**7. Targets for gene therapy of cancer**

ability to repress E2F-responsive promoters such as E2F-1 promoter. It has been shown that Ad vectors that contain transgenes driven by E2F-1 promoter can mediate tumor-selective gene expression in vivo in glioma cells [79]. The promoters of cell cycle genes such as cyclin D, cyclin A, cdc25c, cyclin-dependent kinase inhibitors, p16/INK4, p27, and p14 could be expected to exert cell cycle arrest, thereby increasing the apoptosis when used in vector targeting strategies in proliferating tumor or endothelial cells [80–83]. Additionally, drug-inducible systems such as tat-on/tat-off regulated by tetracycline or rapamycin could provide a wide-dose response

The second strategy of biologic targeting is to engineer the either viral or nonviral vectors in such a way that they can be captured only in tumor tissues, and therapeutic genes are produced only in the environment of the tumor tissue. There have been numerous attempts to modify the vectors with tumor cell-specific ligands that would increase the specific binding to tumor cells and reduce the toxicity. Therefore, targeting DNA complexes to the tumor cell-specific receptors is an attractive strategy. One of the well-known strategies is coating the surface of the complexes with transferrin, an iron-binding plasma protein that is mainly an up-regulated expression on rapidly proliferating cells as tumors [85]. Likewise, coating with EGF has also been reported to cause a 50-fold increase in the transgene expression in hepatocellular carcinoma cells [86]. The suicide gene HSV-TK/PEI complex mixed with a single chain antibody (scFv) against EGFR with a negatively charged oligopeptide tail has exhibited EGFR-

The nonviral systems usually fail in promoting the delivery of DNA to the nucleus. Almost 99% of the internalized DNA from a nonviral vector is degraded in the cytoplasm [88]. Trafficking of exogenous DNA from cytosol to the nucleus may be improved by using the nuclear localization signal (NLS) found in some nuclear proteins [89]. Dermaseptins, a family of antimicrobial peptides that destabilize the membrane, have been successfully linked to NLS of SV40-T antigen and HIV-1 Rev protein [90]. Likewise, mellitin, which is a membrane-active protein, and viral protein r (vpr) of HIV-1, which binds directly to nucleoporins of the nuclear pore complex, have been successfully bound to PEI/DNA complexes to improve nuclear

The selective targeting of viral vectors to specific cells permits the cell-specific expression of transgenes and enables the systemic administration of the vectors. Avoiding the targeting of the native receptor found on immune and inflammatory cell surfaces also reduces the immunity and inflammation to those vectors. Replication-competent retroviral vectors (RCR) based on murine leukemia virus (MLV) represent an attractive system for gene delivery through their ability to replicate and provide long term transgene expression in rapidly proliferating cells [92]. However, the uncontrolled spread of the RCR might cause the infection of nontarget cells. In order to develop tumor-selective RCR vectors, several modifications have been made such as a modification of the envelope protein by insert‐ ing single chain antibodies (scFv) [93] and peptide ligands [94]. Also, the specifically targeted entry of replication-deficient retroviral vectors has been accomplished by combin‐

The capability of an Ad vector to infect a cell is mainly based on CAR and integrin expression. Following the attachment of an adenoviral vector to the target cell via C-

range in the treatment [84].

10 Gene Therapy - Principles and Challenges

**6.4. Transductional targeting**

transport [91].

specific gene transfer in vitro and in vivo [87].

ing cell-specific monoclonal antibodies [95,96].

Current gene therapy studies have mainly focused on introducing the genes into the tumor cells to block the action of oncogene expression and the development of tumor vasculature, or to induce the development of an immune response against the cancer tissue. The major targets of gene therapy are shown on Table 1.


**Table 1.** The major targets of gene therapy of cancer

#### **7.1. Tumor suppressor genes**

Loss of functions of tumor suppressor genes have crucial role in the development and spread of cancer. Therefore, those genes were among the first targets of gene therapy studies. *P53* is mutated in almost 60 percent of solid tumors. Reintroducing wild-type p53 has been one of the common gene therapy approaches within the last two decades. The introduction of wildtype p53 by retroviruses or replication-deficient adenoviral vectors into the cancer cells inhibits tumor growth both in vitro and in vivo [104]. The use of adenoviral vectors carrying p53 has yielded some clinical activities, particularly in patients with head and neck cancers and lung cancers used either as a single agent or in combination with chemotherapy or radiotherapy [105,106]. Likewise, strategies aiming at the activation of p53 pathway in patients with p53 mutated tumors have also been tried. The introduction of small synthetic peptides like CDB3 derived from p53-binding protein 2 or p53 C terminal peptide have been shown to reactivate the mutant p53 functions in vitro [107]. Furthermore, transductions of other family members of p53 like p63 and p73, which are known to transactivate the downstream genes of p53 pathway, have been shown to induce apoptosis of tumor cells [108,109].

RB1 is a tumor suppressor gene involved in cell cycle regulation. Constitutively active RB1 potently inhibits cellular proliferation and induce persistent cell cycle arrest [110]. Since the first cloning of the RB gene at the beginning of the nineties, researchers have tried to activate the tumor suppressor function of the RB pathway. Gene transfer of truncated RB protein, such as RB94, has been shown to restore the RB pathway and to induce potent tumor growth inhibition both in vitro and in vivo [111]. However, these strategies have not been tested in the clinical setting yet.

The restoration of functions of other tumor suppressor genes such as adenomatosis polyposis coli (APC) in colorectal cancer cells [112] and BRCA1 in breast and ovarian cancers [113] has been shown to slow the growth of tumor cells.

#### **7.2. Oncogenes**

The targeting of oncogenes has long been at the focus of drug development studies in cancer. Small molecules of inhibitors of oncogene functions such as tyrosine kinase inhibitors have already been used in the routine treatment of various cancers. Gene therapeutic strategies to suppress oncogene functions are usually focused on the inhibition of those genes at mRNA level. Usually small oligonucleotides or RNA inhibitors such as short-interfering RNA (siRNA), short-hairpin RNA (shRNA), or micro-RNA (miRNA) have been used to interfere the actions of oncogenes [114].

Chemically modified or unmodified small single-stranded DNA molecules, antisense oligo‐ nucleotides inhibit protein translation through the disruption of ribosome assembly or utilization of RNase H enzymes to destroy mRNA. Numerous oligonucleotides and RNA inhibitors have been designed to inhibit oncogenes, including RAS, MYC, BCL-2, or cell signaling molecules survivin, IGF, VEGF, and PKCalphfa, have been tested. Although the efficacy of these oligonucleotides has shown a great diversity, some of them have been tested in phase II/III clinical trials in various cancer types [115]. Oblimersen, an antisense oligonu‐ cleotide targeting Bcl-2, is one of the oldest agents that have already tested in phase III studies of Chronic Lymphocytic Leukemia CLL and multiple myeloma [116,117]. The members of the RAS family of oncogenes have been found mutated in various solid tumors. Therefore, the targeting of RAS would have been a hot topic in the development of recent therapeutics. Targeting RAS with an anti-RAS mRNA plasmid yielded significant tumor inhibition when used alone or in combination with chemotherapy in hepatoma cells [118]. Antisense oligonu‐ cleotides targeting survivin, which are highly expressed in various cancer types, including liver, lung, breast, and prostate, have been employed successfully to inhibit the expression of the gene [119]. The phase I/II clinical trials have also shown some responses in cancers [120].

#### **7.3. Gene-Directed Enzyme/Prodrug Therapy (GDEPT)**

mutated in almost 60 percent of solid tumors. Reintroducing wild-type p53 has been one of the common gene therapy approaches within the last two decades. The introduction of wildtype p53 by retroviruses or replication-deficient adenoviral vectors into the cancer cells inhibits tumor growth both in vitro and in vivo [104]. The use of adenoviral vectors carrying p53 has yielded some clinical activities, particularly in patients with head and neck cancers and lung cancers used either as a single agent or in combination with chemotherapy or radiotherapy [105,106]. Likewise, strategies aiming at the activation of p53 pathway in patients with p53 mutated tumors have also been tried. The introduction of small synthetic peptides like CDB3 derived from p53-binding protein 2 or p53 C terminal peptide have been shown to reactivate the mutant p53 functions in vitro [107]. Furthermore, transductions of other family members of p53 like p63 and p73, which are known to transactivate the downstream genes of p53

RB1 is a tumor suppressor gene involved in cell cycle regulation. Constitutively active RB1 potently inhibits cellular proliferation and induce persistent cell cycle arrest [110]. Since the first cloning of the RB gene at the beginning of the nineties, researchers have tried to activate the tumor suppressor function of the RB pathway. Gene transfer of truncated RB protein, such as RB94, has been shown to restore the RB pathway and to induce potent tumor growth inhibition both in vitro and in vivo [111]. However, these strategies have not been tested in the

The restoration of functions of other tumor suppressor genes such as adenomatosis polyposis coli (APC) in colorectal cancer cells [112] and BRCA1 in breast and ovarian cancers [113] has

The targeting of oncogenes has long been at the focus of drug development studies in cancer. Small molecules of inhibitors of oncogene functions such as tyrosine kinase inhibitors have already been used in the routine treatment of various cancers. Gene therapeutic strategies to suppress oncogene functions are usually focused on the inhibition of those genes at mRNA level. Usually small oligonucleotides or RNA inhibitors such as short-interfering RNA (siRNA), short-hairpin RNA (shRNA), or micro-RNA (miRNA) have been used to interfere

Chemically modified or unmodified small single-stranded DNA molecules, antisense oligo‐ nucleotides inhibit protein translation through the disruption of ribosome assembly or utilization of RNase H enzymes to destroy mRNA. Numerous oligonucleotides and RNA inhibitors have been designed to inhibit oncogenes, including RAS, MYC, BCL-2, or cell signaling molecules survivin, IGF, VEGF, and PKCalphfa, have been tested. Although the efficacy of these oligonucleotides has shown a great diversity, some of them have been tested in phase II/III clinical trials in various cancer types [115]. Oblimersen, an antisense oligonu‐ cleotide targeting Bcl-2, is one of the oldest agents that have already tested in phase III studies of Chronic Lymphocytic Leukemia CLL and multiple myeloma [116,117]. The members of the RAS family of oncogenes have been found mutated in various solid tumors. Therefore, the targeting of RAS would have been a hot topic in the development of recent therapeutics.

pathway, have been shown to induce apoptosis of tumor cells [108,109].

clinical setting yet.

12 Gene Therapy - Principles and Challenges

**7.2. Oncogenes**

the actions of oncogenes [114].

been shown to slow the growth of tumor cells.

Conventional chemotherapeutic drugs are mainly directed to nonspecific direct cell killing. However, dose-limiting toxicities avoid the use of higher doses of those drugs to eradicate the disseminated cancer. However, if the drug was synthesized within the tumor tissue, then the toxicity level would only increase in tumor cells but not other parts of the body. The tumorspecific targeting of drug-metabolizing genes and the systemic use of a prodrug that is converted to a cytotoxic agent by the action of transduced enzyme called gene-directed enzyme/prodrug therapy (GDEPT) enable the achievement of that aim. GDEPT is also known as suicide gene therapy. A lot of drug-metabolizing genes have been used to develop suicide gene therapy/prodrug systems. Cytosine deaminase (CD) and herpes simplex virus 1 thymi‐ dine kinase (HSV1-TK) are the most widely studied ones in cancer gene therapy [121,122]. CD, an enzyme found in fungi and bacteria, converts the nontoxic 5-fluorocyotsine into a toxic chemotherapy drug of 5-fluorouracil. The lack of this enzyme in mammalian cells makes it a convenient gene therapy tool to achieve intaratumoral chemotherapy. Others and we have designed suicide gene therapy vectors to avoid systemic toxicity of 5-FU. We have shown that Lp-driven CD carrying adenoviral vectors (AdLpCD) specifically target the epithelial cancers, including breast, ovary, prostate, and lung [123]. It is possible to achieve a 5-FU dose in tumor tissue as much as 200-fold of the dose when the drug is used intravenously at the standard dose [123]. The 5-FU produced in the infected tumor cells can diffuse into the neighboring tumor cells and kill them even not infected by the vector, which is called bystander effect [124]. Likewise, the combination of CD carrying vectors with conventional chemotherapy or radiotherapy yields synergistic efficacy [125–127].

TK, one of the immediate early (IE) genes of HSV, converts ganciclovir (GCV) into a tri‐ phosphated form of GCV, which is an analogue of purine and inhibits DNA polymerase [128]. HSV1-TK suicide gene therapy loaded onto either adenoviral vectors or retroviral vectors has been used to treat various tumors, including pancreatic cancer, hepatocellular carcinoma, lung cancer, glioma, and leukemia [129–133]. Although the exact mechanism of HSV-TK carrying vectors to kill tumor cells is not completely understood, they can induce apoptosis sensitizing the TNF-related ligands or the sensitization of CD95-L, TNF-related apoptosis inducing ligands may contribute to cell death [134]. The transcriptional targeting of HSV1-TK vectors using tumor-specific promoters has decreased the potential side effects [130]. HSV-TK/GCV prodrug systems have also been modified with other genes such as addition of E-cadherin to increase the bystander effect of the vector [129].

Other prodrug-activating enzymes such as purine nucleoside phosphorylase to convert 6 methylpurine-2-deoxyriboside to 6-methyl purine, cytochrome p450 cyclophosphamide and ifosfamide to active metabolites of phosphoramide mustard and acrolein cyanide, and carboxypeptidase methotrexate-alpha peptides to methotrexate have also been reported to decrease tumor burden in various preclinical models [135–137].

Dying tumor cells during suicide gene therapy could induce a tumor-specific immune response. Therefore, combining prodrug/enzyme systems with an immnuomodulating cytokine would further improve the efficacy. The addition of an IL-2 gene to the HSV-TK has yielded more potent antitumoral activity when compared the each strategy alone [138]. Similarly, GM-CSF, IL-12, and IL-18 have also been used to increase the antitumoral activity of suicide gene therapy [139,140]. Suicide gene therapy also successfully combines with other strategies such as targeting tumor angiogenesis or adoptive transfer [141,142].

#### **7.4. Oncolytic viral vectors**

Viruses have long been recognized tumor cell lytic agents and tried to treat cancer patients. However, the use of unmodified oncolytic viruses usually failed in the clinic. The engineering of those viruses to increase their therapeutic index have been possible in the last two decades. Herpes simplex virus (HSV), adenoviruses, parvoviruses, Newcastle disease virus, and retroviruses have been modified as oncolytic viral vectors.

HSV with its high infective capacity of a large number of cell types has been one of the popular oncolytic agents in the treatment of cancers. By deleting the genes thymidine kinase (TK), ribonucleotide reductase (RR), or ICP34.5 alone or in combination, HSV vectors could be selectively targeted many cancer types [143,144]. In order to further increase the cancer cell specificity of the replicating vector, engineering of the expression of surface glycoproteins, attachment of a novel receptor, or other macromolecules such as bispecific antibodies have been tested [145]. Likewise, tumor cell-specific promoters to drive the immediate–early gene expression, which is essential for viral replication, has been another effective strategy to obtain tumor-selective HSV [146].

Adenoviruses can infect a wide variety of dividing and nondividing normal and tumor cells. They can be engineered to have tumor-selective oncotropic properties or to be conditionally replicative (CRAds) for selective cancer gene therapy.

In type I CRAds, usually a mutant Ad vector that replicates specifically in tumor cells with aberrant cell cycle regulation has been developed. A deletion in the E1B 55-kDa region abrogates the p53 binding of the vector, and therefore, the vector cannot replicate in cells with intact p53 [147]. Therefore, this mutant Ad vector (dl1520) could replicate in only p53-deficient tumor cells. However, further studies revealed that E1B 55-kDa mutant CRAds could also replicate in p53 intact tumor cells [148,149]. The CRAds are already tested in phase II/III clinical trials with some success in patients with p53-deficient tumors [150]. Accordingly, the combi‐ nation of CRAds with conventional treatment modalities provided better tumor control [151]. Although the combination of E1B-55kD mutant Ad vector with chemotherapy has yielded a promising result of 63% partial response in patients with head and neck cancer administered intratumorally [151], no objective responses were seen when the vector used alone [152,153].

Another way to achieve tumor-specific adenoviral replication is to take the advantage of altered cell cycle regulation at G1-S phase checkpoint in which the retinoblastoma 1 (RB1] gene functions. In most of the cancer cells, there is a mutation in RB1 gene. Therefore, an Ad vector having a mutation in the RB-binding site of E1A cannot induce the quiescent cells to pass the checkpoint. A mutant CRAd carrying an E1A deletion, Ad5-∆24, is unable to replicate in normal cells with the wild-type RB1 gene [154]. It has been shown that this E1A mutant Ad vector has strong oncolytic activity in in vitro experiments of glioblastoma cells. Also, a similar vector with E1A mutations at RB-binding sites (dl922-947) has also been shown to have strong antitumor activity in other tumor models such as breast and colon cancer [155]. An additional promising strategy to achieve specific oncolytic activity to the CRAds is the use of tumorspecific promoters that drive the genes of the vector responsible for the replication, referred to as type II CRAds. There have been many replication-competent vectors carrying tumor- or tissue-specific promoters such as prostate-specific antigen (PSA), alphafeto protein (AFP), Tcf4, MUC1, and CEA that have been developed [156–160]. We have designed replicationcompetent adenoviral vectors carrying Lp-driven E1A, which are specifically replicated in various tumor cell lines but not in normal cells [161]. We have also constructed a bicistronic CRAd vector carrying both cytosine deaminase (CD) gene and E1A linked by an IRES component driven by the Lp promoter (AdLpCDIRESE1A) [162]. The new bicistronic construct also has been shown to have significant oncolytic activity in the colon (HTB-38), breast (MCF-7), ovary (Ovcar 5], and prostate (LNCaP) cancer cell lines but not in normal human mammary epithelial cells [162]. Also, the combination of the construct, AdLpCDIRESE1A/ 5flourocytosine system, and chemotherapy has shown synergistic activity [163].

carboxypeptidase methotrexate-alpha peptides to methotrexate have also been reported to

Dying tumor cells during suicide gene therapy could induce a tumor-specific immune response. Therefore, combining prodrug/enzyme systems with an immnuomodulating cytokine would further improve the efficacy. The addition of an IL-2 gene to the HSV-TK has yielded more potent antitumoral activity when compared the each strategy alone [138]. Similarly, GM-CSF, IL-12, and IL-18 have also been used to increase the antitumoral activity of suicide gene therapy [139,140]. Suicide gene therapy also successfully combines with other

Viruses have long been recognized tumor cell lytic agents and tried to treat cancer patients. However, the use of unmodified oncolytic viruses usually failed in the clinic. The engineering of those viruses to increase their therapeutic index have been possible in the last two decades. Herpes simplex virus (HSV), adenoviruses, parvoviruses, Newcastle disease virus, and

HSV with its high infective capacity of a large number of cell types has been one of the popular oncolytic agents in the treatment of cancers. By deleting the genes thymidine kinase (TK), ribonucleotide reductase (RR), or ICP34.5 alone or in combination, HSV vectors could be selectively targeted many cancer types [143,144]. In order to further increase the cancer cell specificity of the replicating vector, engineering of the expression of surface glycoproteins, attachment of a novel receptor, or other macromolecules such as bispecific antibodies have been tested [145]. Likewise, tumor cell-specific promoters to drive the immediate–early gene expression, which is essential for viral replication, has been another effective strategy to obtain

Adenoviruses can infect a wide variety of dividing and nondividing normal and tumor cells. They can be engineered to have tumor-selective oncotropic properties or to be conditionally

In type I CRAds, usually a mutant Ad vector that replicates specifically in tumor cells with aberrant cell cycle regulation has been developed. A deletion in the E1B 55-kDa region abrogates the p53 binding of the vector, and therefore, the vector cannot replicate in cells with intact p53 [147]. Therefore, this mutant Ad vector (dl1520) could replicate in only p53-deficient tumor cells. However, further studies revealed that E1B 55-kDa mutant CRAds could also replicate in p53 intact tumor cells [148,149]. The CRAds are already tested in phase II/III clinical trials with some success in patients with p53-deficient tumors [150]. Accordingly, the combi‐ nation of CRAds with conventional treatment modalities provided better tumor control [151]. Although the combination of E1B-55kD mutant Ad vector with chemotherapy has yielded a promising result of 63% partial response in patients with head and neck cancer administered intratumorally [151], no objective responses were seen when the vector used alone [152,153]. Another way to achieve tumor-specific adenoviral replication is to take the advantage of altered cell cycle regulation at G1-S phase checkpoint in which the retinoblastoma 1 (RB1] gene

strategies such as targeting tumor angiogenesis or adoptive transfer [141,142].

decrease tumor burden in various preclinical models [135–137].

retroviruses have been modified as oncolytic viral vectors.

replicative (CRAds) for selective cancer gene therapy.

**7.4. Oncolytic viral vectors**

14 Gene Therapy - Principles and Challenges

tumor-selective HSV [146].

Different replication-competent viruses are currently being studied for their potential use in cancer gene therapy. The naturally occurring tumor-selective viruses in their replication and cytolysis might have the potential in cancer treatment. Autonomous parvoviruses (APV) have been shown to replicate more efficiently in transformed cells than normal cells [164]. The members of the rodent group of APVs such as LuIII, MVM (minute virus of mice), and H1, which can infect human cells, are currently being studied as vectors for cancer gene therapy. The replication of APV depends on cellular functions expressed during the S phase of the cell cycle. The oncogenic transformation of cells favor the replication of APVs and therefore makes them as oncolytic viruses [165]. The overexpression of the RAS signaling pathway [166] and the defects in the interferon pathway of the transformed cells [167] could possibly enhance the oncolytic activity of the APVs. Further manipulation of the specific targeting of those vectors to achieve tumor-specific transgene expression such as inserting binding sites for the hetero‐ dimer beta-catenin/Tcf transcription factor to the MVM P4 promoter to make it responsive to wnt signaling would make those attractive vectors for cancer gene therapy [168].

Newcastle disease virus (NDV) is an animal virus showing oncolytic activity in transformed cells. In murine tumor xenograft models, the intratumoral administration of NDV has caused significant tumor reduction [169]. Also, the intraperitoneal injection of the virus has resulted in complete regressions of tumor xenografts. A replication-competent strain of NDV, PV701, has been shown to replicate in tumor tissues of patients with solid tumors when administered intravenously [170]. In that phase, trial objective responses have also been achieved at higher and repeated doses of the virus.

The murine hepatitis coronavirus (MHV), an oncolytic virus, is a positive-strand RNA virus displaying strong species specificity with a replication cycle of 10–15 h and efficiently kills cells by fusion of the infected cells with their neighboring cells [171]. Substituting its spike protein by the other species such as porcine amino peptidase could change the host cell tropism of the MHV. The resulting recombinant corona virus pMHV thus only infects porcine cells via the porcine amino peptide N (pAPN) receptor. In vitro studies have shown that the tumor cells could be more susceptible to that recombinant corona virus [172]. It is also likely to further manipulate those vectors by using specific antibodies.

#### **7.5. Tumor vascular targeting therapy**

Unraveling the mechanisms of tumor-induced angiogenesis, which is a key event in tumor growth and metastasis, has opened a new therapeutic era in cancer treatment. The antiangio‐ genic gene therapy approaches have been reported to inhibit the tumor-induced angiogenesis and therefore tumor growth. The main strategies in antiangiogenic gene therapy are targeting specifically the endothelial cells (direct antiangiogenic gene therapy) and interfering with a tumor-derived angiogenic factor or the receptor for it or delivery of genes that encode angiogenesis inhibitors (indirect antiangiogenic therapy).

Proangiogenic cytokines such as vascular endothelial growth factor (VEGF) and basic fibro‐ blast growth factor (bFGF) mainly secreted from tumor cells are required for the new vessel formation. The indirect strategies were mainly focused on the inhibition of proangiogenic cytokines or receptors involved in VEGF pathway or basic fibroblast growth factor (bFGF). VEGF binds two high affinity receptors (VEGFR1/FLT-1 and VEGFR-2/KDR) that are ex‐ pressed on endothelial cells. An adenovirus-mediated transfer of a secreted form of the extracellular domain of the FLT-1 (AdsFLT) has been shown to inhibit the growth of metastatic tumor deposits when administered intravenously to preestablished splenic and liver meta‐ stases from a murine colon carcinoma cell line in syngeneic mice [173].

Likewise, the delivery of genes encoding antiangiogenic proteins such as endostatin, angios‐ tatin, platelet factor 4, interferon alpha, and thrombospondins have also been tested [174]. The intratumoral administration of a plasmid encoding murine endostatin under the control of a CMV promoter has provided elevated concentrations of endostatin high enough to obtain growth arrest of murine renal carcinoma cells and breast cancer model [175]. Likewise, an adenoviral vector carrying human endostatin gene markedly reduced the blood vessel density of the tumor in an orthotopic liver tumor model [176].

The viral vector constructs of other angiogenesis inhibitors such as angiostatitn, thrombo‐ spondin, platelet factor 4, and hepatocyte growth factor antagonists have also been shown to successfully inhibit endothelial cell proliferation and tumor growth [177–180]. However, there are conflicting results regarding the tumor inhibiting activity of antiangiogenic gene therapy modality in experimental models. The combination of antiangiogenic gene therapy with chemotherapy or radiation could be an efficient way of the inhibition of tumor growth [181].

Many vector constructs carrying therapeutic or reporter genes driven by endothelium-specific promoters such as preproproendothelin-1 (PPE-1), VEGFR kinase insert domain receptor (KDR), VEGF, E-selectin, and endoglin/CD105 have been reported to specifically target endothelial cells [182,183]. The replication-competent adenoviral vectors driven by the regulatory elements of FLK-1 and endoglin have successfully been targeted to the dividing endothelial cells, and therefore, this strategy could be used as an antiangiogenic treatment for cancer [184]. The activation of proapoptotic caspases such as caspase 9, driven by endotheliumspecific promoters such as VEGF and FGF, could be another strategy to destroy endothelial cells [185].

Antisense approaches also are being tested for the inhibition of VEGF. A recombinant adenoassociated virus (rAAV) vector encoding an antisense mRNA against VEGF has been shown to inhibit the production of endogenous tumor cell VEGF [186]. The adenovirus-mediated delivery of an uPA uPAR antagonist, which inhibits FGF, has been shown to inhibit angio‐ genesis-dependent tumor growth and metastasis in mice [187].

#### **7.6. Immune system as the target of cancer gene therapy**

The murine hepatitis coronavirus (MHV), an oncolytic virus, is a positive-strand RNA virus displaying strong species specificity with a replication cycle of 10–15 h and efficiently kills cells by fusion of the infected cells with their neighboring cells [171]. Substituting its spike protein by the other species such as porcine amino peptidase could change the host cell tropism of the MHV. The resulting recombinant corona virus pMHV thus only infects porcine cells via the porcine amino peptide N (pAPN) receptor. In vitro studies have shown that the tumor cells could be more susceptible to that recombinant corona virus [172]. It is also likely to further

Unraveling the mechanisms of tumor-induced angiogenesis, which is a key event in tumor growth and metastasis, has opened a new therapeutic era in cancer treatment. The antiangio‐ genic gene therapy approaches have been reported to inhibit the tumor-induced angiogenesis and therefore tumor growth. The main strategies in antiangiogenic gene therapy are targeting specifically the endothelial cells (direct antiangiogenic gene therapy) and interfering with a tumor-derived angiogenic factor or the receptor for it or delivery of genes that encode

Proangiogenic cytokines such as vascular endothelial growth factor (VEGF) and basic fibro‐ blast growth factor (bFGF) mainly secreted from tumor cells are required for the new vessel formation. The indirect strategies were mainly focused on the inhibition of proangiogenic cytokines or receptors involved in VEGF pathway or basic fibroblast growth factor (bFGF). VEGF binds two high affinity receptors (VEGFR1/FLT-1 and VEGFR-2/KDR) that are ex‐ pressed on endothelial cells. An adenovirus-mediated transfer of a secreted form of the extracellular domain of the FLT-1 (AdsFLT) has been shown to inhibit the growth of metastatic tumor deposits when administered intravenously to preestablished splenic and liver meta‐

Likewise, the delivery of genes encoding antiangiogenic proteins such as endostatin, angios‐ tatin, platelet factor 4, interferon alpha, and thrombospondins have also been tested [174]. The intratumoral administration of a plasmid encoding murine endostatin under the control of a CMV promoter has provided elevated concentrations of endostatin high enough to obtain growth arrest of murine renal carcinoma cells and breast cancer model [175]. Likewise, an adenoviral vector carrying human endostatin gene markedly reduced the blood vessel density

The viral vector constructs of other angiogenesis inhibitors such as angiostatitn, thrombo‐ spondin, platelet factor 4, and hepatocyte growth factor antagonists have also been shown to successfully inhibit endothelial cell proliferation and tumor growth [177–180]. However, there are conflicting results regarding the tumor inhibiting activity of antiangiogenic gene therapy modality in experimental models. The combination of antiangiogenic gene therapy with chemotherapy or radiation could be an efficient way of the inhibition of tumor growth [181].

Many vector constructs carrying therapeutic or reporter genes driven by endothelium-specific promoters such as preproproendothelin-1 (PPE-1), VEGFR kinase insert domain receptor

manipulate those vectors by using specific antibodies.

angiogenesis inhibitors (indirect antiangiogenic therapy).

of the tumor in an orthotopic liver tumor model [176].

stases from a murine colon carcinoma cell line in syngeneic mice [173].

**7.5. Tumor vascular targeting therapy**

16 Gene Therapy - Principles and Challenges

The immune system is the most important defense mechanism of the body against cancer. Recent developments in gene therapy have suggested to many cancer therapists that cytokine– chemokine-based gene therapies, tumor antigen-specific vaccination strategies, and genemodified cellular therapies have great potential for future use either in the treatment of an established disease or in the prevention of cancer in people having high risk of developing cancer.

Cytokine/chemokine-based gene therapy has been widely used to induce immune system against tumors. The delivery of immunomodulatory cytokines by gene therapy vectors has opened a new avenue both to decrease the toxicity of these cytokines when used systemically and to augment antitumor immunity. A wide variety of cytokines such as GM-CSF, IFN-a, IFN-g, IL-2, IL-4, IL-12, IL-18, and IL-24 have been tested so far [188–191]. Also, the vector constructs, including the combination of these cytokines, have also been tested in cancer. The coexpression of IL-12 and GM-CSF has been reported to yield significantly more immune response than the either cytokines alone [192]. In particular, implementing the cytokine genes into oncolytic viruses has great potential for use in clinical trials [193]. Chemokines recruit the immune effector cells to the tumor microenvironment. The delivery of chemokines such as CCL-5 using viral vectors has also resulted in significant tumor reduction through increasing tumor infiltration of DCs, macrophages, and CTLs [194].

Tumor-associated antigens (TAA) loaded on to gene therapy vectors have been tested in cancer treatment (DNA vaccines) [195,196]. However, the efficacy of using TAA alone is not enough to get a sufficient immune response to decrease tumor size. Therefore, researchers have focused on the augmentation of the immune response by combining immune cytokines or costimula‐ tory molecules and TAA. This strategy seems much better than using either gene alone. We have previously shown an increased efficacy of an adenoviral vector encoding a fusion protein of CD40L and MUC1 in preclinical models [197]. The addition of prodrug/enzyme system to DNA vaccination further increased the efficacy [198]. This strategy has also been tested in early clinical trials with some success. Vector vaccinations using cytokines or costimulatory molecules and tumor-associated antigens (TAA) have increased the immune responses and caused antitumor responses in preclinical models and even some responses in earlier clinical trials. In a small clinical trial, an attenuated vaccinia vector carrying IL-2 and MUC1 has been found effective in a small group of patients with advanced prostatic cancer [199]. Likewise, a vector vaccine of canary poxvirus encoding B7.1 and CEA has been tested in a group of patients with epithelial tumors [200]. Hundreds of different DNA vaccines have been tested in clinical trials so far [202]. However, no DNA vaccine is available in the market.

Gene therapy vectors have also been used to transduce either autologous tumor cells or dendritic cells. In the earlier studies, irradiated autologous tumor cells transduced to express immunostimulatory molecules have been tested. In a syngeneic colon cancer model, the subcutaneous injection of CT26 colon cancer cells transduced with an adenoviral vector carrying GM-CSF gene has eliminated both the established tumors and prevented the growth of new tumor nodules when rechallenged with tumor cells [201]. Later on, this strategy has also been tested in human tumors. Autologous tumors transduced with GVAX, an adenovirus carrying GM-CSF, have induced tumor-specific immunity in a variety of tumors, including melanoma, prostate, and lung cancers [203]. Although a slight increase in overall survival has been reported in those trials, no significant tumor responses observed [203,204].

The ex vivo transduction of dendritic cells with gene therapy vectors carrying either immunostimulatory genes or TAAs is another promising strategy. When injected subcuta‐ neously, the dendritic cells exposed to vectors migrate to the lymph nodes where they prime cytotoxic T cells and induce a strong immune response. A number of vectors have been designed to activate dendritic cells for the past two decades. We have tested the use of ex vivo transduced dendritic cells with an adenoviral vector carrying a fusion protein of CD40L and MUC1 in a syngeneic mouse tumor model [205]. The intratumoral injec‐ tion of activated dendritic cells induced a potent tumor-specific T-cell response. Further‐ more, the combination of suicide gene therapy of a CD/5FU system and activated dendritic cells caused a more potent immune response and increased tumor response [205]. Like‐ wise, retroviral vectors and lentiviral vectors are both used to transduce dendritic cells [206]. A dendritic cell vaccine based on the ex vivo activation of mononuclear antigen presenting cells by a fusion protein consisting prostatic acid phosphatase and GM-CSF has extended the progression-free survival of patients with advanced prostatic cancer and approved by FDA in 2010 (Provenge®, Dendreon, USA) [207].

Recently, an adoptive therapy of cancer using genetically modified T cells armed with chimeric antigen receptors (CAR) has gained great popularity with the announcement of success in advanced malignancies [208]. CAR is a fusion receptor of an antibody-derived targeting domain and T-cell signaling domain and expressed on T cells by a retroviral vector [209]. CARs target antigens, including proteins, carbohydrates, and glycolipids without antigen processing or HLA recognition. They can be generated in significant quantities ex vivo and used with the minimal risk of autoimmunity or graft versus host disease [210,211]. However, because of the severe side effects, the most troublesome being cytokine-release syndrome, researchers try to obtain better CAR T cells with further refinement of receptor and better targets [212].
