Once the cell has been targeted and the DNA has been internalized via receptor-mediated endocytosis, several intracellular barriers (Fig. 2) need to then be overcome for successful foreign gene expression to occur. To tackle the intracellular barriers, a polyplex needs to:
• Escape from the endosome
• Survive the cytoplasmic environment
• Traffick the cytoplasmic environment targeting the nucleus with subsequent nuclear entry
• Be disassembled in the nucleus so it can be recognized by the cell's transcription machinery and be expressed under control in the target tissue population only
Overcoming the intracellular barriers may be achieved via certain DNA carrier molecules (92) and cell targeting ligands, although exploiting other factors such as additional endosomal releasing mechanisms and nuclear localization signals could strongly improve the capacity of a polyplex to overcome intracellular barriers.
The first major intracellular barrier that greatly impairs the efficiency of gene transfer is the entrapment and degradation of the DNA complex within intracellular vesicles after budding off of the coated pits from the plasma membrane (Fig. 2). Entrapment and degradation can be regarded as 2 separate barriers because overcoming vesicular degradation would only result in the accumulation of the transferred gene in the vesicles, while still limiting further transport into the nucleus. Thus, after cellular internalization, the transferred gene needs to overcome enzymatic degradation during vesicular fusion into lysosomes, and once this is done, it needs to then be released from the vesicles to traffick the cytoplasm and target the nucleus.
Several strategies have been developed to ensure the protection/release of DNA complexes from intracellular vesicles. The strategies involve incorporation or linkage of vesicular destructive elements to polyplexes, which perturb the integrity of vesicle membranes, allowing the luminal contents to spill into the cytoplasm in a nondamaged manner. The fundamental aspect here is to perturb the vesicular membrane, without damaging the DNA complex and other cellular membranes. Thus, the endosome-releasing mechanisms need to become active within the membraneous vesicle, without rupturing the cellular membraneous structures and organelles.
Application of endosome (vesicular) releasing agents to the transfection process has shown to augment gene transfer. Lysosomotropic agents, glycerol, virus particles, membrane disruptive proteins and peptides, and photosynthesizing compounds possess properties that can reorganize and disrupt vesicular membranes, promoting the release of polyplexes (see below).
Lysosomotropic agents are weak-base amines that can inhibit lysosomal function specifically (93,94). Examples of such agents are ammonium chloride and the weakly basic alkyla-mines, such as methylamine, propylamine, chloroquine, pro-caine, and spermidine. These agents are termed lysosomotropic
because they accumulate in the endosomes/lysosomes. This accumulation is partly due to the initially low lysosomal pH, and partly because of the continuous pumping of protons into the endosome/lysosome.
In some cell lines, such as erythroid cell line K562, gene expression is strongly enhanced by adding chloroquine to the transfection medium. This enhancement is primarily due to the prevention of intravesicular (lysosomal) degradation, followed by vesicular membrane disruption triggered by osmotic effects (see below). Chloroquine is also the most commonly used lyso-somotropic agent in gene transfer experiments (82). It is believed that chloroquine accumulates in the endosomal/lyso-somal compartment, acting osmotically, vacuolarizing and eventually disrupting the vesicle. In this case, the transferred gene is protected and released from intracellular vesicles, bafi-lomycin or monensin, 2 other agents that also prevent endosomal/lysosomal acidification but do not accumulate or enhance transfection. Regarding chloroquine, its effectiveness in enhancing gene transfer is also dependent on cell type, which ensures vesicular accumulation of the lysosomotropic agent. For example, K562 cells, in comparison to other living cells, have a defect in their vesicular pump system, determining the accumulation of chloroquine in endocytic vesicles (95). The use of chloroquine is limited due to cytotoxic properties.
Glycerol is a trihydric sugar alcohol, being the alcoholic component of fats. Several reports describe the interaction of glycerol with cellular membranes (96,97). Incubation of DNA-po-lylysine complexes in the presence of glycerol has resulted in a substantially enhanced transfection efficiency in primary fibroblasts and some cell lines. Regarding this, glycerol probably acts in intracellular vesicles after DNA complex internaliza-tion, rather than at the cellular membrane. The presence of glycerol alone is not sufficient for efficient gene transfer. For glyc-erol to have its maximum effect, other factors such as DNA complex net positive charge and the type of DNA-binding carrier plays a major role. Thus, a combined action of glycerol and polylysine on vesicular membrane seems to be responsible for enhancing gene transfer (97). It should, however, be noted that many cell types do not show the strong enhancement by glycerol.
Adenovirus particles are capable of inducing endosomal lysis during the process of adenoviral infection. Upon acidification of the endosome, the capsid proteins (e.g., the penton proteins) undergo conformational changes to an active form, capable of disrupting the endosomal membrane causing release of the contents into the cytoplasm. This observation potentiated the exploration of using adenovirus (viruses that enter cells via receptor-mediated endocytosis) as a means of overcoming vesicular entrapment during gene transfer. Adenovirus has been successfully incorporated in DNA complexes, enhancing transfection efficiency 100- to 1000-fold (48,49,73,84,98-105). Incorporation maybe achieved by adding adenovirus freely, where endo-molysis action is in trans, or by directly linking the virus to the DNA complex. Various methods of linkage have been applied as an alternative method for ensuring cointernalization of the adenovirus, which is not always the case when the virus is added in trans. The majority of these methods have been investigated with DNA complexes with polylysine as the DNA-binding element. There are several methods of covalently coupling polyly-sine to the exterior of the adenovirus. Approaches, including an enzymatic transglutaminase method or chemical coupling methods, have been applied successfully. Alternative approaches are those involving noncovalent coupling methods, such as an immunologic linkage strategy, with an anticapsid monoclonal antibody effecting the linkage between the adeno-virus and polylysine DNA-binding moiety (antibody bridge), ionic interactions, or biotin-streptavidin bridge. Such complexes consisting of DNA, adenovirus-polylysine, and ligand-polylysine are referred to as ''ternary complex.''
Only the membrane destabilizing function of the adenovirus capsid is required, thus the viral genome can be inactivated with methoxypsoralen plus irradiation, retending endosome disruption properties (99,102). One problem with the inclusion of adenovirus is that the transferred gene may also enter cells via the adenovirus receptor, thus compromising ligand-specific gene transfer. Strategies to inhibit uptake via adenovirus receptor include coupling of polylysine to periodate oxidized adenovirus. This treatment modifies the adenovirus fiber, which is necessary for virus attachment. An alternative approach is to target an antibody against the adenovirus fiber. Ablating binding to adenovirus receptor does not interfere with subsequent endoso-molyic activity. In addition to human adenovirus, CELO (chicken adenoviral strain) has also been successfully linked to DNA complexes, enhancing receptor-mediated gene transfer, although not as efficient as adenovirus (73). However, a major drawback in using viruses is their inflammatory response of cells to virus entry per se as well as their immunogenicity.
Because many other viruses also enter cells via the endocytic pathways, they may be used as gene transfer enhancers, provided that they display an endosomolytic function. The use of human rhinoviruses (picornaviruses), which are RNA viruses, has been described and investigated for releasing vesicular-entrapped DNA complexes into the cytoplasm (106). One major drawback of using rhinoviruses is the toxicity to human cells due to host protein synthesis machinery shut off. One way to circumvent the viral-induced drawbacks is to use only the endo-somolytic portion of the virus, or synthetic derivatives thereof.
For many biological processes such as entry of viruses and bacterial toxins into cells to be exercised, cellular membrane barriers need to be by passed, and this is usually achieved by membrane reorganization processes. The membrane reorganization process is the result of specific actions of certain membrane-disruptive elements (peptides and proteins). In most cases, the membrane active locate in such elements are peptide domains with amphipathic sequences. Under appropriate conditions the amphipathic sequences can interact with lipid membranes, perturbing them. This characteristic can be specifically used to influence gene transfer; membrane-active peptides can be derived from viral peptide sequences such as the N-terminus of influenza virus hemagglutinin subunit HA-2 or the N-terminus of rhinovirus VP-1, or they may be designed synthetically from the derived peptides by molecular modeling (e.g., GALA, KALA, EGLA, or JTS1) (106-110).
Viruses that enter the cell via the receptor-mediated endocy-tosis pathway have evolved specific mechanisms that ensure the release of their genome from the intracellular vesicles into the cytoplasm. The mechanisms leading to vesicular release are associated with viral proteins that specifically perturb membranes, either in a disruptive or fusogenic manner. The membrane-disruptive peptides are tools of membrane-free viruses (e.g., adenovirus), whereas the membrane-fusion peptides belong to enveloped viruses (e.g., influenza virus). These fuso-genic or endosomolytic protein domains are activated in a pH-dependent manner, and this characteristic can be used to enhance gene transfer (106-110) in a similar manner as whole virus particles, as mentioned above. The virus-derivedpeptides can be incorporated into polylysine-DNA complexes either by covalently linking to polylysine or via biotinylation of the elements, which allows binding to streptavidinylated polylysine. Simple noncovalent ionic interactions between positively charged polylysine-DNA complex and the negatively charged residues of membrane-destabilizing element is another possibility to achieve linkage. The linkage procedures should be also applicable to other DNA-binding elements. The membrane-destabilizing peptides have shown to enhance receptor-mediated gene transfer up to 1000-fold in cell cultures.
Other larger viral proteins have also been investigated for enhancing transfection by endosomal release. Adenovirus penton proteins have been attached to DNA via a synthetic oligo-lysine-extendedpenton-binding adaptor peptide. This complex transfers the DNA into the cell via adenovirus receptor and results in intracellular release (111).
One of the characteristics involved in virulence activity of micro-organism is the formation of substances (toxins, enzymes, and other proteins) that cause damage to the host (112,113). Some of the virulence factors act specifically on cell membranes. This action can be direct, for example, action of streptolysin O, resulting in cell lysis, or indirect, specifically acting on intracellular vesicular membranes, resulting in the release of entrapped virulence factors, having had entered the cell via an endocytic pathway. Various bacterial lysing proteins (cy-tolysins) and other toxins have been used as devices to enhance gene transfer. For example, streptolysin O and staphylococcal alpha toxin have been used standardwise as intracellular delivery reagents. Perfringolysin O has also been used to enhance delivery of DNA into cultured cells (114). For this purpose, the biotinylated protein was bound to DNA-polylysine complexes by a streptavidin bridge. Another well-studied exotoxin capable of enhancing gene transfer is the diphtheria toxin. A recombinant transmembrane domain of this toxin has been coupled to polylysine with subsequent incorporation into DNA-asialoor-osomucoid-polylysine complexes (115). Also, the bee venom component melittin has been applied for enhancing polyplex-mediated transfection (116).
Partial hepatectomy or microtubular disruption by colchicine treatment has shown to prolong gene expression after intravenous injection of DNA-asialoorosomucoid-polylysine complexes (46,47). This prolonged expression is due to the continuous existence and survival of plasmid DNA in cells of treated cells. This survival is due to protection of the DNA complex from the lysosomal or cytoplasmic environment, as a result of DNA complex persistence in the endosomal compartment, and the prolonged gene expression is due to a constant slow supply of plasmid DNA to the nucleus. The slow release is not a direct endosomolytic action. Therefore, how the DNA complex reaches the nucleus in this case remains unclear.
A novel technology, named photochemical transfection or photochemical internalization, has been recently developed and is primarily based on photosensitizing compounds, such as tetra(4-sulfonatophenyl)porphine or aluminium phthalocya-nine, that localize in the membranes of endosomes and lyso-somes (117,118). These become activated upon illumination and induce the formation of reactive oxygen species, destroying endosomal membrane structures releasing endocytosed DNA into the cell cytoplasm. This technology is being currently investigated as a new strategy for cancer therapy (119).
In contrast to DNA-binding elements such as polylysine, there are other polycations that possess specific properties enabling them to combine DNA binding and condensing activity with membrane-perturbing capacity, thus, not requiring the presence of endomolytic agents for enhancing transfection. The membrane-perturbing activity of such DNA-binding elements may be associated with their ionic state, or conformational flexibility, resulting in membrane-specific interactions.
PEI andpolyamidoamine polymers (''dendrimer'') areeffi-cient transfection agents per se (15-17). Most of these agents possess buffering capacity below physiological pH. This buff ering capacity is due to residues of these agents not being pro-tonated at physiological pH, making them efficient ''proton sponges.'' Upon acidification in the intracellular vesicle, the further protonation of the polymers triggers chloride influx, resulting in osmotic swelling (endosome swelling), and thus de-stabilization (rupture) of the intracellular vesicle membrane, resulting in the escape of the DNA complex. Hence, gene transfer is enhanced in the sense that DNA is free to travel to the nucleus.
Histidinylated polylysine, optionally in combination with zinc ions, was also found to have enhanced transfection property, presumably mediated by a related vesicular escape mechanism (120). Moreover, there are designed cationic peptide carriers that can bind nucleic acids and permeabilize lipid bilayers at the same time. One example is the cationic amphiphilic peptide KALA mentioned above (110).
B. Cytoplasmic Trafficking, Nuclear Targeting, and Entry
Upon receptor-mediated endocytosis and release from the endosome, polyplexes need to then overcome cytoplasmic degradation and traffick to and enter the nucleus. These tasks are only partially fulfilledby certain DNA carrier molecules and cell targeting ligands. Other elements such as nuclear localization signals can be incorporated into polyplexes improving nuclear targeting.
The nucleus is separated from the cytoplasm by the nuclear envelope, which consists of 2 chemically distinct membranes, the inner and outer membrane, with the perinuclear cisterna space in-between. The outer membrane is continuous to the endoplasmic reticulum. Nuclear pore complexes (NPCs) span the nuclear envelope, and transport of macromolecules from the cytoplasm to the nucleus occurs through these complexes. Regarding nuclear import, several major processes need to be distinguished: (1) the steps leading to the import process; that is, signals guiding the protein transport to the nuclear membrane, including cytoplasmic recognition of the import protein by transport factors and nuclear pore targeting; (2) the actual molecular mechanism of translocation of the protein from the cytoplasmic side of NPC to the nuclear side of NPC into the nucleus; the nuclear pores being the site of translocation; (3) release of the import protein into the nucleus; and (4) recycling of transport factors.
There are several distinct nuclear import signals that guide the import of proteins into the nucleus. These signals are part of the primary sequence of the protein destined to be targeted into the nucleus. The best characterized ones are the SV40 LTA ''classical'' nuclear localization signal (NLS) and M9 (an import signal of hnRNP AI protein) import signals.
It has been suggested that shuttling of import receptors may be a major process in nuclear transport of proteins. In this hypothesis, the import receptor binds the protein in the cytoplasm. The molecule is then carried through the NPC into the nucleus, where it is released from the transport receptor. The receptor returns to the cytoplasm, ready for transporting the next molecule into the nucleus. In this model, the binding of the transport receptor to the molecule may be regulated by the different environments of the nucleus and cytoplasm.
Four major transport factors are required for the NLS-de-pendent protein import: (1 and 2) importins alpha and beta (kar-yopherin alpha and beta), (3) GTPase Ran/TC4, and (4) nuclear transport factor 2 (NTF2) (p10). These factors interplay for successful transport of proteins into the nucleus. Proteins (with NLS signal) with sizes up to 25 nm interact in the cytoplasm with soluble NLS receptor (importins alpha and beta). The kar-yopherin beta 1 (''importin beta'') mediates docking to nucleoporins, the components of the nuclear pore, located at cytoplasmic face and nucleoplasmic face of the NPC. Nucleopore cytoplasmic filaments may also be involved in the nuclear targeting process. GTPase Ran and p10 (NTF2) are required to translocate the docked NLS peptide into the nucleus. The influenza virus nucleoprotein particle is one example that is taken up by this pathway.
The M9 domain of hnRNP A1 contains 38 amino acid residues (M9 import signal) sufficient for import purposes into the nucleus. It bears no sequence similarity to classical NLS. Nuclear import of ribonucleoprotein A1 is mediated by another distinct import pathway, by binding to karyopherin beta 2 (''transportin''—import receptor of the M9 pathway). It also requires Ran.
It is very likely that, in addition to the classical NLS and the M9-type import, more pathways into the nucleus exist. This may be of an advantage in the sense of regulating import of distinct classes of molecules separately (121,122).
The nuclear envelope represents a major barrier for pol-yplex-mediated gene transfer. In the majority of cell types, transport of DNA into the cell nucleus is inefficient. For example, less than 1% of NIH 3T3 fibroblasts have shown to express p-galactosidase after cytoplasmic injection of reporter gene. However, p-galactosidase has been efficiently expressed when injected into the cytoplasm of primary rat muscle cells. Although how DNA complexes find their way to the nucleus is not fully understood, recent findings suggest that DNA compaction (92), certain physiological ligand pathways such as bFGF-tar-geted complexes (53), or specific DNA sequences, as demon-stratedby intact protein free SV40 DNA (123), may contribute to cytoplasmic trafficking andnuclear targeting of the polyplex.
In dividing cells, DNA may passively enter the nucleus during mitosis when the nuclear membrane is broken down (124). However, many cell types are nondividing with the nuclear membrane staying intact, thus, the transferred gene needs to enter the nucleus differently than in dividing cells. In this case, nuclear entry and trafficking may represent a major interin-tracellular barrier for successful in vivo gene transfer. When DNA is injected into the muscle cell far from the nuclei, expression decreases. This may be the result of cytoplasmic sequestration, preventing nuclear accumulation of DNA. The movement of DNA toward the nucleus may be inhibited by its binding to cytoplasmic elements, and/or entrapment within the cytoskele-tal mesh (125). However, polymers such as PEI and PLL have been suggested to promote gene delivery from the cytoplasm to the nucleus, possibly attributed to intrinsic nuclear targeting activity and protective mechanisms against cytosol nucleases (126,127).
Once the polyplex has trafficked to the nucleus, it then needs to enter through the nuclear membrane into the nucleus. Although rupture of the nuclear envelope (as occurs during cell division) is not a necessity for some PEIpolyplexes to penetrate the nucleus (92,128), transfection efficiency of polyplexes in general is critically dependent on cell division (129). High transfection efficacy has been observed when polyplexes are added to cells in late S or G2 phase, and low transfection when polyplexes are added in G1 phase. Thus, transfecting nondivid-ing cells in vivo may be considered as an additional barrier in vivo in terms of nuclear entry of polyplexes.
Thus, understanding the mechanisms of nuclear entry is critical and will be a large step toward designing gene delivery systems, more efficient in transferring the DNA into the cell nucleus.
Expression of transferred DNA has been inhibited by wheat germ agglutinin (WGA), suggesting that DNA may enter the nucleus by the WGA-sensitive process common to large karyo-philic proteins and RNA (125) (WGA blocks the NPC machinery because of the presence of N-acetylglucosamin residues on nucleoporins). However, recently it has been shown that nuclear localization of DNA does not require the addition of cytoplasmic protein factors necessary for protein import (130). Nevertheless, DNA entry appears to be regulated by NPC. The NPC accommodates both passive diffusion and active transport. Molecules smaller than 20 nm in diameter passively diffuse through NPC into and out of the nucleus. Larger macro-molecules require active transport for nuclear entry. The exact mechanism by which exogeneous DNA passes through the NPC has not yet been determined, although it may be similar to the transport of proteins, larger than 15 Kda, actively into the nucleus.
Certain viruses such as hepatitis B virus also use the active nuclear import mechanism. Viral core particles containing synthesized DNA bind to the nuclear pore complex; the viral poly-merase, which is covalently linked to the viral DNA, acts as NLS. Core particles (35 nm) exceed the maximal diameter of the nuclear pores, thus requiring disassembly before the viral genome is internalized.
The efficient transport of DNA complex into the nucleus is an active process that probably requires a nuclear localization signal. Several NLSs have been identified in proteins with a nuclear fate, the sequences being very basic, with >50% of their amino acids being lysine residues [e.g., Phe-Lys-Lys-Lys-Arg-Lys-Val, directing the nuclear import of SV-40 large T antigen, or the NLS (Lys-Lys-Lys-Tyr-Lys-Leu-Lys) within HIV-1 matrix protein]. Substitution of 1 of the lysine residues results in total failure of nuclear import. Thus, it is possible that DNA-binding elements rich in lysine (e.g., polylysine) may play a role in the nuclear import of DNA complexes. Regarding this, injection of DNA-polylysine mixtures into the cytoplasm of mouse ES cells has lead to transgenic animals with about 50% efficiency (compared with intranuclear injection of naked DNA). In contrast, injection of naked DNA into the cytoplasm does not lead to transgenesis. To make the DNA import process more efficient, it may be necessary to direct the transport to the nu cleus by incorporating NLSs into the DNA complex. This is strongly supported by recent findings (131) that show that incorporation of a single SV40 LTA NLS into a DNA plasmid molecule can dramatically enhance transfection efficiency.
To sum it up, it would be desirable to engineer DNA complexes with specific nuclear targeting and translocating elements, enabling the DNA complex to be (1) recognized in the cytoplasm by nuclear import receptors (by interacting with transport motifs either directly on the DNA sequence or on a protein from the DNA complex), (2) targeted to the nuclear pore complex, (3) translocated efficiently through the nuclear pore, and (4) released at the nucleoplasmic face of the nuclear pore complex into the appropriate nuclear compartment for subsequent disassembly and expression.
The cell nucleus is crowded, containing large amounts of DNA, RNA, and protein. In addition, nuclear processes such as replication, transcription, translation, and DNA repair processes are constantly active in specific compartments, resulting inanuclear jam(132,133). Within all the mass of cellular DNA, RNA, and proteins and the nuclear processes, the DNA complex needs to become (either disassembled or reorganized in a suitable fashion) exposed, enabling the nuclear expression machinery to recognize and express it. Premature disassembly and DNA release from the carrier molecule in the cytoplasm may prevent efficient transfer to the nucleus, hence obligerate expression. Polyplexes most likely disassemble in the nucleus, which is a characteristic desirable for successful gene transfer and expression (92).
Furthermore, the anatomical location (intranuclear depot) of polyplexes within the nucleus may be crucial for efficient expression (53). For example, bFGF is believed to be retained in the nucleus within discrete storage depots. Similar distributions have been reported for bFGF-targeted polyplexes with a relatively poor level of specific gene expression, suggesting difficult availability of the DNA for transcription (53). Indeed, intranuclear trafficking of transcription factors contributes to transcriptional control, andthus, has implications forbiological control (134,135); hence, it may determine intranuclear trafficking/localization if used as carrier molecules (90,91).
Whether incorporation of intranuclear targeting elements in polyplexes affects nuclear localization and hence, expression, remains to be considered for investigation. Thus, both efficient uncoating and intranuclear trafficking is another challenge to be tackled.
Several factors endanger the persistence of transferred DNA within the nucleus: (1) degradation by intranuclear nucleases; (2) DNA loss, mainly during cell division, although DNA may also be rapidly lost even when cells are not dividing; (3) loss of transfected cell because of apoptotic, inflammatory, or immune response; (4) silencing of the introduced gene by transcriptional shut-off; and (5) possibly inefficient intranuclear trafficking.
Protection of DNA may be achieved via the proper ratio of DNA-DNA-binding carrier (e.g., polylysine). Saturating the binding capacity of the DNA backbone with the DNA-binding moiety may avoid the access of DNA to nucleases and increase the stability of the DNA in the nucleus (nuclear retention, resulting in prolonged expression).
DNA loss other than degradative loss can be prevented by including specific sequences to the transferred DNA that ensure either integration of the DNA into host chromosome, or extrachromosomal replication of the transferred DNA with equal segregation to daughter cells. These persistence-ensuring sequences can be derived from certain viruses or from chromosomes.
Retroviruses and adeno-associated virus (AAV) stably insert their genome into host genome. The integration mechanisms have been characterized [retrovirus: LTR sequences, integrase protein; AAV: inverted terminal repeat (ITR) sequences, rep protein], and may be exploited by incorporation of the corresponding nucleic acid and protein elements into a viruslike particle (DNA complex). During its lysogenic cycle, AAV integrates into a specific site, denoted AAVS1, on human chromosome 19 (136). This property has been used to achieve site-specific integration of plasmid DNA in 293 cells. When the AAV-encoded recombinase rep is supplied in cis or trans, plasmids containing the AAV ITRs are integrated at AAVS1. In this way, DNA can be directed to a specific region of a human chromosome, and thus may avoid the problem of random inser-tional mutations created by integrative vectors such as ret-roviral vectors.
Other viruses such as herpes virus (e.g., Epstein Barr Virus, EBV) can persist in infected cells without integrating their genome into the host. This persistence is partially due to replica-tive property of viral DNA via cis-acting origin of replication, which is activated by the trans-acting gene product of the viral EBNA-1, and additional nuclear retention mechanisms. The viral-persistence mechanisms can be used in designing extra-chromosomal-replicating DNA constructs (episomal vectors) by integrating the appropriate sequence elements, recognizable in mammalian cells, into the DNA construct to be transferred (e.g., EBV Ori P, EBNA-1). DNA constructs containing these sequences have the ability to replicate once per cell cycle with nuclear retention without interfering with the host chromosomes. An alternative to using viral origin of replication, human genomic sequences may be used to mediate DNA construct replication (137).
Other origins of replication characterized include those of bovine papilloma virus or SV40 (138). However, these replication origins also require viral proteins for activation, which may have oncogenic or toxic properties. The viral replication origins are species specific and may also replicate more than once per cell cycle, resulting in mutations in the transferred gene.
The next challenge is the generation of artificial chromosomes for maintenance of large genomic sequences. The basic DNA sequence requirements for human chromosome function are believed to be similar to those identified in yeast, which include a centromere, telomeres, and origins of replication. These elements have been used to construct yeast artificial chromosomes (YACs) (139), and their transfer has been demonstrated by spheroblast fusion. Researchers are now trying to construct better artificial chromosomes (140). It is important here to de fine the minimal sequence requirements for functional mammalian chromosomal elements.
Human artificial chromosomes (HACs) may serve as valuable DNA constructs, containing the requirements for achieving gene expression persistence. An opening door toward this goal is the recent generation of mitotically and cytogenetically stable artificial chromosome derived from transfecting human HT 1080 cells with alpha satellite plus telomere DNA and genomic carrier DNA, forming de novo microchromosomes with functioning centromeres (141). The only sequence that has been shown to form de novo centromeres after transfection is alphoid DNA with telomere DNA and genomic carrier DNA. However, centromeric function of human chromosome is not always associated with alphoid repeats. There are other nonalphoid geno-mic regions characterized by neocentromic activity (sometimes, a centromere can appear at a new position in a chromosome by a process called ''centromere activation''). Transfection experiments with DNA present at the neocentro-mere should demonstrate whether this DNA can also form centromeres de novo.
With advancement, HACs could be used to introduce therapeutic genes into cells. To use HACs for human gene therapy, efficient methods for delivery will be required. The receptor-mediated gene transfer system is a very promising choice capable of targeting and transferring any large DNA construct into cells. Bacterial artificial chromosomes have already been delivered into mammalian cells using psoralen-inactivated adenovirus/PEI carrier (103).
Persistence of the transferred gene is required, but not necessarily sufficient. It does not necessarily mean that gene expression is going to be efficient and long-lasting. A failure of gene therapy in clinical trials may not always be due to gene delivery, but rather due to novel confrontations at the expression level. Specific expression cassettes determine the efficiency of expression after the gene has been transferred and maintained in the nucleus. The major elements of the expression cassette are those that ensure strong, controllable, (switchable), and cell specific (tissue restricted) expression of the therapeutic gene. Previous studies using viral promoters (e.g., cytomegalovirus promoter/enhancer) have observed a transcriptional shut-off (''silencing'') of the introduced expression cassettes. This can be avoided by the use of natural, cell-specific promoter/enhancer sequences. These are also attractive due to the opportunity of transcriptional targeting, as a further filter for specificity
(142). Some examples are steroid-inducible promoters, tumor-specific promoters, muscle-specific promoters, hypoxia response elements, hepatitis B virus-derived promoters for liver-specific expression, and multidrug-resistant gene promoter
(143). Here lies the attractiveness of nonviral receptor-targeted polyplexes in gene therapy, for the reason that any gene construct regardless of the size may be transferred.
Was this article helpful?