The miscibility of milk sphingomyelin and cholesterol is affected by temperature and surface pressure in mixed Langmuir monolayers

Food Chemistry 224 (2017) 114–123

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

The miscibility of milk sphingomyelin and cholesterol is affected by temperature and surface pressure in mixed Langmuir monolayers Ken Cheng a, Marie-Hélène Ropers b, Christelle Lopez a,⇑ a b

STLO, UMR1253, INRA, Agrocampus Ouest, 35000 Rennes, France INRA, UR1268 Biopolymères Interactions Assemblages, 44316 Nantes, France

a r t i c l e

i n f o

Article history: Received 28 July 2016 Received in revised form 12 November 2016 Accepted 12 December 2016 Available online 18 December 2016 Keywords: Milk fat globule membrane Lipid domain Melting temperature Langmuir film

a b s t r a c t The miscibility of milk sphingomyelin (milk-SM) and cholesterol was investigated in this study. The effect of different physical states of milk-SM on its interactions with cholesterol was determined by the recording of isotherms of compression of Langmuir films for temperatures above and below the gel to La phase transition of milk-SM (Tm
34 °C). For T = 15 °C < Tm, the liquid expanded (LE) to liquid condensed (LC) phase transition of milk-SM monolayers was observed at surface pressures of 10–15 mN/ m. For T = 43 °C > Tm, the milk-SM molecules were in a LE phase regardless of the surface pressure applied. A phase diagram pressure – milk-SM/cholesterol composition was established. This study demonstrated that both temperature and surface pressure affected the miscibility between the milkSM and cholesterol. The strongest attractive forces (i.e. condensing effect) were identified for 30 mol% cholesterol when the milk-SM was in the LE phase state. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction The biological membrane surrounding milk fat globules is currently poorly understood despite its importance in the functional properties of milk lipids, in the mechanisms of milk lipid digestion and in nutritional and health benefits. Increasing the knowledge surrounding the packing and miscibility of lipid components (i.e. polar lipids, cholesterol) is of primary importance in order to better understand e.g. the functions of the milk fat globule membrane (MFGM) and the properties of emulsions containing lipid droplets coated with milk polar lipids. The composition and architecture of the MFGM result from the mechanisms of fat globule secretion from the mammary epithelial cells (Keenan & Mather, 2006). The MFGM contains membranespecific proteins and a highly complex assortment of polar lipids: sphingolipids (mainly sphingomyelin, milk-SM: 20–45% of polar lipids depending on mammal species), phosphatidylcholine (PC: 19–23.7%), phosphatidylethanolamine (19.3–32.7%), phosphatidylserine (1.9–19.7%) and phosphatidylinositol (6.1–13.6%) (Lopez, 2011; Rombaut, Dewettinck, & Van Camp, 2007). Apart from these lipids, the MFGM also contains about 30 wt% of cholesterol in the membrane lipid fraction (Mesilati-Stahy & Argov-

⇑ Corresponding author at: INRA, UMR1253 STLO, 65 rue de Saint-Brieuc, 35000 Rennes, France. E-mail address: [email protected] (C. Lopez). http://dx.doi.org/10.1016/j.foodchem.2016.12.035 0308-8146/Ó 2016 Elsevier Ltd. All rights reserved.

Argaman, 2014 – i.e. about 45 mol%). The MFGM is structured as a trilayer of polar lipids and proteins (Keenan & Mather, 2006; Lopez, 2011). The inner monolayer is in contact with the triacylglycerol core of the fat globules and originates from the endoplasmic reticulum of the epithelial cells. The outer bilayer envelops the fat globules during their secretion through the apical plasma membrane of the mammary cells. Previous authors have reported an asymmetry in the localization of polar lipids with milk-SM, PC and cholesterol being preferably located in the outer bilayer of the MFGM (Deeth, 1997). Studies performed in situ in milk by confocal microscopy have revealed a phase separation of polar lipids in the outer bilayer of the MFGM with the formation of lipid domains (Gallier, Gragson, Jiménez-Flores, & Everett, 2010; Lopez, Madec, & Jiménez-Flore, 2010; Lopez & Ménard, 2011; Nguyen et al., 2016; Zou et al., 2015). These lipid domains were assumed to be formed by the lateral segregation of saturated polar lipids with a high phase transition temperature, mainly milk-SM, and could also contain cholesterol (Gallier et al., 2010; Lopez, 2011; Lopez et al., 2010; Nguyen et al., 2016). The role played by cholesterol in the formation and biophysical properties of milk-SM rich domains in the outer bilayer of the MFGM is still poorly understood (Murthy, Guyomarc’h, & Lopez, 2016a; Murthy, Guyomarc’h, Paboeuf, Vié, & Lopez, 2015). Recent studies performed by atomic force microscopy have shown that cholesterol affects the morphology of milk-SM domains (Guyomarc’h et al., 2014; Murthy et al., 2015, 2016a), and reduces the resistance to perforation of the membrane

K. Cheng et al. / Food Chemistry 224 (2017) 114–123

115

bilayers when studied as models of the MFGM (Guyomarc’h et al., 2014; Murthy et al., 2016a). Also, compression isotherms of the MFGM polar lipid monolayers revealed the condensing effect of the cholesterol, with concomitant changes in the topography of the Langmuir-Blodgett monolayers (Murthy et al., 2015). The structural and functional roles of cholesterol in the MFGM remain poorly understood (Guyomarc’h et al., 2014; Murthy et al., 2015, 2016a), despite cholesterol being known to play a fundamental role in the organization of cell membranes. Cholesterol has been reported to be involved in several mechanisms due to its close packing association with saturated lipids, inducing phase separation (McMullen, Lewis, & McElhaney, 2004; Quinn & Wolf, 2009; Wolf, Koumanov, Tenchov, & Quinn, 2001) and forming tightly packed microdomains in membranes called ‘‘rafts”, specifically with sphingolipids (Ramstedt & Slotte, 1999; Simons & Ikonen, 1997). The role played by the milk-SM in the activity of the gastric lipase at the surface of lipid droplets has been demonstrated (Favé et al., 2007) and the role of cholesterol and physical state of SM in the hydrolysis rate of SM by sphingomyelinase has been reported (Contreras, Sot, Ruiz-Argüello, Alonso, & Goñi, 2004; Jungner, Ohvo, & Slotte, 1997; Ruiz-Argüello, Veiga, Arrondo, Goni, & Alonso, 2002). The biophysical properties of the lipid domains present in the MFGM could therefore modulate the activity of lipolytic enzymes during milk fat globule digestion. Authors have recently shown that milk-SM improves lipid metabolism in high fat dietfed mice (Norris, Jiang, Ryan, Porter, & Blesso, 2016) and that the milk-SM is involved in the reduction of cholesterol absorption in the intestine (Eckhardt, Wang, Donovan, & Carey, 2002; Noh & Koo, 2004; Nyberg, Duan, & Nilsson, 2000). Further studies are necessary to better understand the interactions between the milk-SM and cholesterol that could be involved in several functions, particularly in the gastrointestinal tract. These studies need to be performed under experimental conditions adapted to the biological conditions. Most of the biophysical experiments reported in the scientific literature were performed at room temperature (i.e. 20 °C). However, temperature governs the physical state of the milk-SM (temperature of phase transition Tm
34 °C; Murthy et al., 2015) and MFGM polar lipids (Murthy, Guyomarc’h, & Lopez, 2016b) and milk lipids are digested at 37 °C in the gastrointestinal tract, i.e. at a temperature above the Tm of milk-SM. The objective of this work was to determine whether the milkSM and cholesterol were miscible and how temperature influenced their miscibility. The original approach taken in this study was to investigate the effect of the phase state of milk-SM molecules, through changes in temperature (i.e. T < Tm or T > Tm of milkSM), on the interactions between milk-SM and cholesterol. The thermotropic phase behaviour of the milk-SM was determined by differential scanning calorimetry and X-ray diffraction. The miscibility of cholesterol and the milk-SM was investigated by the isotherms of compression of Langmuir monolayers of several milkSM/cholesterol mixtures. For the first time, this work demonstrated the condensing effect of cholesterol on milk-SM monolayers with a maximum area of condensation recorded for 30 mol% cholesterol when the milk-SM was in the LE phase state.

1% C20:0, 19% C22:0, 33% C23:0, 20% C24:0, 3% C24:1 n-9 (Filippov, Orädd, & Lindblom, 2006; Guyomarc’h et al., 2014). PIPES buffer (1,4-piperazinediethanesulfonic acid) with an ionic strength and pH similar to milk was used to investigate the thermotropic phase behaviour of milk-SM bilayers and for Langmuir monolayers experiments. PIPES buffer 10 mM (1,4-piperazinediethane sulfonic acid; purity P 99%; Sigma) with NaCl 50 mM (Panreac, Germany, purity > 99%) and CaCl2 5 mM (Panreac, Germany, purity > 98%) was adjusted to pH 6.7 using NaOH 5M. Purified water (Milli-Q waters) was used for buffer preparation.

2. Materials and methods

2.4. Langmuir film balance experiments

2.1. Materials

The solutions of milk-SM and cholesterol (with concentrations of around 1 mg/ml) were prepared by dissolving the investigated components in chloroform (analytical grade, Carlo Erba reagents, Val de Reuil, France). Mixed solutions were prepared from the respective stock solutions to the required molar ratio Xcholesterol/ (1-X)milk-SM. Surface pressure (p) – area (A) isotherms were recorded with a KSV Langmuir trough (KSV 3000, KSV, Finland) with a total area = 70 cm2 (70  10 cm) placed on an anti-

Sphingomyelin from bovine milk (milk-SM; >99%) and cholesterol (chol; >99%) were purchased from Avanti Polar Lipids (Alabaster, AL, USA) and were used without further purification. Sphingomyelin has sphingosine as the hydrophobic backbone (mainly 18:1), together with an amide-linked acyl chain. The acyl chain composition of milk-SM is as follows: 19% C16:0, 3% C18:0,

2.2. Thermotropic phase behaviour of milk-SM bilayers determined by differential scanning calorimetry Milk-SM bilayers were prepared by hydrating the milk-SM powder in PIPES buffer to obtain 20 wt%. The suspension was heated at 60 °C and thoroughly mixed in a vortex stirrer in order to ensure the formation of multilamellar vesicles. Before analysis, the samples were kept at 20 °C for at least 24 h. Differential scanning calorimetry (DSC) measurements were performed with a DSC Q1000 (TA Instruments, New Castle, DE). An aliquot of the milk-SM sample was loaded into DSC hermetically sealed aluminum pans (TA Instruments). An empty, hermetically sealed, aluminum pan was used as reference. The calorimeter was calibrated with indium (DH = 28.41 J/g; Melting point = 156.66 °C). The DSC pans were introduced into the calorimeter at 20 °C, and then cooled down to 0 °C at 2 °C/min. Heating scans were run at a rate of 2 °C/min from 0 °C to 80 °C. Data analysis was performed using TA Universal Analysis program. The melting transition temperature (Tm) was taken at the peak maximum on heating. 2.3. Identification of lipid phases by synchrotron radiation X-ray diffraction X-ray scattering experiments were performed on the SWING beamline at synchrotron SOLEIL (Gif-sur-Yvette, France). On the X-ray beamline operating at 15 keV, a two-dimensional detector allowed the recording of XRD patterns in the range 0.08–1.8 Å1, thus covering both the small and wide-angle regions of interest. XRD at small angles allows identification of the lamellar organization of milk-SM molecules while XRD at wide angles provides information on the lateral packing of their sphingoid base and acyl chains. Each diffraction pattern, displaying a series of concentric rings, was integrated circularly to yield the intensity as a function of the radial scattering vector q. The scattered intensity was reported as a function of the scattering vector q = 4 p sinh/k, where 2h is the scattering angle and k the wavelength of the incident beam. The channel to scattering vector q calibration of the detector was carried out with pure tristearin (b 2L form) and silver behenate as previously reported (Murthy et al., 2015). Small volumes (around 20 ll) of samples containing fully hydrated milk-SM vesicles (20 wt%) were loaded into thin quartz capillaries of 1.5 mm diameter (GLAS W. Muller, Berlin, Germany) and inserted into the set-up. The samples were characterized at 15 °C (i.e. for T < Tm of milk-SM) and at 45 °C (i.e. for T > Tm of milk-SM).

116

K. Cheng et al. / Food Chemistry 224 (2017) 114–123

vibration table and housed in a homemade box. Surface pressure p was measured with the accuracy ±0.1 mN/m, using a Wilhelmy plate made of Whatman filter paper (Whatman Ltd, USA), connected to an electrobalance. The temperature of the subphase (PIPES buffer with NaCl and CaCl2) was controlled thermostatically to within 0.1 °C by a circulating water system fixed at 10 °C or 60 °C. The real temperature undergone by molecules spread at the air-water interface was measured at the top of the water surface. It reached 15 °C for a bath temperature set at 10 °C and 43 °C when set at 60 °C (room temperature 20 °C). The temperature of the subphase was equilibrated for 10 min before spreading of the milk-SM, cholesterol or milk-SM/cholesterol solutions. Spreading solutions were deposited onto the buffer subphase with a Hamilton micro syringe, precise to 0.5 ll. After being spread, the monolayers were left to equilibrate for 10 min before compression was initiated with a barrier speed of 30 mm/min. The experiments were stopped at surface pressure p = 50 mN/m. The results are reported as surface pressure (p) – area (A) isotherms where the area in Å2 is expressed per molecule. In the case of milk-SM/ cholesterol mixtures, the concentration and molecular weight were recalculated according to their proportions in the mixtures. The miscibility of milk-SM and cholesterol in the mixed monolayers and the magnitude of area condensation due to cholesterol addition in the mixed monolayers were determined as proposed by Gaines (1966). To analyze the miscibility of milk-SM and cholesterol in the mixed monolayers, plots of the mean area per molecule (A12) versus film composition (Xcholesterol) were considered. If the components of a monolayer are immiscible or ideally miscible, A12 is a linear function of the composition according to Eq. (1):

AIdealmix ¼ ðAchol x Xchol Þ þ ½AmilkSM xð1  Xchol Þ 12

ð1Þ

AIdeal-mix 12

Wherein is the mean area per molecule for ideal mixing; Achol, Amilk-SM are molecular areas of the respective components cholesterol and milk-SM in their pure films at a given surface pressure p; Xchol and Xmilk-SM = 1  Xchol are the mole fractions of, respectively, cholesterol and milk-SM. Eq. (1) implies that ideal molecular areas decrease linearly when the percentage of cholesterol increases. According to Gaines (1966), if milk-SM and cholesterol are immiscible or perfectly miscible, experimental molecular areas will respect Eq. (1). Any negative deviation from Eq. (1) provides evidence of miscibility of the two molecular species because of attractive forces, leading to lower mean molecular areas, and any positive deviation reveals immiscibility. To verify the magnitude of area condensation due to cholesterol addition in the mixed monolayers, the excess areas per molecule AExcess versus film composition (XChol) were calculated on the basis of Eq. (2):

AExcess ¼ A12  AIdealmix 12

ð2Þ

3. Results and discussion 3.1. Thermotropic phase behaviour of milk-SM bilayers and identification of lipid phases The combination of DSC with XRD allowed us to identify the limits of the phase transition of milk-SM bilayers and to determine the phase state, respectively. On heating of milk-SM multilamellar vesicles which were fully hydrated in PIPES buffer, the thermogram exhibited a broad asymmetric multicomponent endotherm composed of a sharp peak with a maximum at Tm = 34.2 ± 0.1 °C and with broad components on both sides of the main peak (Fig. 1). The width of the endothermic event recorded between the onset and completion temperatures spanned over 10 °C, e.g. from about 30 to 40 °C. This type of endotherm pattern of milk-

SM is in agreement with previous work (Filippov et al., 2006; Murthy et al., 2015). This complex thermotropic behaviour has been ascribed to the melting of long-chain saturated fatty acids of milk-SM species (C16:0, C22:0, C23:0, C24:0; >80% of milk-SM fatty acids), which have various melting points and different packing states in the gel phase due to molecular heterogeneity in the acyl chain length and to the gel phase polymorphism of individual milk-SM species (Sripada, Maulik, Hamilton, & Shipley, 1987). The phase state of milk-SM bilayers was identified by XRD above and below the endotherm recorded by DSC, at similar temperatures as those used for Langmuir monolayer experiments, e.g. T = 15 °C < Tm and T = 45 °C > Tm of milk-SM (Fig. 1). At 15 °C, XRD patterns recorded at small angles showed the formation of a lamellar organization with a thickness of 75.6 Å, and a single peak at wide angles (q = 1.516 Å1; d = 4.15 Å) characteristic of an hexagonal packing of the chains in the gel phase. At 45 °C, the small angle XRD pattern showed the presence of a lamellar phase (d = 79 Å), while the absence of diffraction peaks was recorded at wide angles, revealing that milk-SM bilayers were organized in a liquid crystalline La phase. 3.2. Interfacial properties of milk-SM and interactions with cholesterol in monolayers spread at the water/air interface The phase state of sphingomyelins (i.e. gel or La fluid phase), which depends on their chemical composition and temperature, was postulated as a key factor affecting the interactions with cholesterol (Smaby, Brockman, & Brown, 1994). Hence, the interfacial properties of the milk-SM and its interactions with cholesterol have been investigated in Langmuir monolayers for T = 15 °C < Tm and T = 43 °C > Tm, with Tm = 34.2 ± 0.1 °C as reported in this study for fully hydrated milk-SM bilayers. 3.2.1. Monolayer properties investigated for T = 15°C < Tm of milk-SM Milk-SM, cholesterol and milk-SM/cholesterol mixtures were spread over the buffer surface to form monolayers at the buffer/ air interface. The surface pressure (p) – area (A) isotherms recorded at 15 °C for milk-SM, cholesterol and the mixed systems with different molar ratios Xchol/(1  X)milk-SM are reported in Fig. 2-A. The surface pressure of the p - A curve recorded for the cholesterol monolayer (Xchol = 1) sharply increased at approximate area 43 Å2/molecule, showing that it is in a solid phase during the compression, in relation to the planar and rigid structure of the molecule. Our results are in good agreement with the literature: authors have reported a value of about 40–43 Å2/molecule for cholesterol at 20 °C with a water sub-phase (Hac-Wydro & Dynarowicz-Latka, 2008; Lancelot & Grauby-Heywang, 2007; Gaines, 1966). In the p - A isotherm recorded at 15 °C for milkSM monolayer, the surface pressure of the p - A curve recorded for the milk-SM monolayer (Xchol = 0) increased at approximate area 82 Å2/molecule, again in agreement with literature data (about 78–85 Å2/molecule for egg-SM and brain-SM, Smaby, Kulkami, Momsen, & Brown, 1996). The surface pressure did not increase continuously. The changes in the p - A curve slope characterized for p = 10 mN/m correspond to a two-dimensional and pressure-induced phase transition between a liquid expanded (LE) phase observed for low pressures and a liquid condensed (LC) phase observed for higher pressures. The existence of a LE to LC phase transition has previously been reported for other SM species (Hac-Wydro & Dynarowicz-Latka, 2008; Lancelot & GraubyHeywang, 2007; Smaby et al., 1994) and for synthetic saturated sphingomyelins below their temperature of phase transition, Tm (Ramstedt & Slotte, 1999). At 15 °C, the milk-SM molecules spread out in monolayers can therefore exist in different phases depending on surface pressure. The LE state is characterized by a fluid and disordered organization of the acyl chains in which only the

K. Cheng et al. / Food Chemistry 224 (2017) 114–123

117

Fig. 1. Thermotropic and structural phase behaviour of milk sphingomyelin (milk-SM). Differential scanning calorimetry (DSC) thermogram recorded on heating of 20 wt% milk-SM bilayers at 2 °C/min. Identification of the lipid phases performed at 15 °C and 45 °C by X-ray diffraction at both small (SAXS) and wide (WAXS) angles.

polar groups are constrained to be in contact with the sub-phase, whereas the LC phase is characterized by an organization of the acyl chains in an ordered crystalline sub-cell (Gaines, 1966; Kaganer, Möhwald, & Dutta, 1999). In the LC phase, the molecules are arranged near to their closest possible packing; the milk-SM molecules can hardly be further compacted which explains why the slope of the p - A curve is higher than in the LE phase. The addition of cholesterol into the milk-SM film strongly influenced the shape and the position of the p - A curves (Fig. 2-A). The isotherm of the mixed monolayers became steeper in comparison to the p A curves of the milk-SM, particularly for Xchol P 0.5. In addition, it shifted position towards a lower mean molecular area and started to resemble the isotherm recorded for cholesterol (Xchol = 1). The miscibility of milk-SM and cholesterol in the mixed monolayers was investigated by considering the plots of the mean area per molecule (A12) versus cholesterol mole fraction in the films (Xchol) at six different surface pressures (p = 5, 10, 15, 20, 25 and 30 mN/m). Fig. 2-B shows A12 values as a function of film composition (solid lines) and the plot of ideal mixing (AIdeal-mix dashed ;12 lines). For most monolayer compositions that have been investigated at 15 °C, there are either positive or negative deviations from the ideal behaviour. The deviations depend on the surface pressure p (p < 10 mN/m or p P 10 mN/m) and on the film composition (Xchol) (Fig. 2-B). For p = 5 mN/m and p = 10 mN/m, i.e. surface pressures for which milk-SM is in the LE state, the deviations from ideal behaviour are positive at Xchol = 0.05 and negative in a large range of monolayer composition (0.1 6 XChol < 1). In this case, mixed monolayers are more condensed than ideal ones. This suggests that the interactions between cholesterol and milk-SM in their mixed films are more attractive than the milk-SM/milk-SM and cholesterol/cholesterol intermolecular forces in the respective singlecomponent films. As cholesterol’s cross-sectional surface area is very sensitive to changes in surface pressure, the experimentally observed changes in the areas of mixed films can be attributed almost exclusively to configurational changes of milk-SM, i.e.

ordering that cholesterol imparts to the milk-SM molecules. For the surface pressures p = 5 mN/m and p = 10 mN/m, cholesterol induced an interfacial area condensation of the milk-SM molecules, revealing its ‘‘condensing effect”. This condensing effect of cholesterol has already been reported for other sources of SM and polar lipids (Smaby et al., 1994; Hac-Wydro & Dynarowicz-Latka, 2008; Quinn & Wolf, 2009; Wydro, 2011, 2013). This is a consequence of a beneficial separation of milk-SM molecules by cholesterol that was attributed to a decrease of the electrostatic forces existing between polar groups of milk-SM molecules (Hac-Wydro & Dynarowicz-Latka, 2008; Smaby et al., 1994). For surface pressures p > 10 mN/m, for which milk-SM is in the LC state, the negative and positive deviations were highly dependent on the film composition (Xchol). The positive deviations from ideal behaviour were found at low cholesterol fractions (Xchol = 0.05) and high cholesterol fractions (from Xchol P 0.5 at p P 20 mN/m), revealing that milk-SM and cholesterol were not miscible. In these p - composition areas, the interactions between milk-SM and cholesterol were found to be more repulsive than the milk-SM/milk-SM and cholesterol/cholesterol intermolecular forces in the respective single-component films. For surface pressures p > 10 mN/m, the negative deviations were found at intermediate mole fractions between 0.1 6 Xchol < 0.5 showing that cholesterol induced an interfacial area condensation of the molecules. The magnitude of the area of condensation due to cholesterol addition was considered by plotting the excess area per molecule (AExcess) versus film composition (Xchol) (Fig. 2-C). For Xchol = 0.05, AExcess values were positive whatever the pressure applied, meaning that milk-SM and cholesterol did not form homogeneous mixtures but separated as a consequence of repulsions. For Xchol > 0.05, AExcess values were negative for low surface pressures and became positive as the surface pressure increased. Fig. 2-C shows that the strongest negative deviation from ideal behaviour appears at p = 5 mN/m and p = 10 mN/m for monolayers with Xchol = 0.3, meaning that the cholesterol/milk-SM attractive interactions were

118

K. Cheng et al. / Food Chemistry 224 (2017) 114–123

Fig. 2. Interfacial properties of milk-sphingomyelin (milk-SM) and interactions with cholesterol investigated in mixed Langmuir monolayers at 15 °C. (A) the surface pressure – area (p-A) isotherms of mixed monolayers formed at the air / buffer interface by milk-SM and cholesterol, (B) The mean area per molecule (A12) versus composition (Xcholesterol) plots for mixed monolayers of milk-SM and cholesterol at different constant surface pressures, and (C) The excess area per molecule (AExcess) versus composition (Xcholesterol) for mixed monolayers of milk-SM and cholesterol at different constant surface pressures.

the strongest for monolayers containing 30 mol% of cholesterol. At surface pressures relevant to biological membranes (25 mN/m) and for Xchol = 0.30, the AExcess value was null. The interactions between cholesterol and milk-SM were not favoured in comparison to cholesterol/cholesterol or milk-SM/milk-SM. Cholesterol and milk-SM were not fully miscible in these conditions. Interestingly, our results agree with previous studies dealing with two other natural SM molecules from egg and brain (HacWydro & Dynarowicz-Latka, 2008; Smaby et al., 1994). The largest negative deviation observed for 20–30 mol% cholesterol is the same as that found in egg- or brain-SM/cholesterol monolayers (Hac-Wydro & Dynarowicz-Latka, 2008; Smaby et al., 1994), in multicomponent monolayers imitating outer and inner leaflet of

human erythrocyte membrane (Wydro, 2013) and in monolayers of a polar lipid extract from MFGM (Murthy et al., 2015). However, discrepancies also exist. For example, cholesterol induces egg-SM condensation at 20 °C for surface pressures between 5 and 35 mN/m, whatever the Xchol/(1-X)egg-SM molar ratio in the monolayer (Hac-Wydro & Dynarowicz-Latka, 2008; Smaby et al., 1994), i.e. within a wider p-composition range than that for milk-SM as found in the present study. The reductions in average area per molecule induced by cholesterol were lower in the cholesterol/ milk-SM system than the values reported by Smaby et al. (1994) for egg-SM and bovine brain SM. These discrepancies could be related to the fatty acid composition of SM (chain length and unsaturation), i.e., the origin of SM (Ramstedt & Slotte, 1999;

K. Cheng et al. / Food Chemistry 224 (2017) 114–123

Smaby et al., 1994). The short-distance van der Waals interactions between the chains of SM increase with the chain length. It is assumed that they are more favoured than SM/cholesterol interactions in systems where SM molecules have long chain length (as it is the case for milk-SM) than in systems where SM molecules have shorter chain length (egg-SM and brain-SM) or carry unsaturation (brain-SM). 3.2.2. Monolayer properties investigated for T = 43°C > Tm of milk-SM The p – A isotherms recorded at 43 °C for milk-SM, cholesterol and the investigated mixed systems are presented in Fig. 3-A. At 43 °C, the temperature is over the melting temperature of milkSM (Fig. 1) and the acyl chains are expected to be in a fluid state. The surface pressure of the p - A curve recorded for the milk-SM

119

monolayer (Xchol = 0) increased at approximate area 104 Å2/molecule and is characterized by a continued and moderate variation of the slope until the collapse of the monolayer. This is in accordance with the formation of a LE phase with acyl chains in a fluid phase. The surface pressure of the p - A curve recorded for the cholesterol monolayer (Xchol = 1) at 43 °C increased at approximate area 44 Å2/molecule with a significant slope, showing that cholesterol was still in the LC phase in accordance with its melting temperature (Gaines, 1966). Hence, for T = 43 °C, milkSM and cholesterol are in two different phases whatever the surface pressure applied, i.e. fluid LE phase of milk-SM acyl chains and crystallized LC state for cholesterol. Addition of cholesterol into the milk-SM monolayer strongly influenced the position of the p - A curves and the slope of the iso-

Fig. 3. Interfacial properties of milk-sphingomyelin (milk-SM) and interactions with cholesterol investigated in mixed Langmuir monolayers at 43 °C. (A) the surface pressure – area (p-A) isotherms of mixed monolayers formed at the air / buffer interface by milk-SM and cholesterol, (B) The mean area per molecule (A12) versus composition (XCholesterol) plots for mixed monolayers of milk-SM and cholesterol at different constant surface pressures, and (C) The excess area per molecule (AExcess) versus composition (XCholesterol) for mixed monolayers of milk-SM and cholesterol at different constant surface pressures.

120

K. Cheng et al. / Food Chemistry 224 (2017) 114–123

therm curves (Fig. 3-A). The miscibility of milk-SM and cholesterol in the mixed monolayers was analyzed as previously described by considering the plot of the mean area per molecule (A12) versus cholesterol mole fraction (Xchol) (Fig. 3-B). The A12 values were determined from the p - A curves at seven different surface pressures (p = 5, 10, 15, 20, 25, 30 and 35 mN/m). The magnitude of area change due to cholesterol addition was considered by plotting the excess area per molecule (AExcess) versus film composition (Xchol) (Fig. 3-C). At 43 °C, the slight positive deviations observed between A12 and the ideal behaviour AIdeal-mix correspond to two 12 limited areas: (i) Xchol = 0.15, low pressure surface (p = 5 and 10 mN/m) and (ii) Xchol > 0.7, high surface pressures (p = 35 mN/m). Such positive deviations correspond to immiscibility between the different components with repulsive interactions between milkSM and cholesterol. Outside these two areas, the milk-SM and cholesterol mixtures showed a good miscibility with attractive forces. The highest negative AExcess values (6 Å2/molecule) were found for Xchol = 0.3 whatever the surface pressure applied. They correspond to an interfacial area of condensation of the milk-SM molecules induced by cholesterol for T = 43 °C > Tm of milk-SM.

3.2.3. The temperature affects the interfacial properties of milk-SM and the interactions between milk-SM and cholesterol Most of the studies investigating the interfacial properties of phospholipids or sphingolipids in Langmuir monolayers are performed at room temperature, i.e. below the melting phase transition temperature Tm of the saturated lipids (Lancelot & GraubyHeywang, 2007). However, temperature governs many mechanisms, e.g. the activity of enzymes is conditioned by the temperature-dependent physical state of lipids. Few authors investigated the role of temperature and then the role played by the phase state of sphingomyelins (Li, Smaby, Momsen, Brockman, & Brown, 2000). The originality of our work was to determine the Tm for milk-SM bilayers and to investigate the interfacial properties of milk-SM and its interaction with cholesterol for T < Tm and T > Tm, e.g. at temperatures for which milk-SM is in two different phases. The p – A isotherms recorded for milk-SM monolayers at temperatures above and below the melting temperature, showed that both the temperature and the surface pressure affect the phase in which milk-SM is present in monolayers. First, the area per milkSM molecule was higher for T > Tm than for T < Tm at a given surface pressure (Fig. 4) due to the melted state of acyl chains and differences in the hydration levels of the polar group with temperature. For example, the molecular area of milk-SM measured for p = 25 mN/m, a surface pressure relevant to biological membranes, was 47.5 Å2/molecule at 15 °C and 69.5 Å2/molecule at 43 °C, i.e. an expansion of 46% from 15 to 43 °C. Second, the LE to LC two-dimensional phase transition, induced by compression of the monolayer at T < Tm, is observed in milk-SM monolayers at low surface pressures (p = 10 mN/m at 15 °C and p = 15 mN/m at 20 °C) and below the melting temperature. The LE to LC phase transition was not observed at 43 °C (i.e. for T > Tm), where the milk-SM molecules were in a LE phase whatever the surface pressure applied (Fig. 4). The existence of LE to LC phase transition has also previously been reported for monolayers of natural SM, e.g. for egg-SM monolayers at T = 20–24 °C with p = 18 mN/m (Hac-Wydro & Dynarowicz-Latka, 2008; Li, Momsen, Smaby, Brockman, & Brown, 2001; Smaby et al., 1994, 1996), for bovine brain SM at T = 24 °C with p = 12 mN/m (Li et al., 2001; Smaby et al., 1996) and semi-synthetic saturated SM with acyl chains (C16:0, C18:0 and C24:0) between 10 and 30 °C (Ramstedt & Slotte, 1999; Smaby et al., 1996; Li et al., 2000). Heterogeneities in the acyl chains composition of natural SM molecules, as it is the case for milk-SM, leads to a broad two-dimensional LE to LC

transition as compared to synthetic SM (e.g. C16:0-SM, C18:0SM) (Smaby et al., 1996). At surface pressures relevant to biological membranes, our experiments revealed that temperature affected the phase in which milk-SM was present. For T = 15 °C < Tm, milk-SM monolayers displayed a LC phase, whereas at T = 43 °C > Tm, milk-SM monolayers displayed a LE phase. The physical state in which milk-SM is present at surface pressures relevant to biological membranes such as the MFGM or at the surface of emulsion droplets could have technofunctional consequences (e.g. physical stability of the emulsion) and biological implications (e.g. hydrolysis by enzymes) that remain to be elucidated. Temperature affects the interactions of milk-SM with cholesterol as well. Fig. 5 summarizes the effect of the surface pressure p and composition (Xchol) of the mixed layers on the interactions between milk-SM and cholesterol determined for T = 15 °C < Tm and T = 43 °C > Tm. For T = 15 °C < Tm, the films of milk-SM are condensable by cholesterol through attractive interactions especially where milk-SM was in a LE phase, e.g. for p < 10 mN/m or over a large pressure range if cholesterol is in a reasonable amount (0.1 < Xchol < 0.5). Surprisingly, there is a small area (0 < Xchol < 0.1) where the tiny amount of cholesterol is detrimental for the homogeneity of the monolayer. Inversely, an area of poor mixing is obtained at high surface pressures for low cholesterol fraction and increases to lower surface pressures as the amount of cholesterol increases in the mixture. For T = 43 °C > Tm, milk-SM and cholesterol are miscible across a wider range of surface pressure and Xchol. The liquid-like arrangement of the flexible hydrophobic chains of milk-SM is thus important for the formation of homogeneous domains with cholesterol. The magnitude of the condensing effect, determined by Aexcess values, is the highest for Xchol = 0.3 in the LE phase state of milkSM, irrespective of the surface pressure at 43 °C and for p 6 15 mN/m at 15 °C (Fig. 5). The fact that this ratio is independent of temperature and surface pressure in the LE phase state of milk-SM may be a sign of the minimization of geometrical constraints. At the pressure occurring in biological membranes, about 25– 30 mN/m (Blume, 1979; Marsh, 1996; Seelig, 1987), a phase separation between cholesterol and milk-SM is expected to occur at T = 15 °C < Tm within a wide Xchol range (Xchol < 0.1 and Xchol > 0.2) and a large area of miscibility between cholesterol and milk-SM is expected at T = 43 °C > Tm. These results explain the formation of milk-SM rich domains in the MFGM characterized at room temperature where milk-SM chains are in a condensed state and unable to adapt to the condensed state of cholesterol molecules (Lopez et al., 2010; Lopez, 2011). The miscibility of milk-SM and cholesterol for T = 43 °C > Tm measured in this study is in agreement with the presence of liquid ordered (Lo) phase domains revealed in the MFGM above the melting temperature of milk-SM (Et-Thakafy, Guyomarc’h, & Lopez, 2017; Nguyen et al., 2016). Our results are also interesting with regard to the potential role played by the milk-SM rich domains in the mechanisms of milk fat globule digestion in the gastrointestinal tract. At the physiological temperature in humans, usually 37 °C which is higher than the melting transition temperature Tm of milk-SM, we can expect a miscibility of a large molar ratio of cholesterol in milk-SM, particularly for Xchol = 0.3. The fact that the inclusion of cholesterol into a SM monolayer increased the extent of SM hydrolysis by sphingomyelinase (Jungner et al., 1997) led to the conclusion that the physical state of milk-SM is of importance for the reactivity of sphingomyelinase during digestion. Furthermore, these results could at least partially explain the role of milk-SM in the reduced intestinal absorption of cholesterol upon digestion (Noh & Koo, 2004; Norris et al., 2016). However, these results need to be confirmed as the human physiological temperature is slightly lower than the tem-

K. Cheng et al. / Food Chemistry 224 (2017) 114–123

121

Fig. 4. Interfacial properties of milk-sphingomyelin (milk-SM) investigated in Langmuir monolayers at various temperatures below and above its phase transition temperature Tm
34 °C. (A) the surface pressure – area (p-A) isotherms of monolayers formed at the air/buffer interface by milk-SM, and (B) Differential scanning calorimetry thermogram recorded on heating of milk-SM bilayers to determine Tm and area per molecule calculated at different surface pressures as indicated in the figure. LC: liquid condensed phase, LE: liquid expanded phase, ⁄: LE to LC phase transition induced by surface pressure.

Fig. 5. Effect of the surface pressure and composition (cholesterol mole fraction; Xcholesterol) of the mixed layers of milk-SM and cholesterol on the interactions between milkSM and cholesterol determined (A) at 15 °C and (B) at 43 °C. The diameter of the bubbles is proportional to the excess area per molecule (AExcess) value. Black colour = negative values of AExcess corresponding to attractive interactions with an interfacial area condensation; grey colour = positive values of AExcess corresponding to repulsive forces between milk-SM and cholesterol; orange colour = AExcess values delimiting the condensing effect of cholesterol from the demixtion in the monolayer. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

122

K. Cheng et al. / Food Chemistry 224 (2017) 114–123

perature of completion of milk-SM melting and some condensed zones of milk-SM could remain and lead to a local phase separation. 4. Conclusions Characterizing the interactions between milk-SM and cholesterol above and below the melting temperature of milk-SM was important in highlighting the role of the physical state of these two biologically active molecules. Cholesterol molecules were found to be incorporated into milk-SM monolayers providing that the milk-SM molecules were in a fluid state (T > Tm = 34 °C). Cholesterol has a condensing effect for milk-SM molecules under these conditions, especially at Xchol = 0.3 and at surface pressures relevant for biological membranes. The physical state of the milk-SM is the key factor for controlling the interactions between milk-SM and cholesterol and the lipid domains in the MFGM; it is thus of importance to investigate the biological phenomena under appropriate temperature conditions. This study provides a foundation for future investigations aiming to evaluate e.g. the hydrolysis of milk-SM by sphingomyelinase, the mechanisms involved in the digestion of milk fat globules, or the milk-SM/cholesterol complexes formed in the gastrointestinal tract upon digestion of milk lipids. Acknowledgements This work was exclusively funded by the INRA (France). The authors thank the SOLEIL synchrotron for allocating beamtime on the SWING beamline (proposal 20090250; C. Lopez), as well as all the members of the Scientific Users Committee of INRA. Javier Pérez (Swing beamline scientist, SOLEIL synchrotron, France), Pierre Roblin (INRA, SOLEIL synchrotron, France) and Claudie Bourgaux (CNRS, France) are acknowledged for the temperature controlled XRD set-up. Christelle Lopez warmly acknowledges John Hannon for revising the English language and for useful comments on the manuscript. References Blume, A. (1979). A comparative study of the phase transitions of phospholipid bilayers and monolayers. Biochimica et Biophysica Acta (BBA) – Biomembranes, 557(1), 32–44. Contreras, F.-X., Sot, J., Ruiz-Argüello, M.-B., Alonso, A., & Goñi, F. M. (2004). Cholesterol modulation of sphingomyelinase activity at physiological temperatures. Chemistry and Physics of Lipids, 130, 127–134. Deeth, H. C. (1997). The role of phospholipids in the stability of milk fat globules. Australian Journal of Dairy Technology, 52, 44–46. Eckhardt, E. R. M., Wang, D. Q. H., Donovan, J. M., & Carey, M. C. (2002). Dietary sphingomyelin suppresses intestinal cholesterol absorption by decreasing thermodynamic activity of cholesterol monomers. Gastroenterology, 122, 948–956. Et-Thakafy, O., Guyomarc’h, F., & Lopez, C. (2017). Lipid domains in the milk fat globule membrane: Dynamics investigated in situ in milk in relation to temperature and time. Food Chemistry, 220, 352–361. Favé, G., Lévêque, C., Peyrot, J., Pieroni, G., Coste, T. C., & Armand, M. (2007). Modulation of gastric lipolysis by the phospholipid specie: Link to specific lipase-phospholipid interaction at the lipid/water interface? The FASEB Journal, 21(800), 1. Filippov, A., Orädd, G., & Lindblom, G. (2006). Sphingomyelin structure influences the lateral diffusion and raft formation in lipid bilayers. Biophysical Journal, 90, 2086–2092. Gaines, G. L. (1966). In Insoluble monolayers at liquid-gas interfaces, chapter 6: Mixed monolayers. New York: Wiley-Interscience. Gallier, S., Gragson, D., Jiménez-Flores, R., & Everett, D. W. (2010). Surface characterization of bovine milk phospholipid monolayers by Langmuir isotherms and microscopic techniques. Journal of Agricultural & Food Chemistry, 58, 12275–12285. Guyomarc’h, F., Zou, S., Chen, M., Milhiet, P.-E., Godefroy, C., Vié, V., & Lopez, C. (2014). Milk sphingomyelin domains in biomimetic membranes and the role of cholesterol: Morphology and nanomechanical properties investigated using AFM and force spectroscopy. Langmuir, 30, 6516–6524.

Hac-Wydro, K., & Dynarowicz-Latka, P. (2008). The impact of sterol structure on the interactions with sphingomyelin in mixed Langmuir monolayers. The Journal of Physical Chemistry B, 112, 11324–11332. Jungner, M., Ohvo, H., & Slotte, J. P. (1997). Interfacial regulation of bacterial sphingomyelinase activity. Biochimica Biophysica Acta, 1344, 230–240. Kaganer, V. M., Möhwald, H., & Dutta, P. (1999). Structure and phase transitions in Langmuir monolayers. Reviews of Modern Physics, 71, 779–819. Keenan, T. W., & Mather, I. H. (2006). Intracellular origin of milk fat globules and the nature of the milk fat globule membrane. In P. F. Fox & P. L. H. McSweeney (Eds.), Lipids. Advanced dairy chemistry (3rd ed., pp. 137–171). New-York, NY, USA: Springer. Lancelot, E., & Grauby-Heywang, C. (2007). Comparison of the interaction of dihydrocholesterol and cholesterol with sphingolipid or phospholipid Langmuir monolayers. Colloids and Surfaces B: Biointerfaces, 59, 81–86. Li, X. M., Momsen, M. M., Smaby, J. M., Brockman, H. L., & Brown, R. E. (2001). Cholesterol decreases the interfacial elasticity and detergent solubility of sphingomyelins. Biochemistry, 40(5954), 5963. Li, X. M., Smaby, J. M., Momsen, M. M., Brockman, H. L., & Brown, R. E. (2000). Sphingomyelin interfacial behavior: The impact of changing acyl chain composition. Biophysical Journal, 78, 1921–1931. Lopez, C. (2011). Milk fat globules enveloped by their biological membrane: Unique colloidal assemblies with a specific composition and structure. Current Opinion in Colloid & Interface Science, 16, 391–404. Lopez, C., Madec, M.-N., & Jiménez-Flore, s. R. (2010). Lipid rafts in the bovine milk fat globule membrane revealed by the lateral segregation of phospholipids and heterogeneous distribution of glycoproteins. Food Chemistry, 120, 22–33. Lopez, C., & Ménard, O. (2011). Human milk fat globules: Polar lipid composition and in situ structural investigations revealing the heterogeneous distribution of proteins and the lateral segregation of sphingomyelin in the biological membrane. Colloids & Surfaces B: Biointerfaces, 83, 29–41. Marsh, D. (1996). Lateral pressure in membranes. Biochimica et Biophysica Acta (BBA) – Reviews on Biomembranes, 1286(3), 183–223. McMullen, T. P. W., Lewis, R. N. A. H., & McElhaney, R. N. (2004). Cholesterol– phospholipid interactions, the liquid-ordered phase and lipid rafts in model and biological membranes. Current Opinion in Colloid & Interface Science, 8, 459–468. Mesilati-Stahy, R., & Argov-Argaman, N. (2014). The relationship between size and lipid composition of the bovine milk fat globule is modulated by lactation stage. Food Chemistry, 145, 562–570. Murthy, A. V. R., Guyomarc’h, F., & Lopez, C. (2016a). Cholesterol decreases the size and the mechanical resistance to rupture of sphingomyelin rich domains, in lipid bilayers studied as a model of the milk fat globule membrane. Langmuir, 32, 6757–6765. Murthy, A. V. R., Guyomarc’h, F., & Lopez, C. (2016b). The temperature-dependent physical state of polar lipids and their miscibility impact the topography and mechanical properties of bilayer models of the milk fat globule membrane. Biochimica et Biophysica Acta (BBA) – Biomembranes, 1858, 2181–2190. Murthy, A. V. R., Guyomarc’h, F., Paboeuf, G., Vié, V., & Lopez, C. (2015). Cholesterol strongly affects the organization of lipid monolayers studied as models of the milk fat globule membrane: Condensing effect and change in the lipid domain morphology. BBA: Biomembranes, 1848, 2308–2316. Nguyen, H. T. H., Madec, M.-N., Ong, L., Kentish, S. E., Gras, S. L., & Lopez, C. (2016). The dynamics of the biological membrane surrounding the buffalo milk fat globule investigated as a function of temperature. Food Chemistry, 204, 343–351. Noh, S. K., & Koo, S. I. (2004). Milk sphingomyelin is more effective than egg sphingomyelin in inhibiting intestinal absorption of cholesterol and fat in rats. Journal of Nutrition, 134, 2611–2616. Norris, G. H., Jiang, C., Ryan, J., Porter, C. M., & Blesso, C. N. (2016). Milk sphingomyelin improves lipid metabolism and alters gut microbiota in high fat diet-fed mice. Journal of Nutritional Biochemistry, 30, 93–101. Nyberg, L., Duan, R. D., & Nilsson, A. (2000). A mutual inhibitory effect on absorption of sphingomyelin and cholesterol. Journal of Nutritional Biochemistry, 11, 244–249. Quinn, P. J., & Wolf, C. (2009). The liquid-ordered phase in membranes. Biochimica et Biophysica Acta, 1788, 33–46. Ramstedt, B., & Slotte, J. P. (1999). Interaction of cholesterol with sphingomyelins and acyl-chain-matched phosphatidylcholines: A comparative study of the effect of the chain length. Biophysical Journal, 76, 908–915. Rombaut, R., Dewettinck, K., & Van Camp, J. (2007). Phospho and sphingolipid content of selected dairy products as determined by HPLC coupled to an evaporative light scattering detector (HPLC-ELSD). Journal of Food Composition and Analysis, 20, 308–312. Ruiz-Argüello, M. B., Veiga, M. P., Arrondo, J. L. R., Goni, F. M., & Alonso, A. (2002). Sphingomyelinase cleavage of sphingomyelin in pure and mixed lipid membranes. Influence of the physical state of the sphingolipid. Chemistry and Physics of Lipids, 114, 11–20. Seelig, A. (1987). Local anesthetics and pressure: A comparison of dibucaine binding to lipid monolayers and bilayers. Biochimica et Biophysica Acta (BBA) – Biomembranes, 899(2), 196–204. Simons, K., & Ikonen, E. (1997). Functional rafts in cell membranes. Nature, 387, 569–572. Smaby, J. M., Brockman, H. L., & Brown, R. E. (1994). Cholesterol’s interfacial interactions with sphingomyelins and phosphatidylcholines: Hydrocarbon chain structure determines the magnitude of condensation. Biochemistry, 33, 9135–9142.

K. Cheng et al. / Food Chemistry 224 (2017) 114–123 Smaby, J. M., Kulkami, V. S., Momsen, M., & Brown, R. E. (1996). The interfacial elastic packing interactions of galactosylceramides, sphingomyelins, and phosphatidylcholines. Biophysical Journal, 70, 868–877. Sripada, P. K., Maulik, P. R., Hamilton, J. A., & Shipley, G. G. (1987). Partial synthesis and properties of a series of N-acyl sphingomyelins. Journal of Lipid Research, 28, 710–718. Wolf, C., Koumanov, K., Tenchov, B., & Quinn, P. J. (2001). Cholesterol favors phase separation of sphingomyelin. Biophysical Chemistry, 89, 163–172.

123

Wydro, P. (2011). The magnitude of condensation induced by cholesterol on the mixtures of sphingomyelin with phosphatidylcholines – study on ternary and quaternary systems. Colloids and Surfaces B: Biointerfaces, 82, 594–601. Wydro, P. (2013). The influence of cholesterol on multicomponent Langmuir monolayers imitating outer and inner leaflet of human erythrocyte membrane. Colloids and Surfaces B: Biointerfaces: Biointerfaces, 103, 67–74. Zou, X., Guo, Z., Jin, Q., Huang, J., Cheong, L., Xu, X., & Wang, X. (2015). Composition and microstructure of colostrum and mature bovine milk fat globule membrane. Food Chemistry, 185, 362–370.