**8. Non-coding RNA**

#### **8.1 MicroRNA**

46 Hepatocellular Carcinoma – Basic Research

BMI1 mediates stemness features in HCC cells. In HCC, SP cells expressing BMI1 showed enhanced tumorigenic potential compared to the corresponding non-SP cells. Knockdown of BMI1 markedly abolished the tumor-initiating ability of SP cells in non-obese diabetic/severe combined immunodeficiency mice, leading to a 100 fold decrease of tumorigenic activity (Chiba et al., 2008). Such decrease in tumorigenic activity was accompanied with a reduction of SP cell number in different HCC cell lines. It is shown that BMI1 mediates the suppression of INK4A/ARF and drives self-renewal. Inhibition of BMI1 resulted in the derepression of INK4A/ARF, and in turn disrupted self-renewal in SP cells, hence suppressed SP cells survival upon long time culture. BMI1 additionally regulates diverse cellular processes including cell cycle, apoptosis and senescence by the repression of

Since there are numerous targets of BMI1 in human genome, it is predictable that the oncogenic effect of BMI1 should not simply depend on INK4A and ARF repression. BMI1 was able to cooperate with activated RAS to transform hepatocytes into malignant cells (Xu et al., 2009). Furthermore, BMI1 expression in HCC is significantly associated with the expression of ABC transporter B1 (ABCB1) which was consistently reported to generate the multiple drug-resistant phenotype (Effendi et al., 2010). It is possible that one of the downstream targets of BMI1 is ABCB1. The ability to eliminate cancer progenitor cell in HCC by BMI1 inhibition is a potent anticancer mediation, and potentially provides a cure

Polycomb repressive complex 2 (PRC2) is another modifier of the chromatin structure, which determines the activity of gene expression. PRC2 primarily regulates gene expression by inducing the methylation on lysine 9 and lysine 27 of histone 3, and plays important roles during development and tumorigenesis. In this complex, EZH2 is the catalytic subunit directly involves in transferring methyl-group to the histone tails (Kirmizis et al., 2004). The level of EZH2 is important in determining the PRC2 activity in cells. EZH2 regulates cell proliferation, and its expression is often augmented during tumorigenesis. EZH2 overexpression can be observed in many cancers, including prostate, breast and pancreas cancer, and most often high level of EZH2 is correlated to aggressiveness of the malignancies (Tsang et al., 2011). EZH2 is rarely detected in normal hepatocytes, but is

EZH2 is involved in numerous cellular processes and signaling pathways and evidences suggested that EZH2 promotes cancer development by repressing diverse tumorsuppressors. Recently, EZH2 reportedly activated Wnt signaling in HCC. Concurrent overexpression of EZH2 and β-catenin was observed in more than 30% of human HCC, which is associated with tumor progression (Cheng et al., 2011). In HCC cells, EZH2 is found to frequently occupy the promoter of numerous Wnt pathway antagonists. As such, these antagonists are silenced, relieving the inhibitory effect on Wnt/β-catenin signaling. In immortalized hepatocytes, ectopic expression of EZH2 activated Wnt/β-catenin signaling and triggered cell proliferation (Cheng et al., 2011). Conversely, downregulation of EZH2 inhibited β-catenin signaling, resulting in the retardation of HCC cell growth. Study showed that knockdown of EZH2 by lentivirus-based shRNA inhibited tumor growth in vivo (Chen et al., 2007), demonstrating a potent effect against HCC by targeting EZH2. The significance

frequently detected in HCC cell lines and HCC tissues (Chen et al., 2007).

the INK4A/ARF expression (Xu et al., 2009).

for HCC patients.

**7.2 EZH2** 

MicroRNAs are small non-coding RNAs that regulate the translation of many genes. They not only regulate normal cell development but also play important roles in cancer development and progression by affecting cell survival, angiogenesis and metastasis. Many studies illustrated the potential of manipulating microRNA expression in cancer therapy. It is believed that microRNA-based remedy can have a huge impact on cancer cells, as they regulate whole programs of gene expression via suppressing hundreds of genes simultaneously (Farazi et al., 2011). In human HCC, numerous microRNAs are identified to give major contributions, either having oncogenic or tumor suppressing ability. Here, those microRNAs having great potential as HCC therapeutic targets will be discussed.

OncomiR is a novel term coined for microRNAs possessing proto-oncogenic effects in cancers. In HCC, a number of oncomiRs are identified and their roles are characterized. Among them, the roles of microRNA-21 (miR-21) in HCC development were welldocumented (Liu et al., 2010). miR-21 is universally overexpressed in majority of cancers and is phenomenally involved in approximately all tumorigenic processes. miR-21 is able to induce cell transformation, mediate cancer cell growth, cell cycle and self-renewal, prevent apoptosis, promote metastasis and generate drug-resistance (Liu et al., 2010). High expression of miR-21 is correlated with advanced tumor stage, frequent metastasis and poor patient prognosis. In HCC, miR-21 is overexpressed, and has been proved to promote malignant cell growth and spreading by targeting tumor suppressor PTEN and inducing FAK phosphorylation (Meng et al., 2007). Furthermore, miR-21 induces resistance to anticancer effect of interferon-α and 5-fluorouracil in HCC (Tomimaru et al., 2010). Importantly, inhibition of miR-21 is able to reduce the aggressiveness in HCC, and relieves the suppressive effect to several tumor suppressor genes targeted. Besides miR-21, targeting other oncomiRs such as miR-29 (Xiong et al., 2010) and miR-221 (Pineau et al., 2010) is also robust in reversing the malignant phenotypes in HCC.

Novel Therapeutic Targets for Hepatocellular Carcinoma Treatment 49

There are two other lncRNAs recently reported to be overexpressed in HCC cell lines and tissues, namely the metastasis associated lung adenocarcinoma transcript 1(MALAT-1) and HOX transcript antisense RNA (HOTAIR). Both of them have been previously implicated in other malignancies regarding their capacity to promote cancer metastasis. In HCC, their expression is correlated to the prognosis of the patients which predicts tumor recurrence after liver transplantation. Inhibition of MALAT-1 in vitro effectively reduced cell viability, motility, invasiveness, and HCC cells became sensitive to pro-apoptotic signal (Lai et al., 2011). Similarly, knockdown of HOTAIR decreased cell viability and cell invasiveness, as well as sensitized HCC cells to tumor necrosis factor- α induced apoptosis, and cytotoxic effect of doxorubicin and cisplatin (Yang et al., 2011). Their roles in HCC progression are important

Accumulating evidences support that the development of HCC is based on the cancer stem cell (CSC) model. In this hypothesis, there is only a subset of cells within a tumor or in the cell pool that sustains malignant growth. Such cellular subset is referred to as the cancerinitiating cells or cancer-propagating cells (Visvader, 2011). It provides the explanations for cancer initiation, local recurrence, metastasis and therapy resistance which raised enormous controversies in the past. CSCs have been identified in many cancer types including HCC (Mishra et al., 2009). In this regard, CSC should be the principal target of HCC therapy. However, conventional methods such as chemotherapy and radiotherapy are ineffective because of the CSC resistant properties, as well as their pro-angiogenic effect. Studies are

Numerous surface markers for HCC stem cells were identified, and they include CD133, CD90, CD44, CD13 and EpCAM. Although their roles in liver CSC are unclear, studies showed that targeting these markers can specifically harm CSC with high efficacy. It is reported that tumorigenicity and invasive capacity of liver CSC were impaired by targeting CSC surface marker EpCAM, leading to reduction of CSC pool (Yamashita et al., 2009). Besides, inhibition of CD44 in HCC cells could enhance apoptosis, reduce tumorigenicity and invasion. Interestingly, isoforms of CD44 are differentially expressed between HCC and normal hepatocytes. Targeting of the CD44 isoforms prevalent in HCC was able to selectively deplete HCC cells without harming normal cells (Miletti-González et al., 2005). Therefore, direct targeting of CSC-specific markers may also be a promising therapeutic

Cancer stem cells share various common characters with somatic and embryonic stem cells. Many signaling pathways observed exclusively in stem cells can also be detected in cancer stem cells. These signalings include Wnt/β-catenin, Hedgehog and Notch signaling (Mishra et al., 2009). Disturbing the signalings involved in normal stem cell fate reportedly decreased the self-renewal and proliferating capabilities of CSCs. For example, small molecule inhibitor of hedgehog pathway could reduce the likely CSC with progenitor

which provide a rational base to take into consideration during HCC therapy.

vigorously conducted to develop effective methods to extinguish CSC in HCC.

**9. Liver cancer stem cell** 

**9.1 Cancer stem cell markers** 

strategy to eradicate liver CSC.

**9.2 Stemness signaling** 

On the other hand, various microRNAs with tumor suppressing capacity are lost or underexpressed in HCC. Restoration of their expression in HCC is another effective approach against HCC. Previous studies suggested a strong association between low miR-26 expression and both prognosis and response to interferon therapy in patients with HCC (Ji et al., 2009). Low miR-26 expression is highly associated with tumor formation in vivo, and replenishing miR-26 in liver tumors with the use of gene therapy could generate potent antitumor effects. miR-26a in particular is substantially reduced in MYC-induced HCC (Ji et al., 2009). The replacement therapy of miR-26 is considered safe, because miR-26 is expressed by most normal cells and is unlikely to be toxic. Other than miR-26, miR-122 is significantly downregulated in liver cancer with intrahepatic metastasis. Restoration of miR-122 reduced cell migration, invasion and colony formation ability in vitro, and tumorigenesis, angiogenesis and metastasis in vivo (Coulouarn et al., 2009). Currently, more microRNAs with antitumor effect including miR-199a/b-3p were identified and potentially play a pivotal role in the combat against HCC (Hou et al., 2011).

To note, increased activity of RNA-induced silencing complex (RISC) was observed in HCC (Yoo et al., 2011). The role of RISC is critical in facilitating activity of RNAi including microRNA-mediated target silencing. The components of RISC including AEG-1 and SND1 were both overexpressed in HCC, which leads to the hyperactivity of RISC. Increased RISC activity resulted in an accelerated degradation of numerous tumor suppressor genes that are the target of various oncomiRs. Report showed that inhibition of RISC activity by knocking down SND1 abrogated cell growth in HCC cells in vitro and in vivo (Yoo et al., 2011). It not only unveils a new microRNA associated pro-tumorigenic mechanism, but also provides an additional approach to disrupt microRNA-mediated tumorigenic effect during HCC remedy.

#### **8.2 Long non-coding RNA**

In eukaryotes, there are abundant amount of transcripts which are long in length and lack any substantial open reading frame as well as protein coding capacity. Increasing evidences suggested that these long non-coding RNAs (lncRNA) play critical role in cellular processes such as development, via the modulation of chromatin structure. Some of them possess the ability to modulate cancer epigenome and contribute to different pathological conditions such as tumor invasion and metastasis. A better understanding in the oncogenic mechanisms of lncRNA will unveil a new direction in cancer therapy.

Highly upregulated in liver cancer (HULC) is an lncRNA that is frequently overexpressed in HCC. siRNA knockdown of HULC in HCC cell lines was able to alter the expression of genes described in the context of HCC (Panzitt et al., 2007). Reduction of cellular HULC upregulated genes participates in diverse biological processes including cell differentiation, cell adhesion, protein phosphorylation and tumor suppression. Another study reported that HULC was also expressed in metastasized tumor nodules in liver originated from colorectal cancer, but not in primary colorectal cancer (Matouk et al., 2009). It suggested that expression of HULC might be a pre-requisite for any tumor formed in liver. The importance of HULC in HCC is further supported by the observation that HULC expression is strongly linked to HBV infection (Matouk et al., 2009). Due to the high specificity of HULC for cancer located in liver, it is worthy of studying its potential role in managing HCC.

There are two other lncRNAs recently reported to be overexpressed in HCC cell lines and tissues, namely the metastasis associated lung adenocarcinoma transcript 1(MALAT-1) and HOX transcript antisense RNA (HOTAIR). Both of them have been previously implicated in other malignancies regarding their capacity to promote cancer metastasis. In HCC, their expression is correlated to the prognosis of the patients which predicts tumor recurrence after liver transplantation. Inhibition of MALAT-1 in vitro effectively reduced cell viability, motility, invasiveness, and HCC cells became sensitive to pro-apoptotic signal (Lai et al., 2011). Similarly, knockdown of HOTAIR decreased cell viability and cell invasiveness, as well as sensitized HCC cells to tumor necrosis factor- α induced apoptosis, and cytotoxic effect of doxorubicin and cisplatin (Yang et al., 2011). Their roles in HCC progression are important which provide a rational base to take into consideration during HCC therapy.
