To better understand the structures formed by the interaction of cationic lipids with DNA, we undertook a systematic analysis to determine the biophysical characteristics of cationic lipid:DNA complexes. Four model cationic lipids with different net cationic charge were found to interact in similar ways with DNA when that interaction was compared in terms of the apparent molar charge ratio of lipid to DNA. When DNA was present in charge excess over the cationic lipid, the complex carried a net negative charge as determined by zeta potential measurements. Under these conditions, some DNA was accessible to ethidium bromide, and free DNA was observed in agarose gels and in dextran density gradients. Between a lipid:DNA charge ratio of 1.25 and 1.5:1, all the DNA became complexed to cationic lipid, as evidenced by its inaccessibility to EtBr and its complete association with lipid upon agarose gel electrophoresis and density gradient separations. These complexes carried a net positive charge. The transition between negatively and positively charged complexes a occurred over a very small range of lipid to DNA ratios. Employing a fluorescent lipid probe, the addition of DNA was shown to induce lipid mixing between cationic lipid-containing vesicles. The extent of DNA-induced lipid mixing reached a maximum at a charge ratio of about 1.5:1, the point at which all the DNA was involved in a complex and the complex became positively charged. Together with freeze-fracture electron micrographs of the complexes, these biophysical data have been interpreted in light of the existing models of cationic lipid:DNA complexes.

Systemic delivery of cationic lipid-plasmid DNA (pDNA) complexes induces an acute inflammatory response with adverse hematologic changes and liver damage. Immunostimulatory CpG motifs in the pDNA are known to contribute substantially to this response. Here we constructed a pDNA vector (pGZB) that has been depleted of 80% of the CpG motifs present in the original vector. Compared with the unmodified vector, systemic administration of pGZB induced considerably fewer changes in blood parameters, reduced levels of inflammatory cytokines, and decreased liver damage. Despite the extensive sequence modifications within pGZB, there were still robust levels of transgene expression. Furthermore, in contrast to the transient expression observed from the unmodified vector, we observed sustained or increasing expression for up to 49 days from pGZB in the lung and liver of immunocompetent BALB/c mice. Studies adding CpG motifs in trans or in cis indicate that the reduced CpG content of pGZB was the major determinant for its persistent expression. This combination of decreased toxicity and sustained expression suggests that CpG-depleted pDNA vectors can greatly improve the safety and efficacy of synthetic gene delivery systems.

A major limitation associated with systemic administration of cationic lipid:plasmid DNA (pDNA) complexes is the vector toxicity at the doses necessary to produce therapeutically relevant levels of transgene expression. Systematic evaluation of these toxicities has revealed that mice injected intravenously with cationic lipid:pDNA complexes develop significant, dose-dependent hematologic and serologic changes typified by profound leukopenia, thrombocytopenia, and elevated levels of serum transaminases indicative of hepatocellular necrosis. Vector administration also induced a potent inflammatory response characterized by complement activation and the induction of the cytokines IFN-gamma, TNF-alpha, IL-6, and IL-12. These toxicities were found to be transient, resolving with different kinetics to pretreatment levels by 14 days posttreatment. The toxic syndrome observed was independent of the cationic lipid:pDNA ratio, the cationic lipid species, and the level of transgene expression attained. Mechanistic studies determined that neither the complement cascade nor TNF-alpha were key mediators in the development of these characteristic toxicities. Administration of equivalent doses of the individual vector components revealed that cationic liposomes or pDNA alone did not generate the toxic responses observed with cationic lipid:pDNA complexes. Only moderate leukopenia was associated with administration of cationic liposomes or pDNA alone, while only mild thrombocytopenia was noted in pDNA-treated animals. These results establish a panel of objective parameters that can be used to quantify the acute toxicities resulting from systemic administration of cationic lipid:pDNA complexes, which in turn provides a means to compare the therapeutic indices of these vectors.

An inflammatory response is invariably associated with administration of gene transfer complexes composed of cationic lipids and plasmid DNA (pDNA). In the lung, an influx of neutrophils and elevated levels of several proinflammatory cytokines such as TNF-alpha, IFN-gamma, IL-6, and IL-12 characterize this dose-dependent response. The induction of these cytokines was shown previously to be due in part to the presence of unmethylated CpG dinucleotides in the bacterially derived pDNA. We have eliminated 270 of 526 CpG dinucleotides in a reporter plasmid (pCFA-CAT) and tested the inflammatory response to cationic lipid:pDNA complexes containing the modified vector (pGZA-CAT) after intravenous (i.v.) or intranasal (i.n.) delivery into BALB/c mice. Compared to the unmodified vector, the CpG-reduced pGZA-CAT was found to be significantly less immunostimulatory, as the levels of IL-12, IFN-gamma, and IL-6 in the serum 24 h after i.v. delivery were reduced by 40 to 75%. Similar reductions in cytokine levels were also observed in the bronchoalveolar lavage fluids (BALF) after i.n. administration, while the levels of reporter gene expression were not affected by the modifications. We have also investigated known inhibitors of the CpG signaling pathways in order to decrease the inflammatory response. Two such inhibitors, chloroquine and quinacrine, greatly reduced the induction of IL-12 from mouse spleen cells in vitro and inhibited cytokine production in the lung by approximately 50% without affecting gene expression. These results illustrate that use of a less immunostimulatory pDNA vector or inhibitors of CpG immunostimulation can reduce significantly the toxicity associated with cationic lipid:pDNA complexes thereby increasing the therapeutic index of this synthetic gene transfer vector.

Advances in gene therapy vectors and techniques hold promise for treatment of many inherited and acquired diseases. For lung indications, especially those involving the epithelium, delivery of the gene therapy vehicle ideally will involve the use of an aerosol. Aerosol delivery of transgenes using cationic lipids is currently limited by the ability to generate highly concentrated formulations of lipid:DNA complexes that are stable and retain their activity following aerosolization. We have examined many of the variables inherent in aerosolizing cationic lipid gene delivery vehicles and have devised a new formulation that incorporates small amounts of a polyethylene glycol-containing lipid. This formulation has allowed the preparation of concentrated dispersions of cationic lipid:plasmid DNA (pDNA) complexes (> 20 mM pDNA) at approximately 10-fold higher concentrations than previously reported. Most of the pDNA in these formulations was bound to the lipid component and thereby protected from nebulizer-induced shearing; the pDNA also maintained full biological activity both in vitro and in vivo. This new formulation thus represents a significant improvement over current methods to prepare concentrated, active cationic lipid gene delivery vectors, and provides a new tool with which to test gene transfer to the lung.