Trojan horse at cellular level for tumor gen

Guillaume Collet a, Catherine Grillon a,, Mahdi Nadima Centre de Biophysique Molculaire, UPR4301 CNRS, Rue Charles Sadron, 45071, Orlans, cedeb Libragen-Induchem Company, 3, rue des satellites, Bat. Canal Biotech, 31400, Toulouse, Franc

s denst onrapatioomadled

1. Introduction

roblempublished by the Internationalthe Wotinue, tcases in

ogressiefcieovativtumoresign o

The main goal in anti-cancer approaches is to maximize efcacy ofcancer treatments and to minimize systemic toxicity.

on tumor microenvironment in addition to tumor cells themselves.

Gene 525 (2013) 208216

Contents lists available at SciVerse ScienceDirect

Gen

j ourna l homepage: www.eapproaches of gene therapies, the image of the trojan horse is largelyused by strategies that combine, in the same engineered entity, atargeting unit and a specic drug/gene delivery system. Inspired bythe GreekMythology, the trojan horse was the source for the presentedapproaches and revisited for numerous applications. The trojan horseis an engineered specic tool in form of: a liposome, an exosome, aspecialized cell or a modied virus in order to specically reach thetumor site (Fig. 1). Hidden Odysseus's army symbolized the various

Abbreviations: AAV, adeno-associated viruses; Ad, adenovirus; CTGVT, cancertargeting gene-viro-therapy; EGFR, epidermal growth factor receptor; EPCs, endotheli-al precursor cells; GAOVT, gene armed oncolytic virus therapy; GCV, ganciclovir; HRE,hypoxia responsive element; HSV-TK, herpes simplex virus thymidine kinase; MDR1,multi drug resistance protein; MSCs, mesenchymal stem cells; NK, natural killers;OVs, oncolytic viruses; PAMAM, highly branched polyamidoamine; PEG, polyethyleneglycol; PEI, poly(ethylenimine); PLL, poly(L-lysine); PU-PEI, polyurethane-short

branch polyethylenimine; QD, quantum dots; SPION,nanoparticles; TERT, telomerase reverse transcriptase;TK, thymidine kinase; TRAIL, tumor necrosis factor-relaVEGF, vascular endothelial growth factor. Corresponding author at: Centre de Biophysique Mo

Charles Sadron, 45071 Orlans cedex 2, France. Tel.: +325 54 59.

E-mail addresses: guillaume.collet@cnrs-orleans.frcatherine.grillon@cnrs-orleans.fr (C. Grillon), mahdi.nadimclaudine.kieda@cnrs-orleans.fr (C. Kieda).

0378-1119/$ see front matter 2013 Elsevier B.V. Alhttp://dx.doi.org/10.1016/j.gene.2013.03.057f new therapies aimingduring the past decades.

However the main obstacles to tissue and cell-specic gene deliverystill remains. Biotechnology allows new strategies that improve andsucceed to change the means of cancer treatments. Among elaboratedto overtake the limitations faced by researchdeath worldwide. Statistical analysisAgency for Research on Cancer fromreveals that if the estimated trends concases will raise from 12.7 million new2030 (Bray et al., 2012).

Facing such an alarming disease prhave been developed but expectedreached. Thus the need to develop inn

Progress in the knowledge aboutaspects of cancer has facilitated the drld Health Organizationhe incidence of all cancer2008 to 21.2 million by

on, varieties of therapiesncy has still not beene strategies is crucial.biology and molecular

This had led for example to the development of newmethods for deliveryof therapies by targeting the tumor-associated vasculature, providingthus promising antitumor effects with minimal systemic toxicity.

Gene therapy was born 50 years ago thanks to Dr. W. Szybalski andDr. E. Szybalska's pioneer experiments (Szybalska and Szybalski, 1962)reporting the rst gene transfer to mammalian cells. This eld has beensteadily developing and fast growing keeping its main focus on cancer.Moreover, a large number of anti-tumor strategies have been described.Cancer remains a global health p and a major cause of Remaining a challenge, new approaches are now developed focusinga b s t r a c ta r t i c l e i n f o

Article history:Accepted 7 March 2013Available online 28 March 2013

Keywords:Nucleic acid deliveryExosomeLiposomeVirusProgenitor cellTumor targeting

Among innovative strategietheir well-known limitatiotherapy is highly dependenhorse concept to carry thetumor site. Recent investiglimitations bringing new prThis review describes recentadvantage of available knowsuperparamagnetic iron oxideTHL, Trojan Horse Liposome;ted apoptosis-inducing ligand;

lculaire, UPR4301 CNRS, Rue3 2 38 25 78 04; fax: +33 2 38

(G. Collet),@cnrs-orleans.fr (M. Nadim),

l rights reserved.e therapiesa,b, Claudine Kieda a

x 2, Francee

veloped for cancer treatments, gene therapies stand of great interest despitein targeting, delivery, toxicity or stability. The success of any given gene-the carrier efciency. New approaches are often revisiting the mythic trojan

eutic nucleic acid, i.e. DNAs, RNAs or small interfering RNAs, to pathologicns are focusing on engineering carrying modalities to overtake the aboveise to cancer patients.vances and perspectives for gene therapies devoted to tumor treatment, takingge in biotechnology and medicine.

2013 Elsevier B.V. All rights reserved.

e

l sev ie r .com/ locate /genetherapeutic genes or interfering RNAs that are supposed to accuratelytarget the gene of interest.

In gene therapy approaches, the nal protein level modulationis obtained by exogenous DNA or mRNA delivery on the one hand,giving rise to specic protein expression. On the other hand, smallRNAs (siRNAs, miRNAs) can be used to inhibit or modulate endogenousprotein translation. To be efcient, these molecules have to resist thedegradation in the circulation and further reach either cytoplasm (for

y. Sd veliumncerplexNAwind foct h

209G. Collet et al. / Gene 525 (2013) 208216tumor cells

Blood vessel

healthy cells

Customized particles

engineered virus

A B

endotheliumphysiological

endotheliumpathological

(i,e, polyplex, liposome,exosomes)

Fig. 1. An overview of various trojan horse strategies developed for cancer gene therapgene delivery to cancer cells. This double side scheme presents, in the upper part, a blooThe lower part refers to a blood vessel in a pathological context of cancer where endotheSuch leaky and chaotic tumor vessels are supposed to give an access for therapy to cahorses approaches have been developed. Numerous customized molecules such as polythe tumor area when injected in the blood stream, carrying either a DNA or interfering Ractivity. Homing cells like EPCs or MSCs (C) are used to target neoangiogenic sites. Follothey will express the transgene, then acting on tumor cells. Targeted cells can be usetherapies to cancer cells, these trojan horses combine regulatory safety locks to proteRNA) or nucleus (for DNA) as well as escape endosomal degradation. Asthese active molecules are not predisposed to overcome physiologicalbarriers to be delivered to the target cells, carriers are needed.Indeed, the latter should be able to facilitate nucleic acid stability in thecirculation, intracellular delivery and, when required, import into thenucleus (Wang et al., 2012).

This review describes various approaches elaborated for tumortreatment addressed to the cell and sub-cellular levels and using thetrojan horse tricky concept.

2. When the horse becomes a molecule

Non-viral gene delivery systems amongwhich are cationic polymerssuch as poly(ethylenimine) (PEI)/poly(L-lysine) (PLL), dendrimers, car-bon nanotubes, Superparamagnetic Iron Oxide Nanoparticles (SPION)or Quantum Dots (QD) are usually positively charged allowingcompression of anionic nucleic acids in solution and interaction withnegatively charged cell membranes. Thus, they may act as moleculartrojan horses to allow gene delivery.

The most widely used are cationic polymers, PLL and PEI, thatcombine with DNA into particulate complex, called polyplex whichenable the gene transfer into cells.

PEI allows polyplexes to efciently escape the degradation with-in endosomes (Kichler et al., 2001). Polyurethane-short branchpolyethylenimine (PU-PEI) or PEI are myristilated to help targetingbrain tumor sites. They are used as a therapeutic-delivery vehicle inthe treatment of glioblastoma either by delivering tumor necrosisfactor-related apoptosis-inducing ligand (TRAIL) in intracranial U87glioblastoma-bearing mice (Li et al., 2011) or by delivering miRNAs(miR145) both in vitro in human glioblastoma-associated cancer stemlike cells (CD 133+) and in vivo in an orthotopic model induced bytransplantation of the same cells in immunocompromised mice (Yanget al., 2012). Such polyplexes can be substituted by polyethylene glycolcarrier cell-based delivery of

virusvector armed targeting cells

DC

(i.e. EPCs)

chematic representation of approaches revisiting the trojan horse strategy for specicssel in a physiological context, harboring a continuous and well organized endothelium.is disorganized, with tumor cells taking place in the vessel wall with endothelial cells.cells which can be targeted as well as tumor endothelium. In this aim, various trojanes, liposomes or even exosomes (A) can be used and derivatized to reach preferentially(siRNA, miRNA). Viruses (B) can also be good carriers and bring an additional oncolyticg systemic application, engineered cells are recruited to the tumor environment wherer oncolytic viral particles production (D) upon recruitment in the tumor. To addressealthy cells. (Figure produced using Servier Medical Art).(PEG) to enhance their stability. PLL alone has poor transfection ability(Pouton et al., 1998). However PEG coating can increase both trans-fection efcacy and circulation half-life (Lee et al., 2002). Liu et al.in a recent study showed that PLL when associated with PEG andpoly(lactic-co-glycolic acid) gets highly efcient to deliver adriamycinand siRNA into hepatic carcinoma cells (Liu et al., 2012a). Similarly,poly(lactic-co-glycolic acid) substituted-nanoparticles, having encap-sulated an anti-miRNA to inhibit miR-155 activity, were shown toefciently slow down the growth of pre-B-cell tumors (Babar et al.,2012). To improve the therapeutic efcacy of these tools, physical/chemical technics are utilized to modify them. Chen et al. (2012)linked PEI-PEG to superparamagnetic iron oxide nanoparticles and asingle-chain variable fragment CD44v6 (scFvCD44v6-PEG-g-PEI-SPION).Such engineered vectors act as a cancer-targeting as well as magneticresonance detectable nanocarrier. Tested in siRNA delivery, these com-plexes provide both imaging and therapeutic modalities. PEI-basedcomplexes could also be combined with physical tools as sonoporationto enhance the transfer of therapeutic molecules as shown for miRNAs(Chen et al., 2011).

Besides these synthetic macromolecules, natural cationic polymerscan also deliver foreign genetic materials into cancer cells. Chitosan, abiodegradable and biocompatible linear aminopolysaccharide is awell-established vector for DNA delivery (Rudzinski and Aminabhavi,2010). Chitosan and its derivatives are also adequate to deliver RNA asshown in siRNA-mediated silencing of gene expression in breast cancercells (Pille et al., 2006).

Cationic lipids are also largely used as gene delivery systems(Miele et al., 2012). Due to their positive charge, they naturallymake complexes with DNA to form lipoplexes, protecting nucleicacid from degradation and facilitating their delivery inside cells.Novel classes of cationic lipids are currently synthetized, such ascyclen-based cationic lipids which allow genetic material transferinto different tumor cell lines (Huang et al., 2011a). Interestingly,

210 G. Collet et al. / Gene 525 (2013) 208216cationic lipids modied with dexamethasone to target glucocorticoidreceptor-bearing cells and carrying the tumor suppressor gene p53induce apoptosis and regression of tumor growth (Mukherjee et al.,2009).

Numerous nanoparticles are being developed depending on size,shape, charge, and surfactant coating to facilitate their uptake by cells.These functionalized nanoparticles are for example fullerenes and car-bon nanotubes, polymericmicelles, polymeric nanospheres, dendrimers,polymer-coated nanocrystals, nanoshells, SPION and QDs.

Single-walled carbon nanotubes carrying siRNA were shownefcient in carrying, releasing and delivering siRNA in vitro into mam-malian HeLa cells (Kam et al., 2005). A study performed by Zhanget al. shows the silencing of telomerase reverse transcriptase (TERT) ex-pression, a key enzyme for the stabilization of chromosomes. Releasedfrom nanotube sidewalls, TERT siRNA inhibited the correspondinggene expression in various tumor types both in vitro and in vivo(Zhang et al., 2006).

Highly branched polyamidoamine (PAMAM) dendrimers are a newclass of polymers in spherical conformation and soluble in aqueous solu-tion. Conjugated with Angiopep-2, which target low-density lipoproteinreceptor-related protein-1 expressed on brain capillary endothelial andglial cells, these particles cross the blood-brain barrier, carrying the ther-apeutic DNA to glial tumor cells (Huang et al., 2011b). In addition,Taratula et al. demonstrated the ability of SPION-complexed dendrimersto deliver siRNA in cancer cells (Taratula et al., 2011).

Recently, quantum dots functionalized by -cyclodextrin coupledto amino acids were designed to facilitate the delivery of siRNA.Silencing the multi-drug resistance 1 gene, these siRNA-carring QDsare promising vehicles for nucleic acid delivery as they were shownto reverse the multidrug resistance in the case of HeLa cells (Liet al., 2012).

Designed microbubbles can specically deliver nucleic acids totumor sites when exposed to ultrasound. Indeed, when loadedwith an active molecule such as nucleic acid, ultrasounds triggermicrobubbles destruction and release of their content. Using thismethod, Carson et al. reported the efcacy of epidermal growth factorreceptor (EGFR)-directed siRNA in reducing squamous cell carcinomagrowth (Carson et al., 2012). Similarly, customized microbubbleswere used to reduce drug resistance of cancer cells by deliveringmulti drug resistance protein 1 (MDR1) siRNA which led to the de-crease of P-glycoprotein activity. As inhibition of MDR1 also increasescancer cell sensitivity to drugs, a strategy combining gene silencingand chemotherapy could be proposed for cancer treatments takingadvantage of this synergistic effect (He et al., 2011).

3. When the horse becomes a liposome

Selective delivery systems are still lacking. Ideally, they shouldspecically target the disease niche and allow an efcient transfer ofthe therapeutic nucleic acid inside the targeted cells. Such strategiesare designed in the purpose of avoiding the toxicity due to thesystemic distribution of injected molecules.

In this respect liposomes as vectors brought promise (Bangham,1995). In fact, liposomes, vesicles with an aqueous compartmentlimited by a phospholipid bilayer, can protect nucleic acids fromenzymatic degradation. According to their lipid composition theycan allow delivery of nucleic acids into cells by interacting with thenegatively charged cell membrane.

Liposomes were rst used for protein and drug delivery(Gregoriadis and Ryman, 1971) and more than ten years later forgene transfer (Cline, 1985; Soriano et al., 1983). Neutral liposomesthe production of which is easy were studied in anticancer therapiesfor in vitro and in vivo delivery of nucleic acids because of their lowtoxicity and immunogenicity (Halder et al., 2006; Landen et al.,2005). However, due to their lack of surface charges, they led to low

transfection efcacy. Therefore, cationic liposomes were developedand optimized, improving entry into cells and protection of nucleicacid against degradation (Spagnou et al., 2004). The Trojan HorseLiposome (THL) Technology was largely developed for non-viralgene transfer and is based on pegylated cationic liposomes (Boado,2007). It is suitable for brain targeting as opposed to conventional de-livery systems. Indeed, viruses or cells do not cross the blood-brainbarrier which is only permeable to lipophilic molecules smaller than400 Da (Patel et al., 2009). Furthermore, encapsulated moleculescan be targeted to specic cells if the liposomes are substituted ontheir outer membrane by either monoclonal antibodies (against insu-lin or transferrin receptors, for example) or other specic molecules,leading to so-called immunoliposomes. A successful therapeuticincrease of the brain beta-glucuronidase activity was obtained byadministration of non-viral plasmid DNA encoding lysosomal enzymeencapsulated in liposomes bearing monoclonal antibody to mousetransferrin receptor (Zhang et al., 2008).

This approach has been further improved by limiting gene expres-sion to specic regions of the brain by a specic promoter. Xia et al.reported a study on rats with experimental Parkinson's diseaseand treated with liposomes containing a plasmid DNA coding for theglia-derived neurotrophic factor and targeted with a monoclonal anti-body to the rat transferrin receptor. The expression of the transgenewas conned to catecholaminergic cells which was achieved by condi-tioning the gene expression to the rat tyrosine hydroxylase promoter(Xia et al., 2008). This intravenous non-viral therapy displayed almostno toxic side effect in rat (Zhang and Pardridge, 2009).

Such THLs can also be used for small RNA delivery (Boado, 2007;Pardridge, 2010). Zhang et al. reported a 90% increase in survival timeof mice with intracranial brain tumors following weekly treatmentwith monoclonal antibodies-targeted THLs that encapsulated anexpression plasmid encoding RNAi knocking down the EGFR (Zhanget al., 2004). Recently, liposomes were shown to be efcient vectorsfor sens/antisens miRNA able to regulate protein expression. This wasillustrated by the targeting of liposome-encapsulated anti-miR-296 toendothelial cells, inhibiting v3 integrin expression and leading toa decrease of angiogenesis both in vitro and in vivo (Liu et al., 2011).An additional example is given bymiR-7-containing cationic liposomeswhich were shown to be effective in suppressing EGFR expressionin vitro and in reducing tumor volume in vivo in a mouse xenograftmodel (Rai et al., 2011).

4. When the horse becomes an exosome

Exosomes can be described as naturally cell-produced liposomes.These small-sized vesicles produced by various cells are composed bya lipid bilayer membrane and contain proteins, nucleic acids, miRNAsin a composition depending on the cell type (Fevrier and Raposo,2004; Thery et al., 2002). They are natural mediators allowing specictransfer of molecules between cells, independently of cell contact.Exosomes are found in biological uids as serum, lymph, saliva butalso milk, allowing a trans-individual transfer of information (Guet al., 2012). Besides their role in physiology, they have been shownto be involved in pathological conditions. Several natural infectiousprocesses use exosomes as natural trojan horses. In prion disease,abnormally folded prion protein was found associated with exosomes,suggesting their contribution in the spreading of prions throughoutthe organism (Fevrier et al., 2004). More recently, they have beenshown to be largely involved in neurodegenerative disorders (Ghidoniet al., 2008; Rajendran et al., 2006; Schneider and Simons, 2012).Moreover, the trojan exosome hypothesis proposes that retrovirusesincluding HIV can take advantage of the intercellular vesicle trafcand exosome exchange to move between cells in the absence of fusionevents in search of adequate target cells (Izquierdo-Useros et al., 2010).

Cancer cells also produce their own exosomes in excessive amountas compared to normal cells. They were shown to mediate tumor

growth by suppressing immune response: inhibiting the functions of

show that exosomes can be efcient tools to make nucleic acid crossthe blood-brain barrier.

211G. Collet et al. / Gene 525 (2013) 208216Cancer cells could also be directly targeted. For example, thedeliveryof tumor-suppressive miRNA was achieved to desired sites usingexosomes and the gene-silencing effect on recipient cell was obtained,resulting in the inhibition of cancer progression (Kosaka et al., 2012).

According to these recent results, the concept of utilizing exosomesas natural vectors for gene delivery is promising. In fact, the non-immunogenic property of autologous exosomes allow nucleic acid toreach the brain through a natural targeting which can be modulatedby engineering the exosome producing cells.

5. When the horse becomes a cell: the cell-mediated gene therapy

Some cells are known to be able to target specically a well-dened part from the organism. Endothelial precursor cells (EPCs)are among these specic cells. They are devoted to reach selectivelyneoangiogenesis areas as well as vascular remodeling regions.From the 1990s, Asahara and colleagues reported the existenceof CD34+expressing cells in the blood of adult mice, which coulddifferentiate in vitro into endothelial cells (Asahara et al., 1997).They showed later on that this EPCs mobilization depends onchemoattracting signals such as vascular endothelial growth factor(VEGF) (Asahara et al., 1999b). Moreover EPCs contribute to postnatalphysiological and pathological neovascularization as well (Asaharaet al., 1999a), thus presenting a perfectly adapted tool for tumortargeting (Varma et al., 2012).

Then the trojan horse was making sense once more, revisiting theAsahara's pioneer works on EPCs and using gene transfer techniquesthat provide the possibility to arm cells with therapeutic genesprior systemic injection (Asahara et al., 2000; Debatin et al., 2008;Dudek, 2010).

Alternatively to EPCs, mesenchymal stem cells (MSCs) were consid-ered as potential candidates for gene transfer (Tang et al., 2010).The homing of MSCs toward tumors has prompted extensive researchto test their use for cancer-specic gene delivery (Dwyer et al., 2010; Sunet al., 2011). Depending on the site to target, such approach could be ex-tended to neural stemcells (Zhao et al., 2012),macrophages (Burke et al.,T cells and natural killers (NK), the differentiation of precursors tomature antigen-presenting cells, increasing the number and/or activityof immune suppressor cells (Zhang and Grizzle, 2011). The exchangeof exosomes between cancer cells and tumor stromamay also promotethe transfer of oncogenes and onco-miRNAs from cell to cell (Kharazihaet al., 2012).

Exosomes are considered as carriers both in pathological as wellas physiological conditions. Indeed, exosomes were found to containendogenous miRNAs. Consequently, exosomes provide these funda-mental regulators of various biological activities the mean to bespecically targeted to distant sites. Taking advantage of this naturaltargeting process, exosomes were loaded with small RNAs to reachspecic sites/cells in order to modify desired protein expression(Lee et al., 2012; O'Loughlin et al., 2012; Tan et al., 2012).

Using exosome properties, attempts to control the immunesystem were developed. Plasma exosomes were used as vectors forgene delivery in order to carry exogenous siRNA to human monocytesand lymphocytes. This was shown to successfully lead to the selectivegene silencing in the case of mitogen-activated protein kinase 1(Wahlgren et al., 2012). In addition, exosomes were shown to transfermiRNAs from T cells to antigen-presenting cells, modulating geneexpression in recipient cells (Mittelbrunn et al., 2011). This effect wasalso demonstrated in vivo. Dendritic cell-derived exosomes expressinga neuron-specic peptide, loadedwith siRNA, and injected intravenouslywere shown to deliver siRNA specically to brain cells, leading to aspecic gene knockdown (Alvarez-Erviti et al., 2011). These results2002) or neutrophils (Tazzyman et al., 2009), and generally to all cellsthat have homing properties to a targeted site (Dudek, 2010; Tabatabaiet al., 2011).

Various methods and vectors can be used to engineer cells makingthem express therapeutic transgenes. The viral constructs are oftenpreferred also not excluding the possibilities offered by non-viralmethods. For such applications, vectors are mainly based on adenovi-ruses, retroviruses, adeno-associated viruses (Herrlinger et al., 2000;Okada and Ozawa, 2008; Ozawa et al., 2000; Ramezani et al., 2003;Stender et al., 2007), and more recently baculoviruses (Zhao et al.,2012). The choice of the viral type of vectors depends on the expectedeffect, like integration into the genome of the recipient cell, ability totransduce, immunogenic potential, level of transgene expression anddurability of expression.

Once the cell vehicle is determined, the nucleic acid to transfer(cDNA for protein expression or small RNAs, i.e. siRNA or miRNA,for modulation of protein expression) is chosen as a function of thedesired type of tumor treatment. Various therapeutic genes havebeen reported including the prodrug-activating enzymes (cytosinedeaminase, carboxylesterase, thymidine kinase), treatments appliedin combination with the proper pro-drug molecules (Aboody et al.,2000, 2006a; Bak et al., 2010; Conrad et al., 2011; Zhao et al., 2012).Various genes expressing interleukins (IL-2, IL-4, IL-12, IL-23) (Elzaouket al., 2006; Nakamura et al., 2004; Yuan et al., 2006), interferon-(Dickson et al., 2007; Sims et al., 2008), apoptosis-promoting genessuch as TRAIL (Ehtesham et al., 2002; Shah et al., 2005), soluble VEGFreceptor (sFlk1) as VEGF-trap (Davidoff et al., 2001) and even expressionof fractalkine by cells (Xin et al., 2007) were successfully appliedand led to either tumor growth inhibition and/or in improvement ofanimal survival. In addition to the tumor development delay broughtby 131Iodide-based therapy using natrium-iodide symporter deliveredby MSCs, such stategy has been shown to allow an imaging modalitywhen used with 123I for scintigraphy or 124I for PET (Dwyer et al.,2011; Knoop et al., 2011).

Further improvements of such cell-mediated approaches werereported by Zhao et al. describing the use of neural stem cells to targeta glioma tumor armed by baculoviral vector to introduce the herpessimplex virus thymidine kinase (HSV-TK) suicide gene. Then, the TKgene product in combination with the pro-drug ganciclovir (GCV)produces a potent toxin which affects replicative cells and inhibitstumor growth (Bak et al., 2010; Zhao et al., 2012). Baculoviral vectoradvantages are rst due to their non-integration and transient trans-gene expression in human cells, both dividing and non-dividing cellsincluding human embryonic stem cells and MSCs (Bak et al., 2010;Zeng et al., 2007). Baculoviral vectors are presented as a safe classof gene delivery vectors because they do not replicate nor cause anytoxicity in mammalian cells (Hu, 2008; Wang and Balasundaram,2010).

To improve the regulation specicity, Conrad et al. and Niess et al.reported the possibility to engineer MSCs to express the therapeuticgene (TK-GCV couple) under the selective control of Tie2 promoter/enhancer (Conrad et al., 2011; Niess et al., 2011). Actively recruitedto growing tumor vasculature, the construction drives the therapeuticgene expression only in the context of angiogenesis. This selectiveexpression restricted to a tumor-specic toxic environment confersanother degree of control to render the approach safer.

In a validation purpose, authors injected the engineered cellsinside or close to the tumor. Nevertheless, the best adapted strategyfor clinical application would be the systemic injections, takingadvantage of the natural targeting specic potential of engineeredcells to reach the tumor site (Aboody et al., 2006a, 2006b; Dicksonet al., 2007; Studeny et al., 2002).

6. When the horse becomes a virus

Viruses are known for their extremely high efciency to transfer

genetic material into cells. They represent for this reason a tempting

212 G. Collet et al. / Gene 525 (2013) 208216delivery tool for gene therapy. Besides previous classical methods,viruses used as trojan horse for tumor targeted approaches andviral gene therapy appear as a potential new strategy in anti-cancertreatments. Ability to carry and introduce a transgene does notwarrantee success since virus-based therapies efcacy is impairedby poor control of targeting. Viral vectors for gene therapy shouldinfect only desired tissue to limit the toxicity towards surroundingcells. To date, reality is still far from this. Nevertheless, research isactively performed in this purpose.

Among viruses used for gene therapy, retroviruses have been with-drawn from clinical trials because of uncontrolled insertionmutagenesiswhich led to adverse events such as disruption and transcriptionalactivation of genes, including oncogenes and transmission via germcells (Kay et al., 2001). Adenovirus (Ad) is awidely used vector for cancergene therapy, while adeno-associated viruses (AAV) are emerging forcancer gene therapy with promising results.

Ad are recognized as tools of choice for cancer gene therapy : theycan be produced at very high titers, have relatively high capacity fortransgene insertion and efciently transduce both quiescent andactively dividing cellswithout incorporation of viral DNA into the host ge-nome (Alemany et al., 2000; Majhen and Ambriovic-Ristov, 2006). Usingadenoviruses, various cancer gene therapy strategieswere developed andclassied intove categories of effects, includingmutation compensation,molecular chemotherapy, genetic immunopotentiation, oncolytic agentsand inhibition of angiogenesis (Majhen and Ambriovic-Ristov, 2006).

Mutation compensation aims to correct deregulated geneexpression of either tumor suppressor genes (i.e. p53, PTEN or BRCA1)or oncogenes (ErbB2) (Ding et al., 2008, 2012). Thus, wild-type p53gene encoding Ad vectors can be used to restore a proper cell controland induce apoptosis (Pisters et al., 2004) while blocking oncogeneexpression such as c-myc can be achieved by introducing antisenseoligonucleotides in Ad-vector (Chen et al., 2001). Both strategies canbe combined as demonstrated by Irie et al. using Ad vector restoringp53 and reducing the overexpressed ErbB2 by expressing an anti-ErbB2 ribozyme within the same entity (Irie et al., 2006).

The molecular chemotherapy, also known as suicide gene therapy,is based on the introduction into cells of pro-drug activating enzymessuch as cytosine deaminase, carboxylesterase or TK. These therapiesare used in combination with the proper pro-drugs (i.e. GCV for TK)since the expressed gene products are able to generate the active formof the toxic drug (Song, 2005). Moreover, this approach is reported toinduce a lateral diffusion of the activated pro-drug, leading to the socalled bystander effect which means that one treated cell causes thedeath of nearby untreated cells.

Due to the weak immunogenicity induced by tumor cells, theimmunopotentiation approach aims to increase the anti-tumor cellspecic immune response. Dendritic cells can be transduced by Adcoding for a tumor-specic antigen then activate cytotoxicT-lymphocytes and induce a protective immunity against tumorcells (Kaplan et al., 1999). Barajas et al. reported activation of NKcells and inhibition of angiogenesis as antitumor mechanisms linkedto injection of adenovirus expressing IL-12 (Barajas et al., 2001), asIL-12 was shown to inhibit angiogenesis by mediating NK recruitment(Bielawska-Pohl et al., 2010).

More elaborated and used for clinical application, oncolytic virus-es (OVs) are genetically modied viruses designed to kill cancer cells,taking advantage of their abnormalities (i.e. p53 deciency) withoutaffecting normal tissues. Resulting from cell lysis, OVs provide theadditional benet of local amplication and highly immunogenicresponse made from the tumor cell epitopes released (Hemminki,2012). Designed to be intravenously administered (although intra-tumoral way is used too), this approach provides a real trojanhorse. Furthermore, an additional modality has been developed byusing OVs as carrier for various protein encoding genes, so calledarmed viruses (Guo and Fang, 2009), such as interleukins, i.e.

IL-12 (Zhu et al., 2012) or IL-24 (Liu et al., 2012b), TRAIL to enhancethe anti-tumor activity, (Guo et al., 2006), or shRNA for targets (Konet al., 2012) such as IL-8 (Yoo et al., 2008), VEGF (Yoo et al., 2007), orsurvivin (Shen et al., 2009).

Naturally oncolytic viruses called reoviruses are of great interestfor cancer therapy (Maitra et al., 2012). The basis of the ability ofreovirus to target and kill tumor cells without infecting normalnonproliferating cells lies in its ability to usurp the highly activatedsignaling pathway found in tumor cells (Lal et al., 2009; Wilcoxet al., 2001). This is most clearly established for Ras or elementsfrom its downstream pathways but ongoing evidences reveal theinvolvement of other complex molecular mechanisms which remainto be claried. Based on this oncolytic property, reovirus-basedtherapies are currently in clinical trials for the treatment of variouscancers (Galanis et al., 2012; Morris et al., 2012).

To supply nutrients and oxygen for cell proliferation, tumorinitiates angiogenesis leading to neo-formation of blood vessels. Dueto its involvement in tumor growth, angiogenesis constitutes a targetof choice to inhibit tumor development. Thus, an adenoviral vectorwas designed by Popkov et al. to deliver a recombinant single-chainantibody fragment (pAd-2S03) able to inhibit Tie-2 receptor (Popkovet al., 2005), leading to tumor growth inhibition by reducing tumorvasculature.

One limitation of such adenovirus-based gene therapy resides inthe low level of selectivity of the treatment towards cancer cells.Thus, other control must operate such as transductional targeting,when the vector is directed toward desired cells or transcriptionaltargeting which connes the transgene expression to certain tissues(Majhen and Ambriovic-Ristov, 2006).

The attachment and entry of Ad are mediated by bers whichconfer their tropisms to the viral particles. Addition of a RGD motifallows RGD-Ad to recognize and bind v-integrins, a class of mem-brane protein overexpressed in various cancer types (Dmitrievet al., 1998), leading to a better infectivity of cancer cells (Buskenset al., 2003; Kanerva et al., 2002). To enhance the Ad particles infec-tivity, combination of different serotypes, chimerism and targetingligand(s) incorporation in ber molecules can be designed. To thisend, Shinozaki et al. engineered a hybrid Ad5/35 virus, where theserotype 5 bers, that are poorly infective for endothelial cells,have been replaced by the serotype 35 bers to allow targetingtumor vasculature (Shinozaki et al., 2006). Only found in theangiogenesis-rich border region of the tumor this virus was sug-gested as a useful vector in anticancer strategies for the targetingof tumor endothelial cells.

At the transcriptional targeting level, the gene of interest is to beunder the control of a tumor-specic promoter (Robson and Hirst,2003). Thus expression is restricted to tumor cells by hTERT (Song,2005; Zhu et al., 2012) or to prostate cancer by DD3 promoters(Ding et al., 2012). Okada et al. described an Ad-RGD mediatedHSV-TK/GCV system where HSV-TK gene was controlled either bymelanoma-specic tyrosinase promoter or by TERT (Okada et al.,2005). Consequently, made to be safe, this vector system for suicidegene therapy is efcient and ts perfectly with the trojan horseapproach. Moreover, gene expression could also be advantageouslydependent upon the conditions which are characteristic of thetumor microenvironment such as hypoxia (Kwon et al., 2010; Xieet al., 2009). A hypoxia/HIF-dependent replicative adenovirus displayshypoxia-conditioned cytolytic activity towards hypoxic but notnormoxic cells (Cho et al., 2004; Post and Van Meir, 2003).

In this constant quest for a tight regulation and tumor specicity,new approaches are proposed combining available mechanisms. Itgives rise to so-called Cancer Targeting Gene-Viro-Therapy (CTGVT)or Gene Armed Oncolytic Virus Therapy (GAOVT). Both are based onthe insertion of an antitumor gene into an oncolytic virus. Thesesophisticated combination strategies are addressed to the tumorcells, which ultimately are attacked from different sides to be killed

(Cai et al., 2012; Liu et al., 2012b). They have a much higher

213G. Collet et al. / Gene 525 (2013) 208216antitumor activity than either gene therapy or oncolytic virotherapyalone (Liu et al., 2012c).

Nevertheless, immunity reduces drastically virotherapy efcacy.Liver and spleen take up the major fraction of injected virusesfrom the circulation. In addition to this innate barrier, the adaptativeimmunity barriers are built in reaction to systemic delivery by pro-duction of antiviral antibodies. Immunity compromises the possibilityof multiple virotherapeutic injections (Power and Bell, 2008).

Attemps were undertaken to limit the visibility of the viral anti-gens by the immune system. Two approaches were mainly reported:the covalent binding of PEG to protect Ad from antibody neutraliza-tion (O'Riordan et al., 1999; Yao et al., 2009), the loading into a carriercell to be protected and hidden. Additional benets are provides bythe carrier cell targeting properties (Power and Bell, 2008).

Among the currently developed viral vectors, AAV is considered asa promising vector because of its lack of pathogenicity and toxicity,its ability to infect dividing and non-dividing cells of various tissueorigins, its very low immunogenicity and long-term expressionwithout integration into host chromosome (Ponnazhagan et al.,2001). Basically, approaches can be divided similarly to adenoviralapproaches: antiangiogenesis, immunomodulation, suicide gene thera-py and repair of damaged tumor cells (Li et al., 2005). Comparably toadenovirus, specic tumor cell targeting is made possible by controlof transduction as well as transcription with tumor cell-specicenhancers/promoters (Nicklin et al., 2003). Hypoxia and hypoxiaresponsive element (HRE) triggering sequence is used to allow atumor hypoxia selective transcription of the transgene (Ruan et al.,2001). Various therapeutic approaches can be reported such as expres-sion of a soluble Flt1 to sequester the VEGF and inhibit angiogenesis(Mahendra et al., 2005) but also angiostatin and endostatin(Ponnazhagan et al., 2004), or TK for combination with GCV (Pan et al.,2012).

The viral vectors reviewed here are not an exhaustive list, butrepresent those which are used in current clinical trials or underadvanced preclinical development.

7. When the horse becomes a virus-producing cell

For their efciency to infect cells and to transfer genetic material,viruses are tools of choice to introduce transgenes into cells. Nevertheless,safe clinical applications remain the challenge.

The well-known homing of cells such as EPCs or MSCs is alreadyused in cancer therapy to reach the tumor site, being carried by theblood to the site of interest (Asahara et al., 1997; Studeny et al.,2002). This cell mediated targeting complies with host tolerance.

The idea of making both approaches fused in a unique strategycould bring a breakthrough for more efcient therapy (Coukos et al.,1999; Guo et al., 2008; Harrington et al., 2002; Power and Bell,2008). The cell provides the vehicle to reach the tumor site and offersthe platform for locally virus release allowing viral particles to trans-duce surrounding tumor cells. In such situation, cells serve as trojanhorse vehicles to evade antiviral mechanisms encountered in thebloodstream (i.e. antibodies, cells) and to avoid uptake by off-targettissues. Such carrier cell-based delivery of virus combination is fullyincluded in the trojan horse approaches.

Involved in the development of such a strategy, Power and Bellreported that its success is tightly dependent on three coordinatedsequential phases which are (1) ex vivo loading, (2) stealth deliveryand (3) virus production at the tumor site (Power and Bell, 2008).Power et al. brought the proof-of-concept that viruses-infected carri-er cells, when intravenously injected, escape from antiviral defensesand reach the tumor site during the eclipse phase (prior to viralprotein synthesis and virion release) (Power et al., 2007). In thetumor site, the protective vehicle role of carrier cells shifts tovirus factory able to produce up to hundreds of virions in the tumor

environment where they infect the surrounding tumor cells thusaccomplishing their therapeutic effect (oncolytic killing for example).Moreover, thanks to the absence of immune response, several injec-tions can be performed to enhance therapeutic efcacy as opposedto naked viruses (Power et al., 2007).

However, efcient delivery of viruses/vectors to tumor sitesremains a challenge that tightly depends on the carrier. Deng andJia hypothesized the use of EPCs to achieve oncolytic viruses delivery(Deng and Jia, 2008). Asahara et al. described that, when systemicallyadministered, EPCs should migrate via peripheral blood to home tothe site of tumor neovasculature (primary and metastatic) to exerttherapy (Asahara et al., 1997). Currently, several technical problemshave to be solved before clinical application, such as EPC expansionin keeping their progenitor cell characteristics and limitation of lyticviral activity in EPCs before they reach the tumor site.

Alternatively to EPCs other cells have been used such as MSCs(Komarova et al., 2006; Pereboeva et al., 2003; Stoff-Khalili et al.,2007), activated T cells (Ong et al., 2007), monocytes, endothelial cells(outgrowth endothelial cells) (Jevremovic et al., 2004), peripheralblood mononuclear cells (Iankov et al., 2007) and even tumor cells(Garcia-Castro et al., 2005; Raykov et al., 2004).

To restrict the therapy to cancer cells, a dual targeting of transgeneexpression to cancer cells using both transcriptional and mRNAtranslational control was developed (Stoff-Khalili et al., 2008). Thiscombines a cancer specic gene transcriptional control using theCXCR4 gene promoter and cancer specic mRNA translational controlutilizing a 5-UTR element from the broblast growth factor-2 mRNA.Replicative specicity demonstrated signicant improvement in tumorselectivity.

This very promising combination of a carrier cell and a virus opensgreat new avenues for therapies. Nevertheless, many hurdles remainbefore such strategy can be clinically applied to patients. The choice ofthe carrier cell is a prominent obstacle in order to achieve an efcienttargeting of the tumor site. Studies on stem cells and precursors cellsraise the prerequisite of phenotype stability in vivo as well as in vitro.Moreover the lytic viral activity has to be controlled in the carriercells until they reach their tumor targets (Deng and Jia, 2008).

8. Concluding remarks

The last ve decades have brought huge improvement in cancertherapy and genetic engineering still opens new horizons throughbetter targeting, addition of safety locks to secure patient treatmentand to overcome reported limitations. In fact, among various trojanhorse types, oncolytic virus studies are the most completed andseveral are already in phase II or III clinical trials (Schmidt, 2011).

Interdisciplinary research is needed to develop innovative therapeuticapproaches for an effective transfer systemand a selective gene targeting.According to newprogresses and promising results, synergy effect shouldbe obtained by combining the scientic tool boxes. Success shouldcome with a judicious assembly of a controlled trafcking to a specictargeting, and a tight regulation making therapeutic delivery highlyrestricted to cancer cells, keeping in mind the investigations to ndnew angles from which to attack and defeat the tumor. Future willshow if these trojan horse like therapeutic approaches will bringimprovements in gene therapy efcacy to help cancer treatments.

Acknowledgments

The authors are grateful to Agata Matejuk for revising theEnglish text.

References

Aboody, K.S., et al., 2000. Neural stem cells display extensive tropism for pathology inadult brain: evidence from intracranial gliomas. Proc. Natl. Acad. Sci. U. S. A. 97,

1284612851.

214 G. Collet et al. / Gene 525 (2013) 208216Aboody, K.S., et al., 2006a. Development of a tumor-selective approach to treat metastaticcancer. PLoS One 1, e23.

Aboody, K.S., et al., 2006b. Targeting of melanoma brain metastases using engineeredneural stem/progenitor cells. Neuro Oncol. 8, 119126.

Alemany, R., Balague, C., Curiel, D.T., 2000. Replicative adenoviruses for cancer therapy.Nat. Biotechnol. 18, 723727.

Alvarez-Erviti, L., Seow, Y., Yin, H., Betts, C., Lakhal, S., Wood, M.J., 2011. Deliveryof siRNA to the mouse brain by systemic injection of targeted exosomes. Nat.Biotechnol. 29, 341345.

Asahara, T., et al., 1997. Isolation of putative progenitor endothelial cells for angiogenesis.Science 275, 964967.

Asahara, T., et al., 1999a. Bone marrow origin of endothelial progenitor cells responsiblefor postnatal vasculogenesis in physiological and pathological neovascularization.Circ. Res. 85, 221228.

Asahara, T., et al., 1999b. VEGF contributes to postnatal neovascularization by mobilizingbone marrow-derived endothelial progenitor cells. EMBO J. 18, 39643972.

Asahara, T., Kalka, C., Isner, J.M., 2000. Stem cell therapy and gene transfer for regeneration.Gene Ther. 7, 451457.

Babar, I.A., et al., 2012. Nanoparticle-based therapy in an in vivo microRNA-155 (miR-155)-dependent mouse model of lymphoma. Proc. Natl. Acad. Sci. U. S. A. 109,E1695E1704.

Bak, X.Y., Yang, J., Wang, S., 2010. Baculovirus-transduced bone marrow mesenchymalstem cells for systemic cancer therapy. Cancer Gene Ther. 17, 721729.

Bangham, A.D., 1995. Surrogate cells or Trojan horses. The discovery of liposomes.Bioessays 17, 10811088.

Barajas, M., et al., 2001. Gene therapy of orthotopic hepatocellular carcinoma in ratsusing adenovirus coding for interleukin 12. Hepatology 33, 5261.

Bielawska-Pohl, A., et al., 2010. The anti-angiogenic activity of IL-12 is increased iniNOS-/-mice and involves NK cells. J. Mol. Med. (Berl) 88, 775784.

Boado, R.J., 2007. Blood-brain barrier transport of non-viral gene and RNAi therapeutics.Pharm. Res. 24, 17721787.

Bray, F., Jemal, A., Grey, N., Ferlay, J., Forman, D., 2012. Global cancer transitionsaccording to the Human Development Index (20082030): a population-basedstudy. Lancet Oncol. 13 (8), 790801.

Burke, B., Sumner, S., Maitland, N., Lewis, C.E., 2002. Macrophages in gene therapy:cellular delivery vehicles and in vivo targets. J. Leukoc. Biol. 72, 417428.

Buskens, C.J., et al., 2003. A genetically retargeted adenoviral vector enhances viraltransduction in esophageal carcinoma cell lines and primary cultured esophagealresection specimens. Ann. Surg. 238, 815824 (discussion 825-6).

Cai, Y., Liu, X., Huang, W., Zhang, K., Liu, X.Y., 2012. Synergistic antitumor effect of TRAILand IL-24 with complete eradication of hepatoma in the CTGVT-DG strategy. ActaBiochim. Biophys. Sin. (Shanghai) 44, 535543.

Carson, A.R., et al., 2012. Ultrasound-targeted microbubble destruction to deliver siRNAcancer therapy. Cancer Res. 72 (23), 61916199.

Chen, J.P., et al., 2001. Molecular therapy with recombinant antisense c-myc adenovirusfor human gastric carcinoma cells in vitro and in vivo. J. Gastroenterol. Hepatol. 16,2228.

Chen, Z.Y., Liang, K., Qiu, R.X., Luo, L.P., 2011. Enhancing microRNA transfection toinhibit survivin gene expression and induce apoptosis: could it be mediated bya novel combination of sonoporation and polyethylenimine? Chin. Med. J. (Engl)124, 35923594.

Chen, Y., et al., 2012. Development of an MRI-visible nonviral vector for siRNA deliverytargeting gastric cancer. Int. J. Nanomedicine 7, 359368.

Cho, W.K., et al., 2004. Oncolytic effects of adenovirus mutant capable of replicatingin hypoxic and normoxic regions of solid tumor. Mol. Ther. 10, 938949.

Cline, M.J., 1985. Perspectives for gene therapy: inserting new genetic information intomammalian cells by physical techniques and viral vectors. Pharmacol. Ther. 29,6992.

Conrad, C., et al., 2011. Linking transgene expression of engineered mesenchymal stemcells and angiopoietin-1-induced differentiation to target cancer angiogenesis.Ann. Surg. 253, 566571.

Coukos, G., et al., 1999. Use of carrier cells to deliver a replication-selective herpessimplex virus-1 mutant for the intraperitoneal therapy of epithelial ovarian cancer.Clin. Cancer Res. 5, 15231537.

Davidoff, A.M., et al., 2001. Bone marrow-derived cells contribute to tumorneovasculature and, when modied to express an angiogenesis inhibitor, can restricttumor growth in mice. Clin. Cancer Res. 7, 28702879.

Debatin, K.M., Wei, J., Beltinger, C., 2008. Endothelial progenitor cells for cancer genetherapy. Gene Ther. 15, 780786.

Deng, W., Jia, J., 2008. Endothelial progenitor cells as cellular vehicles to deliveroncolytic virus therapies to metastatic tumors: the Trojan horse approach.Med. Hypotheses 70, 842844.

Dickson, P.V., et al., 2007. Intravascular administration of tumor tropic neural progenitorcells permits targeted delivery of interferon-beta and restricts tumor growth in amurine model of disseminated neuroblastoma. J. Pediatr. Surg. 42, 4853.

Ding, L., et al., 2008. Somatic mutations affect key pathways in lung adenocarcinoma.Nature 455, 10691075.

Ding, M., et al., 2012. Prostate cancer-specic and potent antitumor effect of a DD3-controlled oncolytic virus harboring the PTEN gene. PLoS One 7, e35153.

Dmitriev, I., et al., 1998. An adenovirus vector with genetically modied bers demon-strates expanded tropism via utilization of a coxsackievirus and adenovirus receptor-independent cell entry mechanism. J. Virol. 72, 97069713.

Dudek, A.Z., 2010. Endothelial lineage cell as a vehicle for systemic delivery of cancergene therapy. Transl. Res. 156, 136146.

Dwyer, R.M., Khan, S., Barry, F.P., O'Brien, T., Kerin, M.J., 2010. Advances in mesenchy-mal stem cell-mediated gene therapy for cancer. Stem Cell Res. Ther. 1, 25.Dwyer, R.M., et al., 2011. Mesenchymal stem cell-mediated delivery of the sodiumiodide symporter supports radionuclide imaging and treatment of breast cancer.Stem Cells 29, 11491157.

Ehtesham, M., et al., 2002. Induction of glioblastoma apoptosis using neural stem cell-mediated delivery of tumor necrosis factor-related apoptosis-inducing ligand.Cancer Res. 62, 71707174.

Elzaouk, L., Moelling, K., Pavlovic, J., 2006. Anti-tumor activity of mesenchymal stemcells producing IL-12 in a mouse melanoma model. Exp. Dermatol. 15, 865874.

Fevrier, B., Raposo, G., 2004. Exosomes: endosomal-derived vesicles shipping extracellularmessages. Curr. Opin. Cell Biol. 16, 415421.

Fevrier, B., et al., 2004. Cells release prions in association with exosomes. Proc. Natl.Acad. Sci. U. S. A. 101, 96839688.

Galanis, E., et al., 2012. Phase II trial of intravenous administration of Reolysin((R))(Reovirus Serotype-3-dearing Strain) in patients with metastatic melanoma. Mol.Ther. 20, 19982003.

Garcia-Castro, J., et al., 2005. Tumor cells as cellular vehicles to deliver gene therapiesto metastatic tumors. Cancer Gene Ther. 12, 341349.

Ghidoni, R., Benussi, L., Binetti, G., 2008. Exosomes: the Trojan horses of neurodegeneration.Med. Hypotheses 70, 12261227.

Gregoriadis, G., Ryman, B.E., 1971. Liposomes as carriers of enzymes or drugs: a newapproach to the treatment of storage diseases. Biochem. J. 124, 58P.

Gu, Y., et al., 2012. Lactation-related microRNA expression proles of porcine breastmilk exosomes. PLoS One 7, e43691.

Guo, W., Fang, B., 2009. Potentiation of oncolytic virotherapy with armed shRNA.Cancer Biol. Ther. 8, 9293.

Guo, W., et al., 2006. Combination effect of oncolytic adenovirotherapy and TRAILgene therapy in syngeneic murine breast cancer models. Cancer Gene Ther. 13,8290.

Guo, Z.S., Thorne, S.H., Bartlett, D.L., 2008. Oncolytic virotherapy: molecular targets intumor-selective replication and carrier cell-mediated delivery of oncolytic viruses.Biochim. Biophys. Acta 1785, 217231.

Halder, J., et al., 2006. Focal adhesion kinase targeting using in vivo short interferingRNA delivery in neutral liposomes for ovarian carcinoma therapy. Clin. CancerRes. 12, 49164924.

Harrington, K., et al., 2002. Cells as vehicles for cancer gene therapy: the missinglink between targeted vectors and systemic delivery? Hum. Gene Ther. 13,12631280.

He, Y., et al., 2011. Ultrasound microbubble-mediated delivery of the siRNAs targetingMDR1 reduces drug resistance of yolk sac carcinoma L2 cells. J. Exp. Clin. CancerRes. 30, 104.

Hemminki, A., 2012. Portrait of a leader in immunotherapeutics: oncolytic viruses fortreatment of cancer. Hum. Vaccin. Immunother. 8.

Herrlinger, U., et al., 2000. Helper virus-free herpes simplex virus type 1 ampliconvectors for granulocyte-macrophage colony-stimulating factor-enhanced vaccinationtherapy for experimental glioma. Hum. Gene Ther. 11, 14291438.

Hu, Y.C., 2008. Baculoviral vectors for gene delivery: a review. Curr. Gene Ther. 8,5465.

Huang, Q.D., et al., 2011a. Cyclen-based cationic lipids for highly efcient gene deliverytowards tumor cells. PLoS One 6, e23134.

Huang, S., et al., 2011b. Dual targeting effect of Angiopep-2-modied, DNA-loadednanoparticles for glioma. Biomaterials 32, 68326838.

Iankov, I.D., et al., 2007. Infected cell carriers: a new strategy for systemic delivery ofoncolytic measles viruses in cancer virotherapy. Mol. Ther. 15, 114122.

Irie, A., et al., 2006. Growth inhibition efcacy of an adenovirus expressing dualtherapeutic genes, wild-type p53, and anti-erbB2 ribozyme, against human bladdercancer cells. Cancer Gene Ther. 13, 298305.

Izquierdo-Useros, N., et al., 2010. HIV and mature dendritic cells: Trojan exosomesriding the Trojan horse? PLoS Pathog. 6, e1000740.

Jevremovic, D., et al., 2004. Use of blood outgrowth endothelial cells as virus-producingvectors for gene delivery to tumors. Am. J. Physiol. Heart Circ. Physiol. 287,H494H500.

Kam, N.W., Liu, Z., Dai, H., 2005. Functionalization of carbon nanotubes via cleavabledisulde bonds for efcient intracellular delivery of siRNA and potent gene silencing.J. Am. Chem. Soc. 127, 1249212493.

Kanerva, A., et al., 2002. Gene transfer to ovarian cancer versus normal tissues withber-modied adenoviruses. Mol. Ther. 5, 695704.

Kaplan, J.M., et al., 1999. Induction of antitumor immunity with dendritic cells trans-duced with adenovirus vector-encoding endogenous tumor-associated antigens.J. Immunol. 163, 699707.

Kay, M.A., Glorioso, J.C., Naldini, L., 2001. Viral vectors for gene therapy: the art ofturning infectious agents into vehicles of therapeutics. Nat. Med. 7, 3340.

Kharaziha, P., Ceder, S., Li, Q., Panaretakis, T., 2012. Tumor cell-derived exosomes:a message in a bottle. Biochim. Biophys. Acta 1826, 103111.

Kichler, A., Leborgne, C., Coeytaux, E., Danos, O., 2001. Polyethylenimine-mediatedgene delivery: a mechanistic study. J. Gene Med. 3, 135144.

Knoop, K., et al., 2011. Image-guided, tumor stroma-targeted 131I therapy of hepato-cellular cancer after systemic mesenchymal stem cell-mediated NIS gene delivery.Mol. Ther. 19, 17041713.

Komarova, S., Kawakami, Y., Stoff-Khalili, M.A., Curiel, D.T., Pereboeva, L., 2006.Mesenchymal progenitor cells as cellular vehicles for delivery of oncolytic adenoviruses.Mol. Cancer Ther. 5, 755766.

Kon, T., et al., 2012. Oncolytic virus-mediated tumor radiosensitization in mice throughDNAPKcs-specic shRNA. Transl. Cancer Res. 1, 414.

Kosaka, N., et al., 2012. Exosomal tumor-suppressive microRNAs as novel cancertherapy: exocure is another choice for cancer treatment. Adv. Drug Deliv. Rev.65 (3), 376382.

215G. Collet et al. / Gene 525 (2013) 208216Kwon, O.J., Kim, P.H., Huyn, S., Wu, L., Kim, M., Yun, C.O., 2010. A hypoxia- and {alpha}-fetoprotein-dependent oncolytic adenovirus exhibits specic killing of hepatocellularcarcinomas. Clin. Cancer Res. 16, 60716082.

Lal, R., Harris, D., Postel-Vinay, S., de Bono, J., 2009. Reovirus: rationale and clinical trialupdate. Curr. Opin. Mol. Ther. 11, 532539.

Landen Jr., C.N., et al., 2005. Therapeutic EphA2 gene targeting in vivo using neutralliposomal small interfering RNA delivery. Cancer Res. 65, 69106918.

Lee, H., Jeong, J.H., Park, T.G., 2002. PEG grafted polylysine with fusogenic peptide forgene delivery: high transfection efciency with low cytotoxicity. J. Control. Release79, 283291.

Lee, Y., El Andaloussi, S., Wood, M.J., 2012. Exosomes and microvesicles: extracellularvesicles for genetic information transfer and gene therapy. Hum. Mol. Genet. 21,R125R134.

Li, C., Bowles, D.E., van Dyke, T., Samulski, R.J., 2005. Adeno-associated virus vectors:potential applications for cancer gene therapy. Cancer Gene Ther. 12, 913925.

Li, J., et al., 2011. The use of myristic acid as a ligand of polyethylenimine/DNAnanoparticles for targeted gene therapy of glioblastoma. Nanotechnology 22, 435101.

Li, J.M., et al., 2012. Multifunctional QD-based co-delivery of siRNA and doxorubicin toHeLa cells for reversal of multidrug resistance and real-time tracking. Biomaterials33, 27802790.

Liu, X.Q., Song, W.J., Sun, T.M., Zhang, P.Z., Wang, J., 2011. Targeted delivery of antisenseinhibitor of miRNA for antiangiogenesis therapy using cRGD-functionalized nano-particles. Mol. Pharm. 8, 250259.

Liu, P., Yu, H., Sun, Y., Zhu, M., Duan, Y., 2012a. A mPEG-PLGA-b-PLL copolymer carrierfor adriamycin and siRNA delivery. Biomaterials 33, 44034412.

Liu, X., et al., 2012b. Gene-viro-therapy targeting liver cancer by a dual-regulatedoncolytic adenoviral vector harboring IL-24 and TRAIL. Cancer Gene Ther. 19,4957.

Liu, X.Y., Li, H.G., Zhang, K.J., Gu, J.F., 2012c. Strategy of cancer targeting gene-viro-therapy (CTGVT) a trend in both cancer gene therapy and cancer virotherapy.Curr. Pharm. Biotechnol. 13, 17611767.

Mahendra, G., et al., 2005. Antiangiogenic cancer gene therapy by adeno-associatedvirus 2-mediated stable expression of the soluble FMS-like tyrosine kinase-1receptor. Cancer Gene Ther. 12, 2634.

Maitra, R., Ghalib, M.H., Goel, S., 2012. Reovirus: a targeted therapeuticprogress andpotential. Mol. Cancer Res. 10, 15141525.

Majhen, D., Ambriovic-Ristov, A., 2006. Adenoviral vectorshow to use them in cancergene therapy? Virus Res. 119, 121133.

Miele, E., Spinelli, G.P., Di Fabrizio, E., Ferretti, E., Tomao, S., Gulino, A., 2012. Nanoparticle-based delivery of small interfering RNA: challenges for cancer therapy. Int. J.Nanomedicine 7, 36373657.

Mittelbrunn, M., et al., 2011. Unidirectional transfer of microRNA-loaded exosomesfrom T cells to antigen-presenting cells. Nat. Commun. 2, 282.

Morris, D.G., et al., 2012. REO-001: a phase I trial of percutaneous intralesionaladministration of reovirus type 3 dearing (Reolysin(R)) in patients with advancedsolid tumors. Invest. New Drugs. http://dx.doi.org/10.1007/s10637-012-9865-z.

Mukherjee, A., et al., 2009. Selective cancer targeting via aberrant behavior of cancercell-associated glucocorticoid receptor. Mol. Ther. 17, 623631.

Nakamura, K., et al., 2004. Antitumor effect of genetically engineered mesenchymalstem cells in a rat glioma model. Gene Ther. 11, 11551164.

Nicklin, S.A., et al., 2003. Transductional and transcriptional targeting of cancer cellsusing genetically engineered viral vectors. Cancer Lett. 201, 165173.

Niess, H., et al., 2011. Selective targeting of genetically engineered mesenchymal stemcells to tumor stroma microenvironments using tissue-specic suicide geneexpression suppresses growth of hepatocellular carcinoma. Ann. Surg. 254,767774 (discussion 774-5).

Okada, T., Ozawa, K., 2008. Vector-producing tumor-tracking multipotent mesenchy-mal stromal cells for suicide cancer gene therapy. Front. Biosci. 13, 18871891.

Okada, Y., Okada, N., Mizuguchi, H., Hayakawa, T., Nakagawa, S., Mayumi, T., 2005.Transcriptional targeting of RGD ber-mutant adenovirus vectors can improvethe safety of suicide gene therapy for murine melanoma. Cancer Gene Ther. 12,608616.

O'Loughlin, A.J., Wofndale, C.A., Wood, M.J., 2012. Exosomes and the emerging eld ofexosome-based gene therapy. Curr. Gene Ther. 12, 262274.

Ong, H.T., Hasegawa, K., Dietz, A.B., Russell, S.J., Peng, K.W., 2007. Evaluation of T cellsas carriers for systemic measles virotherapy in the presence of antiviral antibodies.Gene Ther. 14, 324333.

O'Riordan, C.R., et al., 1999. PEGylation of adenovirus with retention of infectivityand protection from neutralizing antibody in vitro and in vivo. Hum. Gene Ther.10, 13491358.

Ozawa, C.R., Springer, M.L., Blau, H.M., 2000. A novel means of drug delivery: myoblast-mediated gene therapy and regulatable retroviral vectors. Annu. Rev. Pharmacol.Toxicol. 40, 295317.

Pan, J.G., Zhou, X., Luo, R., Han, R.F., 2012. The adeno-associated virus-mediated HSV-TK/GCV suicide system: a potential strategy for the treatment of bladder carcino-ma. Med. Oncol. 29, 19381947.

Pardridge, W.M., 2010. Preparation of Trojan horse liposomes (THLs) for gene transferacross the bloodbrain barrier. Cold Spring Harb. Protoc. http://dx.doi.org/10.1101/pdb.prot5407.

Patel, M.M., Goyal, B.R., Bhadada, S.V., Bhatt, J.S., Amin, A.F., 2009. Getting into thebrain: approaches to enhance brain drug delivery. CNS Drugs 23, 3558.

Pereboeva, L., Komarova, S., Mikheeva, G., Krasnykh, V., Curiel, D.T., 2003. Approaches toutilize mesenchymal progenitor cells as cellular vehicles. Stem Cells 21, 389404.

Pille, J.Y., et al., 2006. Intravenous delivery of anti-RhoA small interfering RNA loaded innanoparticles of chitosan in mice: safety and efcacy in xenografted aggressivebreast cancer. Hum. Gene Ther. 17, 10191026.Pisters, L.L., et al., 2004. Evidence that transfer of functional p53 protein results inincreased apoptosis in prostate cancer. Clin. Cancer Res. 10, 25872593.

Ponnazhagan, S., Curiel, D.T., Shaw, D.R., Alvarez, R.D., Siegal, G.P., 2001. Adeno-associatedvirus for cancer gene therapy. Cancer Res. 61, 63136321.

Ponnazhagan, S., et al., 2004. Adeno-associated virus 2-mediated antiangiogenic cancergene therapy: long-term efcacy of a vector encoding angiostatin and endostatinover vectors encoding a single factor. Cancer Res. 64, 17811787.

Popkov, M., Jendreyko, N., McGavern, D.B., Rader, C., Barbas III, C.F., 2005. Targetingtumor angiogenesis with adenovirus-delivered anti-Tie-2 intrabody. Cancer Res.65, 972981.

Post, D.E., Van Meir, E.G., 2003. A novel hypoxia-inducible factor (HIF) activatedoncolytic adenovirus for cancer therapy. Oncogene 22, 20652072.

Pouton, C.W., Lucas, P., Thomas, B.J., Uduehi, A.N., Milroy, D.A., Moss, S.H., 1998. PolycationDNA complexes for gene delivery: a comparison of the biopharmaceutical propertiesof cationic polypeptides and cationic lipids. J. Control. Release 53, 289299.

Power, A.T., Bell, J.C., 2008. Taming the Trojan horse: optimizing dynamic carrier cell/oncolytic virus systems for cancer biotherapy. Gene Ther. 15, 772779.

Power, A.T., et al., 2007. Carrier cell-based delivery of an oncolytic virus circumventsantiviral immunity. Mol. Ther. 15, 123130.

Rai, K., et al., 2011. Liposomal delivery of MicroRNA-7-expressing plasmid overcomesepidermal growth factor receptor tyrosine kinase inhibitor-resistance in lungcancer cells. Mol. Cancer Ther. 10, 17201727.

Rajendran, L., et al., 2006. Alzheimer's disease beta-amyloid peptides are released inassociation with exosomes. Proc. Natl. Acad. Sci. U. S. A. 103, 1117211177.

Ramezani, A., Hawley, T.S., Hawley, R.G., 2003. Performance- and safety-enhancedlentiviral vectors containing the human interferon-beta scaffold attachment regionand the chicken beta-globin insulator. Blood 101, 47174724.

Raykov, Z., Balboni, G., Aprahamian, M., Rommelaere, J., 2004. Carrier cell-mediateddelivery of oncolytic parvoviruses for targeting metastases. Int. J. Cancer 109,742749.

Robson, T., Hirst, D.G., 2003. Transcriptional targeting in cancer gene therapy. J. Biomed.Biotechnol. 2003, 110137.

Ruan, H., Su, H., Hu, L., Lamborn, K.R., Kan, Y.W., Deen, D.F., 2001. A hypoxia-regulatedadeno-associated virus vector for cancer-specic gene therapy. Neoplasia 3,255263.

Rudzinski, W.E., Aminabhavi, T.M., 2010. Chitosan as a carrier for targeted delivery ofsmall interfering RNA. Int. J. Pharm. 399, 111.

Schmidt, C., 2011. Amgen spikes interest in live virus vaccines for hard-to-treatcancers. Nat. Biotechnol. 29, 295296.

Schneider, A., Simons, M., 2012. Exosomes: vesicular carriers for intercellular commu-nication in neurodegenerative disorders. Cell Tissue Res.

Shah, K., Tung, C.H., Breakeeld, X.O., Weissleder, R., 2005. In vivo imaging of S-TRAIL-mediated tumor regression and apoptosis. Mol. Ther. 11, 926931.

Shen, W., Wang, C.Y., Wang, X.H., Fu, Z.X., 2009. Oncolytic adenovirus mediatedSurvivin knockdown by RNA interference suppresses human colorectal carcinomagrowth in vitro and in vivo. J. Exp. Clin. Cancer Res. 28, 81.

Shinozaki, K., et al., 2006. Efcient infection of tumor endothelial cells by a capsid-modied adenovirus. Gene Ther. 13, 5259.

Sims Jr., T.L., et al., 2008. Neural progenitor cell-mediated delivery of interferonbeta improves neuroblastoma response to cyclophosphamide. Ann. Surg. Oncol.15, 32593267.

Song, J.S., 2005. Adenovirus-mediated suicide SCLC gene therapy using the increasedactivity of the hTERT promoter by the MMRE and SV40 enhancer. Biosci.Biotechnol. Biochem. 69, 5662.

Soriano, P., et al., 1983. Targeted and nontargeted liposomes for in vivo transfer torat liver cells of a plasmid containing the preproinsulin I gene. Proc. Natl. Acad.Sci. U. S. A. 80, 71287131.

Spagnou, S., Miller, A.D., Keller, M., 2004. Lipidic carriers of siRNA: differences inthe formulation, cellular uptake, and delivery with plasmid DNA. Biochemistry43, 1334813356.

Stender, S., et al., 2007. Adeno-associated viral vector transduction of human mesen-chymal stem cells. Eur. Cell. Mater. 13, 9399 (discussion 99).

Stoff-Khalili, M.A., et al., 2007. Mesenchymal stem cells as a vehicle for targeted deliveryof CRAds to lung metastases of breast carcinoma. Breast Cancer Res. Treat. 105,157167.

Stoff-Khalili, M.A., et al., 2008. Cancer-specic targeting of a conditionally replicativeadenovirus using mRNA translational control. Breast Cancer Res. Treat. 108,4355.

Studeny, M., Marini, F.C., Champlin, R.E., Zompetta, C., Fidler, I.J., Andreeff, M., 2002.Bone marrow-derived mesenchymal stem cells as vehicles for interferon-betadelivery into tumors. Cancer Res. 62, 36033608.

Sun, X.Y., Nong, J., Qin, K., Warnock, G.L., Dai, L.J., 2011. Mesenchymal stem cell-mediated cancer therapy: a dual-targeted strategy of personalized medicine.World J. Stem Cells 3, 96103.

Szybalska, E.H., Szybalski, W., 1962. Genetics of human cess line. IV. DNA-mediatedheritable transformation of a biochemical trait. Proc. Natl. Acad. Sci. U. S. A. 48,20262034.

Tabatabai, G., Wick, W., Weller, M., 2011. Stem cell-mediated gene therapies for malignantgliomas: a promising targeted therapeutic approach? Discov. Med. 11, 529536.

Tan, A., Rajadas, J., Seifalian, A.M., 2012. Exosomes as nano-theranostic delivery plat-forms for gene therapy. Adv. Drug Deliv. Rev. 65 (3), 357367.

Tang, C., Russell, P.J., Martiniello-Wilks, R., Rasko, J.E., Khatri, A., 2010. Concise review:nanoparticles and cellular carriers-allies in cancer imaging and cellular gene ther-apy? Stem Cells 28, 16861702.

Taratula, O., Garbuzenko, O., Savla, R., Wang, Y.A., He, H., Minko, T., 2011.Multifunctional nanomedicine platform for cancer specic delivery of siRNA by

superparamagnetic iron oxide nanoparticlesdendrimer complexes. Curr. DrugDeliv. 8, 5969.

Tazzyman, S., Lewis, C.E., Murdoch, C., 2009. Neutrophils: key mediators of tumourangiogenesis. Int. J. Exp. Pathol. 90, 222231.

Thery, C., Zitvogel, L., Amigorena, S., 2002. Exosomes: composition, biogenesis andfunction. Nat. Rev. Immunol. 2, 569579.

Varma, N.R., et al., 2012. Endothelial progenitor cells (EPCs) as gene carrier system forrat model of human glioma. PLoS One 7, e30310.

Wahlgren, J., et al., 2012. Plasma exosomes can deliver exogenous short interferingRNA to monocytes and lymphocytes. Nucleic Acids Res. 40, e130.

Wang, S., Balasundaram, G., 2010. Potential cancer gene therapy by baculoviraltransduction. Curr. Gene Ther. 10, 214225.

Wang, T., Upponi, J.R., Torchilin, V.P., 2012. Design of multifunctional non-viralgene vectors to overcome physiological barriers: dilemmas and strategies. Int. J.Pharm. 427, 320.

Wilcox, M.E., et al., 2001. Reovirus as an oncolytic agent against experimental humanmalignant gliomas. J. Natl. Cancer Inst. 93, 903912.

Xia, C.F., Boado, R.J., Zhang, Y., Chu, C., Pardridge, W.M., 2008. Intravenous glial-derived neurotrophic factor gene therapy of experimental Parkinson's diseasewith Trojan horse liposomes and a tyrosine hydroxylase promoter. J. GeneMed. 10, 306315.

Xie, M., et al., 2009. A novel triple-regulated oncolytic adenovirus carrying PDCD5 geneexerts potent antitumor efcacy on common human leukemic cell lines. Apoptosis14, 10861094.

Xin, H., et al., 2007. Targeted delivery of CX3CL1 tomultiple lung tumors bymesenchymalstem cells. Stem Cells 25, 16181626.

Yang, Y.P., et al., 2012. Inhibition of cancer stem cell-like properties and reducedchemoradioresistance of glioblastoma using microRNA145 with cationicpolyurethane-short branch PEI. Biomaterials 33, 14621476.

Yao, X., et al., 2009. Systemic administration of a PEGylated adenovirus vector with acancer-specic promoter is effective in a mouse model of metastasis. Gene Ther.16, 13951404.

Yoo, J.Y., et al., 2007. VEGF-specic short hairpin RNA-expressing oncolytic adenovi-rus elicits potent inhibition of angiogenesis and tumor growth. Mol. Ther. 15,295302.

Yoo, J.Y., et al., 2008. Short hairpin RNA-expressing oncolytic adenovirus-mediatedinhibition of IL-8: effects on antiangiogenesis and tumor growth inhibition.Gene Ther. 15, 635651.

Yuan, X., Hu, J., Belladonna, M.L., Black, K.L., Yu, J.S., 2006. Interleukin-23-expressingbone marrow-derived neural stem-like cells exhibit antitumor activity againstintracranial glioma. Cancer Res. 66, 26302638.

Zeng, J., Du, J., Zhao, Y., Palanisamy, N., Wang, S., 2007. Baculoviral vector-mediatedtransient and stable transgene expression in human embryonic stem cells.Stem Cells 25, 10551061.

Zhang, H.G., Grizzle, W.E., 2011. Exosomes and cancer: a newly described pathwayof immune suppression. Clin. Cancer Res. 17, 959964.

Zhang, Y., Pardridge, W.M., 2009. Near complete rescue of experimental Parkinson'sdisease with intravenous, non-viral GDNF gene therapy. Pharm. Res. 26,10591063.

Zhang, Y., Zhang, Y.F., Bryant, J., Charles, A., Boado, R.J., Pardridge, W.M., 2004. Intrave-nous RNA interference gene therapy targeting the human epidermal growth factorreceptor prolongs survival in intracranial brain cancer. Clin. Cancer Res. 10,36673677.

Zhang, Z., et al., 2006. Delivery of telomerase reverse transcriptase small interferingRNA in complex with positively charged single-walled carbon nanotubessuppresses tumor growth. Clin. Cancer Res. 12, 49334939.

Zhang, Y., Wang, Y., Boado, R.J., Pardridge, W.M., 2008. Lysosomal enzyme replace-ment of the brain with intravenous non-viral gene transfer. Pharm. Res. 25,400406.

Zhao, Y., et al., 2012. Targeted suicide gene therapy for glioma using human embryonicstem cell-derived neural stem cells genetically modied by baculoviral vectors.Gene Ther. 19, 189200.

Zhu, W., Wei, L., Zhang, H., Chen, J., Qin, X., 2012. Oncolytic adenovirus armed with IL-24Inhibits the growth of breast cancer in vitro and in vivo. J. Exp. Clin. Cancer Res. 31, 51.

216 G. Collet et al. / Gene 525 (2013) 208216

Trojan horse at cellular level for tumor gene therapies1. Introduction2. When the horse becomes a molecule3. When the horse becomes a liposome4. When the horse becomes an exosome5. When the horse becomes a cell: the cell-mediated gene therapy6. When the horse becomes a virus7. When the horse becomes a virus-producing cell8. Concluding remarksAcknowledgmentsReferences