Facts

  • Hepatocellular carcinoma (HCC) is a primary malignant disease of the liver with poor prognosis.

  • With the exception of the multi-kinase inhibitor, sorafenib, there are no effective systemic treatments available for HCC.

  • Several miRNAs are differentially expressed in HCC and may be pathogenically relevant.

  • Genetically modified animal models have been developed to demonstrate the direct role of miRNAs in the initiation and/or progression of cancer.

  • The in vivo therapeutic efficacy of certain miRNA mimics or anti-miRNA oligonucleotides has been evaluated.

Open Questions

  • Do the animal models that are available faithfully represent the etiology and natural history of human HCC?

  • Are methods for the in vivo systemic delivery of miRNAs/anti-miRNAs sufficiently developed, and can they be optimized?

  • Are miRNA/anti-miRNA molecules effective against liver cancer?

  • What are the clinical settings in which miRNA/anti-miRNA molecules may be successfully applied?

MicroRNAs (miRNAs) constitute a large class of short RNAs (e.g., 20–24 nucleotides in length), which have key roles in cell development and differentiation by mediating the post-transcriptional regulation of protein-coding genes.1 Over the past decade, the critical role of miRNAs in tumorigenesis has been widely investigated and their important regulatory action in several biological processes that are frequently altered in cancer has been described. At present, miRNAs have a well-recognized role in human carcinogenesis, including hepatocarcinogenesis, and accumulating experimental evidence indicates that they may act as oncogenes or tumor suppressor genes.2, 3 It is also well known that the expression of several miRNAs is deregulated in human HCC compared with normal tissue.4, 5 In this context, miRNAs have also emerged as novel molecules or targets for tumor therapy, and liver cancer represents an excellent model for their testing.

HCC is the most common primary liver malignant disease, and the third cause of cancer-related deaths, worldwide.6 It has a poor prognosis and the therapeutic options currently available are largely ineffective. Molecular studies have revealed several pathogenic mechanisms at the basis of liver cancer. For example, HCC pathogenesis involves multiple genetic and epigenetic alterations that lead to the deregulation of several signaling pathways, including the p53, PI3K/Akt/mTOR, Ras/Raf/MEK/ERK, IGFR, and Wnt/β-catenin pathways (for reviews, see Gramantieri et al.7 and Aravalli et al.8). Recent studies using next-generation sequencing technology have further refined our knowledge of the signaling pathways involved in liver carcinogenesis.9, 10, 11 Regulation of these cellular signaling pathways by miRNAs indicates that this class of short RNAs may present a new set of potentially therapeutic tools for HCC.12, 13

miRNAs and HCC

Several studies have revealed that the expression of certain miRNAs is deregulated in human HCC compared with matched nonneoplastic tissue (for a comprehensive review, see Negrini et al.14). Moreover, molecular and cellular studies have established that aberrantly expressed miRNAs can affect crucial cancer-associated pathways.2, 3

Cell proliferation and survival

The discovery of modulated gene targets has helped recognize the contribution of miRNAs to cancer-associated pathways. Unrestricted cellular proliferation and prolonged survival have critical roles in the process of hepatocarcinogenesis. In the cell cycle context, abnormal expression of miRNAs has been shown to alter the functionality of several cell cycle regulators, including RB1, cyclin-dependent kinases (CDKs), cyclins, and CDK inhibitors. RB1 is a target of miR-335.15 miR-335 is downregulated during mesenchymal stem cell differentiation16 and upregulated by the Wnt signaling pathway,16 which is often activated in HCC. RB1 is also a target of miR-221, a miRNA that is frequently overexpressed in HCC.17, 18 These observations suggest that expression of RB1, a key G1–S cell cycle transition checkpoint protein, may be aberrantly regulated by altered miRNA expression. Among the other proteins that control cell cycle progression (Figure 1), cyclins are reported to be upregulated in HCC, whereas negative cell cycle regulators are often downregulated compared with surrounding parenchyma.19 Overexpression of cyclinD1/CDK4 has been detected in ∼60% of HCC cases.20 Among CDK inhibitors, p16/INK4A is functionally inactivated owing to deletions in the short arm of chromosome 9 in about 20% of HCCs21 or by promoter methylation in 30–70% of HCC cases.22 All of these genes are also reported to be affected by altered expression of miRNAs (Figure 1). For example, miR-124 and miR-203, which inhibit the growth of HCC cells via downregulation of CDK6,23 are methylated and silenced in HCC cell lines and primary tumors.23 Direct targeting of cyclin D1, CDK6, and E2F3 by miR-195 can also block the G1–S transition, whereas inhibition of miR-195 promotes cell-cycle progression.24 MiR-195 can also suppress the ability of HCC cells to form colonies in vitro, and the development of tumors in nude mice. In most HCC tissues and cell lines, miR-195 expression is reduced,24 and this may contribute to the upregulation of cyclinD1/CDK that is observed in HCC. MiR-26 is another miRNA whose expression is reduced in HCC,25 thereby preventing its induction of cell cycle arrest in liver cancer cells via the direct targeting of cyclins D2 and E2.26 Members of the KIP family of CDK inhibitors, CDKN1A/p21, CDKN1B/p27, and CDKN1C/p57, also act as tumor suppressors in HCC by negatively affecting cell-cycle progression. CDKN1B/p27 and CDKN1C/p57 proteins are downregulated in HCC compared with surrounding cirrhosis.27 One mechanism that leads to a reduction in CDKN1C/p57 expression is the loss of maternal allele methylation at the KvDMR1-imprinted locus at 11p15.5. This has been found in 20–50% of HCCs.28 Another mechanism involved in the downregulation of p27 and p57 is the overexpression of miR-221/222, which occurs in about 70–80% of HCCs.18 Owing to their inhibition of RB1 expression as well, levels of miR-221/222 appear to be major factors in cell-cycle control. CDKN1A/p21 is downregulated by oncogenic members of the miR-17-92 family.29, 30, 31, 32, 33 Among the three clusters of the miR-17-92 family of the human genome, those on chromosomes 13 and 7 are upregulated in HCC.34

Figure 1
figure 1

Aberrant miRNA expression in liver cancer promotes cell-cycle progression. A simplified pathway of cell cycle regulation is presented that shows the effects of various aberrantly expressed miRNAs in liver cancer. Upregulated miRNAs are indicated with a red up arrow, and downregulated miRNAs are indicated with a green down arrow. Upregulated miRNAs exert a stronger inhibitory effect on their targets, whereas downregulated miRNAs mediate weaker effects on their targets. Taken together, multiple aberrant miRNAs act on this pathway to support cell cycle progression

Full size image

Disruption of apoptosis in HCC has been extensively reviewed.35, 36 In this context, miRNAs contribute to the altered expression of members of the BCL2 family (Figure 2). For example, upregulation of antiapoptotic members may result from reduced expression of miRNAs, with miR-122 controlling BCL-W and BCL-XL37 while the latter is also controlled by let-7 members.38 MCL1 is a target of miR-101,39 miR-193b,40 and miR-29,41, 42 and all of these miRNAs are downregulated in HCC. Conversely, proapoptotic members may be inhibited by overexpressed miRNAs. For example, the BH3-only proteins, BMF and BID, are targeted by miR-221 and miR-25, respectively.29, 30, 43 These mechanisms have the potential to allow miR-221 to protect cells from ‘anoikis’, a form of apoptosis induced by the detachment of anchorage-dependent cells from the surrounding extracellular matrix. MiR-25 may also impair the TGF-β tumor suppressor pathway. Both pathways represent critical steps in the process of metastasis.44, 45 miR-483 is a miRNA located within intron 2 of the IGF2 gene, and it is highly expressed in fetal liver, yet is barely detectable in adult liver.46 Moreover, it is overexpressed in 30–40% of HCCs either independently or in combination with IGF2. MiR-483-3p has also been shown to promote cell survival by repressing translation of the p53-inducible BH3-only protein, PUMA/BBC3.47

Figure 2
figure 2

Aberrant miRNA expression in liver cancer affects the functionality of apoptosis pathways. A simplified apoptosis pathway is presented that shows the effects of various miRNAs that are aberrantly expressed in liver cancer. The aberrant expression of multiple miRNAs supports cell survival

Full size image

Apoptosis is also controlled by molecules not belonging to the BCL2 family. For example, the inhibitor of apoptosis protein (IAP), survivin, is overexpressed in HCC and is controlled by miR-203,48 a miRNA silenced by aberrant DNA methylation in HCC tissue.23 Another miRNA, miR-21, appears to promote cell survival under stressful conditions. Indeed, miR-21 is upregulated in HCC, as well as in many other human neoplasms, and is associated with poor prognosis49, 50 and resistance to chemotherapy.51, 52, 53 These effects appears to be linked to the survival advantage imparted by miR-21 via the direct targeting of proapoptotic genes, such as the tumor suppressor lipid-phosphatase, PTEN, which controls the phosphatidyl inositol 3-phosphate kinase (PI3K) signaling pathway54, 55, 56 and programmed cell death 4 (Pdcd4) protein.53, 57, 58, 59, 60, 61, 62 The latter also has a role in TGF-β-induced apoptosis.

At the nexus of cell cycle and apoptosis signaling is a molecular pathway that is controlled by the tumor-suppressor protein, p53. Genetic alterations in the TP53 gene have been extensively described in HCC,21, 63, 64 and execution of the p53 pathway in HCC is affected by miRNA expression at various levels (Figure 2). For example, the loss of miR-122 expression in HCC cells leads to repression of p53 via upregulation of cyclin G1.65, 66 MiR-145 also induces p53 activity by an unknown mechanism,67 and is commonly downregulated in HCC.65 In many cases, miRNAs that are downstream effectors of p53 are downregulated owing to p53 loss of function. Although not fully proven in HCC, this situation is exemplified by miR-34a that is downregulated in some HCCs, most likely owing to the presence of a nonfunctional p53. However, in other HCCs, miR-34a is upregulated.54, 68 Lack of miR-34a induction can promote several biological effects through increased expression of CDK4, cyclin E2, BCL2, and the tyrosine kinase receptor, MET.69, 70, 71, 72, 73, 74 Other examples of p53-inducible miRNAs include miR-145, whose lack of expression leads to the activation of MYC,75 or the cluster, miR-192/215.76, 77 Similar to miR-34, either miR-145 or miR-192/215 can promote cell-cycle arrest and/or apoptosis, colony suppression, and can increase CDKN1A/p21 levels.67, 76, 77

Invasion and metastasis

Advanced tumor features, such as the ability of cancer cells to promote uncontrolled angiogenesis and to invade tissues and blood vessels, are also affected by miRNA deregulation. One of the signaling pathways that confers invasive potential to HCC cells is mediated by the HGF/MET axis. Hepatocyte growth factor (HGF) binds to its transmembrane tyrosine kinase receptor, MET, and promotes hepatocyte proliferation, migration, survival, and angiogenesis. HGF also mediates invasive growth during embryonic development and tumorigenesis.78, 79 Overexpression of MET is found in 40–70% of HCCs80, 81, 82, 83 and is regulated by miR-199-3p (indicated earlier as miR-199a*), miR-34a, miR-23b, and miR-1.84, 85, 86, 87, 88 All these miRNAs are downregulated in HCC, thereby contributing to the upregulation of c-MET. Among the downstream effectors of RTKs, the overexpression of RAS, but rarely activating point mutations, has been demonstrated in HCC.89, 90 All members of the RAS family have been shown to be modulated by various members of the let-7 family, and the latter are downregulated in HCC,65 as well as in several other human cancers,91, 92, 93, 94, 95, 96, 97 thus suggesting that let-7 potentially contributes to the upregulation of RAS.

As previously mentioned, the PI3K/AKT/PTEN pathway promotes cell survival as well as invasion and metastasis. Correspondingly, the targeting of PTEN by miR-21 and miR-221 can favor an invasive phenotype.98, 99 Indeed, the silencing of PTEN and PDCD4 by miR-21 has resulted in a decrease in apoptosis and an increase in cell invasion. MiR-21, which is upregulated in HCC and in most human malignancies, is linked to invasion and metastasis via its targeting of multiple tumor/metastasis suppressor genes, including tropomyosin 1 (TPM1), maspin, tissue inhibitor of metalloproteinase 3 (TIMP3) gene, and RHOB.54, 100, 101, 102, 103 In addition, miR-21-mediated inhibition of RECK, a membrane-anchored glycoprotein that negatively regulates matrix metalloproteinase-9, leads to increased cell invasion.98 In the context of the PI3K/AKT pathway, activation of the mTOR pathway has been found to be common to several human cancers, including HCC.104 Moreover, when it is overexpressed, it is associated with poor prognosis, invasion, and metastasis.105 Despite this, a very low rate of genetic alterations that affect the mTOR pathway in HCC has been reported.104 Recently, mTOR was identified as a direct target of miR-199a-3p, and to be inversely correlated with miR-199a-3p in HCC.84 The miR-199/214 cluster is of particular interest as it is downregulated in the majority of HCCs that have been studied,4, 54, 65, 84, 106, 107 as well as in other human malignancies,108, 109 in cancer-derived cell lines,110 and in experimental neoplastic and pre-neoplastic conditions.111

Other deregulated miRNAs that have a role in the invasive and metastatic properties of HCC cells include miR-122, miR-139, and miR-151. For example, loss of miR-122 enhances cell migration and invasion, and its restoration in metastatic liver cancer cells has been shown to significantly reduce migration, invasion, and anchorage-independent growth in vitro, as well as tumorigenesis, angiogenesis, and intrahepatic metastasis in vivo.66, 112, 113 Invasive and metastatic properties of HCC are also linked to the loss of control of ADAM17 by miR-122. Indeed, specific silencing of ADAM17 resulted in a dramatic reduction in migration and invasion in vitro, and reduced tumorigenesis, angiogenesis, and local invasion in the livers of nude mice.114 These observations are similar to those associated with the restoration of miR-122. Downregulation of miR-122, which has been detected in about 70% of HCCs,5, 54, 65 is also associated with the development of intrahepatic metastases114 and a shorter time to recurrence.66 Reduced expression of miR-139 has also been detected in HCC,106, 115 and has specifically been linked to the acquisition of metastatic properties.115 In HCC cells, overexpression of miR-139 has significantly reduced cell migration and invasion in vitro, and the incidence and severity of lung metastases in vivo. These effects were linked to interactions between miR-139 and the 3′ untranslated region of rho-kinase 2 (ROCK2). In human metastatic HCC, expression of miR-139 is reduced and levels of miR-139 inversely correlate with levels of ROCK2 protein in human HCC samples.115 MiR-151 is frequently amplified and overexpressed in HCC in combination with its host gene, focal adhesion kinase (FAK).116 It is possible that the host gene, FAK, cooperates with miR-151 to enhance cell motility and spreading effects, and this would suggest that miR-151 has a critical role in tumor invasion and metastasis. MiR-151-5p targeting of RhoGDIA, a putative metastasis suppressor, has also been found to increase HCC cell migration and invasion, as well as the activation of Rac1, Cdc42, and Rho GTPases.116

The β-catenin pathway has a central role in the development of liver cancer and is an essential component of both intercellular junctions and canonical Wnt signaling. Thus, β-catenin is important for the regulation of cell proliferation, differentiation, and stemness19, 117, 118, 119, 120, 121 (Figure 3). In HCC, the Wnt/β-catenin pathway has been found to be abnormally activated through gain-of-function mutations at the N-terminus of β-catenin (present in 12–26% of HCCs),21 by deletions, mutations, or epigenetic alterations of the E-cadherin gene, and by loss-of-function mutations in the AXIN1 or AXIN2 genes (present in 8–13% of HCCs).21, 122 The capacity for nuclear β-catenin to promote the epithelial to mesenchymal transition has been shown to induce invasive and metastatic properties of tumor cells.123 Correspondingly, the silencing of β-catenin reverses the epithelial–mesenchymal transition and represses metastatic potential.124 In a recent review of miRNA regulation of the Wnt/β-catenin pathway,125 it is apparent that miR-21, miR-200, miR-315, and miR-135 directly regulate the Wnt/β-catenin core. MiR-135 and miR-315 activate β-catenin by inhibiting the negative regulators, APC and axin, respectively.126, 127 MiR-200a has been shown to regulate β-catenin levels either directly128 or indirectly by modulating ZEB1/2, and this downregulation of miR-200a in HCC has been reported in various studies,4, 65, 106, 129, 130 thereby supporting its role in mediating an increase in nuclear β-catenin levels and the activation of the pathway in cancer cells. MiR-34a has also been shown to be a negative regulator of the Wnt pathway based on its targeting of WNT1.131, 132 In HCC where miR-34 is downregulated, repression of WNT1 is released. Moreover, in addition to invasion and metastasis, the Wnt/β-catenin pathway is also important for maintaining stemness.118, 119, 120 In this context, it has been proposed that upregulation of miR-181b supports this function at two levels, by repressing the Wnt/β-catenin inhibitor, NLK, and by blocking the hepatic differentiation transcription factors, CDX2 and GATA6,133 suggesting that liver cancer cells may arise from stem cells that have lost control over their self-renewal potential.

Figure 3
figure 3

Aberrant miRNA expression in liver cancer supports the activation of various signaling pathways. Simplified diagrams of RAS/MAPK, PI3K/AKT/mTOR, and WNT/β-catenin signaling pathways are shown, along with the effects of various miRNAs that are aberrantly expressed in liver cancer. As a result, cell proliferation, survival, and maintenance of stemness are enhanced

Full size image

As described above, miRNAs appear to have an essential role in the modulation of complex cross-talk that exists between pathways affecting HCC development and progression. Their understanding may facilitate the development of novel, targeted therapeutic strategies against HCC.

Animal Models of Liver Cancer

The importance of deregulated miRNAs in malignant cell transformation and tumor development has been observed in several studies. In particular, the development of animal models that are genetically modified at miRNA loci have proven the involvement of miRNAs in the initiation and progression of cancers, and provided in vivo models to test the efficacy of miRNA-based therapeutic approaches. Most of the models that have been developed are related to hematopoietic diseases, but HCC models have also been developed (Table 1).

Table 1 miRNA-based animal models
Full size table

miRNA-specific models of HCC

For HCC, various cancer genes, including tumor suppressor genes, oncogenes, and hepatitis B virus or hepatitis C virus (HCV) viral genes, have been employed for the development of mouse models.134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144 More recently, miRNA-based mouse models predisposed to HCC have been reported. A transgenic mouse model characterized by overexpression of miR-221 in the liver was developed. In these mice, male animals exhibit a strong predisposition for HCC, which includes the emergence of spontaneous nodular liver lesions with age and a strong acceleration of tumor development after treatment with the carcinogen, N-nitrosodiethylamine (DEN). Notably, the TG221 mouse was the first transgenic animal model to develop a solid tumor following overexpression of a miRNA.145 Among the miRNAs that are downregulated in HCC, two studies have reported the development of miR-122 knockout mice. These mice are characterized by hepatic inflammation, fibrosis, and the development of spontaneous liver tumors with age. These studies also demonstrated the tumor suppressor function of miR-122 in the liver, as well as its importance in liver metabolism and hepatocyte differentiation.146, 147

In addition to these miRNA-specific models, the importance of deregulated miRNA machinery components has also recently been suggested. For DDX20, a DEAD box protein component of miRNA-containing ribonucleoprotein complexes, it suppresses NF-κB activity by regulating miR-140 function.148 Correspondingly, miR-140 knockout mice149 are prone to HCC after treatment with DEN compared with control animals, and they have a phenotype that is similar to that for DDX20 deficiency. Taken together, these results suggest that miR-140 may act as a tumor suppressor by regulating the NF-κB pathway.149 These reports confirm the importance of miRNA deregulation in liver cancer.

Traditional HCC preclinical models

In addition to genetically modified mouse models, traditional preclinical models of liver cancer have been established with chemical treatment. There are two main methods of HCC induction using chemicals: (i) administration of carcinogens alone, such as DEN, over a prolonged period of time or by intraperitoneum injection,150 or (ii) administration of carcinogens (CCl4 and DEN) followed by a promotion phase (partial hepatectomy) or administration of alcohol or phenobarbital.151, 152 DEN is frequently used as a carcinogenic agent;152 however, the target organ in which it induces malignant tumors is species specific. Thus, mice treated with DEN develop liver tumors as well as gastrointestinal,153 skin, lung,154 and haematopoietic malignancies.155 In addition, the time needed to develop HCC after a single DEN injection depends on the administered dose, and the gender, age, and strain of the treated mice.156

Other models of chemically induced liver disease involve the administration of CCl4 in drinking water, by inhalation, or by subcutaneous or intraperitoneal injections.151, 157 CCl4 treatment has been shown to induce an advanced stage of liver cirrhosis and related complications, such as ascites decompensation, in both rats and mice.157, 158 However, development of HCC has not been reported. In a two-stage chemical model of HCC, CCl4 administration is accompanied by alcohol supplementation in drinking water.159 However, only after 104 weeks does this combination induce the development of HCC and metastases.

miRNA-Based Therapies

A comparative analysis of global gene expression patterns for murine and human HCC has shown that many models can reproduce key biological and molecular events observed in the human condition.160 Accordingly, mouse models have been widely used for the testing of potential therapeutic targets and for preclinical studies (see Li et al.161 for a recent review). The efficacy of miRNA-mediated HCC prevention and therapy has also been evaluated in recent years using various strategies involving miRNA inhibition or miRNA replacement.

miRNA inhibition

Krutzfeldt and colleagues were the first to demonstrate that miRNAs could be effectively silenced in mice using specific molecules. In this study, the liver-specific expression of miR-122 was inhibited with an intravenous administration of chemically modified single-stranded RNA, known as ‘antagomirs’, that were complementary to the target miRNA.162 The safety of this silencing approach was established in nonhuman primates, where no evidence of liver toxicity was detected.163 On the basis of these promising results, many studies of miRNA inhibition were subsequently performed. The tumor-promoting activity of miR-221 in HCC was also demonstrated in vivo. In one notable study, systemic delivery of a cholesterol-tagged anti-miR-221 to orthotopic HCC tumors was found to reduce tumor cell proliferation and promote survival.164 In a second study, anti-miR-221 oligonucleotides were found to exhibit significant antitumor activity in the TG221 transgenic mouse model, where a significant reduction in the number and size of tumors in the treated mice was confirmed with histopathological analyses. Significant downregulation of miR-221 levels has also been detected in liver tissues of anti-miR-221-treated mice, thereby confirming the ability of these molecules to inhibit endogenous miR-221145 (Table 2). The importance of manipulating miRNA levels as a therapeutic tool for treating liver cancer is also supported in the work by Lim et al.,143 where anti-miR inhibition of miR-494 (a miRNA overexpressed in human HCC) resulted in a significant reduction in tumor size in a primary MYC-driven liver tumor mouse model.

Table 2 miRNA-based therapies in HCC preclinical models
Full size table

HCV promotes HCC by inducing chronic liver disease and cirrhosis. MiR-122 has been found to support HCV replication.165, 166, 167 Correspondingly, treatment of chronically infected nonhuman primates with an LNA molecule complementary to the miR-122 sequence led to a long-lasting and well-tolerated suppression of HCV viremia. These results support the hypothesis that miR-122 represents a potential target for antiviral intervention and possible liver disease sequels.168, 169 Currently, an anti-miR-122 oligonucleotide, ‘miravirsen SPC3649’ (Santaris Pharma, A/S, Hørsholm, Denmark), is being evaluated in a phase 2 clinical trial for the treatment of HCV infection (ClinicalTrials.gov Identifier NCT01200420)170 (Figure 4). However, the concomitant inhibition of the tumor-suppressor activity of miR-12265, 146, 147 must also be considered.

Figure 4
figure 4

miRNA-based approaches that are moving toward clinical trials. An overview of the many preclinical studies that have produced promising results, as well as the clinical trials that are being planned or are underway

Full size image

miRNA replacement

The miRNA restoration is another strategy that has been used to elucidate the functions of miRNAs, to identify their roles in tumorigenesis, and to provide preclinical tools for the treatment of cancer. One anticancer approach involves the restoration of miRNAs that are generally downregulated in cancer cells. In vivo xenograft models have been used to test the tumor suppressor activity of several miRNAs that were delivered by adeno-associated viruses (AAV), nano-sized lipid particles, or by cholesterol-conjugated modified oligonucleotides (Table 2). Hou et al.171 used both cholesterol-conjugated small RNAs and an AAV delivery system to effectively restore miR-199a/b-3p expression in HCC tissues of HCC-bearing nude mice. As a result, tumor growth was inhibited without evidence of toxicity. Lentivirus-mediated expression of miR-199a was also able to inhibit tumor growth in a nude mouse xenograft model,172 thereby confirming the antitumor effect of this miRNA in HCC and its therapeutic potential.

To provide a better model of human HCC, genetically engineered or chemically induced mouse models have been employed. For example, using MYC-driven mouse models of HCC, the tumor suppressor activity of miR-26a26 and miR-122146 have been demonstrated. In a DEN-induced HCC mouse model, the systemic administration of miR-124 led to a reduction in tumor growth and tumor size via induction of apoptosis and deregulation of a miRNA/inflammatory feedback loop circuit involving hepatocyte nuclear factor 4alpha (HNF4α).173 Taken together, these studies indicate that miRNA mimics represent a promising approach for the treatment of cancer, and could be translated into clinical practice. In fact, MRX34 (Mirna Therapeutics, Inc., Austin, TX, USA), a liposome-formulated mimic of the tumor suppressor miR-34a, is the first miRNA mimic that has entered a multicenter phase I clinical trial aimed at evaluating safety in patients with primary liver cancer or with liver metastasis from other cancers (ClinicalTrials.gov Identifier: NCT01829971)174 (Figure 4).

Oncolytic viruses

A distinct approach that has become interlaced with miRNA expression in HCC is the field of oncolytic viruses. These vectors have tumor-specific activity that results in the cytolysis of cancer cells, whereas normal tissues exhibit minimal toxicity in vivo. Thus, the use of oncolytic viruses represents a promising approach for the treatment of cancer (for a recent review, see Patel et al.175 and Callegari et al.176). The differential expression of miRNAs in neoplastic versus normal tissues has also been used to improve the safety of oncolytic virus-based therapies. For example, to overcome liver toxicity associated with the systemic delivery of oncolytic adenoviruses, regulatory miRNA sequences have been inserted in the genome of conditionally replicating adenoviruses (CRAds). In one case, CRAds regulated by miR-122 were generated to provide liver-specific suppression of viral replication without compromising the replication capacity of the modified virus in cancer tissues in vivo.177, 178, 179 Other uses of oncolytic viruses have included the generation of a let-7-dependent adenovirus that is able to replicate in tumor cells, and not in normal liver cells, in a HCC xenograft model.180 Similarly, a miR-199-dependent CRAd was generated to replicate in HCC tumor cells and not in normal liver cells. This virus was able to control tumor growth in a subcutaneous xenograft model in nude mice, and in HCCs arising in immune-competent mice, without causing significant hepatotoxicity181 (Table 2).

Conclusions

HCC is the most common type of liver cancer diagnosed, and treatment options currently depend on tumor size and staging. For patients with advanced, non-resectable HCC, the only systemic therapy available involves the multi-kinase inhibitor, sorafenib.182, 183 This molecule targets the tyrosine kinase receptors VEGFR and PDGFR, as well as the serine–threonine kinases c-RAF and BRAF. Randomized phase 3 trials with sorafenib have shown an improvement in median overall survival and progression-free survival for HCC patients.182, 184 However, more effective treatments that have less toxic side effects are still needed. The use of miRNAs or anti-miRNAs as antitumor therapeutic molecules has attracted a lot of attention. At present, their short-term safety has been proven in several instances, albeit their efficacy as anticancer agents remains to be fully confirmed. Various studies have demonstrated that miR-26a, miR-199, miR-122, anti-miR-221, and miR-494 reduce tumor growth, and miR-34a has entered a clinical trial. For anti-miR-122, its capacity to inhibit HCV replication has the potential to prevent the development of HCC. Future investigations may also focus on the use of miRNA/anti-miRNA oligonucleotides for the prevention of HCC owing to the limited presence, or absence, of side effects associated with these approaches. To this end, animal models that are predisposed to HCC represent important instruments for testing the effectiveness of miRNA/anti-miRNA molecules.

Animal models of HCC have been established and are widely used for preclinical and translational studies. Moreover, the study of several HCC murine models has provided a comprehensive characterization of the molecular mechanisms of liver carcinogenesis and has also resulted in the development of new therapeutic strategies.185 Compared with human HCC, genetically modified mouse models generally develop HCC without the development of liver cirrhosis. However, this condition is present in >80% of human HCC cases. The administration of CCl4 has been shown to induce advanced stages of liver cirrhosis, although this is not accompanied by the development of HCC.157 Therefore, CCl4 treatment of genetically modified mice may lead to the development of models that more closely represent the human condition where HCC develops in cirrhotic livers. Future studies will be needed to investigate these possibilities.