Dependence of Purple Membrane Bump Curvature on pH and Ionic

Jul 14, 2014 - Masashi Sonoyama,. †,|| and Shigeki Mitaku. †,⊥. †. Department of Applied Physics, Graduate School of Engineering, Nagoya Unive...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/JPCB

Dependence of Purple Membrane Bump Curvature on pH and Ionic Strength Analyzed Using Atomic Force Microscopy Combined with Solvent Exchange Yasunori Yokoyama,*,† Kosuke Yamada,† Yosuke Higashi,† Satoshi Ozaki,† Haorang Wang,† Naoki Koito,† Naoya Watanabe,‡,§ Masashi Sonoyama,†,|| and Shigeki Mitaku†,⊥ †

Department of Applied Physics, Graduate School of Engineering, Nagoya University, Nagoya 464-8603, Japan SII NanoTechnology, Inc., Matsudo 270-2222, Japan



S Supporting Information *

ABSTRACT: Purple membrane (PM), which is a membrane patch formed by the self-assembly of the membrane protein bacteriorhodopsin (bR) with archaeal lipids, is a good subject for studying the mechanism for the supramolecular structural formation of membrane proteins. Several studies have suggested that PM is not simply planar but that it has a curvature. Atomic force microscopy (AFM) studies also indicate the presence of dome-like structures (bumps) on the cytoplasmic surface of PM. PM must have a curvature to form the bump structures; therefore, bump formations will be related to a mechanism for supramolecular structural formation via self-assembly. To elucidate the effect of an asymmetric distribution of charged residues between two aqueous domains on the bump curvature, AFM topography of identical PM sheets were examined with variation of the solvent ionic strength and pH using a newly constructed solvent circulation system. The radius and height distributions of the bumps on the identical PM sheets indicated a linear correlation. The bump curvature, which was simply estimated by the slope of the distribution, became smaller with increasing KCl concentration, which suggests that tension at the cytoplasmic surface caused by electrostatic repulsive force between negatively charged amino acid residues becomes weaker by the electrostatic shielding effect. AFM observations revealed that the bump curvature remained even at high KCl concentration where the Debye length is within a few Angstroms; therefore, the contribution of the intrinsic difference between the domain sizes of bR between two sides was confirmed. Interestingly, the bump curvature was significantly increased by the addition of CaCl2 and then decreased with a similar dependency to KCl at higher CaCl2 concentration. The effect of pH on the bump curvature was also examined, where the curvature increased and reached a maximum at pH 9, while it decreased above pH 10, at which point the two-dimensional crystalline lattice of bR began to disassemble. These experimental results indicate that the bump curvature is strongly influenced by electrostatic interactions. A plausible model for bump structure formation by electrostatic repulsive force is presented based on these results.

1. INTRODUCTION Purple membrane (PM), which is found in the cytoplasmic membrane of halophilic archaebacteria, is a membrane patch formed by the self-assembly of membrane protein bacteriorhodopsin (bR) and archaeal lipids.1,2 bR molecules, which play a role as a light-driven proton pump, are abundantly expressed under anaerobic conditions and generate a proton-motive potential across the membrane using light energy, which is directly coupled to ATP synthesis from the light energy.2 bR molecules in the PM are spontaneously arranged in a hexagonal symmetry of bR trimers with identical orientation toward the membrane (N-terminus at the extracellular side) and form a two-dimensional (2D) crystalline array that ranges up to a few micrometers in the long axis length of the membrane patch.1,3 Although it is considered that the formation of bR complexes such as the bR trimer and the 2D crystal is not necessary for its proton pumping function,4 light-induced denaturation of bR © 2014 American Chemical Society

has been reported under conditions where the 2D crystalline array is also disassembled.5−12 This result suggests that the molecular complex formation contributes to the highly effective bR functionality. PM should be a good subject for studying the structural formation mechanism of a large complex of membrane proteins via self-assembly because there are several reports that imply PM is not simply planar but has a curvature. Henderson and coworkers reported that there are often crystalline distortions such as membrane undulation and that calculations are necessary to remove the distortions from the diffraction data to achieve a more precise structural determination.13 Kouyama et al. demonstrated that the treatment of PM with detergent Received: April 13, 2014 Revised: July 2, 2014 Published: July 14, 2014 9322

dx.doi.org/10.1021/jp5036234 | J. Phys. Chem. B 2014, 118, 9322−9328

The Journal of Physical Chemistry B

Article

causes inside-out vesicles with a diameter of 50 nm.14 Czégé and Reinisch reported that the curvature of PM changed transiently during its functional cycle with a pH dependent manner.15 Atomic force microscopy (AFM) examination of PM sheets supported on a mica substrate indicated that dome-like structures (bumps) were often observed on the membrane surface in solution.16 Both immunostaining techniques and pattern recognition revealed geometric differences in cracks between the cytoplasmic (CP) and extracellular (EC) surfaces of PM after drying and also revealed that the bump structures were present only on the CP surface.17 The PM sheet must have a curvature to form a bump structure; therefore, bump formations should be related to a mechanism for supramolecular structural formation via the self-assembly of membrane protein and lipids. There are two possible mechanisms for the bump formation of PM; a difference in the domain sizes of bR between the CP and EC sides and a difference in the electrostatic repulsion at the two sides because bR molecules are buried in the membrane with the same orientation. In the former case, PM should have a curvature if the cross-sectional area of the CP side of bR molecules is larger than those of the EC side. AFM observation of PM sheets consisting of mutant strains such as D85N and D85T, which assume the CP-open form above pH 9 even in darkness, showed that the PM sheet was globally bent at pH 9.7.18,19 However, in the latter mechanism, another possibility is a difference in the distribution of charged amino acid residues between two soluble domains of bR (the CP and EC sides). The distribution of charged amino acid residues of bR in PM is shown in Supporting Information Figure S1. Numbers of negatively charged amino acids are arranged in the CP soluble domain to acquire protons from the aqueous medium because bR molecules unidirectionally translocate protons from the CP to the EC side.20 It is also reported that the curvature of the PM sheet is decreased by cleaving the Cterminus peptide located in the CP soluble domain that has four negatively charged amino acid residues,21 which suggests some contribution of the electrostatic repulsive force at the CP soluble domain to form the PM curvature. However, there is little information regarding the balance of these factors that are related to curvature formation in PM. In this work, AFM observations of PM sheets were conducted under various ionic strength and pH conditions, assuming that the curvature of bumps is due to the domain size effect and that the effect of salt concentration and pH is due to the electrostatic effect. A solvent circulation system was constructed and combined with the AFM apparatus to observe the topographic changes of identical PM sheets upon change of the medium conditions. The topography of identical PM sheets was observed with incremental variation of the salt concentration using this system, and the effects of electrostatic shielding on the curvature of bumps were examined. The influence of solvent pH upon the bump structure was similarly elucidated. On the basis of these results, the effects of the ionic strength and pH of the solvent are discussed with respect to the bump structure formation mechanisms.

concentration of bR for AFM measurements was adjusted to approximately 10 μM. AFM observation of PM sheet in an aqueous environment was conducted using an AFM station (SII NanoTechnology Inc., Japan, SPI3800N) combined with a scanner (SII NanoTechnology Inc., Japan, SPA400). AFM topography was measured in contact mode (AFM mode) with a silicon nitride cantilever (Olympus Co., Japan TR400PSA-1; spring constant of 0.08 N/m). A freshly cleaved mica substrate was washed twice with 300 μL of imaging buffer (150 mM KCl, 20 mM phosphate buffer, pH 7.0), and then 300 μL of PM suspension was placed on the substrate. After 15 min of incubation, the excess PM suspension was removed, and the substrate was rinsed twice with 150 μL of the imaging buffer. The specimen was then mounted on a custom-made AFM solution cell holder that allows solution exchange and was then immersed in the imaging buffer. The AFM image in this work was taken just before the AFM tip separates from the approach area by setting maximum force reference value of the cantilever in the acceptable range, so that the AFM tip can trace the surface with very weak force without any sample damages. The radius and height distribution obtained in the present study was completely involved in that acquired by the tapping mode. The solvent circulation system for the AFM measurements consists of the AFM cell holder and two peristaltic pumps that are used to send (pull) the solvent to (from) the cell holder (Figure 1). The AFM topography of identical PM sheets under

Figure 1. AFM apparatus combined with a solvent circulation system to capture the topography of identical PM sheets with variation of the solvent conditions.

various solvent conditions can be obtained with this system because the medium solution can be exchanged without removal of the AFM tip away from the sample surface. The time for complete solvent exchange was determined to be 60 min by preliminary measurement of the solution pH increase with a pH indicator. AFM topography was measured 60 min after the solvent exchange, which was determined by the analyses against AFM topography after various length of elapsed time, for thermal equilibrium to be achieved. The KCl and CaCl2 salt concentrations were varied from 0.15 to 3 M and from 0.15 to 1 M, respectively. For measurements with CaCl2, 20 mM Tris buffer (pH 7.0) was used instead of phosphate buffer. AFM topography of identical PM sheets was obtained in the pH range from 7 to 12 using 20 mM universal pH buffer (containing boric acid, citric acid, and trisodium phosphate),23 to adjust the pH at each measurement condition. The image correction of the AFM topography was carried out

2. MATERIALS AND METHODS PM of Halobacterium salinarum, strain R1M1, was purified according to the standard method established by Oesterhelt and Stoeckenius22 and then resuspended in 20 mM phosphate buffer containing 150 mM KCl (pH 7.0). The final 9323

dx.doi.org/10.1021/jp5036234 | J. Phys. Chem. B 2014, 118, 9322−9328

The Journal of Physical Chemistry B

Article

using the measurement software (SII NanoTechnology Inc., Japan, NanoNavi II). All AFM images in this work were obtained by the tilt correction only, as shown in Supporting Information Figure S2. The number of bumps in each PM sheet ranged from 24 to 45.

3. RESULTS AND DISCUSSION To determine the effect of ionic strength and pH on the bump structures, AFM topography of identical PM sheets was observed using a solvent circulation system. Figure 2a shows the change in the AFM topography of identical PM sheets with the change in the KCl concentration. Several bump structures were observed on the PM sheets, as previously reported.16 The results shown in Figure 2a demonstrate that AFM topography can be obtained, even after multiple solvent exchanges, without damage from the AFM tip, because no detachment or rupture of the PM sheets occurred during the measurements. The bump structures did not disappear with an increase of the KCl concentration up to 3 M. The radius and height distributions of the bump structures on identical PM sheets were obtained by cross-sectional analyses of each bump, and the results are shown in Figure 2b. There is a linear correlation between the radius and height of the bump structures, which was also observed under other solvent conditions. These results suggest that the curvature of the bump structures was almost constant at each solvent condition. Figure 2c shows the dependence of curvature on the KCl concentration, which was simply estimated from the slope of the bump height vs bump radius obtained from identical PM sheets. The error bars represent the standard deviation of the linear correlation fitting. The curvature of the bump structure decreased with increasing KCl concentration, which suggests that the CP surface tension due to electrostatic repulsion is weakened by the electrostatic shielding effect under higher KCl concentration, at which the thickness of the diffuse electric double layer is decreased. The bump curvature remained at even high KCl concentration, at which the Debye length is within a few Angstroms, which suggests there is another factor that contributes to the bump formation of PM other than the electrostatic repulsive force. Figure 3 shows the dependence of the bump curvature on the ionic strength by the addition of KCl and CaCl2 to elucidate differences in the bump structures caused by cation valence. The large error range caused by the addition of CaCl2 suggests that the structural fluctuation of the PM sheet became larger and that the bump can assume various curvatures. Figure 3 shows that the bump curvature was significantly increased by the addition of a small amount of CaCl2, while the curvature was decreased with a similar dependence to that for KCl addition when the ionic strength became larger. This suggests that there is another effect of Ca2+ on the bump curvature besides the electrostatic shielding effect. In this context, the specific interaction between Ca2+ and the lipid headgroup was reported previously, where phase separation of the mixed phospholipid bilayer consisting of lipids with phosphatidylserine (acidic) and phosphatidylcholine (neutral) head groups occurred by the addition of 1−2 μM Ca2+.24 It was also reported that the interbilayer distance of dipalmitoylphosphatidylcholine (DPPC) multilamellar vesicles was significantly expanded by the addition of 1 mM CaCl 2 . 25 These experimental results are explained as the result of specific interaction between Ca2+ and the negatively charged headgroup of lipids. In the case of PM, it is also known that PM can bind various multivalent cations and that there are some specific

Figure 2. (a) AFM topography of identical PM sheets at pH 7 upon change in the KCl concentration from 150 mM to 3.0 M using the solvent circulation system. (b) Typical radius and height distributions of bump structures obtained with the solvent circulation system. [KCl] = 150 mM, pH 7. (c) Dependence of bump curvature on the KCl concentration obtained by the slope of the bump height against the bump radius.

binding sites.26,27 It is reported that Ca2+ can bind to PM with 5−10-fold concentration of bR.26 This strongly supports that the specific interaction between Ca2+ and PM occurred at the Ca2+ concentration range of ∼1 mM. Therefore, almost all negatively charged head groups of the lipids in PM should be bound to Ca2+ at considerably lower Ca2+ concentrations than those used in the present experiments. The distribution of the 9324

dx.doi.org/10.1021/jp5036234 | J. Phys. Chem. B 2014, 118, 9322−9328

The Journal of Physical Chemistry B

Article

Figure 3. Dependence of bump curvature on the ionic strength at pH 7: KCl (open circles) and CaCl2 (filled circles).

charged amino acid residues of bR is given in Table 1. Assuming that all negatively charged groups of the lipids are Table 1. Distribution of Charged Amino Acid Residues in Each Regiona CP soluble domain CP interface region hydrocarbon core EC interface region EC soluble domain

positively charged

negatively charged

net charge

3 6 2 3 0

9 2 3 3 1

−6 +4 −1 ±0 −1

Figure 4. (a) AFM topography of identical PM sheets containing 150 mM KCl with change in the pH from 7 to 12 using the solvent circulation system. (b) Dependence of bump curvature on the pH.

a

The regions are the same as shown in Supporting Information Figure S1, which is constructed from a model by Mitsuoka et al.35 (Protein Data Bank ID code: 2at9).

good agreement with that previously reported.16 The bump curvature increased without change in the PM thickness below pH 9, which is related to a change in the electrostatic interactions in the soluble domain or at the PM interface because no structural changes of bR in PM monitored by midIR and near-UV CD were evident below pH 9.10 Numbers of reports regarding pH titration experiments of the charged amino acid residues of bR have been reported, showing that deprotonation pKa values of Glu204,30 Asp96,31 and the Schiff’s base32 were 9.7, 11.4, and 13.3, respectively. It seems that the most similar pKa to the bump curvature increase below pH 9 is that of the deprotonation of Glu204, which is a member of the proton release group located at the EC proton channel.33 However, it is independent to the bump curvature increase because a crystal structure of bR at alkaline pH demonstrated that no large structural change was induced at pH 10 besides some residues near the proton release group.34 As the pH is increased to 9, lysine residues should begin to deprotonate because the pKa for free lysine is approximately 10. Lysine residues have a tendency to be located at the interface region of PM (Supporting Information Figure S1), and some residues are directly bound to the negatively charged head groups of the PM lipids.35 Therefore, it is reasonable that the bump curvature increases with an increase from neutral pH to pH 9 because the electrostatic repulsive force between negative charges begins to appear at the interface region. Above pH 10, the bump curvature and the number of bumps decreased because the 2D crystalline lattice of bR began to be disassembled.10 Figure 5 shows a plausible mechanism for bump formation via the electrostatic repulsive force in the soluble domain and interface regions at the CP side. Many negatively charged amino acid residues are originally located in the CP soluble

neutralized by Ca2+ binding, six positively charged residues in the CP interface region that were originally bound to the negatively charged lipids via electrostatic attractive force are then left behind in the CP interface region. Consequently, new repulsive force between the positively charged amino acid residues should be generated in the CP interface region. It is thus reasonable that the electrostatic interaction becomes approximately 5-fold stronger in the membrane interface region because the dielectric constant in the interface region is considered to be 12−18,28,29 compared with that for water at ca. 90. It is therefore likely that Ca2+ ions bound to the negatively charged head groups cause a new repulsive force in the CP interface region, which results in considerable increase of curvature of the bump structures. The effect of pH upon the bump curvature was also elucidated using AFM with the solvent circulation system. Figure 4a shows the change in the AFM topography with variation of the solvent pH from 7 to 12. The membrane patch was destroyed by the AFM tip when the pH was increased up to 12, which indicates the membrane is destabilized at extremely high pH. Figure 4b shows the pH dependence of the bump curvature obtained from identical PM sheets. The curvature increased and reached a maximum at pH 9, although it then decreased to the original level above pH 10. The pH correlation for the bump curvature and membrane thickness is shown in Supporting Information Figure S3, in addition to the degree of 2D crystallinity (molar ellipticity at 600 nm, θ600nm) estimated using visible circular dichroism (CD) spectroscopy.10 The PM thickness in the pH range between 7 and 10 was in 9325

dx.doi.org/10.1021/jp5036234 | J. Phys. Chem. B 2014, 118, 9322−9328

The Journal of Physical Chemistry B

Article

Figure 5. Plausible mechanisms for salt effects on PM bump curvature. (a) Bump curvature of PM under salt-free conditions. Strong electrostatic repulsive forces exist in the CP soluble domain due to the asymmetric distribution of negatively charged amino acid residues. (b) Decrease of bump curvature caused by the addition of KCl. The repulsive force weakens by the electrostatic shielding effect of KCl. The remaining curvature is possibly caused by the intrinsic difference between the bR domain sizes of the CP and EC sides. (c) Increase of the bump curvature due to the addition of a small amount of CaCl2 (1 μM−1 mM).24,25 New repulsive interaction is generated in the CP interface region from the specific interaction between negatively charged lipid head groups and Ca2+. (d) Decrease of the bump curvature at higher CaCl2 concentrations. The repulsive interaction is weakened due to the electrostatic shielding effect of CaCl2.

soluble domain are thus considered to disappear due to Ca2+ binding. However, the bump curvature was significantly increased. It is because the electrostatic repulsive force in the membrane interface region should be stronger than that in the soluble domain. This concept corresponds to the experimental result of the curvature increase below pH 9 at which the 2D crystalline lattice is still retained. Taking into account the membrane anchoring property by the snorkeling effect, where positively charged amino acids with long alkyl chains such as lysine and arginine in the membrane−water interface region face the negative charges of the phospholipids,36−38 the present result suggests that the attractive force-dominant environment via electrostatic interaction in the interface regions makes a significant contribution to stabilize both the membrane bilayer consisting of individual membrane proteins and the structural formation of large membrane−lipid complexes formed by selfassembly.

domain of bR to maintain its functional efficiency (Supporting Information Figure S1 and Table 1). Supramolecular structure formation by lipid-mediated strong attractive forces causes a strong repulsive force to arise in the CP surface, which results in a difference in the membrane surface tension between the CP and the EC sides (Figure 5a). The strong repulsive force in the CP soluble domain is considered to be the cause of bump formation because the bump curvature is decreased by electrostatic shielding effects due to the addition of KCl (Figure 5b). In addition, PM was not planar and still had a curvature even at high KCl concentrations, at which the Debye length was within a few Angstroms (Figure 2c). It is reasonable that the remaining curvature of PM at extremely high KCl concentration, where the electric repulsive force at the CP soluble domain is almost shielded, is due to the intrinsic difference in the cross-sectional areas of bR between the CP and the EC sides because changes in the bR configuration such as the opening motion at the CP side are reported to affect the PM curvature and the global shape of PM sheet.15,18,19 A rough estimation of the contribution of the electrostatic repulsive force was tried from the KCl concentration dependence (Figure 2c), showing that as many as 40% of bump curvature is reduced at KCl concentration of 3 M at which all repulsive interaction of the soluble domain is supposed to be shielded. This strongly indicates the significance of the electrostatic repulsive force on the bump formation. As the other possible contributions of the bump curvature, the hydrophobic interaction between some amino acid residues and lipid at the membrane interface regions should not be excluded. The other contributions to the bump curvature should be elucidated in the next project. The influence of Ca2+ on the bump curvature is more dramatic (Figures 5c,d). It is likely that the binding of Ca2+ to the negatively charged head groups of the PM lipids leads to the generation of new repulsive forces between the positive charges in the CP interface region (Figure 5c). The repulsive forces between the clusters of negative charges at the CP

4. CONCLUSIONS The effects of salt concentration and pH on the curvature of bump structures on the CP surface of PM were investigated using AFM technique. AFM topography of identical PM sheets was collected with variation of the solvent conditions using a solvent circulation system. Increasing the KCl concentration resulted in a reduction of the bump curvature due to weakening of the CP surface tension caused by the electrostatic shielding effect. The addition of Ca2+ caused a significant increase of the bump curvature, even at very low Ca2+ concentrations. At higher Ca2+ concentration, a similar decrease in curvature as that of KCl was observed. This experimental evidence is interpreted to indicate the generation of a new repulsive interaction in the CP interface region caused by the specific interaction between Ca2+ and negatively charged phospholipid head groups. The contribution of the electrostatic repulsive force against the bump curvature is roughly estimated, showing 9326

dx.doi.org/10.1021/jp5036234 | J. Phys. Chem. B 2014, 118, 9322−9328

The Journal of Physical Chemistry B

Article

(11) Yokoyama, Y.; Sonoyama, M.; Mitaku, S. Structural changes in bacteriorhodopsin in purple membranes induced by irreversible photobleaching with heterogeneous and homogeneous stability. Photochem. Photobiol. 2010, 86, 297−301. (12) Yokoyama, Y.; Negishi, L.; Kitoh, T.; Sonoyama, M.; Asami, Y.; Mitaku, S. Effect of lipid phase transition on molecular assembly and structural stability of bacteriorhodopsin reconstituted into phosphatidylcholine liposomes with different acyl-chain lengths. J. Phys. Chem. B 2010, 114, 15706−15711. (13) Henderson, R.; Baldwin, J. M.; Downing, K. H.; Lepault, J.; Zemlin, F. Structure of purple membrane from Halobacterium halobium: Recording, measurement and evaluation of electron micrographs at 3.5 Å resolution. Ultramicroscopy 1986, 19, 147−178. (14) Kouyama, T.; Yamamoto, M.; Kamiya, N.; Iwasaki, H.; Ueki, T.; Sakurai, I. Polyhedral assembly of a membrane protein in its threedimensional crystal. J. Mol. Biol. 1994, 236, 990−994. (15) Czégé, J.; Reinisch, L. The pH dependence of transient changes in the curvature of the purple membrane. Photochem. Photobiol. 1991, 54, 923−930. (16) Müller, D. J.; Schabert, F. A.; Büldt, G.; Engel, A. Imaging purple membranes in aqueous solutions at sub-nanometer resolution by atomic force microscopy. Biophys. J. 1995, 68, 1681−1686. (17) Fisher, K. A.; Yanagimoto, K.; Stoeckenius, W. Oriented adsorption of purple membrane to cationic surfaces. J. Cell Biol. 1978, 77, 611−621. (18) Baumann, R. P.; Schranz, M.; Hampp, N. Bending of purple membrane in dependence on the pH analyzed by AFM and single molecule force spectroscopy. Phys. Chem. Chem. Phys. 2010, 12, 4329− 4335. (19) Baumann, R. P.; Eussner, J.; Hampp, N. pH dependent bending in and out of purple membranes comprising bR-D85T. Phys. Chem. Chem. Phys. 2011, 13, 21375−21382. (20) Zhivkov, A. M. Change of purple membranes geometry induced by protein adsorption. Colloid Surf., B 2007, 56, 170−173. (21) Zhong, S.; Li, H.; Chen, X. Y.; Cao, E. H.; Jin, G.; Hu, K. S. Different interactions between the two sides of purple membrane with atomic force microscope tip. Langmuir 2007, 23, 4486−4493. (22) Oesterhelt, D.; Stoeckenius, W. Isolation of the cell membrane of Halobacterium halobium and its fractionation into red and purple membrane. Methods Enzymol. 1974, 31, 667−678. (23) Carmody, W. R. Easily prepared wide range buffer series. J. Chem. Educ. 1961, 38, 559−560. (24) Tokutomi, S.; Lew, R.; Ohnishi, S. Ca2+-induced phase separation in phosphatidylserine, phosphatigylethanolamine and phosphatidylcholine mixed membranes. Biochim. Biophys. Acta 1981, 643, 276−282. (25) Inoko, Y.; Yamaguchi, T.; Furuya, K.; Mitsui, T. Effects of cations on dipamitoyl phosphatidylcholine/cholesterol/water systems. Biochim. Biophys. Acta 1975, 413, 24−32. (26) Váró, G.; Brown, L. S.; Needleman, R.; Lanyi, J. K. Binding of calcium ions to bacteriorhodopsin. Biophys. J. 1999, 76, 3219−3226. (27) Tuzi, S.; Yamaguchi, S.; Tanio, M.; Konishi, H.; Inoue, S.; Naito, A.; Needleman, R.; Lanyi, J. K.; Saitô, H. Location of a cation-binding site in the loop between helices F and G of bacteriorhodopsin as studied by 13C NMR. Biophys. J. 1999, 76, 1523−1531. (28) Flewelling, R. F.; Hubbell, W. L. Hydrophobic ion interactions with membranes. Thermodynamic analysis of tetraphenylphosphonium binding to vesicles. Biophys. J. 1986, 49, 531−540. (29) Wimley, W. C.; White, S. H. Experimentally determined hydrophobicity scale for proteins at membrane interfaces. Nat. Struct. Biol. 1996, 3, 842−848. (30) Balashov, S. P.; Imasheva, E. S.; Govindjee, R.; Ebrey, T. G. Titration of asparatate-85 in bacteriorhodopsin: What it says about chromophore isomerization and proton release. Biophys. J. 1996, 70, 473−481. (31) Száraz, S.; Oesterhelt, D.; Ormos, P. pH-induced structural changes in bacteriorhodopsin studied by Fourier transform infrared spectroscopy. Biophys. J. 1994, 67, 1706−1712.

that as many as 40% of bump curvature is reduced by electrostatic shielding effects. This indicates the significance of the electrostatic repulsive interaction on the bump formation. The pH dependence of the bump curvature can also be explained by the same concept. In summary, a plausible model is proposed for the changes in bump curvature caused by the electrostatic repulsive force, as shown in Figure 5.



ASSOCIATED CONTENT

S Supporting Information *

Amino acid sequence of bR; image correction procedure of AFM topography; bump curvature and membrane thickness. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*(Y.Y.) E-mail: [email protected]. Phone/Fax: +81-52-789-4465. Present Addresses §

(N.W.) Hitachi High-Tech Science Co., Tokyo 105-0003, Japan. || (M.S.) Division of Molecular Science, Faculty of Science and Technology, Gunma University, Kiryu 376-8515, Japan. ⊥ (S.M.) Toyota Physical and Chemical Research Institute, Nagakute 480-1192, Japan. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank Mr. Y. Shirakawabe of Hitachi High-Tech Science Co. and Ms. A. Nihei of Food Safety Commission Secretariat, Cabinet Office for continuous technical support. This work was supported in part by the SENTAN, JST, and by the JSPS KAKENHI Grant Numbers 20370058 and 26390046.



REFERENCES

(1) Henderson, R. The purple membrane from Halobacterium halobium. Annu. Rev. Biophys. Bioeng. 1977, 6, 87−109. (2) Oesterhelt, D.; Stoeckenius, W. Rhodopsin-like protein from the purple membrane of Halobacterium halobium. Nat. New Biology 1971, 233, 149−152. (3) Lanyi, J. K. Bacteriorhopsin. Annu. Rev. Physiol. 2004, 66, 665− 688. (4) Dencher, N. A.; Heyn, M. P. Bacteriorhodopsin monomers pump protons. FEBS Lett. 1979, 108, 307−310. (5) Etoh, A.; Itoh, H.; Mitaku, S. Light-induced denaturation of bacteriorhodopsin just above melting point of two-dimensional crystal. J. Phys. Soc. Jpn. 1997, 66, 975−978. (6) Mukai, Y.; Kamo, N.; Mitaku, S. Light-induced denaturation of bacteriorhodopsin solubilized by octyl-β-glucoside. Protein Eng. 1999, 12, 755−759. (7) Yokoyama, Y.; Sonoyama, M.; Mitaku, S. Irreversible photobleaching of bacteriorhodopsin in a high-temperature intermediate state. J. Biochem. 2002, 131, 785−790. (8) Yokoyama, Y.; Sonoyama, M.; Mitaku, S. Inhomogeneous stability of bacteriorhodopsin in purple membrane against photobleaching at high temperature. Proteins 2004, 54, 442−454. (9) Sasaki, T.; Sonoyama, M.; Demura, M.; Mitaku, S. Photobleaching of bacteriorhodopsin solubilized with Triton X-100. Photochem. Photobiol. 2005, 81, 1131−1137. (10) Yokoyama, Y.; Sonoyama, M.; Nakano, T.; Mitaku, S. Structural changes of bacteriorhodopsin in the purple membrane above pH 10 decreases heterogeneity of the irreversible photobleaching components. J. Biochem. 2007, 142, 325−333. 9327

dx.doi.org/10.1021/jp5036234 | J. Phys. Chem. B 2014, 118, 9322−9328

The Journal of Physical Chemistry B

Article

(32) Druckmann, S.; Ottolenghi, M.; Pande, A.; Pande, J.; Callender, R. H. Acid−base equilibrium of the Schiff base in bacteriorhodopsin. Biochemistry 1982, 21, 4935−4959. (33) Richter, H. T.; Brown, L. S.; Needleman, R.; Lanyi, J. K. A linkage of the pKa’s of asp-85 and glu-204 forms part of the reprotonation switch of bacteriorhodopsin. Biochemistry 1996, 35, 4054−4062. (34) Okumura, H.; Murakami, M.; Kouyama, T. Crystal structures of acid blue and alkaline purple forms of bacteriorhodopsin. J. Mol. Biol. 2005, 351, 481−495. (35) Mitsuoka, K.; Hirai, T.; Murata, K.; Miyazawa, A.; Kidera, A.; Kimura, Y.; Fujiyoshi, Y. The structure of bacteriorhodopsin at 3.0 Å resolution based on electron crystallography: implication of the charge distribution. J. Mol. Biol. 1999, 286, 861−882. (36) Krishnakumar, S. S.; London, E. The control of transmembrane helix transverse position in membranes by hydrophilic residues. J. Mol. Biol. 2007, 374, 1251−1269. (37) Standberg, E.; Killian, J. A. Snorkeling of lysine side chains in transmembrane helices: how easy can it get? FEBS Lett. 2003, 544, 69−73. (38) Standberg, E.; Morein, S.; Rijkers, D. T.; Liskamp, R. M.; van der Wel, P. C.; Killian, J. A. Lipid dependence of membrane anchoring properties and snorkeling behavior of aromatic and charged residues in transmembrane peptides. Biochemistry 2002, 41, 7190−7198.

9328

dx.doi.org/10.1021/jp5036234 | J. Phys. Chem. B 2014, 118, 9322−9328