- Email: [email protected]
Effect of a Decapeptide (VPDLLADLLK) on the Phase Transition of Dimyristoylphosphatidylcholine Lipid Bilayer
Journal of Colloid and Interface Science 240, 24–29 (2001) doi:10.1006/jcis.2001.7644, available online at http://www.idealibrary.com on
Effect of a Decapeptide (VPDLLADLLK) on the Phase Transition of Dimyristoylphosphatidylcholine Lipid Bilayer J. Shobini and A. K. Mishra1 Department of Chemistry, Indian Institute of Technology Madras, Chennai 600 036, India Received June 14, 2000; accepted April 21, 2001
therefore interesting to study the nature of interactions between peptides and lipids in model systems. The study of model systems consisting of peptides, either chemically synthesized or naturally occurring, interacting with the bilayers has proven to be a valuable contribution to a better understanding of these mechanisms (4). The synthetic peptide approach has the advantage of being flexible: the effects of small changes in the molecule on the lipid–peptide interactions can be determined. The ability of a substance to alter the phase-transition characteristics of a phospholipid model membrane is indicative of its association with the lipid either by some simple method or by more complicated means (5). An understanding of phase transitions and the fluidity of the phospholipid membrane is important in both the manufacture and exploitation of liposomes, since the phase behavior of a liposome membrane determines such properties as permeability, fusion, aggregation, and protein binding, all of which can markedly affect the stability of liposomes and their behavior in biological systems (6). Several peptides such as (synthetic) poly-(L-lysine) (7), neuropeptides such as neuromidicin B (8), antibacterial peptides such as nisin (9), and an antibiotic peptide iturin A (10) are known to interact differently with the dimyristoylphosphatidylcholine (DMPC) lipid bilayer. They show a wide variety of changes in the physical properties of the membrane, such as fluidity, permeability, and phase-transition temperature. Some peptides, such as nisin, also showed a perturbing effect on the hydrophobic region of the bilayer (9). In this work we report for the first time the effect of a synthetic decapeptide (VPDLLADLLK) on the phase-transition behavior of DMPC liposome membrane. This is a synthetic peptide, having hydrophobic and hydrophilic amino acid residues. A BLAST search (11) of this sequence produces 114 matches having significant alignments (60–80%) with many proteins (many of them are enzymes). Circular dichroism spectroscopy has been used to monitor the conformational changes of the peptide; differential scanning calorimetry and fluorescence spectroscopy using 1,6diphenyl-1,3,5-hexatriene (DPH) as an extrinsic fluorescence probe have been used to monitor the phase-transition behavior of the lipid bilayer. The fluorescence anisotropy of DPH in the lipid bilayer membrane has been extensively used in literature to study membrane-related changes, primarily as a fluidity probe (6, 12).
VPDLLADLLK is a synthetic decapeptide, which shows a difference in conformation in various environments. Circular dichroism spectral studies show that it exists in an unordered conformation in the aqueous phase, and in dimyristoylphosphatidylcholine (DMPC) lipid bilayer, it exhibits an α-helical structure. The membrane property modification due to the peptide incorporation has been studied by using differential scanning calorimetry and fluorescence spectroscopy. With incorporation of the peptide the average steady-state anisotropy of DPH in the membrane decreases slightly in the gel state but remains more or less the same in the liquid crystalline state. The peptide incorporation causes a shift in the phase-transition temperature from 23 to 26◦ C for 15 mol% and 29◦ C for 30 mol% of the peptide, which is accompanied by a decrease in the sharpness and a broadening of the DSC thermogram. This preferential stabilization of the more ordered gel phase by the peptide could be due to the hydrophobic mismatch between the length of the peptide and the length of the hydrophobic segment of the DMPC bilayer. ° C 2001
Academic Press
Key Words: peptide; lipid; CD; DSC; fluorescence spectroscopy; DMPC.
INTRODUCTION
The fundamental role that lipid–peptide interactions play in a wide range of biological processes is acquiring growing experimental support. For example, there is increasing evidence in favor of a direct interaction of signal peptides and of mitochondrial presequences, with the membrane lipids playing a role at some stage of the membrane insertion or translocation process of precursor proteins (1, 2). Furthermore, lipid–peptide interactions have been implied to catalyze the binding of regulatory peptides to their receptor and to assist in the selection of the proper receptor subtype in the case of a peptide like opiod peptide (3). Nevertheless, detailed knowledge of the basic mechanisms underlying these interactions is lacking. Proteins are major constituents of biological membranes, and the biological activities of a variety of peptides and proteins are dependent on their ability to bind to cell membranes. It is
1 To whom correspondence should be addressed. E-mail: [email protected]. ernet.in.
0021-9797/01 $35.00
C 2001 by Academic Press Copyright ° All rights of reproduction in any form reserved.
24
EFFECT OF A DECAPEPTIDE ON THE PHASE TRANSITION OF A LIPID BILAYER
25
MATERIALS AND METHODS
The peptide was obtained as a gift from Professor S. Durani (IIT-Bombay, India). The purity of the peptide was checked by mass spectrometry and used for further studies. Dimyristoylphosphatidylcholine and 1,6-diphenyl-1,3,5hexatriene were obtained from Sigma Chemical Co. (U.S.A.) and used without purification. Triply distilled water and distilled solvents were used. The liposome was prepared by a conventional solvent evaporation method (6). Peptide can be incorporated by adding a measured amount of peptide to the lipid solution, before preparation of liposomes, in a molar ratio yielding a desired final lipid/peptide ratio. The concentrations of the probe were maintained at 1×10−6 M and that of the lipid at 1 × 10−4 M to yield the desired lipid to probe ratio of 100. Since the peptide is highly hydrophobic and sparingly soluble in water, it is expected that this is quantitatively transferred to the lipid phase. The experiments were performed at 0, 15, and 30 mol% of the peptide. Ultraviolet circular dichroism spectra were recorded by using a JASCO, J-720 spectropolarimeter. Fluorescence measurements were recorded in a Hitachi f-4500 spectrofluorometer. Corrected spectra were recorded with the slit widths of 5 nm/5 nm for all measurements. Appropriate background corrections with blank were done from the bilayer with and without the peptide, and the scattering characteristics were not altered upon addition of the peptide. The OD values are low (>0.05) at the excitation wavelength, thereby eliminating the possibility of any inner filter effect. Time-resolved fluorescence decay measurements were carried out by using a single-photon counting instrument. The excitation source was the third harmonic of NdYAG laser at 355 nm. The lifetime values are fitted between χ 2 values of 1.03 and 1.1. All calorimetric work was performed on a NETZSCH DSC 204 with a low-temperature accessory interfaced to a computer. DSC samples were prepared by codissolving the appropriate amounts of lipid and peptide in chloroform, evaporating the solvent under a stream of dry nitrogen, and then further drying in vacuum overnight. One to three milligrams of the dry sample was weighed into the DSC sample pan, an equal mass of buffer was added, and the pan was sealed (13). The sample was then mixed in the calorimeter by cycling the temperature from 273 to 310 K. The thermograms were recorded at a scan speed of 1.25 K/min. RESULTS AND DISCUSSIONS
Circular Dichroism Spectroscopy Figure 1 shows the CD spectra of the peptide in an aqueous medium and in the lipid bilayer at room temperature (25◦ C), at which the lipid bilayer membrane is in the liquid crystalline phase. The spectrum in water is characterized by a negative minimum below 200 nm, diagnostic of unordered conformation (14). The
FIG. 1. CD spectrum of the decapeptide in water and in the DMPC lipid bilayer at room temperature.
appearance of a double minimum at 210 and 222 nm clearly shows the formation of an α-helix (14, 15) in the lipid bilayer membrane. This conformational change can be attributed to the binding of the peptide with the membrane. The peptide-induced changes of the gel–liquid crystalline phase transition in the DMPC membrane were studied by using DSC and DPH fluorescence. However, while using DPH as a probe, we felt that it was necessary to ensure that specific interactions between the peptide and DPH were absent in the membrane and that the DPH fluorescence reflected changes only as a function of the membrane physical state. DPH Fluorescence in Membrane The decapeptide has no characteristic absorption and emission spectral properties, particularly in the region of DPH. Hence the spectral properties of DPH are not affected by the peptide, and DPH can be conveniently used as a probe to monitor the peptideinduced changes on the membrane. DPH is known to incorporate spontaneously into the hydrocarbon region of the membrane (16). Figure 2 shows the emission spectra of DPH in water and in the peptide. DPH in water is very weakly fluorescent. This week emission is neither enhanced nor quenched by the peptide, showing that there is no possible interaction between the peptide and DPH in this medium. Figure 3 shows the emission spectrum of DPH incorporated into the DMPC membrane in the presence and absence of peptide. Increasing amounts of peptide did not cause any significant changes in the emission maximum and fluorescence intensity of DPH in the membrane. This indicates that the incorporations of peptide and DPH are independent processes. The possibility of expulsion of DPH by the peptide is discounted, as it would have resulted in a loss of fluorescence intensity.The absence of any specific interactions between DPH and the peptide in the membrane can be further checked with
26
SHOBINI AND MISHRA
(K sv ) were found to be 0.12 M−1 for DMPC and 0.11 and 0.10 M−1 for 15 and 30 mol% of the peptide respectively. These values are fairly low as expected, indicating inaccessibility of the quencher to the probe. If the peptides were causing any displacement of DPH from the core membrane region toward quencher-accessible interfaces, there should have been an increase in K sv , which was not observed. The lack of fluorescence intensity reduction, the constancy in the partition coefficient values, and the lifetime data and the invariance of K sv clearly show that the membrane-incorporated DPH is unaffected by the presence of peptide, and DPH fluorescence anisotropy is expected to report only the lipid-related changes of the bilayer. DPH Anisotropy in the Membrane FIG. 2. Emission spectrum of DPH in water in the presence and absence of peptide: [DPH] = 1 × 10−6 M, λex = 355 nm.
the partition coefficient of DPH. This was done by using the fluorescence technique of Zhijian and Haugland (17). The slope of the linear double-reciprocal plots in the presence and absence of peptide gave K p values for DPH of 1.3 × 106 (±0.05) in the absence of peptide, and 1.25 × 106 (±0.05) and 1.2 × 106 (±0.07) for 15 and 30 mol% of peptide, respectively. As seen from the values, there is no significant change in the K p value. The fluorescence decay of DPH in the membrane was monoexponential with almost similar lifetimes of 8.4 (±0.4 ns) in pure DMPC and 8.7 (±0.4 ns) for 15 and 30 mol% of the peptide in the DMPC membrane. Thus DPH-reported anisotropy changes faithfully reflected fluidity changes in the membrane. A further proof of this independence of DPH and peptide incorporation comes from the fluorescence-quenching studies using I− as a hydrophilic quencher and using the Stern–Volmer quenching method (18). The Stern–Volmer quenching constants
DPH is known to partition equally well into solid or fluid lipid domains (19); hence the microviscosity obtainable from the anisotropy is known to reflect the weight average of all lipid domains (16). Although conversion of the anisotropy parameter to the microviscosity parameter by using the Perrin equation and its modifications for DPH in membrane is known to be erroneous, variation of the anisotropy parameter itself with membrane property changes provides a useful insight into the physical properties of the lipid bilayer membrane (18). In the present work we have reported the variation of the anisotropy as an indicator of the membrane fluidity change without attempting to convert this parameter to microviscosity. The average steady-state anisotropy value of the DPH fluorescence in the DMPC membrane is 0.30 (0.29; Shinitzky et al. (12)) in the gel state and 0.13 (0.15; Shinitzky et al. (12)) in the liquid crystalline state. The higher value in the gel state reflects the lower membrane fluidity expected from the state. When DPH is added to DMPC liposomes containing different amounts of the peptide at 27◦ C, a progressive and marked increase in the anisotropy is observed (Fig. 4). The value of the
FIG. 3. Emission spectrum of DPH in water and in the DMPC lipid bilayer in the presence and absence of peptide: lipid/probe = 1 : 100, λex = 355 nm.
FIG. 4. Average anisotropy of DPH in the DMPC lipid bilayer with increasing amounts of the peptide at room temperature: λex = 355 nm, λem = 435 nm.
27
EFFECT OF A DECAPEPTIDE ON THE PHASE TRANSITION OF A LIPID BILAYER
30 mol% of peptide appears to be close to the anisotropy of DPH in the gel state. The gel-to-liquid crystalline phase transition for DMPC is at 23◦ C (6). Thus a shift in the phase transition behavior with DMPC incorporation is expected. The effect of temperature on the system is discussed below. Temperature-Dependent Studies Differential scanning calorimetry (DSC). DSC is used to study the thermotropic aspects of the peptide–lipid interaction by examining the change in the melting point, the enthalpy, and the shape of the DSC trace. During heating the lipid initially undergoes a pretransition from an ordered gel state, where the lipids are ordered and tilted (Lβ 0 state), to the Pβ 0 , where the lipids are still ordered and in a gel state but the tilt is minimal. The pretransition occurs before the main transition and is small compared to the main one. During the main transition lipids undergo transition from the ordered Pβ 0 to a fluid disordered state Lα (20). Figure 5 shows the DSC thermograms of the DMPC lipid bilayer in the presence and absence of the decapeptide. The DSC data show that the peptide broadens the pretransition at 15 mol% and completely abolishes it at 30 mol% of peptide. There is a gradual shift of the main transition temperature from 23 to 26◦ C at 15 mol% and 29◦ C at 30 mol% of peptide with decreased height and increased half width. The changes in enthalpy and the peak half heights for DMPC as lipid bilayer in the presence and absence of peptide are given in Table 1. There is a marginal increase in both the enthalpies and the half width of the DSC curves with increasing incorporation of the peptide. These results possibly indicate that the association of the peptide with the lipid occurs more readily in the gel state than in the liquid crystalline state; i.e., the peptide increases the order of the lipid. While DSC gives broad changes in the membrane structural organization, with phase transition, the peptide-induced
FIG. 5. DSC thermogram of DMPC lipid bilayer in the presence and absence of the peptide.
TABLE 1 Changes in Enthalpy and Peak Half Width Accompanying the Phase Transition of the DMPC Lipid Bilayer in the Presence and Absence of the Peptide Mol% of peptide
1H (kJ mol−1 )
Peak half width (◦ C)
0 15 30
26.0 27.3 29.1
0.45 0.53 0.93
microenvironmental changes can be monitored by using fluorescence anisotropy of DPH. The ability of DPH in reporting Tc is well known (6, 12, 16, 19). Temperature-dependent studies of DPH anisotropy. The variation of anisotropy with temperature is often reported as plots of ln(r) vs 1/T, as such plots relate to changes in the flow activation energy (12). These plots are given in Figs. 6a and 6b. There is a small and progressive decrease in the anisotropy in the pure gel state, from 0 to 30 mol% of the peptide, showing a marginal decrease in the fluidity caused by the peptide. In the pure liquid crystalline state, there is no change in anisotropy until 15 mol% of the peptide, and a marginal decrease for 30 mol% of the peptide. A precise value of the phase-transition temperature can be obtained from the derivative plot as shown in Fig. 6b. This plot faithfully reproduces the shift in Tc as obtained from the DSC studies. It is also seen that the progressive blurring of the phase-transition behavior as observed in DSC studies is also reflected in Fig. 6b. Probable Location of the Peptide in the Bilayer Peptide has the effect of increasing the stability of the gel state at higher temperatures. Figure 7 shows the helical wheel
FIG. 6. (a) Plot of ln (r) vs 1/T for DPH in DMPC lipid bilayer in the presence and absence of peptide and (b) derivative plot of ln (r) vs 1/T: lipid/ probe = 1 : 100, λex = 355 nm, λem = 435nm.
28
SHOBINI AND MISHRA
FIG. 7. (a) Helical wheel representation the decapeptide. (b) Ribbon diagram of the decapeptide obtained through homology modeling.
diagram and the ribbon representation of the helical structure of the peptide. The helical structure was built by homology modeling based on the sequence homology between the decapeptide and the sequences of some of the proteins whose segments matchs those of the decapeptide (results communicated). The hydrophobic segment, which comprises the helix, could span the hydrophobic region of the membrane while the hydrophilic residues could project outside the bilayer as shown in Fig. 8. The length of the hydrophobic region of the peptide is around ˚ and the tail part is around 9 A. ˚ The thickness of the 30 A, ˚ (21). hydrophobic segment of the DMPC bilayer is around 29 A The spacing between the phosphorous atom in the lipid head˚ (22). group and the first C atom of the acyl chain is about 5 A
Zhang et al. (23) reported mismatch-dependent effects on the upward shift of the main transition temperature of PC bilayers with artificial, single membrane-spanning peptides. Killian (24) has reported the possibilities of this hydrophobic mismatch of the phase transition of membranes using many systems. Most possibly a similar line of argument can be extended to the present system. The overall hydrophobicity of the peptide in its helical form results in the facile incorporation of the peptide in the membrane, and the hydrophobic mismatch could cause inhomogeneities in the lateral distribution of peptide molecules and modify the temperature range of the coexistence of lipid phases. This could form a possible explanation for both the upward shift in Tc and the broadening of the phase-transition profile. This behavior of the decapeptide contrasts with that of a small synthetic tripeptide LFV (25), which has been shown to occupy a domain in the neck region in the DMPC lipid bilayer membrane, similar to cholesterol, which causes a broadening of Tc with no shift. SUMMARY
The decapeptide undergoes a conformational transition from an unordered state in water to a more ordered α-helical state in the lipid bilayer. Peptide incorporation into the membrane leads to changes in the phase-transition property. Extrinsic fluorescence probing studies using the fluidity probe DPH show that the peptide is not displacing the probe; hence the location of the peptide in the bilayer is in a region different from that of the probe. Temperature-dependent studies such as DSC and steadystate fluorescence anisotropy show a gradual upward shift in the phase-transition temperature. This shows that the peptide stabilizes the gel phase more than the liquid crystalline phase. This preferential stabilization of the gel phase could be due to the hydrophobic mismatch. ACKNOWLEDGMENTS The authors thank Professor S. Durani, IIT-Bombay, for the peptide, Professor P. Balram, IISc-Bangalore, for the CD spectra, the National Center for Ultrafast Processes, University of Madras, for the lifetime measurements, and Dr. N. Chandra, Bioinformatics Center, IISc Bangalore, for the modeling studies. J.S. thanks CSIR New Delhi for the scholarship.
REFERENCES
FIG. 8. Schematic representation of the probable location of the peptide in the bilayer membrane. The shaded portion represents the helical portion of the peptide, and the bilayer is represented by the circle and the chains.
1. Briggs, M. S., and Gierasch, L. M., Adv. Protein Chem. 38, 109 (1986). 2. Roise, D., and Schatz, G., J. Biol. Chem. 263, 4509 (1988). 3. Sargent, D. F., Bean, J. W., and Schwyzer, R., Biophys. Chem. 31, 183 (1988). 4. Jain, M., and Zakim, D., Biochim. Biophys. Acta. 906, 33 (1987). 5. Epand, R. M., and Sturtevant, J. M., Biophys. Chem. 19, 355 (1984). 6. New, R. R. C., “Liposomes, a Practical Approach.” OUP, New York (1991). 7. Lasch, H. P., Schultz, C. P., and Naumann, D., Biophys. J. 75, 840 (1998). 8. Polverini, E., Casadio, R., and Neyroz Pand Masotti, L., Arch. Biochem. Biophys. 349, 225 (1998). 9. Bonev, B. B., Chan, W. C., Bycroft, B. W., Roberts, G. C., and Watts, A., Biochemistry 39, 11425 (2000).
EFFECT OF A DECAPEPTIDE ON THE PHASE TRANSITION OF A LIPID BILAYER 10. Grau, A., Ortiz, A., de Godos, A., and Gomez-Fernandez, J. C., Arch. Biochem. Biophys. 377, 315 (2000). 11. Altschul, S. F., Thomas, L. M., Alejandro, A. S., Jinghui, Z., Zheng, Z., Webb, M., and David, J. L., Nucleic Acids Res. 25, 3389 (1997). 12. Shinitzky, M., and Barenholz, Y., Biochim. Biophys. Acta 515, 367 (1978). 13. Jacobs, R. E., and White, S. H., Biochemistry 25, 2605 (1986). 14. Brahms, S., and Brahms, J., J. Mol. Biol. 138, 149 (1980). 15. Simonetti, M., Falcigno, L., Paolillo, L., and Bello, C. D., Biopolymers 41, 461 (1997). 16. Cranney, M., Robert, B., Jones, C. G. R., Richards, J. T., and Thomas, E. W., Biochim. Biophys. Acta 735, 418 (1983). 17. Zhijian, H., and Haugland, R. P., Biochem. Biophys. Res. Commun. 181, 166 (1991).
29
18. Lackowicz, J. R., “Principles of Fluorescence Spectroscopy.” Elsevier, New York (1983). 19. Lentz, R. B., Barenholz, Y., and Thompson, T. E., Biochemistry 15, 4521 (1976). 20. Small, D. M., and Hanahan, D. J. (Eds.), “Handbook of Lipid Research,” Vol. 4, pp. 475–522. Plenum Press, New York (1986). 21. Tahara, Y., Masaykki, M., Ohnishi, S., Fujiyoshi, Y., Kikuchi, M., and Yamamoto, Y., Biochemistry 31, 8747 (1992). 22. Pearson, R. H., and Pascher, I., Nature 281, 499 (1979). 23. Zhang, Y. P., Lewis, R. N., Hodges, R. S., and McElhomey, M. C., Biochemistry 31, 11579 (1992). 24. Killian, J. A., Biochim. Biophys. Acta 1376, 401 (1998). 25. Shobini, J., and Mishra, A. K., Spectrochem. Acta A 56, 2239 (2000).
Effect of a Decapeptide (VPDLLADLLK) on the Phase Transition of Dimyristoylphosphatidylcholine Lipid Bilayer J. Shobini and A. K. Mishra1 Department of Chemistry, Indian Institute of Technology Madras, Chennai 600 036, India Received June 14, 2000; accepted April 21, 2001
therefore interesting to study the nature of interactions between peptides and lipids in model systems. The study of model systems consisting of peptides, either chemically synthesized or naturally occurring, interacting with the bilayers has proven to be a valuable contribution to a better understanding of these mechanisms (4). The synthetic peptide approach has the advantage of being flexible: the effects of small changes in the molecule on the lipid–peptide interactions can be determined. The ability of a substance to alter the phase-transition characteristics of a phospholipid model membrane is indicative of its association with the lipid either by some simple method or by more complicated means (5). An understanding of phase transitions and the fluidity of the phospholipid membrane is important in both the manufacture and exploitation of liposomes, since the phase behavior of a liposome membrane determines such properties as permeability, fusion, aggregation, and protein binding, all of which can markedly affect the stability of liposomes and their behavior in biological systems (6). Several peptides such as (synthetic) poly-(L-lysine) (7), neuropeptides such as neuromidicin B (8), antibacterial peptides such as nisin (9), and an antibiotic peptide iturin A (10) are known to interact differently with the dimyristoylphosphatidylcholine (DMPC) lipid bilayer. They show a wide variety of changes in the physical properties of the membrane, such as fluidity, permeability, and phase-transition temperature. Some peptides, such as nisin, also showed a perturbing effect on the hydrophobic region of the bilayer (9). In this work we report for the first time the effect of a synthetic decapeptide (VPDLLADLLK) on the phase-transition behavior of DMPC liposome membrane. This is a synthetic peptide, having hydrophobic and hydrophilic amino acid residues. A BLAST search (11) of this sequence produces 114 matches having significant alignments (60–80%) with many proteins (many of them are enzymes). Circular dichroism spectroscopy has been used to monitor the conformational changes of the peptide; differential scanning calorimetry and fluorescence spectroscopy using 1,6diphenyl-1,3,5-hexatriene (DPH) as an extrinsic fluorescence probe have been used to monitor the phase-transition behavior of the lipid bilayer. The fluorescence anisotropy of DPH in the lipid bilayer membrane has been extensively used in literature to study membrane-related changes, primarily as a fluidity probe (6, 12).
VPDLLADLLK is a synthetic decapeptide, which shows a difference in conformation in various environments. Circular dichroism spectral studies show that it exists in an unordered conformation in the aqueous phase, and in dimyristoylphosphatidylcholine (DMPC) lipid bilayer, it exhibits an α-helical structure. The membrane property modification due to the peptide incorporation has been studied by using differential scanning calorimetry and fluorescence spectroscopy. With incorporation of the peptide the average steady-state anisotropy of DPH in the membrane decreases slightly in the gel state but remains more or less the same in the liquid crystalline state. The peptide incorporation causes a shift in the phase-transition temperature from 23 to 26◦ C for 15 mol% and 29◦ C for 30 mol% of the peptide, which is accompanied by a decrease in the sharpness and a broadening of the DSC thermogram. This preferential stabilization of the more ordered gel phase by the peptide could be due to the hydrophobic mismatch between the length of the peptide and the length of the hydrophobic segment of the DMPC bilayer. ° C 2001
Academic Press
Key Words: peptide; lipid; CD; DSC; fluorescence spectroscopy; DMPC.
INTRODUCTION
The fundamental role that lipid–peptide interactions play in a wide range of biological processes is acquiring growing experimental support. For example, there is increasing evidence in favor of a direct interaction of signal peptides and of mitochondrial presequences, with the membrane lipids playing a role at some stage of the membrane insertion or translocation process of precursor proteins (1, 2). Furthermore, lipid–peptide interactions have been implied to catalyze the binding of regulatory peptides to their receptor and to assist in the selection of the proper receptor subtype in the case of a peptide like opiod peptide (3). Nevertheless, detailed knowledge of the basic mechanisms underlying these interactions is lacking. Proteins are major constituents of biological membranes, and the biological activities of a variety of peptides and proteins are dependent on their ability to bind to cell membranes. It is
1 To whom correspondence should be addressed. E-mail: [email protected]. ernet.in.
0021-9797/01 $35.00
C 2001 by Academic Press Copyright ° All rights of reproduction in any form reserved.
24
EFFECT OF A DECAPEPTIDE ON THE PHASE TRANSITION OF A LIPID BILAYER
25
MATERIALS AND METHODS
The peptide was obtained as a gift from Professor S. Durani (IIT-Bombay, India). The purity of the peptide was checked by mass spectrometry and used for further studies. Dimyristoylphosphatidylcholine and 1,6-diphenyl-1,3,5hexatriene were obtained from Sigma Chemical Co. (U.S.A.) and used without purification. Triply distilled water and distilled solvents were used. The liposome was prepared by a conventional solvent evaporation method (6). Peptide can be incorporated by adding a measured amount of peptide to the lipid solution, before preparation of liposomes, in a molar ratio yielding a desired final lipid/peptide ratio. The concentrations of the probe were maintained at 1×10−6 M and that of the lipid at 1 × 10−4 M to yield the desired lipid to probe ratio of 100. Since the peptide is highly hydrophobic and sparingly soluble in water, it is expected that this is quantitatively transferred to the lipid phase. The experiments were performed at 0, 15, and 30 mol% of the peptide. Ultraviolet circular dichroism spectra were recorded by using a JASCO, J-720 spectropolarimeter. Fluorescence measurements were recorded in a Hitachi f-4500 spectrofluorometer. Corrected spectra were recorded with the slit widths of 5 nm/5 nm for all measurements. Appropriate background corrections with blank were done from the bilayer with and without the peptide, and the scattering characteristics were not altered upon addition of the peptide. The OD values are low (>0.05) at the excitation wavelength, thereby eliminating the possibility of any inner filter effect. Time-resolved fluorescence decay measurements were carried out by using a single-photon counting instrument. The excitation source was the third harmonic of NdYAG laser at 355 nm. The lifetime values are fitted between χ 2 values of 1.03 and 1.1. All calorimetric work was performed on a NETZSCH DSC 204 with a low-temperature accessory interfaced to a computer. DSC samples were prepared by codissolving the appropriate amounts of lipid and peptide in chloroform, evaporating the solvent under a stream of dry nitrogen, and then further drying in vacuum overnight. One to three milligrams of the dry sample was weighed into the DSC sample pan, an equal mass of buffer was added, and the pan was sealed (13). The sample was then mixed in the calorimeter by cycling the temperature from 273 to 310 K. The thermograms were recorded at a scan speed of 1.25 K/min. RESULTS AND DISCUSSIONS
Circular Dichroism Spectroscopy Figure 1 shows the CD spectra of the peptide in an aqueous medium and in the lipid bilayer at room temperature (25◦ C), at which the lipid bilayer membrane is in the liquid crystalline phase. The spectrum in water is characterized by a negative minimum below 200 nm, diagnostic of unordered conformation (14). The
FIG. 1. CD spectrum of the decapeptide in water and in the DMPC lipid bilayer at room temperature.
appearance of a double minimum at 210 and 222 nm clearly shows the formation of an α-helix (14, 15) in the lipid bilayer membrane. This conformational change can be attributed to the binding of the peptide with the membrane. The peptide-induced changes of the gel–liquid crystalline phase transition in the DMPC membrane were studied by using DSC and DPH fluorescence. However, while using DPH as a probe, we felt that it was necessary to ensure that specific interactions between the peptide and DPH were absent in the membrane and that the DPH fluorescence reflected changes only as a function of the membrane physical state. DPH Fluorescence in Membrane The decapeptide has no characteristic absorption and emission spectral properties, particularly in the region of DPH. Hence the spectral properties of DPH are not affected by the peptide, and DPH can be conveniently used as a probe to monitor the peptideinduced changes on the membrane. DPH is known to incorporate spontaneously into the hydrocarbon region of the membrane (16). Figure 2 shows the emission spectra of DPH in water and in the peptide. DPH in water is very weakly fluorescent. This week emission is neither enhanced nor quenched by the peptide, showing that there is no possible interaction between the peptide and DPH in this medium. Figure 3 shows the emission spectrum of DPH incorporated into the DMPC membrane in the presence and absence of peptide. Increasing amounts of peptide did not cause any significant changes in the emission maximum and fluorescence intensity of DPH in the membrane. This indicates that the incorporations of peptide and DPH are independent processes. The possibility of expulsion of DPH by the peptide is discounted, as it would have resulted in a loss of fluorescence intensity.The absence of any specific interactions between DPH and the peptide in the membrane can be further checked with
26
SHOBINI AND MISHRA
(K sv ) were found to be 0.12 M−1 for DMPC and 0.11 and 0.10 M−1 for 15 and 30 mol% of the peptide respectively. These values are fairly low as expected, indicating inaccessibility of the quencher to the probe. If the peptides were causing any displacement of DPH from the core membrane region toward quencher-accessible interfaces, there should have been an increase in K sv , which was not observed. The lack of fluorescence intensity reduction, the constancy in the partition coefficient values, and the lifetime data and the invariance of K sv clearly show that the membrane-incorporated DPH is unaffected by the presence of peptide, and DPH fluorescence anisotropy is expected to report only the lipid-related changes of the bilayer. DPH Anisotropy in the Membrane FIG. 2. Emission spectrum of DPH in water in the presence and absence of peptide: [DPH] = 1 × 10−6 M, λex = 355 nm.
the partition coefficient of DPH. This was done by using the fluorescence technique of Zhijian and Haugland (17). The slope of the linear double-reciprocal plots in the presence and absence of peptide gave K p values for DPH of 1.3 × 106 (±0.05) in the absence of peptide, and 1.25 × 106 (±0.05) and 1.2 × 106 (±0.07) for 15 and 30 mol% of peptide, respectively. As seen from the values, there is no significant change in the K p value. The fluorescence decay of DPH in the membrane was monoexponential with almost similar lifetimes of 8.4 (±0.4 ns) in pure DMPC and 8.7 (±0.4 ns) for 15 and 30 mol% of the peptide in the DMPC membrane. Thus DPH-reported anisotropy changes faithfully reflected fluidity changes in the membrane. A further proof of this independence of DPH and peptide incorporation comes from the fluorescence-quenching studies using I− as a hydrophilic quencher and using the Stern–Volmer quenching method (18). The Stern–Volmer quenching constants
DPH is known to partition equally well into solid or fluid lipid domains (19); hence the microviscosity obtainable from the anisotropy is known to reflect the weight average of all lipid domains (16). Although conversion of the anisotropy parameter to the microviscosity parameter by using the Perrin equation and its modifications for DPH in membrane is known to be erroneous, variation of the anisotropy parameter itself with membrane property changes provides a useful insight into the physical properties of the lipid bilayer membrane (18). In the present work we have reported the variation of the anisotropy as an indicator of the membrane fluidity change without attempting to convert this parameter to microviscosity. The average steady-state anisotropy value of the DPH fluorescence in the DMPC membrane is 0.30 (0.29; Shinitzky et al. (12)) in the gel state and 0.13 (0.15; Shinitzky et al. (12)) in the liquid crystalline state. The higher value in the gel state reflects the lower membrane fluidity expected from the state. When DPH is added to DMPC liposomes containing different amounts of the peptide at 27◦ C, a progressive and marked increase in the anisotropy is observed (Fig. 4). The value of the
FIG. 3. Emission spectrum of DPH in water and in the DMPC lipid bilayer in the presence and absence of peptide: lipid/probe = 1 : 100, λex = 355 nm.
FIG. 4. Average anisotropy of DPH in the DMPC lipid bilayer with increasing amounts of the peptide at room temperature: λex = 355 nm, λem = 435 nm.
27
EFFECT OF A DECAPEPTIDE ON THE PHASE TRANSITION OF A LIPID BILAYER
30 mol% of peptide appears to be close to the anisotropy of DPH in the gel state. The gel-to-liquid crystalline phase transition for DMPC is at 23◦ C (6). Thus a shift in the phase transition behavior with DMPC incorporation is expected. The effect of temperature on the system is discussed below. Temperature-Dependent Studies Differential scanning calorimetry (DSC). DSC is used to study the thermotropic aspects of the peptide–lipid interaction by examining the change in the melting point, the enthalpy, and the shape of the DSC trace. During heating the lipid initially undergoes a pretransition from an ordered gel state, where the lipids are ordered and tilted (Lβ 0 state), to the Pβ 0 , where the lipids are still ordered and in a gel state but the tilt is minimal. The pretransition occurs before the main transition and is small compared to the main one. During the main transition lipids undergo transition from the ordered Pβ 0 to a fluid disordered state Lα (20). Figure 5 shows the DSC thermograms of the DMPC lipid bilayer in the presence and absence of the decapeptide. The DSC data show that the peptide broadens the pretransition at 15 mol% and completely abolishes it at 30 mol% of peptide. There is a gradual shift of the main transition temperature from 23 to 26◦ C at 15 mol% and 29◦ C at 30 mol% of peptide with decreased height and increased half width. The changes in enthalpy and the peak half heights for DMPC as lipid bilayer in the presence and absence of peptide are given in Table 1. There is a marginal increase in both the enthalpies and the half width of the DSC curves with increasing incorporation of the peptide. These results possibly indicate that the association of the peptide with the lipid occurs more readily in the gel state than in the liquid crystalline state; i.e., the peptide increases the order of the lipid. While DSC gives broad changes in the membrane structural organization, with phase transition, the peptide-induced
FIG. 5. DSC thermogram of DMPC lipid bilayer in the presence and absence of the peptide.
TABLE 1 Changes in Enthalpy and Peak Half Width Accompanying the Phase Transition of the DMPC Lipid Bilayer in the Presence and Absence of the Peptide Mol% of peptide
1H (kJ mol−1 )
Peak half width (◦ C)
0 15 30
26.0 27.3 29.1
0.45 0.53 0.93
microenvironmental changes can be monitored by using fluorescence anisotropy of DPH. The ability of DPH in reporting Tc is well known (6, 12, 16, 19). Temperature-dependent studies of DPH anisotropy. The variation of anisotropy with temperature is often reported as plots of ln(r) vs 1/T, as such plots relate to changes in the flow activation energy (12). These plots are given in Figs. 6a and 6b. There is a small and progressive decrease in the anisotropy in the pure gel state, from 0 to 30 mol% of the peptide, showing a marginal decrease in the fluidity caused by the peptide. In the pure liquid crystalline state, there is no change in anisotropy until 15 mol% of the peptide, and a marginal decrease for 30 mol% of the peptide. A precise value of the phase-transition temperature can be obtained from the derivative plot as shown in Fig. 6b. This plot faithfully reproduces the shift in Tc as obtained from the DSC studies. It is also seen that the progressive blurring of the phase-transition behavior as observed in DSC studies is also reflected in Fig. 6b. Probable Location of the Peptide in the Bilayer Peptide has the effect of increasing the stability of the gel state at higher temperatures. Figure 7 shows the helical wheel
FIG. 6. (a) Plot of ln (r) vs 1/T for DPH in DMPC lipid bilayer in the presence and absence of peptide and (b) derivative plot of ln (r) vs 1/T: lipid/ probe = 1 : 100, λex = 355 nm, λem = 435nm.
28
SHOBINI AND MISHRA
FIG. 7. (a) Helical wheel representation the decapeptide. (b) Ribbon diagram of the decapeptide obtained through homology modeling.
diagram and the ribbon representation of the helical structure of the peptide. The helical structure was built by homology modeling based on the sequence homology between the decapeptide and the sequences of some of the proteins whose segments matchs those of the decapeptide (results communicated). The hydrophobic segment, which comprises the helix, could span the hydrophobic region of the membrane while the hydrophilic residues could project outside the bilayer as shown in Fig. 8. The length of the hydrophobic region of the peptide is around ˚ and the tail part is around 9 A. ˚ The thickness of the 30 A, ˚ (21). hydrophobic segment of the DMPC bilayer is around 29 A The spacing between the phosphorous atom in the lipid head˚ (22). group and the first C atom of the acyl chain is about 5 A
Zhang et al. (23) reported mismatch-dependent effects on the upward shift of the main transition temperature of PC bilayers with artificial, single membrane-spanning peptides. Killian (24) has reported the possibilities of this hydrophobic mismatch of the phase transition of membranes using many systems. Most possibly a similar line of argument can be extended to the present system. The overall hydrophobicity of the peptide in its helical form results in the facile incorporation of the peptide in the membrane, and the hydrophobic mismatch could cause inhomogeneities in the lateral distribution of peptide molecules and modify the temperature range of the coexistence of lipid phases. This could form a possible explanation for both the upward shift in Tc and the broadening of the phase-transition profile. This behavior of the decapeptide contrasts with that of a small synthetic tripeptide LFV (25), which has been shown to occupy a domain in the neck region in the DMPC lipid bilayer membrane, similar to cholesterol, which causes a broadening of Tc with no shift. SUMMARY
The decapeptide undergoes a conformational transition from an unordered state in water to a more ordered α-helical state in the lipid bilayer. Peptide incorporation into the membrane leads to changes in the phase-transition property. Extrinsic fluorescence probing studies using the fluidity probe DPH show that the peptide is not displacing the probe; hence the location of the peptide in the bilayer is in a region different from that of the probe. Temperature-dependent studies such as DSC and steadystate fluorescence anisotropy show a gradual upward shift in the phase-transition temperature. This shows that the peptide stabilizes the gel phase more than the liquid crystalline phase. This preferential stabilization of the gel phase could be due to the hydrophobic mismatch. ACKNOWLEDGMENTS The authors thank Professor S. Durani, IIT-Bombay, for the peptide, Professor P. Balram, IISc-Bangalore, for the CD spectra, the National Center for Ultrafast Processes, University of Madras, for the lifetime measurements, and Dr. N. Chandra, Bioinformatics Center, IISc Bangalore, for the modeling studies. J.S. thanks CSIR New Delhi for the scholarship.
REFERENCES
FIG. 8. Schematic representation of the probable location of the peptide in the bilayer membrane. The shaded portion represents the helical portion of the peptide, and the bilayer is represented by the circle and the chains.
1. Briggs, M. S., and Gierasch, L. M., Adv. Protein Chem. 38, 109 (1986). 2. Roise, D., and Schatz, G., J. Biol. Chem. 263, 4509 (1988). 3. Sargent, D. F., Bean, J. W., and Schwyzer, R., Biophys. Chem. 31, 183 (1988). 4. Jain, M., and Zakim, D., Biochim. Biophys. Acta. 906, 33 (1987). 5. Epand, R. M., and Sturtevant, J. M., Biophys. Chem. 19, 355 (1984). 6. New, R. R. C., “Liposomes, a Practical Approach.” OUP, New York (1991). 7. Lasch, H. P., Schultz, C. P., and Naumann, D., Biophys. J. 75, 840 (1998). 8. Polverini, E., Casadio, R., and Neyroz Pand Masotti, L., Arch. Biochem. Biophys. 349, 225 (1998). 9. Bonev, B. B., Chan, W. C., Bycroft, B. W., Roberts, G. C., and Watts, A., Biochemistry 39, 11425 (2000).
EFFECT OF A DECAPEPTIDE ON THE PHASE TRANSITION OF A LIPID BILAYER 10. Grau, A., Ortiz, A., de Godos, A., and Gomez-Fernandez, J. C., Arch. Biochem. Biophys. 377, 315 (2000). 11. Altschul, S. F., Thomas, L. M., Alejandro, A. S., Jinghui, Z., Zheng, Z., Webb, M., and David, J. L., Nucleic Acids Res. 25, 3389 (1997). 12. Shinitzky, M., and Barenholz, Y., Biochim. Biophys. Acta 515, 367 (1978). 13. Jacobs, R. E., and White, S. H., Biochemistry 25, 2605 (1986). 14. Brahms, S., and Brahms, J., J. Mol. Biol. 138, 149 (1980). 15. Simonetti, M., Falcigno, L., Paolillo, L., and Bello, C. D., Biopolymers 41, 461 (1997). 16. Cranney, M., Robert, B., Jones, C. G. R., Richards, J. T., and Thomas, E. W., Biochim. Biophys. Acta 735, 418 (1983). 17. Zhijian, H., and Haugland, R. P., Biochem. Biophys. Res. Commun. 181, 166 (1991).
29
18. Lackowicz, J. R., “Principles of Fluorescence Spectroscopy.” Elsevier, New York (1983). 19. Lentz, R. B., Barenholz, Y., and Thompson, T. E., Biochemistry 15, 4521 (1976). 20. Small, D. M., and Hanahan, D. J. (Eds.), “Handbook of Lipid Research,” Vol. 4, pp. 475–522. Plenum Press, New York (1986). 21. Tahara, Y., Masaykki, M., Ohnishi, S., Fujiyoshi, Y., Kikuchi, M., and Yamamoto, Y., Biochemistry 31, 8747 (1992). 22. Pearson, R. H., and Pascher, I., Nature 281, 499 (1979). 23. Zhang, Y. P., Lewis, R. N., Hodges, R. S., and McElhomey, M. C., Biochemistry 31, 11579 (1992). 24. Killian, J. A., Biochim. Biophys. Acta 1376, 401 (1998). 25. Shobini, J., and Mishra, A. K., Spectrochem. Acta A 56, 2239 (2000).