942
Bioconjugate Chem. 2004, 15, 942−947
Bacteriorhodopsin Conjugates as Anchors for Supported Membranes Manoj K. Sharma, Harsha Jattani, and M. Lane Gilchrist, Jr.* Departments of Chemical and Biomedical Engineering, The City College of the City University of New York, 140th Street and Convent Avenue, New York, New York 10031. Received December 26, 2003; Revised Manuscript Received April 23, 2004
The sophistication of supported lipid bilayer membranes has increased steadily as new applications are being explored. In general, tethered lipids are used to anchor the lipid bilayer to the substrate. Here we describe a new type of anchoring system for supported lipid bilayers that is based on biotinPEG3400-bacteriorhodopsin conjugates. Amine-based coupling was used to construct the polymer conjugates, followed by fluorophore labeling to enable confocal imaging. The bacteriorhodopsin-based anchoring system was used to construct solid-supported vesicles from streptavidin-coated microspheres. This method could provide a new route for the stability enhancement of supported lipid bilayer membrane assemblies.
The means of constructing supported membranes for the immobilization of membrane proteins has progressed rapidly from simple physical adsorption of lipid bilayers to more sophisticated approaches involving polymer tethering (1-6). In the cases involving tether-supported membranes, a wide variety of tethering molecules have been employed, ranging from polymers to peptides (1, 5). Thus far, primarily single lipid moieties connected at the end of tethers have been used to anchor membranes to solid supports (7-10). In the present study, our aim was to construct biotin-PEG3400-bacteriorhodopsin (bR) conjugates and use them to anchor lipid bilayer membranes to streptavidin-coated microspheres.1 As our ultimate goal is to construct robust, supported membrane particles for immobilization of membrane proteins such as biocatalysts, the potential benefits of bacteriorhodopsin anchoring are multifold and linked to greater control over the stability of supported bilayers. Shown schematically in Figure 1, the use of bacteriorhodopsin (26.5 kDa without retinal) as a building block provides a means for more extensive anchor-membrane interactions relative to lipid-mediated tethering. The energetics of the removal of the tethered lipophilic moiety from the membrane are expected to be more highly unfavorable for the case of bR relative to a lipid anchor. This can be argued on the basis of anchor to membrane interaction surface area differences. Supported membrane assemblies on surfaces and particles fabricated in this fashion should exhibit greater stability under shear or processing flows. In addition, due to the increase in anchoring efficiency that could be obtained from the bR anchor, it is conceivable that fewer tethering points would be needed for a given supported membrane area. This would allow for greater native lipid areas for the intro1 Abbreviations: bR, bacteriorhodopsin; PEG, polyethylene glycol; NHS, N-hydroxylsuccinimide; SPA, succinimidyl propionate; DMSO, dimethyl sulfoxide; TR, Texas Red; POPC, L-Rphosphatilylcholine; DOPG, 1,2-dioleoyl-sn-glycero-3-[phosphorac-(1-glycerol)] (sodium salt); OG, octyl β-glucoside; SSV, solidsupported vesicle or spherically supported vesicle; MALDI-TOFMS, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry.
Figure 1. Schematic diagram of solid-supported vesicles that integrate bacteriorhodopsin conjugates as lipid bilayer membrane anchors.
duction of membrane proteins into supramolecular assemblies of this type. A further benefit of bR anchoring is that reactive sites on the outer surface of the bR opposite the tether can be exploited to further stabilize the assemblies via cross-linking or by ligating large hydrophilic polymers such as glycans. There are a few precedents related to the present study that have involved bacteriorhodopsin or rhodopsin conjugation that warrant mention here. In one such study, Sirokman and Fasman explored the amine-based PEGylation of bR in conferring water solubility to the protein, providing an analysis of likely conjugation sites (11). In addition, their studies indicated that PEG5000-bR retained similar proton-pumping activity relative to that
10.1021/bc034231h CCC: $27.50 © 2004 American Chemical Society Published on Web 07/03/2004
Technical Notes
of unlabeled bR when the conjugate was reconstituted into lipid bilayers. In more recent work, Puu et al. utilized biotinylation of bR to allow for the localization of the protein in surface-deposited proteoliposomes, which was verified via AFM imaging of streptavidin-colloidal gold probes (4). Also related to the present study is the work of Bieri et al. in which rhodopsin that was biotinylated at the extracellular glycan moiety was used to tether the protein in a specific orientation in a biosensor microarray (12). While these studies demonstrate the feasibility of conjugation of bR and tethering via closely related rhodopsin, our work explores the first use of bR as a building block for a potentially improved means to anchor supported membranes. This paper describes the construction and use of biotin-PEG3400-bacteriorhodopsin conjugates in the formation of solid-supported membranes on particles. We have used amine-based coupling to conjugate biotinPEG3400 to bR. The conjugates were characterized using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS) after purification. The PEG-bR conjugate was further labeled with Texas Red to facilitate localization via fluorescence imaging. Supported membranes containing biotinPEG3400-bacteriorhodopsin were constructed through self-assembly onto streptavidin-coated beads to form solid-supported vesicles (SSVs). Confocal microscopy was used to image these supramolecular complexes, establishing the use of biotin-PEG3400-bacteriorhodopsin as a membrane anchor. MATERIALS AND METHODS
Materials. Bacteriorhodopsin was obtained as lyophilized powder of the Purple Membrane from Halobacterium salinarum strain S9 (Munich Innovative Biomaterials (MIB) GmbH, Germany). Biotin-PEG-SPA (MW 3400) was obtained from Nektar Therapeutics, Huntsville, AL. OG was obtained from Pierce Biotechnology Inc., Rockford, IL. Texas Red-X, succinimidyl ester, and β-BODIPY 530/550 C5-HPC were obtained from Molecular Probes, Eugene, OR. Sephadex G-50 was obtained from Amersham Biosciences Inc, Piscataway, NJ. Streptavidin-coated silica beads (3.18 µm diameter) were obtained from Bangs Laboratories Inc., Fishers, IN. Lipids POPC (egg, chicken), DOPG, and cholesterol (>98% pure) were purchased from Avanti Polar Lipids, Inc. Alabaster, AL. Bio-Beads SM-2 adsorbents were obtained from BioRad Laboratories, Hercules, CA. Sequencing-grade trypsin (Bovine Pancreatic) was obtained from Roche Allied Science, Indianapolis, IN. RapiGest was purchased from Waters Corp., Waltham, MA. Labeling of bR with Biotin-PEG and Texas Red. Purple membrane (3.45 mg bR) and biotin-PEG-SPA (38.2 mg) were each dissolved in 200 µL of 50 mM Na2HPO4 buffer stocks, pH 9.0 with 100 mM OG. The two solutions were mixed together and incubated at room temperature with gentle shaking for 2 h. TR-X succinimidyl ester stock (16.7 mg/mL) was prepared in DMSO and 30 µL was added to 200 µL of the reaction mixture. The remaining reaction mixture was saved for SDSPAGE and MALDI-TOF-MS analysis after the addition of 15 µL of hydroxylamine (1.5 M) stop reagent. TR-Xcontaining reaction mixture was further incubated in the dark at room temperature for 2 h before the addition of 15 µL of hydroxylamine (1.5 M) stop reagent. Excess reactants were separated over a Sephadex G-50 column (1 cm × 50 cm) equilibrated with 50 mM Na2HPO4 buffer
Bioconjugate Chem., Vol. 15, No. 4, 2004 943
without any OG. bR fractions (labeled and unlabeled) collected from the first peak were pooled together and concentrated using Amicon Ultra-15 filters (30 kDa cutoff, 2800g, 30 min). bR aggregates formed on filter membrane were resolubilized using 100 mM OG. To achieve greater purification, traces of excess fluorescent probe TR and unattached biotin-PEG were further removed by another G-50 column. bR fractions from the first peak were pooled together again and exchanged with Hepes buffer A (Hepes, 20 mM; EDTA, 1 mM; NaCl, 100 mM, pH 7.5) using Amicon Ultra-15 filters (2800g, 30 min) in three rinse-spin cycles. The final retentate contained both labeled and unlabeled bR, was redissolved in 100 mM OG (25 µM protein, 600 µL total volume), and was stored at 4 °C until required. Tethering of TR-Labeled Biotin-PEG-bR to Streptavidin-Coated Silica Microspheres. A 200 µL microsphere stock solution (∼2 mg microspheres, biotin binding capacity 0.032 µg/mg microspheres) was washed three times with Hepes buffer A with 100 mM OG and centrifuged (200g, 10 min). After the third spin cycle the microspheres were mixed with 50 µL of the stock containing TR and biotin-PEG-labeled bR and incubated at room temperature for 3 h with gentle shaking. BiotinPEG-bR tethered beads were centrifuged (200g, 10 min) to remove excess unattached bR complexes and rinsed with Hepes buffer A containing 100 mM OG. This washspin cycle was repeated three times to completely remove any nonspecifically adsorbed molecules from the microsphere surface. Formation of Solid-Supported Vesicles. POPC, DOPG, and cholesterol were mixed in molar ratio 9:1:7 in chloroform (total lipid concentration 10 mg/mL). Green β-BODIPY 530/550 C5-HPC lipid probe was also added to the mixture at a ratio of 1:500 (probe:lipid). The lipid solution was subdivided in 250 µL parts in 4 mL vials and solvent was evaporated using a vacuum. The vials containing the lipid films were then stored under argon at -20 °C until required. For SSV preparation, one lipid formulation containing 2.5 mg of lipids was dissolved in 500 µL of Hepes buffer A with 100 mM OG. This lipid suspension was then used to resuspend the centrifuged biotin-PEG-bR tethered beads that were fabricated in previous steps. The mixture was incubated at 5 °C for 1 h. To remove detergent and form SSVs, 80 mg of wet BioBeads was added to the mixture kept at 5 °C with gentle shaking. After 3 h, a fresh lot of 80 mg of Bio-Beads was added and mixture was incubated for additional 2 h. BioBeads were then decanted from the solution containing regular and solid-supported lipid vesicles (SSVs). SSVs were centrifuged (200g, 10 min) and washed with Hepes buffer A. Other Methods. MALDI-TOF mass spectrometry analysis was conducted at the Columbia University Protein Core Facility on a Perceptive Biosystems delayedextraction reflector instrument. Trypsin digestion was carried out overnight at 37 °C in 0.5% RapiGest SF, at concentrations of bR and sequencing-grade trypsin of 0.5 and 0.05 mg/mL, respectively. The detergent was removed via ultrafiltration and a 10 min, 100 °C denaturing step preceded the digestion. SDS-PAGE was done on Bio-Rad Mini PROTEAN II apparatus using 4-15% Linear Gradient Ready Gel Tris-HCl Gels. Bio-Safe Coomassie stain from Bio-Rad Laboratories was used to detect protein bands. Gel densitomograms were obtained using a 16 bit Umax flatbed scanner, with image processing carried out using Igor Pro 4.0 (Wavemetrics, Inc). UV-Vis analysis was done on a Hitachi U-3010 spectrophotometer. The extent of fluorophore labeling was
944 Bioconjugate Chem., Vol. 15, No. 4, 2004
Technical Notes
determined by comparison of the Texas Red absorbance relative to protein concentration. Samples were imaged using a Leica TCS-SP four channel confocal microscope at the Mt. Sinai School of Medicine, using 488 nm excitation for β-BODIPY lipid probe and 568 nm for the Texas Red. The transmittance confocal channel was used to position the imaging plane at the center of the microspheres. Image analysis of confocal data was carried out with Igor Pro 4.0 (Wavemetrics, Inc). RESULTS AND DISCUSSION
Our strategy to conjugate bacteriorhodopsin for anchoring membranes was adapted from the earlier work of Sirokman and Fasman (11) with a distinctly different aim. In our application, predominantly single-site labeling is desired. Furthermore, low yield is tolerable, because only the biotinylated bR conjugates will be assembled into the supramolecular complexes. Therefore, in contrast to the earlier bR PEGylation study, in which moderate double labeling was obtained, we did not employ two reaction cycles nor include nonreactive PEG to increase the reactivity of the protein (11). The first step of the conjugation was carried out at pH 9 at a 90fold molar excess of biotin-PEG3400-SPA. To enable fluorescence imaging of the biotin-PEG3400-bR conjugates, an amine-based labeling (at 5-fold molar excess) with Texas-Red succinimidyl ester was conducted after PEGylation, also at pH 9. Separation of the excess reactants was carried out using gel filtration followed by ultrafiltration to remove all traces of Texas Red dye, thus minimizing artifacts in the confocal imaging studies. Figure 2 displays MALDI-TOF-MS traces of the bR starting material (trace A), the bR-PEG3400-biotin conjugate (trace B), and the biotin-PEG3400-bR-Texas Red conjugate (trace C). The inset of Figure 2 displays SDSPAGE densitomograms that correspond to the MALDITOF-MS traces: bacteriorhodopsin starting material (trace A′), the bR-PEG3400 conjugate (trace B′), and the molecular weight markers (trace M). Traces B (MALDITOF-MS) and B′ (SDS-PAGE) each contain two peaks, a lower molecular weight peak from the unreacted bR and an additional peak of higher molecular weight that we assign to the conjugate. In trace B, we assign the conjugate peak (denoted by an asterisk (*)) to the broad band located approximately 3400 Da higher in molecular weight. A very weak band that appears to correspond to bR-(PEG3400-biotin)2 is evidenced near 34 100 Da in trace B. We assign this feature to a small amount of double labeling of bR. As MALDI-TOF characterization does not provide direct quantitative information about conjugate yield due to differences in desorption of various species, SDS-PAGE studies were also employed. SDSPAGE densitomograms were obtained for separate bioconjugation trials. Multipeak fitting gave estimated yields of the bR-PEG3400-biotin conjugates that ranged from 13 to 25%. Figure 3 shows a schematic of the amino acid sequence of bR, with reference to the secondary structures of bR as determined by X-ray crystallography in recent work by Belrhali et al. (1999; PDB 1qhj) (13). The seven lysine residues are indicated by the circles, and the residues that comprise the transmembrane helices are located inside the rectangles. We note that the N-terminus is posttranslationally modified and does not contain a primary amine (14). The most likely polymer conjugation sites are Lys 129 and Lys 159 (indicated by arrows), as these are found on loops interconnecting the transmembrane helices and are solvent accessible to bulky reactants such as the SPA-PEG3400-biotin employed here
Figure 2. MALDI-TOF-MS and SDS-PAGE gel densitometry results for the biotin-PEG3400-bacteriorhodopsin conjugate. Trace A was obtained from the bR starting material. Trace B was obtained from the biotin-PEG3400-bR conjugate. The peak denoted by the asterisk (*) in the main figure and inset is assigned to the biotin-PEG3400-bacteriorhodopsin conjugate. Trace C was obtained from the biotin-PEG3400-bR-Texas Red conjugate. In the inset, SDS-PAGE gel densitometry results are shown for samples corresponding to traces A and B of the MALDI-TOF-MS results. Trace A′ was obtained from the bR starting material. Trace B′ was obtained from the biotinPEG3400-bR conjugate. Trace M was from the molecular weight markers (20, 25, and 37 kDa, left to right).
(11). Furthermore, in aforementioned studies carried out by Sirokman and Fasman with amine-reactive PEG5000, chymotrypsin digestion studies were consistent with loop lysine labeling (11). Their findings were further bolstered by cogent arguments also based on the structure of bR. Early work with small-molecule labeling indicated that Lys 129 was the most likely site (15, 16). To address this question directly, we have performed a trypsin digestion of the bR-PEG3400-biotin conjugate, followed by MALDI-TOF-MS characterization of the resulting fragments. The degree of difficulty of trypsin fragmentation is especially high for membrane proteins such bR, as most cleavage sites are buried and, in addition, give rise to highly hydrophobic peptides (14). Furthermore, the addition of detergents to expose cleavage sites dramatically reduces trypsin activity. To obtain greater digestion yields, we employed a new, acidcleavable denaturant (RapiGest SF, Waters Corp.) that does not greatly inhibit proteases and allows for higher trypsin reaction yields. In our experiments, run over a wide set of conditions, including ranges of RapiGest SF concentration and trypsin:bR ratio, we obtained peptides with missed cleavages almost exclusively. Table 1 lists the tryptic digest fragments detected that contain either K129 or K159. The corresponding MALDI-TOF-MS traces are shown in Figure 4. Trace A is from bR (control) and traces B and B′ (3 × B) are from the bR-PEG3400-biotin conjugate. Prominent in trace A are the peaks at 17 578 and 24 856 m/z, assigned to 1-159 and 1-227, respectively, formed from loop site cleavage. Inspection of the bR-PEG3400-biotin conjugate trace magnified 3× shows
Technical Notes
Bioconjugate Chem., Vol. 15, No. 4, 2004 945
Figure 3. Schematic diagram of the structure of bacteriorhodopsin. The amino acid residues located in transmembrane helices are indicated by the rectangular boxes. Table 1. Trypsin Fragments of Bacteriorhodopsin Detected That Contain Potentially Accessible Lysines trypsin fragment no.
bR obsd (m/z)a
bR-PEG3400-biotin observed (m/z)a
T6-7
3 566
3 600
T5-7
8 479
8 481
T1-7
9 025
9 017
T1-7-PEG3400-biotin
10 826
T1-12
12 355
12 432
T3-7
13 262
13 307
T1-7
17 578
17 662
T1-7-PEG3400-biotin T1-12
21 200 24 856
T1-12-PEG3400-biotin
24 885 28 551
assigned sequence or conjugate (expected MW, species) 130-159 (3 645, [M + H]+) 83-159 (8 470, [M + H]+) 1-159 (9 034, [M + H]+2) 1-159-PEG3400-biotin (9 034, [M + H]+2) 1-227 (24 882, [M + H]+2) 41-159 (13 198, [M + H]+) 1-159 (17 487, [M + H]+) 1-159-PEG3400-biotin (20 870, [M + H]+) 1-227 (24 882, M+) 1-227-PEG3400-biotin (28 282, [M + H]+)
a The moderate errors in m/z relative to the expected MWs are attributed to external calibration of the spectra and some degree of amino acid side reactions such as methionine oxidation.
broad peaks shifted over by approximately 3400 Da from peaks assigned to sequences 1-159 and 1-227. These peaks are assigned to 1-159-PEG3400-biotin and 1-227PEG3400-biotin conjugated fragments. The doubly charged 1-159-PEG3400-biotin species was also detected as a broad peak centered at 10 826 m/z. We were unable to detect any other PEGylated species, including conjugates of 83-159 and 41-159, a consequence, presumably, of lower yields of intrahelical site cleavage that precluded detection. Due to the high yield of 1-159-PEG3400-biotin species in the trypsin digest, we infer that the predominant PEG3400-biotin conjugation site is K129, as PEG3400biotin labeling at K159 would be expected to inhibit trypsin due to steric hindrance. The other five lysines are thought to be less solvent accessible and are the likely sites of conjugation of the
hydrophobic Texas Red fluorophore. Evidence for successful fluorophore labeling was first seen in the SDSPAGE gels, as the bands from bR labeled with Texas Red generated a different hue due to the colorimetric properties of this chromophore combined with the bromophenol blue protein stain. MALDI-TOF-MS of the biotinPEG3400-bR-Texas Red conjugate is shown in trace C of Figure 2. Two broad peaks with complex line shapes were evidenced, centered at approximately 27 700 and 31 500 Da. These features are shifted by over 600 Da from the corresponding peaks arising from bR and the bR-PEG3400-biotin conjugate (found in trace B). These shifts are consistent with the successful conjugation of Texas Red (ligated molecular weight 702 Da) to a fraction of both the unreacted bR starting material and the bRPEG3400-biotin conjugate, giving rise to broad composite
946 Bioconjugate Chem., Vol. 15, No. 4, 2004
Technical Notes
Figure 4. MALDI-TOF-MS results for the tryptic digest reaction products of the biotin-PEG3400-bacteriorhodopsin conjugate. Trace A was obtained from a tryptic digest of the bR starting material. Trace B was obtained from a tryptic digest of biotin-PEG3400-bR conjugate, conducted under the same conditions as in trace A. For closer examination, the region from 15 000 to 32 000 of trace B is shown magnified three times in the top trace.
peaks comprised of fluorophore-labeled and unlabeled molecules. Furthermore, UV-Vis analysis of the conjugate revealed that approximately 1-2 Texas Red fluorophores were conjugated to each protein molecule. This was based on protein concentration determined by UV absorption, combined with the Texas Red absorption intensity of the conjugate. To demonstrate the use of bR in membrane anchoring, we have constructed solid supported vesicles (SSVs) based on streptavidin-coated silica microspheres (nominal size 3.18 µm). The first step in SSV assembly is to bind the biotin-PEG3400-bR-Texas Red conjugates to the microspheres via biotin-streptavidin complex formation. In these experiments, the biotinylated conjugates were added in excess of a 1:1 molar ratio of the total number of streptavidin sites (from manufacturer’s specifications) to aim for complete coverage. Removal of excess bR and bR conjugates is facilitated by the ease of settling the particles by low-speed centrifugation. To demonstrate formation of the supported bilayers on the particles, a green fluorescent phosphatidylcholine analogue (β-BODIPY 530/550 C5-HPC) was incorporated into the supported bilayer phospholipids. This reagent was incorporated in the formulation of negatively charged phospholipids (at a 1:500 probe:lipid ratio) that was combined with the bR-terminated particles. The phospholipids mixture also contained octyl glucoside detergent at 100 mM, which is above the concentration at which membranes become completely solublized and lipids form mixed micelles with detergent (17, 18). The final reconstitution steps were to form supported lipid bilayers on the particle surface upon removal of detergent by Bio-Beads (adsorbent), followed by centrifugation/washing steps to remove excess regular liposomes.
Figure 5. Confocal microscopy of solid-supported vesicles (SSVs) doped with a green fluorescent phosphatidylcholine analogue (β-BODIPY 530/550 C5-HPC) anchored by biotinPEG3400-bR-Texas Red conjugates. The images were obtained in a single XY imaging plane (∼0.2 µm thickness) at the center of the particles. The top left particle is the nonspecific adsorption negative control, and the desired bR-anchored SSV particle is at the bottom right. The particles are each shown in three channels: β-BODIPY 530/550 (top panel A, excitation 488 nm, detection 500-570 nm), Texas red (middle panel B, excitation 568 nm, detection 580-660 nm), and reflectance mode (bottom panel C). Shown below the emission images are intensity profiles taken from horizontal slices bisecting the images at the positions of the arrows.
Confocal imaging results from a single XY plane obtained from bR-anchored, solid-supported vesicles (right, bottom) and a nonspecific adsorption control (left, top) are shown in Figure 5. These images were obtained from a region of the slide that contained both a (nonspecific adsorption) control particle and a desired bR-anchored SSV particle. The β-BODIPY 530/550 (green) emission
Technical Notes
image is shown in panel A and the Texas Red emission image is shown below in panel B. For reference, the reflectance mode confocal image is given in panel C. Intensity profiles, taken horizontally from the images at the position of the arrows, are shown immediately below each dye emission image. At this resolution, the bRanchored, solid-supported vesicle (right, bottom) exhibits colocalization of Texas Red and β-BODIPY 530/550 emissions in an approximately homogeneous ring around the particle. The presence of this pattern suggests that we have formed the bR anchored SSVs as intended in the design, with spatially overlapping emissions from the biotin-PEG3400-bR-Texas Red conjugate and β-BODIPY 530/550 from the lipid layer. As a negative control, we also probed the nonspecific adsorption of β-BODIPY 530/ 550 containing lipids to the streptavidin microspheres under conditions similar to bR-anchored, SSV formation. In this case no tethering to the microspheres was employed. In general, nonspecific adsorption of the green fluorescent lipids was evidenced on average at lower levels than for the desired bR-anchored, solid-supported vesicles. This is reflected in the comparison of the intensities of representative particles shown in cross section in panel A. Moreover, the control microsphere in panel B (left, top) shows no Texas Red emission, as expected. This attribute was used to distinguish SSV particles from control particles on the same slide. CONCLUSIONS
Bacteriorhodopsin-PEG3400-biotin conjugates were constructed via primary amine coupling. The conjugates were characterized with MALDI-TOF-MS; single-site labeling occurred predominantly under these conditions. In cases in which the bR-PEG3400-biotin conjugates were subsequently labeled with Texas Red for imaging, one or two sites per protein were tagged. Lysine 129, situated on an interhelical loop, was the most likely site for conjugation, as determined by trypsin digestion studies. The confocal imaging results provide direct evidence that we have formed supported bilayers anchored with bacteriorhodopsin conjugates over a majority of the surfaces of the particles. In addition, β-BODIPY 530/550 containing lipid was localized on the particle surface, providing further evidence for supported vesicle formation mediated by the bR anchor formation. The present use of bR as an anchoring moiety may provide a new direction in the goal of enabling greater control over the stability of supported membranes positioned onto particles or onto surfaces. Continuing with this approach, we aim to fabricate particles that provide highly robust, yet nativelike microenvironments for the immobilization of membrane proteins. Further studies are currently underway to assess enhancements in stability relative to lipid tethering and to construct assemblies with varying amounts of bilayer tethering. In parallel, we are working to co-localize other membrane proteins in bRanchored bilayers and examine retention of native function of the protein and activity in the assemblies. ACKNOWLEDGMENT
Confocal microscopy was performed at the Mt. Sinai Medical School-Microscopy Shared Research Facility, supported, in part, with funding from NIH-NCI shared resources grant (1 R24 CA095823-01). Mary Ann Gawinowicz of the Columbia University Protein Core Facility is gratefully acknowledged for vital guidance in the tryptic digestion and MALDI-MS analysis.
Bioconjugate Chem., Vol. 15, No. 4, 2004 947 LITERATURE CITED (1) Cheng, Y., Ogier, S. D., Bushby, R. J., and Evans, S. D. (2000) Discrete membrane arrays. J. Biotechnol. 74 (3), 15974. (2) Groves, J. T., Ulman, N., and Boxer, S. G. (1997) Micropatterning fluid lipid bilayers on solid supports. Science 275 (5300), 651-3. (3) Kalb, E., Frey, S., and Tamm, L. K. (1992) Formation of supported planar bilayers by fusion of vesicles to supported phospholipid monolayers. Biochim. Biophys. Acta. 1103 (2), 307-16. (4) Puu, G., Artursson, E., Gustafson, I., Lundstrom, M., and Jass, J. (2000) Distribution and stability of membrane proteins in lipid membranes on solid supports. Biosens. Bioelectron 15 (1-2), 31-41. (5) Sackmann, E. (1996) Supported membranes: Scientific and practical applications. Science 271 (5245), 43-8. (6) Tamm, L. K., and McConnell, H. M. (1985) Supported phospholipid bilayers. Biophys J 47 (1), 105-13. (7) Wagner, M. L., and Tamm, L. K. (2000) Tethered polymersupported planar lipid bilayers for reconstitution of integral membrane proteins: Silane-poly(ethylene glycol)-lipid as a cushion and covalent linker. Biophys. J. 79 (3), 1400-14. (8) Wagner, M. L., and Tamm, L. K. (2001) Reconstituted syntaxin1a/SNAP25 interacts with negatively charged lipids as measured by lateral diffusion in planar supported bilayers. Biophys. J. 81 (1), 266-75. (9) Shen, W. W., Boxer, S. G., Knoll, W., and Frank, C. W. (2001) Polymer-supported lipid bilayers on benzophenonemodified substrates. Biomacromolecules 2 (1), 70-9. (10) Kiessling, V., and Tamm, L. K. (2003) Measuring distances in supported bilayers by fluorescence interference-contrast microscopy: polymer supports and SNARE proteins. Biophys. J. 84 (1), 408-18. (11) Sirokman, G., and Fasman, G. D. (1993) Refolding and proton pumping activity of a poly(ethylene glycol)-bacteriorhodopsin water-soluble conjugate. Protein Sci. 2 (7), 116170. (12) Bieri, C., Ernst, O. P., Heyse, S., Hofmann, K. P., and Vogel, H. (1999) Micropatterned immobilization of a G protein-coupled receptor and direct detection of G protein activation. Nat. Biotechnol. 17 (11), 1105-8. (13) Belrhali, H., Nollert, P., Royant, A., Menzel, C., Rosenbusch, J. P., Landau, E. M., and Pebay-Peyroula, E. (1999) Protein, lipid and water organization in bacteriorhodopsin crystals: A molecular view of the purple membrane at 1.9 A resolution. Structure Fold Des. 7 (8), 909-17. (14) Gerber, G. E., Gray, C. P., Wildenauer, D., and Khorana, H. G. (1977) Orientation of bacteriorhodopsin in Halobacterium halobium as studied by selective proteolysis. Proc. Natl. Acad. Sci. U.S.A. 74 (12), 5426-30. (15) Henderson, R., Jubb, J. S., and Whytock, S. (1978) Specific labeling of the protein and lipid on the extracellular surface of purple membrane. J. Mol. Biol. 123 (2), 259-74. (16) Heberle, J., and Dencher, N. A. (1992) Surface-bound optical probes monitor protein translocation and surface potential changes during the bacteriorhodopsin photocycle. Proc. Natl. Acad. Sci. U.S.A. 89 (13), 5996-6000. (17) Paternostre, M. T., Roux, M., and Rigaud, J. L. (1988) Mechanisms of membrane protein insertion into liposomes during reconstitution procedures involving the use of detergents. 1. Solubilization of large unilamellar liposomes (prepared by reverse-phase evaporation) by Triton X-100, octyl glucoside, and sodium cholate. Biochemistry 27 (8), 266877. (18) Ollivon, M., Eidelman, O., Blumenthal, R., and Walter, A. (1988) Micelle-vesicle transition of egg phosphatidylcholine and octyl glucoside. Biochemistry 27 (5), 1695-703.
BC034231H
