**1. Introduction**

454 Non-Viral Gene Therapy

Yang, P.T., Hoang L., Jia, W.W. & Sharsgard, E.D. (2011). In utero gene delivery using chitosan-DNA nanoparticles in mice. *Journal of Surgical Research*, (in press) Zhu, D., Jin, X., Leng, X., Wang, H., Bao, J., Liu, W., Yao, K. & Song, C. (2010). Local gene

nanoparticles. *International Journal of Nanomedicine,* 5(1): 95-102.

delivery via endovascular stents coated with dodecylated chitosan-plasmid DNA

Human diseases can be treated by the transfer of therapeutic genes (transgene) into specific cells or tissues of patients to correct or supplement defective, causative genes. Gene therapy offers a solution to controlled and specific delivery of genetic materials (DNA and RNA) to targeted cells. The success of gene therapy depends on the ability to deliver these therapeutic materials to targeted site. Viral vectors (e.g. adenovirus) are very effective in term of transfection efficiency, but they have limitations *in vivo*, particularly by their safety concern and non tissue-specific transfection. Non-viral gene transfer systems are limited by their lower gene transfer efficiency, low tissue specificity and transient gene expression.

Chitosan is a polysaccharide usually obtained from deacetylation of chitin, which may be extracted from various sources, particularly from exoskeletons of arthropods such as crustaceans. The goal of this chapter is to introduce the readers to chitosan as a DNA/small interfering RNA (siRNA) delivery vector, as well as different variable strategies to improve cellular transfection and its potential clinical application. the first section is to present of chapter (section 1). The second section presents the discussion about barriers to DNA/siRNA delivery *in vitro* and *in vivo*. It is important to have a clear overview of obstacles to the *in vivo* treatment with DNA/siRNAs. Different *in vivo* administration routes will encounter different physiological barriers, and complications may be furthered by different cells in organs and tissues (section 2). The third section provides the readers with an understanding of the key steps of cellular internalization of DNA/siRNA non-viral vectors. Internalization of non-viral vector-based DNA/siRNA delivery system into cells typically occurs through endocytosis (section 3). the fourth section describes chitosan as a vector for gene therapy (section 4) followed by chitosan structure and physicochemical

Chitosan-DNA/siRNA Nanoparticles for Gene Therapy 457

be unsatisfactory. Research efforts to improve the *in vivo* DNA-delivery efficacy of non-viral

Systemic gene delivery involves a systemic approach in which exogenous genes are delivered to cells in a certain tissues, and secreted gene products are released into the circulatory system where they could modulate disease processes throughout the body. Systemic non-viral gene delivery has become an attractive alternative to viral vectors because of their safety, versatility and ease of preparation (Li & Huang, 2006). Genes can be delivered systemically (intramuscularly, intravenously, subcutaneously or, in animals, intraperitoneally). Otherwise, hydrodynamic-based gene delivery through systemic DNA injection offers a convenient, efficient and powerful means for high-level gene expression in animals (Liu & Knapp, 2001; Suda & Liu, 2007). This method is expected to be evaluated in patients soon (Romero et al., 2004). The limitations of the systemic approach to gene therapy are essentially the advantages of local delivery: exposure of non-target tissues to the therapeutic agent may have toxic effects or may compromise the immune system of the patient. Certain proteins will likely require very high levels of synthesis to achieve

Ideally, gene therapy must protect DNA against degradation by nucleases in intercellular matrices so that the availability of macromolecules is not affected. Transgenes should be brought across the plasma membrane and into the nucleus of targeted cells but should have no detrimental effects. Hence, interaction with blood components, vascular endothelial cells and uptake by the reticuloendothelial system must be avoided. For DNA-based gene therapy to succeed, small-sized systems must internalize into cells and pass to the nucleus. Also, flexible tropisms allow applicability to a range of disease targets. Last but not least, such systems should be able to escape endosome-lysosome processing for endocytosis.

The discovery of small interfering RNAs (siRNAs) has given renewed vision to the treatment of incurable diseases and genetically-associated disorders. Short double stranded (ds) RNA of 21-23 bp was cleaved by the RNAse III-like protein Dicer and incorporated into RNA-induced silencing complexes (RISC) (Hammond et al., 2000). Chemically-synthesized siRNAs and short hairpin RNA (shRNA) expression plasmids, which are sequence-specific for mRNA targeting, are methods commonly employed to mimic Dicer cleavage (Chen et al., 2007). However, siRNAs are susceptible to nuclease destruction and cannot penetrate the cell membrane because of their highly-charged backbone. An effective delivery system would enclose siRNA in carriers for protection and transport to the cytoplasm of targeted cells but should have no detrimental effects such as specific and non-specific off-targeted effects. Off-target effects can be divided into two categories: specific and non-specific offtargeted effects. Off-targeted effects may cause inflammation including interferon response,

Turning siRNA into drugs is a 3-step process. The design and *in vitro* screening of target siRNAs are followed by incorporating stabilizing chemical modifications in lead siRNAs, as required, and end in the selection as well as *in vivo* evaluation of delivery technologies that are appropriate for the target cell type/organ and disease setting (Vaishnaw et al., 2010). After nearly 10 years of study and development, many problems have been resolved, such as improving the stability of siRNAs, and avoiding 2 types of off-target effects. A recent

vectors are ongoing.

therapeutic function.

**2.4 Barriers to siRNA delivery** 

cell toxicity, and unintended gene knockdown.

**2.3 Barriers to DNA delivery** 

behaviour (section 5), general strategies for chitosan modification (section 6), chitosan-DNA delivery system (section 7), chitosan-siRNA delivery system (section 8), and potential applications of chitosan–DNA/siRNA nanoparticles (section 9). Our current research will be summarized in the section of conclusion.

### **2. Barriers to gene delivery using non-viral vectors**

#### **2.1 Viral gene vectors**

Gene transfer can occur through 2 delivery systems: viral or non-viral vectors. Viral gene therapy consists of using viral vectors which, given their structure and mechanisms of action, are good candidates or models to carry therapeutic genes efficiently, leading to longterm expression. Viruses are obvious first choices as gene transport. They have the natural ability to enter cells and express their own proteins. Nowadays, most viral vectors used are retroviruses, herpes virus, adenoviruses and lentiviruses. Unfortunately, certain viral vectors (for example, adenoviruses) can elicit a robust cellular immune response against viral and some transgenic proteins, so their use has been limited to studies in immunecompromised animals (Seiler et al., 2007). Adeno-associated viruses (AAV), which have been considered safe, appear to be immunogenic in several experimental settings (Vandenberghe & Wilson, 2007) and in a clinical trial (Mingozzi & High, 2007). Some serious adverse events have occurred with viral gene therapy. One patient died of fatal systemic inflammatory response syndrome after adenoviral gene transfer in 1999 (Raper et al., 2003). Two children developed leukemia-like clonal lymphocyte proliferation after recombinant retroviral gene transfer in 2000 (Hacein-Bey-Abina et al., 2003), and 1 of them died after unsuccessful chemotherapy late in 2004. Attention focused recently on the tragic death of a young female patient in a gene therapy study (intra-articular injection of AAV vectors) of severe RA in 2007 (Kaiser, 2007).

#### **2.2 Non-viral gene vectors**

Non-viral gene transfer systems offer several potential advantages over virus vectors. They are non-infectious, relatively non-immunogenic, have low acute toxicity, can accommodate large DNA plasmids or RNA, and may be produced on a large scale (Castanotto & Rossi, 2009; Gary et al., 2007). Non-viral gene therapy has been explored by physical approaches (transfer by gene gun, electroporation, ultrasound-facilitated and hydrodynamic delivery) as well as chemical approaches (cationic lipid-mediated gene delivery and cationic polymermediated gene transfer). Numerous chemical non-viral gene transfer systems have been proposed, including naked DNA, cationic liposomes, histones, and polymers (Gao et al., 2007; Ulrich-Vinther, 2007). The main drawback of non-viral vectors as gene carriers is their typically low transfection efficiency (Gao et al., 2007; Giannoudis et al., 2006). Furthermore, the *in vivo* delivery of non-viral liposome/plasmid DNA complex triggers an immune response (Sakurai et al., 2008). Non-viral gene therapy with cationic liposomes has already been tested in clinical trials that dealt with the treatment of inherited genetic disorders (for example, cystic fibrosis) (Hyde et al., 2000) and cancer (Ramesh et al., 2001). Synthetic and natural cationic polymers (positively-charged) have been widely used to carry DNA or siRNA (both negatively-charged) and condense it into small particles, facilitating cellular internalization via endocytosis through charge-charge interactions with anionic sites on cell surfaces. Although existing non-viral vectors have been found to enable DNA expression after *in vivo* delivery, the efficiency and duration of ensuing gene expression have proven to be unsatisfactory. Research efforts to improve the *in vivo* DNA-delivery efficacy of non-viral vectors are ongoing.

#### **2.3 Barriers to DNA delivery**

456 Non-Viral Gene Therapy

behaviour (section 5), general strategies for chitosan modification (section 6), chitosan-DNA delivery system (section 7), chitosan-siRNA delivery system (section 8), and potential applications of chitosan–DNA/siRNA nanoparticles (section 9). Our current research will be

Gene transfer can occur through 2 delivery systems: viral or non-viral vectors. Viral gene therapy consists of using viral vectors which, given their structure and mechanisms of action, are good candidates or models to carry therapeutic genes efficiently, leading to longterm expression. Viruses are obvious first choices as gene transport. They have the natural ability to enter cells and express their own proteins. Nowadays, most viral vectors used are retroviruses, herpes virus, adenoviruses and lentiviruses. Unfortunately, certain viral vectors (for example, adenoviruses) can elicit a robust cellular immune response against viral and some transgenic proteins, so their use has been limited to studies in immunecompromised animals (Seiler et al., 2007). Adeno-associated viruses (AAV), which have been considered safe, appear to be immunogenic in several experimental settings (Vandenberghe & Wilson, 2007) and in a clinical trial (Mingozzi & High, 2007). Some serious adverse events have occurred with viral gene therapy. One patient died of fatal systemic inflammatory response syndrome after adenoviral gene transfer in 1999 (Raper et al., 2003). Two children developed leukemia-like clonal lymphocyte proliferation after recombinant retroviral gene transfer in 2000 (Hacein-Bey-Abina et al., 2003), and 1 of them died after unsuccessful chemotherapy late in 2004. Attention focused recently on the tragic death of a young female patient in a gene therapy study (intra-articular injection of AAV vectors) of

Non-viral gene transfer systems offer several potential advantages over virus vectors. They are non-infectious, relatively non-immunogenic, have low acute toxicity, can accommodate large DNA plasmids or RNA, and may be produced on a large scale (Castanotto & Rossi, 2009; Gary et al., 2007). Non-viral gene therapy has been explored by physical approaches (transfer by gene gun, electroporation, ultrasound-facilitated and hydrodynamic delivery) as well as chemical approaches (cationic lipid-mediated gene delivery and cationic polymermediated gene transfer). Numerous chemical non-viral gene transfer systems have been proposed, including naked DNA, cationic liposomes, histones, and polymers (Gao et al., 2007; Ulrich-Vinther, 2007). The main drawback of non-viral vectors as gene carriers is their typically low transfection efficiency (Gao et al., 2007; Giannoudis et al., 2006). Furthermore, the *in vivo* delivery of non-viral liposome/plasmid DNA complex triggers an immune response (Sakurai et al., 2008). Non-viral gene therapy with cationic liposomes has already been tested in clinical trials that dealt with the treatment of inherited genetic disorders (for example, cystic fibrosis) (Hyde et al., 2000) and cancer (Ramesh et al., 2001). Synthetic and natural cationic polymers (positively-charged) have been widely used to carry DNA or siRNA (both negatively-charged) and condense it into small particles, facilitating cellular internalization via endocytosis through charge-charge interactions with anionic sites on cell surfaces. Although existing non-viral vectors have been found to enable DNA expression after *in vivo* delivery, the efficiency and duration of ensuing gene expression have proven to

summarized in the section of conclusion.

**2.1 Viral gene vectors** 

severe RA in 2007 (Kaiser, 2007).

**2.2 Non-viral gene vectors** 

**2. Barriers to gene delivery using non-viral vectors**

Systemic gene delivery involves a systemic approach in which exogenous genes are delivered to cells in a certain tissues, and secreted gene products are released into the circulatory system where they could modulate disease processes throughout the body. Systemic non-viral gene delivery has become an attractive alternative to viral vectors because of their safety, versatility and ease of preparation (Li & Huang, 2006). Genes can be delivered systemically (intramuscularly, intravenously, subcutaneously or, in animals, intraperitoneally). Otherwise, hydrodynamic-based gene delivery through systemic DNA injection offers a convenient, efficient and powerful means for high-level gene expression in animals (Liu & Knapp, 2001; Suda & Liu, 2007). This method is expected to be evaluated in patients soon (Romero et al., 2004). The limitations of the systemic approach to gene therapy are essentially the advantages of local delivery: exposure of non-target tissues to the therapeutic agent may have toxic effects or may compromise the immune system of the patient. Certain proteins will likely require very high levels of synthesis to achieve therapeutic function.

Ideally, gene therapy must protect DNA against degradation by nucleases in intercellular matrices so that the availability of macromolecules is not affected. Transgenes should be brought across the plasma membrane and into the nucleus of targeted cells but should have no detrimental effects. Hence, interaction with blood components, vascular endothelial cells and uptake by the reticuloendothelial system must be avoided. For DNA-based gene therapy to succeed, small-sized systems must internalize into cells and pass to the nucleus. Also, flexible tropisms allow applicability to a range of disease targets. Last but not least, such systems should be able to escape endosome-lysosome processing for endocytosis.

#### **2.4 Barriers to siRNA delivery**

The discovery of small interfering RNAs (siRNAs) has given renewed vision to the treatment of incurable diseases and genetically-associated disorders. Short double stranded (ds) RNA of 21-23 bp was cleaved by the RNAse III-like protein Dicer and incorporated into RNA-induced silencing complexes (RISC) (Hammond et al., 2000). Chemically-synthesized siRNAs and short hairpin RNA (shRNA) expression plasmids, which are sequence-specific for mRNA targeting, are methods commonly employed to mimic Dicer cleavage (Chen et al., 2007). However, siRNAs are susceptible to nuclease destruction and cannot penetrate the cell membrane because of their highly-charged backbone. An effective delivery system would enclose siRNA in carriers for protection and transport to the cytoplasm of targeted cells but should have no detrimental effects such as specific and non-specific off-targeted effects. Off-target effects can be divided into two categories: specific and non-specific offtargeted effects. Off-targeted effects may cause inflammation including interferon response, cell toxicity, and unintended gene knockdown.

Turning siRNA into drugs is a 3-step process. The design and *in vitro* screening of target siRNAs are followed by incorporating stabilizing chemical modifications in lead siRNAs, as required, and end in the selection as well as *in vivo* evaluation of delivery technologies that are appropriate for the target cell type/organ and disease setting (Vaishnaw et al., 2010). After nearly 10 years of study and development, many problems have been resolved, such as improving the stability of siRNAs, and avoiding 2 types of off-target effects. A recent

Chitosan-DNA/siRNA Nanoparticles for Gene Therapy 459

reach its target, the cell by endocytosis. In this respect it is well accepted that the polyelectrolyte complex polycation-DNA exhibiting a net positive charge binds to negatively charged cell membrane. (4) After the internalization the following crucial step in gene delivery with cationic polymers is the escape of the polymer/DNA complexes from the endosome. (5) The inefficient release of the DNA/polymer complex from endocytic vesicles into the cytoplasm is indicated as one of the primary causes of poor gene delivery. (6) and (7) the following step, the nuclear envelope is the ultimate obstacle to the nuclear entry of plasmid DNA. This obstacle is also considered crucial and two main mechanisms were proposed to explain how plasmid DNA enters into the nucleus: (i) a passive DNA entry into the nucleus during cell division when the nuclear membrane is temporarily disintegrated or

Cationic polymers, such as chitosan, are promising candidates for DNA transport in nonviral delivery systems (Kean & Thanou, 2010; Tong et al., 2009). Chitosan, a linear polysaccharide composed of randomly distributed β-(1-4)-linked D-glucosamine (deacetylated unit) and N-acetyl-D-glucosamine (acetylated unit), and has once been considered as an attractive gene transfer candidate for its superior biocompatibility, superior biodegradability and low cell toxicity. In recent years, with more researching methods involved, a more accurate and subtle view on the process of the entry of chitosan/DNA complexes into the cell nucleus has been developed. The enabling characteristics of Chitosan-DNA nanoparticles include biocompatibility, multiple ligand affinity, and a capacity of taking up large DNA fragments, while remaining small in size (Techaarpornkul et al., 2010). Chitosan and its derivatives, as favorable non-viral vectors involved in plasmid DNA delivery, have attracted attention in the field of siRNA delivery *in vitro* and *in* vivo (Andersen et al., 2009; Howard et al., 2009). Chitosan was once believed to be less effective than most other non-viral vectors because of its low stability and buffering capacity. However, recent technological advances in the chemical modification of chitosan have instituted improvements of its transfection efficiency without disturbing its biocompatibility and biodegradability. It has been demonstrated that transfection level is closely related to the molecular weight of polymers (Godbey et al., 1999). Chitosan (10-150 kDa), with a specific degree of deacetylation, allows maximum transgenic expression *in vitro*  (Lavertu et al., 2006). Another strategy for improving transfection is to take advantage of the mechanism of ligand-mediated uptake by cells to promote targeting and internalization, enhancing transfection efficiency. Ligand-mediated transfection has been shown to facilitate DNA internalization into cells via membrane receptors both *in vitro* and *in vivo.* Cell-specific ligand modification such as galactose, transferrin, folate and mannose can also effectively enhance the specificity of transfection through receptor-mediated endocytosis. Galactose ligand modification has been used to target HepG2 cells through the interaction with asialoglycoprotein receptors (ASGP-R) (Gao et al., 2003). A transferrin receptor is found on many mammalian cells, therefore it can be used as a universal ligand (Dautry-Varsat, 1986). Folate is not only over-expressed on macrophage surfaces, but is also over-expressed on many human cancer cell surfaces (Lee et al., 2006). Antigen presenting cells (APCs), the ideal targets of DNA vaccine, such as macrophages and immature dendritic cells are the target cells of mannose ligand (Kim et al., 2006). The specificity of these modifications can be

(ii) an active transport of the DNA through the nuclear pores.

demonstrated through ligand competitive inhibition experiments.

**4. Chitosan as a vector for gene therapy** 

anti-influenza study showed that the anti-viral activity of siRNA as found to be due to active siRNA. However, a different non-targeting control siRNA also had significant antiviral activity (Mook et al., 2007). siRNA targeting vascular endothelial growth factor for patients with age-related macular degeneration (AMD) are currently in clinical trial. But further study showed that the inhibition is a siRNA classic effect, which is sequence- and target-independent (Jackson & Linsley, 2010). The off-target effect can be minimized by optimizing the rules and algorithms for siRNA design (Vaishnaw et al., 2010). However, several other factors limit the utility of siRNAs as therapeutic agents, such as competition with endogenous RNA, induction of immune responses, degradation in lysosomes after endocytosis (Dominska & Dykxhoorn, 2010; Wang et al., 2010). Unprotected, naked siRNAs are relatively unstable in blood and serum and have short half-lives *in vivo* (Gao et al., 2009). Naked siRNAs do not freely cross cellular membranes because of their large molecular weight (~13 kDa) and strong anionic charge. They are rapidly degraded by nuclease. Physiological barriers hinder siRNAs from reaching their targets, thereby reducing their therapeutic efficacy. Moreover, siRNA molecules have unfavorable physicochemical properties (negative charge, large molecular weight and instability). Therefore, they need delivery systems to overcome physiological obstacles and prolong vascular circulation by reducing renal clearance, protecting them from serum nucleases, improving their effective bio-distribution as well as targeted cellular uptake with endosomal escape and, finally, promoting trafficking to the cytoplasm and loading onto RISC. Therefore, delivery systems are required to facilitate siRNA access to intracellular sites of action.Barriers to siRNA delivery depend on the targeted organ and routes of administration. For example, intravenous (IV) administration is the most commonly used technique. The endothelial wall in the vasculature presents the primary delivery barrier to siRNAs. The endothelial barrier is often altered by inflammatory processes (e.g., RA, infection) (Moghimi et al., 2005). siRNAs leave a blood vessel to enter tissue. After reaching target cells, they undergo internalization via endocytosis, escape from endosomes, and release into the cytosol and, finally, load onto RISC. At the same time, siRNAs undergo elimination. The mononuclear phagocyte system is responsible for removing circulating foreign particles from the bloodstream by the phagocytosis of resident macrophages (Moghimi et al., 2005).

#### **3. Cellular internalization of non-viral vector delivery system**

There are seven steps should be overcome before the expression of exogenous DNA. They are (1) complexation, (2) *in vivo* administration, (3) endocytosis, (4) escape from endolysosome, (5) release of DNA, (6) trafficking through cytoplasm and (7) finally import of DNA into nucleus. (If siRNA is used as exogenous nucleotide, the last two steps can be ignored; but if vector-expressed siRNA is used, the process remains the same.) During each step, many factors may come into play, inducing toxicity, immunogenicity or affecting transfection efficiency. (1) During complexation, the non-viral vectors-DNA interaction is driven mainly by the electrostatic interaction between the polycation and the charged phosphate groups leading to reversible linear to globule transition of DNA. The ability of the non-viral vectors to condense DNA into nanoparticles is often critical for transfection efficiency since DNA must be protected from DNase degradation. (2) Different *in vivo* administration routes will meet different physiological barriers. Therefore, it is suggested that the corresponding primary cells and similar physiological barriers should be tested *in vitro* as far as possible, before *in vivo* administration is attempted. (3) The following step is to

anti-influenza study showed that the anti-viral activity of siRNA as found to be due to active siRNA. However, a different non-targeting control siRNA also had significant antiviral activity (Mook et al., 2007). siRNA targeting vascular endothelial growth factor for patients with age-related macular degeneration (AMD) are currently in clinical trial. But further study showed that the inhibition is a siRNA classic effect, which is sequence- and target-independent (Jackson & Linsley, 2010). The off-target effect can be minimized by optimizing the rules and algorithms for siRNA design (Vaishnaw et al., 2010). However, several other factors limit the utility of siRNAs as therapeutic agents, such as competition with endogenous RNA, induction of immune responses, degradation in lysosomes after endocytosis (Dominska & Dykxhoorn, 2010; Wang et al., 2010). Unprotected, naked siRNAs are relatively unstable in blood and serum and have short half-lives *in vivo* (Gao et al., 2009). Naked siRNAs do not freely cross cellular membranes because of their large molecular weight (~13 kDa) and strong anionic charge. They are rapidly degraded by nuclease. Physiological barriers hinder siRNAs from reaching their targets, thereby reducing their therapeutic efficacy. Moreover, siRNA molecules have unfavorable physicochemical properties (negative charge, large molecular weight and instability). Therefore, they need delivery systems to overcome physiological obstacles and prolong vascular circulation by reducing renal clearance, protecting them from serum nucleases, improving their effective bio-distribution as well as targeted cellular uptake with endosomal escape and, finally, promoting trafficking to the cytoplasm and loading onto RISC. Therefore, delivery systems are required to facilitate siRNA access to intracellular sites of action.Barriers to siRNA delivery depend on the targeted organ and routes of administration. For example, intravenous (IV) administration is the most commonly used technique. The endothelial wall in the vasculature presents the primary delivery barrier to siRNAs. The endothelial barrier is often altered by inflammatory processes (e.g., RA, infection) (Moghimi et al., 2005). siRNAs leave a blood vessel to enter tissue. After reaching target cells, they undergo internalization via endocytosis, escape from endosomes, and release into the cytosol and, finally, load onto RISC. At the same time, siRNAs undergo elimination. The mononuclear phagocyte system is responsible for removing circulating foreign particles from the bloodstream by the

phagocytosis of resident macrophages (Moghimi et al., 2005).

**3. Cellular internalization of non-viral vector delivery system** 

There are seven steps should be overcome before the expression of exogenous DNA. They are (1) complexation, (2) *in vivo* administration, (3) endocytosis, (4) escape from endolysosome, (5) release of DNA, (6) trafficking through cytoplasm and (7) finally import of DNA into nucleus. (If siRNA is used as exogenous nucleotide, the last two steps can be ignored; but if vector-expressed siRNA is used, the process remains the same.) During each step, many factors may come into play, inducing toxicity, immunogenicity or affecting transfection efficiency. (1) During complexation, the non-viral vectors-DNA interaction is driven mainly by the electrostatic interaction between the polycation and the charged phosphate groups leading to reversible linear to globule transition of DNA. The ability of the non-viral vectors to condense DNA into nanoparticles is often critical for transfection efficiency since DNA must be protected from DNase degradation. (2) Different *in vivo* administration routes will meet different physiological barriers. Therefore, it is suggested that the corresponding primary cells and similar physiological barriers should be tested *in vitro* as far as possible, before *in vivo* administration is attempted. (3) The following step is to reach its target, the cell by endocytosis. In this respect it is well accepted that the polyelectrolyte complex polycation-DNA exhibiting a net positive charge binds to negatively charged cell membrane. (4) After the internalization the following crucial step in gene delivery with cationic polymers is the escape of the polymer/DNA complexes from the endosome. (5) The inefficient release of the DNA/polymer complex from endocytic vesicles into the cytoplasm is indicated as one of the primary causes of poor gene delivery. (6) and (7) the following step, the nuclear envelope is the ultimate obstacle to the nuclear entry of plasmid DNA. This obstacle is also considered crucial and two main mechanisms were proposed to explain how plasmid DNA enters into the nucleus: (i) a passive DNA entry into the nucleus during cell division when the nuclear membrane is temporarily disintegrated or (ii) an active transport of the DNA through the nuclear pores.
