Dnalipid Interactions

Mixed in solution with cationic lipids (CLs), DNA spontaneously forms CL-DNA aggregates of submicron size. These DNA-lipid aggregates, sometimes called ''lipoplexes,'' (88) are routinely used for cell transfection in vitro. More important, they are used primarily as potential gene delivery vehicles for in vivo gene therapy [for recent reviews, see (89-94) and references therein]. Under appropriate conditions these aggregates reveal complex underlying thermodynamic phase behavior. There is a practical paradox here. We use stable equilibrium structures to reveal the forces that cause aggregation and assembly; we use this knowledge of forces to create the unstable preparations likely to be most efficient in transfec-tion.

Lipoplexes for transfection were first proposed by Felgner and coworkers (95,96). The guiding idea was to overcome the electrostatic repulsion between cell membranes (containing negatively charged lipids) and negative DNA by complexing DNA with positively charged CL. Preliminary experimental data showed that at least some lipoplexes deliver DNA through direct fusion with the cell membrane (97). More often, however, lipoplex internalization probably proceeds through endocytosis after initial interaction with the cell's membrane.

Prior to the attempts to use lipoplexes for transfection, studies of DNA aggregated with multivalent cations and coated with negatively charged liposomes were also explored as possible vectors. It was hoped that CL-DNA complexes would no longer require an additional complexing agent, and that also, the transfection efficiency would be higher. The complex's lipid coating could protect the tightly packed DNA cargo during its passage to the target cells.

Although not confronted with the immunological response, risked by the alternative viral vector strategy, the use of li-poplexes in gene therapy is still hampered by toxicity of the CL and low in vivo tranfection efficiency, despite the in vitro efficiency of some CL formulations. This discrepancy can be attributed to the multistage and multibarrier process the complexes must endure before transfection is achieved. These steps typically include passage in the serum, interaction with target and other cells, internalization, complex disintegration in the cytoplasm, transport of DNA into the nucleus, and ultimately expression.

In the search for increasingly more potent gene delivery vectors, the intimate relationship between the lipoplex's phase structure (or morphology) and its transfection efficiency probably serves as the greatest motivation for their study. How is transfection affected by lipoplex morphology? How may this structure be controlled? Experiment and theory of the past decade shed some light on such fundamental questions. They may give perspective for future strategies to design CL-based nonviral vectors.

To this end, we present our current understanding of the structure and phase behavior of CL-DNA complexes. We review the relation of structure to transfection efficiency and, more specifically, to the way the complex formation overcomes one barrier to DNA release into the cytoplasm.

A. Structure of CL-DNA Complexes

In general, the structures of CL-DNA composite phases can be viewed as morphological hybrids of familiar pure-lipid and pure-DNA phases. A first example is the lamellarlike structure initially proposed by Lasic at al. (99,100). The first comprehensive and unambiguous evidence for this structure came from a series of studies by Radler et al. (101-105). From high-resolution synchrotron X-ray diffraction and optical microscopy, they reported the existence of novel lamellar CL-DNA phase morphologies. In particular, one complex structure was shown to consist of lamellar multilayer. In this case smecticlike stacks of mixed bilayers, each composed of a mixture of CL—for example, dioleoyltrimethylammonium propane (DOTAP)—and neutral ''helper'' lipid—for example, dioleoylphosphatidylcholin (DOPC)—with monolayers of DNA strands intercalated within the intervening water gaps (Fig. 19A), like a multilipid bilayer La phase (106). Helper lipids are often added for their fusogenic properties. Dioleoyl-phosphatidylethanol amine (DOPE), for example, is conjectured to promote transfection. In addition, because pure (syn-

0 0

Post a comment