Figure 25 The phase diagram of a lipid-DNA mixture involving ''curvature-loving'' helper lipid. The spontaneous curvature of the helper lipid is c —1/25 A_1. For the cationic lipid, the spontaneous curvature taken is c = 0 A — 1. The bending rigidity for both lipids is Kc = 10 kBT. The symbols S, H, B, I, and D denote, respectively, the Lca, HCI, La, HII, and uncomplexed (naked) DNA phases. The broken line marks the single HcII phase. (Reprinted by permission from Ref. 143, Biophysical Society.)

faces, the membrane may corrugate so as to optimize its contact with DNA (see Fig. 19A). If the membrane is further softened, finally, a transition may occur to the HcÏI phase. In this respect, the membrane corrugation in the Lca complex may be regarded as a further stabilization of the lamellar complex, and a delay to the onset of the Lca ^ HCI transition.

A possible consequence of membrane corrugation in the Lca phase is an induced locking between neighboring galleries. This follows the formation of ''troughs'' in a gallery, induced by the interaction of the membrane with DNA in adjacent galleries. This imposes ''adsorption sites'' for the DNA in the 2 neighboring galleries, which propagates the order on. The formation of these troughs, as well as a very weak electrostatic interaction between galleries, may thus correlate between the positions of DNA in different galleries (128,144,145). Limited experimental evidence supports this notion. In cryotransmission electron microscopy (cryo-TEM) studies of the Lca phase, spatial correlations were found between DNA strands in different galleries (146). In another series of X-ray studies, the corrugation and charge density modulation in an L^-like complex, in which the membranes are in the gel phase, were measured (147). Further support for the possible formation of corrugations is gained from computer simulations of lipid-DNA complexes (128).

In order to assess the extent of membrane corrugation, a balance of forces between many degrees of freedom should be taken into account. The free energy minimum now depends on the local membrane composition—dictating membrane properties such as local charge density, spontaneous curvature and bending elasticity—and the extent of local deformation around the DNA. Theoretical predictions show that for a wide range of conditions, both stiff and soft membranes can show corrugations that are stable with respect to thermal undulations of the membranes (145). The spacings between galleries and between DNA molecules are also predicted to change somewhat with respect to the case where no corrugations are allowed (144). For the conditions in which the troughs are shallow or absent altogether, one may anticipate the formation of phases where DNA in different galleries are positionally uncorrelated, while orientational order is preserved. These structures were predicted theoretically and termed ''sliding phases'' (103,104,148-150,146).

F. Lipoplex Structure and Transfection Efficiency

In recent years a large number of CL-DNA formulations have been proposed as vectors. However, the fate of the CL-DNA complex once administered, its interaction with the cell membrane, and entry into the cell and subsequently into the cell nucleus, is likely complex and largely unresolved. The poorly understood process of DNA release once in the cell interior must be important (151-153). For example, it has been shown from action in the nucleus that DNA expression is diminished when it is tightly complexed with lipids (156). Hints to the mechanism of the intracellular release of lipoplexes come from experimental evidence in vitro, showing that other added polye-lectrolytes may compete with DNA and subsequently replace it in the complex (154). This kind of replacement, by natural polyelectrolytes, may be one way in which DNA is released in cells (155). Another possible mechanism is the fusion of complex lipids with lipid membranes in the cell (89,104).

Only a limited number of experiments have probed the relationship between the structure of CL-DNA complexes and the transfection efficiency. One emergent theme attributes an important role to complex frustration and destabilization in promoting transfection.

Experimental studies show that the 2 ordered complex structures, Lca and Hcn, behave differently inside living cells. Furthermore, a correlation was found between the structure of the lipoplexes formed and the transfection efficiency. The structure formed depends in turn on the specific choice and relative amount of HL, CL, and DNA. The HcII complex was found (in the studied cases) to be a more potent vector than Lca (157). Further information is gained from fluorescence studies of cell cultures with both complex types internalized in fibroblast L cells. These indicate that the Lca complex is more stable inside the cells, while the HcII more readily disinte-grates—its lipids fusing with the cell's own (endosomal or plasma) membranes—resulting in DNA release. This is in accord with the theoretical findings that the Lca complex structure is rather flexible toward changes in the system's compositional parameters, due to its ability to tune both the membrane composition and the DNA-DNA spacing, while this tuning is more limited in the HcII phase.

The picture is further substantiated by a series of studies by Barenholtz and coworkers (90,152,153,158). In general it was shown that maximal transfection efficiency could be achieved in complexes that were formed in the excess lipid regime (with p in the range of 2-5). This correlated well with the point of maximal size heterogeneity of the complexes.

These instabilities were shown to occur concomitantly with an increase in the amount of membrane defects that were in turn mainly attributed to the appearance of several coexisting structures in solution (e.g., Hcn and Lca in DOTAP/DOPE li-poplexes, or micellar and lamellar phases in DOSPA/DOPE-based lipoplexes). This is in accordance with the theoretical prediction that the regions of most phase diversity and the largest number of coexisting phases occurs at high p (and low values (see Fig. 25) (90,114,140).

Other evidence seems to agree with these notions. For example, some successful formulations, such as BIV, are also probably metastable (97,99,110). This may suggest that it is in fact their instability that helps them to release their DNA cargo once they are inside the cell. Attempts have also been made to destabilize lipoplexes more specifically only once they are already internalized in the cells (rather than en route in the serum). Reduction-sensitive cationic lipids were designed, and the subsequent lipoplexes that are formed were shown to undergo large structural changes when exposed to the cyto-plasmic reductive systems. The lipoplexes are thus destabilized and the previously packaged DNA is released into the cytosol (92,159-161). A decrease in the toxicity of the CL and increased transfection efficiency are thus achieved (162).

Destabilizing lipoplexes is not the only barrier to transfec-tion. For example, entry of DNA into the nucleus through the nuclear pore complex is inefficient for large pieces of DNA. It has been shown that the cell own nuclear import machinery may be used to increase transfection efficiency dramatically, by attaching a peptide containing a nuclear localization signal (NLS) to the DNA (163,164). Furthermore, the size of the complexes also seems to play a crucial role in determining transfection efficiencies (90,91,97,99). Here, the repulsive interaction between like-surface charge of the complex due to over/undercharging (excess lipid or DNA) can aid in stabilizing the complexes, once they are formed, from fusing further. Another strategy to controlling the interaction between aggregates and the stability of the aggregate in vivo is to modify the composition of the outer wrapping sheath of the lipoplex. The caveat is that the li-poplexes are not stabilized to such a degree that they can no longer disintegrate once inside the cells. For example, short-chain lipids possessing a PEG headgroup (or a derivative thereof) have been used to increase the stability of the li-poplexes in the bloodstream, while not interfering with the endosomal unwrapping once the lipoplexes are internalized in cells (165).

More generally, we can expect that understanding how to control and manipulate the formation of specific phases on the one hand, while better understanding the multistage trans-fection mechanism and the parameters (conditions) affecting it on the other, should aid in the design of more potent lipid-based gene delivery vectors in the future. These, together with control over the coating and targeting of the complexes, may render these vectors as useful vehicles in gene therapy.

0 0

Post a comment