Fast Collisional Lipid Transfer Among Polymer-Bounded Nanodiscs

Some styrene/maleic acid (SMA) copolymers solubilise membrane lipids and proteins to form polymer-bounded nanodiscs termed SMA/lipid particles (SMALPs). Although SMALPs preserve a lipid-bilayer core, they appear to be more dynamic than other membrane mimics. We used time-resolved Förster resonance energy transfer and small-angle neutron scattering to determine the kinetics and the mechanisms of phospholipid transfer among SMALPs. In contrast with vesicles or protein-bounded nanodiscs, SMALPs exchange lipids not only by monomer diffusion but also by fast collisional transfer. Under typical experimental conditions, lipid exchange occurs within seconds in the case of SMALPs but takes minutes to days in the other bilayer particles. The diffusional and second-order collisional exchange rate constants for SMALPs at 30 °C are kdif = 0.287 s−1 and kcol = 222 M−1s−1, respectively. Together with the fast kinetics, the observed invariability of the rate constants with probe hydrophobicity and the moderate activation enthalpy of ~70 kJ mol−1 imply that lipids exchange through a “hydrocarbon continuum” enabled by the flexible nature of the SMA belt surrounding the lipid-bilayer core. Owing to their fast lipid-exchange kinetics, SMALPs represent highly dynamic equilibrium rather than kinetically trapped membrane mimics, which has important implications for studying protein/lipid interactions in polymer-bounded nanodiscs.


Kinetics of phospholipid exchange among SMALPs
The transfer of lipid molecules among nanoparticles can take place through (i) desorption and interparticle diffusion of lipid monomers through the aqueous phase 1-3 and (ii) lipid exchange through particle collisions [4][5][6][7] . If the particles making up the two populations that exchange lipid molecules are identical in size and shape, the observed rate constant resulting from monomer diffusion takes the form 4,5,7 : where k dif is the diffusional exchange rate constant and , ∘ and , are the bulk solution concentrations of lipid in the donor and acceptor populations, respectively. For second-order ("bimolecular") collision-dependent lipid transfer, the observed rate constant reads: where k col is the second-order collisional exchange rate constant. If both of the above processes are at play, the overall observed rate constant is given by the sum of equations (1) and (2) For SMALPs, including higher-order collisional events did not further improve the fit (Supplementary Table 1), suggesting that collisions among more than two SMALPs can be neglected. For example, third-order ("termolecular") collisional transfer would require an additional term proportional to , 2 4,5,7 .

Concentration-dependent TR-FRET decays
Upon mixing fluorescently labelled and unlabelled SMALPs, NBD-PE and Rh-PE redistribute among all available SMALPs. This dilution of the fluorescent probe leads to dequenching of NBD-PE, which manifests in an exponential increase in the fluorescence emission intensity at 530 nm according to: Here, is the intensity at time t, C and < are the baseline-corrected original and final intensities, respectively, and m is the slope of the final baseline, which accounts for linear 3 signal drift at long times. For global data analysis, equation (3) was inserted into equation (4) to yield: In this global fitting equation, k dif and k col are global fitting parameters, whereas C , < , and m are local (i.e., c L -specific) fitting parameters. Best-fit parameter values and 95% confidence intervals were derived by nonlinear least-squares fitting in Excel spreadsheets as detailed elsewhere 8 .

Temperature-dependent TR-SANS decays
Upon mixing h-and d-SMALPs, lipid exchange leads to an exponential decrease in the SLD with time, as given by: where I(t) is the signal intensity at time t, C and < are the initial and final intensities, respectively. According to transition-state theory 9 , the second-order collisional rate constant can be expressed as a function of temperature, T, according to: with R being the universal gas constant, N A Avogadro's number, h Planck's constant, and ΔH ‡ and ΔS ‡ the activation enthalpy and entropy, respectively. M denotes the unit "molar", with , = 1 M being the standard concentration. At the c L values used for TR-SANS, the relative contribution of monomer diffusion to the overall lipid transfer rate is very small (Fig. 1c), so that the first term on the right-hand side of equation (3) can be neglected. Thus, insertion of equation (7) into equation (3) and further into equation (6) yields: In this global fitting equation, ΔH ‡ and ΔS ‡ are treated as global fitting parameters, whereas C and < are local (i.e., T-specific) fitting parameters. Again, best-fit parameter values and 95% confidence intervals were derived by nonlinear least-squares fitting in Excel spreadsheets 8 . In the light of the broad temperature range across which phospholipids in SMALPs transition from the gel to the liquid-crystalline (i.e., fluid) state (Supplementary Figure 1), TR-SANS data were fitted in two ways, namely, once including and once excluding the data acquired at 11.1 °C. However, no significant differences were found in the best-fit values of ΔH ‡ or ΔS ‡ . 4 The entropic component at arbitrary c L values is related to the above standard-state value at c L = 1 M through: The apparent values of the molar activation enthalpy and entropy were corrected for the temperature dependence of the buffer viscosity as detailed elsewhere 10 .

Diffusion-limited collisional lipid transfer
In order to estimate the lipid-exchange efficiency of SMALP/SMALP collisions, we determined the diffusion-limited collisional rate constant that would be applicable to nonionic particles as 11 : where η denotes the dynamic viscosity of the buffer.

Supplementary Figure
Supplementary

Supplementary Tables
Supplementary