Polyfunctional Vectors for Gene Delivery

Barbara Demeneix

UMR CNRS 8572 Muséum National d'Histoire Naturelle Paris, France

Jean-Paul Behr

Universite Louis Pasteur de Strasbourg Illkirch, France


Gene therapy relies on nucleic acid carriers. Viral diversity provides a broad palette of possibilities, thus fulfilling most requirements for gene delivery in clinical situations. However, parallel evolution of viruses and their hosts has made foreign protein particles as well as infected cells effective targets for the immune system. Although the immune response to cells can eventually be turned into therapeutic benefit, immune responses to particles excludes repetitive treatment, which is the only realistic therapeutic approach for a chronic disease. In contrast, artificial carriers can be developed without polypeptide components. They can even be coated with an inert layer, thus escaping most of the immune surveillance. Unfortunately, at best only a few out of a million copies of the gene reach the target cell nucleus. Limited biodistribution and ineffective intracellular trafficking are mainly responsible for this low efficiency.


The routing of a foreign gene to the nucleus of a cell is a complex multistage process that requires a multifunctional vector (Fig. 1). The core of synthetic vectors is invariably a polycation capable of inducing DNA condensation. In effect, conversion of a filiform molecule into a compact particle improves both its chemical stability and physical properties. Plasmid DNA compaction by cationic liposomes or polymers is a quasi-irreversible process that leads to microprecipitates containing hundreds of DNA molecules per particle. For transfection of cells in cul ture, large complexes are advantageous because they sediment onto the cells. As expected, however, their in vivo transfection properties are weak due (among others) to diffusion hindrance.

Unlike the polycationic species mentioned above, oligocations such as spermine or cationic detergents interact with DNA reversibly. Equilibration ensures comparable sizes for all complexes, and entropy tends to direct the system toward the largest number of condensed DNA particles. As a consequence, each particle will be made of a single plasmid molecule (i.e., the smallest possible particle). Unluckily, the other consequence of reversible binding is that DNA complexes do not withstand dilution or binding to polyanions such as proteoglycans (Fig. 1), and hence cannot be used as DNA vectors (1,2).

A chemical solution to this dilemma was found, based on in situ chemical conversion of the cationic detergent into a ca-tionic lipid (Fig. 2) (3). This 2-step process leading to monomo-lecular and stable DNA particles was validated using cysteine-based cationic detergents as condensing agents, and thiol air oxidation into disulfide as the conversion reaction (4). As shown in Fig. 2, the particles all have the same size (25 nm) that corresponds to the volume of a single molecule of plasmid DNA. Moreover, such particles are capable of moving through an agarose gel in electrophoresis conditions, in contrast to classical cationic lipid/DNA complexes, which remain in the wells. To our surprise, the particles moved even faster than plasmid DNA itself (5) (see also Fig. 5). Improved in vivo diffusion within tissues can thus be expected. Intracellular trafficking may be favored, too, especially as noncondensed plasmid DNA was shown to be immobile in the cytoplasm (6). Finally, the size of theparticles remains compatible with active nuclear pore crossing, which may facilitate transfection of postmitotic cells.

DNA release. ö and nuclearfâçolizi

DNA release. ö and nuclearfâçolizi

DNA condensation

* binding ta : cl clustering of ISmnf^^nuRole proteoglycans escape from the vacuole

DNA condensation

* binding ta : cl clustering of ISmnf^^nuRole proteoglycans escape from the vacuole

Figure 1 Gene delivery with synthetic vectors is a complex multistep process. See the color insert for a color version of this figure.

0 0

Post a comment