|Surgical organ perfusion method for somatic gene transfer: An experimental study on gene transfer into the kidney, spleen, lung and mammary gland|
|Prev||Chapter 2. Review of the literature||Next|
Essentially, gene transfer involves the delivery of therapeutic protein encoding nucleic acid to the target cells. The DNA consists of one or more genes to be expressed in the target cell and the sequences controlling their expression. The administration of genes as therapy requires transportation vehicles, called vectors, which encapsulate the gene and carry it into the target cell. The vector binds to the target cell membrane, and after this internalization takes place. The administered genome is then transported into the cell nucleus, where it is integrated into the genome of the cell or remains outside the genome as an episome in the nucleus, depending on the type of vector used. Every step in this process and the ultimate expression of the gene constitute potential limitations of gene therapy. Some characteristics of the most commonly investigated vectors are presented in table 2.
Table 2. Characteristics of the most commonly used gene transfer vectors.
|Vector||Capacity for DNA-carrying/kb||Duration of expression||Inflammatory response||In vivo gene delivery efficiency|
The recombinant retroviruses are RNA viruses that are deleted in the gag, pol and env viral sequences, which makes the replication of the viruses defective and decreases their immunogenicity. They can accommodate up to 5 kb of exogenous material (Crystal 1995b). Retroviruses are able to efficiently infect dividing cells, and the exogenous genetic material is integrated into the genome of the target cell. The natural retroviral life cycle is generally nonlytic for the cell. The retroviral vectors are produced in a packaging cell line that contains the gag, pol and env sequences providing the proteins necessary to package the virus. The vector enters the target cell via a specific receptor. In the cytoplasm, the reverse transcriptase enzyme carried by the vector converts the RNA into proviral DNA, which is integrated into the genome of the target cell, where the transgene product is expressed.
Theoretically, the target cell genome is permanently modified, which would be an advantage when treating hereditary and chronic diseases. The prerequisite for a successful transfection procedure is that the target cell must proliferate, and the retrovirus is therefore not a very feasible vector for in vivo transduction into, for instance, pulmonary or renal cells, which have quite silent cell turnover.
Another disadvantage is related to the potential mutagenicity of retroviral vectors due to their random integration into the host’s genome, which may cause insertional mutagenesis in case it disrupts a tumor suppressor gene or activates an oncogene. It is also difficult to prepare high titers of viral stocks, and the retrovirus cannot carry very large amounts of transgenic material. Besides, the expression of the transgene is difficult to control. Retroviruses have been used mostly in ex vivo gene transfer trials (Crystal 1995b). So far, ex vivo retrovirus-mediated marker gene transfer in humans has been used for T cells, tumor infiltrating lymphocytes (TIL), stem cells in blood and bone marrow and neoplastic cells derived from solid tumors, hematopoietic cell lines and synovial cells. Therapeutic genes have been transferred into fibroblasts, T cells in ADA deficiency and HIV, cord blood cells, placental cells, tumor cells and hepatocytes in familial hypercholesterolemia by retroviruses. (Rosenberg et al. 1993, Moullier et al.1993, Oldfield et al. 1993, Medin et al. 1996, Jorgensen et al. 1997.) The marker gene expression after retroviral gene transfer or mRNA has been observed for variable lengths of time, ranging from several weeks to 36 months.
Initially, adenoviruses were evaluated as vectors for CFTR gene transfer, because of their natural tropism for pulmonary epithelial cells. In nature, adenoviruses exist as more than 40 serotypes, which cause common clinical syndromes ranging from pharyngitis to diarrhea. The infections are usually not severe, and the virus is not associated with cancer initiation. Adenovirus is a large double-stranded DNA virus, which contains a 36 kb genome that consists of early regulatory proteins encoding genes and a late structural protein gene. Adenoviruses of serotype 2 and 5 used in gene transfer trials have been shown to bind to cultured epithelial cells through attachment to specific receptors, which are the same as those used by the coxsackie B virus (Crystal 1997). The use of recombinant adenoviral vectors as in vivo gene transfer vehicles has been investigated in most detail (Wilson 1996).
The advantages of adenovirus as a gene transfer vector are that it is safe and can grow high titers (up to 1012 infectious particles per ml versus 106 for retroviruses) of the purified recombinant virus, which can infect efficiently a broad range of differentiated non-dividing cells in vivo (Curiel et al. 1996). The first-generation recombinant adenovirus is deleted in the early E1 regulatory region, which normally functions to activate other viral genes. This renders the virus unable to replicate. The E3 region is usually also deleted, to obtain room for the transgene (Trapnell & Gorziglia 1994). This engineering of the virus involves replacement of the viral structural protein encoding part of the genome with a marker gene (e.g. β -galactosidase) or therapeutic gene cDNA (Kozarsky & Wilson 1993). The virus does not integrate into the cell chromosome, but remains as an episome in the nucleus. Besides, it can accommodate a large amount of exogenous information.
The disadvantage of the first-generation adenovirus is its immunogenicity, which, together with its epichromosomal location, makes gene expression transient (Yang et al. 1994, Zsengeller et al. 1995, Otake 1998). Several studies have pointed out that transgene expression lasts from only a few days to some weeks. It has been observed that adenoviral gene transfer elicits a cytotoxic T cell response that destroys the transfected cells and induces humoral immunity due to elevated IgG and IgA antibodies (Yang et al. 1994, Van Ginkel et al. 1995). These antiviral antibodies may neutralize the vector when administered a second time, and this may preclude repetitive dosage. In human trials, an acute local and systemic inflammatory response has been observed in vascular endothelium and lung tissue after adenoviral administration (Channon & George 1997).
To overcome these inflammatory and toxic problems, a second- and third-generation adenovirus have been created. In these vectors, more viral proteins encoding the sequence’s E4 region (and possibly E2) are deleted to reduce the expression of viral proteins. On the other hand, this yields lower titers of the vector and makes it more vulnerable to contamination (Channon & George 1997).
Human first-generation adenoviruses have been most widely used as in vivo gene transfer vectors (Wilson 1996). They have been used for in vivo human gene transfer into nasal and pulmonary cells (CFTR cDNA), vascular endothelial cells, kidney (Heikkilä et al. 1996, Sukhatme et al. 1997), heart, liver, central nervous system, muscle and hematopoietic cells and cancer cells (Jaffe et al. 1992, Engelhardt et al. 1993, Chen 1994, Chang et al. 1995, Cusack et al. 1996, Simons et al. 1999). Merrick et al. (1996) compared the infectivity of a replication-defective type 5 recombinant adenovirus in various vascular endothelial cells in in vitro culture, organ culture and in vivo. They observed that, in culture, infectivity was good in both porcine and human vascular endothelial cells (up to 90% cells were infected at an AdV concentration of 1x1010 pfu ml–1), but the in situ gene delivery into uninjured vascular endothelium was markedly poorer, suggesting that some mechanisms other than vectors or target cells underlie the poorer in vivo transduction efficiency. In a trial of gene transfer into malignant glioma cells via a catheter inserted into the tumor, Puumalainen et al. (1998a) compared the gene transfer efficiencies of retro- and adenoviruses. They found adenoviruses to be more efficient vectors for gene transfer into glioma cells than retroviruses, <0.01–11% and <0.01–4%, respectively.
Adeno-associated virus (AAV), a human parvovirus, is a small single-stranded DNA virus that can infect both dividing and non-dividing cells. The wild-type AAV has many hosts and it will lead to a productive viral infection in virtually all cell lines, provided that an appropriate helper virus (such as adenovirus) is present. If a helper virus is absent, AAV will remain latent in the host’s genome. The generation of recombinant AAV is much like that of retroviruses, but more complicated than that of adenoviruses. The packaging capacity of recombinant AAV is only 4.9 kb, which makes it difficult to insert large genes, such as the CFTR gene. The integration of the wild type viral genome usually occurs at a specific site in human chromosome 19, but the integration may not be site-specific in immortalised cells lines (Flotte & Carter 1995, Moullier 1997). The advantages of AAV are that it is relatively non-toxic and non-immunogenic and results in long-lasting expression. The transgene expression obtained is usually much longer than that of adenoviruses.
Murphy et al. (1997) obtained 6-month normalization of hyperglycemia, insulin resistance and correction of the serum leptin level by recombinant AAV-leptin cDNA gene transfer in transgenic obese type II diabetic mice. Successful in vivo reporter gene transfer into brain cells has been attained using AAV vectors, and transgene expression control was possible by simultaneous doxycycline administration (Haberman et al. 1998). In addition, there are reports of long-term expression following the delivery of AAV vectors into muscle, heart, liver and lung (Flotte et al. 1993, Koeberl et al. 1997, Fisher et al. 1997).
Other viral vectors have been less extensively studied. Herpes simplex virus type I has a large genome (150kb) and is able to transfer large intact genes. It has been used for gene transfer into neurons, brain tumors (Kennedy 1997), various tumor cells and B cells (Levatte et al. 1998). Carew et al. (1998) reported effective IL-2 and lacZ reporter gene transfer by HSV amplicon vectors in murine squamous cell carcinomas after intra-arterial vector delivery. The disadvantage of HSV is that it may become latent in neural cells and that there is so far little information of the fate or stability of the vector. On the other hand, latency may be an advantage for stable gene expression in chronic diseases.
Cytomegalovirus, baculovirus and poxvirus have been used earlier as gene transfer vectors in a few studies, but they have many disadvantages. Vaccinia virus (a double-stranded DNA poxvirus) has been used for gene transfer into the lungs (Hogan et al. 1998) and urinary bladder. Lee et al. (1994) used successfully vaccinia virus recombinants in intravesical instillation for gene transfer into normal bladder urothelium and transitional cell carcinoma cells in vivo. They suggested that this method could be used to introduce genes of immunogenic antigens and cytokines to elicit a host immunologic response against superficial bladder cancer.
Recently, a lentiviral vector based on the human immunodeficiency virus (HIV) has been used for gene transfer into lymphocytes, brain, retina (Miyoshi et al. 1997), muscle and liver (Naldini 1998). In this vector, all the viral sequences non-essential for transduction have been deleted. Olsen (1998) has reported a new viral vector derived from equine infectious anemia virus (EIAV).
Eukaryotic cells can, under suitable conditions, take up exogenous DNA, part of which becomes located in the nucleus. This process is, however, usually insufficient for gene therapy, and a wide variety of gene transfer facilitating methods other than immunogenic viral vectors have therefore been developed, including 1) liposome fusion, 2) calcium phosphate, 3) microinjection, 4) electroporation, 5) polycations, 6) particle bombardment, and 7) receptor-mediated methods (Felgner et al. 1987). The synthetic methods designed for gene transfer are extremely varied. To put it simply, these vectors generally use natural mechanisms of mammalian cells for the uptake and intracellular transport of macromolecules. Multimolecular aggregates are generally formed with plasmid DNA, which subsequently bind to cell surfaces and trigger endocytosis of the vector for the transgene to be transported into the cell nucleus. The major advantage of these methods is that they are relatively non-immunogenic, except for the possible immune response which the transgene itself may elicit in the recipient.
Liposomes used in human gene transfer have various compositions, but they usually include synthetic cationic lipid bilayers complexed to the negatively charged plasmid to be transferred into the target cell by a cell-membrane fusion event or endocytosis (Felgner et al. 1987). Plasmid-liposome complexes have many advantages as gene transfer vectors. Because of the lack of proteins, they are relatively non-immunogenic. They can carry exogenous material of essentially unlimited size. They cannot replicate or recombine to form infectious agents. Their disadvantage is their relatively low transduction efficiency compared to viral vectors. However, their transfection efficiency has been improved by simultaneous delivery of agents that prevent DNA degradation within endosomes (Budker et al. 1996). Plasmid-liposome complex-mediated gene transfer, called lipofection, has been used for gene transfer into the spleen, lung, arterial and liver cells (Alton et al. 1993, Baru et al. 1995, Laitinen et al. 1997, Schmid et al. 1998). The liposome-mediated gene transfer efficiency has been improved by complexing viral particles with liposomes, such as the Sendai (HVJ) virus (Tomita et al. 1992, Yonemitsu et al. 1997). Using this method, Tomita et al. (1992) reported renal gene transfer into 15% of glomerular cells after intra-arterial infusion of vectors four days after treatment. By conjugating liposomes with antibodies or ligands, better targeted lipofection can be achieved. Recently, polyethylenimine, a cationic polymer-plasmid complex, has been used for gene transfer into lung (Ferrari et al. 1997, Goula et al. 1998).
Calcium phosphate precipitation has been used for the transfection of plasmid DNA into some cultured cells, such as hepatocyte cell lines, and in vivo, but the efficiency has not been very high, the transfection frequencies being less than 1% (Chang & Wu 1994). This method was primarily reported in the first gene transfer trial by Graham and Van Der Eb (1973). Calcium phosphate precipitates have been used together with the adenoviral vector to enhance transfection efficiency (Fasbender et al. 1998) in the airway epithelial cells.
Particle bombardment has been recently adopted as a technology for gene delivery. Originally, this method was developed for foreign gene delivery into higher plants. Basically, the method involves transfer of plasmid DNA into a target cell that is coated unto the surface of microparticles of, for example, gold or tungsten (Biewenga et al. 1997). These particles are then accelerated by a particular driving force, e.g. by establishing a high-voltage discharge between two electrodes. The ultimate transfection efficiency depends on the combination of the “ballistic” parameters and the characteristics of the target tissue. The main advantage is that it is a mechanical way to transfer a gene across the plasma membrane, which is why the transfection efficiency is less dependent on the characteristics of the target cell. Compared to the other methods, particle acceleration-mediated transformation is less effective than the viral methods, but more effective than lipofection or calcium phosphate precipitation. This method has been used for gene transfer into brain tissue (Jiao et al. 1993), muscle (Zelenin et al. 1997) and cancer cells (Hui & Chia 1997). Guo et al. (1996) compared particle bombardment, lipofection, calcium phosphate precipitation and retroviral transfection in vitro for lacZ and luciferase gene transfer into rat oligodendrocytes and found that the most effective of these four methods was particle bombardment-mediated transfection with gene gun accelerated DNA-coated 0.95 µm gold particles.
Receptor-mediated gene transfer facilitates the vector’s entry into the target cell by utilizing specific receptors on the cell surface. For example, a synthetic glycoprotein called mannosylate polylysine has been used as a targeting ligand to introduce functional genes into macrophages that express the mannose receptor (Ferkol et al. 1996).