Up‐regulation of miR‐10b‐3p promotes the progression of hepatocellular carcinoma cells via targeting CMTM5

Abstract In this study, we investigated how miR‐10b‐3p regulated the proliferation, migration, invasion in hepatocellular carcinoma (HCC) at both in vitro and in vivo levels. CMTM5 was among the differentially expressed genes (data from TCGA). The expression of miR‐10b‐3p and CMTM5 was detected by qRT‐PCR and Western blot (WB). TargetScan was used to acquire the binding sites. Dual‐luciferase reporter gene assay was used to verify the direct target relationship between miR‐10b‐3p and CMTM5. WB analysis proved that miR‐10b‐3p suppressed CMTM5 expression. Furthermore, proliferation, invasion and migration of HCC cells were measured by MTT assay, colony formation assay, transwell assay and wound‐healing assay, respectively. Kaplan‐Meier plotter valued the overall survival of CMTM5. Finally, xenograft assay was also conducted to verify the effects of miR‐10b‐3p/CMTM5 axis in vivo. Up‐regulation of miR‐10b‐3p and down‐regulation of CMTM5 were detected in HCC tissues and cell lines. CMTM5 was verified as a target gene of miR‐10b‐3p. The overexpression of CMTM5 contributed to the suppression of the proliferative, migratory and invasive abilities of HCC cells. Moreover, the up‐regulation of miR‐10b‐3p and down‐regulation of CMTM5 were observed to be associated with worse overall survival. Lastly, we have confirmed the carcinogenesis‐related roles of miR‐10b‐3p and CMTM5 in vivo. We concluded that the up‐regulation of miR‐10b‐3p promoted the progression of HCC cells via targeting CMTM5.


| INTRODUCTION
As one of the most prevalent malignancies worldwide, hepatocellular carcinoma (HCC) is also the leading cause of cancer-associated mortality. 1 The intractable disease evolves more sophisticated due to the dysregulation of different genes by promoting the development and progression of HCC. 2 Although improvements have been made in surgery and other treatment methods, it remains at a low level in terms of the 5-year overall survival rate of patients with HCC. The lack of accurate and non-invasive diagnostic tools makes early diagnosis and remission of HCC difficult, resulting in poor prognosis. 3 Hence, it is critical to define the mechanisms of hepatocarcinogenesis at molecular level and to develop novel strategies for HCC early diagnosis and prognosis prediction.
MicroRNAs (miRNAs) play an important role in malignancy by targeting various tumour suppressors and oncogenes. They also take part in cancer stem cell biology, angiogenesis and epithelialmesenchymal transition, which could also influence carcinogenesis. 4 carcinoma, pancreatic cancer, malignant glioma and HCC. [5][6][7][8] Furthermore, overexpression of miR-10b-3p was observed in tumorous tissue specimens of HCC compared with non-tumour adjacent tissues. 9 The patients with higher miR-10b levels showed shorter survival time. 10 These findings suggest the possible oncogenic role of miR-10b-3p in HCC. However, the molecular mechanism remains largely unknown. Given the overexpression of miR-10b-3p in cancer tissues, further investigation into the role and molecular mechanisms of miR-10b-3p in cancer development and progression was needed.
CMTM5, also called CKLF-like MARVEL transmembrane domain containing member 5, is a member of CMTM family that could inhibit tumour growth. 11 Being broadly expressed in human tissues, CMTM5 is usually down-regulated in carcinoma tissues. 12,13 Restoration of CMTM5 may contribute to better morphological transformation, but the antitumour mechanism remains unclear. 14 It was proposed that the understanding of the antitumour mechanism of CMTM5 was critical before it became a new target in the gene therapies for tumours.
Previously, few studies have been performed to investigate the correlation between miRNAs and CMTMs. Our findings indicated that miR-10b-3p directly targeted CMTM5 and negatively regulated its expression, which therefore inhibited the proliferation, migration and invasion of HCC cells. These results provided new insights into the mechanisms by which miR-10b-3p modulated the development of HCC by interacting with CMTM5.
Tissue specimens of HCC and adjacent non-tumour tissues (n = 30) were obtained from China-Japan Union Hospital of Jilin University.
Written informed consent was accessed from all patients (n = 350).
This study was also approved by the ethics committee of China-Japan Union Hospital of Jilin University.

| Cell transfection
Cell transfection was conducted using Lipofectamine 3000 reagent.
HepG2 cells were cultured under normal conditions and were inoculated uniformly to 96-well plates at appropriate concentrations (approximately 3 9 10 5 cells/mL in this case). After adherent cell culture, cell transfection was conducted to miR-10b-3p mimics, inhibitors and CMTM5 overexpression plasmids. The cells of the normal group were treated with Lipofectamine only. The mimics and inhibitor etc. were diluted in MEM medium without serum, and then the Lipofectamine 3000 reagent was added to the medium. After incubation for 5 minutes, the diluted Lipofectamine 3000 was mixed with the mixture of last step, which was added to the culture plate of HepG2 cells and incubated at 37°C for 5 hours. Thereafter, the mixture was incubated in MEM medium with 10% FBS for another 48 hours.

| Microarray analysis
The microarray data obtained from TCGA Database (https://cance rgenome.nih.gov/) was used to screen out the differentially expressed genes in HCC. The data included 18 samples (9 tumour tissues and 9 adjacent non-tumorous tissues). Affy and limma packages (R packages) were used to read the expression measures and screen out differently expressed genes that were with fold change value greater than 2 or smaller than -2 (P < .05).

| Luciferase reporter assay
A bunch of HepG2 cells were seeded onto 24-well plates to grow in 5% CO 2 at 37°C till 70% confluence. Then cells were cotransfected with miR-10b-3p mimics or control mimics with wild-type or mutated-type 3 0 UTR of CMTM5 using Lipofectamine 3000 transfection reagent (Life Technologies, Gaithersburg, MD, USA). Thirty-six hours after transfection, cells were washed with PBS. Luciferase activity was then determined using the Dual-Luciferase Reporter Assay System (Promega) and a microplate fluorescence reader (Bio- Tek, Winooski, Vermont, USA).

| Colony formation assay
Transfected cells were lysed and then seeded onto 6-well plates.
When colonies were visible after approximately 2 weeks, they were washed with PBS and fixed with 4% paraformaldehyde for 20 minutes before they were stained with GIMSA for 30 minutes. Finally, cells were air-dried. Colonies were counted under an Olympus CK2 phase-contrast inverted microscope (Olympus, Tokyo, Japan). We randomly chose 5 fields for observation and repeated the experiment for three times.

| Cell invasion assay
Cells were initially starved for 24 hours before the suspension. Then they were lysed and washed with PBS before being resuspended with serum-free media. 1 9 10 5 cells were plated in the upper well, and 600 lL of DMEM medium containing 10% FBS was served as the chemoattractant in the lower chambers. After 24 hours, invading cells were fixed with 4% paraformaldehyde, stained with 0.1% crystal violet, air-dried and photographed using a light microscope. The number was also counted under the microscope as well.

| In vivo experiments
We purchased 16 healthy and specific pathogen-free BALB/C mice

| Survival analysis
The survival outcomes of 320 patients were analysed. High level was defined as a fold change value bigger than 2, whereas a low level was defined as a fold change value smaller than -2. A total of 320 patients were included in the following study (details of the survival analysis have been given in the main text), and among them, 228 were with high miR-10b-3p levels, and 92 were with low miR-10b-3p levels; and 218 were with low CMTM5 levels, and 102 were with high CMTM5 levels.

| Statistical analysis
The statistical analysis and graphical depiction were performed with whereas lower level of miR-10b-3p was seen in inhibitor group.
The modulation of CMTM5 expression did not affect the expression of miR-10b-3p ( Figure 2E). On the other hand, the expression of CMTM5 was higher in CMTM5 group and inhibitor group, whereas lower in mimics group. The cotransfection of CMTM5 and mimics did not affect the expression of CMTM5 ( Figure 2F). Western blot results showed that miR-10b-3p could suppress CMTM5 ( Figure 2G, P < .01). Therefore, we concluded that miR-10b-3p was able to suppress CMTM5 in HCC cell line HepG2.

| miR-10b-3p promoted HepG2 cell proliferation, migration and invasion by suppressing CMTM5
Enforced expression of CMTM5 dramatically reduced cell proliferation (in both MTT assay and colony formation assay) in HepG2 cells, whereas enforced expression of miR-10b-3p significantly promoted cell proliferation. The cotransfection of mimics and CMTM5 overexpression plasmids led to no significant change in cell proliferation.
These results suggested that CMTM5 could impair the proliferation, migration and invasion of HepG2 cells.

| MiR-10b-3p and CMTM5 predicted overall survival in patients with HCC
A total of 320 patients were included in the follow-up study. The overall survival rates of the patients were shown in Figure 4.
Patients with low expression of miR-10b-3p were accompanied with better overall survival ( Figure 4A, P = .0006), whereas patients with low expression of CMTM5 were accompanied with worse overall survival ( Figure 4B, P = .0015). The results revealed that miR-10b-3p had correlation with poor prognosis, whereas CMTM5 was connected with good prognosis for patients with HCC.
3.5 | miR-10b-3p promoted HCC tumour growth by suppressing CMTM5 In vivo Twenty-five days after the tumour harvest, we found bigger tumour sizes in mimics group and smaller tumour sizes in CMTM5 group ( Figure 5A). In vivo experiment results suggested that the tumour growth was significantly suppressed in CMTM5 overexpressed group but significantly promoted in mimics group ( Figure 5B). The cotransfection of mimics and CMTM5 overexpression plasmids caused no change in tumour growth. The results suggested that miR-10b-3p could promote tumour growth by suppressing CMTM5 In vivo.
F I G U R E 2 CMTM5 was a direct target of miR-10b-3p in HCC cells. A, The binding sequences of miR-10b-3p and CMTM5 (at position 239-245) were illustrated. B, Dual-luciferase reporter gene assay was conducted to verify that miR-10b-3p significantly masked the wild-type 3 0 UTR but not the mutated-type 3 0 UTR of CMTM5. **P < .01, compared with the mimics control group. C, The transfection efficiency of miR-10b-3p mimics and inhibitor in HepG2 cells was confirmed. D, The transfection efficiency of CMTM5 overexpression plasmids in HepG2 cells was confirmed. E, qRT-PCR results revealed that mimics group and CMTM5 + miR group showed significantly up-regulated miR-10b-3p in HepG2 cells. U6 was used as the internal control. F, qRT-PCR results revealed that cells of CMTM5 group showed significantly upregulated CMTM5 expression, whereas those of mimics group showed downregulated CMTM5 expression in HepG2 cells. GAPDH was used as the internal control. G, Western blot analysis results confirmed that CMTM5 was downregulated by miR-10b-3p mimics in HepG2 cells. *P < .05, **P < .01, compared with the control group. Mimics: miR-10b-3p mimics; CMTM5: CMTM5 overexpression; CMTM5 + miR: CMTM5 overexpression plus miR-10b-3p mimics Human algorithm suggest a potential target relationship between the two. An earlier study revealed that miR-10b might target RhoC, uPAR, MMP-2 and MMP-9. 15 Other researches implied CSMD1 and CADM1 could also be regulated by miR-10b. 5,18 Most of these genes acted as tumour suppressors, so we assumed that miR-10b could suppress certain anticancer factors including CMTM5. In our study, we proved that CMTM5 was a target gene of miR-10b-3p in HepG2 cell line.
The ectopic gene expression could influence cell activities. After we modified the expression of miR-10b-3p and CMTM5, we found the altered cell proliferation, invasion and migration in single-transfected group, whereas no obvious alteration on cell activities was found.
As we have found that the modification of miR-10b-3p and CMTM5 expressions changed cell activities, we then supposed that they may also be associated with prognosis. Some studies demonstrated that miR-10b-3p level was significantly higher in HCC patients with worse overall survival outcome, indicating higher miR-10b-3p expression level was an independent predictor of poor prognosis, 10 Overexpression of CMTM5 and CMTM3, which are two members of CMTM family, inhibited tumour progression both in vitro and In vivo. 11,20 The two studies indicated that miR-10b and CMTM5 could also be useful biomarkers for HCC metastasis, which can be further investigated in the following studies.
In conclusion, our study indicated that up-regulation of miR-10b-3p could promote the progression of HCC cells by suppressing CMTM5 expression. Our findings may cast new light on the novel therapy targets of HCC.

CONFLI CT OF INTEREST
The authors confirm that there are no conflict of interests. F I G U R E 5 MiR-10b-3p promoted HCC tumour growth by suppressing CMTM5 in vivo. A, Twenty-five days after the tumour harvest, bigger tumours in mimics group and smaller tumours were seen in CMTM5 group. The cotransfection of mimics and CMTM5 overexpression plasmids did not affect the tumour size. B, The tumour growth was significantly suppressed in CMTM5 overexpressed group but significantly promoted in mimics group. The cotransfection of mimics and CMTM5 overexpression plasmids did not cause the change in tumour growth rate. *P < .05, compared with the control group. Mimics: miR-10b-3p mimics; CMTM5: CMTM5 overexpression; CMTM5 + miR: CMTM5 overexpression plus miR-10b-3p mimics O R C I D Baozhen Sun http://orcid.org/0000-0001-8878-3924