**4. Gene transfer systems of cancer gene therapy**

There are three main ways of transferring genes into the tumor cells: nonviral vectors, viral vectors, and cell-based vehicles. For most of the tumors, a relatively short-term expression of therapeutic genes may be sufficient to kill the tumor cells. Rapid clearance of viral vectors from the blood stream has enabled the development of synthetic gene delivery vectors. However, an important drawback for these approaches is to carry the DNA of interest to the distant metastatic deposits. The nonviral gene delivery vectors have usually been injected locally to the tumors. Although local injection is reasonable for tumors as melanoma, head and neck cancers, or peritoneal carcinomatosis; it is not suitable in patients with hematogenous meta‐ stases. The limitations of the viral vectors are also valid for the nonviral vectors for gene therapy. They have to survive through the blood stream to be arrested in the target tumor tissue, to extravasate, and to bind to specific cells and to enter the cells and then to reach the nucleus.

#### **4.1. Nonviral vectors**

*Plasmid DNA*, which is mostly used as nonviral gene therapy modality, is easily degraded by nucleases [11]. Therefore, some strategies to reduce the size and prevent the degradation have been developed. The most commonly used agents for gene delivery are *cationic lipids* [12]. The cationic head group of the lipids binds to DNA and the lipid tail enables the collapse of the DNA lipid complex [13]. Cationic lipid DNA complexes (lipoplexes) (LPD/DNA) enter the target cell through an endosomal pathway. However, the transgene expression efficiency is very low with lipoplexes. It has been shown that only a very small portion of the systemically injected DNA could be reached to tumor tissue [14].

Lipid-based formulations of gene delivery have been predominantly limited to the intratu‐ moral or local applications. The systemic administration carries the potential risk of adverse inflammatory and immune reactions. The development of systemic lipid delivery systems with the modifications to reduce the systemic toxicity could have the potential for clinical use in cancer gene therapy. In an animal model of breast cancer, folate-targeted lipid–protamine DNA complexes (LPD-PEG-folate) have been shown to reduce the tumor volume and increase the survival when administered systemically [14].

*Neutral liposomes* composed of DOPC (1,2-dioleyl-sn-phosphatidyl choline) and DOPE (1,2 dioleyl-sn-phosphatidyl ethanol amine) and polycationic carrier proteins as protamine, polylysine, polyarginine, polyhistidine, or polyethynilemine (PEI) are also suitable to carry the DNA [15–18]. The *hydrophobic polymers*, such as polyethylene glycol (PEG), polyhy‐ droxy propylmethacrylamide (pHPMA), and polyvinyl pyrrolidine (pVPyrr), have also been used to mask the positive charge of DNA to extend its half-life in the blood [19,20]. Both the neutral liposomes and hydrophobic polymers yield less toxicity when administered systemically. The leaky nature of the blood vessels of the tumors allows the influx of macromolecules as polymer shielded DNA into the tumor. The PEGylation of plasmid DNA has been reported to circulate in the blood several hours and passively accumulate in the subcutaneous tumors in animals [21].

#### **4.2. Viral vectors**

For the safety of the procedure and the increased therapeutic efficacy, the genes of interest should be expressed in only target cells or tissues. Sparing of the normal cells and tissues is one of the keystones in their clinical use. The target specificity of the vectors could be achieved

There are three main ways of transferring genes into the tumor cells: nonviral vectors, viral vectors, and cell-based vehicles. For most of the tumors, a relatively short-term expression of therapeutic genes may be sufficient to kill the tumor cells. Rapid clearance of viral vectors from the blood stream has enabled the development of synthetic gene delivery vectors. However, an important drawback for these approaches is to carry the DNA of interest to the distant metastatic deposits. The nonviral gene delivery vectors have usually been injected locally to the tumors. Although local injection is reasonable for tumors as melanoma, head and neck cancers, or peritoneal carcinomatosis; it is not suitable in patients with hematogenous meta‐ stases. The limitations of the viral vectors are also valid for the nonviral vectors for gene therapy. They have to survive through the blood stream to be arrested in the target tumor tissue, to extravasate, and to bind to specific cells and to enter the cells and then to reach the

*Plasmid DNA*, which is mostly used as nonviral gene therapy modality, is easily degraded by nucleases [11]. Therefore, some strategies to reduce the size and prevent the degradation have been developed. The most commonly used agents for gene delivery are *cationic lipids* [12]. The cationic head group of the lipids binds to DNA and the lipid tail enables the collapse of the DNA lipid complex [13]. Cationic lipid DNA complexes (lipoplexes) (LPD/DNA) enter the target cell through an endosomal pathway. However, the transgene expression efficiency is very low with lipoplexes. It has been shown that only a very small portion of the systemically

Lipid-based formulations of gene delivery have been predominantly limited to the intratu‐ moral or local applications. The systemic administration carries the potential risk of adverse inflammatory and immune reactions. The development of systemic lipid delivery systems with the modifications to reduce the systemic toxicity could have the potential for clinical use in cancer gene therapy. In an animal model of breast cancer, folate-targeted lipid–protamine DNA complexes (LPD-PEG-folate) have been shown to reduce the tumor volume and increase

*Neutral liposomes* composed of DOPC (1,2-dioleyl-sn-phosphatidyl choline) and DOPE (1,2 dioleyl-sn-phosphatidyl ethanol amine) and polycationic carrier proteins as protamine, polylysine, polyarginine, polyhistidine, or polyethynilemine (PEI) are also suitable to carry the DNA [15–18]. The *hydrophobic polymers*, such as polyethylene glycol (PEG), polyhy‐ droxy propylmethacrylamide (pHPMA), and polyvinyl pyrrolidine (pVPyrr), have also been

by the targeting of those specific to the tumor cells or tissues.

**4. Gene transfer systems of cancer gene therapy**

injected DNA could be reached to tumor tissue [14].

the survival when administered systemically [14].

nucleus.

**4.1. Nonviral vectors**

4 Gene Therapy - Principles and Challenges

Viruses have the natural ability to deliver the nucleic acids within its own genome to specific cell types, including cancer cells. This ability makes those attractive and popular gene-delivery vehicles. Retroviruses, adenoviruses, adeno-associated viruses, herpes simplex virus, poxvi‐ ruses, and baculoviruses are commonly modified and used as gene therapy vectors in cancer. Additionally, chimeric viral-vector systems combining the properties of two or more virus type are also developed.

*Retroviral vectors* derived from retroviruses contain a linear single-stranded RNA of around 7– 10 kb and have a lipid envelope. The viral particles enter the mammalian cells expressing appropriate receptors for retroviruses [22]. After entering the cell, the viral reverse transcrip‐ tase transcribes the virus RNA into double-stranded DNA (dsDNS). The dsDNA transcribed in the cytoplasm forms a nucleoprotein preintegration complex (PIC) by binding cellular proteins [23]. The PIC migrates to the nucleus and thereby integrates the host genome. The ability of transgene expression in only dividing cells is an advantage of retroviral vectors for cancer gene therapy to avoid undesired expression in nondividing cells of surrounding tissues. The incorporation of retroviral genes into the host genome provides long-term expression of transgenes. Although this is advantageous, a nonspecific incorporation of viral DNA could impair the function of host gene or induce aberrant expression of a cellular oncogene [24]. Although retroviral vectors have been the most widely used gene transfer vehicles in the clinic, the risk of insertional oncogenesis seen in the trial of X-SCID infants in 2003 has limited the use of retroviral gene transfer systems in humans [25]. The possibility of generating replicationcompetent retroviruses is another safety issue regarding the clinical use of those vectors [26].

*Lentiviral vectors* derived from retroviruses can cause stable integration of the transgene into the host genome with long-term gene expression. The ability of transducing both dividing and nondividing cells make those vectors more suitable and efficient gene transfer vehicle over retroviruses. Targeting strategies of vectors at the level of cell entry and transgene transcription improved the use of lentiviral vectors in gene therapy trials [27]. However, the biosafety concerns of random integration to the host genome as in retroviruses are the limitations of those vectors.

*Adenoviral vectors* are widely used to introduce the therapeutic genes into the tumor cells. They can infect a broad range of cell types, transfer the genes being not dependent on cell division, and have high titers and high level of gene expression [28]. The most widely used serotypes of adenoviruses to develop vectors in human cancer gene therapy studies are type 5 (Ad5) and type 2 (Ad2). They have the capacity of approximately 8–10 kb of therapeutic genes with firstgeneration vectors and up to 36 kbp with gutless third generation adenoviral vectors [29]. However, along with the immunogenic potential, the broad range of host cells by adenovirus limits its systemic use in human cancer gene therapy trials [30]. Targeting strategies have enabled the use of adenoviral vectors in human gene therapy trials. Adenoviral vectors cannot integrate to cellular genome and express the transgene episomally. They cannot induce random mutations. However, the transgene expression is limited to 7–10 days postinfection [31]. Therefore, repeated administrations of the vector are needed to achieve sustainable responses in cancer treatment. Adenoviruses could be engineered either as replication deficient by deleting the immediate early genes of E1 or replication-competent keeping the E1 region. Replication-competent adenoviral vectors will be further discussed in the section of oncolytic viruses.

*Adeno-associated viruses (AAV)* are simple viruses with approximately single-stranded DNA of 4.7 kb in size [32]. They belong to parvovirus family and require a helper virus such as adenovirus or herpes virus for lytic replication and release from the cell [33]. They can infect a wide variety of cells independent of cell cycle. This property makes AAV as suitable vectors for cancer gene therapy. Furthermore, unlike adenoviruses, they elicit little immune response when infect the normal host cells. Another advantage of AAV over adenoviruses is their ability to integrate the transgene into a particular spot on the 19th chromosome of human cells [34]. Unlike retroviruses, AAV cannot induce mutations. However, the major drawback of AAV is its limited cargo capacity of approximately 4 kbp of therapeutic genes. AAV could transduce certain cell types. Therefore, targeting strategies such as modification of viral capsid proteins, binding monoclonal antibodies, or bispecific proteins have been developed to improve the efficiency of AAV systems in cancer gene therapy [35,36].

*Baculoviruses* are enveloped viral particles with a large dsDNA of approximately 80–180 kb. They naturally infect insect cells. There have been no diseases related to baculoviruses in humans. Along with their highly safety profile in humans, they seem very useful gene therapy vehicles with their highly large cargo capacity of approximately 40 kb with possible multiple inserts, easy manipulation, and production [37]. *Autographa californica* multiple nucleopoly‐ hedrovirus (*Ac*MNPV) is the most widely used types of baculovirus in gene therapy studies. It has a circular dsDNA genome of 135 kb [38]. They can easily transduce mammalian cells, including many types of cancer cells, and cause high transgene expression in the host cell [39]. They are already approved for the production of human vaccine components such as Cervarix (GlaxoSmithKline) in cervical cancer and Provenge (Dendreon) in prostatic cancer [40].

*Herpes simplex virus (HSV)* is a large DNA virus with approximately 152 kb of dsDNA genome. It has a natural tropism to nerve tissues and cannot integrate into the host genome [41]. The HSV vectors can be designed in three different types as amplicons, replication-defective, and replication-competent vectors [42]. In general, the replication-competent HSV vectors are used as oncolytic agents in cancer gene therapy studies [43].

*Poxviruses* were the first viruses to be used as gene therapy vectors. They have been used in the in vitro production of proteins and as live vaccines. The attenuated forms of poxviruses have been developed and used in the development of genetic cancer vaccine trials [44]. The immunostimulatory properties of poxviruses make them preferable agents to induce immun‐ ity against tumors. In particular, the attenuated MVA virus derived from chorioallantoid vaccinia Ankara (CVA), a Turkish smallpox vaccine strain, has been widely used in cancer vaccine development strategies [45].
