J. Phys. Chem. B 2010, 114, 12685–12691
12685
Acetonitrile Induces Nonsynchronous Interdigitation and Dehydration of Dipalmitoylphosphatidylcholine Bilayers Fu-Gen Wu, Nan-Nan Wang, Le-Fu Tao, and Zhi-Wu Yu* Key Laboratory of Bioorganic Phosphorous Chemistry and Chemical Biology (Ministry of Education), Department of Chemistry, Tsinghua UniVersity, Beijing 100084, P. R. China ReceiVed: May 8, 2010; ReVised Manuscript ReceiVed: August 27, 2010
The formation mechanism of the interdigitated (LβI) phase and the responses of the individual groups of phospholipids to the phase transition are of basic concern within the community of lipid research. In this work, we studied the effect of acetonitrile (CH3CN) on the structure and phase behavior of dipalmitoylphosphatidylcholine (DPPC) bilayers by using differential scanning calorimetry, synchrotron X-ray diffraction, and Fourier transform infrared spectroscopy. We found that the two processes (i.e., the interdigitation and dehydration of the DPPC bilayers) occur nonsynchronously at two different CH3CN concentrations (4 and 12 wt %). A detailed submolecular picture for the formation mechanism of the LβI phase was provided during the Lβ′ to LβI phase transition at c(CH3CN) ) 4 wt %: the conformation state and the hexagonal packing mode of the lipid acyl chains and the hydration properties of the lipid polar groups do not change, and the only difference is that the formed LβI phase has a tighter lipid acyl chain packing than that of the Lβ′ phase. When c(CH3CN) > 12 wt %, the added CH3CN molecules selectively dehydrate the interfacial carbonyl groups. Thus, two different kinds of LβI phases differing only in the hydration states of the interfacial carbonyl groups of phospholipids exist in the c(CH3CN) regions of 5-12 and 13-40 wt %, respectively. The strong ability of acetonitrile to induce interdigitation in the lipid bilayers has been discussed in the viewpoint of its toxicity. 1. Introduction The phospholipid bilayer serves as a structural framework for cell membranes and is often considered as a primary target for small guest molecules. Manipulating the structure and properties of cell membranes by small molecules is an issue that is important in fundamental biophysical studies, environment toxicology, and biomedical applications (e.g., anesthesia, cryopreservation, and permeation enhancement). Some frequently studied small polar organic molecules include DMSO,1–8 ethanol,9–17 glycerol,18–24 urea,25–27 formamide,28–31 acetone,32–34 etc. These small solutes can significantly affect the membrane structures (a typical example is to induce the formation of water pores in membranes,4,5,35 membrane properties (e.g., membrane permeability, membrane dynamics, and mechanics), and phase behaviors (e.g., phase stability, transition kinetics, and inducement of new phases, such as the interdigitated phase). Acetonitrile is a less frequently studied small molecule with a medium polarity and is miscible with water. It has a moderate liquid range, a low viscosity, and a low chemical reactivity and can dissolve a wide range of ionic and nonpolar substances. On the basis of its properties, acetonitrile is widely used in liquid chromatography, electrochemistry, purification of organic compounds, and manufactures of DNA oligonucleotides, pharmaceuticals, and photographic films. Acetonitrile has a modest toxicity, and its wide applications in laboratories and industries increase its probability to affect the health of human bodies by inhalation, ingestion, and skin absorption. Hence, acetonitrile molecules may have chances to interact with cell membranes and exert their toxic influences on the structure and properties of phospholipid membranes. * To whom correspondence should be addressed. Phone: (+86) 10 6279 2492. Fax: (+86) 10 6277 1149. E-mail:
[email protected].
The interaction of acetonitrile with dipalmitoylphosphatidylcholine (DPPC) has been studied by Kinoshita and Yamazaki.32 By using differential scanning calorimetry (DSC), they found that the main transition (i.e., tilted rippled gel, Pβ′, to lamellar liquid crystalline, LR, phase transition) temperature of DPPC decreased with an increase in acetonitrile concentration from 0 to 6.0 vol % and increased above 6.0 vol %. X-ray diffraction data revealed that a phase transition from tilted lamellar gel (Lβ′) to interdigitated gel (LβI) phase occurred at 5.0 vol %, and the lipids were completely in the LβI phase above 6.0 vol % at 20 °C. Although numerous efforts have been made to study the transition mechanisms of phospholipids, some long-standing and challenging questions still remain. For example, although the Lβ′-LβI phase transition processes in phosphatidylcholine (PC) bilayers have been investigated,36–42 the formation mechanisms of the interdigitated phase are still not completely resolved, especially with regard to what happens to the individual groups of the PC molecules during the transition process. We are particularly curious to know the submolecular details that are crucial to understand the transition mechanisms of phospholipids. To this end, it is necessary for us to investigate the role the individual groups/portions of an amphiphilic molecule play in the transformation process from one phase to another. In this contribution, we selected acetonitrile, which represents the kind of solutes that cannot form conventional hydrogen bonds with PCs, to study its effect on the structure and phase behavior of DPPC bilayers by using DSC, synchrotron smalland wide-angle X-ray diffraction (SAXS and WAXS), and Fourier transform infrared (FTIR) spectroscopy. We will emphasize the formation mechanisms of the interdigitated phase at the submolecular level.
10.1021/jp104190z 2010 American Chemical Society Published on Web 09/13/2010
12686
J. Phys. Chem. B, Vol. 114, No. 39, 2010
2. Experimental Section 2.1. Sample Preparation. DPPC was purchased from Avanti Polar Lipids (Birmingham, AL). Acetonitrile (g99.6%) was from Beijing Beihua Fine Chemical Co., Ltd. (Beijing, China). Double-deionized water with a resistivity of 18.2 MΩ cm was used for the preparation of the lipid samples. The dry DPPC powder was dispersed into a CH3CN-H2O mixed solvent with a lipid/solvent ratio of 1/3 (w/w). The concentrations of CH3CN in the CH3CN-H2O mixtures were 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 20, 25, 30, and 40 wt %. Homogeneous lipid dispersions were prepared by repeated thermal cycles between 20 and 60 °C. 2.2. DSC. Calorimetric data were obtained with a differential scanning calorimeter DSC821e (Mettler-Toledo Co., Switzerland) equipped with the high-sensitivity sensor HSS7, which has a temperature precision of (0.1 °C and a heat flow precision of (0.01 mW. The heating rate was 0.5 °C/min. 2.3. Synchrotron X-ray Diffraction. SAXS and WAXS experiments were performed at the beamline BL16B1 of the Shanghai Synchrotron Radiation Facility (SSRF) (λ ) 1.24 Å). A Rayonix SX-165 CCD detector was used to collect the X-ray scattering data. A standard silver behenate sample was used for the calibration of diffraction spacings. X-ray scattering intensity patterns were recorded during 60 s exposure of the samples to the synchrotron beam. To obtain the SAXS and WAXS data simultaneously, we fixed the sample-to-detector distance at 473.6 mm. A Linkam thermal stage (Linkam Scientific Instruments, UK) was used for temperature control ((0.1 °C). The X-ray powder diffraction intensity data were analyzed using the program Fit2D. The precision of the calculated scattering vector (the q value) is estimated to be (0.1 nm-1. 2.4. FTIR Spectroscopy. FTIR spectra were recorded using a Nicolet 5700 Fourier transform infrared spectrometer with a DTGS detector in the range of 4000-900 cm-1 with a spectral resolution of 2 cm-1 and a zero filling factor of 2. The precision of the wavenumber measurement is better than 0.1 cm-1, but the determination of the peak position may have some uncertainties, depending on the band shape. Samples were coated onto the inner surfaces of a pair of CaF2 windows, which were mounted on a Linkam heating-cooling stage for temperature control ((0.1 °C). The spectra were recorded every 30 s, and each spectrum consisted of 16 scans. 3. Results and Discussion 3.1. DSC. Figure 1 shows the DSC results of DPPC in CH3CN-H2O mixed solvents. The results show two thermal events at low concentrations of acetonitrile (c(CH3CN) < 5 wt %). In pure water, the onset and peak temperatures of the first endothermic transition are 35.0 and 36.1 °C, respectively. Those of the second endothermic transition are 41.8 and 42.3 °C, respectively. These results are in good agreement with the published data, and the two thermal events have been identified as transitions from the tilted lamellar gel (Lβ′) to the tilted rippled gel (Pβ′) and then to the lamellar liquid crystalline (LR) phase.7,43 By adding CH3CN to c(CH3CN) ) 1 wt %, the onset and peak temperatures of the Lβ′ to Pβ′ pretransition shift to 33.1 and 34.6 °C, respectively. When c(CH3CN) is increased to 4 wt %, the peak temperature of the pretransition process (Tp) decreases almost linearly with the increase of c(CH3CN) (the corresponding temperature data are shown in Figure 2A). Upon increasing c(CH3CN) to 5 wt %, an “abnormal” and complex phase transition behavior was observed, with a diffused peak at ∼37 °C and two overlapped peaks in the temperature region of 39-41 °C. The two overlapped peaks indicate the coexistence
Wu et al.
Figure 1. DSC results of DPPC in CH3CN-H2O mixtures. The heating rate is 0.5 °C/min.
of the original phase and a new phase. The results suggest that a special event has occurred by increasing c(CH3CN) from 4 to 5 wt % at 20 °C. For the Pβ′-to-LR main transitions, we found that when c(CH3CN) e 5 wt %, the peak temperature of the main transitions (Tm) decreases almost linearly, but with a much smaller amplitude as compared with those of the pretransition processes (see Figure 2A). The evident downward shift of the pretransition temperature suggests that the increase in CH3CN in the mixed solvent favors the formation of the Pβ′ phase at the expense of the Lβ′ phase. This can be explained as a result of the weakening interactions between the headgroups with increasing concentration of CH3CN in the headgroup region. Yu and Quinn once formulated three conditions for the formation of the rippled phase.7 The weakening in the interactions between headgroups allows the occurrence of steric repulsion and, thus, membrane surface undulation at lower temperatures. The main transition, which mainly involves the lipid acyl chain melting, is only slightly influenced by the addition of the CH3CN molecules. Its decreasing manner is also in agreement with the weakening of the interactions between headgroups. When c(CH3CN) > 5 wt %, the pretransition processes disappear, leaving only one phase transition process. This single DSC peak is assigned to the interdigitated gel phase (LβI) to LR phase transition, as evidenced by the SAXS/WAXS results (see section 3.2). Meanwhile, as shown in Figure 2A, the Tm behaves in a manner of first increase and then slight decrease.
Interdigitation and Dehydration of DPPC Bilayers
J. Phys. Chem. B, Vol. 114, No. 39, 2010 12687
Figure 2. (A) Dependencies of the peak temperatures of the pretransition (Lβ′ to Pβ′) and main transition (Pβ′ or LβI to LR) processes of DPPC in CH3CN-H2O mixtures. (B) The corresponding enthalpy changes of the pre- and main transitions.
In general, the change of Tm is not significant in all the DPPC dispersions in the c(CH3CN) region of 0-40 wt %, indicating that the addition of CH3CN molecules only slightly perturbs the melting process of DPPC molecules. Shown in Figure 2B are the corresponding enthalpy changes of the pre- and main transitions. The enthalpy change for the pretransition process is 5.2 ( 0.2 kJ/mol at c(CH3CN) ) 0 wt %, and it decreases to 3.2 ( 0.2 kJ/mol at c(CH3CN) ) 4 wt %. The decrease in the enthalpy change, together with the decrease in the pretransition temperature, may also indicate weakening of the interactions between the lipid headgroups with increasing concentration of CH3CN in the headgroup region. The average enthalpy change for the Lβ′-to-LR transition (which includes the enthalpy values of the Lβ′-Pβ′ and Pβ′-LR transitions) is 39.8 ( 0.4 kJ/mol when c(CH3CN) ) 0-4 wt %, which is smaller than that (43.9 ( 1.1 kJ/mol) for the LβIto-LR transition when c(CH3CN) > 4 wt %. The larger enthalpy change for the LβI-to-LR transition as compared with that for the Lβ′-to-LR transition may indicate that the LβI phase has a tighter acyl chain packing as compared with that of the Lβ′ phase, which is further confirmed by the WAXS data shown in section 3.2. 3.2. SAXS and WAXS. To investigate the structural changes of DPPC in the CH3CN-H2O mixed solvents at 20 °C, we carried out SAXS and WAXS measurements, and the results are shown in Figure 3. In Figure 3, we can see from the left part of the figure (the small-angle region) that, for the samples studied, the scattering vectors (q) of the diffraction orders show a ratio of 1:2:3:4:5 or 1:2:3 (with the periodical orders marked with numbers in the figure), showing that all the samples are lamellar-structured. From the q values, we can obtain the repeat distances (the d values) using the equation d ) 2π/q. The thus obtained d values are 6.4 nm for the DPPC in H2O (c(CH3CN) ) 0 wt %); 6.1 nm for c(CH3CN) ) 2 wt %; and 5.0 nm for c(CH3CN) ) 5, 15, and 40 wt %. The results show that when c(CH3CN) > 4 wt %, the lamellar repeat distances of the phospholipids were almost conserved. For the wide-angle region (the right part of Figure 3), the X-ray scattering intensity profiles of c(CH3CN) ) 0 and 2 wt % are characterized by a sharp peak centered at a spacing of 0.42 nm and a broad shoulder on the high-angle side, which arises from a packing of the tilted acyl chains in a quasihexagonal lattice,7,44,45 whereas for c(CH3CN) ) 5, 15, and 40 wt %, the wide-angle X-ray scattering peaks all reside at a spacing of 0.41 nm and the symmetric feature of the intensity profiles indicates that the lipids are in the nontilted gel phase.
Figure 3. Synchrotron X-ray diffraction results of some selected samples of DPPC in CH3CN-H2O mixtures (c(CH3CN) ) 0, 2, 5, 15, and 40 wt %) at 20 °C.
That is, by increasing the concentration of CH3CN to c(CH3CN) > 4 wt %, the lipid acyl tails convert from the original tilted gel (Lβ′) to a nontilted gel phase at 20 °C. According to the phase transition trend in the c(CH3CN) region of 0-4 wt % (Figure 1), we may expect the lipid samples with c(CH3CN) > 4 wt % to be in the Pβ′ phase at 20 °C. However, the obtained WAXS results clearly demonstrate that the lipids at the high concentration range are in a nontilted gel phase (the LβI phase). This is quiet unexpected, and the formation mechanism of the LβI phase at the critical CH3CN concentration will be discussed later. Previous studies on the neat DPPC aqueous dispersions have revealed that the lipid bilayer thickness (dl) and water layer thickness (dw) in the Lβ′ phase are 4.6 and 1.8 nm (the d value is around 6.4 nm, where d ) dl + dw), respectively, and in the Pβ′ phase, they are 4.73 and 2.24 nm (the d value is around 7.0 nm), respectively.46,47 Our present X-ray data show that when c(CH3CN) > 4 wt %, the d value is only 5.0 nm, a big drop from 6.4 nm in the Lβ′ phase in the absence of acetonitrile. The value 5.0 nm is very close to the lipid bilayer thickness (dl ) 4.6 or 4.7 nm) in the tilted gel phase (Lβ′ or Pβ′ phase). Moreover, the WAXS data reflecting the carbon-carbon packing information show that the average carbon-carbon spacing in the nontilted gel phase is smaller than that in the Lβ′ phase, which means a tighter lipid acyl chain packing in the former than that in the latter.
12688
J. Phys. Chem. B, Vol. 114, No. 39, 2010
Wu et al.
Figure 4. FTIR absorption spectra of DPPC in CH3CN-H2O mixtures with a concentration range from 0 to 40 wt %: (A) 3000-2800, (B) 1800-1700, (C) 1525-1425, and (D) 1150-950 cm-1. For clarity, the molecular structure of DPPC was also shown in the figure.
All these results strongly suggest the formation of an interdigitated structure (the LβI phase), in which the lipid molecules from opposing monolayers are interpenetrated or interdigitated. A similar conclusion was also drawn by Kinoshita and Yamazaki, who found that a phase transition from Lβ′ to LβI occurred at c(CH3CN) ) 5.0 vol % (or 4.0 wt %).32 In addition, in the interdigitated phase of the DPPC bilayers (when c(CH3CN) > 4 wt %), the observed d-spacing value (0.41 nm) in the WAXS profile is the same as that of the LβI phase in the DPPC-ethanol system,41 which may also support the formation of the LβI phase in our system. 3.3. FTIR. Having detailedly studied the phase behavior of the DPPC-CH3CN system from the macroscopic (DSC) and structural aspect (X-ray), we would like to gain submolecular information of the solute-dependent phase behavior of DPPC in the CH3CN-H2O mixed solvents at 20 °C by using FTIR spectroscopy. The corresponding IR results are shown in Figure 4. By monitoring the characteristic IR vibrations of the polar headgroups (PO4- and N+-CH3), the interface groups (CdO), and the apolar tail groups (CH2s), we can detect the changes of the different regions of the self-assembled phospholipid membranes during the solute-induced phase transition process. The 3000-2800 cm-1 region (Figure 4A) contains the CH2 asymmetric and symmetric stretching vibrations (denoted as νasCH2 and νsCH2, respectively), which reflect the conformational order of the lipid methylene chains. During the Lβ′-toLβI transition, the IR bands of νasCH2 and νsCH2 are almost fixed at 2917.6 and 2849.5 cm-1, respectively. As the peak position changes of the two stretching vibrations have been used frequently to follow the conformational order and the transgauche isomerization of the CH2 chains in lipid tail regions,48,49 the observed results in this work indicate an ordered lipid acyl tail packing, and such state is conserved upon the CH3CNinduced Lβ′-to-LβI phase transition. The dynamic evolution of the CdO stretching band (νCdO) in the 1760-1700 cm-1 region upon increasing c(CH3CN) from 0 to 40 wt % gives information on the alterations of the hydration state and hydrogen bonding network in the lipid interface region; the results are shown in Figure 4B. The corresponding wavenumber changes of νCdO are depicted in Figure 5. In the c(CH3CN) region of 0-12 wt %, the νCdO bands center at ∼1730 cm-1, and only a very small frequency drift (within 1 cm-1) was observed, indicating that the hydration of the CdO groups is almost unperturbed with the addition of CH3CN molecules in this concentration range. When c(CH3CN) > 12 wt %, evident upward wavenumber shifts were observed,
Figure 5. The change of the wavenumber of the CdO stretching vibration (νCdO) with c(CH3CN) at 20 °C. The regions of the Lβ′ and LβI phases, and the subregions of the two LβI structures, LβI (1) and LβI (2), are indicated in the figure.
and at the final CH3CN concentration studied (40 wt %), the νCdO band centers at ∼1737 cm-1, corresponding to a blue shift of 7 cm-1. The large change in the wavenumber of the νCdO band indicates that the interfacial CdO groups are significantly dehydrated by the added CH3CN molecules. Figure 4C contains the CH2 scissoring band (δCH2) centered at ∼1468 cm-1 in the absence of acetonitrile. Upon increasing c(CH3CN) from 0 to 40 wt %, no significant changes in the band shape and band position occur. This band is sensitive to the intermolecular forces and can serve as a key band for examining the state of packing of the methylene chains in various phases.48,50,51 The single sharp peak centered at ∼1468 cm-1 indicates that the methylene trans-zigzag planes are packed in a hexagonal state in both of the Lβ′ and LβI phases. The results show that the hexagonal packing mode of the lipid acyl tails is hardly affected by the addition of CH3CN molecules. The other two important bands are the IR vibrations of the PO4- and N(CH3)3+ groups in the DPPC head region. These two bands are sensitive to the change of the conformation and hydration of the lipid headgroups. Again, we can see that almost no band position shift or band shape change occurs in the symmetric stretching vibration of the PO4- group (νsPO2-) at ∼1088 cm-1 and the asymmetric stretching vibration of the N(CH3)3+ group (νasN+-CH3) at ∼971 cm-1 (Figure 4D). The results show that the lipid head region does not change in the hydration state by the addition of CH3CN molecules. It is noticed that the asymmetric stretching vibration of the PO4- group (νasPO2-) is not selected for the above analysis.
Interdigitation and Dehydration of DPPC Bilayers
J. Phys. Chem. B, Vol. 114, No. 39, 2010 12689
Figure 6. A schematic model illustrating the structural changes of the DPPC bilayers upon the addition of CH3CN at 20 °C.
The reason is that a number of peaks mainly assigned to the CH2 wagging progressions of all-trans methylene chains51,52 reside on the shoulder of the νasPO2- band in the region of 1300-1200 cm-1 (data not shown), which deviates the band contour and band position of the νasPO2- band. The above FTIR results show that for all the IR vibrations (νasCH2, νsCH2, νCdO, δCH2, νsPO2-, and νasN+-CH3) studied, only the νCdO band begins to undergo upward wavenumber shift at c(CH3CN) ) 12 wt %. This means that above this threshold concentration, CH3CN can selectively dehydrate the CdO groups in the lipid polar region. This can be regarded as a solute-induced nonsynchronicity phenomenon in phospholipids (i.e., the different parts of the phospholipids respond differently to the added solutes). More interestingly, the dehydration of the interfacial CdO groups begins at c(CH3CN) ) 12 wt %, which is in the concentration region (5-40 wt %) of the LβI phase. Accordingly, we divide the c(CH3CN) region into two sections, 5-12 and 13-40 wt %, corresponding to two different kinds of LβI phases. The LβI phase (denoted as LβI (1) in Figure 5) in the low c(CH3CN) region of 5-12 wt % has the same hydration state of all the polar groups (CdO, PO4-, and N(CH3)3+) as that in the Lβ′ phase, whereas the LβI phase (denoted as LβI (2) in Figure 5) in the high c(CH3CN) region of 13-40 wt % differs only in the hydration state of the CdO groups, but not other polar groups, compared with the LβI (1) phase. To the best of our knowledge, this is the first observation of two distinct structures of the LβI phase. Several earlier studies have reported the coexistence of two kinds of liquid crystalline phases in a number of systems under certain conditions.53–55 For example, in the fluids of dipolar oblate ellipsoids, there are two types of liquid crystalline phases (ferroelectric liquid and antiferroelectric liquid) differing in the ferroelectric properties in two different temperature regions,53 but our work demonstrates that two interdigitated phases differing in the interfacial hydration state of phospholipid bilayers are present in two different solute (acetonitrile) concentration regions. Moreover, the FTIR data show that the conformation state and the hexagonal packing mode of the lipid acyl chains and the hydration properties of the lipid polar groups do not change during the phase transition from Lβ′ to LβI at c(CH3CN) ) 4 wt %, although the WAXS data demonstrate a tighter lipid acyl chain packing in the latter phase state. Thus, altogether, we have provided a relatively detailed submolecular picture for the formation mechanism of the interdigitated phase. From an overall viewpoint, we found that the two processes (i.e., the interdigitation and dehydration of the DPPC bilayers induced by the addition of CH3CN molecules) occur nonsynchronously. 3.4. Interactions between CH3CN and DPPC Bilayers. By combining the DSC, X-ray diffraction, and FTIR results, we
gave a schematic model to show the structural changes of the DPPC bilayers upon the addition of CH3CN molecules at 20 °C (Figure 6). In the model, the added CH3CN molecules are located just below the lipid headgroups due to the amphiphilicity nature and act as spacers in the lipid-solvent interface, which enlarges the distance between the neighboring lipid polar groups in the same lipid monolayer and induces the formation of the LβI phase at c(CH3CN) ) 4 wt %. However, in the c(CH3CN) region of 0-12 wt %, the CH3CN molecules do not have direct interaction with any of the polar groups (the interfacial CdO groups and the headgroups PO4- and N(CH3)3+) and do not perturb the hydration layer of these polar groups as revealed by our FTIR data. This means that the influences exerted by the addition of CH3CN molecules on the phase state change (from Lβ′ to LβI) may depend solely on the location of the solutes and the corresponding geometrical effects. Upon increasing c(CH3CN) to above 12 wt %, more CH3CN molecules insert into the lipid-solvent interface and eventually replace the preexisting water molecules that previously hydrate the CdO groups, leading to the dehydration of the interfacial CdO groups. The WAXS data, however, do not signal any influence of CH3CN on the carbon-carbon spacings of the lipid acyl chains, showing that at such high concentrations, the CH3CN molecules still do not lead to the change of the packing tightness of the lipid acyl chains. In the two c(CH3CN) regions of 5-12 and 13-40 wt %, although the interfacial CdO groups show different hydration states, the constant d value (5.0 nm) and the conserved IR bands of νasCH2/νsCH2 and δCH2 indicate that the increase in CH3CN molecules does not change the membrane thickness or the ordering and packing of the lipid acyl chains. The amphiphilicity nature causes the CH3CN molecules to locate at the lipid-solvent interface and controls the dehydration of the lipid CdO groups at high CH3CN concentrations, resulting in the competitive interactions between CH3CN and water with the lipid CdO groups. The dehydration of the interfacial CdO groups may influence the dipole interactions between solvent and lipids and the dielectric properties of the solvent around lipid polar groups and may further regulate membrane diffusion, elasticity, and permeability abilities. Here, we will emphasize that in our samples studied, the water contents are sufficient to fully hydrate the lipid polar groups. The limiting water contents of DPPC (that is, the required weight ratios of water in the DPPC-water system to completely hydrate the lipid polar groups) for the Lβ′, Pβ′, and LR phases are 27, 31, and 36 wt %, respectively.46,47 In our studied systems, the sample with the lowest water content is c(CH3CN) ) 40 wt %. By taking the lipid/solvent ratio of 1/3 (w/w) into account, the calculated water content in the whole system is 45 wt %, which is larger than the limiting water contents of DPPC in all three phases (the Lβ′, Pβ′, and LR phases). That is, in our studied systems with the c(CH3CN) value ranging from 0 to 40 wt %,
12690
J. Phys. Chem. B, Vol. 114, No. 39, 2010
the water content is sufficient to fully hydrate the lipid polar groups. Thus, the observed dehydration effect is exclusively caused by the added CH3CN molecules that bind to the lipid-solvent interface and replace the preexisting water molecules hydrating the CdO groups. On the basis of our present experimental data, it is difficult to accurately determine the number and orientation of CH3CN molecules in the membrane interfacial region. However, since the CH3CN molecule has a large solubility of alkanes and can partition into the interfacial region between water and the segments of the terminal methyl groups of the membrane in the LβI phase,32 if we assume that in the LβI(1) phase, the CH3CN molecules are exclusively residing at the interfacial region, we can estimate the number of acetonitrile molecules in the membrane interface at c(CH3CN) ) 5 wt % to be 2.7 per lipid and at c(CH3CN) ) 12 wt % to be 6.4 per lipid. This means that when more than two acetonitrile molecules per lipid are incorporated into the membrane interfacial region, the interdigitation will be induced. When c(CH3CN) increases to above 12 wt % (in the LβI(2) phase), more acetonitrile molecules will be incorporated into the interfacial region, with some adhering to the CdO groups and causing them to dehydrate. At the very high concentration (e.g., 40 wt %), the distribution of acetonitrile molecules in the interlamellar aqueous region or in the bulk solution must be considerable. Yamazaki et al.32,56 have explained the formation mechanism of the LβI phase from the Lβ′ phase by considering the chemical potential difference of the two phases, ∆µ ) µ(LβI) - µ(Lβ′). Particularly, the χ parameter (the interaction energy parameter) was introduced to explain the induction mechanism of the LβI phase through the equation χ ) ∆G/2kBT, where ∆G is the free energy increase associated with the contact of segments with solvent, or the free energy decrease associated with the contact between segments, and kB and T are the Boltzmann constant and absolute temperature, respectively.32 They proposed that in the absence of small organic solvents, such as CH3CN, the χ parameter of the segment of the alkyl chain in the LβI structure, which is mainly determined by the terminal methyl group of the alkyl chain, is large, and ∆µ > 0. While in the presence of small organic solvents, such as CH3CN, these molecules partition into the interfacial region between water and the segment of the terminal alkyl chains of the membrane in the LβI phase. Since the organic solvents have a large solubility of alkanes, the χ parameter of the segment of the alkyl chain in the LβI structure may decrease with an increase in concentrations of such solvents, resulting in the decrease of ∆µ. Above a critical concentration, ∆µ is negative, and thereby, the LβI phase forms. Hence, the decrease in the interaction energy (the χ parameter) between the segments of the terminal alkyl chain of the phospholipid and surrounding solvents is one of the most important factors to induce the LβI phase. Our previous studies have shown that during the thermotropic phase transition processes, the different parts of the amphiphilic lipid molecules can undergo nonsynchronous changes,50,51,57 and our present work gives another type of nonsynchronicity where the phospholipid DPPC molecules change nonsynchronously upon the perturbation of the solute concentration. That is, the acyl chain rearrangement and the dehydration of the interfacial region upon addition of CH3CN take place nonsynchronously at two different solute concentrations: the abrupt change in the former is 4 wt %, and that in the latter is 12 wt %. Finally, it is valuable to discuss and compare the results obtained from the DPPC-CH3CN-H2O system with those from the DPPC-DMSO-H2O system because DMSO is also an
Wu et al. aprotic polar solute that cannot form the normal hydrogen bonds with the PC polar groups, which is quite similar with the case of CH3CN. As DMSO molecules hydrogen-bond strongly with water and exhibit unfavorable interactions with the polar headgroups of PCs, they also prefer to remain either directly below the headgroups (which is related to the amphiphilicity nature of the DMSO molecule) or in the aqueous phase at the low DMSO concentrations.3,8 However, large differences between the effect of DMSO and CH3CN on the phase behaviors of PCs have been found. The temperatures of the pre- and main transitions of DPPC increase linearly with the increasing DMSO content. The difference in the increasing rates of the two transition temperatures results in the disappearance of the Pβ′ phase and the direct transition between the Lβ′ and LR phases.7 That is, DMSO stabilizes the Lβ′ and Pβ′ phases of DPPC at the expense of the LR phase. However, in our DPPC-CH3CN-H2O system, the added CH3CN molecules decrease the two transition temperatures almost linearly with increasing c(CH3CN) at low CH3CN concentrations ( 12 wt %, although the lipids are still in the LβI phase, the added CH3CN molecules selectively dehydrate the interfacial CdO groups. Such selective interaction can be viewed as a solute-induced nonsynchronicity phenomenon observed in the self-assembled phospholipid aggregates. Moreover, we provided the submolecular evidence that there are two structures of the LβI phase differing only in the hydration states of the interfacial CdO groups of the self-assembled phospholipids. They exist in the c(CH3CN) regions of 5-12 and 13-40 wt %, respectively. Finally, the abilities of acetonitrile to induce interdigitation and dehydration in DPPC bilayers have been discussed in the viewpoints of its toxicity and its influence on membrane permeation. Acknowledgment. This work was supported by grants from the Natural Science Foundation of China (NSFC: 20633080 and 20973100) and a “973” National Key Basic Research Program of China (Grant No. 2006CB806203). The SAXS and WAXS data were collected at the beamline BL16B1 of the Shanghai Synchrotron Radiation Facility (SSRF) with the assistances of the station scientists. References and Notes (1) Yu, Z. W.; Quinn, P. J. Mol. Membr. Biol. 1998, 15, 59–68. (2) Gordeliy, V. I.; Kiselev, M. A.; Lesieur, P.; Pole, A. V.; Teixeira, J. Biophys. J. 1998, 75, 2343–2351. (3) Sum, A. K.; de Pablo, J. J. Biophys. J. 2003, 85, 3636–3645. (4) Notman, R.; Noro, M.; O’Malley, B.; Anwar, J. J. Am. Chem. Soc. 2006, 128, 13982–13983. (5) Gurtovenko, A. A.; Anwar, J. J. Phys. Chem. B 2007, 111, 10453– 10460. (6) Gurtovenko, A. A.; Onike, O. I.; Anwar, J. Langmuir 2008, 24, 9656–9660. (7) Yu, Z. W.; Quinn, P. J. Biophys. J. 1995, 69, 1456–1463. (8) Yu, Z. W.; Quinn, P. J. Biochim. Biophys. Acta 2000, 1509, 440–450. (9) Terama, E.; Ollila, O. H. S.; Salonen, E.; Rowat, A. C.; Trandum, C.; Westh, P.; Patra, M.; Karttunen, M.; Vattulainen, I. J. Phys. Chem. B 2008, 112, 4131–4139. (10) Chanda, J.; Chakraborty, S.; Bandyopadhyay, S. J. Phys. Chem. B 2006, 110, 3791–3797. (11) Feller, S. E.; Brown, C. A.; Nizza, D. T.; Gawrisch, K. Biophys. J. 2002, 82, 1396–1404.
J. Phys. Chem. B, Vol. 114, No. 39, 2010 12691 (12) Patra, M.; Salonen, E.; Terama, E.; Vattulainen, I.; Faller, R.; Lee, B.; Holopainen, J.; Karttunen, M. Biophys. J. 2006, 90, 1121–1135. (13) Dickey, A. N.; Faller, R. Biophys. J. 2007, 92, 2366–2376. (14) Dickey, A. N.; Yim, W. S.; Faller, R. J. Phys. Chem. B 2009, 113, 2388–2397. (15) Trandum, C.; Westh, P.; Jørgensen, K.; Mouritsen, O. G. Biochim. Biophys. Acta 1999, 1420, 179–188. (16) Westh, P.; Trandum, C. Biochim. Biophys. Acta 1999, 1421, 261–272. (17) Trandum, C.; Westh, P.; Jørgensen, K.; Mouritsen, O. G. Biophys. J. 2000, 78, 2486–2492. (18) Biondi, A. C.; Disalvo, E. A. Biochim. Biophys. Acta 1990, 1028, 43–48. (19) Williams, W. P.; Quinn, P. J.; Tsonev, L. I.; Koynova, R. D. Biochim. Biophys. Acta 1991, 1062, 123–132. (20) Yu, Z. W.; Tsvetkova, N. M.; Tsonev, L. I.; Quinn, P. J. Biochim. Biophys. Acta 1995, 1237, 135–142. (21) Swamy, M. J.; Marsh, D. Biophys. J. 1995, 69, 1402–1408. (22) Chen, L.; Yu, Z. W.; Quinn, P. J. Biophys. Chem. 2001, 89, 231–238. (23) Takahashi, H.; Ohta, N.; Hatta, I. Chem. Phys. Lipids 2001, 112, 93–97. (24) Westh, P. Biophys. J. 2003, 84, 341–349. (25) Sanderson, P. W.; Lis, L. J.; Quinn, P. J.; Williams, W. P. Biochim. Biophys. Acta 1991, 1067, 43–50. (26) Feng, Y.; Yu, Z. W.; Quinn, P. J. Chem. Phys. Lipids 2002, 114, 149–157. (27) Costa-Balogh, F. O.; Wennerstro¨m, H.; Wadso¨, L.; Sparr, E. J. Phys. Chem. B 2006, 110, 23845–23852. (28) Walter, A.; Gutknecht, J. J. Membr. Biol. 1986, 90, 207–217. (29) McIntosh, T. J.; Magid, A. D.; Simon, S. A. Biochemistry 1989, 28, 7904–7912. (30) Li, S. J.; Kinoshita, K.; Furuike, S.; Yamazaki, M. Biophys. Chem. 1999, 81, 191–196. (31) Jedlovszky, P.; Mezei, M. J. Phys. Chem. B 2003, 107, 5322–5332. (32) Kinoshita, K.; Yamazaki, M. Biochim. Biophys. Acta 1996, 1284, 233–239. (33) Kinoshita, K.; Asano, T.; Yamazaki, M. Chem. Phys. Lipids 1997, 85, 53–65. (34) Kinoshita, K.; Yamazaki, M. Biochim. Biophys. Acta 1997, 1330, 199–206. (35) Gurtovenko, A. A.; Anwar, J. J. Phys. Chem. B 2007, 111, 13379– 13382. (36) Nambi, P.; Rowe, E. S.; McIntosh, T. J. Biochemistry 1988, 27, 9175–9182. (37) Ohki, K.; Tamura, K.; Hatta, I. Biochim. Biophys. Acta 1990, 1028, 215–222. (38) Komatsu, H.; Rowe, E. S. Biochemistry 1991, 30, 2463–2470. (39) Yamazaki, M.; Miyazu, M.; Asano, T. Biochim. Biophys. Acta 1992, 1106, 94–98. (40) Yamazaki, M.; Miyazu, M.; Asano, T.; Yuba, A.; Kume, N. Biophys. J. 1994, 66, 729–733. (41) Adachi, T.; Takahashi, H.; Ohki, K.; Hatta, I. Biophys. J. 1995, 68, 1850–1855. (42) Kranenburg, M.; Vlaar, M.; Smit, B. Biophys. J. 2004, 87, 1596– 1605. (43) Lewis, R. N. A. H.; Mak, N.; McElhaney, R. N. Biochemistry 1987, 26, 6118–6126. (44) Tardieu, A.; Luzzati, V.; Reman, F. C. J. Mol. Biol. 1973, 75, 711– 733. (45) Brady, G. W.; Fein, D. B. Biochim. Biophys. Acta 1977, 464, 249– 259. (46) Inoko, Y.; Mitsui, T. J. Phys. Soc. Jpn. 1978, 44, 1918–1924. (47) Marsh, D. CRC Handbook of Lipid Bilayers; CRC Press: Boca Raton, 1990. (48) Lewis, R. N. A. H.; McElhaney, R. N. Methods Mol. Biol. 2007, 400, 207-226. (49) Wu, F. G.; Chen, L.; Yu, Z. W. J. Phys. Chem. B 2009, 113, 869–872. (50) Wu, F. G.; Wang, N. N.; Yu, Z. W. Langmuir 2009, 25, 13394– 13401. (51) Wu, F. G.; Wang, N. N.; Yu, J. S.; Luo, J. J.; Yu, Z. W. J. Phys. Chem. B 2010, 114, 2158–2164. (52) Senak, L.; Moore, D.; Mendelsohn, R. J. Phys. Chem. 1992, 96, 2749–2754. (53) Ayton, G.; Wei, D. Q.; Patey, G. N. Phys. ReV. E 1997, 55, 447–454. (54) Horton, J. C.; Donald, A. M.; Hill, A. Nature 1990, 346, 44–45. (55) Galindo, A.; Jackson, G.; Photinos, D. J. Chem. Phys. Lett. 2000, 325, 631–638. (56) Furuike, S.; Levadny, V. G.; Li, S. J.; Yamazaki, M. Biophys. J. 1999, 77, 2015–2023. (57) Wu, F. G.; Luo, J. J.; Yu, Z. W. Langmuir 2010, 26, 12777–12784. (58) Kinoshita, K.; Li, S. J.; Yamazaki, M. Eur. Biophys. J. 2001, 30, 207–220.
JP104190Z