Direct Observation of Protein Motion during the Photochemical

Langmuir 2007, 23, 7225-7228

7225

Direct Observation of Protein Motion during the Photochemical Reaction Cycle of Bacteriorhodopsin Zolta´n Ba´lint,† Gergely A. Ve´gh,† Anca Popescu,# Mihai Dima,# Constanta Ganea,# and Gyo¨rgy Va´ro´*,† Institute of Biophysics, Biological Research Center of the Hungarian Academy of Sciences, Szeged, Hungary, H-6726, and Department of Biophysics, “Carol DaVila” UniVersity of Medicine and Pharmaceutics, Bucharest, Romania, 050474 ReceiVed March 7, 2007 Platinum-coated, conductive atomic force microscope cantilevers were used to deposit electrophoretically purple membranes from Halobacterium salinarum on the bottom part of the cantilevers. By illuminating the bacteriorhodopsincontaining purple membranes, the protein goes through its photochemical reaction cycle, during which a conformational change happens in the protein, changing its shape and size. The size change of the protein acts upon the cantilever by causing its deflection, which can be monitored by the detection system of the atomic force microscope. The shape of the signal, the action spectrum of the deflection amplitude, and the blue light inhibition of the deflection all prove that the origin of the signal is the conformational change arising in the bacteriorhodopsin during the photocycle. From the size of the signal, the magnitude of the protein motion could be estimated. Using polarized light, the orientation of the motion was determined, relative to the transition moment of the retinal.

Introduction The process of proton pumping of the light-driven retinal protein bacteriorhodopsin (BR) is accompanied by protein structural change induced by the photon absorption-initiated reaction cycle. The BR is tightly packed in the purple membrane (PM), a part of the cell membrane of Halobacterium salinarum.1 The structures of BR and its intermediates are known with high resolution.2-5 Retinal is bound to lysine 216 through a protonated Schiff-base,6 which plays a key role in the proton transporting process.7 Its orientation inside the protein is well defined.2,5 Previously, it was determined that the light linearly polarized parallel to the retinal is absorbed and initiates the photocycle.8 A detailed description of the photocycle can be found in review papers.9-11 Briefly, the photocycle proceeds as follows: After absorbing the light quanta, a charge separation along the retinal occurs in the femtosecond time domain,12 followed by an alltrans to 13-cis isomerization in several picoseconds, reaching the K state. In the K f L transition, a local rearrangement around the retinal occurs in less than 10 µs. In the L f M1 transition, about 100 µs after the excitation, the Schiff-base of the retinal * Corresponding author. † Biological Research Center of the Hungarian Academy of Sciences. # “Carol Davila” University of Medicine and Pharmaceutics. (1) Oesterhelt, D.; Stoeckenius, W. Methods Enzymol. 1974, 31, 667-678. (2) Luecke, H.; Schobert, B.; Richter, H. T.; Cartailler, J. P.; Lanyi, J. K. J. Mol. Biol. 1999, 291, 899-911. (3) Edman, K.; Nollert, P.; Royant, A.; Belrhali, H.; Pebay-Peyroula, E.; Hajdu, J.; Neutze, R.; Landau, E. M. Nature 1999, 401, 822-826. (4) Sass, H. J.; Bu¨ldt, G.; Gessenich, R.; Hehn, D.; Neff, D.; Schlesinger, R.; Berendzen, J.; Ormos, P. Nature 2000, 406, 649-653. (5) Lanyi, J. K.; Schobert, B. J. Mol. Biol. 2003, 328, 439-450. (6) Stoeckenius, W.; Lozier, R. H.; Bogomolni, R. A. Biochim. Biophys. Acta 1979, 505, 215-278. (7) Brown, L. S.; Dioumaev, A. K.; Needleman, R.; Lanyi, J. K. Biochemistry 1998, 37, 3982-3993. (8) Cze´ge´, J.; De´r, A.; Zima´nyi, L.; Keszthelyi, L. Proc. Natl. Acad. Sci. U.S.A. 1982, 79, 7273-7277. (9) Krebs, M. P.; Khorana, H. G. J. Bacteriol. 1993, 175, 1555-1560. (10) Lanyi, J. K. Int. ReV. Cytol. 1999, 187, 161-202. (11) Lanyi, J. K. Annu. ReV. Physiol. 2004, 66, 665-688. (12) Groma, G. I.; Colonna, A.; Lambry, J. C.; Petrich, J. W.; Va´ro´, G.; Joffre, M.; Vos, M. H.; Martin, J. L. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 79717975.

deprotonates by transferring its proton to the acceptor Asp 85, and a proton is released from the release group on the surface of the membrane close to the external medium.13,14A conformational change in the BR occurs during the M1 f M2 f N transitions, when the protein switches from the extracellular to the cytoplasmic conformation in about 100 µs.15 The accessibility of the Schiff-base changes from the extracellular to the cytoplasmic side.7 The Schiff-base is reprotonated by the proton donor Asp 96 on the cytoplasmic side, reflected by the appearance of intermediate N. In the N f O f BR transition, the retinal reisomerizes to its original all-trans form, a proton is taken up from the cytoplasmic side, and the proton from the acceptor is transferred to the proton release group in the millisecond time domain.10,16 At low pH, below the pKa of the proton acceptor group, the photocycle does not translocate protons across the membrane, and the M intermediate is missing from the photocycle.17 A powerful tool to identify the M intermediate, characteristic of the proton-pumping activity of the BR, is the blue light-induced inhibition.18,19 When the PMs are allowed to air-dry, the BR still preserves its photocycle up to the M intermediate, and, once the sample is rehydrated, the protein regains its original proton-pumping function.20-22 Using the permanent electric dipole and the net negative charge of the PM, highly oriented membrane samples were prepared electrophoretically on a conductive surface,20,23 (13) Brown, L. S.; Sasaki, J.; Kandori, H.; Maeda, A.; Needleman, R.; Lanyi, J. K. J. Biol. Chem. 1995, 270, 27122-27126. (14) Spassov, V. Z.; Luecke, H.; Gerwert, K.; Bashford, D. J. Mol. Biol. 2001, 312, 203-219. (15) Va´ro´, G.; Lanyi, J. K. Biochemistry 1990, 29, 2241-2250. (16) Oesterhelt, D.; Tittor, J.; Bamberg, E. J. Bioenerg. Biomembr. 1992, 24, 181-191. (17) Va´ro´, G.; Lanyi, J. K. Biophys. J. 1989, 56, 1143-1151. (18) Dancsha´zy, Z.; Drachev, L. A.; Ormos, P.; Nagy, K.; Skulachev, V. P. FEBS Lett. 1978, 96, 59-63. (19) Brown, L. S.; Zima´nyi, L.; Ottolenghi, M.; Needleman, R.; Lanyi, J. K. Biochemistry 1993, 32, 7679-7685. (20) Keszthelyi, L.; Ormos, P.; Va´ro´, G. Acta Phys. Acad. Sci. Hung. 1982, 53, 143-157. (21) Ganea, C.; Gergely, C.; Ludmann, K.; Va´ro´, G. Biophys. J. 1997, 73, 2718-2725. (22) Va´ro´, G.; Lanyi, J. K. Biophys. J. 1991, 59, 313-322. (23) Va´ro´, G. Acta Biol. Acad. Sci. Hung. 1981, 32, 301-310.

10.1021/la700666p CCC: $37.00 © 2007 American Chemical Society Published on Web 05/16/2007

7226 Langmuir, Vol. 23, No. 13, 2007

which helped to determine the charge motions inside the protein during its function. Change in the ultraviolet light scattering during the photocycle indicated the existence of protein motion during the protonpumping process.24,25 The conformational change in the second part of the photocycle was determined by electron diffraction on deep-frozen PMs and BR crystals.26,27 After the high-resolution crystalline structure of the intermediates was determined by X-ray diffraction, the changes in the protein structure were calculated, and it was deduced that the conformational change happens at helices F and G.5,26,28,29 Although these structures were determined from crystallized proteins, it is assumed that the changes in the native membrane could be identical or similar. The first observation by atomic force microscopy (AFM) of the change in the image of the BR during the photocycle was effectuated by Mu¨ller et al. by scanning the PMs during a strong illumination of the sample.30,31 Another way to observe the conformational change of the BR was to position the AFM cantilever at the top of the protein and register its changes after a strong illumination of the sample.32-34 All the above-mentioned experiments register the changes occurring in the BR perpendicular to the membrane surface, while the important conformational changes happen inside the membrane. A new concept in the conformational measurement was the deposition of BR proteoliposomes on microcantilever sensors.35 It was possible to monitor the conformational changes during the retinal removal (bleaching) of the protein. In this paper, we report the study of the conformational changes occurring inside the membrane during the photochemical reaction cycle of the BR by measuring the deflection change of a conductive AFM cantilever, on which PMs were electrophoretically deposited. From the measurements it was possible to estimate the size of the area increase of the active protein and the relative direction of the motion compared to the orientation of the retinal. Materials and Methods PMs were isolated from Halobacterium salinarum strain S9 according to a standard procedure.1 The preparation was further washed by centrifugation in freshly prepared tridistilled water to remove any traces of salt from the suspension. After the last centrifugation, the pellet was diluted to contain about 6 mg/mL BR. The BR-containing PMs were deposited on a conductive AFM cantilever (type CSC37/Ti-Pt Micromasch, Tallinn, Estonia) by the standard procedure described previously,23,36 modified to the small size. Briefly, 20 µL of suspension was pipetted between the Pt cathode and the three cantilever anodes, placed at a distance of 0.5 mm. A 400 nA current deposited the BR on the bottom part of the cantilevers in about 3 min, forming a PM multilayer of 0.2-0.5 OD. The top part of the cantilevers, which is used to reflect the (24) Cze´ge´, J. Acta Biochim. Biophys. Hung. 1987, 22, 463-478. (25) Cze´ge´, J.; Reinisch, L. Biophys. J. 1990, 58, 721-729. (26) Subramaniam, S.; Gerstein, M.; Oesterhelt, D.; Henderson, R. EMBO J. 1993, 12, 1-8. (27) Subramaniam, S.; Hirai, T.; Henderson, R. Philos. Trans. R. Soc. London, Ser. A 2002, 360, 859-874. (28) Luecke, H.; Schobert, B.; Richter, H. T.; Cartailler, J. P.; Lanyi, J. K. Science 1999, 286, 255-260. (29) Lanyi, J. K. Mol. Membr. Biol. 2004, 21, 143-150. (30) Muller, D. J.; Buldt, G.; Engel, H. A. J. Mol. Biol. 1995, 249, 239-243. (31) Persike, N.; Pfeiffer, M.; Guckenberger, R.; Fritz, M. Colloids Surf., B 2000, 19, 325-332. (32) Lewis, A.; Rousso, I.; Khachatryan, E.; Brodsky, I.; Lieberman, K.; Sheves, M. Biophys. J. 1996, 70, 2380-2384. (33) Rousso, I.; Khachatryan, E.; Brodsky, I.; Nachustai, R.; Ottolenghi, M.; Sheves, M.; Lewis, A. J. Struct. Biol. 1997, 119, 158-164. (34) Rousso, I.; Khachatryan, E.; Gat, Y.; Brodsky, I.; Ottolenghi, M.; Sheves, M.; Lewis, A. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 7937-7941. (35) Braun, T.; Backmann, N.; Vogtli, M.; Bietsch, A.; Engel, A.; Lang, H. P.; Gerber, C.; Hegner, M. Biophys. J. 2006, 90, 2970-2977. (36) Va´ro´, G.; Keszthelyi, L. Biophys. J. 1983, 43, 47-51.

Ba´ lint et al.

Figure 1. Deflection of a BR-coated AFM cantilever upon yellow light illumination measured in air (solid line) and in water (dashed line). The inset is the deflection of a black ink-coated cantilever upon yellow light illumination. Measuring conditions: temperature 30 °C, relative humidity in air 45%. position sensing laser diode, remained clear because it was not immersed in the suspension. The deposited BR was allowed to air-dry. By determining the resonant frequency of the cantilevers before and after the deposition, the approximate quantity of the deposited material could be estimated. Because the deposition was not localized to the tip of the cantilever, but rather along the whole cantilever, this calculation gave the lower limit of the deposited quantity.37 AFM measurements were performed with an Asylum MFP-3D head and a Molecular Force Probe 3D controller (Asylum Research, Santa Barbara, CA). The MFP-3D head was mounted on a Zeiss Axiovert 200 microscope, equipped with a 100 W halogen lamp, which illuminated the cantilever from below, through the objective. In the light-path band-pass or interference, filters were inserted to select the desired wavelength. The length of the light excitation was between 100 ms and several seconds. Much shorter excitation would not produce a measurable signal, as the reaction time of the AFM is in the range of several milliseconds. With a polarizer inserted in the light path, the polarization of the excitation light could be controlled. The position of the polarizer was calibrated in such a way that, at 0 and 180°, the plane of polarization was parallel to the cantilever. The light intensity at each wavelength and polarization angle was measured with an Ophir Nova laser power monitor with a PD 300-UV-SH head (Ophir Optronics, Ltd., Jerusalem, Israel). The deflection was monitored with an acquisition rate of 500 points per second for 10 s, during which the cantilever did not touch any hard surface. The measurement was externally synchronized with the shutter of the light microscope. The deflection signal was measured, keeping the cantilever in air or in water, far from the hard surface. In air, the humidity and the temperature of the closed chamber of the AFM was controlled by purging the chamber with a partially humidified air flow. The temperature of the sample was 30 °C, and the relative humidity was 45%, if not otherwise mentioned. When measured in water, the cantilever was immersed in freshly distilled water, having a pH of around 6. Since, during the sample preparation, the PM suspension had a pH of around 7, it can be assumed that, during the measurement, the pH was close to neutral.

Results and Discussion Illumination of the BR-deposited cantilever with yellow light (λ > 500 nm) produced an upward deflection of it. The signal from the sample in air (Figure 1, solid line) is much larger than that measured in water (Figure 1, dashed line). Several controls were made to prove that the signal originated from the PMs. An uncoated, clean cantilever did not show any deflection during illumination (not shown). If the bottom surface of the cantilever was painted with black ink and air-dried, its (37) Ilic, B.; Craighead, H. G.; Krylov, S.; Senaratne, W.; Ober, C.; Neuzil, P. J. Appl. Phys. 2004, 95, 3694-3703.

Detection of Protein Motion by AFM

Figure 2. Deflection of a BR-coated AFM cantilever upon 577 nm monochromatic light illumination measured in air (solid line) and in water (dashed line). Measuring conditions were the same as those for Figure 1.

Figure 3. Light intensity dependence of the cantilever deflection upon yellow light illumination measured in air (diamonds) and in water (circles). The lines are provided just to conduct the eye. Measuring conditions were the same as those for Figure 1.

illumination in air resulted in a downward deflection of less than 1 nm (Figure 1, inset). These control measurements eliminated the possibility of the signal resulting from either heating the cantilever or the interference of the excitation light with the position-sensing photodiode. The BR Signals and the Action Spectrum. To eliminate any artifact coming from the possible multiple excitations of the BR and its intermediates by a rather strong polychromatic light, monochromatic light was produced with interference filters placed in the filter holder of the microscope. The signal measured during a 577 nm excitation is shown in Figure 2. The amplitude difference between the signal measured in air (Figure 2, solid line) and that measured in water (Figure 2, dashed line) can be partially explained by the difference in the decay rate of the photocycle. In water, after switching off the light, the signal decays with a time constant of around 6-7 ms, the decay time of the BR photocycle in suspension at low and neutral pH.38 In air, the signal decays in about 300 ms, the average decay time of the photocycle at 50% relative humidity.21,36 The ratio of the two decay times is about 40-50. This is responsible for the difference between the light intensity dependence of the signal generated by yellow light excitation in water (Figure 3, circles) or in air (Figure 3, diamonds). The log-log plot of the light intensity dependence shows that the generated signal is far from saturation.

Langmuir, Vol. 23, No. 13, 2007 7227

Figure 4. Action spectrum of the relative deflection amplitude measured in air (circles) and in water (squares). The solid line is the absorption spectrum of the BR. Measuring conditions were the same as those for Figure 1.

If the changes in the BR producing the deflection are the same in both cases, the ratio of the two decay times would suggest a much larger amplitude ratio, instead of the observed ratio of 3:6 (Figures 1 and 2). The difference can be explained by the changed photocycle of the dried BR. In air, the photocycle stops at the early M intermediate, and the conformational change in the protein is hindered.21,36 Contrary to earlier assumptions that the abovementioned hindrance totally abolishes the conformational changes, these measurements prove that there is only a partial hindrance. The wavelength dependence of the signal amplitude, normalized to the incident light intensity (Figure 4), follows the shape of the BR absorption spectrum (solid line), resulting in the action spectrum of the sample for signals measured both in air (circles) and in water (squares). The presented absorption spectrum was measured on a dark-adapted PM suspension. The light-adapted spectrum is about 5 nm red-shifted, which is a minor change compared to the scatter of the measured points. The use of the dark-adapted spectrum is more suitable because the dried BR cannot be totally light-adapted.21,36 The correlation of the action spectrum with the absorption spectrum is strong evidence that the measured deflection signal is related to the changes in the light-excited BR protein. The Blue Light Effect. To identify the intermediate, which is responsible for generating the observed deflection, the effect of double light excitation was investigated by using a green light and a blue light, with the sample in air. To perform the measurements, a special double-peak interference filter (405 and 590 nm) was combined with a yellow and blue glass filter to have the desired 405 nm, 590 nm, or the superposition of these two. Separately, both the green (Figure 5, dashed line) and blue lights (Figure 5, dashed-dotted line) generate a positive deflection because they excite the ground-state BR. When, upon the green light excitation, the blue light is turned on, it results in an amplitude decrease (Figure 5, solid line). The green light accumulates a lot of the M intermediate because the photocycle in air is very slow. The blue light excites this M intermediate and drives it back to the ground state.18 The appearance of the blue light inhibition proves the accumulation of the blue-shifted M intermediate, demonstrating that the signal amplitude is directly connected to its accumulation. Measurements with Polarized Light. To determine the direction of the size change, the exciting light was linearly polarized, and the polarization angle relative to the cantilever was changed. Zero degrees corresponds to the polarization parallel to the cantilever. The very dried sample (relative humidity, 5%) (38) Ludmann, K.; Gergely, C.; Va´ro´, G. Biophys. J. 1998, 75, 3110-3119.

7228 Langmuir, Vol. 23, No. 13, 2007

Figure 5. Deflection of a BR-coated AFM cantilever upon illumination with 590 nm light (dashed line) and 405 nm light (dashed-dotted line); during the 590 nm illumination, the 405 nm light was switched on and off (solid line). The blue light inhibition can be observed. Measuring conditions were the same as those for Figure 1.

Figure 6. The dependence of the deflection of the BR-coated cantilever on the polarization angle of the exciting light, measured at different relative humidities. The signals were normalized at their maximum. The zero angle corresponds to the light polarized parallel to the cantilever. Measuring conditions: temperature 30 °C.

showed maximal deflection when the light was polarized perpendicular to the cantilever (Figure 6, dashed line) with an amplitude change of about 40%. In this case, the retinal is perpendicular to the main direction of the protein motion. With the cantilever immersed in water, the maximum deflection occurs with the polarized light parallel to the cantilever (Figure 6, solid line). The amplitude change is almost 70%. In the wet sample, the main protein motion is parallel to the retinal molecule. At intermediate humidity, a mixture of the two types of curves can be observed (Figure 6, dotted line). The appearance of the effect of water, already at relatively low humidity, and its increase signify that binding only a few water molecules can initiate the transition between the two extremes. The water molecules inside the PM form H-bonds with the protein and make its structure resemble that of its hydrated state. Characterization of the Conformational Changes during the Photocycle. From the change in the resonant frequency of the cantilever after the BR deposition, the mass of deposited PMs could be estimated, resulting in a minimum of 59.6 ng.

Ba´ lint et al.

Knowing the molecular mass of the BR, the area of the BR, and the dimensions of the cantilever, the number of deposited layers was calculated to be a minimum of 156, which corresponds to an optical density of a minimum of 0.15, in good agreement with that estimated during the deposition (0.2-0.5 OD). This low optical density and the optical path of the microscope assures that, even at low light intensity, the whole sample is uniformly excited. It was considered that the straight cantilever was bent uniformly by the increase in the size of the BR. In air, 2/3 of the 300 × 35 × 2 µm cantilever was illuminated, and, during maximal excitation, the cantilever was deflected with about 870 nm, which produced an elongation of 8.7 nm on its exterior part (for details of the calculation, see the Supporting Information). From the shape of the intensity dependence of the signal (Figure 3), the estimation suggested that roughly 30% of the BR molecules were excited, when measured in air, by using maximum yellow light intensity. If it is considered that the BR molecules have a roughly circular area and uniformly cover the surface of the cantilever, the calculated elongation would result from a 0.66 pm diameter increase of one BR molecule. Considering that, in water, using the same light intensity, the excitation is about 50 times smaller (about 0.6%) and the maximum deflection is about 150 nm, the diameter increase of one molecule would be 5.7 pm, almost 1 order of magnitude larger compared to that measured in air. If it is assumed that the length of a covalent bond is approximately 100 pm, the several picometer increase observed in the diameter of the protein means an insignificant shift in the position of the atoms. It is important to note that this dilatation is not a fluctuating random motion of the molecule, but it is synchronized to the photocycle. All these calculations are very approximate, permitting just the estimation of the order of magnitude of the motions. In air, the smaller dilatation of the protein could be explained in two ways: either the conformational change of the BR is not similar to that occurring in water, or only a small fraction of the excited molecules go through the conformational change. On the basis of the literature,21,36 we are inclined to accept the first variant. All the excited molecules go through the same conformational change, which is much smaller than that in water because the dried protein is more rigid. From the deflection signal of the cantilever, when BR was deposited on one side of it and illuminated by continuous light, the conformational change of the protein was detected. From the measured signal, the size and direction of the changes could be estimated. In the BR, the effect of the point mutation on the conformational change of the retinal protein can be studied. This measuring technique makes it possible to monitor the conformational changes occurring in other proteins, if the protein can be deposited on the cantilever and the conformational change can be triggered by an external signal. Acknowledgment. The National Science Research Fund of Hungary (OTKA T048706) supported this work. A.P., M.D., and C.G. performed this research in the frame of the RomanianHungarian Bilateral Cooperation, 2006-2007, under number RO-4/05 by TEÄ T Hungary and ANCS Romania. Supporting Information Available: Estimation of the size increase of the BR molecule caused by its conformational change during the photocycle. This material is available free of charge via the Internet at http://pubs.acs.org. LA700666P