COMPUTER-SIMULATION OF LIPID DIFFUSION IN A 2-COMPONENT BILAYER - THEEFFECT OF ADSORBING MACROMOLECULES

We have modelled the effects of macromolecular adsorption upon lipid l ateral diffusion in a two-component lipid bilayer or monolayer, which is at a temperature above both of the main transition temperatures. On e set of lipids (binders, b) can bind to the macromolecules with a fre e energy of binding, F(B), while the other set does not bind (non-bind ers, nb). We assumed that no phase separation of the lipids occurs in the absence of adsorbed macromolecules. We represented the lipid bilay er/monolayer by a triangular lattice, each site of which is occupied b y a lipid molecule. Adsorbed macromolecules were represented by hexago ns covering n(H) sites, and we defined a probability per unit of time, p, that a hexagon attempts to adsorb onto the lattice. We considered two sizes of hexagons, n(H) = 7 (Size-1) and n(H) = 19 (Size-2) and di sallowed or permitted adsorbed hexagons to move laterally on the latti ce. We calculate the lipid relative diffusion coefficients, D(nb), and D(b), for three characteristic time-regimes, (i) tau(c) much less tha n tau(a), (ii) tau(c) almost-equal-to tau(a), and (iii) tau(c) much gr eater than tau(a), where tau(c) and tau(a) are the times for proteins to adsorb/desorb or for lipids to move from site to site, respectively . We obtain analytical expressions for D(nb) and D(b) in the first cas e and calculate them using computer simulation in the other two cases. We found that (i) D(alpha)(iii) less-than-or-equal-to D(alpha)(ii) le ss-than-or-equal-to D(alpha)(i) (alpha = nb, b); (ii) D(alpha) could d isplay a shoulder as a function of F(B) for low values of p; (iii) com pared to cases in which lateral diffusion was disallowed, the lateral diffusion of absorbed hexagons appeared to have little effect on D(nb) , but could cause D(b) to increase by 50%. (iv) Scatter in the calcula ted values of D via simulation appeared to be largest for Size-I hexag ons, and could be understood as a consequence of the large interfacial region between areas free of hexagons and areas 'covered' by hexagons . Our results suggest that it is advisable to measure D(b), since D(nb ) might show little change from 1.0 for the values of F and p appropri ate to the system being studied.