DNA molecules condense into compact structures in the presence of a critical concentration of multivalent cations. To probe the contribution of electrostatic forces to condensation, we used mixtures of water with methanol (MeOH), ethanol (EtOH), and isopropanol (iPrOH) to vary the dielectric constant epsilon from 80 to 50. The condensation of pUC18 plasmids by hexaammine cobalt (III), Co(NH3)(3+)6, was monitored by total intensity and dynamic light scattering, electron microscopy, and CD. The total scattering intensity increased as epsilon went from 80 to 70, and the decreased as epsilon decreased further. Ultraviolet spectrophotometry confirmed that the loss of intensity at low epsilon was not due to the particles' settling out of solution. The rate as well as the extent of condensation increased as epsilon was lowered from 80 to 70, and also depended on the species of alcohol (MeOH < EtOH < iPrOH). The hydrodynamic radii RH of the particles, however, remained roughly the same at 300-350 A and was independent of the species of alcohol. RH increased below epsilon = 70. The critical concentration of Co(NH3)6(3+) required to induce DNA condensation decreased from 21 microM to about 16 microM as the dielectric constant decreased from 80 to 70, and decreased moderately with the nonpolarity of the alcohol. The fraction of DNA charge neutralized at the onset of DNA condensation was calculated by a modification of Manning's two-variable counterion condensation theory to be 0.90 +/- 0.01, independent of epsilon. By electron microscopy we observed that the condensed particles changed from about 93% toroids at epsilon = 80 to 89% rods at epsilon = 70 and 98% rods at epsilon = 65. At epsilon lower than 65, DNA collapsed into a network of multistranded fibers. The morphology of condensed DNA particles, whether toroids, rods, or fibers, was independent of the alcohol species. CD spectra in ethanol-water mixtures indicated that both closed circular and linearized plasmids were in the B conformation when condensed with Co(NH3)6(3+) at epsilon > or = 70, although the closed circular molecules exhibited a weak psi-DNA spectrum. A transition from the B to A form took place between epsilon = 70 and 60, well above the normal dielectric constant of epsilon = 40 for this transition, indicating that ethanol and Co(NH3)6(3+) synergistically promote the B-A transition. We interpret these results to mean that alcohols have both electrostatic and structural effects on DNA, leading to three regimes of condensation. At the lowest alcohol concentrations the B conformation is stable and condensation is relatively slow, allowing time for the packing adjustments necessary to form toroids.(ABSTRACT TRUNCATED AT 400 WORDS)

Multivalent cations condense DNA in vitro, but it had been thought that a valence of at least + 3 was required in aqueous solution. We have found that Mn2+ can produce toroidal condensates of supercoiled plasmid DNA, but not of linearized plasmid. Mg2+ does not cause condensation, and neither MgCl2 nor NaCl can negate the effect of MnCl2, indicating that the condensation mechanism with Mn is not primarily electrostatic. Supercoiled MnDNA is more extensively digested than the linear form by S1 nuclease. Supercoiling appears to cooperate with Mn2+ in stabilizing helix distortions and also provides a "pressure" that enhances lateral association.

Although growth factor proteins display potent tissue repair activities, difficulty in sustaining localized therapeutic concentrations limits their therapeutic activity. We reasoned that enhanced histogenesis might be achieved by combining growth factor genes with biocompatible matrices capable of immobilizing vectors at delivery sites. When delivered to subcutaneously implanted sponges, a platelet-derived growth factor B-encoding adenovirus (AdPDGF-B) formulated in a collagen matrix enhanced granulation tissue deposition 3- to 4-fold (p < or = 0.0002), whereas vectors encoding fibroblast growth factor 2 or vascular endothelial growth factor promoted primarily angiogenic responses. By day 8 posttreatment of ischemic excisional wounds, collagen-formulated AdPDGF-B enhanced granulation tissue and epithelial areas up to 13- and 6-fold (p < 0.009), respectively, and wound closure up to 2-fold (p < 0.05). At longer times, complete healing without excessive scar formation was achieved. Collagen matrices were shown to retain both vector and transgene products within delivery sites, enabling the transduction and stimulation of infiltrating repair cells. Quantitative PCR and RT-PCR demonstrated both vector DNA and transgene mRNA within wound beds as late as 28 days posttreatment. By contrast, aqueous formulations allowed vector seepage from application sites, leading to PDGF-induced hyperplasia in surrounding tissues but not wound beds. Finally, repeated applications of PDGF-BB protein were required for neotissue induction approaching equivalence to a single application of collagen-immobilized AdPDGF-B, confirming the utility of this gene transfer approach. Overall, these studies demonstrate that immobilizing matrices enable the controlled delivery and activity of tissue promoting genes for the effective regeneration of injured tissues.

Tissue repair is driven by migratory macrophages and fibroblasts that infiltrate injury sites and secrete growth factors. We now report the enhancement of skeletal muscle repair by targeting transgene delivery to these repair cells using matrix-immobilized gene vectors. Plasmid and adenovirus vectors immobilized in collagen-gelatin admixtures were delivered to excisional muscle wounds, and when encoding either fibroblast growth factor-2 (FGF2) or FGF6 transgenes, produced early angiogenic responses that subsequently remodeled into arteriogenesis. FGF2 gene delivery enhanced the number of CD31(+) endothelial cells present at treatment sites > 6-fold by day 14, and muscular arteriole density up to 11-fold by day 21 (P<0.0001). Muscle repair was also enhanced, as FGF gene-treated wounds filled with regenerating myotubes expressing the marker CD56 (an average 20-fold increase in CD56 expression versus controls, P<0.0001). These responses required transfection of a threshold level of repair cells, achievable only in injured muscles, and were transgene-driven, as neither platelet-derived growth factor-B (PDGFB) gene nor FGF2 protein delivery produced equivalent responses. In conclusion, using biomatrices to direct gene delivery to repair cells allows for relatively complex regenerative processes such as arteriogenesis and myogenesis, and therefore represents a promising approach to treating injured and ischemic muscle.

Several growth factor proteins have been evaluated as therapeutic agents for the treatment of chronic dermal wounds. Unfortunately, most have failed to produce significant improvements in wound healing, in part due to ineffective delivery and poor retention in the wound defect. It has been proposed that gene therapy might overcome the limitations of protein therapy via ongoing transcription and translation, thus prolonging the availability of the therapeutic protein. Reasoning that it would be of further benefit to ensure retention of the DNA vector as well as the therapeutic protein within the wound defect, we have evaluated matrix-enabled gene transfer for cutaneous wound repair (Gene Activated Matrix). Formulations consisting of bovine type I collagen mixed with adenoviral or plasmid gene vectors have been evaluated in 3 in vivo models. The therapeutic transgenes employed encode human platelet-derived growth factor-A or -B, proteins key to each phase of normal wound repair. Increased granulation tissue formation, vascularization, and reepithelialization have been shown compared to controls treated with collagen alone or collagen containing a reporter gene vector. Further enhancements of the tissue repair response have been achieved by combining matrix-enabled gene transfer with molecular targeting, in which the DNA vector is conjugated to a growth factor ligand (basic fibroblast growth factor). These promising results support the clinical evaluation of gene activated matrices for the treatment of chronic dermal wounds.