Genetic pharmacology: progresses in siRNA delivery and therapeutic applications.

In the RNA interference process, the catalytic degradation of an endogenous mRNA results from the Watson-Crick complementary recognition by either a small silencing synthetic double-stranded ribonucleotide (siRNA) or by a small hairpin RNA (shRNA) produced in the cell by transcription from a DNA template. This interference process ideally results in an exquisitely specific mRNA suppression. The present review is dedicated to siRNAs. It describes the mechanism of RNA silencing and the main siRNA delivery techniques, with a focus on siRNA self-complexing to cationic lipids to form nanoparticles, which are called lipoplexes. The addition to lipoplexes of an anionic polymer leads to the ternary formulation APIRL (Anionic-Polymer-Interfering-RNA-Lipoplexes) with increased in vivo stability and biological efficacy. In terms of clinical development, the review focuses on therapeutic applications by intravenous delivery to the liver and inflammatory joints, and to localized siRNA delivery to the ocular sphere.

Long-term efficacy of ciliary muscle gene transfer of three sFlt-1 variants in a rat model of laser-induced choroidal neovascularization.

Inhibition of vascular endothelial growth factor (VEGF) has become the standard of care for patients presenting with wet age-related macular degeneration. However, monthly intravitreal injections are required for optimal efficacy. We have previously shown that electroporation enabled ciliary muscle gene transfer results in sustained protein secretion into the vitreous for up to 9 months. Here, we evaluated the long-term efficacy of ciliary muscle gene transfer of three soluble VEGF receptor-1 (sFlt-1) variants in a rat model of laser-induced choroidal neovascularization (CNV). All three sFlt-1 variants significantly diminished vascular leakage and neovascularization as measured by fluorescein angiography (FA) and flatmount choroid at 3 weeks. FA and infracyanine angiography demonstrated that inhibition of CNV was maintained for up to 6 months after gene transfer of the two shortest sFlt-1 variants. Throughout, clinical efficacy was correlated with sustained VEGF neutralization in the ocular media. Interestingly, treatment with sFlt-1 induced a 50% downregulation of VEGF messenger RNA levels in the retinal pigment epithelium and the choroid. We demonstrate for the first time that non-viral gene transfer can achieve a long-term reduction of VEGF levels and efficacy in the treatment of CNV.

Cytosolic phospholipase A2α gene silencing in the myeloid lineage alters development of Th1 responses and reduces disease severity in collagen-induced arthritis.

OBJECTIVE: Several lines of evidence implicate cytosolic phospholipase A(2)α (cPLA(2)α) as a critical enzyme in inflammatory disorders, including rheumatoid arthritis. Since cells from the myeloid compartment regulate local and systemic disease pathogenesis, the present study was undertaken to examine the effect of cPLA(2)α inhibition in experimental arthritis, using a delivery system tailored to target monocyte functions by RNA interference (RNAi). METHODS: Mice with collagen-induced arthritis (CIA) were injected intravenously with an anti-cPLA(2)α small interfering RNA (siRNA) sequence (siPLA2) formulated as lipoplexes with the RPR209120/DOPE cationic liposome and a carrier DNA. The clinical course of joint inflammation was assessed, and the immunologic balance was analyzed by measuring T helper cell frequencies and cytokine expression. Biodistribution studies of siRNA were also performed. RESULTS: Weekly systemic injection of siPLA2 lipoplexes significantly reduced the incidence and severity of CIA, in both preventive and curative settings, as compared with findings in control animals. Histologic scores for inflammation and cartilage damage were reduced. The clinical effect was associated with local inhibition of tumor necrosis factor α secretion and lower cPLA(2)α expression and activity. The siPLA2 lipoplexes enabled triggering of in vivo RNAi-mediated gene silencing of cPLA(2)α in CD11b+ cells recovered from the spleen. While the treatment had no effect on anti-type II collagen (anti-CII) antibodies, CII-specific T helper cells producing interferon-γ, but not interleukin-17, in draining lymph node cells were decreased. CONCLUSION: Our findings indicate that systemic RNAi-mediated cPLA(2)α gene silencing in CD11b+ cells is effective in the treatment of CIA, and Th1 suppression is one of the potential underlying mechanisms, whereas Th17 suppression is not.

Current data on ATP-containing liposomes and potential prospects to enhance cellular energy status for hepatic applications.

The pharmacological use of adenosine triphosphate (ATP), although promising, is restricted due to poor cellular penetration and drastic hydrolysis that is markedly accelerated in vivo by ectoenzymes. In the literature, liposomes have proven efficient in offering a physical barrier to extracellular enzymes and favor penetration into cells. First, this review addresses the issues raised by ATP development in pharmaceutics. Second, studies conducted with ATP liposomally entrapped (lipo-ATP) are described, including pharmaco-technical formulation engineering and related models of assessment. Finally, potential directions for research to better target ATP penetration into the liver are considered. Lipo-ATP were formulated for a number of applications, including sepsis-related disorders; spermatozoid alteration; brain ischemia episodes; and ophthalmic, cardiac, and hepatic use. Key formulation parameters need to be carefully considered to optimize stability and entrapment yield value, and to define the manufacturing process. Positive lipids, such as stearylamine, increase entrapment yield value by electrostatic interaction with negatively charged ATP. A freezing-thawing step in the manufacturing process considerably increases entrapment yield value. Lipo-ATP were assessed using cell culture, isolated organs, and animal experimental models. Very promising results were obtained with antimyosin PEGylated immunoliposomes using isolated rat hearts and experimental myocardial infarction in rabbits. In hepatic applications, lipo-ATP are effective in preventing liver injury during shock and to improve the energy status of cold-stored rat liver, in particular, if liposomes are loaded with apolipoprotein E (ApoE). For liver delivery, liposome size needs to be lower than 100 nm to allow diffusion through the Disse space, but liposome flexibility and lipid content may also influence liver uptake. The role of the liposome charge remains unclear. ApoE and the ligand for the asialoglycoprotein receptor [ASGPr) were both used in the literature, but the ASGPr seems more promising. Ligand-ASGPr interaction is based on the sugar preference (N-acetylgalactosamine>>galactose), the antennary structure (tetra>tri>di>monoantennary), and sugar spacing. Numerous high-affinity ligands have been extracted or designed to target hepatocytes, which can be classified according to their origin (i.e., natural, hemisynthetic, or synthetic). Synthetic ASGPr, such as Gal-C4-Chol (cholesten-5-yloxy-N-(4-((1-imino-2-D-thiogalactosylethyl)formamide), are composed of a lipid anchor (e.g., cholesteryl), a spacer (C2 to C6 chain), and a sugar head (galactose or lactose). The formulation includes ligand incorporation, by either simple preincubation or covalent graft, onto preformulated liposomes or direct mixing with other lipids. The ligand-loaded liposomes encapsulated pharmacological agents, markers, or plasmid DNA. Interesting results were obtained with antitumor or antioxidant agents to promote drug penetration in cell culture (e.g., primary rat hepatocyte or HepG2) and specific targeting to hepatocyte in isolated perfused liver (pharmacokinetic studies). Effectiveness was demonstrated in experimental models (e.g., tumor-bearing animals and hepatotoxic models). These targeted formulations were less toxic than standard formulations and controls. A development scheme that can be applied to other drugs, which may benefit from improved hepatic targeting, is proposed to optimize liposome characteristics and ligand structure, including verifications such as the displacement-binding test, ligand incorporation, cell internalization, tissue diffusion, organ and receptor specificity, and efficiency in experimental models.

Cationic lipid-mediated gene transfer: effect of serum on cellular uptake and intracellular fate of lipopolyamine/DNA complexes.

Most of the cationic lipids used for gene transfer experiments drastically lose their efficiency in the presence of serum. We used a cationic lipid with a spermine head group and its fluorescent analog to study the cellular uptake and the intracellular fate of lipoplexes in the presence and absence of serum. We found that the amount of DNA and lipid taken up by the cells was not related to the efficacy of the gene transfer. When the lipofection was performed in the presence of serum, lipoplexes were contained within small intracellular vesicles. In the absence of serum, the vesicles were larger and heterogeneous in size and shape. By analysis of their size distribution, we showed that lipoplexes preformed in the absence of serum tended to aggregate. This aggregation was inhibited in the presence of serum. We used a carbonate formulation that led to the preformation of large particles: those large particles gave a high lipofection efficiency in the presence of serum and their intracellular distribution was identical to that observed in the absence of serum.

Inhibition of osteolysis and increase of bone formation after local administration of siRNA-targeting RANK in a polyethylene particle-induced osteolysis model.

Receptor activator of nuclear factor kappa-B (RANK) and RANK-ligand are relevant targets for the treatment of polyethylene particle-induced osteolysis. This study assessed the local administration of siRNA, targeting both human RANK and mouse Rank transcripts in a mouse model. Four groups of mice were implanted with polyethylene (PE) particles in the calvaria and treated locally with 2.5, 5 and 10 µg of RANK siRNA or a control siRNA delivered by the cationic liposome DMAPAP/DOPE. The tissues were harvested at day 9 after surgery and evaluated by micro-computed tomography, tartrate-resistant acid phosphatase (TRAP) immunohistochemistry for macrophages and osteoblasts, and gene relative expression of inflammatory and osteolytic markers. 10 µg of RANK siRNA exerted a protective effect against PE particle-induced osteolysis, decreasing the bone loss and the osteoclastogenesis, demonstrated by the significant increase in the bone volume (P<0.001) and by the reduction in both the number of TRAP(+) cells and osteoclast activity (P<0.01). A bone anabolic effect demonstrated by the formation of new trabecular bone was confirmed by the increased immunopositive staining for osteoblast-specific proteins. In addition, 5 and 10 µg of RANK siRNA downregulated the expression of pro-inflammatory cytokines (P<0.01) without depletion of macrophages. Our findings show that RANK siRNA delivered locally by a synthetic vector may be an effective approach for reducing osteolysis and may even stimulate bone formation in aseptic loosening of prosthetic implants.

Synthesis, activity, and structure--activity relationship studies of novel cationic lipids for DNA transfer.

We have designed and synthesized original cationic lipids for gene delivery. A synthetic method on solid support allowed easy access to unsymmetrically monofunctionalized polyamine building blocks of variable geometries. These polyamine building blocks were introduced into cationic lipids. To optimize the transfection efficiency in the novel series, we have carried out structure-activity relationship studies by introduction of variable-length lipids, of variable-length linkers between lipid and cationic moiety, and of substituted linkers. We introduce the concept of using the linkers within cationic lipids molecules as carriers of side groups harboring various functionalities (side chain entity), as assessed by the introduction of a library composed of cationic entities, additional lipid chains, targeting groups, and finally the molecular probes rhodamine and biotin for cellular traffic studies. The transfection activity of the products was assayed in vitro on Hela carcinoma, on NIH3T3, and on CV1 fibroblasts and in vivo on the Lewis Lung carcinoma model. Products from the series displayed high transfection activities. Results indicated that the introduction of a targeting side chain moiety into the cationic lipid is permitted. A primary physicochemical characterization of the DNA/lipid complexes was demonstrated with this leading compound. Selected products from the series are currently being developed for preclinical studies, and the labeled lipopolyamines can be used to study the intracellular traffic of DNA/cationic lipid complexes.

Advantages of bioluminescence imaging to follow siRNA or chemotherapeutic treatments in osteosarcoma preclinical models.

Osteosarcoma is the most common malignant primary bone tumor for which pertinent preclinical models are still needed to develop new therapeutic strategies. As osteosarcoma growth is strongly supported by bone resorption, previous studies have inhibited the cytokine receptor activator of nuclear factor-kappaB ligand using antibodies or recombinant proteins. However, its expression has not yet been inhibited using genetic approaches using small interfering RNA. To optimize the delivery of small interfering RNA to its cellular target and demonstrate their efficiency in vivo, two new osteosarcoma models expressing the firefly luciferase enzyme were developed. These luciferase-expressing osteosarcomas showed conserved osteolytic and osteogenic activities in mice and were detectable by in vivo bioluminescence imaging. In comparison with measurement of tumor volume, bioluminescence analysis enabled earlier tumor detection and revealed extensive cell death in response to ifosfamide treatment. Finally, by targeting the luciferase expression into osteosarcoma, we established a protocol for in vivo administration of small interfering RNA combined with cationic liposome.

Differential characterization of neutrophil cytochrome p30 and cytochrome b-558 by low-temperature absorption and resonance Raman spectroscopies.

Cytochrome p30, a novel hemoprotein isolated from rabbit peritoneal neutrophils [Escriou, V., Laporte, F., Garin, J., Brandolin, G. & Vignais, P. V. (1994) J. Biol. Chem. 269, 14007-14014] has been characterized by low-temperature (77 K) absorption and resonance Raman spectroscopies. The spectral data have been compared with those obtained with neutrophil cytochrome b-558. At room temperature, the absorption difference spectra (reduced minus oxidized) of cytochrome p30 and cytochrome b-558 could not been distinguished from each other. However, at 77 K, significant differences were observed. In particular, the alpha band of cytochrome p30 was split whereas that of cytochrome b-558 was symmetrical, but particularly broad. The resonance Raman spectra of cytochrome p30 provided evidence for the presence of two hemes both in the ferric and ferrous states. One of them was a six-coordinated low-spin heme either oxidized or reduced whereas the other one was a high-spin heme, five-coordinated in the reduced state and six-coordinated in the oxidized state. It is probable that two histidine residues constitute the axial ligands of the six-coordinated low-spin heme of cytochrome p30. The resonance Raman spectra of cytochrome b-558 allowed the detection of a six-coordinated low-spin heme, similar to that found in cytochrome p30. The component typical of the high-spin heme of cytochrome p30 was however absent in the spectra of oxidized and reduced cytochrome b-558.

Assessment of the flavoprotein nature of the redox core of neutrophil NADPH oxidase.

The latent NADPH oxidase activity of purified cytochrome b(558) from rabbit peritoneal neutrophils was expressed in a cell-free system consisting of either gel-filtrated cytosol from resting neutrophils, or a mixture of the three cytosolic activation factors, namely p47, p67 and the G protein Rac1. The cell-free system was supplemented with arachidonic acid and GTPgammaS. With gel-filtrated cytosol, the oxidase activity was relatively high (22 moles O(2)(-)/s/mole heme b in the absence of added FAD), and enhanced by less than one fourth upon addition of FAD. In contrast, with the purified cytosolic activation factors the rate of O(2)(-) production was low (8 moles O(2)(-)/s/mole heme b), and enhanced more than two-fold by a saturating concentration of FAD. The specificity of FAD was demonstrated by the lack of effect of FMN. FAD was determined together with heme b and the oxidase activity in eluates from a Sepharcryl column at the last step of the purification of cytochrome b(558). In the eluted fraction that contained both the maximal inducible oxidase activity and the highest amount of heme b, the molar amount of FAD was 20 times less than that of heme b. It is concluded that cytochrome b(558) is an NADPH-dependent flavocytochrome oxido-reductase (NADPH oxidase) in which one part of FAD is firmly bound and another, loosely attached. On the other hand, there may exist a parallel pathway of electron transfer from NADPH via distinct FAD dehydrogenase(s) to the heme b component of the NADPH oxidase.

Novel method for covalent fluorescent labeling of plasmid DNA that maintains structural integrity of the plasmid.

We have developed a chemical strategy for covalent coupling of fluorophores to plasmid DNA. A p-azido-tetrafluoro-benzyl-lissamine conjugate was synthesized and purified. This conjugate was used to covalently associate fluorescent molecules to plasmid DNA by photoactivation. In contrast to nick-translated plasmid DNA, plasmid-lissamine conjugates appeared on gel as supercoiled DNA. Reporter gene was expressed after transfection of the plasmid-lissamine conjugates in NIH 3T3 cells, although gene transfer efficiency was decreased by 60% as compared with unlabeled DNA. Intracellular traffic of plasmid-lissamine conjugates was studied in transfected cells. After cytoplasmic microinjection, fluorescent plasmid did not diffuse from the site of injection and appeared to be progressively degraded in the cytoplasm.

Purification and physical properties of a novel type of cytochrome b from rabbit peritoneal neutrophils.

The main redox component of the O2- generating oxidase complex in neutrophils is believed to be a b-type cytochrome, named cytochrome b558. In the course of purification of cytochrome b558 from rabbit peritoneal neutrophils, another hemoprotein with an apparent molecular mass of 30 kDa, referred to as p-30, was isolated. Although the spectrum of p-30 was virtually identical to that of cytochrome b558, its redox potential, Em,7 = -4 +/- 10 mV, was much less negative than that of cytochrome b558 (-270 +/- 5 mV). The alkaline pyridine hemochrome from purified p-30 was typical of a b-type cytochrome. The 20 N-terminal amino acid residues and some tryptic peptides isolated from p-30 did not show any significant sequence homology to the human phagocyte cytochrome b558 or to mitochondrial and microsomal cytochromes, except for the N-terminal region which displayed some homology to that of rat liver P-450. After subcellular fractionation, p-30 was found to be located in the plasma membrane and the granule fractions, similarly to cytochrome b558. Upon neutrophil activation, part of p-30 was transferred from granules to the plasma membrane.

Triple helix formation on plasmid DNA determined by a size-exclusion chromatographic method.

Triple-helix-forming oligodeoxynucleotides are receiving considerable attention due to their potential applications for the inhibition of specific genes in vivo. However, their development is impaired by the lack of triple helix formation under physiological conditions. It is thus crucial to be able to quantitatively assay triple helix formation of various oligodeoxynucleotides on different target sequences. Usual methods to detect triple helix formation are restricted under the experimental conditions that can be studied. In addition, quantitative techniques are limited. We present a novel method for rapid detection and quantification of triple helix formation between an oligodeoxynucleotide and a plasmid carrying a target sequence. The oligodeoxynucleotide was radiolabeled and, after incubation with the target plasmid, the unbound oligodeoxynucleotide was separated from the mixture of plasmids and plasmid-bound oligodeoxynucleotides by rapid gel filtration spun columns. The formation of a triple helix between a target plasmid and several oligodeoxynucleotides was demonstrated and compared. Temperature, sequence and ionic dependencies, and kinetics of association were analyzed. This new technique can be used under a variety of conditions and should allow the rapid determination of optimal conditions required for triple helix formation, as well as the easy selection of an oligodeoxynucleotide that specifically binds with the highest affinity to a target double-stranded sequence.

Critical assessment of the nuclear import of plasmid during cationic lipid-mediated gene transfer.

BACKGROUND: Cationic lipid-mediated gene transfer is a promising approach for gene therapy. However, despite the significant amount of lipoplexes internalized by target cells, transgene expression remains too low. Obstacles to nuclear accumulation of plasmid DNA include: the passage of DNA across the cellular membrane, the dismantling of nucleolipidic particles in the cytoplasm and the nuclear import of plasmid DNA. The purpose of the present study was to evaluate the impact of cell status on cationic lipid-mediated transfer. METHODS: Cells were either growth-arrested (by aphidicolin) or synchronized (by a classical double-thymidine block protocol) and cationic lipid-mediated transfection of these cells was evaluated. For the study of the nuclear import of plasmid DNA, two techniques were developed: microinjection of plasmid DNA into intact cells, and the use of cells permeabilized with digitonin. RESULTS: When CV-1 cells were growth-arrested by aphidicolin, cationic lipid-mediated gene transfer was inhibited. Hela cells were synchronized and incubated with lipoplexes at different times after release of the block. Gene expression was greatly enhanced when cells underwent mitosis. When transfection was performed during the early period after block release, when fewer than 5% of the cells had divided, gene expression was carefully quantified and could be attributed to cells that escaped cell cycle block. However, by direct analysis of nuclear import of GFP-coding plasmid using cytoplasmic microinjection, GFP expression could be detected in a few cells that had not divided. CONCLUSIONS: Cationic lipid-mediated gene transfer efficiency increased when cells underwent mitosis. However, when cells did not divide, gene transfer was not completely abolished. Nuclear import of plasmid was greatly facilitated by a mitotic event. In non-mitotic cells, nuclear envelope crossing by plasmid DNA could be detected but was a very rare event.

Cationic lipid-mediated gene transfer: analysis of cellular uptake and nuclear import of plasmid DNA.

Cationic lipids are widely used for gene transfer in vitro and show promise as vectors for in vivo gene therapy applications. However, there is limited understanding of the cellular mechanisms involved in nonviral gene transfer. We investigated two major steps that could be limiting barriers to cationic lipid-mediated gene transfer in vitro. We used a fluorescent plasmid to study the cellular uptake and the intracellular fate of lipoplexes during in vitro transfection of fibroblast cells and found that 100% of the cells take up lipoplexes. The intracellular staining observed with lipoplexes was clearly different from that obtained with endocytosed fluorescent dextran. This suggests that cells readily take up lipoplexes by a mechanism that could be different from endocytosis in our conditions. However, the escape of DNA from intracellular vesicles could be a major limiting barrier to gene transfer. Direct injection of plasmid DNA into the nucleus and cytoplasm of cells indicated that DNA traffic from the cytoplasm to the nucleus might be also an important limiting step.

Intracellular fate and nuclear targeting of plasmid DNA.

One of the major steps limiting nonviral gene transfer efficiency is the entry of plasmid DNA from the cytoplasm into the nucleus of the transfected cells. The nuclear localization signal (NLS) of the SV40 large T antigen is known to efficiently induce nuclear targeting of proteins. We have developed two chemical strategies for covalent coupling of NLS peptides to plasmid DNA. One method involves a site-specific labeling of plasmid DNA by formation of a triple helix with an oligonucleotide NLS peptide conjugate. After such modification with one NLS peptide per plasmid molecule, plasmid DNA remained fully active in cationic lipid-mediated transfection. In the other method, we randomly coupled 5-115 p-azidotetrafluorobenzyllissamine-NLS peptide molecules per plasmid DNA by photoactivation. Oligonucleotide-NLS and plasmid lissamine-NLS conjugates interacted specifically with the NLS-receptor importin alpha. Plasmid-lissamine-NLS conjugates were not detected in the nucleus, after cytoplasmic microinjection. Plasmids did not diffuse from the site of injection and plasmid-lissamine-NLS conjugates appeared to be progressively degraded in the cytoplasm. The process of plasmid DNA sequestration/degradation stressed in this study might be as important in limiting the efficiency of nonviral gene transfer as the generally recognized entry step of plasmid DNA from the cytoplasm into the nucleus.

Efficient purification of plasmid DNA for gene transfer using triple-helix affinity chromatography.

Plasmid DNA used for nonviral therapeutic gene transfer or nucleic acid vaccination has to be highly purified devoid of contaminating components such as bacterial proteins, endotoxins, or bacterial chromosomal DNA. We have developed a new affinity chromatography technique for plasmid DNA purification: triple-helix affinity chromatography (THAC). This technique is based on the sequence-specific interaction of an oligonucleotide forming a triple-helix with plasmid DNA. The oligonucleotide was covalently linked to a chromatographic matrix, thus providing a reusable affinity support. By inserting a suitable homopurine sequence in the plasmid DNA, it is possible to obtain a triple-helix interaction that will only be stable at mild acidic pH and that will dissociate in alkaline conditions. A crude lysate from a recombinant E. coli, or a pre-purified plasmid DNA, is thus applied at acidic pH on to a THAC column. After extensive washing of the column, purified plasmid DNA is eluted using an alkaline buffer. The binding conditions of the plasmid DNA on to the column have been optimized, as well as the hybridization sequence and the linker group between the matrix and the third strand oligonucleotide. The THAC technique makes it possible to purify in one step supercoiled plasmid DNA, and to significantly reduce the level of contaminating RNA, endotoxins and chromosomal DNA. In particular, a 100-fold reduction of chromosomal DNA contamination over that obtained with conventional techniques can be achieved through a single additional THAC step. Further improvements of THAC technology are possible, and we anticipate that this technique can be scaled up for integration into a full commercial-scale DNA production process.

Synthetic DNA-compacting peptides derived from human sequence enhance cationic lipid-mediated gene transfer in vitro and in vivo.

Cationic lipids can deliver genes efficiently in vitro, but are generally inhibited by the presence of serum, and their efficiency in vivo is much lower than in vitro. An attractive strategy is to induce strong DNA compaction by its association with proteins, before addition of lipids. However the use of whole proteins might present both production and immunological limitations. We have devised a system in which DNA is associated with short peptides derived from human histone or protamine, before the addition of a cationic lipid or polymer. Peptides strongly associating with DNA confer to such peptide-DNA-lipid particles an enhanced in vitro transfection efficiency over that observed with classical DNA/lipid lipoplexes, and particularly confer the capacity to transfect in the presence of serum. This acquisition of serum resistance is cell type-independent, and observed with all four lipopolyamines tested and polyethylenimine. Precompacting DNA with a histone H1-derived peptide enhances cationic lipid RPR 115335-mediated gene transfer in an in vivo model of Lewis lung carcinoma. Apart from their use in peptide-DNA-lipid association, such peptides could be useful as part of chimeric gene delivery vectors presenting a DNA-binding moiety that can be easily associated with other functional domains.