Intraperitoneal linear polyethylenimine (L-PEI)-mediated gene delivery to ovarian carcinoma nodes in mice.

Linear polyethylenimine (L-PEI) is an efficient transfection agent for ovarian carcinoma cells in vitro and ex vivo. In the present work, we go a step further and evaluate the efficacy of L-PEI in human ovarian tumor nodes developed in mice. PEI/DNA complexes were administered intraperitoneally instead of intravenously to avoid sequestering of complexes in the lung and liver and to allow transfection of nonvascularized tumor nodes. Plasmid biodistribution was studied by PCR and gene expression was characterized using complementary luciferase and beta-galactosidase assays. Intraperitoneal (i.p.) injection of L-PEI/DNA complexes allowed the straightforward distribution of plasmid in the whole peritoneal cavity. Gene expression occurred in many organs, but tumor nodes appeared as preferential sites for transgene expression. The i.p. delivery route allowed repeated injections and administration of large amounts of DNA (up to 400 mug) without signs of toxicity, even for doses well beyond the intravenous lethal dose. Transgene expression was dose-dependent and transient. However, multiple injections allowed its persistence to increase. These results provide encouraging elements towards the development of PEI-based gene therapy protocols for the treatment of advanced stage ovarian carcinoma.

Development of plasmid and oligonucleotide nanometric particles.

Nucleic acids delivery vectors have shown promising therapeutic potential in model systems. However, comparable clinical success is delayed essentially because of their poor biodistribution and of their ineffective intracellular trafficking. The size of condensed DNA particles is a key determinant for in vivo diffusion, as well as for gene delivery to the cell nucleus. Towards this goal, we have developed cationic thiol-detergents that individually compact plasmid DNA molecules into anionic particles. These particles are then 'stabilized' by air-induced dimerization of the detergent into a disulfide lipid on the template DNA. The particles all measure approximately 30 nm, which corresponds to the volume of a single molecule of plasmid DNA. The gel electrophoretic mobility of the anionic particles was found to be higher than that of the plasmid DNA itself. Similarly, particles formed with a 31-mer oligonucleotide measured 19 nm. Improved in vivo diffusion, as well as improved intracellular trafficking may be inferred from the faster migration of the complexes. Moreover, the size of the particles remains compatible with nuclear pore crossing. Finally, in an attempt to improve the biodistribution of these particles, we have coated the monomolecular particles with a poly(ethylene glycol) corona.

Towards synthetic viruses.

Gene delivery is too complex to be performed with a single carrier molecule. Synthetic multicomponent vectors are being designed that mimic key properties of viruses. Some solutions, such as diverting cell-anchoring molecules or the endogenous nuclear import machinery from their normal function, are directly copied from bacteria and viruses. Some other solutions are original ones: monomolecular genome condensation via detergent dimerization or endosome disruption by the proton sponge effect are not exploited by natural cell invaders. All these components, however, still have to be assembled into a unique supramolecular system, an 'artificial virus'.

Recognition of DNA by strand invasion with oligonucleotide-spermine conjugates.

Modified oligonucleotides bearing spermine groups (ODN-sper) with increased binding affinity to DNA have been synthesized. The ability of these ODN-sper to bind within superhelical double-stranded DNA by strand invasion has been studied. The uptake by a supercoiled plasmid was 3 fold higher for the ODN-sper than for the unmodified oligonucleotides.

Dimerizable cationic detergents with a low cmc condense plasmid DNA into nanometric particles and transfect cells in culture.

The size of condensed DNA particles is a key determinant for in vivo diffusion and gene delivery to cells. Gene molecules can be individually compacted by cationic thiol detergents into nanometric particles that are stabilized by oxidative conversion of the detergent into a gemini lipid. To reach the other goal, gene delivery, a series of cationic thiol detergents with various chain lengths (C(12)-C(16)) and headgroups (ornithine or spermine) was prepared, using a versatile polymer-supported synthetic strategy. Critical micelle concentrations and thiol oxidation rates of the detergents were measured. The formation and stability of complexes formed with plasmid DNA, as well as the size, xi-potential, morphology, and transfection efficiency of the particles were investigated. Using the tetradecane/ornithine detergent, a solution of 5.5 Kpb plasmid DNA molecules was converted into a homogeneous population of 35 nm particles. The same detergent, once oxidized, exhibited a typical lipid phase internal structure and was capable of effective cell transfection. The particle size did not increase with time. Surprisingly, the gel electrophoretic mobility of the DNA complexes was found to be higher than that of plasmid DNA itself. Favorable in vivo diffusion and intracellular trafficking properties may thus be expected for these complexes.

DNA condensation by an oxidizable cationic detergent. Interactions with lipid vesicles.

Cationic amphiphile-mediated delivery of plasmid DNA is the non-viral gene transfer method most often used. In the present work, we considered a new cysteine-detergent, ornithinyl-cysteinyl-tetradecylamide (C(14)-CO), able to convert itself, via oxidative dimerization, into a cationic cystine-lipid. By using fluorescence techniques, we first characterized the structure of complexes of plasmid DNA with C(14)-CO molecules either kept as monomers, or oxidized into dimers. Both forms are able to condense DNA, with the formation of hydrophobic micelle-like domains along the DNA chain. Domains with a larger molecular order were obtained with dimeric C(14)-CO/DNA complexes. In a second step, the interactions of these complexes with lipid vesicles considered as membrane models were investigated. In the presence of vesicles, we observed a decondensation of the DNA involved in complexes obtained with C(14)-CO monomers. With anionic vesicles, the DNA is released into the bulk solution, while with neutral vesicles, it remains bound to the vesicles via electrostatic interactions with inserted C(14)-CO molecules. In sharp contrast, the complexes with C(14)-CO dimers are unaffected by the addition of either neutral or anionic vesicles and show no interaction with them. These results may partly explain the low transfection efficiency of these complexes at the +/-charge ratios used in this study.

Induction of potent human immunodeficiency virus type 1-specific T-cell-restricted immunity by genetically modified dendritic cells.

A novel technology combining replication- and integration-defective human immunodeficiency virus type 1 (HIV-1) vectors with genetically modified dendritic cells was developed in order to induce T-cell immunity. We introduced the vector into dendritic cells as a plasmid DNA using polyethylenimine as the gene delivery system, thereby circumventing the problem of obtaining viral vector expression in the absence of integration. Genetically modified dendritic cells (GMDC) presented viral epitopes efficiently, secreted interleukin 12, and primed both CD4(+) and CD8(+) HIV-specific T cells capable of producing gamma interferon and exerting potent HIV-1-specific cytotoxicity in vitro. In nonhuman primates, subcutaneously injected GMDC migrated into the draining lymph node at an unprecedentedly high rate and expressed the plasmid DNA. The animals presented a vigorous HIV-specific effector cytotoxic-T-lymphocyte (CTL) response as early as 3 weeks after a single immunization, which later developed into a memory CTL response. Interestingly, antibodies did not accompany these CTL responses, indicating that GMDC can induce a pure Th1 type of immune response. Successful induction of a broad and long-lasting HIV-specific cellular immunity is expected to control virus replication in infected individuals.

Inhibition of hepadnaviral replication by polyethylenimine-based intravenous delivery of antisense phosphodiester oligodeoxynucleotides to the liver.

Antisense oligodeoxynucleotides (ODNs) appear as attractive anti-hepatitis B virus (HBV) agents. We investigated in vivo, in the duck HBV (DHBV) infection model, whether linear polyethylenimine (lPEI)-based intravenous delivery of the natural antisense phosphodiester ODNs (O-ODNs) can prevent their degradation and allow viral replication inhibition in the liver. DHBV-infected Pekin ducklings were injected with antisense O-ODNs covering the initiation codon of the DHBV large envelope protein, either in free form (O-ODN-AS2) or coupled to lPEI (lPEI/O-ODN-AS2). Following optimization of lPEI/O-ODN complex formulation, complete O-ODN condensation into a homogenous population of small (20-60 nm) spherical particles was achieved. Flow cytometry analysis showed that lPEI-mediated transfer allowed the intrahepatic delivery of lPEI/O-ODN-AS2 to increase three-fold as compared with the O-ODN-AS2. Following 9-day therapy the intrahepatic levels of both DHBV DNA and RNA were significantly decreased in the lPEI/O-ODN-AS2-treated group as compared with the O-ODN-AS2-treated, control lPEI/O-ODN-treated, and untreated controls. In addition, inhibition of intrahepatic viral replication by lPEI/O-ODN-AS2 was not associated with toxicity and was comparable with that induced by the phosphorothioate S-ODN-AS2 at a five-fold higher dose. Taken together, our results demonstrate that phosphodiester antisense lPEI/O-ODN complexes specifically inhibit hepadnaviral replication. Therefore we provide here the first in vivo evidence that intravenous treatment with antisense phosphodiester ODNs coupled to lPEI can selectively block a viral disease-causing gene in the liver.

Conjugation of folate via gelonin carbohydrate residues retains ribosomal-inactivating properties of the toxin and permits targeting to folate receptor positive cells.

Conjugation of folate to proteins permits receptor-mediated endocytosis via the folate receptor (FR) and delivery of the conjugate into the cytoplasm of cells. Since many cancers up-regulate the FR it has enabled the targeting of toxins to tumor cells resulting in specific cell death. However, current conjugation methods rely on chemistries that can affect certain catalytic subunits, such as the A-chain of the plant toxin gelonin. As a result many folate-targeted toxins are a compromise between receptor/ligand interaction and toxin activity. We describe the first example of folate conjugated to a protein via carbohydrate residues, using a novel SH-folate intermediate. The folate-gelonin conjugate retains over 99% of toxin activity in a cell-free translational assay compared with unmodified gelonin and is able to bind the FR at the same affinity as free folic acid (10(-10) m). Additionally, the conjugate exhibits prolonged inhibition of protein synthesis in FR positive cell lines in vitro. Folate linked to gelonin via amino conjugation exhibits the same affinity for FR as free folic acid but the toxin is 225-fold less active in a cell-free translational assay. The effect of different conjugation methods on toxin activity and the implications for folate targeting of other glycoproteins are discussed.

Improvement of nonviral p53 gene transfer in human carcinoma cells using glucosylated polyethylenimine derivatives.

Polyethylenimine (PEI) derivatives are potent polycationic nonviral vectors for gene transfer. The gene transfer efficiency of glucosylated and galactosylated PEI derivatives was optimized using green fluorescent protein gene as reporter gene in FaDu and PANC3 human carcinoma cell lines. Glucosylated or galactosylated PEI derivatives were found to be slightly less cytotoxic than unsubstituted PEI. Gene transfer efficiency was found to be related to DNA/cell number ratio and optimal gene transfer efficiency was achieved at 4 microg DNA/10(5) cells. PEI-DNA complexes were found to enter cells rapidly and were detected into cytoplasmic vesicles 2 hours post-transfection. Green fluorescent protein gene expression was detected 4-6 hours after transfection and reached maximal value 24 hours post-transfection. The results achieved demonstrated that glucosylated PEI yield higher and longer gene transfer efficiency than unsubstituted PEI. Using glucosylated PEI allowed to achieve significant gene transfer in more than 10% of the total cell population for more than 4 days. These data were then applied to p53 gene transfer in PANC3 cells bearing p53 gene deletion and consequently unable to initiate apoptosis. Using glucosylated PEI, p53 gene transfer was successfully achieved with subsequent recovery of p53 mRNA expression and transient P53 protein expression. P53 protein functionality was further demonstrated because transfected cells underwent apoptosis.

Ca2+-sensitive cytosolic nucleases prevent efficient delivery to the nucleus of injected plasmids.

BACKGROUND: Efficient gene delivery by synthetic vectors is a major challenge in gene therapy. However, inefficient nuclear delivery of cDNA is thought to be a major limiting step in gene transfer using non-viral vectors. It is commonly thought that, in the cytosol, cDNA has to be released from its vector before importation to the nucleus. The stability of naked cDNA in the cytoplasm is not well established. METHODS: cDNA plasmids, either free or complexed with poly(ethyleneimine) (PEI), were microinjected into the cytoplasm of mammalian cells and their turnover was assessed by fluorescence in situ hybridization (FISH). Incubations of cDNA plasmids in cytosolic extracts were also performed. RESULTS: FISH experiments showed that naked cDNA rapidly fade with time when injected into the cytosol. Fading was not observed when naked cDNA plasmids were injected into the nucleus. Incubation of naked cDNA in a cytosolic fraction isolated from mammalian cells reproduced cDNA degradation as observed in microinjection experiments. Nuclease inhibitors, including aurin tricarboxylic acid or Zn2+, prevented in vitro cDNA degradation. The cytosolic nuclease activity was optimal at physiological pH and physiological Ca2+ concentration. By contrast, it was insensitive to Mg2+ or Na+ concentrations. Finally, cDNA complexation with PEI or addition of oligonucleotides prevented in vitro cDNA degradation. CONCLUSION: Altogether, these experiments suggest that cDNA digestion by cytosolic nucleases occur when the decomplexed transgene is present in the cytosol. We propose that the inefficient transfer of cDNA into the nucleus during transfection with synthetic vectors may result from rapid digestion of naked cDNA by a Ca2+-sensitive cytosolic nuclease.

Polyethylenimine-mediated gene transfer into pancreatic tumor dissemination in the murine peritoneal cavity.

Although peritoneal dissemination of cancer cells often occurs at the advanced stages of pancreatic, gastric or ovarian cancers, no effective therapy has been established. Cationic lipid-mediated gene transfer into peritoneal dissemination may offer a prospect of safe therapies, but vector improvements are needed with regard to the efficiency and specificity of the gene transfer. In this study, the intraperitoneal injection of plasmid DNA:polyethylenimine (PEI) complexes into mice was evaluated as a gene delivery system for the peritoneal disseminations. The luciferase and beta-galactosidase genes were used as marker genes. PEI was more efficient than the cationic lipids examined in this study in vivo, and the transgene was preferentially expressed in the tumors. Although PCR analysis showed that the injected DNA was delivered to various organs, the distributed DNA became undetectable by 6 months after the gene transfer. Blood chemistry and histological analysis showed no significant toxicity in the injected mice. This study demonstrated that the intraperitoneal injection of DNA:PEI is a promising delivery method to transduce a gene into disseminated cancer nodules in the peritoneal cavity.

Nonviral gene delivery: Towards artificial viruses.

Several limitations to nonviral gene delivery have been overcome. Small nanometric particles have been obtained by condensation ofDNA with a polymerizable cation followed by DNA template-directed homopolymerization of the vector. Targeting of specific cell types has been achieved by using polyethylenimine (PEI), a cationic polymer, coupled to cell ligands such as galactose or small RGD peptide that allows cell entry by a receptor-mediated endocytosis pathway. Escape of the DNA complexes from the endosomes is favored by the 'proton sponge' effect as a consequence of the high buffering capacity of PEI. The last barrier to gene delivery, i.e. the nuclear membrane, can be crossed at the level of the nuclear pore complexes, by using nuclear localization signal DNA molecule conjugates.

Synthesis and hybridization properties of oligonucleotides containing polyamines at the C-2 position of purines: a pre-synthetic approach for the incorporation of spermine into oligodeoxynucleotides containing 2-(4,9,13-triazatridecyl)-2'-deoxyguanosine.

We have developed a synthesis of spermine-containing oligonucleotides (ODN-sper) which allows incorporation of multiple polyamine residues. This approach was based on the pertrifluoroacetylated 5'DMT-dGsper phosphoramidite synthon. Its coupling yield with resin-bound ODN decreased dramatically when close to the 3'-end. Optimization of the coupling conditions allowed 22-mer ODNs containing up to six spermine residues to be synthesized. Several ODNs of different sequences with 1-4 pendent spermines could be purified and their hybridization properties were evaluated. Duplex melting temperatures increased linearly with the number of polyamine residues (deltaTm/sper = 3.0 +/- 0.2 degrees C in 100mM NaCl). This compares very favorably with values reported for duplexes of similar initial stability containing other cation-substituted bases. Moreover, the stability increase was neither sequence nor position-dependent, and even contiguous spermine residues did not cross-talk. Extrapolation based on these findings leads to the conclusion that a duplex formed with a 22-mer oligonucleotide containing seven spermine residues would be as stable as genomic DNA, which highlights its potential for DNA strand invasion.

Systemic linear polyethylenimine (L-PEI)-mediated gene delivery in the mouse.

BACKGROUND: Several nonviral vectors including linear polyethylenimine (L-PEI) confer a pronounced lung tropism to plasmid DNA when injected into the mouse tail vein in a nonionic solution. METHODS: and results We have optimized this route by injecting 50 microg DNA with excess L-PEI (PEI nitrogen/DNA phosphate = 10) in a large volume of 5% glucose (0.4 ml). In these conditions, 1-5% of lung cells were transfected (corresponding to 2 ng luciferase/mg protein), the other organs remaining essentially refractory to transfection (1-10 pg luciferase/mg protein). beta-Galactosidase histochemistry confirmed alveolar cells, including pneumocytes, to be the main target, thus leading to the puzzling observation that the lung microvasculature must be permeable to cationic L-PEI/DNA particles of ca 60 nm. A smaller injected volume, premixing of the complexes with autologous mouse serum, as well as removal of excess free L-PEI, all severely decreased transgene expression in the lung. Arterial or portal vein delivery did not increase transgene expression in other organs. CONCLUSIONS: These observations suggest that effective lung transfection primarily depends on the injection conditions: the large nonionic glucose bolus prevents aggregation as well as mixing of the cationic complexes and excess free L-PEI with blood. This may favour vascular leakage in the region where the vasculature is dense and fragile, i.e. around the lung alveoli. Cationic particles can thus reach the epithelium from the basolateral side where their receptors (heparan sulphate proteoglycans) are abundant.

Ovarian carcinoma cells are effectively transfected by polyethylenimine (PEI) derivatives.

As a prerequisite to nonviral gene therapy approaches of ovarian carcinoma, we evaluated the possibility of transfecting established tumor cell lines (SKOV3, IGROV1) as well as primary mesothelial and tumor cells by various polyethylenimine (PEI) derivatives. Several PEI-based vectors were able to effectively transfect these cells, as shown by high luciferase expression levels (10(8) to 10(9) relative light units per milligram of cell protein) that corresponded with 25-50% of green fluorescent protein-positive cells after 24 hours. However, unpredictable differences were observed among the vectors and cell types that a posteriori justified the screening procedure. We also showed that cells that were not transfected after the first experiment remained transfectable in a subsequent transfection experiment to a level similar to that of the initial population. This experiment does not support the emergence of a transfection-resistant cell population and opens the door to multiple therapeutic gene deliveries. Although efficacy and cell targeting still remain to be improved, PEI derivatives appear to be promising molecules for the development of nonviral gene therapy of ovarian carcinoma.

Polyethylenimine shows properties of interest for cystic fibrosis gene therapy.

Before being considered for a cystic fibrosis (CF) gene therapy trial, any gene delivery agent must be able to show that it produces low levels of toxicity as well as being able to protect the DNA from nuclease degradation. Here we show that complexes of linear polyethylenimine (L-PEI) and DNA can repeatedly be administered to animals (up to 21 consecutive days) without eliciting an immune response against PEI/DNA particles or inducing toxic side effects due to accumulation of PEI in the lungs. However, the host response to the exogenous protein resulted in some decrease of expression. PEI-mediated transfection was unaffected by treatment of the complexes with DNase (frequently used to reduce the viscosity of lung secretions in CF patients). Taken together, these properties make L-PEI a valuable vector for gene therapy of CF.

In vivo delivery to tumors of DNA complexed with linear polyethylenimine.

Synthetic gene delivery vectors have shown promise in several organs, including brain and lung. Tumor cell targeting, however, is still hindered by their low efficacy. A linear polyethylenimine (L-PEI, Exgen 500) was found to be effective in vivo. Our first attempts to use L-PEI for intratumoral gene delivery were not successful, presumably because of poor diffusion of the complexes within the tumor mass after injection with a syringe. Here we show that L-PEI-mediated transfection can be strongly enhanced when the complexes are delivered slowly into a solid tumor mass, using a micropump. Furthermore, L-PE/DNA complexes actively transfect pseudocystic tumor cells when injected into the cyst cavity. In both cases L-PEI induced a significant and long-lasting (> or =15 days) expression of the reporter gene. Finally, even though systemic delivery of L-PEI/DNA complexes leads to high levels of expression in the lung, this method is not adapted for transfection of subcutaneous tumors implanted in the thigh nor for transfection of lung metastases. Altogether, these results show that L-PEI has promising features for transfection of tumor cells, provided that the mode of delivery is adapted.

Gene transfer with synthetic virus-like particles via the integrin-mediated endocytosis pathway.

The interaction between cationic DNA-containing particles and cell surface anionic proteoglycans is an efficient means of entering cultured cells. Therapeutic in vivo gene delivery levels, however, require binding to less ubiquitous molecules. In an effort to follow adenovirus, thiol-derivatized polyethylenimine (PEI) was conjugated to the integrin-binding peptide CYGGRGDTP via a disulfide bridge. The most extensively conjugated derivative (5.5% of the PEI amine functions) showed physical properties of interest for systemic gene delivery. In the presence of excess PEI-RGD, plasmid DNA was condensed into a rather homogeneous population of 30-100 nm toroidal particles as revealed by electron microscopy images in 150 mM salt. Their surface charge was close to neutrality as a consequence of the shielding effect of the prominent zwitterionic peptide residues. Transfection efficiency of integrin-expressing epithelial (HeLa) and fibroblast (MRC5) cells was increased by 10- to 100-fold as compared with PEI, even in serum. This large enhancement factor was lost when aspartic acid was replaced by glutamic acid in the targeted peptide sequence (RGD/RGE), confirming the involvement of integrins in transfection. PEI-RGD/DNA complexes thus share with adenovirus constitutive properties such as size and a centrally protected DNA core, and 'early' properties, i.e. cell entry mediated by integrins and acid-triggered endosome escape.

Gene delivery: a single nuclear localization signal peptide is sufficient to carry DNA to the cell nucleus.

Translocation of exogenous DNA through the nuclear membrane is a major concern of gene delivery technologies. To take advantage of the cellular import machinery, we have synthesized a capped 3.3-kbp CMVLuciferase-NLS gene containing a single nuclear localization signal peptide (PKKKRKVEDPYC). Transfection of cells with the tagged gene remained effective down to nanogram amounts of DNA. Transfection enhancement (10- to 1,000-fold) as a result of the signal peptide was observed irrespective of the cationic vector or the cell type used. A lysine to threonine mutation of the third NLS amino acid completely abolished these remarkable features, suggesting importin-mediated translocation. Our hypothesis is that the 3-nm-wide DNA present in the cytoplasm is initially docked to and translocated through a nuclear pore by the nuclear import machinery. As DNA enters the nucleus, it is quickly condensed into a chromatin-like structure, which provides a mechanism for threading the remaining worm-like molecule through the pore. A single NLS signal is thus sufficient, whereas many signals on a gene would actually inhibit entry, the same DNA molecule being threaded through adjacent pores.