The Role of Cholesterol
The plasma membrane forms barrier in eukaryotic cells, it usually consists of a range of lipids and proteins. The lipids have different chain length, level of saturation and headgroup. These arrange non-covalently with hydrophobic force1 and Singer and Nicholson2 first described their structure using the classic fluid mosaic model. The membrane is described as fluid lamellar liquid crystalline (Lα) with acyl chains of the lipids are disordered while protein diffuses laterally between the layers. Later evidence shows that lipid could segregate laterally in certain conditions and forms distinct domains and were reported by different groups of researchers. Philips et al.3 in 1970 using differential scanning calorimetry (DSC) technique. At around the same time Shimshick and McConnell4 used electron spin resonance (ESR) to map the phase diagrams for the lateral phase separations and Schmidt and Barenholz5 using nuclear magnetic resonance (NMR) to suggest that sphingolipid could form microdomains.
Later Jain and White suggested the concept of the Plate Model6 to try to explain the experimental evidence gathered. The model suggests that the separation between the ordered region and the disordered region is due to the natural consequence of lattice deformation and intermolecular interaction. And it is suggested in some reviews7 that these mixing of lipid might be due to the presence of microdomains. In 1988, Simon and van Meer8 proposed the hypothesis of segregated lipid microdomains called ‘lipid rafts' are presences within the plasma membrane. One of the key in vivo evidence is the isolation of detergent-resistant membrane (DRM)9,10 from both animal and human cells. The raft is described as cholesterol and sphingolipid rich with specific lipid anchored protein presence. These rafts are proposed to exist in the ordered lamellar liquid crystalline phase (Lo), while the non-raft regions are in disordered lamellar liquid crystalline phase (Ld). Ipsen and Karlstrom11 first described the distinction the ordered and disordered phases. The ordered phase has the properties in between of the gel (Lβ) and liquid-crystalline (Ld) phase. Like in the gel (Lβ) phase, it has a lower proportion of gauche form in the hydrocarbon chains while like in the liquid-crystalline (Ld) phase it can diffuse freely both rotational and translational direction. This behavior is thought to be due to the high concentration of chol in the microdomains. Chol is thought to reduce and broaden the enthalpy between the gel and crystalline phase12,13 and forming a ‘condensed complex'14.
A wide range of phospholipids are present in the plasma membranes, each has a different chain length and degree of saturation and different headgroup. But they all share common features of a polar hydrophilic heard group and a hydrocarbon hydrophobic tail usually of 12-20 carbon. These are the building block of the membrane and can determine the overall shape, geometry and structure. When the level of unsaturation increases in the phospholipids, a bigger chain volume is resulted. Due to this property three types of phospholipids can be separated using head group: chain volume ratio. Type 0, Type I and Type II.
Different structure could be formed with different shape and type of the phospholipids. In solution, there are three major form that the phospholipids will arrange amongst itself; bilayer, micelle and liposome. In the bilayer structure, different phases could be present with different phospholipids and conditions, including the lamella and inverse hexagonal phases, in this project, all the phases present will be lamella phase.
This phase is made up of repeating units of flat stacks of bilayers with water in between. According to the difference of physical properties of the hydrocarbon chains, such as the order of the chain and the lateral diffusion between the bilayers, there are several sub-sets of lamellar phase characterized, these are the lamellar crystalline phase (Lc), the lamellar gel phase (Lβ) disordered fluid lamellar phase (Lα).
The Lc phase is the must rigid phase with the bilayers in a crystalline state, with lateral correlation between the bilayers. In the case where the bilayer is in a crystalline state but there is no lateral correlation, then the phase is classified as sub-gel.
In the Lβ phase, the hydrocarbon chains pack hexagonally into a 2D lattice with all the chain completely rigid and extended to an all-trans conformation. When the chain is parallel to the bilayer normal, therefore decrease the cross section and thickens the bilayer, all motion within are restricted 15 (see figure 1). the phase is referred as Lβ while if it is tilted (usually at 30°(REF)) it is referred as Lβ'. The typical intra-hydrocarbon chain distance in an Lβ phase is 4.2Å with a cross section area of 21 Å2.
When the temperature is increased above the Tm, the hydrocarbon chains starts to melt and the system becomes more disordered, which results in the Lα phase, however in the case of Lβ' a pre-transition phase might occur during this process and form a ripple phase (Pβ'). The Tm for different lipids will vary, as this value is strongly dependent on the structure of the lipid, in most cases the longest acyl chain and smallest headgroup will result in the highest Tm.
In the Lα phase, each phospholipids will undergo rapid lateral diffusion and long axis rotation, and trans-gauche isomerism will also occur, resulting a shorter chain length and a expanded cross section area, the bilayer also become thinner. From going to all-trans to gauche conformation, the position of the gauche will effect the chain dynamics as different conformation will have different chain extension and interaction between neighboring chains.
There are two 18C hydrocarbon chains on DSPC with a cis double bond at the C9-C10 position on both of the hydrocarbon chains. The cis-unsaturation at the chain results in a much lower chain melting temperature Tm. (-20°C) (REF), it is believed that DOPC do not usually interact with chol and it is always in La phase as it melts at very low temperature, this is because chol pack closely to the phospholipid and require the lipid's hydrocarbon chain to be in the all-trans conformation which in the case of DOPC, is impossible because of the cis double bond.
A range of experiments had been done to determine the different interactions and affinities of chol towards different type of phospholipids16. It was founded that the affinity of chol towards the phospholipids change with different headgroup and hydrocarbon chain, generally increase with the level of saturation, this is thought to because it would be easier for the chol to partition between the unsaturated lipid to form the chol rich liquid ordered microdomains.
The existence of lipid rafts in vivo are difficult to prove but evidence of such phases of Ld and Lo co-existence had been proved using different spectroscopy technique.
When using real cells to perform experiments have proven to be very difficult, model membrane of a different range mixture of lipid compositions had been used to mimic the biological environment of lipid rafts. Usually this is made up of an unsaturated lipid with a low Tm (DOPC, POPC), a saturated lipid with a high Tm (DPPC, SM) and chol in excess water. As mention above, chol prefers to interact with long and saturated lipids, so in the model system, partitioning of the chol with the saturated lipids are favored to the unsaturated domains. It is suggested that the saturated lipids and chol rich area is present transiently in the membrane with a higher molecular order 17,18. The investigation of the co-existence of the fluid-fluid phase is very important as it gives evidence if these rafts can be found in the plasma membrane.
In the model, as chol is partition with different degree into the two phospholipids, two different domains are formed. It was first reported in a binary system using DPPC-d62 with chol. The Lo is composed of chol rich domain while the Ld phase is composed of chol poor domain19,20. The ordering effect19 of chol can be seen here as the membrane thickness21and rigidity22 is increase. In this area of interest, the combination of DPPC, chol and DOPC is used most frequently; saturated DPPC binds with chol strongly while unsaturated DOPC stays relatively ‘unreactive' as binding is not favorable.
Different mixtures of the lipids will have different phases at equilibrium, and these are described using a framework with the distinction of the different phases as a function of temperatures and quantities. Data used to construct the diagram can either be obtained using spectroscopic approaches like fluorescence, ESR and NMR, or by using thermodynamic techniques like DSC23. First the phase boundaries are determined then tie line are estimated and drawn. In a binary system, the tie line are isotherms while in a ternary system the although the tie lines are also isothermal, the direction has no relation to the composition axes, in addition it changes with temperature and therefore needs to be determined using data from experiments24.
End point spectra: spectral subtractions method
The end point spectra for a binary mixture can be obtained using the subtraction method24. Within a set temperature, there will be a number of spectra that shows phase co-existence with different sample compositions, each of these spectra contain a set of sub-spectra with different fractions. The end point spectra can be obtained by subtracting different fractions of one spectrum from another19. For ternary systems, although the samples are less likely to lie within the tie lines, but for a small range of temperature, it is possible that the sample will stay close to the tie lines and the same spectral subtraction method can also be used.
Various physical spectroscopy techniques like florescence microscopy and atomic force microscopy (AMF) had been used in this area whist the former are the most frequently used. Other techniques like Nuclear Magnetic Resonance, X-ray diffraction, Differential Scanning Calorimetry and electron spin resonance are also frequently used in this area. There are advantages and disadvantages with each technique, for example Giant unilamellar vesicle (GUV) are directly observed using fluorescence microscopy to identify the different phase boundaries, temperature can be changed to see the effect and the transition takes places, though as there are limited shapes and colour of these domain, to made identification of the composition in the domain not possible. Moreover, the vesicle has to be up to a certain size to be observed. In the following sections, details of different techniques are discussed.
Fluorescence microscopy and atomic force microscopy
These techniques allow us to look into the lateral structure of the lipid bilayer of spatial resolutions ranging from nanometer (AFM) to micrometers (fluorescence microscopy). Probes are needed for fluorescence as it has a lateral resolution limit of around 300 nm. For AFM, probes are not required and as it as a resolution in Å, it can view lipid domain with phase separations in planar membranes25. For AMF, mono- or bilayer are used whist for fluorescence both mono/bilayer and GUVs can be used. By using these techniques the physical information like size, time evolution or shape of the lipid domains can be obtained. However, there is a problem with the fluorescence microscopy which the images obtained is dependent on the lipid domain composition rather than the phase state26. The shapes and colours of the image are also limited, which gives no actual information regarding the distribution of the lipids in the domains. Using this technique, only the phase separations can be identified while the actual content inside the domain cannot be observed.
Thermodynamic technique: DSC
This technique gives us the information within the enthalpy change in the lipid system, when the gel phase turns into the liquid phase a sharp signal is obtained27. Studies also show that as cholesterol level increases the transition enthalpy will decrease and broaden in a binary system, Mannock et al.28 have studied the effect of chol in egg sphingomyelin and discovered that when chol is present it causes the sphingomyelin transition endotherm to broaden and the enthalpy to decrease, the length of the lipid chain length also contribute as a factor in the heat capacity during the transition. In DPPC/Chol systems it is found that when the chol concentration is higher than 22mol%, the system will be only in Lo phase19.
X-ray diffraction spectroscopy
These can be used to define the structure and lamellar bilayer spacing of the lipid phases. Quinn et al. had identified three co-existing region of Lß, Ld and Lo in MLVs made up with DPPC/DOPC/Chol and estimated that chol was distributed equally in both the Ld and Lß phases29. Different chain ordering of different phases can be obtained using wide angle X-ray scattering (WAXS). For example in the Lo phase, the chain scattering intensity is concentrated on the equator while for the Ld phase is it more spread towards the meridian in a broad arc30. For Small Angle X-ray Scattering (SAXS), it is used to measure lamellar bilayer spacing and this spacing gives an indication of the chol and phospholipids chain packing.
Solid State Nuclear Magnetic Resonance (SSNMR)
Although many different methods had been used to investigate the lipids mixture, this is going to be the primarily technique of this project. One advantage of using SSNMR is that this technique does not require intrusive probes. Probes can be problematic even at a very low concentration as it can give out misleading information if it partitions between different phases. Quantitative measurement of the co-existing phases can be obtained using SSNMR. 2H NMR can be used to investigate the system's chain dynamic as it can reflects the conformational disorder of the chain group, it is also particular useful in determining the tie-line endpoints as it provide a quantitative measure of the two phases. Additional information about the motion of the phospholipids headgroup can be obtained using 31P NMR as there is only one phosphorus atom in the phospholipids, and the interactions such as dipolar coupling and chemical shift anisotropy (CSA) depends on the orientation of the headgroup.
Each headgroup of the phospholipids contains one spin 1/2 31P nucleus, which is 100% abundant; therefore it gives a strong signal. There is a strong electronic interaction between the nuclear and the election, in solution this can be averaged out due to fast rotation of the molecule, but in stolid state this can be cause a large effect due to restricted motion, therefore a CSA powder pattern is used to represent all possible orientation of the phospholipids. The powder pattern has a principle values making up of chemical-shift tensor, and the value of these tensor can be estimated from the spectrum, an example is shown below in figure 5.
As the 31P CSA pattern is headgroup geometry and its local dynamics sensitive, by changing the position of the sample with respect the external magnetic field of the NMR, information of the headgroup conformation can be obtained.32 The linewidth and lineshape can be used to identify different phases and they are temperature dependent. For characteristic Ld phase, the linewidth is narrow and sharper and so as the lineshape, while for the Lß phase both the linewidth and lineshape is boarder. This effect is seen, as the CSA pattern is sensitive to the local motions that each different phase processes.33
MAS 31P NMR
MAS NMR is used to partially remove the CSA contribution to the lineshape. The sample is spin at the magic angle θm of 54.74 degree (cos2θm=1/3). By doing this the nuclear dipole-dipole interaction caused by nuclei magnetic moments averages to zero while the CSA nuclear-electron interaction will average to a non-zero value. The result is a spectrum comparable to one in the solution state, also by removing the CSA contribution, the isotropic chemical shift can be seen, this is very useful as the phase behavior of the system can be reflected by the relative intensity of the chemical shift.
The Static 31P CSA NMR has been used extensively throughout the past decades on investigating the headgroup dynamics and structure of the phospholipid bilayer systems. (ref) Despite the advantages the MAS NMR could offer, the work done on this area of research remains relatively limited. Recently there are an increase interested in this area and several lipid systems like egg sphingomyelin (SM)/DOPC/Chol34,35 and DPPC/DOPC/Chol system.36,37 MAS has a lot of advantage like removing overlapping of the power pattern and improving the resolutions, and it also requires less amount of sample comparing to the static NMR. In order to completely remove the Ds of the sample, the spinning rate has to be faster than 12KHz, but as the rapid spin rate could create heat du to frictional force, therefore a slower spin rate are used and therefore sidebands are resulted. The isotropic full width at half maximum (FWHM) and intensity are investigated to distinguish the different phases. For example, the FWHM of the gel phase will be wider; this is resulted from the slower and more restricted exchange of the conformation while in the liquid phase the FWHM is much narrower due to more conformation motion in the head group.
Deuterium is quadrupolar has a spin number I of 1, therefore it has 2I+1=3 level of Zeeman energy and two allowed transitions. As it is quadrupolar, the most dominated effect presence is the electronic quadrupolar interaction rather then CSA or dipolar coupling in the 31P case. As the natural abundance of deuterium is very small, pre-deuterated lipids are used as they are commercially available, the deuterium could be introduced to different areas of the lipid chain so different part of the chain could be investigated. DSPC-d70 is used in my project; both the sn-1 and sn-2 chains are deuterated at C2-C18. As only DSPC is deuterium labeled, therefore only the signal contributed by DSPC will show on the spectra, therefore made the analysis of the phase behavior of the system easier.
Different phases have different kind of motion within the hydrocarbon chain and characteristic powder patterns are resulted. For example, in the Lb phase, motion are restricted and therefore gives a board peak with a doublet in the middle given by the methyl group from the phospholipid acyl chain, while in the La phase, there are much more rotational freedom and the rate of motion is much faster than the NMR timescale therefore the peaks sharpen. The quadrupolar splitting could also use to determine the relative degree of motion within the hydrocarbon chain; the more motion-restricted region has a smaller quadrupolar splitting, the typical value for the methylene in an Lb is 50kHz while for the La is 25kHz.
This is a mathematical way to describe the quadrupolar splitting, when the C-2H bond is aligned parallel with the external fiend, the value of the quadrupolar splitting can be used to extract the order parameters. There is however one major problem relating to this method as there are a large amount of overlapping in spectra, a numerical method of ‘de-paking' is introduced to resolve this problem by removing the anisotropy effect in the system38. Different approach to this numerical method has been proposed including fast Fourier methods39, complete analytical solutions and original iterative method. Order parameter is important as it gives comparison between different phases, information of the motion in the bilayer can be qualitatively measured as order parameters is sensitive to conformational dynamics and molecular orientation.
The order parameter can be calculated using these following equations (ref) The relationship between the observed quadrupolar splitting and order parameter is Where is a constant, is the angle between the C-2H bond and the external magnetic field and After de-paking are performed, the angle will become 0 as now the C-2H bond are parallel to the field When , , putting this back to equation (X) gives Rearranging equation (X) gives Materials and Experimental
All phospholipids were purchased from Avanti Polar Lipids (Birmingham, USA), with >98% purity in anhydrous form, they were used without further purification. HPLC grade water and cholesterol were supplied from Sigma Aldrich Chemical, Gillingham, UK. The cholesterol were provided in the monohydrate form with >99% purity.
Analytical grade cyclohexane and methanol were supplied by VWR International, Lutterworth, UK.
All phospholipids were dissolved in cyclohexane with a few drops of methanol then lyophilised under vacuum in order to obtain the anhydrous form. The phospholipids/ cholesterol mixtures were then made up by mixing appropriate molar ratio of each component. The molar ratios were calculated assuming each phospholipid were two water molecules hydrated. Appropriate amount of HPLC grade H2O was then added to the mixture assuming excess hydration requires ten water molecules per lipid. Each sample was made up to 30 mg. The mixture was then centrifuged, followed by ten freeze thaw cycle using liquid nitrogen and hair dryer to ensure complete mixing.
The prepared samples were stored in freezer and transferred to 4mm zirconia NMR rotors when studied.
Solid State NMR Spectroscopy
All NMR data were acquired on a Bruker DRX (Karlsurhe, Germany) 600 MHz NMR spectrometer, which operates at 14.09T with a 31P resonance of 242.9 MHz and a 2H resonance at 62.1 MHz.. The recycle delay for 31P is 2.0s and for 2H is 0.5s. ____ scans were perform with the MAS 31P NMR and _____ for the 2H. Magic angle spinning were performed at 3-5kHz.Standard pulse programs were used to acquire single pulse and proton decoupled spectra. Deuterium spectra were acquired using a standard phase-cycled quadrupole echo pulse sequence. A ten-minute break was allowed before changing to a different temperature to ensure the system to be equilibrated.
x-ray diffraction measurements were carried out using a WAXS beamline in Diamond light source, Oxfordshire (beamline I22).
Throughout the years a lot of different lipids are used with chol to make up a binary membrane model, the behaviors are studied and therefore mapping the phase diagrams. Some systems are particularly popular like SM/chol40,41 and DPPC/chol42. Different techniques were employed and quite different phase diagram were resulted due to problems discussed above. Huang et al. had used 13C and 2H SSNMR to study the binary system of DPPC/chol and DSPC/Chol, for 13C SSNMR regions of hydrocarbon chains like the methelyene group and the carbonyl were investigated, the changes results in phase transitions made useful for mapping the phase boundaries in the phase diagram. For 2H SSNMR the quadrapolar splitting was analyzed against different concentration of chol and temperature, it was found that as chol concentration increases the transition was broaden and the mid-point was shifted indicating a more ordered phase. The phase diagrams were proposed.
In my project, different sets of binary system made up of DSPC/chol are investigated with varying chol concentration at a range of different temperature. Data are then compared with the binary phase diagram proposed by Huang at el. The DSPC system is chosen as relatively a lot more of detail work had been done on DPPC. While both are phosphatidylcholine(PC), DSPC has a slightly longer hydrocarbon chain and a slightly higher Tm; this could affect the behavior of the membrane. Five sets of binary mixture were made by varying the concentration of chol from 90% to 50%, such a high percentage of chol is chosen although there are relatively only a small amount of chol in the plasma membrane, but as when other lipids like DSPC and DOPC are present, chol has a tendency to partition into the Lo phase while other lipids are in the La phase. This means that in fact the effective concentration of chol is much higher as it seems. In addition the amount of work in higher level of chol are relatively limited, therefore it could be beneficial to look at theses area of the phase diagram.
31P MAS NMR
The sample were first heated to 343K (70°C) to ensure the lipids are fully mixed and then nine experiment were recorded while the sample cools to 278K (5°C), then finally back to room temperature 296K (23°C). On initial inspection of the spectra, there is a broadening effect at the lower temperature scans and the intensity of the isotropic peak also seems to decrease with temperature. (fig. 10) This indicates the presence of an Lb phase at low temperature and a more disordered La phase at higher temperature.
All of the spectra are fitted using a computer simulation Topspin to obtain the physical spectroscopic parameters, which represent the tensor in a numerical form. From this key parameters like the axiality of the CSA tensor can be obtained.
The effect of temperature and chol level are investigated, for all chol composition, there is a general trend of as temperature increase the CSA parameter decreases, the system goes from ordered to disordered (reduction of CSA means more headgroup dynamics). When the level of chol in the system increase, the value of the CSA decreases, for example at the Tm of DSPC (323K) for 50% chol Ds=26.22 while for 90% chol Ds= 24.57, this effect could be due to the increasing averaging between two different phases as the level of Lo increases at higher chol level.
There is a general trend for all chol level that as temperature increases the intensity increase, then it reaches a maximum 30-40 °C, as temperature increase further; the intensity starts to level off. As the level of chol increases, the temperature which the intensity reaches maximum decreases. For 90% chol, intensity reaches maximum at 30°C, while for 50% chol the intensity reaches at 40 °C.
The FWHM is the measure of the width across at the half height of the isotropic peak. This value is shown to be temperature and chol level dependent as well (fig. 12). However there is a lot of errors associated while measuring the linewidth, therefore might lead to inaccuracy and difficulties on analysing the data, an more accurate mathematical approach could be used as an alternative way to look at the data (see moment analysis section below)
Form the graph we can see that there is a dramatic decrease in the linewidth for all chol concentration as temperature increase from 5 to 30/40 °C. This effect is more prone for the higher chol concentration sample. In addition for the higher chol concentration sample the linewidth minimum are reached at a lower temperature, for 80% chol, the linewidth reaches minimum at 30 °C while for 50% chol is at 40 °C. Above 40 °C there is no significant changes in the linewidth.
Although we could obtain information of the phase behaviors of the bilayer system by measuring the linewidth, the accuracy of this method is not very high and percentage of error could be hard to determined. Moreover there are problems associated with extracting the principle values of the tensor under strong inhomogeneous and poor signal to noise ratio. An alternative method was proposed to analysis the lineshape by calculating the moments43, and later it was proposed that by doing moment calculations the principal values of chemical shielding can be obtained44.
Computer software simulations were used to aid the analysis of the spectra to estimate the physical parameters, the program Bruker Topspin 2.0 were used to fit the spectra. The estimated came out to be reasonably accurate as overall a overlap of above 90% were achieved. Information like the isotropic chemical shift (diso), axiality of the CSA tensor (ds) and asymmetric parameter (h) were estimated.
These parameters could be used to calculated moments, as mentioned above this is an extremely powerful method to analyze the lineshape of the spectra. For 31P, a spin half nucleus, the first moment (M1) is just simply the diso, while the higher moments includes terms like h and ds. General forms of the equations are given below45.
As the chemical shift is responsible for most of the interaction in a spin half nucleus, the observed frequency (v) are given by:
(1) a and b are polar angles of the chemical shielding tensor with respect to the z-axis of the laboratory frame. And moment is define as:
(2) Therefore substituting equation (1) into (2) we get
(3) After integrating equation (3) and normalized the results to we get:
Moments higher than the third moment were not used here as they depend on the previous moments of M2 and M3.
Using the software Bruker Topspin 2.0, fitting of the spectra were performed and an average of 90% overlap were obtained; this means the approximation is reasonably accurate to carry on the moment analysis. From the fitting the different physical parameters needed to perform moment calculations were estimated. Using the equation above the second and third moment are calculated for all of the binary systems.
(see table 1)
Table 1. physical spectroscopic parameters obtained from the simulation program for the binary systems and the second and third moment calculated.
The results from the moment analysis for both M2 and M3 follow the same trend as we had for the linewidth analysis, the moment decrease when temperature is increased. Although the parameter used for the calculation are only estimated and the fittings are not 100% accurate, the moment analysis agrees strongly with the linewidth analysis above which confirms the trend which linewidth does decrease with increasing temperature. This shows that there is a phase transition at around 20°C
Ternary system: model membrane
The model membrane is widely used when investigating phase behavior as it can reflect the underlying physical interactions of the phospholipids that make up the membrane. Traditionally, the phospholipids system used to make up a model membrane is DOPC/DPPC/chol in excess water. In this case, DSPC are used instead of DPPC, they both share the same headgroup of choline, as they are only different in the level of saturation at the hydrocarbon chain, therefore the effect could be investigated. The effect of chol on DSPC was discussed in the above section. By understanding the phase behavior in the binary systems could help us to understand the more complex behavior by addition of another lipid. It is believed that due to the physical properties of DOPC, there will be minimum interaction with the chol and stay in the La phase throughout, while the saturated lipid (DSPC/DPPC) interacts with the chol and form different phases, some even describe unsaturated lipid like DOPC as ‘unreactive' towards chol whilst saturated lipids like DPPC/DSPC as ‘active'46. Different method had been employed to construct the ternary phase diagram of the DPPC/DOPC/chol system. Veatch et al. has proposed a ternary diagram for DOPC/DPPC/chol using fluorescence microscopy, the lipid phase could be observe directly from the GUV, it was reported that two co-existing liquid phases, one rich in saturated lipid (DPPC) with chol and one rich in unsaturated lipid (DOPC), were observed in a wide range of temperature and lipid compositions47.
Although an example of using DSPC in a ternary system is rare, a ternary phase diagram of DSPC/DOPC/chol were proposed by Zhao et al. using the technique of light microscope48 .
(1) D, C. Nature 2005, 437, 640-647.
(2) S.J, S.; G.L., N. Science 1972, 175, 720-31.
(3) Phillips, M. C.; Ladbrooke, B. D.; Chapman, D. Biochim. Biophys. Acta. 1970, 196, 35-44.
(4) Shimshick; E.J.; McConnell, H. M. Biochemistry 1973, 12, 2351-2360.
(5) Schmidt, C. F.; Barenholz, Y.; Thompson, T. E. Biochemistry 1977, 16, 2649-2655.
(6) Jain, M. K.; White III, H. B. Adv. Lipid Res. 1977, 15, 1-60.
(7) Thompson, T. E.; Tillack, T. W. Annu. Rev. Biophys. Biophys. Chem. 1985, 14, 361-386.
(8) Simons, K.; van Meer, G. Biochemistry 1988, 27, 6197-6202.
(9) Brown, D.; Rose, J. Cell 1992, 68.
(10) Brown, D. A.; London, E. Biochem Biophys Res Commun 1997, 240, 1-7.
(11) Ipsen, J. H.; Karlstrom, G.; Mouritsen, O. G.; Wennerstrom, H.; Zuckermann, M. J. Biochim. Biophys. Acta 1987, 905, 162-172.
(12) Ohvo-Rekila, H.; Ramstedt , B.; Leppimaki , P.; Slotte, J. P. Prog Lipid Res 2002, 41, 66 -97.
(13) McMullen, T. P. W.; McElhaney, R. N. Curr Opin Colloid Interface Sci 1996, 1, 83 -90.
(14) McConnell, H. Biophys. 2005, 88.
(15) Gennis, R. B. pringer advanced texts in chemistry 1989.
(16) Silvius, J. P. Biochim Biophys Acta 2003, 1610, 174-183.
(17) Simons, K.; Ikonen, E. Nature 1997, 387, 569-572.
(18) London, E. Biochim. Biophys. Acta. 2005, 1746, 203-220.
(19) Vist, M. R., and J. H. Davis. Biochemistry 1990, 29, 451-464.
(20) Davis, J. H. NMR studies of chol orientational order and dynamics, and the phase equilibria of chol/phospholipid mixture, In physics of NMR Spectroscopy in Biology and Medicine. ; North Holland, Amsterdam, 1986.
(21) Chachaty, C.; Rainteau, D.; Tessier, C.; Quinn, P. J.; Wolf, C. Biophys. J. 2005, 88, 4032-4044.
(22) Maraviglia; B.; Davis, J. H.; Bloom, M.; Westerman, J.; Wirtz, K. W. A. Biochim. Biophys. Acta. 1982, 686, 137-140.
(23) Goni, F. M.; Alonso, A.; Bagatolli, L. A.; Brown, R. E.; Marsh, D.; Prieto, M.; Thewalt, J. L. Biochimica et Biophysica Acta 2008, 1781, 665-684.
(24) Davies, J. H.; Clair, J. J.; J., J. Biophysical Journal 2009, 96.
(25) Connell, S. D.; Smith, D. A. Mol. Membr. Biol. 2006, 23, 17-28.
(26) Batatolli, L. A. Biochim. Biophys. Acta 2006, 1758, 1541-1556.
(27) Wang, B.; Tan, F. Science in China (Series B) 1999, 42, 8-13.
(28) Mannock, D. A.; McIntosh, T. J.; Jiang, X.; Covey, D. F.; McElhaney, R. N. Biophys. J. 2003, 84, 1038-1046.
(29) Chen, L.; Yu, Z.; Quinn, P. J. Biochim. Biophys. Acta 2007, 1768 2873-2881.
(30) Mills, T. T.; Tristram-Nagle, S.; Heberle, F. A.; Morales, M. F.; Zhao, J.; Wu, J.; Toombes, G. E. S.; Nagle, F. J.; Feigenson, G. W. Biophys. J. 2008, 95, 682-690.
(31) Smith, I. C. P.; Ekiel, I. H. Phosphours-31 NMR of Phospholipids in Membranes.In Phosphorus-31 NMR; Principles and applications. D Gorenstein, ed.; Academic Press Inc. London, England, 1984.
(32) Griffin, R. G.; Powers, L.; Pershan, P. S. Biochemistry 1978, 17, 2718-2722.
(33) Dufourc, E. J.; Mayer, C.; Stohrer, J.; Althoff, G.; Kothe, G.Biophys. J. 1992, 61, 42-57.
(34) Holland, G. P.; McIntyre, S. K.; T.M., A. Biophysical Journa 2006, 90, 4248-4260.
(35) Costello, A. L.; Alam, T. M. Biochimica et Biophysica Acta 2008, 1778, 97-104.
(36) Clarke, J. A.; Herron, A. J.; Seddon, H. M.; Law, R. V. Biophysical Journal 2006, 90, 2383-2393.
(37) Clarke, J. A.; Seddon, H. M.; Law, R. V. Soft matter 2009, 5.
(38) Bloom, M.; Davis, J. H.; Mackay, A. L. Chemical Physics Letters 1981, 80, 198-202.
(39) McCabe, M. A.; Wassall, S. R. Journal of Magnetic Resonance Series B 1995, 106, 80-82.
(40) Cullis, P. R.; Hope, M. J. Biochem. Biophys. Acta.1990, 597, 533-542.
(41) Guo, W.; Kurze, V.; Huber, T.; Afdhal, N. H.; Beyer, K.; Hamilton, J. A. Biophysical Journal2002, 80, 1465-1478.
(42) Huang, T. H.; Lee, C. W. B.; Das Gupta, S. K.; Blume, A.; Griffin, R. G. Biochemistry1993,32, 13277-13287.
(43) Van Vleck, J. H. Phys. Rev.1948, 74, 1168.
(44) Maricq, M. M.; Waugh, J. S. J. Chem. Phys. 1979, 70,3300.
(45) Herreros, B.; Metz, A. W.; Harbison, G. S. Solid State Nuclear Magnetic Resonance 2002, 16, 141-150.
(46) McConnell, H.; Radhakrishnan, A. PNAS 2006, 103, 1184-1189.
(47) Veatch, S. L.; Keller, S. L. 2003, 85, 3074-3083.
(48) Zhao, J.; Wu, J.; Heberle, F. A.; Mills, T. T.; Klawitter, P.; Huang, G.; Costanza, G.; Feigenson, G. W.Biochimica et Biophysica Acta 2007, 1768, 2764-2776.