Enzymatic Cross-Linking of Purple Membranes Catalyzed by Bacterial

c/o Technical University Darmstadt, D-64287 Darmstadt, Germany, and Fachbereich ... Technologie, Fachhochschule Darmstadt, Hochschulstrasse 2, D-64289...
1 downloads 0 Views 432KB Size
Download PDF
Biomacromolecules 2001, 2, 233-238

233

Enzymatic Cross-Linking of Purple Membranes Catalyzed by Bacterial Transglutaminase A. Seitz,† F. Schneider,† R. Pasternack,‡ H.-L. Fuchsbauer,§ and N. Hampp*,† Institute for Physical Chemistry, University of Marburg, D-35032 Marburg, Germany, N-Zyme BioTec, c/o Technical University Darmstadt, D-64287 Darmstadt, Germany, and Fachbereich Chemische Technologie, Fachhochschule Darmstadt, Hochschulstrasse 2, D-64289 Darmstadt, Germany Received October 2, 2000

It was found that bacterial transglutaminase (TGase) facilitates selective cross-linking of bacteriorhodopsin (BR) in purple membrane (PM) form under mild conditions. Fluorescent probes were used to detect that the membrane protein BR may act as a glutamine donor as well as a lysine donor for TGase. The binding sites were determined to be Gln-3 as the reactive glutamine, and Lys-129 is the corresponding lysine residue. Upon incubation of PM with TGase, cross-linking of PM patches can be achieved without an additional spacer molecule. To our knowledge, this is the first time that an intermembrane cross-linking of membranebound proteins is reported. Furthermore, this finding may provide the ability to achieve covalent linkage of complete purple membrane patches to synthetic polymers. 1. Introduction Bacteriorhodopsin (BR) is found as two-dimensional crystals, so-called purple membranes (PM), in Halobacterium salinarum. Due to its efficient photochemistry, its photoelectric and proton pumping properties and its excellent stability against chemical, thermal, and photochemical degradation, a technical utilization of BR was considered rather soon after its discovery. Up to this point many different proposals for possible technical applications have been published (for a review see, e.g., refs 1 and 2). For applications of BR that are based on its light-induced proton transport or its photoelectric properties, the oriented coupling of PM sheets onto a substrate and the ordered interlinkage of PM sheets have to be considered as key steps.3 Oriented covalent coupling of proteins to substrates may be achieved by genetic engineering of proteins, e.g., by introducing a cysteine residue or a histidine tag.4 The utilization of antibodies to achieve a controlled orientation of BR in PM form onto a surface was described first in ref 5. Both methods allow only monomolecular oriented layers of PM to be obtained. Chemical cross-linking of PM by bifunctional reagents such as glutardialdehyde6 or by fabricating functionalized substrates7 is not efficient due to (i) undesired side reactions of the low molecular cross-linking agents and (ii) their low site specificity. In addition to these disadvantages, unused coupling reagent has to be removed after the reaction. Transglutaminases have been used repeatedly to crosslink proteins under mild conditions. They catalyze the formation of N-(γ-L-glutamyl)-L-lysine isopeptide bonds8 in * To whom correspondence should be addressed: Tel.: [+49] (6421) 282 5775. Fax: [+49] (6421) 282 5798. E-mail: [email protected]. † University of Marburg. ‡ N-Zyme BioTec. § Fachhochschule Darmstadt.

a two-step reaction. In the first step the amide group of a glutamine is cleaved accompanied by the formation of an acyl-enzyme complex under deliberation of ammonium ions. The -amino group of the reaction partner lysine is then linked to the acyl-enzyme complex in the second step and the TGase is released. In the present work we describe the cross-linking of BR in membrane-bound form (PM sheets) with bacterial transglutaminase (TGase) from StreptoVerticillium mobaraense. The bacterial enzyme was selected because it differs from the vertebrate enzymes as it is not cofactor dependent9,10 and therefore provides the advantage that in a simple enzymatic procedure a highly selective covalent cross-linking of membrane-bound BR molecules without any low-molecular linker molecules is accomplished. Bacterial TGase in a two-step process catalyses the intermolecular cross-linking of Gln-3 with Lys-129. The results obtained with BR may be applied also to other membranebound molecules. 2. Materials and Methods Chemicals and Biomaterials. All chemicals were obtained in the finest grade available from Merck (Darmstadt, Germany) or Sigma/Aldrich (Munich, Germany) and were used without further purification. Purple membranes (PM) containing wild-type bacteriorhodopsin (BR-WT) or the variant BR-D96N/D85N were prepared according to the standard procedure of Oesterhelt and Stoeckenius.11 The variant BR-D96N/D85N differs from the wild type by the exchange of the aspartic acids (D) in positions 85 and 96 for asparagine (N).12 This variant was chosen because of its blue color, which allows it to be distinguished visually from the purple-colored wild type (BR-WT). The exchanged amino acids are located in the inner part of the molecule close to the retinal-binding pocket. Any influence on the TGase-catalyzed reactions on the outer side of the BR

10.1021/bm0056207 CCC: $20.00 © 2001 American Chemical Society Published on Web 02/17/2001

234

Biomacromolecules, Vol. 2, No. 1, 2001

molecule can be excluded because the exchanged amino acids cannot be accessed by the TGase. PM suspension was filtered through 5 µm nylon filters before it was used. The concentration of BR was determined spectroscopically using the extinction coefficient of 570 ) 62 700 L mol-1 cm-1 at 570 nm. Transglutaminase from S. mobaraense (TGase) was isolated according to the procedure published by Gerber et al.,13 and its activity was determined as described by Grossowicz et al.14 The fluorescent TGase substrate monodansylcadaverin (5N-(5′-N′,N′-dimethylamino-1′-naphthalenesulfonyl)diamidopentane)15 is available from Sigma/Aldrich. The fluorescent TGase substrate CBZ-Gln-Gly-C-DNS (1-N-(carbobenzoxyL-glutaminylglycyl)-5-N-(5′-N′,N′-dimethylamino-1′-naphthalenesulfonyl)diamidopentane) was obtained according to the procedure described by Pasternack et al.16 Monodansylcadaverin is a lysine-analogue molecule which is linked under the catalysis of TGase to glutamine residues. For this reason the name Glnprobe is used for it throughout this paper. By analogy, CBZ-Gln-Gly-C-DNS is referred to as Lysprobe. Gelelectrophoresis. SDS-PAGE was done in 12.5% gels. Since neither BR nor TGase contains disulfur bridges, mercaptoethanol was omitted in the gel buffer. The gels were silver stained,17 and fluorescence images were taken with a digital CCD camera (Coolpix 950, Nikon) prior to silver staining. The gels were irradiated with light from a 254-nm filtered UV lamp. To suppress excitation light, a filter combination (BG39, GG475, Schott) with a maximal transmission at 510 nm, the emission wavelengths of the fluorescent probes, was placed in front of the camera. Ultracentrifugation. Linear sucrose gradients for ultracentrifugation were prepared with a homemade gradient mixer. Ultracentrifugation was done in a swing-out rotor (TS41.14 in a Centrikon T1080, Kontron) at 4 °C and 34 000 U/min (∼140000g) for 18 h. TGase-Catalyzed Reactions. For the determination of reactive glutamines BR-WT (5.7 × 10-5 mol/L) was incubated with Glnprobe (5.6 × 10 -3 mol/L) and TGase (1.7 U/mL) at 37 °C in 40 mM Tris/HCl buffer, pH 7.0. The reactions were stopped after the desired reaction time by heating of the mixture to 80 °C for 2 min. The samples were then centrifuged, and the pellets were washed three times with water in order to remove unbound Glnprobe and the inactivated TGase. Then the pelleted material was resuspended in 50 mM Tris/HCl pH 7.0 and analyzed by SDSPAGE. In a similar way the presence of reactive lysines in BR was examined by means of the marker Lysprobe. For this purpose 5.5 mg of Lysprobe was dissolved in 210 µL of DMSO and diluted with 1890 µL of 10 mM hydrochloric acid. BRWT (4.8 × 10 -4 mol/L), Lysprobe (2.1 × 10-3 mol/L) in the DMSO solution described above, and TGase (1.6 U/mL) were incubated at 37 °C in 30 mM Tris/HCl buffer, pH 8.0, and samples were taken and analyzed as described above. Delipidation of BR. A 100-nmol portion of BR suspended in 300 µL of water was mixed with 1 mL of ammonium hydroxide/acetone solution (1:5 v/v) and was incubated at 37 °C for 5 min. After addition of 6 M trichloroacetic acid (ca. 80 µL), the precipitated protein was collected by

Seitz et al.

centrifugation and washed three times with acetone in order to remove residual lipids. Cleavage of BR. Chemical and enzymatic cleavage of BR after TGase-catalyzed labeling with Glnprobe or Lysprobe followed by separation of the polypeptide fragments via highperformance liquid chromatography (HPLC) was done in order to identify the amino acids in BR which serve as substrates for the TGase-catalyzed reaction. Chemical. Cleavage of BR with cyanogen bromide was done according to the procedure described by Lemke and Oesterhelt.18 A 100-nmol portion of delipidated BR was dissolved in 900 µL of concentrated formic acid. After addition of 385 µL of H2O and 130 µL of cyanogen bromide solution (1.4 g/mL in 70% formic acid), the solution was flushed with argon for 2 min. Then the mixture was incubated at 22 °C for 24 h. After addition of 1 mL of H2O, the sample was freeze-dried and redissolved in 200 µL of concentrated formic acid. Enzymatic. The cleavage of BR with chymotrypsin was done as described by Gerber et al.19 BR was suspended in 400 µL of 50 mM phosphate buffer, pH 8, and 0.45 mg of chymotrypsin dissolved in 50 µL of H2O and 50 µL of 50 mM calcium sulfate were added. After incubation at 37 °C for 24 h, the precipitate was washed three times with water and resuspended in 500 µL of 50 mM Tris/HCl buffer, pH 7.0. Liquid Chromatography. The polypeptide fragments obtained after either chemical cleavage or enzymatical digestion were separated by reversed-phase HPLC on a Hewlett-Packard model 1050 instrument equipped with a diode array detector. The column used was a RP18 (NC04, C18, 5 µm, 250 × 4 mm, Bischoff). Samples were prepared according to ref 18, and 100 µL was loaded on the column per run. Elution was done with a gradient starting with 100% formic acid and an increase of the acetonitrile component from 0% to 90% within 50 min. The absorption traces at 280 nm were chosen for the plots shown here. 3. Results Bacteriorhodopsin (BR) is obtained as two-dimensional hexagonal crystalline lattices from H. salinarum, the so-called purple membranes (PM). Most of the 248 amino acids of BR are arranged in seven R-helical domains which span the membrane plane. Those amino acids are not accessible for enzymatic reactions, and only a small number of amino acids are extramembraneous, in particular the C and N termini and the loops which connect the R-helical domains. We analyzed whether there are reactive glutamines and lysines in BR which are substrates for TGase. BR in PM form was used for all investigations reported here. Furthermore, it was shown that the TGase-catalyzed reaction enables linker-free cross-linking between BRs belonging to different PM patches. Reactive Glutamines and Lysines in Membrane-Bound BR. We used low-molecular fluorescent probes to determine whether BR contains amino acid residues which are accessible for and reactive with TGase. The aminopentane group in Glnprobe is structurally related to lysine. It has been used

Selective Cross-Linking by Transglutaminase

Biomacromolecules, Vol. 2, No. 1, 2001 235

Figure 1. Detection of reactive glutamines in BR. The TGasecatalyzed reaction of Glnprobe (monodansylcadaverin) with BR-WT is analyzed by SDS-PAGE (left, silver stain; right, fluorescence). The incubation time is given for each lane.

Figure 2. Detection of reactive lysines in BR. The TGase-catalyzed reaction of Lysprobe (1-N-(carbobenzoxy-L-glutaminylglycyl)-5-N-(5′N′,N′-dimethylamino-1′-naphthalenesulfonyl)diamidopentane) with BRWT is analyzed by SDS-PAGE (left, silver stain; right, fluorescence). The incubation time is given for each lane.

successfully as a TGase substrate to determine reactive glutamines in proteins.15 The modified fluorescent dipetide Lysprobe is known as TGase substrate and is coupled to reactive lysines.16 For the determination of reactive glutamines the TGase labeling of BR-WT with Glnprobe was investigated. Figure 1 shows the SDS-PAGE of the TGase-catalyzed reaction for five different incubation times (0, 2, 4, 8, and 24 h). On the left side, the silver-stained gel is shown. At all incubation times (0-24 h) one band in each lane with an apparent molecular mass of 23 kD is observed, which corresponds to BR.20 At 0 and 2 h no fluorescence is observed, but after 4 h a weak fluorescence band is discovered which further increases with extended reaction time (8 and 24 h). From this experiment it is concluded that at least one reactive glutamine exists in BR. Reactive lysines were determined by TGase-mediated labeling of BR-WT with Lysprobe. Samples were taken after 0, 2, 4, 8, and 30 h and analyzed on a SDS-PAGE (Figure 2). The BR bands on the silver-stained gel are overloaded (compare Figure 1). Due to the relatively low labeling yield of Lysprobe, more BR has to be loaded onto the SDS-PAGE in order to obtain a reliable fluorescence signal. With increasing incubation time, an increase of the fluorescence intensity is observed. This shows that besides reactive glutamines also reactive lysines are found in BR and BR may act as the glutamine as well as the lysine donor in TGase-catalyzed reactions. This result indicates the possibility of a linker-free covalent cross-linking of BR by means of TGase. Determination of the Reactive Sites. From the amino acid sequence of BR (e.g., in ref 21), two potential glutamines on the extracellular side may be identified (Gln-3 in helix A and Gln-75 in helix C). There are six lysines which need to be considered. The first and the only one on the

extracellular side is Lys-129 in helix C. In addition there are five lysines on the cytoplasmatic side: Lys-30 in helix A, Lys-40 and Lys-41 in helix B, Lys-159 in helix E, and Lys-172 in helix F. The seventh lysine in BR is Lys-216, which is the binding site for the chromophoric retinylidene group. Because the color of BR does not change during the TGase-catalyzed reaction, obviously no attachment of the fluorescent label to Lys-216 occurs, which would lead to the complete loss of the purple color of BR.22 For the determination of the binding sites, BR was marked with Glnprobe and Lysprobe, respectively, cleaved either with cyanogen bromide or with chymotrypsin, and the fragments which carry the fluorescent labels were determined. Chymotrypsin cleavage of BR yields three main fragments, one with 25 kD (fragment A), a second one with 17 kD (fragment B), and a third fragment with a molecular mass of 8 kD (fragment C). The two bigger main fragments appear often separated into two bands A1 and A2 as well as B1 and B2, which arise from carboxy-terminal digestion of BR by chymotrypsin at two different sites.23,24 Glutamine Gln-3 is located on fragment C; the other glutamine, Gln-75, is on fragment B. The lysines Lys-30, Lys-40, and Lys-41 are on fragment C, and fragment B comprises Lys-129, Lys-158, Lys-171, and Lys-216. The BR samples labeled with Glnprobe and with Lysprobe, respectively, were digested with chymotrypsin and analyzed by SDS-PAGE (Figure 3). The fluorescence of the gel (right side) was photographed, and then it was silver stained (left side). In the left lane nondigested BR labeled with Glnprobe is analyzed. Chymotrypsin-cleaved BR labeled with Lysprobe and with Glnprobe are on the middle and right lanes. The silver-stained SDSPAGE allows assignment of the fluorescence to one of the three fragments. The fluorescence of the Glnprobe-labeled BR (right lane) is found in fragment C, which leads to the conclusion that the reactive glutamine must be Gln-3. The

236

Biomacromolecules, Vol. 2, No. 1, 2001

Figure 3. Localization of the TGase-reactive amino acids by chymotryptic digestion of fluorescence-labeled BR. BR labeled with the fluorescent probes Lysprobe and Glnprobe were cleaved by chymotrypsin, and the reaction products were analyzed on SDS-PAGE (left, silver stain; right, fluorescence). A molecular weight marker (M) is on the outer left lane of the silver-stained gel. (middle) Reactive lysines can be found on fragment B only, which contains Lys-129, Lys-158, Lys-171, and Lys-216. (right) On fragment C a reactive glutamine (Gln-3) is found.

other glutamine is on fragment B, where no fluorescence is detected in this lane. An additional weak fluorescence is observed in 25 kD fragment A in the lane where Lysprobe-labeled and chymotrypsin-digested BR is analyzed. This fragment results from carboxyterminal digestion by chymotrypsin at two different sides (see refs 24 and 25). In the case where BR is labeled with Glnprobe, this band is not observed. Linkage of the fluorophores obviously affects the digestion activity of chymotrypsin. The main fluorescence of Lysprobe-labeled BR (middle lane) is found in fragment B, and due to this fact all lysines (Lys129, Lys-158, and Lys-171) must be taken into account as possible linkage sites. Upon cleavage of BR with cyanogen bromide, 10 fragments are obtained,18 which comprise the following amino acid sequences: 1-20; 21-32; 33-56; 57-60; 61-68; 69118; 119-145; 146-163; 164-209, and 210-248. The BR fragments were analyzed by means of reversed-phase highperformance liquid chromatography (RP-HPLC).18 BR (100nmol portions) was labeled with either Glnprobe or Lysprobe, then the BR was delipidated and cleaved with cyanogen bromide and the fragments were separated by RP-HPLC. For the determination of the reactive lysines, the effluent of the HPLC run with Lysprobe-labeled BR was collected in four fractions. The collected fractions are marked in the chromatogram (Figure 4). The composition of the fractions

Seitz et al.

Figure 4. Identification of the reactive lysine in BR. The RP-HPLC chromatogram of Lysprobe-labeled and cyanogen bromide cleaved BR is shown, and the four collected fractions are marked. The fluorescence spectra of the collected fractions are given below. Only in fraction 2 is fluorescence observed, and therefore, Lys-129 is determined to be the reactive lysine residue.

is as follows: fraction 1 ) amino acids 21-32 and 33-56, fraction 2 ) amino acids 1-20 and 119-145, fraction 3 ) amino acids 146-163 and 68-118, and the last fraction 4 with the amino acids 164-209. The fragment 210-248 contains no aromatic amino acids and for this reason is not detected at 280 nm. The collected fractions were freeze-dried and resuspended in 50 mM phosphate buffer, pH 9.0. The fluorescence of the samples was stimulated with an excitation wavelength of 330 nm. The numbers of the fluorescence spectra shown in the lower part of Figure 4 correspond to the number of the collected fractions. Only in fraction 2 is fluorescence resulting from the coupled Lysprobe found. Since no other lysine residue except Lys-129 is located on the fragments in this pool, Lys-129 is identified as the only reactive lysine in BR. The corresponding chromatogram of BR labeled with Glnprobe is shown in Figure 5. The fractions were again freezedried, resuspended in 50 mM phosphate buffer pH 9.0, and subjected to fluorescence analysis. Gln-3 appears in fraction 1, and Gln-75 is in fraction 2. The fluorescence spectra of the two fractions collected clearly assign Gln-3 as the reactive Gln residue in correspondence with the assignment derived by SDS-PAGE earlier. TGase-Catalyzed Intermembrane Cross-Linking. The reactive Gln site as well as the reactive Lys site are both located on the extracellular side of the membrane. The question arises whether it is possible to obtain a reaction between two BRs in the same PM (intramembrane cross-

Selective Cross-Linking by Transglutaminase

Figure 5. Identification of the reactive glutamine in BR. The RPHPLC chromatogram of BR labeled with Glnprobe and cleaved with cyanogen bromide is shown in the upper part of the diagram. The indicated fractions were collected and freeze-dried. The fluorescence spectra are shown below. In fraction 1, which contains the fragment with the amino acids 1-20, a strong fluorescence is observed. This confirms Gln-3 as the reactive glutamine.

-linking) or between two BRs in different PMs (intermembrane cross-linking). In the case that a dimerization of BR molecules is obtained in the TGase-catalyzed cross-linking of BR, a band with twice the molecular weight of BR is expected. PMs are cross-linked even if only a single intermembrane linkage between two BR molecules exists. Because each PM contains several thousand BR molecules, the ratio between dimeric and monomeric BR is too low for reliable detection on a SDS-PAGE. In addition it is known from ref 26 that in SDS-PAGE of BR often bands with an apparent molecular mass corresponding to BR dimers and trimers are observed. Because SDS-PAGE is not suitable to check for the crosslinking of BR, ultracentrifugation was employed. It was found that BR-WT and the variant BR-D85N/D96N do have a slightly different bouyant density. Separation of these two BR materials by ultracentrifugation using a linear sucrose gradient from 37.5% to 42.5% (w/w) shows two clearly separated bands. In addition the BR-D85N/D96N species has an intense blue color instead of the purple color of BRWT. In the case that the TGase-catalyzed cross-linking reaction occurs between two BR molecules of the same patch, the bouyant density of the two BR materials should change in a similar way. Only in the case where a cross-linking reaction occurs between BR molecules contained in two different patches should the resulting material have a common bouyant density. In particular, the combination of BR-WT and BR-D85N/D96N should lead to a band in the

Biomacromolecules, Vol. 2, No. 1, 2001 237

Figure 6. Intermembrane cross-linking of BR with TGase. Ultracentrifugation of reaction mixtures of purple BR-WT and blue BR-D96N/ D85N with TGase on a linear sucrose gradient. Samples of 100 µg of BR were analyzed in each tube. With an increasing amount of TGase, the two separated protein bands fuse into one single band which shows the mixed color of both BR types indicating the intermembrane cross-linking of BR.

density gradient which shows the mixed absorption of the two BR bands. The two BR variants were incubated with TGase at 37 °C for 24 h. Then the reactions were stopped by heating of the mixture to 80 °C for 2 min and subjected to ultracentrifugation on a linear sucrose gradient (Figure 6). In each experiment the total concentration of BR was 1 mg/mL (0.5 mg/mL BR-WT + 0.5 mg/mL BR-D96N/D85N). The concentration of TGase was approximately doubled from tube to tube starting with 0.23 U/ml in tube B and ending with 3.7 U/ml in tube F. In tube A a control sample which was incubated without TGase is shown. In the case of the two lowest TGase concentrations, no significant differences compared with the control sample A are observed. The weak offset of the BR bands cannot be counted as the result of a reaction because minor variations in the particular sucrose gradients exist. The reactions with the three higher TGase concentrations (tubes D to F) show remarkable differences. In tube D two additional bands with a buoyant density between BR-WT and BR-D96N/D85N are found. In tube E the band of BR-WT has completely disappeared. The color of the main band and the two weaker bands above the main band indicate that a reaction between the two different BR types has occurred. In tube F only a single, intense band is observed. Its color is a mixture of the blue and purple BR variants. This demonstrates that a cross-linking between PM patches has occurred. 4. Discussion In this paper the reaction of BR with TGase from StreptoVerticillium mobaraense is introduced. It revealed that

238

Biomacromolecules, Vol. 2, No. 1, 2001

Seitz et al.

Because both reactive amino acids are located on the extracellular side, the extracellular side of one PM is linked to the extracellular side of another PM. Due to this fact it is not possible to obtain three-dimensional networks of BRWT with the TGase reaction. Genetically engineered BRs with a single accessible glutamine and lysine binding site on the cytoplasmatic and the extracellular side are expected to open the way to three-dimensional networks of BR. Acknowledgment. We thank Volker Schwass from the MPI for Biochemistry, Martinsried, for preparing the BR materials. Discussion of the experimental results with Dieter Oesterhelt is gratefully acknowledged. This work was supported by the German Ministry for Education and Science (BMBF) under Grant FKZ 01M2988A. References and Notes

Figure 7. Model for the TGase-catalyzed cross-linking of PM patches. (A) The reaction of TGase with BRGln-3 leads to a reactive complex. (B) TGase is released during the cross-linking reaction with a reactive Lys-129. (C) Finally, the flexible extramembraneous parts of the amino acid sequence of BR form the intermembrane linker.

BR in its membrane-bound form may act as a glutamine as well as a lysine donor for TGase. By means of the fluorescent probes monodansylcadaverin (Glnprobe) and CBZ-Gln-GlyC-DNS (Lysprobe), the amino acid residues Gln-3 and Lys129 were identified as the only two reactive amino acids in BR which are substrates for TGase. Of particular interest is the finding that a cross-linking of BR molecules between different PM patches is possible by means of TGase. The surface charges of BR at pH 6.6 were determined to be -2.5 ( 0.2 on the cytoplasmic side and -1.8 ( 0.2 elementary charges/BR on the extracellular side.27 Direct coupling between two PM patches is hindered by these surface charges because the distance between the two PM surfaces needs to become small enough for a reaction to take place. The macromolecule TGase obviously shields the repulsive surface charges. As shown in the model (Figure 7) following the first step, i.e., the formation of the TGaseGln(BR) complex, the amino acid chain of BR might be partially pulled out of the membrane. During the formation of the linkage to Lys-129, TGase is released and the initial conformation may be retained. In particular the amino acid loop where Lys-129 is located may act like a flexible linker long enough to bridge the gap between two PM patches.

(1) Hampp, N. Chem. ReV. 2000, 100, 1755-1776. (2) Oesterhelt, D.; Bra¨uchle, C.; Hampp, N. Q. ReV. Biophys. 1991, 24, 425-478. (3) Hong, F. T. Prog. Surf. Sci. 1999, 62, 1-237. (4) Hohenfeld, I. P.; Wegener, A. A.; Engelhard, M. FEBS Lett. 1999, 442, 198-202. (5) Koyama, K.; Yamaguchi, N.; Miyasaka, T. Science 1994, 265, 762765. (6) Konishi, T.; Packer, L. Biochem. Biophys. Res. Commun. 1976, 72, 1437-1442. (7) Harada, Y.; Yasuda, K.; Nomura, S.; Kajimura, N.; Sasaki, Y. C. Langmuir 1998, 14, 1829-1835. (8) Folk, J. E.; Finnlayson, J. S. AdV. Protein Chem. 1977, 31, 1-133. (9) Ando, H.; Adachi, M.; Umeda, K.; Matsuura, A.; Nonaka, M.; Uchio, R.; Tanaka, H.; Motoki, M. Agric. Biol. Chem. 1989, 53, 26132617. (10) Gorman, J. J.; Folk, J. E. J. Biol. Chem. 1984, 256, 2712-2715. (11) Oesterhelt, D.; Stoeckenius, W. Methods Enzymol. 1974, 31, 667678. (12) Tittor, J.; Schweiger U.; Oesterhelt, D.; Bamberg, E. Biophys. J. 1994, 67, 1682-1690. (13) Gerber, U.; Jucknischke, U.; Putzien, S.; Fuchsbauer, H.-L. Biochem. J. 1994, 299, 825-829. (14) Grossowicz, N.; Wainfang, E.; Borek, E.; Waelsch, H. J. Biol. Chem. 1950, 187, 111-125. (15) Lorand, L.; Campbell, L. K. Anal. Biochem. 1971, 76, 207-220. (16) Pasternack, R.; Laurent, H.-P.; Ru¨th T.; Kaiser, A.; Scho¨n N.; Fuchsbauer, H.-L. Anal. Biochem. 1997, 249, 54-60. (17) Blum, H.; Beier, H.; Gross, H. J. Electrophoresis 1987, 8, 93-99. (18) Lemke, H.-D.; Oesterhelt, D. Eur. J. Biochem. 1981, 115, 595-604. (19) Gerber, G. E.; Anderegg, R. J.; Herlihy, W. C.; Gray, C. P.; Bieman, K.; Khorana, H. G. Proc. Natl. Acad. Sci. U.S.A. 1979, 76, 227231. (20) Bridgen, J.; Walker, I. D. Biochemistry 1976, 15, 792-798. (21) Birge, R. R. Annu. ReV. Phys. Chem. 1990, 41, 683-733. (22) Oesterhelt, D.; Meentzen, M.; Schuhmann, L. Eur. J. Biochem. 1973, 40, 453-463. (23) Gerber, G. E.; Gray, C. P.; Wildenauer, D.; Khorana, H. G. Proc. Natl. Acad. Sci. U.S.A. 1977, 74, 5426-5430. (24) Lemke, H. D.; Oesterhelt, D. FEBS Lett. 1981, 128/2, 255-260. (25) Ovchinnikov, Yu. A.; Abdulaev, N. G.; Feigina, M. Yu.; Kiselev, A. V.; Cabanov, N. A. FEBS Lett. 1979, 100, 219-224. (26) Fimmel, S.; Choli, T.; Dencher, N. A.; Bu¨ldt, G.; Wittmann-Liebold, D. Biochim. Biophys. Acta 1988, 978, 231-240. (27) Alexiev, U.; Marti, T.; Heyn, M. P.; Khorana, H. G.; Scherrer, P. Biochemistry 1994, 33, 298-306.

BM0056207