Topology cubic membranes

Figure 7.5 A multicontinuous G-PCS (the gyroid) differei tiating sieve-elements (see cdso Fig. 7.6). (a) A [211] projection of a double bilayer G-PCS. The match of the theoretical projection is excellent (as well as the correlation between the Fourier transforms (a (experiment) and a (theory)) and it is eeisily seen that a single bilayer G-PCS does not account for the experimental projection (see (c)). (b) Serial section of (a). Note the apparent pleomorphic behaviour of the cubic membrane in (b), which shows co-existing D- and G-morphologies, related by an intersection-free, and thus, topologically constrained transformation, (c bottom) Computer generated projections of the single G [211] and the double bilayer G2 [211] projections. Figures (a) and (b) are modified from [32], with permission.

Besides the asymmetry between monolayers in cytomembranes, two of the more obvious differences between cubic phases and membranes are the unit cell size and the water activity. It has been argued that tire latter must control the topology of the cubic membranes [15], and hence tiiat the cubic membrane structures must be of the reversed type (in the accepted nomenclature of equilibrium phase behaviour discussed in Chapters 4 and 5 type II) rather than normal (type I). All known lipid-water and lipid-protein-water systems that exhibit phases in equilibrium with excess water are of the reversed type. Thus, water activity alone cannot determine the topology of cubic membranes. Cubic phases have recently been observed with very high water activity (75-90 wt.%), in mixtures of lipids [127], in lipid-protein systems [56], in lipid-poloxamer systems [128], and in lipid A and similar lipopolysaccharides [129,130]. 

Recognition of the existence of topologically distinct spaces through the formation of cubic membranes implies that cell space is restricted and predetermined, in the fashion outlined above. This h)rpothesis leads to many consequences. For example, an alternative mechanism for intracellular transport and orgaitisation has been suggested [56] which does not depend upon the formation of vesicles (outlined in Chapter 3), though that is not excluded. 

It is believed that the Gaussian bending modulus k controls the membrane topology. In particular, a negative value of this constant is needed for stable bilayers. A positive value will induce nonlamellar topologies, such as bicontinuous cubic phases. Therefore, it is believed that k is negative for membranes. 

Fig. 24. a) Schematic illustration of the "stretching" of water channel junctions during the continuous transformation between the D and G cubic phases, which occur with no disruption of the bilayer topology. A junction of four water channels in the Qu° phase is converted into two three-way junctions in the Qu° phase, b) Possible mechanism of membrane fusion the monolayers of two apposed lipid bilayers mix to form a stalk intermediate that expands radially to a trans monolayer contact (TMC), leading to rupture as a result of curvature and interstitial stresses and finally to the formation of a fusion pore. 

---