Oncogene (2006) 25, 3857–3865 & 2006 Nature Publishing Group All rights reserved 0950-9232/06 $30.00 www.nature.com/onc REVIEW RNA interference-mediated prevention and therapy for hepatocellular carcinoma

PR Romano, DE McCallus and CJ Pachuk

Nucleonics, Horsham, PA, USA

Hepatocellular carcinoma (HCC) is the fourth leading established infections would therefore significantly cause of cancer-related death and is on the increase reduce new cases of HCC that arise each year. Although worldwide. Hepatocellular carcinoma results from chronic an effective vaccine for HBV has been available for liver disease and cirrhosis most commonly associated with 20 years, new infections in the US adult community chronic B (HBV) or (HCV) infection. continue to increase at a rate of 78 000–80 000 a year. The highest incidences of HCC are found in China and This is likely owing to the fact that the vaccine is Africa, where chronic HBV infection is the major risk underutilized by the adult population. A vaccine is not component. In the United States, Europe and Japan, the currently available for HCV. Small-molecule therapies significant increase in HCC and HCC-related deaths and interferon treatment are available for both viral within the last three decades is mainly attributed to the infections, but these therapies suffer from a number of rise in the number of HCV-infected individuals; smaller drawbacks including the selection of drug-resistant increases of HCC are associated with HBV. Given that mutants, toxicity and limited efficacy. A promising HCV and HBV infection account for the majority of new class of therapeutics based on RNA interference HCCs, therapeutic and prophylactic approaches to (RNAi) technology has already demonstrated efficacy in control or eliminate virus infection may prove effective preclinical studies and has shown the potential to in reducing the occurrence of HCC. Although anti-viral overcome many of the limitations associated with therapies exist for both HBV and HCV infections, they current therapies. The use of RNAi-based therapeutics are ineffective for a significant number of patients. In as anti-hepatitis drugs is the main focus of this paper. addition, some treatments such as interferon therapy are RNA interference is a phenomenon in which genes are dose limiting owing to toxic side effects. Clearly, new silenced in a sequence-dependent manner at the level of approaches are needed. RNA interference (RNAi)-based mRNA degradation (Figure 1) (Fire et al., 1998; Lau approaches may meet this need and have already shown et al., 2001). Short dsRNA (double-stranded RNA promising preclinical results in cell culture and animal molecules), of about 19–23 bp (known as short interfer- models. Although this paper focuses on the potential of ing RNAs or siRNAs), are involved in mediating the RNAi as a prophylactic for HCC development, the degradation of mRNA (Hamilton and Baulcombe, potential use of RNAi-mediated approaches for HCC 1999; Zamore et al., 2000). Single RNA strands derived therapy will also be discussed. from these siRNAs stably associate with RNA-induced Oncogene (2006) 25, 3857–3865. doi:10.1038/sj.onc.1209549 silencing complexes (RISC) found within cells (Gregory et al., 2005; Matranga et al., 2005; Rand et al., 2005). Keywords: HCC; HBV; HCV; RNAi; shRNA; SiRNA Once bound to an siRNA, RISC becomes activated and is capable of binding to mRNAs that share significant sequence homology to the bound siRNA. This binding is mediated by Watson–Crick base pairing between the two RNAs. Activated RISC is then capable of degrad- Introduction ing the target RNA (Hammond et al., 2001). The process is catalytic and once activated, RISC is thought Chronic and C infections are the largest risk to potentiate the degradation of thousands of target factors associated with the development of hepatocel- mRNA molecules. lular carcinoma (HCC). Worldwide, approximately 20– Because RNAi is an innate phenomenon that has 40% of individuals chronically infected with hepatitis B been shown to occur in all cell types examined, it can virus (HBV) will die from liver disease and HCC, theoretically be harnessed to silence selected genes for whereas 5–7% of those chronically infected with therapeutic purposes. RNA interference-based thera- hepatitis C virus (HCV) will die from these complica- peutics may be designed to target viral genes and/or tions. The ability to prevent infection and/or resolve genes associated with disease including cancer. Intro- duction of effector siRNAs into cells has typically been Correspondence:Dr CJ Pachuk, Nucleonics, 702 Electronic Drive, performed in one of two manners. Chemically synthe- Horsham, PA 19044, USA. sized siRNAs have been exogenously delivered to cells E-mail:[email protected] through transfection or electroporation in cell culture or RNAi-mediated prevention and therapy for HCC PR Romano et al 3858 shRNA gene

a Long double-stranded RNA siRNA

Expression Vector

b d Dicer siRNA

shRNA gene Ago2 c shRNAs RISC

e Nucleus mRNA AAAAA

Cytoplasm

AAAAA

Figure 1 RNAi Mechanism. (a) Double-stranded RNA (dsRNA) is introduced into the cytoplasm of the cell via a DNA expression vector, long dsRNA, or short-inhibitory RNAs (siRNAs). (b) The long dsRNA is processed to B21 nucleotide (nt) siRNAs by the enzyme Dicer and, in association with auxiliary protein(s), is transferred to the RNA-induced silencing complex (RISC). (c) DNA expression vectors enter the nucleus where the short-hairpin RNA (shRNA) gene is transcribed. The shRNA is transported to the cytoplasm where siRNAs processed from the shRNAs are bound by RNA-induced silencing complexes (RISC). (d) siRNAs of 19–23 nts in length directly enter the RISC complex. (e) Once bound by RISC, the siRNAs are unwound and one strand of the siRNAs is discarded. The other (guide) strand anneals to its target sequence within an mRNA. This mRNA is cleaved by the ‘Slicer’ component of RISC.

into animals through intravenous (i.v.) or local injec- individual siRNAs in clinical trials before cocktails of tion. Intracellular production of siRNAs has been siRNAs can be used. achieved through the delivery of plasmids or viral Both exogenously administered siRNAs and endo- vectors containing expression units for short hairpin genously expressed shRNAs have been shown to reduce RNAs (shRNAs) in which the siRNA is contained and the RNA levels of pathogenic animal viruses. A wide processed from one of the stems of the hairpin. Both variety of viruses are susceptible to RNAi (for a review types of siRNA technologies offer certain advantages see Leonard and Schaffer, 2005). For use in the clinic, and suffer from distinct limitations. The use of siRNAs the ability to deliver the RNAi molecule (siRNA or theoretically allows one to deliver a known quantity of vector) to the infected cell will be of utmost importance these effector molecules to cells. In addition, siRNAs and the challenge for researchers is to achieve even a have an advantage over shRNA expression vectors in fraction of the success in vivo that has been demon- that they do not need to be taken up by the nucleus in strated in various cell culture models. order to exert an effect. The persistence of unmodified Both HBV and HCV have been shown to be amenable siRNAs has been shown to be limited however, owing to to downregulation by RNAi (Radhakrishnan et al., instability of unmodified RNA, and therefore frequent 2004). This inhibition has been achieved using both dosing will likely be needed to maintain efficacy in a siRNAs and vector-mediated RNAi. The following therapeutic setting. Although certain RNA modifica- review will discuss these results and the hurdles that tions have been shown to increase the stability of face the development of RNAi-based therapeutics for siRNA, some of these modifications result in loss of the treatment of hepatitis B and C. In addition, the siRNA activity. The ideal modifications for therapeutic potential use of RNAi-based therapy for the treatment siRNAs will therefore need to balance stability with of HCC will also be discussed. efficacy. The use of shRNA expression plasmids or viral vectors does not suffer from similar stability problems and can be used to express large quantities of shRNA RNA interference as a therapy for chronic hepatitis B intracellularly over extended periods of time. In addi- virus infection tion, a single expression vector can be designed to encode for multiple shRNAs, thereby increasing the is an attractive target for the potency of the product. Although, multiple chemically development of RNAi-based therapeutics given the limi- synthesized siRNAs can theoretically also be used, tations of current therapies and the poor patient response regulatory requirements necessitate the testing of rate to these therapies. Although oral nucleoside

Oncogene RNAi-mediated prevention and therapy for HCC PR Romano et al 3859 analogues such as lamivudine and adefovir significantly offer a significant advantage over existing treatment lower viral titers and have shown some promise for the options, because they would decrease the visibility of treatment of chronic hepatitis B, they do little in the way hepatocytes to the immune system thereby decreasing of decreasing HBV antigen load (Lau et al., 2000; the severity of chronic inflammation/hepatitis and Papatheodoridis et al., 2002). Reduction of antigenemia secondarily preventing the development of HCC. Such is crucial, as antigenemia is the major factor in the therapies have the potential to be beneficial for healthy development of liver disease. Liver pathology associated carriers as well, a subset of patients who are at risk of with HBV infection is primarily immune-mediated and developing chronic active hepatitis but for whom no is thought to result from cell-mediated immune current therapy exists (Mendy et al., 1999). In addition, responses directed against infected hepatocytes. During a subset of patients exhibiting non-immune-mediated chronic infection, HBV DNA is frequently found to be liver pathology caused by L antigen over expression integrated into host DNA in a form that can be (Huang and Yen, 1993) is also likely to benefit from transcribed by host RNA polymerase II ( II). treatment. Transcription of these integrants, most of which encode In the absence of a vigorous immune response to HBV X protein and HBsAg, and transcription of destroy antigen-harboring hepatocytes, elimination of episomal HBV covalently closed circle (cccDNA) result viral antigen can only be achieved through methods that in significant expression of HBV antigens that con- prevent antigen expression. Therapies employing RNAi tribute to antigen load within infected cells (Figure 2). either alone or in conjunction with antiviral therapies Even after reduction of viral titer by oral nucleoside are ideal in that inhibition of and analogues, antigens continue to be expressed from downregulation of antigen expression could be induced integrants and persisting cccDNA allowing hepatocytes simultaneously and by different mechanisms. Based on to remain susceptible to immune attack. The chronic the immunopathogenesis of HBV-associated liver dis- inflammation induced by the immune response to ease, resolution of HBV infection and antigen expres- persistent antigen expression is thought to contribute sion by non-immune mediated mechanisms such as to the development of HCC. Therefore, treatments that RNAi would be ideal. RNA interference-based strate- target antigen expression as well as viral replication gies are ideal in another way:the selection of escape

HBV HBV

HBsAg X X X X HBXAg

mRNA Core protein

cccDNA

Nucleus HBeAg

Cytoplasm Pre-genomic RNA

Figure 2 Hepatitis B virus (HBV) replication and antigen expression. Hepatitis B virus enters an uninfected hepatocyte. The partially double-stranded DNA genome of HBV is transported to the nucleus where it is converted to covalently closed circle DNA (cccDNA) . Covalently closed circle DNA is used as a template by host RNA polymerase II to produce all of the HBV RNAs expressed during infection. These include the HBV mRNAs that encode all of the HBV antigens and the HBV pregenomic RNA that is encapsidated within viral particles and converted to partially double-stranded DNA within particles via the action of the co-encapsidated Reverse Transcriptase . The HBV proteins encoded by the HBV mRNAs include the surface antigens (large, middle and small surface Antigens), core antigen, e antigen, reverse transcriptase and X. With the exception of e Ag, all viral proteins are needed for the efficient production of viral particles which are secreted from the cell. Secreted virus can then infect other uninfected hepatocytes. Peptides derived from all viral proteins are displayed on the surface of cells via MHC class 1 receptors, rendering infected hepatocytes susceptible to cellular immune responses. It is this immune response against infected cells that results in inflammation and liver damage.

Oncogene RNAi-mediated prevention and therapy for HCC PR Romano et al 3860 mutants can be avoided. Unlike small-molecule thera- diameter) of liver sinusoids and be taken up by peutics, which target a specific viral activity and hence a hepatocytes while avoiding lysosomal degradation. specific portion of the virus, RNAi therapeutics can be devised to cover multiple if not all regions of the virus, Hydrodynamic delivery. To circumvent these delivery making it difficult if not impossible for the virus to issues and enable studies on RNAi function in vivo, generate viable escape mutants. studies in HBV mouse models have initially relied on Antigenemia also plays a role in viral persistence. The hydrodynamic tail vein injection to achieve high-level high amount of surface antigen particles expressed in hepatocyte transfection. This method employs rapid, chronically infected patients is believed to interfere with high-volume injection of a nucleic acid solution through the production of a resolving surface Antigen (sAg)- the mouse tail vein. The build up of hydrostatic pressure specific humoral response. Early antigen (eAg), which is in the inferior vena cava forces the nucleic acid solution not needed for viral infection, is believed to play a role into a number of different organs with the highest in immunotolerance (Chen et al., 2005). Therefore, uptake observed in hepatocytes of liver. Up to 40% of reduction of both of these antigens by RNAi should liver hepatocytes can be transfected with the bulk of the enable the immune system to mount a resolving immune DNA taken up by hepatocyte nuclei (Lui et al., 1999; response to HBV infection. Zhang et al., 2004a).When multiple nucleic acid molecules are co-administered by hydrodynamic injec- In vitro models tion, colocalization of these molecules within the same Evaluation of siRNA efficacy and nucleic acid delivery subset of hepatocytes occurs with high frequency, paradigms has been enabled by the existence of relevant making co-transfection with multiple nucleic acid cell culture models and in vivo mouse models. As HBV molecules possible. Using this mode of delivery, a can not infect cells in culture, in vitro proof of concept number of groups have demonstrated that co-transfec- studies have focused on delivery of HBV-specific tion of HBV expression plasmids with anti-HBV siRNA siRNAs into cells transfected with infectious HBV or shRNA expression plasmids can downregulate HBV clones (Hamasaki et al., 2003) or into cells stably antigen production and HBV replication in vivo. Using expressing wild-type or lamivudine-resistant HBV virus this approach, Giladi et al. (2003) and Klein et al. (2003) (Shlomai and Shaul, 2003; Ying et al., 2003). A number have shown that unmodified siRNAs directed against of RNAi effector molecules designed to target different HBV core and surface antigen mRNAs had significant regions of HBV have been shown to be effective, dose-dependent effects on HBV expression in immuno- indicating that a multitude of different siRNAs and competent mice. Other groups have demonstrated that shRNAs can be potentially utilized for treating HBV. expressed siRNAs derived from pol III-driven shRNA- Results indicate that both HBV-specific siRNAs and pol containing plasmids are also effective in hydrodynamic III (Zhang et al., 2004b; Guo et al., 2005) and pol II co-transfection experiments (McCaffrey et al., 2003; Wu (Ren et al., 2005) expressed shRNAs lead to significant et al., 2005). Additionally, McCaffrey et al. (2003) reductions in HBV replication and expres- showed this effect in both immunodeficient and immune sion. These reductions are RNAi mediated and are not competent mice, indicating that B and T-cell function associated with a nonspecific interferon response (Ren were not required for the expressed siRNA anti-HBV et al., 2005). In one study comparing RNAi with activity. The use of immunodeficient mice, allows long- lamivudine, Guo et al. (2005) have shown that expressed term expression of HBV (Yang et al., 2002). siRNAs targeting the polymerase and surface antigen mRNAs are effective at reducing viral mRNA and Non-hydrodynamic delivery. The above studies all HBsAg production in HepG2.215 cells, whereas, as demonstrate the in vivo effectiveness of RNAi against expected, the polymerase inhibitor lamivudine reduced HBV when hydrodynamic injection is employed to the load of core-related DNA but did not affect RNA or deliver the effector RNAs and DNAs to hepatocytes. To surface antigen levels. These results indicate a potential be therapeutically viable however, clinically relevant advantage of RNAi-based therapies over existing delivery modalities must be developed. Because rapid treatment modalities in that antigenemia and viral degradation of unmodified RNA (Soutschek et al., 2004; replication are both inhibited. Morrissey et al., 2005a) and to a lesser extent DNA (Kawabata et al., 1995) in serum by serum nucleases In vivo models occurs, improving serum stability is key in the develop- In vitro experimentation has proven the effectiveness of ment of systemic delivery systems for nucleic acids. RNAi for targeted knockdown of HBV gene expression Morrissey et al. (2005a) used hydrodynamic injection of and indicates that the virus itself is amenable to an infectious HBV clone to establish virus expression in inhibition by RNAi. For use in therapeutic applications, hepatocytes of C57BL/J6 mice. This was followed 72 h however, delivery of siRNAs or RNAi expression later by standard i.v. injection of 20-OH-modified vectors to hepatocytes of the liver in a clinically relevant siRNA duplexes targeting HBV RNAs. The 20-OH- format is a major obstacle. For in vivo delivery, siRNA, modification has been shown to increase siRNA DNA and viral vectors must survive the transit through stability. A reduction of serum HBV DNA levels by the circulatory system, avoid significant uptake by non- 0.90 log10 was observed at the highest dose of the hepatocytes, pass through the small fenestrae (150 nm in modified siRNA, whereas no statistically significant

Oncogene RNAi-mediated prevention and therapy for HCC PR Romano et al 3861 reduction was seen using standard i.v. injection of the this vector to downregulate both viral replication and unmodified siRNAs. These results highlight some of the viral antigenemia (both sAg and eAg) has been difficulties and promise surrounding delivery of RNAi- demonstrated (Figure 3). Experiments in mouse models based drugs to the liver; although less than 1% of the also indicate that this vector is effective in an in vivo administered modified siRNA reached the liver when setting. Stability and hepatocyte cell targeting properties both a high dose and frequent dosing regime were used, are conferred to the plasmid DNA through complexa- the transfected siRNAs still exerted an anti-viral effect. tion with ligand modified spermine molecules. Using the (Morrissey et al., 2005b) later showed that formulating spermine-based delivery system, the prototype HBV the 20-OH modified siRNAs with specialized liposome vector was successfully delivered to mouse hepatocytes compositions resulted in siRNA lipid particles with where downregulation of HBV proteins and RNA was increased serum stability. In these studies** hydrody- achieved at levels comparable to those obtained using namic delivery of an HBV plasmid was used to establish hydrodynamic delivery. The ability of this multi- HBV expression in immunodeficient mice. Six days post- cistronic vector to work effectively and the ability to hydrodynamic injection, the liposome-formulated mod- target desired cell types using appropriately ligand ified HBV-specific siRNA, was administered by i.v. modified spermines provides evidence for a platform injection. A dose dependent reduction of both serum technology that can be used against a variety of viral HBV DNA and HBsAg was observed for up to 7 days pathogens. post i.v. injection. Formulation of the siRNA in a lipid particle allowed the use an B30-fold lower dose compared to the dose used in the previous study and RNA interference as a therapy for chronic hepatitis C achieved a more significant reduction in HBV expression virus infection (Morrissey et al., 2005a). Although these improvements are significant, the extrapolated dose and dosing Hepatitis C virus replication frequency for humans are not clinically relevant. Infection with HCV is a serious public health problem Soutschek et al. (2004) have also demonstrated the worldwide. It is estimated that 2–3% of the world increased stability and efficacy of chemically modified population is chronically infected with HCV. There are siRNAs delivered by i.v. injection. Cholesterol modified approximately six HCV genotypes, whose sequences siRNAs directed against the endogenous apolipoprotein vary as much as 51% between the different strains B (apoB) gene product were able to significantly reduce (Bukh et al., 1995; Simmonds, 1999). Each genotype expression of the apoB gene product and lower blood consists of multiple variations or quasispecies (Enomoto cholesterol levels in a transgenic mouse model. As with et al., 1995). Identifying siRNAs that effectively target the results of Morrissey et al. (2005b) further optimiza- conserved HCV sequences will therefore be key to the tion of these modified siRNAs is required to achieve development of RNAi-based HCV therapeutics. potency and dosing that will make them viable for Genotypes 1a and 1b are the predominant genotypes therapeutic use. found in North and South America and Europe, In an HBV transgenic mouse model mimicking whereas 1b is the predominant HCV genotype in Asia chronic infection, (Uprichard et al., 2005) have shown (Chayama et al., 1995). Hepatitis C virus, a hepacivirus, that pol III driven siRNA delivered to the liver by is a member of the Flaviviridae family. The viral genome recombinant adenovirus suppressed preexisting HBV is a single-stranded, positive-sense RNA of B9.6 kb, gene expression and replication for at least 26 days. that functions as the only viral mRNA encoding a single These results demonstrate the potential of viral vectors polyprotein of B3000 amino acids that is processed into for delivery to the liver and long-term RNAi-mediated structural proteins (core, p7, and the envelope proteins suppression of target messages when expression systems E1 and E2) and nonstructural (NS) proteins (NS2, NS3, are used to produce intracellular siRNAs. This long- NS4A and B, and NS5A and B) (Figure 4) (reviewed in term expression of siRNA is not only applicable for Bartenschlager and Lohmann, 2000). A negative strand therapeutic use but may also enable the use of expressed replicative intermediate is also produced in cells at lower RNAi for prophylaxis against virus infection when levels than the plus strand RNA. This means that there administered before viral exposure. is only a single plus strand species that can be targeted The advantages presented by using a vectored by RNAi. approach to express siRNA have led Nucleonics to The in vitro study of HCV has relied on the use of develop multi-cistronic plasmids in which a number of HCV subgenomic (Figure 5) and genomic replicons that distinct siRNAs are produced from a single vector. replicate autonomously in human heptatoma cells Expressing multiple HBV-specific shRNAs from a single (Huh7 cells) (Lohmann et al., 1999; Blight et al., plasmid allows effective siRNAs to be produced in the 2002). Commonly used genomic and subgennomic cell at a higher and more uniform dose level over longer HCV replicons are bicistronic constructs, where the periods, and thus provides insurance against potential HCV 50-UTR-IRES (internal ribosome entry site) escape mutants. A prototype HBV vector (containing directs expression of the neomycin phosphotranferase four HBV-specific siRNAs) has been extensively tested gene (neo) followed by the encephalomyocarditis virus in a cell culture system in which an HBV replicon has (ECMV) IRES directing expression of HCV gene been reproducibly inhibited with a calculated IC50 in products and terminating with the HCV 30-UTR. the sub-femtomolar range. As predicted, the ability of Inclusion of the neo gene allows for selection of

Oncogene RNAi-mediated prevention and therapy for HCC PR Romano et al 3862 a shRNA 4 KanR Promoter A

Promoter D shRNA 1

shRNA 3 Promoter B Promoter C Ori shRNA 2

b Replicative c Intermediates Large and middle sAgs

3TC 10 ng 5 ng 1 ng 0.5 ng 0.1 ng Neg Ctrl

10 5 2.51.0 0.5 0.25 0.1 3TC rcDNA dlDNA L

ssDNA M

Figure 3 Plasmid DNA expressing multiple shRNAs specific for HBV is able to decrease viral DNA and surface Ag (sAg) expression. (a) Multi-cistronic plasmid expressing HBV-specific shRNAs. The plasmid contains four promoters driving four unique shRNAs (represented by hairpin loops) targeting different sequences of HBV. Arrows indicate the direction of transcription. KanR, kanamycin resistance gene; Ori, E. coli origin of replication. (b) Plasmid-expressed shRNAs decrease HBV replicative intermediate DNA. Human heptatoma cells (Huh7 cells) were co-transfected with various doses of the shRNA expression vector (0–10 ng) and 500 ng HBV replicon. HBV replicative intermediate DNA was extracted 5 days post-transfection and viral DNA subjected to gel electrophoresis, blotted and probed with P32-labeled unit-length HBV DNA. HBV polymerase inhibitor 3TC, at 50 mM, was used as positive control for replication inhibition. rcDNA, relaxed circular DNA; dlDNA, double linear DNA; ssDNA, single strand DNA. (c) Plasmid- expressed shRNAs decrease HBV Large and Middle sAg expression. A duplicate set of transfected Huh7 cells (as described in a) was lysed and subjected to Western blot analysis. Viral Large (L) and Middle (M) surface proteins were probed using rabbit anti-HBV PreS2 , followed by incubation with goat anti-rabbit IgG antibody conjugated with horseradish peroxidase. Bands were visualized using chemiluminescence. Because of glycosylation, the HBV Large protein shows two bands of 42 and 39 kd., whereas the Middle protein exhibits a triplets of 36, 33 and 30 kd.

Structural Proteins Nonstructural Proteins

core E1 E2p7 NS2 NS3 NS4A NS4B NS5A NS5B

Figure 4 Hepatitis C virus (HCV) genome organization. The 9.6 kb positive strand RNA genome is translated as a 3000 amino acid polyprotein that is processed into 10 independent gene products. The amino-terminal region encodes virion structural proteins whereas the carboxy-terminal region encodes the nonstructural proteins.

transformants and stable propagation of replicon- RNA-dependent RNA polymerase (NS5B) with a high- containing cell lines. error rate. This low fidelity replication machinery results in the production of virus ‘quasispecies’, which contain Challenges to the development of RNA interference-based an average of approximately one nucleotide change per therapeutics replication cycle (Lohmann et al., 2000). Successful anti- The variability of the HCV genome presents significant HCV therapeutics must have the inherent ability to see challenges to the design of therapeutics that effectively and inhibit replication of viral variants and prevent the suppress HCV infection. Hepatitis C virus encodes a generation of virus escape mutants. As the efficacy of an

Oncogene RNAi-mediated prevention and therapy for HCC PR Romano et al 3863 5’ UTR 3’ UTR mids expressing pol III driven sense and antisense transcripts against NS5B, inhibition of HCV RNA replication was maintained for up to 3 weeks. Interest- EMCV neo NS3 4A 4B NS5A NS5B ingly, synthetic siRNAs against NS5B inhibited ( þ )and IRES (À) strand HCV RNA replication, indicating that both the genome strand and the anti-genome strand may be HCV IRES targets for RNAi and/or that targeting one of the Figure 5 Schematic representation of an hepatitis C virus strands has a downstream negative effect on production subgenomic replicon. Additional subgenomic replicons have been made that include the NS2 nonstructural gene. of the other strand. Prabhu et al. (2005) also showed siRNA mediated inhibition of both ( þ )and(À) strand HCV RNA siRNA for a specific target is directly related to sequence synthesis. High level expression of HCV RNA was homology, the use of RNAi therapy to combat HCV achieved using T7 polymerase driven expression of a infection will need to incorporate sequence design full-length HCV cDNA clone. Polymerase III driven paradigms which can circumvent the plasticity of the expression of shRNAs directed against the E2, NS3 and HCV genome. NS5B regions of the HCV genome inhibited gene Research into anti-HCV drug development has been expression as well as production of both the genome hampered by the lack of robust cell culture infection and antigenome strands of HCV. systems and accessible animal models to test drug Wilson and Richardson (2005) have shown that HCV efficacy. A major step towards studying HCV has replicon escape mutants can emerge after multiple recently been achieved with the development of stable siRNA treatments. An 87% reduction of HCV replicon HCV cell culture models, which permit robust produc- replication and gene expression was previously shown tion of infectious HCV particles (Cai et al., 2005; (Wilson et al., 2003) when replicon expressing cells were Lindenbach et al., 2005; Wakita et al., 2005; Zhong transfected with an siRNA targeting the NS5B gene. et al., 2005). These new reagents will aid in the study and Mutations, speculated to have occurred in the HCV design of anti-HCV therapies. Previous HCV work RNA, were thought to be selected for after siRNA using RNAi has focused on the use of HCV-reporter treatment. In a follow-up study, HCV escape mutations fusion systems, and subgenomic and genomic HCV were shown to have indeed occurred. After one replicon models (Lohmann et al., 1999; Blight et al., treatment with the anti-NS5B siRNA, HCV genomes 2000, 2002; Kro¨ nke et al., 2004). Using an HCV NS5B were identified that contained mutations within the gene target sequence fused to a luciferase reporter, siRNA target sequence. A single mutation in the target McCaffrey et al. (2002) used co-hydrodynamic injection sequence had a minimal effect on the efficacy of the of the reporter fusion along with anti-NS5B siRNAs or siRNA, leading to a 4 or 13% loss of activity. After five polIII produced anti-NS5B shRNAs to show knock- consecutive siRNA treatments, genomes were identified down of luciferase expression in mice, demonstrating that contained multiple point mutations within the in vivo functionality of the anti-HCV siRNA. Other siRNA target sequence, and the addition of a second groups have used the sub-genomic replicon system and point mutation within the target sequence led to a demonstrated siRNA efficacy against established repli- 70–80% loss of siRNA activity. This occurred when two cating HCV RNA. Kapadia et al. (2003) showed individual siRNAs targeting different regions within significant inhibition of HCV RNA and protein expres- NS5B were used. Interestingly, when the two anti-NS5B sion with siRNAs targeting NS3 and NS5B. This siRNAs were co-transfected simultaneously into repli- downregulation was independent of an interferon con containing cells, resistance was barely detectable. response. Yokota et al. (2003) used a similar system to These results indicate that simultaneous targeting of test the efficacy of siRNA sequences targeted to the multiple HCV sequences may be one method to avoid HCV 50-UTR which is one of the most conserved the generation of escape mutations in this highly regions of the HCV genome. They also designed pol III mutable virus. Another observation made from these driven expression constructs producing shRNA derived studies was that siRNAs co-transfected with HCV siRNAs and constructs producing sense and anti-sense replicon RNA had a greater efficacy (97–99% inhibi- RNA. Sequence specific effectiveness was found indicat- tion) than siRNAs transfected into cells stably expres- ing that secondary structure within the 50-UTR target sing the HCV replicon (90% inhibition), suggesting that impacts RNAi efficacy, and that plasmids expressing RNAi may be useful as a prophylactic to protect shRNA work better than plasmids producing sense and uninfected cells from HCV infection. anti-sense transcripts. Using a full length replicon In addition to targeting viral RNAs, it may be system, Randall et al. (2003) found that siRNAs possible to inhibit HCV replication by targeting the directed to the 50-UTR region cleared replicating HCV expression of cellular proteins needed for HCV replica- RNA from >98% of the cell culture to below detectable tion. Several cellular proteins have been identified as co- levels. This effect was dose dependent and specific in factors that effect HCV replication by interacting with that nucleotide changes within the target siRNA the 50 and 30 untranslated regions of the HCV genome. sequence abolished the effect. Wilson et al. (2003) Korf et al. (2005) have shown that pol III driven showed that long term suppression of HCV RNA shRNAs targeted to either the HCV-50-UTR or the replication could be established. Using episomal plas- 30-UTR are efficient at downregulating HCV replicon

Oncogene RNAi-mediated prevention and therapy for HCC PR Romano et al 3864 replication. Additionally, the cellular co-factors, proteo- obstacles to delivery of the effector molecule and the some a-subunit 7 (PSMA7), which modulates HCV- efficacy with which it acts. IRES (Kruger et al., 2001) activity in cell culture, and Because of the limited therapeutic options available HU antigen R (HuR), which interacts with the HCV 30- for patients with HCC, numerous groups have searched UTR (Spangberg et al., 2000), were targeted by pol III for molecules that are upregulated in HCC cells and driven shRNAs in replicon expressing cells. Short may provide attractive therapeutic targets. Comparison interfering RNA knock-down of these endogenous gene of the transciptome and proteome of HCC cells products effectively inhibited HCV replicon replication with healthy liver cells has indeed indicated that a in Huh7 cells. Furthermore, combinations of siRNAs variety of messages and proteins are present to a greater against PMSA7, HuR along with anti-HCV siRNAs degree in HCC cells (Blanc et al., 2005; Mao et al., 2005; exhibited additive effects leading to enhanced decreases Wong et al., 2005). However, the particular transcripts in HCV RNA levels. It remains to been seen whether and/or proteins that are increased sometimes vary there are any deleterious effects associated with down- among the individual studies. Not surprisingly, growth regulation of these host cell proteins. These results, factors such as epidermal growth factor (EGF) and its however, along with the work of Wilson and Richard- receptor (EGFR) as well as angiogenic factors such as son (2005) indicate that combination siRNA therapies vascular endothelial growth factor (VEGF) are often may be an effective way to increase the efficacy of found in increased amounts in HCC cells (Suzuki et al., therapeutic siRNAs and decrease the potential of 1996; Miura et al., 1997; Hisaka et al., 1999). An generating drug resistant mutations. increase in the activity of the mitogen activated protein kinase (MAPK) pathway is also seen in a large Endogenous RNA interference targets for treatment of percentage of HCC cells (Schmidt et al., 1997; Ito Hepatocellular carcinoma. The targeting of HBV and et al., 1998). HCV transcripts provides one avenue by which HCC Therefore, RNAi is an obvious candidate for the can be treated. However, the potential for reducing or treatment of patients with HCC through the down- eliminating the intracellular expression of selected genes regulation of proteins associated with malignant cells. The and proteins using RNAi also makes this technology data of Song et al., in which siRNAs targeting the Fas attractive for the treatment of endogenous targets within gene decreased hepatitis in mice, suggest that this HCC cells. Whereas many of the same challenges will be approach may indeed be applicable in an in vivo setting associated with the knock-down of endogenous host (Song et al., 2003). Unfortunately, the variable nature of genes as with viral genes, additional questions will need HCC may make it difficult to identify a single protein that to be addressed before RNAi can be used to attack HCC will be responsible for all (or even most) cases of HCC. It via the reduction of particular endogenous mRNAs. may be necessary to deliver a number of different effector The complex nature of HCC has made the assignment molecules to cover as many potentially causative tran- of genes associated with this disease difficult to general- scripts as possible. This may be accomplished through a ize. As with other types of malignancy, the stage of cocktail of siRNAs or through the use of a multi-cistronic disease will influence the panoply of genes that are up or plasmid expressing a number of shRNAs. Additional down regulated at any given time. The typical patient studies that more clearly identify these endogenous targets presents at a late stage that may provide additional will aid in the design of these effectors.

References

Bartenschlager R, Lohmann V. (2000). J Gen Virol 81: Gregory RI, Chendrimada TP, Cooch N, Shiekhattar R. 1631–1648. (2005). Cell 123:6313–6640. Blanc JF, Lalanne C, Plomion C, Schmitter JM, Bathany K, Guo Y, Guo H, Zhang L, Xie H, Zhao X, Wang F et al. Gion JM et al. (2005). Proteomics 5:3778–3789. (2005). J Virol 79:14392–14403. Blight KJ, Kolykhalov AA, Rice CM. (2000). Science 290: Hamasaki K, Nakao K, Matsumoto K, Ichikawa T, Ishikawa 1972–1974. H et al. (2003). FEBS Lett 543:51–54. Blight KJ, McKeating JA, Rice CM. (2002). J Virol 76: Hamilton AJ, Baulcombe DC. (1999). Science 286:950–952. 13001–13014. Hammond SM, Boettcher S, Caudy AA, Kobayashi R, Bukh J, Miller RH, Purcell RH. (1995). Semin Liver Dis 15: Hannon GJ. (2001). Science 293:1146–1150. 41–63. Hisaka T, Yano H, Haramaki M, Utsunomiya I, Kojiro M. Cai Z, Zhang C, Chang KS, Jiang J, Ahn BC, Wakita T et al. (1999). Int J Oncol 14:453–460. (2005). J Virol 79:13963–13973. Huang ZM, Yen TS. (1993). J Virol 67:7032–7040. Chayama K, Tsubota A, Arase Y, Saitoh S, Ikeda K, Ito Y, Sasaki Y, Horimoto M, Wada S, Tanaka Y, Kasahara Matsumoto T et al. (1995). J Hepatol 23:648–653. A et al. (1998). Hepatology 27:951–958. Chen M, Sallberg M, Hughes J, Jones J, Guidotti LG, Chisari Kapadia SB, Brideau-Andersen A, Chisari FV. (2003). Proc FV et al. (2005). J Virol 79:3016–3027. Natl Acad Sci USA 100:2014–2018. Enomoto N, Sakuma I, Asahina Y, Kurosaki M, Murakami Kawabata K, Takakura Y, Hashida M. (1995). Pharm Res 12: T, Yamamoto C et al. (1995). J Clin Invest 96:224–230. 825–830. Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, Mello Klein C, Bock CT, Wedemeyer H, Wustefeld T, Locarnini S, CC. (1998). Nature 391:806–811. Dienes HP et al. (2003). Gastroenterology 125:9–18. Giladi H, Ketzinel-Gilad M, Rivkin L, Felig Y, Nussbaum O, Korf M, Jarczak D, Beger C, Manns MP, Kruger M. (2005). Galun E. (2003). Mol Ther 8:769–776. J Hepatol 43:225–234.

Oncogene RNAi-mediated prevention and therapy for HCC PR Romano et al 3865 Kro¨ nke J, Kittler R, Buchholz F, Windisch MP, Peitschmann Ren GL, Bai XF, Zhang Y, Chen HM, Huang CX, Wang PZ T, Bartenschlager R et al. (2004). J Virol 78:3436–3446. et al. (2005). Biochem Biophys Res Commun 335:1051–1059. Kruger M, Berger C, Welch PJ, Barber JR, Manns MP, Wong- Schmidt CM, McKillop JH, Cahill PA, Sitzmann JV. (1997). Staal F. (2001). Mol Cell Biol 21:8357–8364. Biochem Biophys Res Commun 236:54–58. Lau DT, Khokhar MF, Doo E, Ghany MG, Herion D, Park Shlomai A, Shaul Y. (2003). Hepatology 37:764–770. Y et al. (2000). Hepatology 32:828–834. Simmonds P. (1999). J Hepatology 31:54–60. Lau NC, Lim LP, Weinstein EG, Bartel DP. (2001). Science Song E, Lee SK, Wang J, Ince N, Ouyang N, Min J et al. 294:858–862. (2003). Nat Med 9:347–351. Leonard JN, Schaffer DV. (2005). Gene Ther 13:532–540. Soutschek J, Akinc A, Bramlage B, Charisse K, Constien R, Lindenbach BD, Evans MJ, Syder AJ, Wolk B, Tellinghuisen Donoghue M et al. (2004). Nature 432:173–178. TL, Liu CC et al. (2005). Science 309:623–626. Spangberg K, Wiklund L, Schwartz S. (2000). Virology 274: Lohmann V, Korner F, Koch J, Herian U, Theilmann L, 378–390. Bartenschlager R. (1999). Science 285:110–113. Suzuki K, Hayashi N, Miyamoto Y, Yanamoto M, Ohkawa Lohmann V, Roos A, Korner F, Koch JO, Bartenschlager R. K, Ito Y et al. (1996). Cancer Res 56:3004–3009. (2000). J Viral Hepatol 7:167–174. Uprichard SL, Boyd B, Althage A, Chisari FV. (2005). Proc Lui F, Song Y, Lui D. (1999). Gene Ther 6:1258–1266. Natl Acad Sci USA 102:773–778. Mao HJ, Li HN, Zhou ZM, Zhao JL, Wan DF. (2005). World Wakita T, Pietschmann T, Kato T, Date T, Miyamoto M, J Gastroenterol 11:2811–2816. Zhao Z et al. (2005). Nat Med 11:791–796. Matranga C, Tomari Y, Shin C, Bartel D, Zamore PD. (2005). Wilson JA, Richardson CD. (2005). J Virol 79:7050–7058. Cell 123:607–620. Wilson JA, Jayasena S, Khvorova A, Sabatinos S, Rodrigue- McCaffrey AP, Meuse L, Pham TT, Conklin DS, Hannon GJ, Gervais IG, Arya S et al. (2003). Proc Natl Acad Sci USA Kay MA. (2002). Nature 418:38–39. 100:2783–2788. McCaffrey AP, Nakai H, Pandey K, Huang Z, Salazar FH, Wong N, Chan KYY, Macgregor PF, Lai PBS, Squire JA, Xu H et al. (2003). Nat Biotechnol 21:634–639. Beheshti B et al. (2005). Clin Cancer Res 11:1319–1326. Mendy ME, Fortuin M, Hall AJ, Jack AD, Whittle HC. Wu HL, Huang LR, Huang CC, Lai HL, Liu CJ, Huang YT (1999). Br J Biomed Sci 56:34–38. et al. (2005). Gastroenterology 128:708–716. Miura H, Miyazaki T, Kuroda M, Oka T, Machinami R, Ying C, De Clercq E, Neyts J. (2003). Biochem Biophys Res Kodama T et al. (1997). J Hepatol 27:854–861. Commun 309:482–484. Morrissey DV, Blanchard K, Shaw L, Jensen K, Lockridge Yang PL, Althage A, Chung J, Chisari FV. (2002). Proc Natl JA, Dickenson B et al. (2005a). Hepatology 41:1349–1356. Acad Sci USA 99:12825–13830. Morrissey DV, Lockridge JA, Shaw L, Blanchard K, Jensen Yokota T, Sakamoto N, Enomoto N, Tanabe Y, Miyagishi M, K, Breen W et al. (2005b). Nat Biotechnol 23:1002–1007. Maekawa S et al. (2003). EMBO Rep 4:602–608. Papatheodoridis GV, Dimou E, Papadimitropoulos V. (2002). Zamore PD, Tuschl T, Sharp PA, Bartel DP. (2000). Cell 101: Am J Gastroenterology 97:1618–1628. 25–33. Prabhu R, Vittal P, Yin Q, Flemington E, Garry R, Zhang G, Gao X, Song YK, Vollmer R, Stolz DB, Robichaux WH et al. (2005). J Med Virol 76:511–519. Gasiorowski JZ et al. (2004a). Gene Ther 11:675–682. Radhakrishnan SK, Layden TJ, Gartel AL. (2004). Virology Zhang XN, Xiong W, Wang JD, Hu YW, Xiang L, Yuan ZH. 323:173–181. (2004b). World J Gastroenterol 10:2967–2971. Rand TA, Petersen S, Du F, Wang X. (2005). Cell 123:621–629. Zhong J, Gastaminza P, Cheng G, Kapakia S, Kato T, Randall G, Grakoui A, Rice CM. (2003). Proc Natl Acad Sci Burton DR et al. (2005). Proc Natl Acad Sci USA 102: USA 100:235–240. 9294–9299.

Oncogene