The monolayers are formed on the water surface when amphiphilic substances (surfactant) are two-dimensionally arranged to a monomolecular layer at the air-water interface. Amphiphilic molecules’ hydrophobic alkyl chains orient toward the air phase and hydrophilic moieties toward the water subphase. When amphiphilic substance is a lipid, the monolayer corresponds to a half part of the biomembrane. Therefore the lipid monolayer is selected as a biomembrane model to investigate cellular functions such as membrane structure, interaction with functional substances, molecular recognition, and transmission of functional materials in and out of the living biological cell membrane    .
Cholesterol (Figure 1(a)) is classified under the steroid group. Structurally, it is composed of a steroid skeleton and a hydrophilic hydroxyl group (-OH). Cholesterol is one of the main components in biomembrane and the important organic compound from the viewpoint of various biological phenomena such as controlling of biomembrane conformation, functioning of integral biomembrane proteins, and metabolism in the living body. In these phenomena, “condensation effect” of cholesterol is widely known. The addition of cholesterol in biomembrane alters the order of biomembrane structure and affects the fluidity of the biomembrane. This also promotes thermal and structural stabilities of biomembrane in a wide temperature range. Many studies on the condensation effect have been performed using phospholipid-cholesterol mixtures in monolayer and bilayer systems, and liposomes   - . The techniques of thermal analysis, ATR-FTIR and NMR spectroscopy, and molecular simulation       showed that the structure of DMPC molecule exchanged from liquid-disordered (Ld) phase to liquid-ordered (Lo) phase with an increase in cholesterol mole fraction.
In this communication, we have reported the cholesterol-mixed effect on the morphology of expandable dimyristoyl phosphatidylcholine (DMPC) (Figure 1(b)) monolayer formed on the water surface for various cholesterol mole fractions by dropping method. To this end, physicochemical methods: surface tension measurement (STm), Brewster angle microscopy (BAM), and infrared external reflection spectroscopy (IERS) have been applied. The results have been interpreted to work out an addition effect of cholesterol in the structural change of DMPC molecules.
2. Materials and Methods
Dimyristoyl phosphatidylcholine (DMPC) and cholesterol (both purities: >99%) were purchased from Avanti polar lipids Inc. (Alabama, USA) and Sigma-Al- drich Corp. (St. Louis, USA), respectively, and were used without further purification. Chloroform (purity: 99%) were from Wako Pure Chem. Ind. Ltd. (Osaka, Japan). Purified water with conductance < 0.07 μS/cm was obtained using a Super Water Purifying System (WL-21P; Yamato Scientific Corp. Ltd., Tokyo, Japan).
Figure 1. Structures of cholesterol (a) and DMPC (b).
2.2. Monolayer Formation
DMPC and cholesterol were dissolved in chloroform to prepare dropping solution. The solution was dropped on a purified water surface with a 100 μl microsyringe (Ge-0583-04; Hamilton Corp., Nevada, USA) to form monolayer. The details of the method of monolayer preparation have been reported in a previously publication . 1 μl drop of the solution was dropped gently on the water surface and the next drop was added after ≥1 min so that previous droplet expanded sufficiently on the water surface. The arrival of equilibrium by proper expanding of DMPC and cholesterol on the surface was confirmed by the constant surface tension value attainment after the addition of each droplet (Section 2.3.). The process of the solution addition was repeated until completion of monolayer formation was done. The completion of the monolayer was confirmed by observing the formation of lenses on the water surface and the value of surface tension (Section 2.3.) did not change any more.
2.3. Surface Tension Measurement (STm)
Surface tension measurement (STm) for each mixed monolayer was carried out on a Surface Tensiometer (CBVP-A3; Kyowa Interface Science Corp. Ltd., Saitama, Japan) with a platinum Wilhelmy plate. Surface tension was recorded at each dropping at 26˚C ± 1˚C. The surface tension values were then converted into surface pressure (π). The π-A isotherm curves for each monolayer were constructed as a function of the dropping volume of the monolayer forming components in terms of molecule numbers.
2.4. Brewster Angle Microscopy (BAM)
Brewster angle microscopy (BAM) was employed for visualizing the morphologies of each DMPC-cholesterol mixed monolayer using a commercial BAM microscope (EMM633K; Filgen Inc., Nagoya, Japan) attached with an USB-CAP type (SD-USB2CUP3; AREA Co. Ltd., Tokyo, Japan) image analysis software. BAM records the monolayer images arise from the refractive index difference between the monolayer and the water phase. BAM device was mounted on a glass dish and p-polarized light of 632.8 nm wavelength from a 10 mW He-Ne laser was irradiated at the Brewster angle of 53.1˚ on each mixed monolayer surface. The reflected light was magnified with a 40 mm focal length lens and detected by a CCD camera (C5948-70; Hamamatsu Photonics, Hamamatsu). The formation process of each mixed monolayer by dropping method was observed on real-time at 26˚C ± 1˚C. The lateral resolution of BAM images was about 1 μm.
2.5. Infrared External Reflection Spectroscopy (IERS)
The structure of DMPC molecule alkyl chains in DMPC monolayer and DMPC- cholesterol mixed monolayer was investigated by infrared external reflection spectroscopy (IERS). The details of the method of IERS measurement have been reported in a previously publication . An FT-IR spectrometer (FT-IR-600Plus; JASCO Int. Co. Ltd.) with a liquid nitrogen-cooled MCT detector and a reflectance apparatus was employed in combination to make an IERS setup. For every measurement, the solution of DMPC or DMPC-cholesterol (cholesterol mole fraction (xChol) 0.4) and chloroform was expanded on the water surface in a poly (tetrafluoroethrene) trough of the size (20 × 62 mm2) at 26˚C˚ ± 1˚C, according to the dropping method as described above in Section 2.2. The measurements of the above prepared monolayers were conducted at π = 35 mN/m. The p-polarized IR light at incident angles of 8˚ - 20˚ and 75˚ was irradiated on each monolayer and reflected light was detected by MCT (Mercury-Cadmium-Telluride) detector to investigate the molecular structure and the orientation. More than 1000 scans were averaged with a resolution of 4 cm−1 for the measurements. The IER spectra obtained were converted to the absorption spectra (k-spectra) by the above reported method  using Lorentz oscillator model (LOM) calculation under the assumptions, first the n∞ values (the real part of complex refractive index) of the limiting of the high-wavenumber are 1.50 and second the thickness of DMPC or DMPC-Chol monolayer is 2 nm.
3. Results and Discussion
3.1. Surface Tension Measurement (STm)
Figure 2 shows π (surface pressure) versus molecular area (A), (π-A) isotherm curves for various DMPC-cholesterol mixed monolayers. The π-A isotherms were constructed for a number of cholesterol mole fractions (xChol): 0.0, 0.2, 0.3, 0.4, 0.5, 0.6, 0.8, 1.0 on the water surface using dropping method. The horizontal axis represents A/nm2 calculated from molecular numbers in the dropping volume and the vertical axis represents π/mN × m−1 recorded after dropping the sample solution on the water surface. The value of π was noted finally when π reached at a constant value.
Figure 2. π-A isotherm curves of various DMPC-cholesterol mixed monolayers (cholesterol mole fraction xChol = 0.0, 0.2, 0.3, 0.4, 0.5, 0.6, 0.8, 1.0) on the water surface at 26˚C, using dropping method.
The DMPC monolayer π-A isotherm curve (○) without cholesterol is the same as reported previously  . The π value increased gradually at about 1.2 nm2 of A and showed a monotonous increase up to 43 mN/m. The shape of the curve was smooth and similar to that obtained by compression method   . The limiting molecular A0 was 0.74 ± 0.02 nm2/molecule corresponded to 40 mN/m of π and the value was very similar (0.73 nm2/molecule) obtained at 40 mN/m in compression method. From the pattern of π-A isotherm curve, it was confirmed that DMPC monolayer by dropping method formed liquid expanded (LE) state where two DMPC molecule’s alkyl chains were disordered partially. The structure of alkyl chains contained gauche conformation, thereby the cross- sectional area would be larger than of the hydrophilic group.
As shown in Figure 2, with an increase of xChol from □ to ■, the curves of DMPC-cholesterol mixed monolayer shifted to the left and approached to a plot pattern of cholesterol monolayer (●). At > xChol = 0.4 (▼), the curves showed steep rise and approached to that of similar to cholesterol monolayer (●), though each A value where each curve started to rise was different. Each A0 for >xChol = 0.4 showed values close to that of cholesterol monolayer, confirmed that at >xChol = 0.4 mixed monolayers are in a condensed state on the water surface.
For the cholesterol monolayer, the π-A isotherm curve (●) in Figure 2 showed that the π increased a little gradually and subsequently sudden steeply up to 30 mN/m. A0 was 0.39 ± 0.02 nm2/molecule at 30 mN/m of π and this value was also very similar (0.40 nm2/molecule at 35 mN/m) in compression method    . The obtained A0 was also identical (0.41 nm2/molecule at 34 mN/m) as reported previously by dropping method .
A0 of DMPC-cholesterol mixed monolayer in this research was compared to that of the ideal mixed monolayer that is calculated from each A0 of DMPC and cholesterol monolayers. Figure 3 is xChol dependence versus A0 plot (●) of DMPC-cholesterol mixed monolayer. The horizontal axis represents xChol and the vertical axis is A0/nm2 determined from π-A isotherm curve (Figure 2). The
Figure 3. Cholesterol mole fraction dependence of limiting molecular area A0. A0 in the case without the cholesterol-mixed effect is indicated by straight line. ●: In this study, ■: Previously reported data of Janout et al. .
straight line (-) is drawn as an ideal additivity on each xChol. In the figure, previously reported data of A0 (■)  versus xChol obtained by compression method have been compared with our results (●). From the data of our plot (●), the addition of cholesterol in the DMPC monolayer showed a negative deviation from the ideal line. This indicates that cholesterol molecules have a condensation effect on DMPC monolayer at various xChol  . The tendency of negative deviation in dropping method (●) and compression method (■) was almost same, although the negative deviation of plot (●) was slightly larger than that of plot (■) at more than xChol = 0.4. Therefore the deviation occurred in both the methods suggests that the DMPC-cholesterol mixed monolayer formation by dropping and compression methods are same and thus cholesterol condensation effect of the mixed monolayer is also same in both methods. The result of cholesterol condensation effect on DMPC monolayer was different from that of dipalmitoyl phosphatidyl choline (DPPC) monolayer, where the condensation effect of cholesterol on DPPC monolayer by dropping method was twice larger than compression method .
3.2. Brewster Angle Microscopy (BAM)
A series of BAM images containing five cholesterol mole fractions (xChol = 0.0, 0.3, 0.4, 0.6, 1.0) of DMPC-cholesterol mixed monolayers are shown in Figure 4. The xChol of Figures 4(a)-(d), Figures 4(e)-(h), Figures 4(i)-(l), Figures 4(m)-(p), and Figures 4(q)-(t), are 0.0, 0.3, 0.4, 0.6, and 1.0, respectively. For each xChol value, BAM images were recorded at four π values as given in the figure.
For xChol = 0.0 (Figures 4(a)-(d), the images showed that DMPC molecules form an expanding monolayer. Therefore the whole image was shown in slight bright contrast at 0.5 mN/m (Figure 4(a). With an increase of π to 10 and 32 mN/m as shown in Figure 4(b) and Figure 4(c), respectively, images became brighter as the DMPC monolayer continued to form homogeneously. At π = 42 mN/m in Figure 4(d), the brightness of image increased further and the image
Figure 4. BAM images of DMPC-cholesterol mixed monolayer at various surface pressures. (a)-(d): DMPC monolayer; (e)-(h): Mixed monolayer of xChol = 0.3; (i)-(l): Mixed monolayer of xChol = 0.4; (m)-(p): Mixed monolayer of xChol = 0.6; (q)-(t): cholesterol monolayer.
was brightest among all the above three π values, thereby confirmed that the completion of monolayer occurred at the water surface. The increase in brightness of the BAM images corresponded to a gradual increase in π values as shown in the π-A isotherm curve of DMPC monolayer (Figure 2, plot (○)), confirmed the formation of LE state. The monolayer formation process was almost the same as reported previously by compression method . The DMPC of an expandable molecule formed an expanded monolayer by both methods, as the aggregation process on the water surface remained the same in both methods.
On the DMPC-cholesterol mixed monolayer at xChol = 0.3 (Figures 4(e)-(h)), images were similar to the DMPC monolayer (Figures 4(a)-(d)). Comparing images Figure 4(e) with Figure 4(a), the image Figure 4(e) was somewhat brighter than Figure 4(a). It is due to that the DMPC molecules were attracted to each other by the addition of cholesterol into the DMPC monolayer. The above change in BAM images corresponded to the addition effect of cholesterol in Figure 2 (plot ◇) which shifted to the left by the addition of cholesterol. Similarly the mixed monolayer of xChol = 0.2 also showed similar changes in BAM images as those of xChol = 0.3 (images not shown).
On the DMPC-cholesterol mixed monolayer at xChol = 0.4 (Figures 4(i)-(l)), the images were different from xChol = 0.0, 0.2, and 0.3. At quite lower value of π = 0.5 mN/m (Figure 4(i)), the image showed line and circle type domain. The appearance of this domain indicated that condensed aggregates of DMPC and cholesterol molecules formed at the water surface. With an increase of π to 2 mN/m (Figure 4(j)), the domain became larger and the image was also brighter than Figure 4(i). At π = 9 mN/m (Figure 4(k)), the homogeneous state appeared and the domain obtained at 2 mN/m disappeared. At higher π = 35 mN/m (Figure 4(l)), the image became still brighter and indicating that the completion of the homogeneous DMPC-cholesterol monolayer formation on the water surface. The above change in BAM images were similar to the following cholesterol monolayer images (Figures 4(q)-(t)) which formed by the cholesterol- type condensed monolayer on the water surface. The change also corresponded to the addition effect of cholesterol in Figure 2 (plot▼) which shifted to the left by the addition of cholesterol and was similar to the cholesterol-like monolayer shape.
On the DMPC-cholesterol mixed monolayer at xChol = 0.6 (Figures 4(m)-(p)), the images were similar to those of xChol = 0.4 (Figures 4(i)-(l)) and also similar to the following cholesterol monolayer images (Figures 4(q)-(t)) which formed by the cholesterol-type condensed monolayer on the water surface, as same as xChol = 0.4 (Figures 4(i)-(l)). The change in BAM images also corresponded to the addition effect of cholesterol in Figure 2 (plot ¨) which shifted to the left by the addition of cholesterol and was similar to the cholesterol-like monolayer shape. Similarly the mixed monolayer of xChol = 0.8 also showed similar changes in BAM images as those of xChol = 0.4 and 0.6 (images not shown).
On the cholesterol monolayer at xChol = 1.0 (Figures 4(q)-(t)), the images showed that cholesterol molecules form a condensed monolayer. At quite lower value of π = 0.5 mN/m (Figure 4(q)), the image showed line and circle type domain, thereby the appearance of condensed aggregates of cholesterol on the water surface. With an increase of π to 2 and 7 mN/m as shown in Figure 4(r) and Figure 4(s), respectively, the number and size of line- and circle-type domains appeared at π = 0.5 mN/m increased and enlarged gradually. The images also became brighter due to high condensation forces of cholesterol molecules. At π = 29 mN/m in Figure 4(t), the domains appeared at <7 mN/m disappeared and the image became homogeneous. Among all images of xChol = 1.0, Figure 4(t) was most bright and thus the formation of homogeneous cholesterol monolayer was completed on the water surface. The monolayer formation process of cholesterol (Figures 4(q)-(t)) was almost the same as those of reported previously by compression method   . Therefore for condensable molecule such as cholesterol, the condensed monolayer is formed through the same aggregation process on the water surface either by dropping or compression methods.
3.3. Infrared External Reflection Spectroscopy (IERS)
Figure 3 showed the dependency of A0 on the added cholesterol in the mixed monolayer. At more than xChol = 0.4, the plot confirmed that A0 values of lipid approached to the cholesterol monolayer. It is specific that DMPC molecule reduced its A0 by the addition of cholesterol, because two DMPC molecule alkyl chains contain gauche conformation in the monolayer. It is predicted that the change in DMPC molecule alkyl chains conformation occurred by the addition of cholesterol. The IERS measurement has been performed for the investigation of alkyl chain structural conformation change occurred in DMPC molecules of the DMPC-cholesterol mixed monolayer.
Figure 5(a) and Figure 5(b) are IER spectra of DMPC monolayer without addition of cholesterol at π = 35 mN/m as shown in Figure 2 (plot ○). Figure 5(a) shows the result measured at the incident angle (IA) of 75˚ and Figure 5(b) at IA 10˚, respectively. The horizontal axis represents wavenumber/cm−1 and the vertical axis represents reflectance/%. The dotted lines (---) show measured spectrum and solid lines (―) are the simulated spectrum. Figure 5(c) is the absorption spectra (k-spectra) obtained from simulation calculation based on the solid experimental line in Figure 5(a) and Figure 5(b) . In Figure 5(c), the vertical axis represents imaginary part of the complex refractive index (k), solid line (―) spectra (kxy) is of the components parallel to the surface plane (in plane), dotted line (---) spectra (kz) is of the components perpendicular to the surface plane (out of plane) .
As shown in the k-spectra of Figure 5(c) (kz), five absorption peaks were assumed in simulation at 2958, 2920, 2897, 2872 and 2852 cm−1. These peaks were assigned to CH3 anti-symmetric stretching vibration band νas(CH3), CH2 anti-symmetric stretching vibration band νas(CH2), Fermi resonance CH3 symmetric stretching vibration band νs(CH3), and CH2 symmetric stretching vibration band νs(CH2), respectively. The two peaks at 2920 and 2852 cm−1 were strong in kxy spectrum assigned to the methylene groups in the two DMPC molecule alkyl chains.
To investigate the form of the two alkyl chains (the hydrophobic group) in the monolayer, we focused on the information of two stretching vibration bands (νas(CH2) and νs(CH2)) of methylene groups (-CH2-) in Figure 5(c). The bands (νas(CH2) and νs(CH2)) appeared at 2920 and 2852 cm−1, respectively are characteristics of gauche conformation of alkyl chain  . Therefore the hydrophobic DMPC molecule alkyl chains in the monolayer are in an irregularly kinked form.
Figure 6(a) shows IER spectra of DMPC-cholesterol (xChol = 0.4) mixed monolayer of π = 35 mN/m at the IA of 8˚. The dotted lines (---) show measured spectrum and solid lines (―) are the simulated spectrum. In Figure 6(b), the solid line (―) shows the absorption spectrum (k-spectrum) obtained from the simulation calculation  conducted using the solid line (―), the experimental IER recording shown in Figure 6(a). The solid lines (―) of Figure 6(b) is the absorption spectrum (k-spectra) (kxy) obtained from simulation calculation based on the solid experimental line in Figure 6(a) . For a comparison, the (kxy) of DMPC monolayer in Figure 5(c) is also described as the dotted lines (---) in Figure 6(b).
The two strong absorption peaks νas(CH2) and νs(CH2) were optimized to fit the experimental IER spectrum, resulting the peak position of kxy (solid lines (―)) at 2919 cm−1 and 2849 cm−1, respectively, for DMPC-cholesterol (xChol = 0.4) mixed monolayer. These peaks were assigned to CH2 anti-symmetric stretching vibration band νas(CH2) and CH2 symmetric stretching vibration band νs(CH2), respectively.
Figure 5. (a) and (b): IER spectra of DMPC monolayer at around a pressure of 35 mN/m. Incident angle is 75˚ (a) and 10˚ (b), respectively. Dotted-line is measured spectrum and solid-line is simulated spectrum. (c): absorption spectra (k-spectra) provided from simulated IER spectra of (a) and (b). Solid-line is in plane (kxy) and dotted-line is out of plane (kz).
Figure 6. (a): IER spectra of DMPC-cholesterol mixed monolayer (xChol = 0.4) at around a pressure of 35 mN/m. Incident angle is 10˚. Dotted-line is measured spectrum and solid-line is simulated spectrum. (b): absorption spectra (k-spectra in plane) provided from the simulation. Solid-line is mixed monolayer and dotted-line is DMPC monolayer kxy in Figure 5(c).
Comparing the two k-spectra of DMPC monolayer with (―) and without cholesterol (---) in Figure 6(b), two differences in peak position and absorption intensity were noticed. The peak positions of DMPC-cholesterol spectrum (―) νas(CH2) and νs(CH2) at 2919 and 2849 cm−1, respectively, were at shorter wavelengths than DMPC spectrum (---) νas(CH2) and νs(CH2) at 2920 and 2852 cm−1, respectively. The absorbance intensities of the former mixed monolayer’s bands were 1.3 times higher than the latter.
Snyder and Strauss  reported in C-C bond of polymethylene a lower (blue shift) in two νas(CH2) and νs(CH2) stretching vibration bands resulted due to change from gauche- to trans- conformation. Therefore the blue shift is suggested that the kinked form (gauche conformation) of DMPC molecules decreased by the addition of cholesterol to the monolayer. Thus addition of cholesterol caused the structural transition from gauche- to trans- conformation. The increase in the absorption intensity of the two peaks (νas(CH2) and νs(CH2)) by the addition of cholesterol also supports the structural change described above. The trans conformation consecutively arranged in the alkyl chain in all-trans zig-zag form with the long axis along the surface normal with parallel to the -CH2-plane at the surface plane. This arrangement makes the peak intensities of νas(CH2) and νs(CH2) higher (Figure 6(b)) on the addition of cholesterol in the k-spectra (in kxy plane) .
The blue shifted stretching frequency of νas(CH2) and νs(CH2) vibration bands of the two DMPC molecule alkyl hydrophobic chains, with the addition of cholesterol in the DMPC monolayer, leads the mixed monolayer more ordered and condensed by a decrease in the area of those alkyl chains with the structural change from disorder to order. This also corresponds to the negative deviation from the ideal line in Figure 3. The structural transition from gauche- to trans- conformation due to the addition of cholesterol occurred in the DMPC molecule alkyl chains. The alkyl chains ordered structure and more condensed state also corresponded to the STm result in Figure 2 that the shape of π-A isotherms curve showed LC state close to that of cholesterol.
In this research, we have investigated the cholesterol addition effect on the DMPC monolayer formed by dropping method. The π-A isotherm curves constructed by STm on various DMPC-cholesterol mixed monolayer showed the addition effect of cholesterol on the expandable DMPC monolayer, as a negative deviation of A0 from the ideal mixed A0. When the amount of cholesterol was >xChol = 0.4 A0 was almost the same as cholesterol. The recorded BAM images also supported STm results, the images containing cholesterol at
The results obtained in this research showed first the ordering of DMPC molecule occurred even at the level of monolayer on the water surface and the ordering was related not only to the general condensation effect of cholesterol but also to the gauche- to trans- conformation change in the two DMPC molecule alkyl chains. The ordering might have induced changes in the density and orientation of hydrophilic group of DMPC molecule at monolayer-water interface. More experiments are needed to obtain further information about the effect of cholesterol on the DMPC molecules in the DMPC-cholesterol mixed monolayer, using several other physicochemical methods such as surface potential measurement and vibration sum frequency generation (VFSG) spectroscopy.
This work was supported by a Grant-in-Aid from the Ministry of Science, Education, Sports and Culture (No. 21750143 and 24550154).