SMMC-7721 cells. Moreover, co-transfection of miR-125a-5p and siHAX-1 promoted the effects of miR-125a-5p or siHAX-1 transfection alone on the dysregulations of HAX-1, Bcl-2, Bax and caspase-9 (Fig. 5G). These findings suggested that CYTOR repressed HCC cell apoptosis via miR-125a-5p/HAX-1 cascade.

Figure 5. Long noncoding RNA CYTOR repressed hepatocellular carcinoma cell apoptosis via miR-125a-5p/HAX-1 cascade. After transfection with siCYTOR, miR-125a-5p, and siCYTOR+miR-125a-5p, HepG2 and SMMC-7721 cells were subjected to (A, B) qRT-PCR analysis of miR-125a-5p and HAX-1 mRNA, (C) Western blot analysis of HAX-1, Bcl-2, Bax, cleaved caspase-9, and total caspase-9, and (D) flow cytometry analysis of cell apoptosis. After transfection with miR-125a-5p or siHAX-1, HepG2 and SMMC-7721 cells were subjected to (E) qRT-PCR analysis of HAX-1 mRNA, (F) caspase-9 activity assessment, and (G) Western blot analysis of HAX-1, Bcl-2, Bax, cleaved caspase-9, and total caspase-9.
CYTOR, cytoskeleton regulator RNA; siCYTOR, siRNA of CYTOR; qRT-PCR, quantitative real-time polymerase chain reaction; HAX-1, HS1-associated protein X-1; NC, negative control. *p<0.05, ^†p<0.01, ^‡p<0.001.

The global incidence of HCC is still increasing in recent years, resulting in a steady increase of the burden of HCC care. To better understand the potential molecular mechanisms of HCC and develop novel promising molecular targets for HCC diagnosis and therapy, research hotpots of HCC have been focused on a famous RNA molecule, lncRNA. In recent years, due to the rapid development of next-generation sequencing technology, lncRNA expression patterns of HCC were investigated by increasing studies.^16,17 In 2015, lncRNA CYTOR (LINC00152) was reported by Li et al.¹⁸ to be increased in HCC, acting as a novel potential biomarker in predicting HCC diagnosis. In the following years, the upregulation of CYTOR in HCC was demonstrated by multiple studies.¹⁹ For instance, CYTOR was found to facilitate cell cycle progression in HCC via miR-193a/b-3p/CCND1 cascade.²⁰ Moreover, CYTOR/miR-215/CDK13 pathway was reported to be contributed to the progression of HCC.²¹ Recently, increased CYTOR in HCC was proved to facilitate HCC cell proliferation through miR-125b/SEMA4C axis.²² Consistent with previous conclusions, we also observed a remarkable upregulation of CYTOR in HCC. And functional assays indicated that CYTOR knockdown caused a remarkable promotion of HCC cell apoptosis. In addition, CYTOR silencing resulted in a significant increase of caspase-9 in HCC cells. Taken together, existing evidence indicates that CYTOR server as oncogene of HCC.

Besides lncRNA, miRNA has also been shown a critical role in the tumorigenesis of HCC. Dysregulation of miRNAs is frequently observed in HCC patients, implying that miRNA might be a promising agent for HCC diagnosis and therapy. In HCC, miR-125a-5p was revealed to be downregulated and acted as a tumor suppressor.²³ Moreover, a recent study has proven that miR-125a-5p was downregulated in HCC and acted as a target of CYTOR in regulating HCC progression.²⁴ In line with previous conclusions, miR-125a-5p was also found to be downregulated in this study in HCC. Moreover, based on competing endogenous RNA theory, lncRNA/miRNA/mRNA axis is one of the most common regulatory mechanisms of tumor pathogenesis.²⁵ In this study, we also proved that CYTOR acted as a sponge of miR-125a-5p in HCC cells. Overexpression of miR-125a-5p could enhance the promotive effects of siCYTOR on HCC cell apoptosis, suggesting CYTOR regulated HCC progression by miR-125a-5p.

The anti-apoptosis effects of HAX-1 have been well demonstrated on various human cancers, including glioblastoma, prostate cancer and breast cancer.^12,26,27 In our previous study, we found that HAX-1 was overexpressed and promoted the proliferation of HCC cells.²⁸ Moreover, it could facilitate HCC cell migration and invasion by inducing epithelial-mesenchymal transition through NF-kB cascade.²⁹ Moreover, the activity of HAX-1 in tumor cells was regulated by multiple miRNAs.^27,30 In this study, HAX-1 was bind and negatively regulated by miR-125a-5p in HCC cells, moreover, knockdown of HAX-1 enhanced the miR-125a-5p overexpression induced HCC apoptosis. This is the first time we have reported the relationship between miR-125a-5p and HAX-1 in HCC. Evidences have shown that HAX-1 could regulate tumor cell invasion and migration through various of mechanisms. It was reported to form a tetramer with Galopa13 via glucocorticoids and Rac, and enhancing tumor cell migration by elevating Galpha13 activity.¹ HAX-1 can also directly bind to uPAR and activate uPAR-related signal transduction processes, thus, promoting tumor cell proliferation.² Other studies have shown that HAX-1 could affect the progression of human cancers by regulating multiple pathways, such as NF-kappa B and AKT1 pathways.^3,4 Thus, we will focus on these pathways in the future to as the potential downstream pathways of HAX-1.

In conclusion, our results provided evidence that CYTOR modulated the apoptosis of hepatoma cells by regulating miR-125a-5p/HAX-1 axis to induce caspase-9 inactivation. In view of the significant dysregulation of CYTOR, it might be a promising biomarker for HCC diagnosis. In addition, CYTOR/miR-125a-5p/HAX-1 axis might be a prospective therapeutic target for HCC.

This work was supported by Science Foundation of CASIC (2020-LCYL-009), the Science Foundation of ASCH (YN202104), and the Hygiene and Health Development Scientific Research Fostering Plan of Haidian District Beijing (HP2021-19-50701).

No potential conflict of interest relevant to this article was reported.

Study concept and design: Z.Y.W. Data acquisition: Z.Y.W., H.H. Data analysis and interpretation: X.N.A. Drafting of the manuscript: Z.Y.W. Critical revision of the manuscript for important intellectual content: Y.W., Z.Y.W. Statistical analysis: X.N.A., Q.Z. Obtained funding: Z.Y.W. Administrative, technical, or material support; study supervision: Y.W., Y.G.Q. Approval of final manuscript: all authors.