**4. Discussion**

Both experimental methods applied in the present study for the measurement of the membrane electrical capacitance yield higher *C<sup>m</sup>* values at 200 mmol/L of sucrose in the aqueous surroundings compared to its value in the sugar-free control sample. The capacitance obtained from measurements on GUVs in sugar-free aqueous solutions *<sup>C</sup><sup>m</sup>* <sup>=</sup> 0.51 <sup>±</sup> 0.04 <sup>µ</sup>F/cm<sup>2</sup> is lower than values ~ 0.55 <sup>÷</sup> <sup>1</sup> <sup>µ</sup>F/cm<sup>2</sup> measured for planar bilayers of different lipid composition and charge [28,50,54–57] (data overview in Ref. [58]). In the present study, we confirm this trend by reporting higher specific electrical capacitance from experiments on Montal–Mueller BLMs compared to the values acquired from GUVs. Fluctuating vesicles are characterized by low-tension membranes [59,60], whereas for planar lipid bilayers the lateral tension is orders of magnitude higher [61,62]. Thus, a reduced membrane thickness of BLMs compared to GUVs is expected due to the tension difference for the two bilayer systems. Patch-clamp experiments on GUVs recently explored the bilayer thinning effect on membrane capacitance in 200 mmol/L sucrose and glucose solutions [58]. They have shown that membrane capacitance can vary with tension by up to 3%. The measurements reported in [58] have been performed in the relatively high-tension regime of mN/m. In the present study, we employed fluctuating free-standing vesicles with low membrane tensions, 10−<sup>6</sup> <sup>÷</sup> <sup>10</sup>−<sup>4</sup> mN/m as deduced from fluctuation spectroscopy [59,63–65]. Considering the vesicle electrodeformation involved in the capacitance measurements, here, membrane tensions are slightly elevated to 10−<sup>3</sup> mN/m [39,66,67], which is orders of magnitude lower than the mN/m tensions applied in [58]. Despite the different tension range in [58], it can be concluded that membrane thinning associated to tension only partially explains the differences in the capacitance data acquired from GUVs and from planar lipid membranes. In our previous study [16] providing the first experimental evidence about the influence of sucrose on the electrical properties of lipid bilayers, we explored the possibility of changes in membrane dielectric constant as a factor affecting the capacitance. Here, we relate the reported increase of the electric capacitance of the bilayer and the corresponding increase in its dielectric permittivity [16] in the presence of ≥200 mmol/L sucrose to possible alterations in membrane structure and organization induced by membrane-sugar interactions [64].

The phosphatidylcholines POPC and SOPC studied here are synthetic mixed-acyl glycerophospholipids. The monounsaturated oleic acid residue (18:1) is positioned identically in both lipids (sn-2 position), while the saturated hydrocarbon chain at sn-1 position is different, in palmitic and stearic acid for POPC and SOPC, respectively. POPC bilayers

undergo a gel-to-liquid crystalline phase transition around −2 ◦C [65]. The main phase transition of SOPC membranes occurs around 7 ◦C [66].

The fluorescence spectroscopy of Laurdan and DPH provide information about the lipid packing at different levels in the lipid bilayer, near the glycerol backbone and the hydrophobic core, respectively. Laurdan responds to the degree of hydration at the glycerol level, while DPH fluorescence anisotropy corresponds to the rotational diffusion of the probe in the hydrophobic region of the bilayer. Considering that the lipid molecules studied here are not characterized by truncated chains, we apply interchangeably the terms ordering and packing [67,68].

The amphiphilic fluorophore Laurdan comprises a naphthalene residue linked by an ester bond (hydrophilic) and a chain of lauric fatty acid (hydrophobic). As a result, it inserts the membrane parallel to lipid molecules in a way that its naphthalene moiety is located at the glycerol backbone of the lipid molecule and more precisely, at the level of *sn*-1 carbonyl [69]. Upon excitation by UV-light, the dipole moment of the fluorescent moiety increases leading to reorientation of the surrounding solvent dipoles. The results presented above (Figures 6 and 7) suggest that at high temperatures, membrane becomes more loosely packed, which imparts higher mobility and increased dipolar relaxation at membrane interface. The reported results indicate that at higher temperatures the Laurdan *GP* values decrease for all measured LUV suspension compositions. Our findings support the expected thermotropic behavior of the lipid bilayer, whose packing order decreases upon increasing the temperature.

Reorientations of *sn*-1 and *sn*-2 chains lead to conformational and hydration changes at the glycerol level related also to reorientation of the lipid headgroups. Laurdan position distribution is characterized by a certain width as well as its possible relocation upon excitation. Thereby, the fluorescence wavelength is related to the fluorophore location depth in membrane. Shorter wavelengths are emitted by Laurdan molecules positioned deeper within the lipid bilayer, while red-shifted emission (longer wavelengths) occur from fluorophores located closer to membrane-water interface. Hence, Laurdan experiments are able to distinguish POPC from SOPC lipid ordering. Our results support the hypothesis about sucrose ordering effect in membrane in the proximity of the glycerol backbone (Figure 5; blue-shifted Laurdan emission signal).

The depolarization of fluorescence has been recognized as a reliable parameter for characterization of dynamic features and thermotropic behavior of the hydrophobic regions of lipid membranes and lipoproteins [31]. Measurements of DPH fluorescence anisotropy allow evaluating the hindrance of the fluorophore mobility in the hydrophobic core of the bilayer related to alterations in packing of the aliphatic chains. We obtain that the increase of the sucrose content in the aqueous surrounding leads to hindering the DPH rotational diffusion as a result from to the formation of more ordered liquid phase.

The relative changes, ∆*rDPH* and ∆*GP*, state that the reduction in rotational diffusion and degree of hydration for the corresponding molecular probe is larger in the presence of the disaccharide. This reduction is more considerable for Laurdan (Figure 9). Therefore, the changes in lipid ordering induced by the presence of sucrose are predominantly at the glycerol level rather than in the hydrophobic core.

In order to further elucidate the effect of sucrose binding [64,70] on the electric properties of lipid bilayers, we study the bilayer dipole potential, which occurs due to the hydrated polar headgroups, the glycerol-ester region of the lipids and the functional group dipoles of the terminal methyl groups of hydrocarbon chains. This membrane parameter is still limitedly understood but undoubtedly recognized as an important regulator of membrane structure and function [71]. Numerous examples can be given in this respect such as the modulation of the hydration force interplaying in membrane–membrane and membrane-ligand interactions or the lipid-mediated cellular signaling in cells. The dipole potential arises in a medium over which the dielectric constant is changing in a large interval—from 2 to 80 [71]. Even if the microscopic nature of the interactions leading to

the creation of the dipole potential remains not completely described, some elaborated theoretical models of its origin account for the important role of interfacial water molecules.

The slight increase in the dipole potential reported here for POPC and SOPC bilayers upon the addition of sodium chloride and small carbohydrates is in agreement with the trend measured for monolayers of dimyristoylphosphatidylcholine on the air/water interface at 20 ◦C, which is below the main transition temperature of the lipid [51]. Four sucrose molecules have been reported to displace three water molecules per lipid, thus producing only a weak effect on the dipole potential or the carbonyl groups in monolayers [51]. In contrast to trehalose, sucrose has not been found to interact directly with the phospholipid groups, thus implying that this disaccharide is not expected to replace water molecules in the tightly bound hydration sphere. Upon increasing the sugar concentration, the water activity in the bulk solution decreases [72]. Furthermore, the refractive index changes observed in vesicles under osmotic stress inferred alterations in the extent of hydration and/or lipid packing of phospholipid molecules [73]. Hence, sucrose would exhibit a colligative effect on fully hydrated bilayers by osmotically extruding water from the bilayer.

Our findings suggest possible alterations of the bilayer relative permittivity in sucrosecontaining aqueous surroundings. They are coherent with our previous results for sucrose solutions with high ionic strength [16]. From the capacitance data and the reported membrane thickness [64], the values of the relative dielectric permittivity have been evaluated to vary from ~2.3 (sugar-free, 10 mmol/L NaCl) to ~3.5 (for sucrose concentrations above ~200 mmol/L and 10 mmol/L NaCl). Following the approach here, we estimate a narrower range of *εrm* changes upon increasing sucrose content in the bulk phase, characterized by an order of magnitude of lower ionic strength. POPC bilayer relative permittivity varies from *εrm* ∼2.4 in 1 mM NaCl up to *εrm* ∼2.9 in the presence of sucrose with concentrations >200 mmol/L in the electrolyte solution.

The above evaluations for the relative dielectric permittivity are performed considering the integral thickness of the bilayer, including its hydrophobic part and headgroup regions. In order to appropriately account for the contribution of the hydrocarbon core and the headgroup regions, the bilayer has to be described by an equivalent circuit of capacitors in series replacing *C<sup>m</sup>* in Equation (5). Attempts to analyze the obtained alteration of *ε<sup>m</sup>* are based on the data available in the literature for the thickness of the headgroup region of PC ~ 9 Å [74] and the thickness of dioleoylphosphatidylcholine (DOPC) bilayers in sugar-containing aqueous solutions concentrations [64]. This approximation is realistic as the head groups of DOPC and POPC are identical and both lipids are characterized by similar hydrophobic lengths. The hydrophobic thickness of DOPC and POPC is reported to be between 27.1 and 27.2 Å [75,76] for the former, and 27.1 Å [77] for the latter. Estimations for the relative dielectric permittivity of the dipolar region of phosphatidylcholine lipid bilayers have given a wide range of values depending on the theoretical model applied [78]. The consideration of the rotating polar headgroups as an ensemble of interacting dipoles embedded in a nonhomogeneous dielectric with explicitly considering the interactions between the nearest neighborhood has derived ~30 for the headgroup dielectric constant [78], thus providing for the capacitance of the headgroup region a value of the order of 15 µF/cm<sup>2</sup> . As shown by Velikonja et al. [44] at high monovalent salt concentrations the relative permittivity in the dipolar headgroup region is decreased due to a saturation effect in orientational ordering of water dipoles. Considering that our study is performed at nearly 500 times lower salt concentrations we hypothesize that the changes in the dielectric properties are much more pronounced in the hydrophobic core than in the dipolar head region. In this case, one should keep in mind that further theoretical investigations could help quantifying the sucrose effect on the dielectric permittivity of the headgroup region. Here, we suppose that the obtained alteration of *ε<sup>m</sup>* is a result mainly of modulation in the hydrocarbon core of the bilayer. Hence, we deduce that in the presence of high sucrose concentrations (≥200 mmol/L) the relative dielectric permittivity of hydrocarbon chains is increased by ~14% and varies between ~1.4 and 1.6 in the studied

sucrose concentration range. Considering the narrower range of *εrm* changes upon increasing sucrose content in the bulk phase at an order of magnitude lower ionic strength than in [16] we formulate a hypothesis for a cooperative effect of sodium chloride and sucrose on membrane properties, including the bilayer specific capacitance and relative permittivity. It is supported by previous results indicating that the presence of sodium chloride modifies the effect of sucrose on the bending rigidity of lipid bilayers. We found that membrane bending elasticity measured in sucrose solutions containing 10 mM NaCl was independent of the disaccharide concentration in the aqueous surroundings [79]. This result has to be considered in the light of the ion-induced modification of sucrose-phosphatidylcholine hydrogen bond network, reported in the literature [80,81]. The capability of disaccharides to replace water molecules [51] and to create a water-like hydrogen bond network in the lipid surroundings contributes to retaining of the molecular properties of lipids [6]. Molecular dynamics simulation results have provided evidence that the hydrogen bond network of phosphatidylcholines and sucrose is partially disrupted in the presence of sodium and chloride ions [81].
