Topography of surface-exposed amino acids in the membrane protein bacteriorhodopsin determined by proteolysis and micro-sequencing

Determination of membrane protein orientation upon liposomal reconstitution down to the single vesicle level

Reconstitution of membrane proteins into liposomal membranes represents a key technique in enabling functional analysis under well-defined conditions. In this review, we provide a brief introduction to selected methods that have been developed to determine membrane protein orientation after reconstitution in liposomes, including approaches based on proteolytic digestion with proteases, site-specific labeling, fluorescence quenching and activity assays. In addition, we briefly highlight new strategies based on single vesicle analysis to address the problem of sample heterogeneity.

Functional cell-free synthesis of a seven helix membrane protein: in situ insertion of bacteriorhodopsin into liposomes

The expression of membrane proteins for functional and structural studies or medicinal applications is still not very well established. Membrane-spanning proteins that mediate the information flow of the extracellular side with the interior of the cell are prime targets for drug development methods that would allow screening techniques or high throughput formats are of particular interest. Here we describe a systematic approach to the liposome-assisted cell-free synthesis of functional membrane proteins. We demonstrate the synthesis of bacteriorhodopsin (bR(cf)) in presence of small unilamellar liposomes. The yield of bR(cf) per volume cell culture is comparable to that of bacteriorhodopsin in its native host. The functional analysis of bR(cf) was performed directly using the cell-free reaction mixture. Photocycle measurements reveal kinetic data similar to that determined for bR in Halobacterium salinarum cell-envelope vesicles. The liposomes can be attached directly to black lipid membranes (BLM), which allows measuring light activated photocurrents in situ. The results reveal a functional proton pump with properties identical to those established for the native protein.

Glycine receptors: lessons on topology and structural effects of the lipid bilayer

The members of the superfamily of nicotinicoid receptors, sometimes referred to as the ligand-gated ion channel superfamily (LGICS), are essential mediators in the propagation of electrical signals between cells at neuronal and neuromuscular synapses. Given the significant sequence and proposed topological similarities between family members, the structural architecture of any one of these neuroreceptors is believed to be archetypic for the family of ligand-gated channels. We have focused our biophysical studies on the glycine receptor (GlyR) since homomeric expression of just the alpha1 chain of the receptor is sufficient to reconstitute native-like activity when expressed in heterologous cells, and we have successfully overexpressed and purified relatively large quantities of this receptor. Our CD data suggests that the historical four transmembrane helix topology model for nicotinicoid receptors may be erroneous. Proteolytic studies as well as chemical modification studies coupled with mass spectroscopy (MS) have provided additional evidence that this model may be inappropriate. While we suggest a novel topological model for the superfamily of nicotinicoid receptors, the absence of high resolution data for the transmembrane regions of these ion channels precludes further refinement of this model. In addition, we observe structural changes in the recombinant alpha1 GlyR as a function of bilayer composition, suggesting that these receptors may be dynamically modulated by cellular control over the properties of the plasma membrane.

Energetics, stability, and prediction of transmembrane helices

We show that the peptide backbone of an alpha-helix places a severe thermodynamic constraint on transmembrane (TM) stability. Neglect of this constraint by commonly used hydrophobicity scales underlies the notorious uncertainty of TM helix prediction by sliding-window hydropathy plots of membrane protein (MP) amino acid sequences. We find that an experiment-based whole-residue hydropathy scale (WW scale), which includes the backbone constraint, identifies TM helices of membrane proteins with an accuracy greater than 99 %. Furthermore, it correctly predicts the minimum hydrophobicity required for stable single-helix TM insertion observed in Escherichia coli. In order to improve membrane protein topology prediction further, we introduce the augmented WW (aWW) scale, which accounts for the energetics of salt-bridge formation. An important issue for genomic analysis is the ability of the hydropathy plot method to distinguish membrane from soluble proteins. We find that the method falsely predicts 17 to 43 % of a set of soluble proteins to be MPs, depending upon the hydropathy scale used.

Coupled proteolytic and mass spectrometry studies indicate a novel topology for the glycine receptor

Members of the heteropentameric ligand-gated ion channel superfamily rapidly mediate signaling across the synaptic cleft. Sequence analysis and limited experimental studies have yielded a topological model containing four transmembrane alpha-helices, labeled M1 to M4, and a large soluble, extracellular N-terminal domain. This model persists to date despite some recent structural studies that suggest it may be inappropriate. In this study, the topology of the glycine receptor was probed by limited proteolysis coupled to mass spectrometry. Of particular note, accessible cleavage sites within the putative M1 and M3 transmembrane helices were identified. Membrane-associated fragments within the postulated globular extracellular N-terminal domain were also observed. This report presents several key details incorporated in a new topological model and is the first direct experimental evidence that a subset of the transmembrane regions are too short to be membrane-spanning alpha-helices; rather, these regions are proposed to be a mix of alpha-helices and beta-sheets. This report is also the first to exploit the capability of mass spectrometry to probe critically the topology of a class of membrane proteins of unknown structure.

Fourier transform infrared spectroscopy study of the secondary structure of the gastric H+,K+-ATPase and of its membrane-associated proteolytic peptides

Membrane topology of the H+,K+-ATPase has been studied after proteolytic degradation of the protein by proteinase K. Proteinase K had access to either the cytoplasmic part of the protein or to both sides of the membrane. Fourier transform infrared attenuated total reflection spectroscopy indicated that membrane-associated domain of the protein represented about 55% of the native protein, meanwhile the cytoplasmic part represented only 27% of the protein. The secondary structure of the ATPase and of its membrane-associated domains was investigated by infrared spectroscopy. The secondary structure of the membrane-associated structures and of the entire protein was quite similar (alpha-helices, 35%; beta-sheets, 35%; turns, 20%; random, 15%). These data were in agreement with 10 alpha-helical transmembrane segments but suggested a participation of beta-sheet structures in the membrane-associated part of the protein. Polarized infrared spectroscopy indicated that the alpha-helices were oriented nearly perpendicular to the membrane plane. No preferential orientation could be attributed to the beta-sheets. Monitoring the amide hydrogen/deuterium exchange kinetics demonstrated that the membrane associated part of the ATPase molecule is characterized by a relatively high accessibility to the solvent, quite different from that observed for bacteriorhodopsin membrane segments.

Immuno-atomic force microscopy of purple membrane

The atomic force microscope is a useful tool for imaging native biological structures at high resolution. In analogy to conventional immunolabeling techniques, we have used antibodies directed against the C-terminus of bacteriorhodopsin to distinguish the cytoplasmic and extracellular surface of purple membrane while imaging in buffer solution. At forces > or = 0.8 nN the antibodies were removed by the scanning stylus and the molecular topography of the cytoplasmic purple membrane surface was revealed. When the stylus was retracted, the scanned membrane area was relabeled with antibodies within 10 min. The extracellular surface of purple membrane was imaged at 0.7 nm resolution, exhibiting a major and a minor protrusion per bacteriorhodopsin monomer. As confirmed by immuno-dot blot analysis and sodium dodecyl sulfate-gel electrophoresis, labeling of the purple membrane was not observed if the C-terminus of bacteriorhodopsin was cleaved off by papain.

Hydrophobic organization of alpha-helix membrane bundle in bacteriorhodopsin

The hydrophobic organization of the intramembrane alpha-helical bundle in bacteriorhodopsin (BRh) was assessed based on a new approach to characterization of spatial hydrophobic properties of transmembrane (TM) alpha-helical peptides. The method employs two independent techniques: Monte Carlo simulations of nonpolar solvent around TM peptides and analysis of molecular hydrophobicity potential on their surfaces. The results obtained by the two methods agree with each other and permit precise hydrophobicity mapping of TM peptides. Superimposition of such data on the experimentally derived spatial model of the membrane moiety together with 2D maps of hydrophobic hydrophilic contacts provide considerable insight into the hydrophobic organization of BRh. The helix bundle is stabilized to a large extent by hydrophobic interactions between helices--neighbors in the sequence of BRh, by long-range interactions in helix pairs C-E, C-F, and C-G, and by nonpolar contracts between retinal and helices C, D, E, F. Unlike globular proteins, no polar contacts between residues distantly separated in the sequence of BRh were found in the bundle. One of the most striking results of this study is the finding that the hydrophobic organization of BRh is significantly different from those in bacterial photoreaction centers. Thus, TM alpha-helices in BRh expose their most nonpolar sides to the bilayer as well as to the neighboring helices and to the interior of the bundle. Some of them contact lipids with their relatively hydrophilic surfaces. No correlation was found between disposition of the most hydrophobic and the most variable sides of the TM helices.

Fourier transform infrared spectroscopy study of the secondary structure of the reconstituted Neurospora crassa plasma membrane H(+)-ATPase and of its membrane-associated proteolytic peptides

We reconstituted purified plasma membrane H(+)-ATPase from Neurospora crassa into soybean phospholipid vesicles (lipid/ATPase ratio of 5:1 w/w). The proteoliposomes contained an active ATPase, oriented inside-out. They were subjected to proteolysis by using Pronase, proteinase K, trypsin, and carboxypeptidase Y. Fourier transform infrared attenuated total reflection spectroscopy indicates that the amount of protein remaining after hydrolysis and elimination of the extramembrane domain of ATPase represents about 43% of the intact protein. The secondary structure of intact ATPase and of the membrane-associated domain of ATPase was determined by infrared spectroscopy. The membrane domain shows a typical alpha-helix and beta-sheet absorption. Polarized infrared spectroscopy reveals that the orientation of the helices is about perpendicular to the membrane. Amide hydrogen/deuterium exchange kinetics performed for the intact H(+)-ATPase and for the membrane-associated domain demonstrate that this part of ATPase shows less accessibility to the solvent than the entire protein but remains much more accessible to the solvent than bacteriorhodopsin membrane segments.

Bacteriorhodopsin expressed in Schizosaccharomyces pombe pumps protons through the plasma membrane

Bacterioopsin (bO) from Halobacterium salinarium ("Halobacterium halobium") has been functionally expressed in a heterologous system, the fission yeast Schizosaccharomyces pombe. Regeneration of bO to bacteriorhodopsin (bR) in S. pombe has been achieved in vivo by addition of the chromophore retinal to the culture medium, as shown for a retinal-negative mutant of H. salinarium (JW5). Western blot analysis revealed that bR is more stable than bO against proteolysis in fission yeast and also in JW5. The light-driven proton pump is expressed in the eukaryotic organism and incorporated into the plasma membrane. Illumination of intact yeast cells leads to acidification of the external medium due to the translocation of H+ from inside to outside of the cell, indicating the same orientation of bR in the yeast plasma membrane as in H. salinarium. The kinetics of proton release into the water phase was observed with the optical pH indicator pyranine. Time-resolved absorbance changes of isolated plasma membrane measured by flash spectroscopy showed rise and decay of the M intermediate during the photocycle similar to those in the homologous system. Photocurrents and photovoltages were recorded with yeast plasma membrane attached to a planar lipid membrane and to a polytetrafluoroethylene (Teflon) film, respectively. Stationary currents measured in the presence of a protonophore showed continuous pumping activity of bR. The action spectrum of the photocurrent and the kinetics of the photovoltage were analyzed and compared with signals obtained from purple membranes. From all these different investigations we conclude that the integral membrane protein bR is correctly folded in vivo into the cytoplasmic membrane of the fission yeast S. pombe.

Conformational preference functions for predicting helices in membrane proteins

A suite of FORTRAN programs, PREF, is described for calculating preference functions from the data base of known protein structures and for comparing smoothed profiles of sequence-dependent preferences in proteins of unknown structure. Amino acid preferences for a secondary structure are considered as functions of a sequence environment. Sequence environment of amino acid residue in a protein is defined as an average over some physical, chemical, or statistical property of its primary structure neighbors. The frequency distribution of sequence environments in the data base of soluble protein structures is approximately normal for each amino acid type of known secondary conformation. An analytical expression for the dependence of preferences on sequence environment is obtained after each frequency distribution is replaced by corresponding Gaussian function. The preference for the alpha-helical conformation increases for each amino acid type with the increase of sequence environment of buried solvent-accessible surface areas. We show that a set of preference functions based on buried surface area is useful for predicting folding motifs in alpha-class proteins and in integral membrane proteins. The prediction accuracy for helical residues is 79% for 5 integral membrane proteins and 74% for 11 alpha-class soluble proteins. Most residues found in transmembrane segments of membrane proteins with known alpha-helical structure are predicted to be indeed in the helical conformation because of very high middle helix preferences. Both extramembrane and transmembrane helices in the photosynthetic reaction center M and L subunits are correctly predicted. We point out in the discussion that our method of conformational preference functions can identify what physical properties of the amino acids are important in the formation of particular secondary structure elements.

In vitro synthesis of bacterio-opsin: integration into microsomal membranes

The translation and membrane integration of bacterio-opsin from Halobacterium salinarium were investigated. Plasmids containing the bacterio-opsin-gene with or without its original presequence were transcribed with the T7-RNA-polymerase and translated in vitro in a wheat germ system. The integration of the expressed bacterio-opsin into dog pancreas microsomes was studied. Both precursor bacterio-opsin and mature bacterio-opsin integrate into the eukaryotic membrane.

Surface-bound optical probes monitor protein translocation and surface potential changes during the bacteriorhodopsin photocycle

Light-induced H+ release and reuptake as well as surface potential changes inherent in the bacterio-rhodopsin reaction cycle were measured between 10 degrees C and 50 degrees C. Signals of optical pH indicators covalently bound to Lys-129 at the extracellular surface of bacteriorhodopsin were compared with absorbance changes of probes residing in the aqueous bulk phase. Only surface-bound indicators monitor the kinetics of H+ ejection from bacteriorhodopsin and allow the correlation of the photocycle with the pumping cycle. During the L550----M412 transition the H+ appears at the extracellular surface of bacteriorhodopsin. Surface potential changes detected by bound fluorescein or by the potentiometric probe 4-[2-(di-n-butylamino)-6-naphthyl]vinyl-1-(3-sulfopropyl)pyridinium betaine (di-4-ANEPPS) occur in milliseconds concomitantly with the formation and decay of the N intermediate. pH indicators residing in the aqueous bulk phase reflect the transfer of H+ from the membrane surface into the bulk but do not probe the early events of H+ pumping. The observed retardation of H+ at the membrane surface for several hundred microseconds is of relevance for energy conversion of biological membranes powered by electrochemical H+ gradients.

Topology of the anion-selective porin Omp32 from Comamonas acidovorans

Limited proteolysis experiments were performed with outer membranes from Comamonas acidovorans to probe the topology of its major protein component, the anion-selective porin Omp32. Proteinase K treatment above a critical temperature of 42 degrees C cleaved the surface-exposed regions of the porin, yielding membrane-embedded fragments which were separated by SDS polyacrylamide gel electrophoresis or reversed phase chromatography. The identification of the proteinase K-sensitive sites was performed by microsequencing. This allowed us to determine six surface-exposed sites of the porin, all located in nonconserved primary structure regions. These results along with the previously determined amino acid sequence and in conjunction with some structural constraints applicable to porins allowed us to propose a chain-folding model of the Omp32 porin. The features of our model are compared with the structure of the Rhodobacter capsulatus porin, recently established by X-ray crystallography (Weiss et al., 1991) and they are used to elucidate the structural basis of the anion selectivity.

Topography of the membrane-bound ADP/ATP carrier assessed by enzymatic proteolysis

The folding of the peptide chain of the beef heart ADP/ATP carrier in the inner mitochondrial membrane was investigated by enzymatic and immunochemical approaches, using specific proteases and polyclonal antibodies directed against the whole protein and specific regions of the carrier. The accessibility of the membrane-bound ADP/ATP carrier to proteases was followed by immunodetection of the cleavage products, using mitochondria devoid of outer membrane (mitoplasts) and inside-out submitochondrial particles (SMP) in the presence of either carboxyatractyloside (CATR) or bongkrekic acid (BA), two specific inhibitors which are able to bind to the outer face or the inner face of the carrier, respectively. Four types of particles were investigated, namely, mitoplasts-CATR, mitoplasts-BA, SMP-CATR, and SMP-BA. Only the ADP/ATP carrier in SMP-BA was cleaved by two specific proteases, namely, trypsin and lysine C endoprotease, at low doses for short periods of time. Two initial cleavage sites were found between Lys-42 and Glu-43, and between Lys-244 and Gly-245. After a longer period of incubation, an additional cleavage site between Lys-146 and Gly-147 could be demonstrated. Despite cleavage of the membrane-embedded carrier, the binding capacity and affinity of SMP for BA were not altered. A number of other proteases tested, including V8 protease, proline C endoprotease, thrombin, alpha-chymotrypsin, and thermolysin had virtually no effect. These results are explained by a dynamic model of the arrangement of the peptide chain of the ADP/ATP carrier.(ABSTRACT TRUNCATED AT 250 WORDS)

High-resolution 13C NMR study of the topography and dynamics of methionine residues in detergent-solubilized bacteriorhodopsin

The proton transport membrane protein bacteriorhodopsin has been biosynthetically labeled with [methyl-13C]methionine and studied by high-resolution 13C NMR after solubilization in the detergent Triton X-100. The nine methionine residues of bacteriorhodopsin give rise to four well-resolved 13C resonances, two of which are shifted upfield or downfield due to nearby aromatic residues. Methionine residues located on the hydrophilic surfaces, on the hydrophobic surface, and in the interior of the protein could be discriminated by studying the effects of papain proteolysis, glycerol-induced viscosity increase, and paramagnetic broadening by spin-labels on NMR spectra. Such data were used to evaluate current models of the bacteriorhodopsin transmembrane folding and tertiary structure. T2 and NOE measurements were performed to study the local dynamics of methionine residues in bacteriorhodopsin. For the detergent-solubilized protein, hydrophilic and hydrophobic external residues undergo a relatively large extent of side chain wobbling motion while most internal residues are less mobile. In the native purple membrane and in reconstituted bacteriorhodopsin liposomes, almost all methionine residues have their wobbling motion severely restricted, indicating a large effect of the membrane environment on the protein internal dynamics.

Transmembrane protein structure: spin labeling of bacteriorhodopsin mutants

Transmembrane proteins serve important biological functions, yet precise information on their secondary and tertiary structure is very limited. The boundaries and structures of membrane-embedded domains in integral membrane proteins can be determined by a method based on a combination of site-specific mutagenesis and nitroxide spin labeling. The application to one polypeptide segment in bacteriorhodopsin, a transmembrane chromoprotein that functions as a light-driven proton pump is described. Single cysteine residues were introduced at 18 consecutive positions (residues 125 to 142). Each mutant was reacted with a specific spin label and reconstituted into vesicles that were shown to be functional. The relative collision frequency of each spin label with freely diffusing oxygen and membrane-impermeant chromium oxalate was estimated with power saturation EPR (electron paramagnetic resonance) spectroscopy. The results indicate that residues 129 to 131 form a short water-exposed loop, while residues 132 to 142 are membrane-embedded. The oxygen accessibility for positions 131 to 138 varies with a periodicity of 3.6 residues, thereby providing a striking demonstration of an alpha helix. The orientation of this helical segment with respect to the remainder of the protein was determined.

Observations concerning topology and locations of helix ends of membrane proteins of known structure

Hydropathy plots of amino acid sequences reveal the approximate locations of the transbilayer helices of membrane proteins of known structure and are thus used to predict the helices of proteins of unknown structure. Because the three-dimensional structures of membrane proteins are difficult to obtain, it is important to be able to extract as much information as possible from hydropathy plots. We describe an "augmented" hydropathy plot analysis of the three membrane proteins of known structure, which should be useful for the systematic examination and comparison of membrane proteins of unknown structure. The sliding-window analysis utilizes the floating interfacial hydrophobicity scale [IFH(h)] of Jacobs and White (Jacobs, R.E., White, S. H., 1989. Biochemistry 28:3421-3437) and the reverse-turn (RT) frequencies of Levitt (Levitt, M., 1977, Biochemistry 17:4277-4285). The IFH(h) scale allows one to examine the consequences of different assumptions about the average hydrogen bond status (h = 0 to 1) of polar side chains. Hydrophobicity plots of the three proteins show that (i) the intracellular helix-connecting links and chain ends can be distinguished from the extracellular ones and (ii) the main peaks of hydrophobicity are bounded by minor ones which bracket the helix ends. RT frequency plots show that (iii) the centers of helices are usually very close to wide-window minima of average RT frequency and (iv) helices are always bounded by narrow-window maxima of average RT frequency. The analysis suggests that side-chain hydrogen bonding with membrane components during folding may play a key role in insertion.