Immobilizing Single Lipid and Channel Molecules in Artificial Lipid

Publication Date (Web): June 1, 2006 ... Simultaneous Optical and Electrical Recording of Single Molecule Bonding to Single Channel Proteins. Toru Ide...
0 downloads 0 Views 240KB Size
6302

Langmuir 2006, 22, 6302-6307

Immobilizing Single Lipid and Channel Molecules in Artificial Lipid Bilayers with Annexin A5 Takehiko Ichikawa,† Takaaki Aoki,‡ Yuko Takeuchi,‡ Toshio Yanagida,§ and Toru Ide*,‡,§ Department of Biophysical Engineering, Graduate School of Engineering Science, Osaka UniVersity, 1-3, Yamadaoka, Suita, Osaka 565-0871, Japan, Creation of InnoVatiVe Technology by Integration of Nanotechnology with Information, Biological and EnVironmental Technologies, Japan Science and Technology Agency, Osaka UniVersity, 1-3, Yamadaoka, Suita, Osaka 565-0871, Japan, and Soft Biosystem Group, Laboratories for Nanobiology, Graduate School of Frontier Biosciences, Osaka UniVersity, 1-3, Yamadaoka, Suita, Osaka 565-0871, Japan ReceiVed December 27, 2005. In Final Form: April 21, 2006 The effects of annexin A5 on the lateral diffusion of single-molecule lipids and single-molecule proteins were studied in an artificial lipid bilayer membrane. Annexin A5 is a member of the annexin superfamily, which binds preferentially to anionic phospholipids in a Ca2+-dependent manner. In this report, we were able to directly monitor single BODIPY 1,2-dihexadecanoyl-sn-glycero-3-phosphoethanolamine (DHPE) and ryanodine receptor type 2 (RyR2) labeled with Cy5 molecules in lipid bilayers containing phosphatidylserine (PS) by using fluorescence microscopy. The diffusion coefficients were calculated at various annexin A5 concentrations. The diffusion coefficients of BODIPYDHPE and Cy5-RyR2 in the absence of annexin A5 were 4.81 × 10-8 cm2/s and 2.13 × 10-8 cm2/s, respectively. In the presence of 1 µM annexin A5, the diffusion coefficients of BODIPY-DHPE and Cy5-RyR2 were 2.2 × 10-10 cm2/s and 9.5 × 10-11 cm2/s, respectively. Overall, 1 µM of annexin A5 was sufficient to induce a 200-fold decrease in the lateral diffusion coefficient. Additionally, we performed electrophysiological examinations and determined that annexin A5 has little effect on the function of RyR2. This means that annexin A5 can be used to immobilize RyR2 in a lipid bilayer when imaging and analyzing RyR2.

Introduction Recent developments in optical microscopy have enabled us to directly observe single fluorescent molecules. In particular, in the past decade, the development of single-molecule imaging techniques has allowed the study of the dynamics of single biological molecules both in vivo and in vitro. This powerful technique has enabled the finding of several biological phenomena that could not have been identified by using multimolecular methods.1-3 These techniques can also be applied to study the dynamic properties of ion channel proteins. Single ion channel current recordings have been performed using the patch-clamp technique or lipid bilayer systems from 25 years ago.4-6 However, imaging of single-channel proteins in lipid bilayers is a far less established technique. Our goal is to establish simultaneous electrical and optical observations of single ion channels and elucidate the correlation between ligand binding and current. Previously, we observed ions passing through channels using a horizontal bilayer membrane formed on a thin agarose layer and simultaneously recorded the single channel’s current.7-9 Others * To whom correspondence should be addressed. E-mail: ide@ phys1.med.osaka-u.ac.jp. Telephone: +81-6-6879-4428. Fax: +81-6-68794428. † Graduate School of Engineering Science, Osaka University. ‡ Japan Science and Technology Agency. § Graduate School of Frontier Biosciences, Osaka University. (1) Ishii, Y.; Ishijima, A.; Yanagida, T. Trends Biotechnol. 2001, 19, 211216. (2) Ishijima, A.; Yanagida, T. Trends Biochem. Sci 2001, 26, 438-444. (3) Sako, Y.; Yanagida, T. Nat. ReV. Mol. Cell Biol. 2003, 4, SS1-5. (4) Neher, E.; Sakmann, B. Nature 1976, 260, 799-802. (5) LaTorre, R.; Alvarez, O. Physiol. ReV. 1981, 61, 77-150. (6) Montal, M. J. Membr. Biol. 1987, 98, 101-115. (7) Ide, T.; Yanagida, T. Biochem. Biophys. Res. Commun. 1999, 265, 595599. (8) Ide, T.; Takeuchi, Y.; Aoki, T.; Yanagida, T. Jpn. J. Physiol. 2002, 52, 429-434. (9) Ide, T.; Takeuchi, Y.; Yanagida, T. Single Mol. 2002, 3, 33-42.

also have observed ion movement and performed single-channel current recordings.10-12 However, the structural dynamic properties in which ion channels correlate the intermolecular signal of ligand binding and the resulting electrical influx is essentially unknown. A major technological obstacle in carrying out simultaneous imaging ligand binding of single-channel and electrical change is the lateral diffusion of the channel protein within the lipid bilayer. This makes long observations nearly impossible. Although it is possible to design a smaller lipid bilayer in order to trap the channel protein within the fluorescent field, the lipid bulk layer encompassing a bilayer cannot be placed in the one molecule detection microscopic view because it scatters the excitation laser beam, resulting in excessive noise. This increases the background of the fluorescent field, making single-molecule channel observation impossible. On the other hand, if the size of the microscopic view is smaller than the size of the bilayer, channel proteins can then easily move out of the microscopic view. This has prevented long-term measurement of ligand binding kinetics. To resolve these problems, we have considered using annexin A5, a membrane-binding protein. Annexins are a family of calcium-dependent proteins that preferentially bind to negatively charged lipids.13-16 Atomicforce microscopic and electron microscopic studies have revealed (10) Harms, G. S.; Orr, G.; Montal, M.; Thrall, B. D.; Colson, S. D.; Lu, H. P. Biophys. J. 2003, 85, 1826-1838. (11) Peng, S.; Publicover, N. G.; Kargacin, G. J.; Duan, D.; Airey, J. A.; Sutko, J. L. Biophys. J. 2004, 86, 134-144. (12) Lu, H. P. Acc. Chem. Res. 2005, 38, 557-565. (13) Gerke, V.; Creutz, C. E.; Moss, S. E. Nat. ReV. Mol. Cell Biol. 2005, 6, 449-461. (14) Rescher, U.; Gerke, V. J. Cell Sci. 2004, 117, 2631-2639. (15) Gerke, V.; Moss, S. E. Physiol. ReV. 2002, 82, 331-371. (16) Swairjo, M. A.; Seaton, B. A. Annu. ReV. Biophys. Biomol. Struct. 1994, 23, 193-213.

10.1021/la0535025 CCC: $33.50 © 2006 American Chemical Society Published on Web 06/01/2006

Immobilizing Single Channels with Annexin A5

that annexin A5 self-assembles and forms p6 or p3 twodimensional (2D) crystals.17-20 It is known that the binding of annexin A5 to phospholipids influences the fluidity of the lipid bilayer.21 Saurel and Cezanne et al. reported that annexin A5 decreased the diffusion of the lipid bilayer using the multimolecular methods NMR and FRAP.22,23 Peng et al. reported that annexin B12 reduced the diffusion of single ryanodine receptor type 2 (RyR2) by tracking the center of Ca2+ flux.24 In our report, we have used singlemolecule tracking techniques to observe the influence of annexin A5 on the lipid fluidity of single lipid proteins and the lateral diffusion of single RyR2 in an artificial planner lipid bilayer. RyR2 is a Ca2+-induced Ca2+ release channel and is the major calcium release channel of the sarcoplasmic reticulum (SR) in cardiomyocytes. RyR2 forms homotetramers with a molecular weight of about 2.3 MDa. Because of the rather large electrical conductance of about 700 pS when the potassium ion is the charge carrier, single-channel current detection of RyR2 is fairly easy.25 We tracked single fluorescently labeled lipids and single RyR2 channel proteins in the artificial planner lipid bilayers containing phosphatidylserine (PS). Furthermore, we demonstrate that annexin A5 reduces the lateral diffusion of single RyR2 channel proteins in a concentration-dependent manner. Last, we measured the K+ current of single RyR2 in the presence of 1 µM annexin A5 and observed that annexin A5 has minimal effect on RyR2 function. Materials and Methods Materials. Phosphatidylethanolamine (PE), phosphatidylserine (PS), and phosphatidylcholine (PC) were purchased from SigmaAldrich (Missouri), n-decane from Wako Pure Chemical Industries (Osaka, Japan), monoclonal anti-ryanodine receptor antibody (antiRyR2) was raised in a mouse against a canine cardiac ryanodine receptor (RyR2) from Affinity Bioreagents (CO), and Cy5monofunctional reactive dye from General Electric (CT). All other chemicals were commercial products of an analytical grade. Purification of RyR2. The purified canine cardiac RyR2 was prepared according to the method of Anderson et al.,26 and the reconstitution was performed according to the method of Lee et al.27 From about 30 g of canine heart, about 0.6 g of heavy sarcoplasmic reticulum (HSR) was isolated. From 7.5 mg of HSR, 0.27 mg of purified RyR2 was obtained. Purified RyR2 was confirmed by 2-15% gradient SDS-PAGE and Western blot against antiRyR2. Fluorescence Labeling with Cy5-Conjugated Antibody. RyR2 antibody was conjugated to Cy5-monofunctional reactive dye following the appended instructions. Free Cy5 was removed using a NAP-10 gel filtration column (General Electric, CT). After conjugation with Cy5, antiRyR2 was reacted with purified RyR2 for 2 h at 4 °C. To remove unreacted antibodies, the labeled sample was washed twice with 5 volumes of the washing buffer following (17) Richter, R. P.; Lai Kee Him, J.; Tessier, B.; Tessier, C.; Brisson, A. R. Biophys. J. 2005, 89, 3372-3385. (18) Mo, Y.; Campos, B.; Mealy, T. R.; Commodore, L.; Head, J. F.; Dedman, J. R.; Seaton, B. A. J. Biol. Chem. 2003, 278, 2437-2443. (19) Oling, F.; Bergsma-Schutter, W.; Brisson, A. J. Struct. Biol. 2001, 133, 55-63. (20) Patel, D. R.; Isas, J. M.; Ladokhin, A. S.; Jao, C. C.; Kim, Y. E.; Kirsch, T.; Langen, R.; Haigler, H. T. Biochemistry 2005, 44, 2833-2844. (21) Megli, F. M.; Selvaggi, M.; Liemann, S.; Quagliariello, E.; Huber, R. Biochemistry 1998, 37, 10540-10546. (22) Saurel, O.; Cezanne, L.; Milon, A.; Tocanne, J. F.; Demange, P. Biochemistry 1998, 37, 1403-1410. (23) Cezanne, L.; Lopez, A.; Loste, F.; Parnaud, G.; Saurel, O.; Demange, P.; Tocanne, J. F. Biochemistry 1999, 38, 2779-2786. (24) Peng, S.; Publicover, N. G.; Airey, J. A.; Hall, J. E.; Haigler, H. T.; Jiang, D.; Chen, S. R.; Sutko, J. L. Biophys. J. 2004, 86, 145-151. (25) Fill, M.; Copello, J. A. Physiol. ReV. 2002, 82, 893-922. (26) Anderson, K.; Lai, F. A.; Liu, Q. Y.; Rousseau, E.; Erickson, H. P.; Meissner, G. J. Biol. Chem. 1989, 264, 1329-1335. (27) Lee, H. B.; Xu, L.; Meissner, G. J. Biol. Chem. 1994, 269, 13305-13312.

Langmuir, Vol. 22, No. 14, 2006 6303

Figure 1. Overview of the lipid bilayer formation apparatus. This apparatus consists of two chambers. A lipid bilayer is formed in a small hole (about 100 µm diameter) at the bottom of the upper chamber. The bottom surface of the lipid bilayer touches the agarosecoated cover glass at the bottom of the lower chamber. centrifugation in a HITACHI RP65 rotor at 190 000 × g for 1 h. After centrifugation, the sample was quickly frozen and stored at -80 °C. Microscopy. The optical system has been previously described in detail.7,8 Figure 1 shows a schematic overview of the lipid bilayer formation apparatus. This apparatus consists of two chambers. Lipid bilayers were formed in a small hole (about 100 µm diameter) at the bottom of the upper chamber. The lipid bilayers touched the agarose-coated cover glass at the bottom of the lower chamber. An oil immersion objective lens (PlanApo, ×100, 1.4 NA, OLYMPUS, Tokyo, Japan) was located just below the lower chamber. Bilayers were illuminated by an evanescent wave with a YAG laser (JDS Uniphase, CA) and a He-Ne Red laser (Melles Griot, CA) according to the method of Tokunaga et al.28,29 Images were recorded using an image intensifier (VS4-1845, Video Scope International, VA) and an electron bombarded CCD camera (C7190-20, Hamamatsu Photonics, Shizuoka Japan), and stored on digital videotape. Video sequences were imported and analyzed by a personal computer using image analysis software. Images were acquired every 33 ms. Measurement Conditions. Lipids contained phosphatidylethanolamine, phosphatidylserine, and phosphatidylcholine at a 5:3:2 ratio (50 mg/mL phospholipids in n-decane). BODIPY 530/550 1,2dihexadecanoyl-sn-glycero-3-phosphoethanolamine (DHPE) (Invitrogen, CA) 0.05% was mixed with the lipids. The bilayer was formed by placing lipids at the hole of the upper chamber. Purified RyR2 with Cy5-conjugated antiRyR2 was added to the upper chamber. Annexin A5 was added to both chambers when BODIPYDHPE was monitored and only the lower chamber when Cy5RyR2 was monitored. Measurements were performed more than 30 min after the addition of annexin A5. The solution contained 250 mM KCl, 20 mM Pipes pH 7, 2 mM CaCl2, and 0.1 mM EGTA in both chambers when BODIPY-DHPE was monitored and in the lower chamber when Cy5-RyR2 was monitored. When Cy5-RyR2 was observed, the upper chamber solution contained 250 mM KCl, 20 mM Pipes pH 7, 0.3 mM CaCl2, and 0.1 mM EGTA. BODIPYDHPE and Cy5-RyR2 were excited with green (532 nm) and red (633 nm) lasers, respectively. These experiments were performed at room temperature. Current records were taken with a patch clamp amplifier (CEZ2400, Nihon Kohden, Japan), digitized at 10 kHz, and recorded with a personal computer. The data were filtered at 1 kHz and analyzed with commercial software (pClamp 9, Axon). Data Analysis. The mean square displacement (〈r2〉) averaged over a trajectory at each time interval (∆t) was calculated from the trajectory of a particle. The diffusion constant, D, was calculated from the slope of the ∆t - 〈r2〉 plot by least-squares fitting. (28) Tokunaga, M.; Kitamura, K.; Saito, K.; Iwane, A. H.; Yanagida, T. Biochem. Biophys. Res. Commun. 1997, 235, 47-53. (29) Wazawa, T.; Ueda, M. AdV. Biochem. Eng. Biotechnol. 2005, 95, 77106.

6304 Langmuir, Vol. 22, No. 14, 2006

Figure 2. Imaging single BODIPY-DHPE and Cy5-antiRyR2 conjugated RyR2 (Cy5-RyR2) molecules. Arrows indicate the position of each molecule.

Results Single-Molecule Detection of BODIPY-DHPE and Cy5RyR2. For single-molecule detection of 1,2-dihexadecanoylsn-glycerol-3-phosphoethanolamine (DHPE) and RyR2, each molecule was labeled with either the fluorescent dye BODIPY or Cy5, respectively. Lipids, which contained phosphatidylethanolamine, phosphatidylserine, and phosphatidylcholine in the ratio of 5:3:2 (50 mg/mL lipid in decane) and 0.05% BODIPYDHPE, was used to form a lipid bilayer at the bottom hole of the upper chamber (see methods). Cy5-RyR2, which was reconstituted in phospholipid vesicles, was added into the upper chamber and incorporated into the lipid bilayers. Fluorescence spots representing single RyR2 and DHPE molecules excited at each wavelength are shown in Figure 2. The fluorescence intensity for each spot was measured. The decay in fluorescence intensity for each spot of DHPE occurred in a single-step manner, indicating that the observed fluorescent signal was caused by a single fluorescent molecule (Figure 3A). The cumulative fluorescence intensity of the spot was plotted into a histogram (Figure 3B). The distribution of the fluorescence intensities for each spot showed a single peak, which was well fitted with a single Gaussian distribution. Taking into account these observations, we concluded that each fluorescent spot represented a single DHPE molecule and that single DHPE molecules were conjugated to BODIPY with a dye-to-lipid ratio of 1:1. Because RyR2 is considered to exist as a homotetramer channel, we estimated the dye-to-protein

Ichikawa et al.

ratio to be 15-20:1 for Cy5-RyR. The Cy5-RyR2 concentration in lipid vesicles used in our experiments was suppressed in order to visualize single or a few receptor proteins within the lipid bilayer. To confirm that single channels were observed, the number of channels in a lipid bilayer was determined by measuring electrical current. Figure 3C shows a single current, while only one spot was observed. Figure 3D is the histogram of the current distribution of Figure 3C. This distribution showed two peaks, indicating an open and a closed state. The distribution was well fitted with a second Gaussian distribution, confirming that one spot represents one channel. Annexin A5 Decreases the Diffusion of Lipid and Channel Molecules in the Lipid Bilayer. Lateral diffusion of single BODIPY-DHPEs and Cy5-RyR2s were monitored in the absence and presence of 0.125, 0.25, 0.5, and 1 µM annexin A5. Typical trajectories of 30 consecutive points recorded at 30 Hz (1 s) are shown in Figure 4. Upper and lower traces are trajectories of BODIPY-DHPEs and Cy5-RyR2s, respectively. Cumulative data of the diffusion coefficient and the histogram of the step size distribution of BODIPY-DHPE and Cy5-RyR2 at each annexin A5 concentration are shown in Figure 5A and B, respectively. The shape of the distribution became sharp with an increasing concentration of annexin A5. This means annexin A5 reduced the step size. The calculated diffusion coefficients for BODIPY-DHPEs and Cy5-RyR2s are listed in Table 1. The reduction rate is defined as the ratio of the diffusion coefficient at no annexin A5 and the diffusion coefficient at a given annexin A5 concentration (0.125, 0.25, 0.5, or 1 µM). Annexin A5 decreased the diffusion coefficients of BODIPY-DHPEs and Cy5-RyR2 in a concentration-dependent manner. In the absence of annexin A5, the diffusion coefficients of BODIPY-DHPE and Cy5-RyR2 were 4.81 ( 2.05 × 10-8 cm2/s (n ) 25) and 2.13 ( 0.809 × 10-8 cm2/s (n ) 24), respectively. In the presence of 1 µM annexin A5, the diffusion coefficients of BODIPYDHPE and Cy5-RyR2 were 2.23 ( 2.41 × 10-10 cm2/s (n ) 18) and 0.95 ( 1.49 × 10-10 cm2/s (n ) 25), respectively. It was revealed that 1 µM annexin A5 decreases the diffusion of PE and RyR2 molecules to 1/200th of their values in the absence of annexin A5. Influence of Annexin A5 on RyR2 Channel Properties. The ability of annexin A5 to limit RyR2 diffusion could make

Figure 3. Analysis of fluorescence intensity of a single BODIPY-DHPE and Cy4-RyR2 molecules. (A) A typical fluorescence intensity change of single BODIPY-DHPE in a lipid bilayer. The spot was photobleached in a single step, suggesting a single molecule. (B) The histogram of the distributions of cumulative fluorescence intensity of the spot in Figure 3A. (C) Single-channel current recording while one spot was observed at +50 mV holding potential. (D) Histogram of the distributions of current amplitude.

Immobilizing Single Channels with Annexin A5

Langmuir, Vol. 22, No. 14, 2006 6305

Figure 4. Typical trajectories of single BODIPY-DHPC and Cy5RyR2 molecules in the lipid bilayers. Thirty data points were recorded at 33 ms intervals. Upper and bottom traces show BODIPY-DHPE and Cy5-RyR5 trajectories, respectively. Lipids consisted of 5:3:2 PE/PS/PC (50 mg/mL in decane). Diffusion of BODIPY-DHPE and Cy5-RyR2 decreases with an increase in annexin A5.

it a useful tool for optical analyses of the function of RyR2 channels reconstituted in lipid bilayers. However, it is not clear whether annexin A5 alters the biological properties of the RyR2 channel. In this study, we further tested the function of RyR2 channels in the absence and presence of annexin A5. We confirmed that the conductance for the K+-conducting channel was constant with or without 1 µM annexin A5 at different holding potentials (-50, -25, 0, 25, and 50 mV). An example of an electrical current trace of single RyR2s at +25 mV in the absence and presence of 1 µM annexin A5 is shown at the top of Figure 6. The amplitude and open probability (Po) are shown at the bottom of Figure 6. Po in the absence and presence of 1 µM annexin A5 were 0.56 and 0.57, respectively. Although further cumulative studies may be required to investigate the subtle effects of annexin A5 on the RyR2, in our settings, we could not find any significant influence of annexin A5 on the function of the RyR2. Thus, annexin A5 can be used in our experimental system, which aims at revealing the relationship of ligand binding and the change of current.

Discussion Annexin A5 Decreases the Diffusion of the Lipid Bilayer in a Concentration-Dependent Manner. In this study, we monitored single BODIPY-1,2-dihexadecanoyl-sn-glycerol-3phosphoethanolamine (DHPE) and Cy5-labeled ryanodine receptor type 2 (RyR2) channels using an optical lipid bilayer system in the absence and presence of annexin A5. The diffusion coefficients of BODIPY-DHPE and Cy5-RyR2 in the absence of annexin A5 in the lipid bilayer containing phosphatidylethanolamine, phosphatidylserine, and phosphatidylcholine in the ratio of 5:3:2 (50 mg/ml phospholipids in n-decane) were calculated as 4.81 ( 2.05 × 10-8 cm2/s (n ) 25) and 2.13 ( 0.809 × 10-8 cm2/s (n ) 24), respectively. On the other hand, the diffusion coefficients of BODIPY-DHPE and Cy5-RyR2 in the presence of 1 µM annexin A5 were 2.23 ( 2.41 × 10-10 cm2/s (n ) 18) and 0.95 ( 1.49 × 10-10 cm2/s (n ) 25), respectively. Figure 5A shows the diffusion coefficients for BODIPY-DHPE and Cy5-RyR2 decrease with an increase in annexin A5 concentration. It was observed that 1 µM annexin A5 decreases the diffusion of PE and RyR2 molecules to 1/200th of their values in the absence of annexin A5. Figure 5B shows the diffusion step-size distribution of BODIPY-PE and Cy5RyR2 under different annexin A5 concentrations. Step size shortens with increasing annexin A5 concentration. In this report, we introduce a method for simultaneous optical and electrical recording of single ion channels. Our next challenge

Figure 5. Cumulative data showing effects of annexin A5 on the lateral diffusion of single PE and RyR2 molecules in bilayers. (A) The diffusion coefficient of PE and RyR2 at each annexin A5 concentration. Circles and triangles show the diffusion coefficients of BODIPY-PE and Cy5-RyR2, respectively. Annexin A5 decreases the diffusion coefficient of both PE and RyR2 in a concentrationdependent manner. (B) The diffusion step size distribution for 33 ms of PE and RyR2 at each annexin A5 concentration. This graph shows that annexin A5 reduces the number of large step size with an increasing concentration of annexin A5.

is to develop a method and technology for simultaneous optical and electrical measurement of single drug bindings to single ion channels. For this purpose, we have to immobilize channel proteins in a membrane because the lateral diffusion of channels will likely interfere with the optical measurement. For example, excited fluorophores attached to proteins and drugs with strong light leads to photobleaching of the dyes in a rather short period. This means the trace of the channel is lost before drug binding. Peng et al. reported an excellent way of immobilizing RyR2 in artificial membranes using annexins. However, their method, in which they determined a position of RyR channels as the center of Ca2+ flux, took 3 s to obtain a single frame and therefore cannot be applied to optical detection of single drug bindings to ion channels because of the low spatial and temporal resolution. We have realized higher spatiotemporal resolution by applying annexin to our electrooptical single-channel recording system.

6306 Langmuir, Vol. 22, No. 14, 2006

Ichikawa et al.

Table 1. Diffusion Coefficients and Reduction Rates for BODIPY-DHPEs and Cy5-RyR2sa annexin (µM) 0 0.125 0.25 0.5 1

DHPE D

(µm2

/s)

4.81 ( 2.05 (n ) 25) 0.798 ( 0.482 (n ) 20) 0.0902 ( 0.123 (n ) 19) 0.0649 ( 0.0431 (n ) 17) 0.0223 ( 0.0241 (n ) 18)

Cy5-RyR 2

reduction rate

D (µm /s)

reduction rate

6.03 53.3 74.1 215

2.13 ( 0.809 (n ) 24) 0.621 ( 0.333 (n ) 26) 0.396 ( 0.126 (n ) 27) 0.0365 ( 0.0545 (n ) 25) 0.0095 ( 0.0149 (n ) 25)

3.42 5.36 58.3 224

a Reduction rate is defined as the ratio of the diffusion coefficient at no annexin A5 and the diffusion coefficient at a given annexin A5 concentration (0.125, 0.25, 0.5, or 1µM)

Figure 6. Typical traces of single RyR2 channel recordings with and without annexin A5 and the histograms of each trace. Experiments were performed at a holding potential of +25 mV. The histograms of each trace are shown below. Open probabilities (Po) of RyR2 with and without 1 µM annexin A5 were 0.56 and 0.57, respectively.

To explain how annexin A5 decreases the lateral diffusion of lipid and channel protein molecules in lipid bilayers in a concentration-dependent manner, we propose the following model. Lipid and channel protein molecules diffuse in the area between the complexes of annexin A5 and the lipid. When the concentration of annexin A5 is increased, the area in which lipid and channel protein molecules can freely diffuse is narrowed because the area of the complexes of annexin A5 and lipid increases. Consequently, the diffusion coefficients of the lipid and channel protein molecules decreased in an annexin A5 concentration-dependent manner. From a recent study with AFM, it was seen that annexin A5 expands gradually, encompassing a greater area, consistent with our model. 17 Additionally, in the presence of 1 µM annexin A5, only in one chamber did we simultaneously observe two types of diffusion of BODIPY-DHPE: stopped and rapidly moving spots (data not shown). On the other hand, in the presence and absence of 1 µM annexin A5 in both chambers, we observed only one type of diffusion, either stopped or rapidly moving spots, respectively. This result indicates that annexin A5 can access only monolayers, consistent with Isas et al.30 Influence of Annexin A5 on Other Channels. The single RyR2 molecule electrophysiological results with and without 1 µM annexin A5 show that annexin A5 has minimal effect on either the open probability (Po) or conductance (Figure 6). In contrast, Eskesen et al. reported that annexin A5 decreases the conductance of gramicidin A, concluding diffusion resistance increases in the gramicidin A channel mouth.31 This is likely because gramicidin A has a small cylindrical radius in the transmembrane region, meaning annexin A5 binding to the lipid can easily access the ion pathway of this peptide channel. Consequently, ion permeation is prevented by annexin A5 and (30) Isas, J. M.; Cartailler, J. P.; Sokolov, Y.; Patel, D. R.; Langen, R.; Luecke, H.; Hall, J. E.; Haigler, H. T. Biochemistry 2000, 39, 3015-3022. (31) Eskesen, K.; Kristensen, B. I.; Jorgensen, A. J.; Kristensen, P.; Bennekou, P. Eur. Biophys. J. 2001, 30, 27-33.

the conductance of this channel decreases. On the other hand, the cylindrical radius of the RyR2 trans membrane region of RyR2 is 10 times as large as that of gramicidin A, therefore, the conductance of RyR2 may not be affected by the complexes of annexin A5 and the lipid, explaining the contrast in our results involving RyR2 and those from other groups investigating gramicidin A. Annexin Can be Used to Analyze RyR2 Properties in the Lipid Bilayer System. Our goal is to establish a simultaneous electrical and optical single ion channel observation system and identify a correlation between ligand binding to the channel protein and electrical current change. Our group has already succeeded in simultaneously imaging and measuring electrical currents of single ion channels.7-9 As a next step, we aim to unravel the correlation between ligand binding and current changing of an ion channel. One of the big obstacles to achieve this goal is the deviation of channel proteins from the microscopic field because channel proteins diffuse laterally in the lipid bilayer. Although it is possible to enlarge the visible field of the microscope and include the whole field of the membrane, it was difficult to detect single molecules in the field including the lipid bulk layer because the bulk layer scattered the laser. Because the size of the lipid bilayer is larger than the size of the microscopic field, channel proteins in the lipid bilayer must be immobilized. Regarding methods for immobilizing channel proteins in the lipid bilayer, two methods are proposed. One involves tethering the channel proteins to the surface of the glass slip. The other is to reduce the fluidity of the lipid bilayer membrane. In the tethering method, Ide et al. suggested the possibility of tethering the lipid with biotinylated antibody to a glass surface coated with biotinylated poly(ethylene glycol).32 However, the method of decreasing the fluidity of the lipid bilayer membrane using annexin A5 has proven to be simpler. In this method, all an experimenter needs is to mix annexin A5 with buffer. Furthermore, purified annexin A5 is commercially available. In this paper, we reveal that annexin A5 decreases the diffusion of fluidity of lipid bilayers as well as the diffusion of RyR2 by tracking single molecules directly. We also observed the electrophysiological influence of annexin A5 on RyR2 and showed annexin A5 has little effect on the function of RyR2. These results indicate that annexin A5 advances the establishment of simultaneous electrical and optical observation systems. Furthermore, these results also have the potential to serve as an important tool for limiting lateral diffusion of membrane proteins in a lipid bilayer system.

Conclusions We have demonstrated that annexin A5 decreases the diffusion coefficient of single lipid and channel protein molecules in a concentration-dependent manner by using an optical lipid bilayer system. The diffusion coefficients of BODIPY-DHPE and Cy5RyR5 in the absence of annexin A5 were calculated as 4.81 ( (32) Ide, T.; Takeuchi, Y.; Aoki, T.; Tabata, K.; Noji, H. e-J. Surf. Sci. Nanotechnol. 2005, 3, 70-73.

Immobilizing Single Channels with Annexin A5

2.05 × 10-8 cm2/s (n ) 25) and 2.13 ( 0.809 × 10-8 cm2/s (n ) 24), respectively. On the other hand, the diffusion coefficients of BODIPY-DHPE and Cy5-RyR2 in the presence of 1 µM annexin A5 were 2.23 ( 2.41 × 10-10 cm2/s (n ) 18) and 0.95 ( 1.49 × 10-10 cm2/s (n ) 25), respectively. It was revealed that 1 µM annexin A5 decreases the diffusion of PE and RyR2 molecules to 1/200th of their values in the absence of annexin A5. Furthermore, we measured the K+ current of single RyR2 in the presence of 1 µM annexin A5 and found that annexin A5 does not have much effect on RyR2 function under the conditions

Langmuir, Vol. 22, No. 14, 2006 6307

used in the present studies. These results indicate the potential to serve as an important tool for limiting lateral diffusion of membrane proteins in lipid bilayer systems. Acknowledgment. The authors thank our colleagues at Osaka University for valuable discussions and Dr. S. C. Shibata and Dr. P. Karagiannis for carefully revising the manuscript. This work was supported by Japan Science and Technology Agency. LA0535025