A splicing variant of Merlin promotes metastasis in hepatocellular carcinoma

Luo, Zai-Li; Cheng, Shu-Qun; Shi, Jie; Zhang, Hui-Lu; Zhang, Cun-Zhen; Chen, Hai-Yang; Qiu, Bi-Jun; Tang, Liang; Hu, Cong-Li; Wang, Hong-Yang; Li, Zhong

doi:10.1038/ncomms9457

Download PDF

Article
Open access
Published: 07 October 2015

A splicing variant of Merlin promotes metastasis in hepatocellular carcinoma

Zai-Li Luo^1,2^na1,
Shu-Qun Cheng³^na1,
Jie Shi³^na1,
Hui-Lu Zhang¹,
Cun-Zhen Zhang³,
Hai-Yang Chen¹,
Bi-Jun Qiu⁵,
Liang Tang¹,
Cong-Li Hu¹,
Hong-Yang Wang^1,5 &
…
Zhong Li^4,6

Nature Communications volume 6, Article number: 8457 (2015) Cite this article

4727 Accesses
27 Citations
1 Altmetric
Metrics details

Subjects

Abstract

Merlin, which is encoded by the tumour suppressor gene Nf2, plays a crucial role in tumorigenesis and metastasis. However, little is known about the functional importance of Merlin splicing forms. In this study, we show that Merlin is present at low levels in human hepatocellular carcinoma (HCC), particularly in metastatic tumours, where it is associated with a poor prognosis. Surprisingly, a splicing variant of Merlin that lacks exons 2, 3 and 4 (^Δ2–4Merlin) is amplified in HCC and portal vein tumour thrombus (PVTT) specimens and in the CSQT2 cell line derived from PVTT. Our studies show that ^Δ2–4Merlin interferes with the capacity of wild-type Merlin to bind β-catenin and ERM, and it is expressed in the cytoplasm rather than at the cell surface. Furthermore, ^Δ2–4Merlin overexpression increases the expression levels of β-catenin and stemness-related genes, induces the epithelium–mesenchymal-transition phenotype promoting cell migration in vitro and the formation of lung metastasis in vivo. Our results indicate that the ^Δ2–4Merlin variant disrupts the normal function of Merlin and promotes tumour metastasis.

PPM1G promotes the progression of hepatocellular carcinoma via phosphorylation regulation of alternative splicing protein SRSF3

Article Open access 21 July 2021

The aberrant upregulation of exon 10-inclusive SREK1 through SRSF10 acts as an oncogenic driver in human hepatocellular carcinoma

Article Open access 16 March 2022

PRPF8 increases the aggressiveness of hepatocellular carcinoma by regulating FAK/AKT pathway via fibronectin 1 splicing

Article Open access 06 January 2023

Introduction

Hepatocellular carcinoma (HCC) is the fifth leading cause of death in patients with cancer. The development of portal vein tumour thrombus (PVTT) in HCC is rarely curable and associated with an extremely poor prognosis¹. Approximately 50–80% of HCCs are accompanied by portal or hepatic vein invasion². Cancer stem cells/tumour-initiating cells are associated with the formation of HCC, PVTT and tumour relapse due to their capacity for self-renewal and differentiation³.

Merlin (moesin-ezrin-radixin-like protein), encoded by the neurofibromatosis type 2 (Nf2) tumour suppressor gene, is a member of the band 4.1 family. As one of the most versatile tumour suppressors, it is capable of integrating several different mechanisms that regulate cell proliferation, motility, survival and signalling pathways⁴. Although mutations in Nf2 gene are primarily associated with schwannomas and meningiomas⁵, it is not reportedly mutated in certain tumours such as those associated with breast, liver and colorectal cancer^6,7. Wild-type Merlin (^wtMerlin) is thought to be involved in cell-specific mechanisms of growth control⁸. In contrast to the narrow spectra of benign schwannoma and meningioma tumours in neurofibromatosis patients, Nf2 heterozygous mice develop a variety of malignant tumours with high rates of metastasis⁹, which are dependent on the Merlin-mediated stabilization of adherens junctions¹⁰, the control of cadherin-mediated cell:cell contact¹¹, and the activation of Rac¹². The specific knockout of Merlin in mouse hepatocytes leads to both cholangiocellular and hepatocellular carcinoma¹³, which suggests that Merlin plays an important role in liver tumorigenesis. One of the mechanisms for the Merlin-mediated inhibition of tumour growth may be the control of organ size through the activation of YAP in mammals^14,15. However, whether Merlin is involved in human HCC metastases remains largely unclear.

In this study, we first observe the expression of Merlin in HCC samples and find that the expression of ^wtMerlin is relatively lower in tumour tissues than in adjacent non-tumour tissues. Next, we find that several spliced forms of Merlin are increased in HCC. Specifically, a splicing variant of Merlin lacking the sequences encoded by exons 2, 3, and 4 is identified and designated ^Δ2–4Merlin. Compared with ^wtMerlin, ^Δ2–4Merlin losses its ability to interact with β-catenin and Ezrin/Radixin/Moesin (ERM) proteins through its N-terminal binding domain. The knockdown of Merlin or the overexpression of ^Δ2–4Merlin promotes cell migration and invasion via an increase in Twist1. Furthermore, ^Δ2–4Merlin also enhances the activation of β-catenin and stemness-related genes. Pulmonary metastatic mouse model shows that ^wtMerlin reduces HCC cell-induced lung metastasis, while ^Δ2–4Merlin promotes distant metastasis. Altogether, our results reveal that ^Δ2–4Merlin functions as a tumour promoter. In particular, we discover that the new variant of Merlin promotes liver cancer metastasis by interfering with the tumour suppression role of ^wtMerlin.

Results

Merlin expression in HCC is associated with patients′ survival

To understand the role of Merlin in HCC, we detected the expression of Merlin with tissue-chips containing 148 samples of HCC with PVTT, 37 samples of HCC without PVTT, 29 samples of tumour-adjacent tissues and 16 samples of PVTT. We obtained and fixed 5-μm thick × φ3-mm sections on slides and stained them with an anti-Merlin antibody designed to target the N terminus of the Merlin protein. The expression of Merlin in these samples was quantified based on the percentage of positive cells and the density of staining by three people using 12 standard points, as previously reported. The overall results show that Merlin levels were expressed in the following order: tumour-adjacent tissues>tumours>PVTT (Fig. 1a,b). A western blot analysis of the tumour (T), non-tumour adjacent (N) and PVTT (P) tissues showed similar results when equivalent protein quantities were used (Fig. 1c). Subsequently, we analysed the relationships between Merlin and certain clinical and pathological characteristics. On the basis of the expression of Merlin in HCC and adjacent non-tumour tissues, the scores for Merlin in each sample were divided into high- and low-expression groups (mid index=5). Using analysis of variance (ANOVA), we found that the expression of Merlin was negatively associated with metastasis and the development of PVTT but was not significantly statistically related to sex, age, tumour size, HBV infection or tumor node metastasis (TNM) staging (Table 1). Kaplan–Meier estimates within 5 years of follow-up revealed that in 148 HCC patients a low expression of Merlin had shorter disease-free durations (P<0.001) and lower overall survival rates (P=0.013; Fig. 1d).

**Figure 1: Low expression of Merlin in HCC and PVTT is associated with a cacoethic prognosis.**

Table 1 Relationship of Merlin protein expression with clinical pathologic characteristics.

Full size table

Merlin limits cell migration and invasion

Tumour metastasis is known to be the major cause of death in cancer. Because the expression of Merlin was negatively associated with the development of PVTT and metastasis (Fig. 1, Table 1), we hypothesized that Merlin might be involved in tumour metastasis. Three shRNAs for Merlin were designed, two of which strongly interfered with the expression of Merlin (Fig. 2a). Thus, the #2 shRNA sequence was inserted into a lentivirus vector that was used to establish stable HCCLM3, Huh-7 and normal-like hepatocyte L02 cells expressing shMerlin (Fig. 2b). An analysis of cell migration showed that silencing Merlin in the L02 and HCCLM3 cells increased the number of migrated cells (Fig. 2c) and facilitated the invasion of HCCLM3 and Huh-7 cells (Fig. 2d). In a mouse pulmonary metastatic model, 1 × 10⁶ HCCLM3 cells expressing either shMerlin or scrambled RNA were injected into the caudal veins of three BALB/C nude male mice. The lungs of the mice were analysed for tumour nodules after 3 months. The HCCLM3 cells with downregulated Merlin exhibited increased numbers and volumes of lung metastatic tumours compared with control cells (Fig. 2e). These results suggest that Merlin plays a negative role in the regulation of cell motility and migration.

**Figure 2: Interfering with the expression of Merlin promotes metastasis in HCC cell lines.**

ERM is responsible for cytoskeleton reorganization and migration, which is regulated by the Rho family. Merlin decreases the activation of Rac1 (ref. 16). To detect whether knockdown of Merlin would regulate cell locomotion via the Rho family, we detected the activation of Rho and Cdc42 by using the specific binding domain GST-RBD for active Rho and GST-CBD for active Cdc42 in HCCLM3 cells. Although the knockdown of Merlin did not affect the expression of Rho and Cdc42 (Fig. 2f, left panels), it increased the activation of Rho but not Cdc42 (Fig. 2f, right panels). Therefore, downregulation of Merlin promotes migration possibly through Rho activation.

Splicing variants of Merlin are amplified in HCC and PVTT

Observing the levels of Nf2 mRNA in HCC, adjacent non-tumour and PVTT specimens by quantitative PCR (qPCR), we surprisingly found that there were no significant differences in expression (Fig. 3a). Considering that Merlin protein levels in HCC tissues were lower than those in adjacent non-tumour tissues, we speculated that the transcription of Merlin might be interrupted in HCC. On the basis of previous studiesshowing frequent splicingevents at the N terminus, a pair of primers was designed encompassing exons (1 to 5) of the Nf2 open reading frame (sense: 5′- CAAGACGTTCACCGTGAGGAT -3′, antisense: 5′- GATTGCAAAGTAGTTCACACCG -3′). Reverse transcription PCR (RT–PCR) results revealed that, in addition to wild-type Nf2, there were several additional fragments present in normal liver tissues, HCC cell lines (HCCLM3, Huh-7) and PVTT-derived cells (CSQT2) (Fig. 3b,c), including at least three bands in the HCC tissues and cells (Fig. 3b,c). After sequencing these bands in Fig. 3c, we found that band 2 had a deleted exon 3, and band 3 lacked exons 2, 3 and 4 (Supplementary Fig. 1a,b). Interestingly, the Merlin variant lacking exons 2,3 and 4 appeared to be increased in CSQT2 cells (Fig. 3c). Western blots confirmed that this variant was sole band in CSQT2 cells (Fig. 3d), suggesting that it may be involved in the PVTT formation. So we selected it for further study and designated it ^Δ2–4Merlin.

**Figure 3: The alternative splicing of Merlin in HCC is determined by PCR and western blotting.**

The difference between type I and type II Merlin is attributable to a difference in their C-terminus sequences wherein type I Merlin ends with exons 15 and 17, while type II Merlin ends with exons 16 and 17 (Supplementary Fig. 1b). To determine the functions of the Merlin variants in HCC, we generated constructs for full-length Nf2 for type I and type II Merlin, as well as ^Δ2–4Merlin from type I and type II Merlin. We found that the Merlin antibody A19 could recognize both type I and II ^wtMerlin and type II ^Δ2–4Merlin, whereas the Merlin antibody C18 could recognize only Type I ^wtMerlin and Type I ^Δ2–4Merlin (Fig. 3e). Subsequently, we found that both Type I and Type II ^wtMerlin were expressed more highly in adjacent tumour tissues than in HCC tissues (Supplementary Fig. 1c). In the C19 antibody used samples in Fig. 1c, we detected with the C18 antibody and found that type I ^Δ2–4Merlin was increased in HCC(T) and PVTT(P), and type I ^wtMerlin was decreased in these HCC tissues (Fig. 3f). To determine the presence of ^Δ2–4Merlin, we generated a polyclonal antibody specific to the combined regions of exons 1 and 5. This novel ^Δ2–4Merlin antibody (Merlin N15) solely targeted ^Δ2–4Merlin (Fig. 3g) and identified ^Δ2–4Merlin expression in HCC and PVTT tissues (Fig. 3h). In comparison, as shown in Fig. 1c and Fig. 3f, the expression of type I ^Δ2–4Merlin was higher than that of type II ^Δ2–4Merlin in HCC. Because type II ^wtMerlin does not function as a tumour suppressor^17,18, we decided to explore the mechanisms of type I ^Δ2–4Merlin and ^wtMerlin in our study of liver tumourigenesis and metastasis.

^Δ2–4Merlin promotes metastatic activity and EMT

To study the role of ^Δ2–4Merlin in tumorigenesis, we generated expression vectors for ^wtMerlin and ^Δ2–4Merlin and expressed them in HCCLM3 cells. A migration assay showed that ^wtMerlin inhibited cell migration and invasion, whereas ^Δ2–4Merlin activated cell migration and invasion (Fig. 4a).

**Figure 4: ^Δ2–4Merlin expression promotes cell migration and induces EMT.**

The epithelium–mesenchymal-transition (EMT) signifies that cancer cells are regaining their capacity to migrate and invade near and remote tissues. We found that Snail and Twist1 were increased in HCC Huh-7 (Fig. 4b left) and LM3 cells (Fig. 4b right) with shMerlin expression. Additional EMT-related genes, such as ZO-1, Vimentin, E-cadherin and β-catenin, were unaffected. In addition, the knockdown of Merlin in CSQT2 cells decreased the expression of β-catenin, Snail, ZO-1 and Twist1 (Fig. 4c). Furthermore, the expression of ^wtMerlin in HCCLM3 cells decreased the expression of Snail, Twist1 and Vimentin, and slightly increased the expression of ZO-1 (Fig. 4d). In contrast, ^Δ2–4Merlin reversed the effects of ^wtMerlin (Fig. 4d). To confirm changes in these EMT-related genes with respect to protein levels, Western blotting was performed to analyse the expression of the EMT genes and β-catenin after stimulation with TGF-β, which is known to induce EMT. Silencing Merlin resulted in increases in Snail and Twist1 expression compared with scrambled RNA-expressing control cells (Fig. 4e). ^wtMerlin suppressed Snail and Twist1 (Fig. 4f). Meanwhile, ^Δ2–4Merlin blocked the endogenous Merlin-induced decrease in these protein levels (Fig. 4f). Vimentin seems not to be affected in the conditions of both overexpression and knockdown of Merlin (Fig. 4e,f). Taken together, these results suggest that Merlin is involved in tumour metastasis, and the loss of Merlin and/or increase of the ^Δ2–4Merlin variant promotes tumour metastasis.

It is known that Merlin is involved in Wnt signalling through its interaction with β-catenin^10,19. In stimulation with TGFβ, HCCLM3 cells showed similar patterns in, an increased expression of β-catenin (Fig. 4e). However, the stable overexpression of ^wtMerlin reduced β-catenin expression, whereas the stable overexpression of ^Δ2–4Merlin increased β-catenin expression (Fig. 4f). These results suggest that ^Δ2–4Merlin may interfere ^wtMerlin-mediated inhibition of β-catenin. In a rescue assay, when ^wtMerlin, ^Δ2–4Merlin and ^wtMerlin/^Δ2–4Merlin were transiently transfected into HCCLM3 cells, ^wtMerlin reduced the expression of β-catenin and increased N-cadherin expression, whereas ^Δ2–4Merlin evidently increased β-catenin expression and reduced N-cadherin expression (Fig. 4g). The transfection of ^Δ2–4Merlin into cells expressing ^wtMerlin blocked the ^wtMerlin-induced reduction of β-catenin, although ^Δ2–4Merlin expression appeared to be lower when it was jointly expressed with ^wtMerlin compared with when it was expressed alone (Fig. 4g).

Merlin regulates stemness feautures of cancer cells

Merlin has been shown to control progenitor homeostasis and tumorigenesis in the liver¹³. Accordingly, we observed spheroid formation in cells in which Merlin was knocked down. Huh-7 cells with shMerlin expression demonstrated increased spheroid numbers, and the number of spheroids with diameters greater than 100 μm was also increased in these cells (Fig. 5a), as well as the expression of stemness-related genes (Fig. 5b). As expected, the knockdown of Merlin in CSQT2 cells, which solely express ^Δ2–4Merlin, decreased the number of spheroids and the expression of stemness genes compared with control cells (Fig. 5c,d). In contrast to the knockdown of Merlin, overexpression of either ^Δ2–4Merlin type I (^Δ2–4Mer I) or ^Δ2–4Merlin type II (^Δ2–4Mer II) in HCCLM3 cells resulted in an increase in spheroid formation (Fig. 5e) and the expression of stemness-related genes compared with control cells and ^wtMerlin expressing cells (Fig. 5f). A side population (SP) analysis of the HCCLM3 cells (0.41%) showed that silencing Merlin reduced the positive rate of stem cells to 0.26%, whereas ^Δ2–4Merlin increased the rate to 0.77%(Fig. 5g). At the same time, the knockdown of Merlin in CSQT2 cells decreased the SP rate (Fig. 5h). Thus, our data suggest that ^Δ2–4Merlin can promote the property of cancer stemness.

**Figure 5: The silencing of Merlin or the expression of ^Δ2–4Merlin promotes stemness activity.**

^Δ2–4Merlin no longer interacts with β-catenin and ERM

Merlin is associated with plasma membrane location, but ^Δ2–4Merlin has lost its ability to anchor to the plasma membrane (Fig. 6a), suggesting that the domain encoded by exons 2–4 is responsible for membrane association. Merlin has been reported to bind to β-catenin and ERM. An immunoprecipitation assay revealed that ^wtMerlin was associated with β-catenin on the plasma membrane, but the knockdown of Merlin caused the accumulation of β-catenin, primarily in the cytosol and nucleus (Fig. 6b). Immunoprecipitation analysis subsequently confirmed that ^wtMerlin in HCCLM3 cells was bound to β-catenin and ERM but that ^Δ2–4Merlin had lost this binding capacity (Fig. 6c). Furthermore, an immunofluorescence assay revealed that ^wtMerlin was generally found to be bound with β-catenin on the plasma membrane, whereas ^Δ2–4Merlin primarily accumulated in the cytosol surrounding the nuclei (Fig. 6d). ^Δ2–4Merlin was increased in the cytosol possibly due to its dissociation with β-catenin and the plasma membrane. Thus, β-catenin was increased in the nuclei (Fig. 6d). A subcellular fractionation assay suggested that ^Δ2–4Merlin promoted the translocation of β-catenin into the nucleus and that ^wtMerlin limited this process (Fig. 6e). Taken together, these results suggest that the ^Δ2–4Merlin variant disrupts the association of ^wtMerlin with β-catenin or ERM, which in turn promotes the translocation of β-catenin into the nuclei.

**Figure 6: ^Δ2–4Merlin loses its functions of anchoring the membrane and binding to ERM and β-catenin.**

To further study the role of ^Δ2–4Merlin in tumour metastasis, we injected HCCLM3 cells expressing either ^wtMerlin or ^Δ2–4Merlin into the caudal veins of five BALB/C nude mice and observed their tumour nodules in the lungs. ^wtMerlin overexpression reduced the size and number of lung nodules compared with control cells (Fig. 6f). However, the overexpression of ^Δ2–4Merlin increased the formation of lung metastasis with respect to both the size and number of tumour nodules (Fig. 6f). Therefore, we conclude that ^Δ2–4Merlin, an increased splicing variant of Nf2, plays a critical role in the liver tumorigenesis and metastasis.

Discussion

Nf2 heterozygous or homozygous mice have multiple malignant tumours. Merlin expression is markedly reduced in human malignant gliomas²⁰. The loss of Nf2 in the livers of mice leads to cholangiocellular and hepatocellular carcinoma¹³. However, less is known about the role of Merlin in the development and progression of human liver cancer. Our data not only track the levels of Merlin in HCC but also provide new evidence that an Nf2 splicing variant in liver cancer is responsible for tumour proliferation and metastasis.

Merlin contains several functional domains. Its nuclear export sequence is encoded by exon 15, and a cytoplasmic retention domain is encoded by exon 2 (ref. 21). The deletion of exon 2 leads to the unrestricted entry of Merlin into the nucleus^21,22. Less abundant variants such as ^Δ2Merlin (missing exon 2), ^Δ3Merlin (missing exon 3) and ^Δ2/3Merlin (missing exons 2 and 3) demonstrate inactivity as tumour suppressors^23,24. ^Δ2Merlin, ^Δ3Merlin, ^Δ2/3Merlin, and even ^Δ15/16Merlin exist in normal pleura²⁵. The dominant effects of ^Δ2/3Merlin are functionally associated with Schwann cell tumorigenesis in the presence of ^wtMerlin²⁶. ^Δ2–4Merlin, which was first described in a malignant pleural mesothelioma cell line, is relatively prevalent (11.8%) and is principally expressed in malignant pleural mesothelioma^25,27. However, the function of the splicing variant is unclear. We first studied the function of ^Δ2–4Merlin in HCC and demonstrated that ^Δ2–4Merlin is increased in HCC and PVTT. The accumulation of abnormal variants of Merlin in HCC should be taken into account when screening targets for cancer treatment. Given that the loss of Merlin occurs in highly malignant and metastatic tumours such as mesothelioma, colorectal cancer, melanoma and neurofibroma^{24,25,28,29,30}, as well as HCC according to our data, we performed exome sequencing for Nf2 gene and found no mutation present in human HCC, PVTT and non-tumour adjacent tissues (Supplementary Table 1). Therefore, it is believed that an increase in Merlin splicing variants is new characteristic associated with the liver metastasis and recurrence.

Aberrant splicing leading to the inactivation of Merlin drew considerable attention two decades ago²⁹, and the loss of Merlin function can be achieved by truncating various locations in the protein³¹. However, most studies to date have focused on ^wtMerlin or knockouts of Nf2 in animal models; few reports have studied the functions of aberrant Merlin variants. Surprisingly, ^Δ2–4Merlin was found to be expressed in HCC, specifically in the PVTT-derived cell line, which inspired us to explore the role of ^Δ2–4Merlin in PVTT formation. Aberrant gene splicing can occur via exon skipping, intron retention and alternative 3′ or 5′ splice sites. The Merlin aberrant gene splicing in HCC is mainly attributable to exon skipping. Cancer-specific aberrant splicing of both oncogenes and tumour suppressor genes is associated with specific cancer types, and cancer shows a preference with respect to the selection of alternative splice-sites and the utilization of alternative splicing types³². Aberrant transcripts, often with truncated or missing domains, contribute to essential phenotypes that are associated with transformed cells³³ and are a potential source of new diagnostic, prognostic, predictive and therapeutic tools for human cancer.

Merlin has a close relationship with migration and motility and serves as a linker between transmembrane proteins and the actin-cytoskeleton. An analysis of primary Schwannoma samples derived from neurofibromatosis patients showed that the activity of Pak1 was highly elevated³⁴. The expression of a dominant-negative form of Merlin mimics Nf2 deficiency and interferes with the ability of wild-type cells to assemble adherens junctions, confirming the specificity of Merlin’s function in contact inhibition¹⁰. Nf2^−/− cells exhibit the characteristics of cells expressing activated Rac, which induces the phosphorylation of Merlin and decreases the association of Merlin with the cytoskeleton³⁵. Merlin also inhibits mixed lineage kinase 3 (MLK3) activity by blocking the Cdc42–MLK3 interaction³⁶. Merlin inhibits MLK3 activity by directly associating with MLK3 at its C-terminal residues from 340–590, preventing MLK3 from activating its downstream effectors B-Raf, extracellular signal-regulated kinase (ERK) and c-Jun N-terminal kinase (JNK), as well as its upstream activator, Cdc42 (ref. 37). In our study, the knockdown of Merlin increased RhoA activity, which is also responsible for cytoskeleton organization and polarization, but did not affect the expression and activation of Cdc42, which is similar to the results observed in HEK293 cells³⁷.

Active Merlin maintains the beta-catenin and N-cadherin complex at the plasma membrane via direct regulation¹⁹. Nf2 knockout mouse embryonic fibroblasts lose their contact inhibition of cell proliferation but exhibit significantly increased canonical Wnt signalling³⁸. Dominant-negative TCF4 or Rac1 mutants, as well as small-molecule inhibition of Wnt, are able to curb Nf2 deficiency-elicited confluent cell proliferation³⁸. In this paper, we found that a variant form of Merlin can interfere with ^wtMerlin-mediated functions and lead to an increase in β-catenin and promote its nuclear accumulation.

In summary, the expression of Merlin was found to be lower in HCC and PVTT versus non-tumour adjacent tissues, which is associated with remote metastasis and the formation of PVTT. In studying this mechanism, an increased number of splicing variants were found in HCC, in which ^Δ2–4Merlin, lacking the sequences encoded by exons 2, 3 and 4, was identified and demonstrated to promote migration and metastasis. Compared with ^wtMerlin, ^Δ2–4Merlin has lost its ability to bind to β-catenin and ERM and promotes migration and invasion via an increase in Twist1. ^Δ2–4Merlin also promotes stemness activity, potentially by increasing the expression of β-catenin and the nuclear accumulation and activities of other stemness genes (Sox2, OCT4, Klf4, C-myc and Nanog). Therefore, this Merlin alternative splicing variant (deletion of exons 2, 3 and 4) represents a novel mechanism for liver metastasis and PVTT formation.

Methods

Patients

Specimens including 185 primary HCC with or without PVTT from HCC patients who received curative surgery in the Eastern Hepatobiliary Surgery Hospital (Shanghai, China) from January 2000 to December 2003 were used for tissue-chips. 44 sets of HCC specimens (HCC, PVTT and non-tumour adjacent tissues) were used for qPCR or western blot analysis. A written informed consent was obtained from each patient and these studies were approved by the Ethics Boards of the Eastern Hepatobiliary Surgery Hospital.

Cell cultures and stable cell line construction

The HCC cell lines HCCLM3, HepG2 and Huh7 as well as normal-type hepatocyte L02 cells were purchased from the Shanghai Cell Bank of the Chinese Academy of Sciences (Shanghai, China). The PVTT cell line CSQT2 was a gift from Prof. Shuqun Cheng, who established it in 2010 (ref. 39). All cells were routinely cultured in DMEM (Invitrogen) supplemented with 10% fetal bovine serum (FBS; Life Technologies) in an atmosphere of humidified air containing 5% CO₂ at 37 °C. After seeding and incubation for 24 h, the cells were transfected with their respective plasmids using jet-PEI (Polyplus, New York, NY, USA) according to the manufacturer’s protocol. Stable cell lines were constructed via the overexpression of plasmids or lentivirus vectors and selected with puromycin. The ^wtMerlin and ^Δ2–4Merlin expression plasmids were constructed by inserting fragments of the human type I Merlin and type II Merlin opening reading frames into the pCDNA^TM3.1/myc-His(-)A-vector using EcorRI/BamHI. For the disruption of Merlin, lentivirus-GV248-shMerlin (5′- ACTTCAAAGATACTGACAT -3′) and control scrambled viruses were prepared by Genechem company (Shanghai, China).

RT–PCR and real-time PCR analysis

Total RNA was extracted from tumour and non-tumour adjacent tissues and HCC cell lines with TRIzol (Invitrogen, Carlsbad, CA, USA). The reverse transcription of 2 μg of total RNA was performed using random primers (Roche Diagnostics, Penzberg, Germany) and SuperScript II reverse transcriptase (Invitrogen) according to the manufacturer’s protocol. Because Merlin splicing frequently occurs at N terminus, two primers, sense: 5′- CAAGAGGAAGCAACCCAAG -3′ and antisense: 5′- GAAGCCAGGAGCACAGAAG -3′, were designed to lie outside the Merlin exon 1 and exon 5 reading frame for the amplification of Merlin transcripts in tumours and cells. Real-time PCR was performed to measure the expression of the transcripts of stemness or EMT-related genes with primers (Supplementary Table 2) in reaction solutions with iTaqSYBR Green Supermix (Bio-Rad Laboratories, Hercules, CA, USA).

Immunoprecipitation and immunoblotting assays

Whole cell extracts and HCC tumour specimens were prepared in lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl , 10% glycerol, 0.2% Nonidet P-40, 1 mM EDTA, 1 mM ethylene glycol-bis[β-aminoethyl ether]- N,N,N′,N′-tetraacetic acid, 1 mM phenylmethylsulfonyl fluoride, 10 mM NaF, 5 mg ml⁻¹ aprotinin, 20 mM leupeptin and 1 mM sodium orthovanadate) and centrifuged at 12,000 × g for 15 min. The supernatants were incubated with 2 μg of anti-Myc tagged or normal mouse immunoglobulin G (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) for 8 h at 4 °C, followed by the addition of Protein A/G Plus-Agarose (Santa Cruz Biotechnology Inc.) for 2 h. Total and binding proteins were detected by western immunoblots. Immunoblots were performed using specific primary antibodies, following with a fluorescein-conjugated secondary antibody and then detected using an Odyssey fluorescence scanner (Li-Cor, Lincoln, NE, USA). The following primary antibodies were used: Merlin/NF2 (1:100, Santa Cruz Biotechnology, A19 sc331 rabbit and C18 sc332 rabbit), Twist1 (1:100, Santa Cruz Biotechnology, sc170453 mouse), β-actin (1:200, Santa Cruz Biotechnology, c-2 mouse), Myc-tag (1:1,000, CST, 71D10 mouse); β-catenin (1:1000, CST, D10A8 rabbit), Vimentin (1:1,000, Abcam, ab8978 mouse), ERM (1:1,000, Epitomics, 2054 rabbit) and Snail (1:1,000, Epitomics, L70G2 rabbit). Uncropped scans of immunoblots are provided in Supplementary Fig. 2.

Immunofluorescence and immunohistochemistry

For immunofluorescence (IF) assay, cells were seeded into wells of glass-bottomed dishes and incubated for 24 h. Next, the cells were fixed with 4% paraformaldehyde (PFA) for 30 min after being rinsed twice with ice-cold PBS. The following primary antibodies were used: anti-Myc-tag (1:100, Cell Signaling Technology, 71D10 mouse) and β-catenin (1:100, CST, D10A8 rabbit). The secondary antibodies Alexa Fluor 488 (1:300, Invitrogen, Z25002 mouse or Z25302 rabbit) and Alexa Fluor 555 (1:300, Invitrogen Z25005 mouse or z25305 rabbit) were used. A Leica confocal microscope was used to capture the images.

For the immunohistochemistry (IHC) assays on tissue microarray (TMA), the collected paraffin embedded specimens including 148 HCC with PVTT, 37 HCC without PVTT and 29 tumour-adjacent tissues specimens were used to prepare the tissue-chips by Shanghai Outdo Biotech Co., Ltd. Tissue slides containing sections (5-μm thick × φ3-mm). Brifly, the tissue sections were dewaxed, rehydrated and then immersed in methanol containing 0.3% hydrogen peroxide (Sinopharm Chemical Reagent Co., Ltd., 10011218) for 30 min to block endogenous peroxidase activity. Subsequently, the sections were heated in a pressure cooker filled with 10 mM ethylenediaminetetraacetic (EDTA Sigma, 34550) buffer (pH 8.0) for 2 min. After cooling, the sections were incubated in 1% blocking serum (Chemicon, 20773) for 30 min to reduce nonspecific binding⁴⁰. A primary antibody against Merlin/NF2 (1:100, Santa Cruz Biotechnology, sc331 rabbit) was used. The nuclei were labeled with DAPI. The secondary antibodies were HRP-conjugated goat anti-rabbit immunoglobulins (1:500, Santa Cruz, ).

Fractionation of cells

To separate the cytosolic and nuclear proteins, the cells were collected in PBS and incubated for 20 min at 4 °C in S100/P100 buffer (20 mM Tris, 150 mM NaCl, 2.5 mM EDTA, 1 mM EGTA, 1 mM benzamidine, 4 μM leupeptin, 0.5 mM PMSF, 1 μM microcystin and 1 mM DTT at pH 7.5). After homogenization using a 26-gauge needle (Becton Dickinson), the nuclear pellets were collected by centrifugation for 2 min at 1,000 × g at 4 °C. The supernatants were collected as cytoplasmic and membrane fractions followed by centrifuging at 100,000 × g for 60 min at 4 °C. Equivalent nuclear, cytoplasmic and membrane fractions were then analysed via SDS–PAGE and immunoblotting.

Spheroid assay

Hepatoma cells were plated at 4 × 10³ per ml in six-well ultralow attachment culture plates with DMEM/F12 (Invitrogen, #11320-033) containing B27 serum-free complement (Invitrogen, #17504-044), insulin (Sigma, solution, bovine, #I0516), bFGF (Invitrogen, #13256-029), and EGF (Sigma, #E4269) and were cultured for 10 days⁴¹. The quantities and sizes of the spheroids were measured in three independent experiments.

Migration and invasion assays

Cell migration assay was performed with a transwell chamber (Corning Incorporated Costar, cat #3422,). Briefly, 1 × 10⁵ cells were seeded onto the upper surface of an 8.0-μm pore size transwell chamber for 24 h. The cell invasion assay was performed with an invasion chamber (BD Biosciences, cat #354480). The cells (1 × 10⁶) were loaded onto the upper chamber and the invasive cells which had invaded the lower surface of the membrane from the extracellular matrix layer after 48 h were fixed with methanol and stained with 0.1% crystal violet dye. Photographs of five randomly selected fields of the fixed cells were captured. The numbers of migrated cells were counted and expressed as mean±s.d. The experiments were repeated independently three times.

Side population cells analysis using flow cytometry (FCM)

The cultured LM3 and CSQT2 cells at 70–80% confluence were digested with 0.25% Trypsin–EDTA and then suspended in 3% FBS/PBS. Volume of 1 × 10⁶ cells per ml were incubated with 20 μg ml⁻¹ Hoechst33342 (Invitrogen, USA) either alone or with 25 μg ml⁻¹ Verapamil (Sigma-Aldrich, USA) at 37 °C for 90 min. Cells were washed and suspended with ice cold 3% FBS/PBS containing 1 μg ml⁻¹ propidium iodide (PI, Sigma-Aldrich, USA) as previously described^42,43. FCM analysis was performed with a FACS apparatus MoFlo XDP (Beckman Coulter, USA).

In vivo metastases assay

Scrambled RNA-, shMerlin-, ^wtMer-, ^Δ2–4Mer- or control vector-expressing HCCLM3 cells (1 × 10⁶) were injected into the caudal veins of 5-week-old male BALB/C nude mice. Each group had five mice. All the mouse groups were killed after 2.5 months. The lungs of each mouse were separated and fixed for H&E staining. The average number of metastatic foci in each group was counted under a microscope. The protocols for the animal experiments were approved by the Ethics Boards of the Eastern Hepatobiliary Surgery Hospital.

Statistical analysis

All the measurement data are expressed as the mean±s.d. The Pearson χ²-test was used to analyze the relationships between the expression of Merlin and the clinicopathologic features. A Kaplan–Meier analysis was used to assess the differences in survival rates. SPSS 16.0 software (SPSS Inc., Chicago, IL, USA) was also used for the above statistical analyses. A value of P<0.05 was considered to be significant.

Additional information

How to cite this article: Luo, Z.-L. et al. A splicing variant of merlin promotes metastasis in hepatocellular carcinoma. Nat. Commun. 6:8457 doi: 10.1038/ncomms9457 (2015).

References

Ishikawa, T. et al. Percutaneous transhepatic portal vein stent placement can improve prognosis for hepatocellular carcinoma patients with portal vein tumor thrombosis. Hepatogastroenterology 61, 413–416 (2014) .
PubMed Google Scholar
Bosch, F. X., Ribes, J., Diaz, M. & Cleries, R. Primary liver cancer: worldwide incidence and trends. Gastroenterology 127, S5–S16 (2004) .
Article Google Scholar
Zhang, J. et al. RMP promotes venous metastases of hepatocellular carcinoma through promoting IL-6 transcription. Oncogene 34, 1575–1583 (2015) .
Article CAS Google Scholar
Pecina-Slaus, N. Merlin, the NF2 gene product. Pathol. Oncol. Res. 19, 365–373 (2013) .
Article CAS Google Scholar
Merel, P. et al. Predominant occurrence of somatic mutations of the NF2 gene in meningiomas and schwannomas. Genes Chromosomes Cancer 13, 211–216 (1995) .
Article CAS Google Scholar
Arakawa, H., Hayashi, N., Nagase, H., Ogawa, M. & Nakamura, Y. Alternative splicing of the NF2 gene and its mutation analysis of breast and colorectal cancers. Hum. Mol. Genet. 3, 565–568 (1994) .
Article CAS Google Scholar
Kanai, Y., Tsuda, H., Oda, T., Sakamoto, M. & Hirohashi, S. Analysis of the neurofibromatosis 2 gene in human breast and hepatocellular carcinomas. Jpn J. Clin. Oncol. 25, 1–4 (1995) .
CAS PubMed Google Scholar
Lallemand, D. et al. Merlin regulates transmembrane receptor accumulation and signaling at the plasma membrane in primary mouse Schwann cells and in human schwannomas. Oncogene 28, 854–865 (2009) .
Article CAS Google Scholar
McClatchey, A. I. et al. Mice heterozygous for a mutation at the Nf2 tumor suppressor locus develop a range of highly metastatic tumors. Genes Dev. 12, 1121–1133 (1998) .
Article CAS Google Scholar
Lallemand, D., Curto, M., Saotome, I., Giovannini, M. & McClatchey, A. I. NF2 deficiency promotes tumorigenesis and metastasis by destabilizing adherens junctions. Genes Dev. 17, 1090–1100 (2003) .
Article CAS Google Scholar
Savio, M. et al. Degradation of p21CDKN1A after DNA damage is independent of type of lesion, and is not required for DNA repair. DNA Repair (Amst) 8, 778–785 (2009) .
Article CAS Google Scholar
Liu, Y., El-Naggar, S., Darling, D. S., Higashi, Y. & Dean, D. C. Zeb1 links epithelial-mesenchymal transition and cellular senescence. Development 135, 579–588 (2008) .
Article CAS Google Scholar
Benhamouche, S. et al. Nf2/Merlin controls progenitor homeostasis and tumorigenesis in the liver. Genes Dev. 24, 1718–1730 (2010) .
Article CAS Google Scholar
Zhang, N. et al. The Merlin/NF2 tumor suppressor functions through the YAP oncoprotein to regulate tissue homeostasis in mammals. Dev. Cell 19, 27–38 (2010) .
Article CAS Google Scholar
Camargo, F. D. et al. YAP1 increases organ size and expands undifferentiated progenitor cells. Curr. Biol. 17, 2054–2060 (2007) .
Article CAS Google Scholar
Manchanda, P. K. et al. Rac1 is required for Prkar1a-mediated Nf2 suppression in Schwann cell tumors. Oncogene 32, 3491–3499 (2013) .
Article CAS Google Scholar
Sherman, L. et al. Interdomain binding mediates tumor growth suppression by the NF2 gene product. Oncogene 15, 2505–2509 (1997) .
Article CAS Google Scholar
Gutmann, D. H., Geist, R. T., Xu, H., Kim, J. S. & Saporito-Irwin, S. Defects in neurofibromatosis 2 protein function can arise at multiple levels. Hum. Mol. Genet. 7, 335–345 (1998) .
Article CAS Google Scholar
Zhou, L. et al. Merlin-deficient human tumors show loss of contact inhibition and activation of Wnt/beta-catenin signaling linked to the PDGFR/Src and Rac/PAK pathways. Neoplasia 13, 1101–1112 (2011) .
Article CAS Google Scholar
Lau, Y. K. et al. Merlin is a potent inhibitor of glioma growth. Cancer Res. 68, 5733–5742 (2008) .
Article CAS Google Scholar
Kressel, M. & Schmucker, B. Nucleocytoplasmic transfer of the NF2 tumor suppressor protein merlin is regulated by exon 2 and a CRM1-dependent nuclear export signal in exon 15. Hum. Mol. Genet. 11, 2269–2278 (2002) .
Article CAS Google Scholar
Laulajainen, M., Melikova, M., Muranen, T., Carpen, O. & Gronholm, M. Distinct overlapping sequences at the carboxy-terminus of merlin regulate its tumour suppressor and morphogenic activity. J. Cell Mol. Med. 16, 2161–2175 (2012) .
Article CAS Google Scholar
Pykett, M. J., Murphy, M., Harnish, P. R. & George, D. L. The neurofibromatosis 2 (NF2) tumor suppressor gene encodes multiple alternatively spliced transcripts. Hum. Mol. Genet. 3, 559–564 (1994) .
Article CAS Google Scholar
Bianchi, A. B. et al. Mutations in transcript isoforms of the neurofibromatosis 2 gene in multiple human tumour types. Nat. Genet. 6, 185–192 (1994) .
Article CAS Google Scholar
Thurneysen, C. et al. Functional inactivation of NF2/merlin in human mesothelioma. Lung Cancer 64, 140–147 (2009) .
Article Google Scholar
Giovannini, M. et al. Schwann cell hyperplasia and tumors in transgenic mice expressing a naturally occurring mutant NF2 protein. Genes Dev. 13, 978–986 (1999) .
Article CAS Google Scholar
Deguen, B. et al. Heterogeneity of mesothelioma cell lines as defined by altered genomic structure and expression of the NF2 gene. Int. J. Cancer 77, 554–560 (1998) .
Article CAS Google Scholar
Hanemann, C. O. Magic but treatable? Tumours due to loss of merlin. Brain 131, 606–615 (2008) .
Article CAS Google Scholar
Bianchi, A. B. et al. High frequency of inactivating mutations in the neurofibromatosis type 2 gene (NF2) in primary malignant mesotheliomas. Proc. Natl Acad. Sci. USA 92, 10854–10858 (1995) .
Article ADS CAS Google Scholar
Rustgi, A. K. et al. Neurofibromatosis 2 gene in human colorectal cancer. Cancer Genet. Cytogenet. 84, 24–26 (1995) .
Article CAS Google Scholar
Jacoby, L. B., MacCollin, M., Barone, R., Ramesh, V. & Gusella, J. F. Frequency and distribution of NF2 mutations in schwannomas. Genes Chromosomes Cancer 17, 45–55 (1996) .
Article CAS Google Scholar
He, C., Zhou, F., Zuo, Z., Cheng, H. & Zhou, R. A global view of cancer-specific transcript variants by subtractive transcriptome-wide analysis. PLoS ONE 4, e4732 (2009) .
Article ADS Google Scholar
Fackenthal, J. D. & Godley, L. A. Aberrant RNA splicing and its functional consequences in cancer cells. Dis. Model Mech. 1, 37–42 (2008) .
Article CAS Google Scholar
Yi, C., Wilker, E. W., Yaffe, M. B., Stemmer-Rachamimov, A. & Kissil, J. L. Validation of the p21-activated kinases as targets for inhibition in neurofibromatosis type 2. Cancer Res. 68, 7932–7937 (2008) .
Article CAS Google Scholar
Shaw, R. J. et al. The Nf2 tumor suppressor, merlin, functions in Rac-dependent signaling. Dev. Cell 1, 63–72 (2001) .
Article CAS Google Scholar
Zhan, Y. & Chadee, D. N. Inhibition of Cdc42-mediated activation of mixed lineage kinase 3 by the tumor suppressor protein merlin. Small GTPases 1, 183–186 (2010) .
Article Google Scholar
Zhan, Y. et al. Regulation of mixed lineage kinase 3 is required for Neurofibromatosis-2-mediated growth suppression in human cancer. Oncogene 30, 781–789 (2011) .
Article CAS Google Scholar
Bosco, E. E., Nakai, Y., Hennigan, R. F., Ratner, N. & Zheng, Y. NF2-deficient cells depend on the Rac1-canonical Wnt signaling pathway to promote the loss of contact inhibition of proliferation. Oncogene 29, 2540–2549 (2010) .
Article CAS Google Scholar
Wang, T. et al. Characterisation of a novel cell line (CSQT-2) with high metastatic activity derived from portal vein tumour thrombus of hepatocellular carcinoma. Br. J. Cancer. 102, 1618–1626 (2010) .
Article CAS Google Scholar
Dong, L. W. et al. Prognostic significance of Beclin 1 in intrahepatic cholangiocellular carcinoma. Autophagy 7, 1222–1229 (2011) .
Article CAS Google Scholar
Chen, H. et al. Low glucose promotes CD133mAb-elicited cell death via inhibition of autophagy in hepatocarcinoma cells. Cancer Lett. 336, 204–212 (2013) .
Article CAS Google Scholar
Chiba, T. et al. Side population purified from hepatocellular carcinoma cells harbors cancer stem cell-like properties. Hepatology 44, 240–251 (2006) .
Article CAS Google Scholar
Shi, G. M. et al. Identification of side population cells in human hepatocellular carcinoma cell lines with stepwise metastatic potentials. J. Cancer Res. Clin. Oncol. 134, 1155–1163 (2008) .
Article CAS Google Scholar

Download references

Acknowledgements

We want to thank Huanlin Sun, Dongping Hu, Dandan Huang, Linna Guo, Shanhua Tang, Shanna Huang, Lexing Yu and Dan Cao for their technical assistance. This work was supported by grants from the National Natural Science Foundation of China (81221061 to H.W., 81272427 and 31471339 to Z.L.and 81302117 to Z-L.L.), the National Key Basic Research Program of China (2013CB910500 to Z.L.), and the Ministry of Science and Technology Key Program (2012ZX10002-009, -011 to H.W.).

Author information

Zai-Li Luo, Shu-Qun Cheng and Jie Shi: These authors contributed equally to this work

Authors and Affiliations

International Cooperation Laboratory on Signal Transduction, Eastern Hepatobiiary Surgery Institute/Hospital, The Second Military Medical University, 225 Changhai Road, Shanghai, 200433, China
Zai-Li Luo, Hui-Lu Zhang, Hai-Yang Chen, Liang Tang, Cong-Li Hu & Hong-Yang Wang
Institute of Biomedical Sciences, Fudan University, 138 Medical College Road, Shanghai, 200032, China
Zai-Li Luo
Department of Surgery, Eastern Hepatobiliary Surgery Institute/Hospital, The Second Military Medical University, 225 Changhai Road, Shanghai, 200433, China
Shu-Qun Cheng, Jie Shi & Cun-Zhen Zhang
Department of Oncology, Cancer Institute, The First Affiliated Hospital of Henan University of Science and Technology, 24 Jinghua Road, Luoyang, 471003, China
Zhong Li
Renji Hospital, School of Medicine, Shanghai Jiao Tong University, 160 Pujian Road, Shanghai, 200127, China
Bi-Jun Qiu & Hong-Yang Wang
Shanghai Center for Engineering Cell Therapy Research, 75 Qianyang Road, Shanghai, 201805, China
Zhong Li

Authors

Zai-Li Luo
View author publications
You can also search for this author in PubMed Google Scholar
Shu-Qun Cheng
View author publications
You can also search for this author in PubMed Google Scholar
Jie Shi
View author publications
You can also search for this author in PubMed Google Scholar
Hui-Lu Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Cun-Zhen Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Hai-Yang Chen
View author publications
You can also search for this author in PubMed Google Scholar
Bi-Jun Qiu
View author publications
You can also search for this author in PubMed Google Scholar
Liang Tang
View author publications
You can also search for this author in PubMed Google Scholar
Cong-Li Hu
View author publications
You can also search for this author in PubMed Google Scholar
Hong-Yang Wang
View author publications
You can also search for this author in PubMed Google Scholar
Zhong Li
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

Z-L.L. performed the majority of the experiments. H.L, S.C, B.Q., C.Z., L.T. and C.H. performed part of the experiments and data analyses. H.C and H.Z. constructed the plasmids. C.Z., J.S. S.C. and S.G. provided the tissue chips, PVTT tissues, clinical information and analyses. Z-L.L., H.W. and Z.L. designed the project. H.W. and Z.L. provided the funding for the project. Z-L.L. and Z.L. wrote and revised the manuscript.

Corresponding authors

Correspondence to Hong-Yang Wang or Zhong Li.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Supplementary information

Supplementary Information

Supplementary Figures 1-2 and Supplementary Tables 1-2 (PDF 458 kb)

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Reprints and permissions

About this article

Cite this article

Luo, ZL., Cheng, SQ., Shi, J. et al. A splicing variant of Merlin promotes metastasis in hepatocellular carcinoma. Nat Commun 6, 8457 (2015). https://doi.org/10.1038/ncomms9457

Download citation

Received: 18 June 2015
Accepted: 24 August 2015
Published: 07 October 2015
DOI: https://doi.org/10.1038/ncomms9457

This article is cited by

RNA splicing: a dual-edged sword for hepatocellular carcinoma
- Anjali Kashyap
- Greesham Tripathi
- Manoj Kumar Kashyap
Medical Oncology (2022)
LINC01348 suppresses hepatocellular carcinoma metastasis through inhibition of SF3B3-mediated EZH2 pre-mRNA splicing
- Yang-Hsiang Lin
- Meng-Han Wu
- Kwang-Huei Lin
Oncogene (2021)
The natural compound neobractatin inhibits tumor metastasis by upregulating the RNA-binding-protein MBNL2
- Juan Zhang
- Zhaoqing Zheng
- Hongxi Xu
Cell Death & Disease (2019)
Hippo pathway affects survival of cancer patients: extensive analysis of TCGA data and review of literature
- Anello Marcello Poma
- Liborio Torregrossa
- Gabriella Fontanini
Scientific Reports (2018)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.