transmembrane protein topology mapping by the substituted cysteine accessibility method (scamtm):...

Methods 36 (2005) 148–171

www.elsevier.com/locate/ymeth

Transmembrane protein topology mapping by the substituted cysteine accessibility method (SCAMTM): Application to lipid-speciWc

membrane protein topogenesis

Mikhail Bogdanov, Wei Zhang, Jun Xie, William Dowhan ¤

Department of Biochemistry and Molecular Biology, University of Texas-Houston, Medical School, Houston, TX 77030, USA

Accepted 15 November 2004Available online 13 May 2005

Abstract

We provide an overview of lipid-dependent polytopic membrane protein topogenesis, with particular emphasis on Escherichia colistrains genetically altered in their lipid composition and strategies for experimentally determining the transmembrane organizationof proteins. A variety of reagents and experimental strategies are described including the use of lipid mutants and thiol-speciWc chem-ical reagents to study lipid-dependent and host-speciWc membrane protein topogenesis by substituted cysteine site-directed chemicallabeling. Employing strains in which lipid composition can be controlled temporally during membrane protein synthesis and assem-bly provides a means to observe dynamic changes in protein topology as a function of membrane lipid composition. 2005 Elsevier Inc. All rights reserved.

Keywords: Membrane protein; Cysteine scanning; Maleimides; Phospholipid; Topogenesis; Phosphatidylethanolamine; Protein topology

1. Introduction

Polytopic membrane proteins account for 25–30% ofall open reading frames in sequenced genomes, and theyare responsible for a wide range of cellular functionssuch as solute transport, biosynthesis, energy produc-tion, intracellular signaling, and cell–cell communica-tion. Moreover, 50% of all drug targets are membraneproteins. High-resolution crystal structures of onlyabout 100 membrane proteins have been determined.Although, the methodology for obtaining high-resolu-tion structures is improving, the need to determine low-resolution organizational information on membraneproteins in a native membrane will continue. PuriWca-tion, crystallization, and structure determination stillremain formable tasks. Crystal structures are static and

* Corresponding author. Fax: +1 713 500 0652.E-mail address: [email protected] (W. Dowhan).

1046-2023/$ - see front matter 2005 Elsevier Inc. All rights reserved.doi:10.1016/j.ymeth.2004.11.002

may be distorted due to puriWcation and crystallizationconstraints and loss of information on interactions withother proteins and the lipid environment. Dynamicaspects of protein structure as a function of physiologi-cal state of the cell are best probed in whole cells ormembranes. Determination of membrane protein orga-nization has mainly relied on in silico approaches of pre-dicting extramembrane domains and transmembrane(TM) segments based on the hydrophobicity of the com-ponent amino acids. However, these hydropathy plotsare only 60–70% accurate in predicting topological orga-nization and therefore, only provide a starting point forthe design of experimental approaches to arrive at a Wnaltopological map of a protein in a membrane.

Due to the importance of determining membrane pro-tein topological organization, an increasing number ofreports over the last few years have utilized diversereagents and methodologies to address protein topology.There are no comprehensive reviews covering the latestapproaches. We will Wrst review factors that may

mailto: [email protected]

mailto: [email protected]

M. Bogdanov et al. / Methods 36 (2005) 148–171 149

determine membrane protein topological organization,and then focus on the methodologies used to establishtopological organization with emphasis on the use of thesubstituted cysteine accessibility method (SCAM) asapplied to determining TM segment orientation(SCAMTM). In SCAM, cysteine is introduced as a singleamino acid replacement in a protein engineered to removeall native cysteines. When applied to polytopic membraneproteins containing TM segments spanning the membranebilayer (SCAMTM), cysteine replacements are positionedin the putative extramembrane domains predicted byhydropathy plots and other information about the targetprotein. The orientation with respect to the plane of thebilayer of each TM segment in a native membrane orreconstituted vesicle system is determined by using a com-bination of reagents directed at cysteine under conditionsof membrane impermeable or membrane permeable to thereagent. Cysteines placed within TM segments are gener-ally inaccessible to thiol-speciWc reagents and combinedwith the labeling pattern of extramembrane domainsdeWnes the length and orientation of TM segments.

1.1. Membrane protein topogenesis

The biogenesis of polytopic membrane proteins withhigh helical content involves the proper positioning ofTM helices, coordinated folding of extramembranedomains, and helical packing within the lipid bilayer.The vast majority of prokaryotic and eukaryotic integraltransmembrane proteins are co-translationally insertedinto the membrane in a signal recognition particle-dependent manner [1]. Insertion occurs continuously orin a step-wise manner either by lateral translocation ofthe TM segments from the translocon complex into thelipid bilayer or by retrograde recruitment back to thetranslocon pore for orientation or proofreading [2]. Withfew exceptions [3], this results in a unique topologydetermined during membrane insertion by interactionbetween topogenic signals within the nascent proteinand extra-protein factors that are only partially under-stood.

A fundamental aspect of the structure of polytopicmembrane proteins is the membrane topology, i.e., thenumber and orientation of TM segments. For transportproteins, channels, and pores the organization of the TMsegments determines function. However, for most pro-teins with catalytic capacity or involvement in signalingand recognition processes, more relevant to function isthe disposition of the extramembrane domains withrespect to the plane of the membrane bilayer. It is gener-ally accepted that the topology of most polytopic mem-brane proteins is established co-translationally duringmembrane insertion and once established is maintainedduring subsequent steps of biogenesis, cellular traYck-ing, and function. However, several recent studies pro-vide an exception to this rule and demonstrate that

initial membrane topology in the bacterial cytoplasmicmembrane [4,5] or the eukaryotic cell endoplasmic retic-ulum membrane [6] is dynamic and may be modiWed bysubsequent lipid-dependent or translocon-dependentfolding events, respectively.

These Wndings have signiWcant implications for themolecular mechanisms by which polytopic proteinsacquire their topology in the membrane. The dynamicaspect of membrane protein topological organizationand the possibility that topology may be a function ofmembrane location in the cell or post-synthetic temporalfactors necessitates eVective methods for determiningmembrane protein topological organization in nativemembranes.

1.2. Lipid mutants as “biological reagents”

The ability to manipulate membrane lipid composi-tion in living cells has made possible the study of the roleof the membrane lipid environment in a broad spectrumof cellular processes including determination of mem-brane protein topology. Genetic manipulation of Esche-richia coli and other bacteria, yeast, and mammaliancells with resulting controlled changes in membrane lipidcomposition has been reviewed in detail elsewhere [7,8].

The major phospholipids of E. coli are phosphatidyl-ethanolamine (PE, 70–75%), phosphatidylglycerol (PG,20–25%), and cardiolipin (CL, 5–10%). Viable strains areavailable with the following changes in these major lipidclasses: 10-fold reduced level of CL [9]; complete lack ofPG and CL with PE comprising 90% of the totalphospholipid and the remainder being primarily phos-phatidic acid and CDP-diacylglycerol [10]; completelack of PE with the remainder being primarily PG andCL [11]. In addition, strains have been engineered inwhich the steady state level of PG plus CL [12] or PE [4]can be regulated in a dose dependent manner as a func-tion of extracellular regulation of the biosyntheticenzymes responsible for their synthesis. Foreign lipidshave been introduced into E. coli either in addition to orin the place of native lipids. The foreign lipids that havebeen introduced into E. coli are phosphatidylcholine[13], phosphatidylinositol [14], and monoglucosyl diacyl-glycerol [15]. Use of strains with altered lipid composi-tion has established a deWned role for phospholipids in:SecA-dependent, TAT-dependent, and FtsY-dependenttranslocation of proteins across the cytoplasmic mem-brane [16–24]; DnaA protein-dependent initiation ofDNA replication [25–27]; sugar transport by the phos-photransfer system [28,29]; cell viability [30–33]; proteintranslocation across the inner membrane [34]; eYcientelectron transport [35]; cell division [36,37]; formation ofdistinct lipid domains in the cell membrane [38]; and sig-nal transduction via the Cpx system [39]. These strainsalso have been used to: uncover the role of lipids as lipo-chaperones [40–42]; establish the role of lipids as factors

150 M. Bogdanov et al. / Methods 36 (2005) 148–171

controlling Wnal membrane protein topology [4,5,23,43]and uncover lipid-dependent topological switches withinmembrane proteins [4,5].

Yeast, being a eukaryote, has a more complex lipidcomposition and possesses organelles. All phospholipidbiosynthetic genes have been identiWed and cloned andyeast strains are available lacking phosphatidylserine,lacking phosphatidylcholine, or containing very lowlevels of PE [44]. In addition, the level of the mitochon-drial-speciWc phospholipids PG and CL has also beengenetically manipulated to eliminate CL [45] alone orboth PG and CL [46]. Strains lacking CL have func-tional but compromised mitochondrial-dependentenergy transducing systems [45] and appear to havereduced or less stable [45,47] interactions between com-plexes of the electron transport chain that are normallyorganized into supermolecular complexes. Strains lack-ing both PG and CL have severely dysfunctional mito-chondria and defects in synthesis of electron transportchain components [48].

At the moment bacteria and yeast are the most tracta-ble organisms for genetic manipulation of lipid metabo-lism to study the role of lipids in supporting cellfunction. The genetics of lipid metabolism in mamma-lian cells is well advanced [49] and the prospects fordeveloping viable somatic cell lines with altered lipidcontent are good.

1.3. Lipid-dependent topogenesis

Simultaneous development of reagent strains of E.coli with altered phospholipid composition and moresophisticated methods for determining protein topology,in particular SCAMTM, revealed that membrane lipidcomposition is a critical determinant of topology. Lac-tose permease (LacY) [4] and phenylalanine permease(PheP) [5] expressed in E. coli mutants lacking PE aredefective in active transport but still carry out facilitatedtransport. These secondary transporters actively accu-mulate substrate against a concentration gradient bycoupling uphill transport of substrate with downhillmovement of a H+. They contain 12 TM segments withthe N- and C-termini facing the cytoplasmic side of themembrane but belong to diVerent families of secondarytransporters. By using SCAMTM these proteins wereshown to display signiWcantly diVerent topological orga-nization in PE-containing cells than in PE-lacking cells.For LacY the six N-terminal TM segments and for PhePthe two N-terminal TM segments assume an invertedtopology with respect to the plane of the membranebilayer in PE-lacking membranes. Moreover, for LacYthe Wnal topological organization appears to be deter-mined solely by the phospholipid composition indepen-dent of cellular protein assembly based on reconstitutioninto proteoliposomes followed by SCAMTM [43]. Evenmore interesting was the observation that induction of

PE synthesis after membrane insertion and folding ofLacY and PheP resulted in a return of native topologicalorganization and transport function. These resultsclearly demonstrated that the lipid composition is adeterminant of TM segment orientation and challengedthe dogma that once TM orientation is established dur-ing assembly it is static and not subject to change.

According to this new topology paradigm, it is possi-ble that speciWc regions of a membrane protein canundergo reversible conformational or TM segment reor-ganizations in vivo, dictated not only by phospholipids[4,5,43], but also by components of the insertion machin-ery, substrates, or other eVectors [6,50–52]. The N-termi-nal signal sequence of a polytopic membrane protein canundergo reorientation after entering into the translocon[50]. If downstream topogenic sequences override the ini-tial topology of a TM segment, the TM segment may beable to re-enter the translocon to reorient itself [2]. Post-translational topological reorientation of viral proteinscan be facilitated by the host endoplasmic reticulumtranslocon [53] or molecular chaperones [54]. The E. colitranslocon component SecG shows an unusual propertyof inverting its orientation in the membrane, which istightly coupled to the SecG function and linked with theATP-driven insertion–deinsertion cycle of SecA [55].The dynamic topological changes of the Tat proteininsertion apparatus might be coupled to the transloca-tion of folded proteins across the cytoplasmic membrane[56].

The direct interaction of positively charged proteinresidues with negatively charged lipids can be dominantin retaining these protein domains on the cytoplasmicside of the membrane providing a structural basis for the“positive inside” rule which is based on the observationthat positively charged residues are four-times moreabundant in the cytoplasmic domains than in translo-cated loops of membrane proteins [23]. At low anionicphospholipid content a higher positive charge is requiredto prevent translocation of cytoplasmic domains whileincreasing anionic phospholipid content results inincreased retention for domains with a lower positivecharge. Placing a negatively charge amino acid within sixresidues from the end of a TM segment can increase itspotential for translocation across the membrane [57].The head groups of zwitterionic phospholipids may con-tribute to the retention of negatively charged residues onthe cytoplasmic side of the membrane by charge pairingbetween the phospholipid amine and the amino acid car-boxylate. PE also dilutes the high negative charge den-sity of the anionic phospholipids PG and CL that wouldincrease the protonated form of acidic amino acids thusfavoring their translocation across the membrane. Thecytoplasmic domains mis-oriented when LacY [4] orPheP [5] are expressed in cells lacking the zwitterionicphospholipid PE contain acidic amino acid residues,which may have a higher potential for translocation in


the absence of PE. Therefore, phospholipids inXuencemembrane protein topology either independently or incooperation with components of the translocon.

1.4. Topological isoforms, mixed topologies, and topological disorders

In recent years, it has become evident that certain nat-urally occurring polytopic proteins exhibit variations inTM topology. There are an increasing number of exam-ples of proteins that are expressed in diVerent topologi-cal forms with diVerent functions. For example, ductinwas found in two diVerent orientations in membranes,one of which serves as the subunit of the vacuolar H+-ATPase and the other serves as a component of themicrosomal connexin channel of gap junctions [58]. Amicrosomal epoxide hydroxylase is found with a diVer-ent topology in the endoplasmic reticulum than in thesinusoidal plasma membrane, where it mediates bile acidtransport [59]. However, current dogma assumes that theinitial topology of a protein in the endoplasmic reticu-lum membrane accurately reXects the topology of theprotein elsewhere in the cell. Do these topological diVer-ences originate co-translationally during membraneinsertion or are they induced by changes in lipid compo-sition as proteins move through diVerent organelles totheir Wnal destination?

Studies examining the synthesis and translocation ofthe prion protein (PrP) at the endoplasmic reticulumhave revealed that it is capable of being made in threetopological forms from the same pool of nascent chains.The majority is completely translocated into the lumen,but another fraction is integrated into the endoplasmicreticulum membrane as single-spanning proteins witheither the N- or the C-terminus in the lumen (NtmPrP orCtmPrP, respectively) [60]. Remarkably, mutations thatincrease the hydrophobicity of residues adjacent to orwithin TM segments result in complete reversal of the(Ntm) PrP topology and cause neurodegenerative dis-ease in either transgenic mice or in some naturally occur-ring inheritable prion diseases [60,61].

The three N-terminal TM helices of glutamate/aspar-tate transporter (GLAST) are encoded by exons 2, 3, and4, respectively. The loss of exon 3 converts the three TMdomains into two and reverses the whole membranetopology of this protein [62]. Moreover, this splice vari-ant of the transporter (GLAST-1a) encodes a functionaltransporter with inverted orientation within the plasmamembrane. Remarkably, in some neurological disordersthe release of glutamate due to anoxia was found to belargely due to an inverse operation of the glutamatetransporter.

It was recently elegantly demonstrated by SCAMTM

that the preexisting, membrane-bound anti-apoptoticprotein Bcl-2 changes membrane topology upon induc-tion of apoptosis. Moreover, this topological change

converts Bcl-2 into a suicide inhibitor of the Bcl-2 familyof proteins, notably Bax and Bak, to mediate anti-apop-totic activity through formation of a dead end non-active supermolecular complex. This complex formationprevents both loss of mitochondrial membrane potentialand release of pro-apoptotic proteins such as cyto-chrome c into the cytosol [63].

It is also important to note that some aspects of pro-tein topology may be expression system dependent. TheP-glycoprotein is localized to mammalian cytoplasmicmembranes, and in its native host the protein exhibits12 TM segments with both the N- and C-terminiexposed to the cytoplasm. When expressed in E. coli, theN-terminal half of the protein assumes the same topol-ogy as in the native host. However, TM segment VII nolonger spans the membrane and TM segments VIII–XIIassume an inverted orientation [64]. Similarly, a citratecarrier of Klebsiella pneumoniae displays 11 TM seg-ments when inserted into dog pancreas endoplasmicreticulum membranes but only nine TM segments whenexpressed in E. coli [65]. One intriguing possibility isthat certain polytopic proteins may require individualphospholipids as specialized membrane components toachieve their proper topology in addition to the basictranslocon components needed for protein insertionand translocation. The topogenic information presentthroughout a polytopic membrane protein might beinterpreted diVerently in prokaryotic and eukaryoticcells, suggesting that problems encountered when tryingto express eukaryotic membrane proteins in prokary-otic hosts may in some cases be related to incorrecttopological organization.

2. Experimental strategies for TM topology assessment

2.1. Overview

The diYculties encountered in the crystallization ofintegral membrane proteins have led to the develop-ment of several alternative approaches for investigationof their structural organization in the membrane. Giventhe enormous number of sequences that are produced ingenome-sequencing projects, it is not realistic to assumethat the structures of all the encoded proteins will begenerated by crystallographic approaches, especially formembrane proteins. The physico-chemical constraintsimposed by the lipid environment and the knownhydrophobicity of individual amino acids provide amethod using hydropathy plots to predict the topologyof a membrane protein [66–69]. Two databases of TMtopologies, MPtopo [70] and TMPDB [71], are availableand can be used to evaluate the reliability of predictedtopologies. The TM topologies in these databases weredetermined experimentally by means of X-ray crystall-ography, NMR, gene fusions, SCAMTM, insertion of


glycosylation sites, and other biochemical methods[147].

However, very often predictive methods generate a mis-leading topology and, due to the simplicity of generatingpredicted topologies, are often cited without further veriW-

cation. Long range-interactions between TM helices, unan-ticipated inter- (translocon and subunits) and intra-protein(salt bridges between charged residues within the hydro-phobic core of the bilayer) interactions, and speciWc lipidprotein interaction are some of the variables not addressedby predictive methods [72]. Therefore, hydropathy analysisof the sequence of polytopic membrane proteins may onlyreveal potential TM segments and their relative orienta-tion as a starting point for designing biochemical experi-ments to establish topological organization.

To verify predicted membrane protein topology mod-els, the existence of all the putative TM domains must beveriWed and the hydrophilic loops must be localized toone side or the other of the membrane. Strategiesemployed are quite varied but utilize the impermeabilityof the membrane bilayer to hydrophilic molecules, thediVerence in properties between the compartments sepa-rated by the membrane, and incorporation into proteinsof a large variety of reporter groups whose orientation ispresumed to reXect the topology of the protein [73,74].Reporter groups can be as simple as a single amino acidsubstitution as in SCAMTM, insertion of a proteolysissite [75], insertion of foreign antigenic reporter epitopes[76], insertion of a glycosylation motif or as complex asfusions of truncated target proteins to reporter proteins.

2.2. Fusions with reporter proteins

An early and still used approach for proteinsexpressed in E. coli is to construct a chimeric proteinbetween successively C-terminal truncated target pro-teins and the N-terminus of a mature reporter protein.The reporters are typically molecules whose properties(for example, enzymatic activity, resistance to protease,antibiotic resistance, and antigenicity) depend on theirsubcellular location. Reporter domains should ideallylack intrinsic topogenic information, be readily identi-Wed, and passively and eYciently follow topogenic infor-mation presented by the nascent target protein fragment.A major drawback of the approach is that it is diYcultto fulWll these requirements without compromising thetopological information of the target protein especiallywhen long range and cooperative interactions areinvolved.

The most extensively used single reporter systems inE. coli have been fusion proteins with alkaline phospha-tase (PhoA) and �-galactosidase (LacZ) [77]. Proteintranslational fusions are obtained by progressively delet-ing the target protein gene from its 3� end and ligatingthe truncated gene to phoA that lacks the promoter andcoding information for its membrane targeting leader

sequence. A complementary approach is to fuse the sametarget protein coding regions to lacZ. Reporter functionis based on the activation of PhoA activity only in theperiplasm where it forms a dimer, acquires Zn+, formsan intrachain disulWde, and in cell lysates is highly resis-tant to proteolysis; PhoA is inactive in the cytoplasmand sensitive to proteolysis. LacZ fusions provide com-plement information since fusions to cytoplasmicdomains are active and fusions to periplasmic domainsshow decreased activity. The characterization of a seriesof membrane protein-PhoA and -LacZ chimeric proteinsyields a reciprocal activity pattern for the reporter pro-teins that is in agreement with the predicted TM topol-ogy of many membrane proteins. Initial screening fortopological location can be done in whole cells based onconversion of substrate analogues of each reporterexpressed in cells to a colored product on agar plates.

The third group of commonly used reporter mole-cules is proteins that confer antibiotic resistance [78].The mature form of �-lactamase (bla gene product),when in the periplasm, confers resistance to antibioticssuch as ampicillin. Since the antibiotic targets are cellwall biosynthetic enzymes, cells expressing cytoplasmic�-lactamase are sensitive to ampicillin. Chloramphenicolacetyltransferase (cat gene product) only confers resis-tance to chloramphenicol in the cytosol where there is asupply of acetyl CoA for inactivation of the antibiotic[79]. Thus, only cells expressing fusion proteins in whichthe mature form of �-lactamase is fused to a periplasmicdomain of a membrane protein or chloramphenicol ace-tyltransferase is fused to a cytoplasmic domain will beresistant to the respective antibiotics.

The major assumption in the fusion approach fordetermining membrane protein topology is that trunca-tion of a membrane protein does not aVect its nativetopology. However, because the complete structure of theprotein of interest is not used and functional tests are notthen available, the eVects of long-range interactionsbetween domains of the protein are not detected by thismethod and can lead to incorrect conclusions. The use of“sandwich” PhoA fusions, in which the reporter isinserted into the membrane protein rather than replacingthe C terminus, have been shown to remedy this problemand may give a more accurate picture of the topology[80]. The entire membrane protein is present and theapproach does not suVer from the drawbacks associatedwith the C-terminal deletion fusion approach. However,inserting fusions into a whole membrane protein mayalter the way the protein folds or inserts into membrane.Therefore, reliance on topological information is safest ifthe chimeric protein retains the original activity.

In general active PhoA is a more reliable marker ofperiplasmic location than active LacZ is of cytoplasmiclocation. The former must reach the periplasm to be acti-vated while the latter when fused to a periplasmicdomain may form the active tetramer in the cytoplasm


and fail to translocate due to its size. The latter problemhas been addressed by joining mature PhoA and the �-fragment of LacZ into a single dual reporter [81]. The �-fragment represents 6% of full length LacZ and is activein the cytoplasm of cells encoding the remaining inactive�-fragment of LacZ. The dual reporter, when fused toperiplasmic domains, produces fusions with high PhoAactivity and, when fused to cytoplasmic domains, pro-duces fusions with high LacZ activity in E. coli strainscapable of �-complementation. Dual indicator platescontaining a blue PhoA-activity dependent chromogenicsubstrate and a red LacZ-activity dependent chromo-genic substrate in conjunction with these reportersallows for initial discrimination between non-informa-tive fusions (white), cytoplasmic fusions (red), periplas-mic fusions (blue) or fusions within TM domains(purple).

Interpretation of reporter activities in terms of topo-logical information is often complicated by variableexpression of the fusion proteins. For example, the lowexpression of periplasmic PhoA may lead to low alkalinephosphatase activity characteristic of cytoplasmicfusions and hence to misinterpretation of experimentaldata unless reporter activities are normalized to the rateof protein synthesis. Therefore, the single reporter fusionapproach involves time-consuming experiments employ-ing pulse labeling, immunoprecipitation, and quantiWca-tion of incorporated radioactivity. Using dual reportersallows for a simple alternative to normalization of PhoAand LacZ activities. It is assumed that the speciWc activi-ties of both the LacZ and PhoA portions of the dualreporter are characteristic of a given fusion point andindependent of the level of expression of the fusion pro-tein; that is, the level of expression of the fusion proteinwill aVect the absolute activity but not the ratio of thetwo reported activities. By normalizing the PhoA andLacZ activities of each fusion to the maximal activityobserved for each reporter enzyme in the set of fusions, acorrection can be made for the intrinsically higher PhoAactivity and the resulting ratio of normalized PhoA toLacZ activities provides readily interpretable informa-tion about subcellular localization of the fusion point.Although the utilization of dual reporters alleviates mostof the above-mentioned problems, this most advancedgenetic approach suVers from lack of detection of long-range eVects on topology.

2.3. Fusions with a glycosylation epitope

An aVective approach to mapping the topology ofproteins present in the endoplasmic reticulum is thelumen speciWc glycosylation machinery. This approachhas become the eukaryotic counterpart of the bacterialPhoA fusion and sandwich techniques. In eukaryoticcells, glycosylation activity is found in the lumen of theendoplasmic reticulum and is carried out by an oli-

gosaccharyl transferase, which catalyzes addition of oli-gosaccharides to the amino group of asparagine residueswithin the consensus sequence Asn-X-Thr/Ser. N-glyco-sylation is a common feature of eukaryotic membraneproteins, and the consensus sequence is usually found inthe largest luminal exposed loops of the protein. SincemodiWcation of the glycosylation site occurs in a com-partment-speciWc manner, the presence of glycosylationprovides information for topological assignment [74,82].Addition of the oligosaccharide chain to a single siteresults in an increase in the apparent molecular mass onSDS–PAGE (sodium dodecyl sulfate–polyacrylamidegel electrophoresis) by about 2.5 kDa. Glycosylation canbe manipulated after synthesis by in situ treatment withglycosidases (Endo H and N-glycosidase F) or by invitro expression of the protein in the presence ofglycosylation inhibitors (tunicamycin) or competitiveacceptor peptides (Ac-Asn-Tyr-Thr and N-benzoyl-Asn-Leu-Thr-N-methylamide) [83].

In the glycosylation scanning mutagenesis approach,consensus glycosylation sites or domains bearing a gly-cosylation site are introduced into membrane proteinsdevoid of glycosylation sites. Plasmids encoding theseengineered proteins are transfected into mammaliancells and localization of the insertion site on the luminalor cytoplasmic side of the membrane is inferred from thepresence or absence of glycosylation of the engineeredprotein, respectively. In the glycosylation fusionapproach, a domain bearing one or more glycosylationsites is fused behind diVerent C-terminal deletionmutants of a membrane protein [84]. The Asn in accep-tor sites is glycosylated only in loops larger than 25 resi-dues, and a sharp cutoV is observed in glycosylation ofsites positioned less than 12 residues upstream or 14 resi-dues downstream of a TM segment [85]. The distanceconstraints imposed on glycosylation sites by the oli-gosaccharyl transferase were used to map the ends of theTM segments on the luminal extramembrane domains ofseveral polytopic membrane proteins [86].

The glycosylation scanning technique assumes thatthe initial topology of a protein in the endoplasmic retic-ulum membrane accurately reXects the topology of theprotein elsewhere in the cell, which, as was mentionedabove, is not always true. Because each strategy has itsown inherent strengths and limitations, a topologicalmodel that is based on the results of several approachesis likely to be more informative.

3. Principles of SCAMTM

3.1. Overview

ModiWcation of thiol groups in proteins has become apowerful technique used to analyze protein structure. Inthe original substituted cysteine accessibility method


(designated as SCAM by Karlin and co-workers [87–92])single cysteine substitutions within a target protein cou-pled with covalent cysteine modiWcation by hydrophilicthiol-speciWc reagents was used to study structure–func-tion relationships and dynamics of membrane proteinfunction (mapping of channel gating residues, identiWca-tion of residues lining a membrane channel, identiWca-tion of residues involved in substrate or ligand binding,etc.). SCAM provides an approach to systematicallymap the residues on the water-accessible surface ofmembrane proteins either at steady state or related toprotein function. SCAM can probe conformationalchanges that result in changes in steric constraints andelectrostatic potential within the vicinity of the substi-tuted cysteines by comparing the rates of reaction withreagents of varying size and charge.

By introducing and modifying cysteines in the poly-topic membrane protein LacY of E. coli, Kaback andco-workers [92] identiWed and deWned the structural rela-tionship between the 12 TM-spanning helices of thisprotein. The movement of TM helices relative to eachother, changes in accessibility of certain residues uponbinding of substrate, and the residues critical to functionwere also demonstrated. The conclusions drawn fromthese extensive biochemical studies resulted in a modelfor the three-dimensional structure of LacY that islargely supported by the recently determined X-ray crys-tal structure [93].

We have adapted SCAM to map and assign TMtopology of polytopic membrane proteins (designedSCAMTM) as a function of membrane lipid composition.In this approach, cysteine residues replace individualamino acids that reside in the putative extracellular orintracellular loops connected to TM segments of a mem-brane protein. The use of SCAMTM was introduced andelegantly applied for Wrst time by [88,94] to establishmembrane protein topology. The orientation withrespect to the membrane is determined using membrane-impermeable thiol reagents with intact membranes orwith membranes that have been permeabilized or disin-tegrated. Alternatively, a combination of thiol reagentscan be used that are membrane-impermeable or mem-brane-permeable due to diVerences in their physicalproperties but with the same chemical reactivity underdiVerent conditions. This method can be used in wholecells [4,94–100], uniformly oriented membrane vesiclesisolated from cells [4,51,65,101–105], intact organelles(vacuoles and mitochondria) [106,107], or in reconsti-tuted proteoliposomes containing uniformly orientedtarget proteins [43,108,109]. A major advantage ofSCAMTM over methods reviewed above is that minimalperturbation of protein structure results from introduc-ing single cysteine residues into a protein. In addition thewhole protein is analyzed, and retention of native func-tion in vivo can be used to verify structural integrity ofthe engineered protein.

Exact protocols and detailed descriptions of method-ology will not be presented since SCAMTM must be tai-lored to speciWc proteins, cell hosts, and hostmembranes. Therefore, a more global approach will bepresented describing the properties of thiol groups andreagents, membrane permeability properties with respectto diVerent reagents, methods of accessing thiols indiVerent compartments, methods of detection and analy-sis, and advantages and limitations of SCAMTM.

3.2. Properties of the cysteinyl thiol

The chemical nature of the reactive portion of a label-ing reagent should be highly reactive with and selectivefor thiol groups and form a stable non-exchangeable ornon-hydrolysable derivative. Maleimides, which areavailable in a wide variety of forms, are particularlysuited for SCAMTM. Maleimide reacts with the ionizedform of a thiol group (Fig. 1A), and this reactionrequires a water molecule as a proton acceptor [88,110].Maleimides are virtually unreactive until they encounteran available thiol group. The pKa of the thiol of cysteinein a water milieu is around 9 and in the hydrocarboncore of the bilayer is around 14 characteristic of cysteinein a nonpolar environment [111]. Therefore, the labelingcharacteristics of intramembrane (unreactive) and extra-membrane (reactive) cysteines would be consistent withtheir localization either in a polar or nonpolar environ-ment, respectively [88,94,96,99,112,113]. This is animportant chemical feature of cysteine residues that isthe main basis of SCAMTM. Examples (discussed later)exist of highly hydrophobic maleimide derivatives thatreact with thiols in a hydrocarbon environment, whichcan be selectively applied to map thiols in TM segments.A possible explanation for the reactivity of the proton-ated thiol group in a lipid environment may be the highconcentration of a hydrophobic maleimide in the hydro-carbon core of the bilayer due to its favorable partition-ing coeYcient. However, conclusions based on reactivityshould be made with caution, since cysteines may bepositioned facing other helices and therefore might beinaccessible but not in a hydrocarbon environment.Local secondary structure or properties of neighboringamino acids may restrict access by thiol reagents. Cys-teine residues facing a hydrophilic pore or near a sub-strate-binding site maybe within a TM segment butchemically reactive due to water channels or pockets[114].

Since the formation of cysteinyl thiolate anions isfavored by increasing the solution pH (optimum pH 8.0–8.5), increasing the pH during labeling should favor thereaction [115]. However, maleimides are known to reactwith primary amines at pH values above 7.5 [116]. There-fore, attempts to increase eYciency of labeling by raisingthe pH of the assay should be thoroughly controlled inorder to ensure that the modiWcation is conWned to


cysteine. An eVective control to rule out non-thiol modi-Wcations is to use a cysteineless target protein. Thus farthe most extreme thiol labeling reaction conditions (mil-limolar concentration of reagent, room temperature, andpH 8.0–8.5) have been utilized without signiWcant modi-Wcation of additional reactive groups on target proteins[94]. Generally, pH 7.0–7.5 is suYcient for eYcient label-ing [4,65,88,94,95,102–104,107,113] and reduces back-ground labeling especially in experiments with orientedinside-out membrane vesicles (ISOV) containing targetproteins with cytoplasmic loops very often containinglysine residues.

3.3. Thiol reactivity in proteins

Cysteine is a relatively hydrophobic, small aminoacid, and its introduction at most positions in a mem-brane protein is likely to be tolerated. Furthermore, cys-teine has little preference for a particular secondarystructure [117,118]. For most water-exposed cysteine res-idues in proteins, the thiol pKa lies in the range of 8–9and formation of cysteinyl thiolate anions is optimum inaqueous rather in a non-polar environment. The processof choosing suitable residues for replacement by cysteineis often empirically determined, and the rationale fordeciding which residues to alter is aided by the followingconsiderations. Secondary structure predicted by com-

puter-aided hydropathy analysis [66–69] (thus far 60–70% reliable) [119] is an initial starting point for the like-lihood that a particular residue is in an extramembranedomain. Replacement of charged residues is generallynot advised because these have a high probability ofbeing topogenic signals or may be involved in long-range interactions. Consideration should be given towhether the replacement will be well tolerated based onstructural and functional information about the protein.If the protein contains stretches of residues of intermedi-ate hydrophobicity that cannot unambiguously be iden-tiWed as membrane spanning, substitutions should bemade approximately every 10 residues.

Ideally the protein under study should be devoid ofall native cysteine residues because these residues mayalso react with thiol-modifying reagents or they mayform disulWde bonds with the engineered cysteines andprevent their interaction with modifying thiol-speciWcreagents. The cysteineless protein serves as the startingtemplate for introducing single cysteine residues atdesired positions as well as a negative labeling control toassure that residues such as lysine are not labeled by thereagents. Alternatively, templates containing naturalcysteines can be utilized in this assay if they do not reactwith the thiol-speciWc reagents. Very often the native cys-teine residues present within TM segments are inaccessi-ble to thiol reagents making it unnecessary to remove

Fig. 1. Thiol-modifying reagents widely used in SCAMTM. (A) The reaction of the thiolate anion of cysteine with maleimide by nucleophilic additionto the double bond of the maleimide ring. (B) The structure of the maleimide- and biotin-containing labeling reagent MPB. (B) The structure of theblocking reagent AMS.


these cysteines. However, the possibility always existsthat engineered cysteines may form disulWdes or changesin protein structure may expose the native cysteines. Alimitation of SCAMTM is with proteins where cysteinepairs form disulWdes critical to the folding of the protein.Removal of these cysteines or introduction of additionalcysteines might cause misfolding. Therefore, a prerequisitefor each cysteine replacement is retention of function thatprovides assurance of retention of near native structure.

The native cysteine residues are usually changed intoalanine or serine residues which are small, commonlyfound in membrane proteins and appear to be toleratedat most positions thus rendering an active protein. Forexample the eight native cysteine residues of LacY weresimultaneously replaced to yield a cysteineless templatethat retained at least 50% of its wild-type activity. Of the417 single cysteine replacements in LacY, only four dis-rupted transport function [92] supporting a near nativestructure for the other replacements. However, loss offunction may not result in signiWcant structural changesif these substitutions only aVect substrate binding or cat-alytic processes. Thus, cysteine-scanning mutagenesispermits topology assessment under conditions in whichthe proteins are active, strongly suggesting the proteinstructure is not seriously altered.

To obtain a minimal topological map a single cysteinereplacement in each of the putative extramembraneloops should be expressed from a plasmid and analyzedin appropriate host cells. In practice, several cysteinereplacements or complete cysteine scanning across extra-membrane loops and into TM segments is required for amore precise mapping of topology. Secondary structure,as discussed in Section 9, may sterically prevent access tocysteines in extramembrane loops, which requires analy-sis of several cysteine replacements along a loop. Cys-teine residues closer to the membrane interface generallyreact slower than those near the center of extramem-brane loops [94] and these diVerences can be used toassign residues at the membrane-aqueous interface. Thehost strain for plasmid expression should be deleted ofthe target protein gene if it contains native cysteines andis expressed at levels high enough to be detected in theassay. Since LacY expression is high in cells induced forlac operon expression, it was deleted in SCAMTM analy-sis [4]. However, the level of chromosomal expression ofPheP was not high enough to be detected in the assay sodeletion was not necessary [5].

The physical and chemical properties of thiol-speciWcreagents and basic properties of membrane proteins pro-vide a rational basis for conclusions about exposure andpositioning of residues that become modiWed. Thehydrocarbon core of a model phosphatidylcholinebilayer is about 27 Å wide with the phospholipid headgroups occupying about 5 Å on either side of the hydro-phobic domain [120]. The majority of TM spanningregions are �-helical so that a minimum of 18 hydropho-

bic amino acids (1.5 Å per turn of the helix) are suYcientto span the hydrocarbon core of the bilayer and another2–3 amino acids are needed to bridge the phospholipidhead group region on either side of the membrane.Recent crystal structure data indicates that TM seg-ments of 18–20 amino acids in length are common butsegments longer than 20 amino acids exist and areobliquely oriented to the plane of the bilayer. The cys-teinyl thiol group lies some 8–10 Å from the peptidebackbone [111] so that maleimides derivatized withbulky residues would have access to cysteine residuesonly about 5–6 amino acids into a TM segment [110].Therefore, for a 20 amino acid TM segment about 8–10residues (12–15 Å) at the center of the TM segmentwould not be accessible to these reagents. For TM seg-ment 7 of UhpT, an E. coli sugar phosphate transporter,accessibility of cysteines by bulky maleimides Wts theabove dimensions while smaller probes reacted with resi-dues more that 10 Å into the hydrocarbon core [91,114].

4. Application of SCAMTM

4.1. Overview

Thiol-speciWc reagents are available in a large varietyof sizes, polarity, and monitoring features, as will be dis-cussed in Section 4.3. Membrane permeable or mem-brane impermeable thiol-reagents can be employed toselectively label the residues from either both sides of themembrane (hydrophilic maleimides) and the residuesfacing the lipid core of the bilayer (hydrophobic malei-mides) or only residues exposed to the outer surface ofthe membrane, respectively. The outer membrane ofE. coli allows small molecules under about 600 Da pas-sage to the periplasm and access to the inner membraneso these reagents can be used in whole cells without theneed to disrupt membrane structure [121]. The doublebond of the maleimide group reacts with the thiol-groupof cysteine to form a thioether bond (Fig. 1A) that is sta-ble to reducing agents such as �-mercaptoethanol (�ME)or dithiothreitol. The reaction rate of diVerent thiols iscontrolled primarily by their surface exposure and prox-imal environment. For example cysteines buried withinthe structure of native proteins are essentially unmodi-Wed by N-ethyl maleimide (NEM, although it can crossthe membrane) or Xuorescein-5-maleimide (FM, mem-brane impermeable) over a time scale of 10 min. How-ever, when the cysteines are exposed by SDSdenaturation of proteins, they became accessible to theaqueous phase and are rapidly modiWed by bothreagents within minutes [88].

The general design of labeling experiments to distin-guish between cysteines located in an extracellular orintracellular domain is outlined in Fig. 2. Extracellular(periplasmic for E. coli) residues are those that are


labeled in intact cells but not in ISOV unless they arepermeabilized. Intracellular (cytoplasmic) residues arethose that are labeled in ISOV but not in intact cellsunless they are permeabilized. Pre-blocking intact cellsor ISOV with a thiol reagent that is transparent in thedetection phase of the procedure allows selective label-ing of luminal cysteines after permeabilization. Theresults of this approach are valid only if the modifyingreagent is thiol-speciWc and membrane impermeable, ori-entation of ISOV is uniform and opposite to that ofcells, and permeabilization does not expose stericallyhindered or water inaccessible cysteine residues. The lat-ter can generally be assumed if cells and ISOV are per-meabilized before reaction with thiol reagents and thenlabeled with a detectable thiol reagent before and afterreaction with a non-detectable blocking reagent. Theblocking reagent should completely prevent labeling and

Fig. 2. General strategy for SCAMTM using impermeable thiolreagents. A target membrane protein containing a single cysteinereplacement exposed either to the extracellular (periplasmic, left halfof circle) or intracellular (cytoplasmic, right half of circle) side of themembrane is expressed in cells. Preparation of uniformly orientedISOV results in an opposite orientation for the same cysteine residues.Half the cells or ISOV is reacted with a detectable thiol reagent to spe-ciWcally label the externally exposed cysteine and the other half isreacted with a non-detectable thiol reagent to protect external cyste-ines in subsequent labeling steps. The latter half of cells or ISOV is per-meabilized to expose the interior cysteine and reacted with a detectablethiol reagent to speciWcally label internal cysteine residues. Cells andISOV are then analyzed for labeling of the target protein by the detect-able thiol reagent. Note that the cysteine labeling (+) and blocking (¡)patterns in ISOV are the mirror image of the patterns in whole cells.

the labeling without blocking should be to the sameextent for external cysteines as seen in the protocol out-lined in Fig. 2. Cysteine residues that are not labeledunder any conditions are either inaccessible due to beinglocated in a TM segment embedded in the hydrocarboncore of the bilayer (intramembrane) or restricted by sec-ondary structure. More details concerning each step,diVerent reagents, considerations for controls, dealingwith inaccessible cysteines, and interpretation of resultswill be expanded in the following sections.

4.2. General protocol for SCAMTM

Single cysteine replacements are expressed in theappropriate host, and the cells are harvested and sus-pended in modiWcation buVer or ISOV are prepared [4]and suspended in modiWcation buVer. Cells or ISOV aretreated under a variety of conditions and with diVerentthiol reagents in order to establish where the cysteine res-idues are located as described in Fig. 2. A maleimide-based thiol reagent is added and the modiWcation reactionterminated by 5–10-fold dilution with buVer alone [107] orby adding a 50–100-fold-excess of either �ME, dithiothre-itol [94,103] or cysteine [105,122] to destroy the unreactedmaleimide. The concentration of thiol reagent and time ofreaction is empirically determined by using the most vig-orous conditions that do not result in labeling of a cyto-solic protein or luminal thiol scavenger as described inSections 5 or 12, respectively. Termination of the reactionis followed immediately by several cycles of centrifugationand washing with thiol quencher [94,123] to remove excesslabeling reagent. Dilution rather than centrifugation wasemployed [65,107] to eliminate additional variability inyield associated with pelleting and resuspending of cells.Excess reagent can also be removed by centrifuging theterminated reaction mixture through small columns of gelWltration resin [4]. The last two procedures should be usedwhen sequential treatments of the sample is required suchas pre-blocking with one thiol reagent followed by reac-tion with a second thiol reagent [4] or to avoid lysis offragile preparations like intact spheroplasts, vacuoles[65,107] and proteoliposomes [43].

After labeling, cells or ISOV are solubilized with theappropriate detergent or detergent mixture such as SDSalone [4,104], Triton X-100 alone [65,105] SDS and Tri-ton-X-100 [113,124], Chaps [124,125], octylglucoside,deoxycholate, cholate, and Tween 20 [102], octylgluco-side [111], �-D-dodecylmaltoside [95], or nonidet P-40and sodium deoxycholate [112,126]; use of detergentsother than SDS may require lysis of cells by sonicationprior to solubilization. Conditions must be empiricallydetermined that yield a non-aggregated soluble targetprotein throughout the remainder of the procedure. Forexample many membrane proteins aggregate if boiled inSDS, and LacY forms irreversible polydisperse aggre-gates if solubilized by Triton X-100 alone.


The thiol reagents react with cysteine residues presentin all other proteins in the membrane. Immunoprecipita-tion of the membrane protein of interest or a rapid puri-Wcation step is necessary to eliminate other labeledproteins. A biotin-maleimide labeled protein can berecovered from cell lysates directly with streptavidin–agarose beads [100], and can then be detected byWestern blotting using a target-speciWc antibody. Forimmunoprecipitation of labeled protein from solubilizedsamples, polyclonal and monoclonal antibodies[4,94,102,107,110,113] have been widely utilized. Anti-gen–antibody complexes can be isolated using precipita-tion with Pansorbin (Staphylococcus aureus cells) [4,113],protein A–agarose [102], or protein A or G–Sepharosebeads [98,107,125,126]. If antibodies speciWc to the pro-tein under study are not available, then epitope tags suchas myc [124,126] or aYnity tags such as His6 [124] can beincorporated at the C-terminus of the target protein foreither immunoprecipitation [124] or isolation by Ni2+

chelated aYnity resin packed into micro-columns orattached to agarose beads [65,103–105,111,116,127,128].Use of aYnity methods with His-tagged proteins andsmall-scale batch puriWcation procedures is becomingthe method of choice since the labeled protein can bedirectly extracted from the resin with SDS-containingbuVers followed by SDS–PAGE [104]. Of course proteinfunction or topology should not be compromised by thepresence of the tag.

Following modiWcation and isolation, the target pro-tein is resolved by SDS–PAGE, transferred to a solidsupport, and detected by Western blotting or one of thefollowing techniques. Thiol reagents are available thatcontain a biotin group [4,94,98], a Xuorescent group[51,88,103,116,122], or a radiolabel [113,129], allowingdetection of labeled proteins by avidin linked to horseradish peroxidase (avidin-HRP) and indirect chemilumi-nescence detection, Xuorescence, or autoradiography,respectively. Signals can be quantiWed using availableImaging systems and software.

A major problem is the variability between samplesdue to mechanical loss during the work up or due todiVerences in expression level of individual replacements.For example, of the 41 cysteine replacements within theABC multidrug transporter LmrA, 40 were expressedand present in the membrane, but at diVerent levels [103].Variability can be corrected for by using a His-taggedtarget protein. The thiol derivative can be detected basedon its properties, and the amount of target protein pres-ent can be detected using an antibody directed againstthe His tag. Since these can be done on the same blot,thiol signal can be normalized for the amount of targetprotein. Alternatively, Western blots of duplicate sam-ples or of the same blot can be probed by antibody spe-ciWc for the target protein to normalize the signal[4,100,111,112]. After recording a Xuorescence or radio-active proWle of a derivatized target protein resolved by

SDS–PAGE, the signal can be normalize to the targetprotein recovered as determined by staining the same gelwith Coomassie brilliant blue [111,129] or silver stain[123]. If no labeling occurs with the thiol reagent, it isimportant to verify that the target protein was expressedand is present on the blot.

4.3. Properties of thiol-speciWc reagents

The choice of thiol-speciWc reagents is inXuenced byphysical–chemical properties, size, chemical reactivity,and monitoring features suitable for cysteine modiWca-tion followed by detection. Following is a description ofthe various thiol-speciWc reagents that have been used.

4.3.1. Labeling with detectable thiol-speciWc reagentsImpermeable thiol reagents that can be easily

detected after modiWcation of target proteins are essen-tial for successful application of SCAMTM. Biotin-linkedmaleimides such as 3-(N-maleimidylpropionyl) biocytin(MPB, Fig. 1B) and (+)-biotinyl 3-maleimidopropio-nomidyl-3,6-dioxaoctanediamine (MPEOB) are particu-larly useful due to their low membrane permeabilityproperties and formation of stable non-hydrolysablebonds with thiols. After SDS–PAGE of target proteinsisolated by immunoprecipitation or aYnity tags, the bio-tinylated target proteins are easily detected using avidin-HRP and chemiluminescence [94,110]. Owing to therelatively long hydrophilic spacer between the biotinylgroup and the reactive maleimide, MPEOB is fully solu-ble in aqueous solution and less likely to penetrate the E.coli cytoplasmic membrane [5]. This reagent was selectedto study reversible TM orientation of PheP in responseto a change in phospholipid composition of E. coli mem-branes [5] and establish TM topology of LacY reconsti-tuted in proteoliposomes [43]. MPB is morehydrophobic than MPEOB and must be added dissolvedin dimethyl sulfoxide so conditions must be establishedfor minimal membrane permeability by MPB. MPB hasbeen extensively utilized in SCAMTM under conditionsof low membrane permeability for detection of externalsurface exposed cysteines [94,97,99,102,107,109,127] andunder conditions of membrane permeability (higher con-centration and longer incubation times) where MPBmodiWes water accessible thiols on both sides of themembrane [65,95,98,130].

FM, which at neutral pH is a dianion and membrane-impermeable, has also been used in SCAMTM [103–105]and has the advantage of being Xuorescent and directlydetectable in gels after SDS–PAGE. Cysteines in puta-tive TM segments were found to be inaccessible to FM,whereas cysteines in polar loop regions were readilylabeled [103,104]. High pH in the assay buVer can help toreduce the membrane permeability of FM by promotingthe ionization of the carboxyl groups in the molecule[104] and was utilized along with a cysteineless control


protein to demonstrate that the labeling was speciWc tocysteine. Oregon Green 488 maleimide carboxylic acid(OGM), another highly negatively charged Xuorescentmaleimide, has also been successfully used in SCAMTM

[96,111,116]. N-Biotinylaminoethyl methanethiosulfonate(MTSEA-biotin) was used to label water exposed cysteineresidues of the serotonin transporter expressed in Intestine407 cells [131] or HeLa cells [100] and to map the extracel-lular boundary for TM segments of human reduced folatecarrier in Chinese hamster ovary cells [126]. MTSEA-bio-tin was also successfully used to map the membranetopology of the renal Na+/K+-ATPase �-subunitexpressed in baculovirus transformed insect cells [125].This reagent has never been tested in bacterial systems.

4.3.2. Membrane impermeable thiol speciWc blocking reagents

Also required for SCAMTM is a set of impermeableblocking reagents that eVectively react with thiolsexposed to solvent but are transparent in the detectionphase of the procedure. These reagents are useful in sev-eral ways. They can be used with membrane imperme-able labeling reagents as outlined in Fig. 2 to restrictdetectable labeling to luminal cysteine residues. Alterna-tively, they can be used with membrane permeablereagents to identify internal cysteines. NEM readilycrosses the membrane bilayer but still labels primarilywater accessible cysteines. MPB and FM are generallyconsidered membrane impermeable, but at high concen-trations and with long incubation times, luminal waterexposed cysteines can be labeled [94,127]. Cells or ISOVcan Wrst be modiWed with a hydrophilic blocking reagentand then treated with radiolabeled NEM or high con-centrations of MPB or FM to access internal waterexposed cysteines. Even under the most stringent condi-tions some of the more hydrophobic labeling reagents(MPB) have limited membrane permeability. Pre-block-ing external cysteines before labeling intact cells or ISOVprovides a means of estimating the degree of labelingdue to this low permeability.

A widely used blocking reagent is 4-acetamido-4�-maleimidylstilbene-2,2�-disulfonic acid (AMS, Fig. 1C)since it possesses two charged sulfonate groups and ishighly soluble in water. The size and the charged nature ofthis reagent results in its demonstrated inability to crossthe cytoplasmic membrane of bacteria or the plasmamembrane of mammalian cells [65,94,95,107,113,124,127].AMS treatment of E. coli cells and spheroplasts modiWeda cysteine residue exposed to the periplasmic side of themembrane but did not modify a cytoplasmic protein,elongation factor Tu [132]. ConWrmation of labeling ofexternal water-exposed cysteines by MPB was achieved byWrst blocking putative external cysteines with AMS inintact E. coli cells, oriented membrane vesicles, proteolipo-somes [4,43,65,95,107,127], intact vacuoles [107], andmammalian cells [94,99,126].

The membrane impermeable, non-Xuorescent, quater-nary amine [2-(trimethylammonium)ethyl]methanethio-sulfonate bromide (MTSET) showed a side-speciWcmembrane modiWcation pattern [103,111,116,123].External cysteines were identiWed by initial preincuba-tion with membrane impermeable and strongly acidic p-chloromercuribenzosulfonate (pCMBS) and subsequentlabeling with fully membrane permeable radiolabeledNEM [101]. Both pCMBS and MTSET react with peri-plasmically exposed cysteines of the E. coli Na+/H+ anti-porter 10-fold faster in right-side out membrane vesicles(RSOV) than in ISOV. Anionic lucifer yellow iodoaceta-mide (LYIA) and positively charged bromotrimethylam-moniumbimane bromide (BTMB) have also been usedas blocking reagents in MPB labeling experiments[98,112]. The diVerence in the labeling between intactcells and ISOV with FM was further conWrmed in pro-tection experiments with 2-[(4�-maleimidyl)ani-lino]naphthalene-6-sulfonic acid (MIANS) [105]. Thetwo positively charged and highly water-soluble thiol-speciWc reagents MTSET and 2-(aminoethyl)methane-thiosulfonate hydrobromide (MTSEA) have been usedas blocking agents in combination with the membraneimpermeable maleimides OGM and FM to verify themembrane topology of cysteines strategically placed indiVerent proteins [96,103,111,116,123]. However, sinceMTSEA is a weak base with a pK of around 8.5, it is pos-sible that it equilibrates across the membrane in itsundissociated form. Therefore, this reagent should beavoided in topology assays [97,101]. The labeling ofexternal residues with FM and NEM [103,104] or 4-acet-amido-4�-[(iodoacetyl)amino]stilbene-2,2�-disulfonic acid(IASD) [123] was also veriWed by prior blocking withMTSET. Finally, FM and MIANS are Xuorescent sothey can be used as primary detectable labeling reagentswith AMS as a blocking agent or as blocking reagents ifa biotinylated-labeling reagent is used.

4.3.3. Labeling thiols within the hydrophobic core of the bilayer

Although the protonated thiol located in the hydro-phobic core of the bilayer should display low reactivitywith thiol reagents, there are several highly hydrophobicthiol reagents that react at appreciable rates with cysteineresidues clearly in TM segments in a hydrophobic envi-ronment. These can be used to distinguish between inac-cessibility due to secondary structure and inaccessibilitydue to location in a TM segment. Since these hydrophobicreagents will also modify water-exposed residues on bothsides of the membrane, they must Wrst be blocked aftermembranes are permeabilized and prior to labeling. Iden-tiWcation of intramembrane cysteine replacements in aprotein was done by using a combination of hydrophilicFM as a blocking reagent and hydrophobic benzophe-none-4-maleimide (BM) as a labeling agent [106,133]. BMis a lipid soluble thiol regent that is large enough to cause


an increase in the molecular weight of relatively smallmembrane proteins or polypeptide fragments, which isdetectable by a shift in mobility during SDS–PAGE. Evi-dence for intramembrane location of engineered cysteineresidues in the dog kidney Na+/K+-ATPase was obtainedby exploiting the hydrophobic properties of CPM (7-diethylamino-3-(4-maleimidylphenyl)-4-methylcoumarin)and the hydrophilic and membrane impermeable proper-ties of AMS [51]. Since CPM is highly Xuorescent andAMS is not Xuorescent, speciWc labeling by the former canbe followed by visualizing the Xuorescent target protein.BM and CPM also react with cysteines exposed to theaqueous phase; therefore, to identify thiols within thehydrophobic core of the membrane bilayer, water exposedthiols should Wrst be blocked on both sides of the mem-brane with a hydrophilic thiol reagent.

5. Membrane permeability

SCAMTM is based on the controlled membrane per-meability by sulfhydryl reagents. Membranes either intheir native state or due to experimental manipulationcan be slightly permeable to labeling reagents. Mem-brane permeability of some reagents such as MPB, FM,and OGM are time and concentration dependent. Celllysis during a labeling experiment will also result inlabeling of intracellular cysteines. Therefore, optimallabeling conditions must be established for each reagent,cell, membrane vesicle, or proteoliposome used.

The membrane permeability of a reagent can be testedby quantiWcation of the degree of labeling of an abundantcytoplasmic protein that is rich in surface exposed cysteineresidues. In E. coli LacZ is ideal for this purpose due to itscontent of cysteine, mobility in a region on SDS–PAGEdevoid of other major proteins, availability of mutantslacking the enzyme as a control, and availability of anti-body against LacZ. In this case a labeling experiment withboth intact cells and permeabilized cells is carried outexcept that soluble proteins rather than the membranefraction are analyzed after immunoprecipitation.

Based on protocols described above, both PE-con-taining and PE-lacking cells derived from diVerentstrains were labeled with and without pretreatment withtoluene (Fig. 3). As discussed in Section 7, controlledtreatment of cells with toluene permeabilizes the innermembrane without lysing cells. All cell types exhibitedstrong labeling of LacZ after permeabilization with tolu-ene (Fig. 3A, lanes 1 and 3). Parental JP6488 cells (PE-containing) showed weak labeling (about 10%) of LacZwithout pretreatment with toluene (lane 2), while PE-lacking cells derived from JP6488 showed almost 50%labeling of LacZ in the absence of toluene (lane 4). Incontrast, AD93 cells showed essential no labeling ofLacZ in the absence of toluene irrespective of the mem-brane lipid composition. This experiment clearly shows a

signiWcant diVerence in the permeability of diVerent hoststrains and emphasizes the need to screen host strains forreagent permeability prior to initiating experiments.

A similar experimental approach can be used to opti-mize the reaction conditions to maximize labeling ofexternal cysteines and minimize labeling of internal cys-teines. Cells or membranes should be treated with vari-ous concentrations of reagent from 10 �M to 1 mM, attemperatures from 0 to 25 °C, and for various lengths oftime from 5 min to 1 h. The following experiment com-pares treatment of strain AD93 with two commerciallyavailable biotinylated maleimides. MPB is usually madeup as a stock solution in dimethyl sulfoxide or dimethyl-formamide due to its low solubility in water [4,94,110]while MPOEB can be used directly from a water stocksolution [5]. Both reagents showed increased permeableof the inner membrane with increasing reagent concen-tration, as indicated by the increase in the amount ofLacZ labeling in PE-containing or PE-lacking cells(Fig. 3B). Membranes were more permeable to MPBthan MPOEB and PE-lacking cells were more permeablethan PE-containing cells to both reagents. However, nosigniWcant labeling of LacZ occurred in all cell typeswhen either reagent was used at 100 �M for 5 min.

Other cytosolic bacterial markers such a glutathione[88] or elongation factor Tu [132] have been used toaccess membrane permeability. The permeability of the

Fig. 3. Control experiments demonstrating membrane impermeabilityof labeling reagents. PE-containing and PE-deWcient cells derived fromtwo diVerent host backgrounds (JP6488 and AD93) were labeled with100 �M MPB (A) or in the presence of varying concentrations of MPBor PMEOB (B) at room temperature for 5 min. In (A), cells weretreated as indicated (+) with a Wnal concentration of 0.5% tolueneprior to MPB treatment. The cells were lysed with detergent and thecytoplasmic fraction retained, after removal of the membrane fractionby centrifugation, for immunoprecipitation with anti-LacZ polyclonalantibody. The immunoprecipitates were subjected to SDS–PAGE andbiotinylated protein was detected using avidin-HRP and chemilumi-nescence.


eukaryotic cytoplasmic membrane can be tested bymonitoring modiWcation of the Ca2+-ATPase, which islocalized in the endoplasmic reticulum with its nucleo-tide-binding domain (containing 13 cysteines) facing thecytoplasm [94].

6. Membrane orientation

Most of thiol-speciWc reagents utilized in TM topol-ogy assays (FM, OGM, and MPB) in intact E. coli cellshave intrinsic permeability with respect to the outermembrane and therefore preparations of spheroplasts orRSOV is usually not required [95,96,105]. In one reportthe periplasmic exposure of cysteine residues to MPBwas enhanced by labeling of E. coli cells in the presenceof 50 �M polymyxin B, which permeabilizes the outermembrane [102]. However, there is utility in obtaining anindependent assessment of TM topology in ISOV. Thethiol labeling and protection patterns in whole cells andISOV should be mirror images of each other. If not,replacement positions inaccessible to reagent in wholecells but accessible in ISOV may be sterically hinderedby cellular components and not truly buried in the mem-brane. Many initial reports using SCAMTM concludedthat cysteines were in intracellular extramembranedomains based on the lack of labeling by membraneimpermeable reagents or were in intramembranedomains by lack of labeling by permeable reagents inwhole cells rather than on a positive result. Comparingintact cells or RSOV with ISOV gives a positive signalwith the same reagent for each cysteine replacement intwo oppositely oriented membrane preparations, thusavoiding the problem of diVerences in reactivity of diVer-ent reagents, scavenging of the label by cytoplasmic thiolgroups, or interference by interaction of the target pro-tein with other cellular components. Therefore, obtain-ing topological results with ISOV consistent withconclusions drawn from whole cell experimentsstrengthens conclusions on topological organizationparticularly for unexpected results.

Conclusions drawn from analysis of ISOV are validonly if it can be shown that the ISOV are sealed, imper-meable to the reagents, and have a uniform orientationopposite to that in whole cells. Fortunately, passingE. coli cells through a French press at a relatively lowpressure of 8000 psi results in a yield of over 90% sealedISOV [4]. Conditions for preparation of membrane vesi-cles varies among strains and may be signiWcantly diVer-ent for cells with changes in membrane lipidcomposition or mutations in membrane function.SCAMTM is sensitive to contamination of the popula-tion of ISOV by RSOV, unsealed membrane vesicles orunbroken cells. Therefore, control experiments shouldbe carried out as outlined below to establish generationof sealed ISOV.

Leader peptidase (Lep) of E. coli is a typical bitopicmembrane protein whose membrane topology has beenextensively studied using various chemical approaches.The orientation of Lep has been established in both PE-containing [134] and PE-deWcient cells [34] with itslarger catalytic domain oriented to the periplasm forprocessing of translocated proteins. Since misorienta-tion of the catalytic domain would be lethal, orientationof Lep is a direct measure of orientation of membranesanalyzed from any viable cell. Lep has only one free cys-teine residue accessible from the cytoplasm, and thedisulWde-linked cysteines in the periplasm are crypticunless reduced (see Fig. 4A). Moreover trypsin cleavagesites in both extramembrane domains (large globular P2in the periplasm and small P1 in the cytoplasm) of Lepprovide sites for limited trypsinolysis that yield uniquesized degradation products. Trypsin treatment of RSOVmade by lysing spheroplasts produces a small cytoplas-mic trypsin-resistant fragment (TRF II) while proteoly-sis of ISOV produces the larger TRF I. Leaky vesicleswould allow access of trypsin to both sides of the mem-brane and result in formation of a 5-kDa fragment. Apolyclonal antibody recognizes all the above forms ofLep on Western blots. As shown in Fig. 4B, trypsintreatment of RSOV made from PE-containing and PE-lacking membranes yields TRF II. The small amount ofTRF I-like material results from autohydrolysis of Lepeven without trypsin addition. Trypsin treatment ofISOV results in only TRF I. In both cases no 5-kDafragment was observed so RSOV and ISOV made fromPE-containing and PE-lacking cells are uniformly ori-ented and sealed.

Analysis of Lep in these vesicles using SCAMTM vali-dated the method using a known membrane protein andveriWed membrane orientation and lack of membranepermeability by the labeling reagent (Fig. 4C). Lep wasnot biotinylated in whole cells using MPB unless perme-abilized by toluene. Pretreatment with �ME increasedbiotinylation due to reduction of the periplasmic disul-Wde bond. The small amount of labeling observed wasvariable and was blocked by AMS, which is less mem-brane permeable than MPB, suggesting this labeling wasdue to some cell lysis. With ISOV, biotinylation occurredwithout toluene treatment and was completely blockedby AMS. The same results were obtained for PE-con-taining and PE-lacking membranes.

Thus Lep, an essential cytoplasmic membrane proteinof E. coli, appears to tolerate drastic changes inphospholipid composition [4,23,34] and has naturalcysteine positions that are clearly extracellular or intra-cellular. For all these reasons, analysis of Lep can serveas a simple diagnostic tool and powerful control fordetermining the sealed state and orientation of wholecells and membrane vesicles as well as for establishingfavorable biotinylation conditions for extramembraneresidues.


7. Labeling internal cysteine residues by disrupting the membrane barrier

Inability of a membrane impermeable reagent to labela cysteine residue in intact cells or vesicles may be due tosteric constraints by the local structure, residency withina TM segment, or residency in a luminal extramembranedomain. Therefore, in order to conclude that a positionis on the luminal side of the membrane, it is necessary to

Fig. 4. Lep as a vehicle for determining intactness and sidedness of ori-ented vesicles. (A) The orientation of leader peptidase (Lep) in wildtype E. coli for RSOV, whole cells (Cells), and ISOV. The larger C-ter-minal domain of Lep is exposed to the periplasm in whole cells or theexterior in RSOV and contains two cysteine residues in disulWde link-age. The smaller cytoplasmic loop that connects the two TM segmentsof Lep contains a single cysteine. Trypsin treatment of RSOV digeststhe periplasmic domain of Lep leaving the smaller TRF II. Trypsintreatment of ISOV digests the cytoplasmic loop leaving the largerTRF I. (B) The fragmentation pattern generated by trypsin treatmentof Lep in either RSOV or ISOV from either +PE or ¡PE cells. (C) Theresults of various treatments of either whole cells or ISOV from either+PE or ¡PE strains with various thiol-speciWc reagents or tolueneprior to immunoprecipitation with Lep-speciWc antibody followed bySDS–PAGE and detection of biotinylated protein using avidin-HRPand chemiluminescence. This Wgure was reproduced with permissionfrom [4]. (Copyright 2002 EMBO) where experimental details can befound.

acquire positive labeling of the residue by performing acomplementary experiment using ISOV as described inSection 6 or by showing that the protein can be labeledin a non-compartmentalized system, i.e., in permeabili-zed cells or after membrane disintegration. Althoughresults using ISOV as a complementary approach tousing whole cells are very convincing, the approachsuVers from being time consuming and the requirementto establish uniform inverted orientation of the vesicles.Direct permeabilization of cells is simpler and can beused in conjunction with ISOV as shown in Section 6.However, the approach must be optimized for each celltype and assumes that only cysteines in luminal domainsare exposed by the treatment. Permeabilization of cells,organelles or oriented vesicles can be combined withprior blocking of external surface exposed cysteineresidues with membrane impermeable reagents so thatsubsequent labeling will only reXect luminal residues(Fig. 2).

The membrane orientations of the cysteines intro-duced into the predicted extracellular or cytoplasmicloops of a fully functional cysteineless human reducedfolate carrier expressed in Chinese hamster ovary cellswere determined by treatments with the MPB, NEM,and AMS combined with the Streptococcus pyogenescholesterol-speciWc pore-forming toxin streptolysin O(SLO) [126] which permeabilizes host plasma mem-branes and facilitates the labeling of internal cysteineswithout aVecting intracellular membranes [97]. In theseexperiments cells expressing constructs with the cysteinesubstitutions within loops predicted to face the cytosolwere treated with SLO (0.5 �g/ml) prior to treatmentwith MPB. A sample of cells was also pretreated withAMS before incubation with SLO to further verify cys-teine accessibilities in the absence or in the presence ofpermeabilization. Biotinylation of cysteines placed in thepredicted intracellular loops was only detected after cellpermeabilization with SLO. Since these cysteines werenot accessible to AMS during the pretreatment, no com-petition for labeling was observed with this reagent.However, the labeling of these residues was abolishedupon pretreatment with the membrane-permeablereagent NEM. The labeling with MTSEA-biotin wasused to deWne the accessibility of internal and externalresidues within the loop Xanking TM segment 2 of ratserotonin transporter in intact Intestine 407 or HeLacells and cells treated with digitonin (0.0025%) for 4 minto permeabilize the plasma membrane [100,131].

The extracellular residues should be accessible to thelabeling agent to the same extent both in intact and per-meabilized cells since in the presence of permeabilizingagent all the extramembrane cysteine residues regardlessof their location should be available for modiWcation.However, this is not always the case as was observedwith using toluene to permeabilize bacterial cells. Tolu-ene has been employed extensively to supply enzymes


within cells with substrate by making holes in the mem-brane without lysing cells [135]. Theoretically, treatingwith high concentrations of toluene would make cellsmore permeable and expose cytoplasmic cysteines tothiol probes. However, higher levels of toluene mightseriously alter membrane protein architecture and thesolvent accessibility of cysteine residues. Therefore, it isimportant to determine a critical concentration of tolu-ene, which allows exposure of cytoplasmically orientedcysteines without interfering with their labeling.

Cells expressing PheP with a single cysteine located inthe periplasmic loop P5 were pretreated with an increas-ing amounts of toluene and the labeling of the proteinwith MPEOB was compared to a parallel sample with-out treatment with toluene (Fig. 5). Since P5 is a peri-plasmic domain, labeling of the single cysteine shouldnot be inXuenced by toluene treatment. Unexpectedly,the higher levels of toluene did aVect labeling byMPEOB (lanes 1, 2, 3, and 4) as compared with no tolu-ene treatment (lane 8). The optimal concentration wasfound to be around 0.5% (lane 5). Accordingly, all tolu-ene-related experiments were then conducted under thisoptimal concentration, which is enough to permeabilizethe cell but not too high to aVect labeling eYciency ofextramembrane domains of the protein of interest. Theoptimal toluene concentration is strain dependent andwas found to be ineVective as a permeabilizing agent(Jun Xie, personnel communication) in E. coli cells lack-ing PE in which the foreign lipid monoglucosyl diacyl-glycerol was introduced at 30–50 mol% of total lipid [15].Therefore, toluene may not be universally eVective inbacteria with lipid compositions signiWcantly diVerentfrom E. coli.

Cell lysis approaches have also been used to gainaccess to internal cysteines. Several reports have usedsolubilization of membranes with detergents[104,106,107,114]. Some cysteine mutants were notlabeled with the thiol-speciWc probe applied from eitherside of the membrane, indicating that the residues wereeither located within the membrane or buried within atightly folded pocket of the protein. However, these cys-teines were readily modiWed after protein denaturationor membrane solubilization with detergent making it is

Fig. 5. Determination of critical concentration of toluene to permeabi-lize cells. E. coli cells expressing PheP with a single cysteine in periplas-mic domain P5 [5] were incubated for 5 min at room temperature inthe presence of varying concentrations of toluene prior to reactionwith 100 �M of MPEOB. The cells were lysed and solubilized mem-brane proteins were subjected to immunoprecipitation with anti-PhePpolyclonal antibody. The immunoprecipitates were subjected to SDS–PAGE and biotinylated proteins were detected using avidin-HRP andchemiluminescence.

diYcult to distinguish between these two possibilities[103,128,129]. Use of detergents to compromise themembrane barrier will not be further reviewed becausesuch treatment has the high potential of exposing sitesburied within the protein structure or the membranebilayer.

Two eVective methods of cell disruption are osmoticshock of spheroplasts [96] and disintegration of cells bysonication. The rate of labeling of cysteine replacementsin the E. coli MotA protein (a component of the Xagella)by FM as a function of time was measured in intactspheroplasts and spheroplasts lysed by dilution intohypotonic medium. Reaction with the reagent in intactspheroplasts was very slow relative to the rapid reactionin disrupted spheroplasts conWrming a cytoplasm loca-tion for the cysteines [122]. To simultaneously label cys-teines exposed to both surfaces of the membrane,membrane ghosts were prepared directly from E. colicells expressing the formate transporter from Oxalob-acter formigenes by dilution of spheroplasts into hypo-tonic media containing OGM [111]. Periplasmically orcytoplasmically oriented cysteines in the E. coli osmo-protectant symporter ProP were also identiWed by a sim-ilar approach. Intact cells were treated with or withoutpre-blocking by membrane impermeable MTSET fol-lowed by reaction with membrane impermeable OGMbefore or after osmotic shock of spheroplasts [116].

Sonication is a rapid technique for cell disruption thatis not labor intensive and can be used to label water-exposed cysteines on both sides of the membranewithout regard for their normal disposition. LacY with asingle cysteine replacement in either the C6 domain(F205C), which in PE-containing cells faces the cyto-plasm, or TM segment VII (I230C), which residueswithin the hydrophobic core of the bilayer, wasexpressed in PE-containing and PE-lacking E. coli cells(Fig. 6). Samples of each cell type were split. One set wassonicated in the presence of MPEOB to simultaneouslylabel cysteines exposed to both sides of the membrane.Samples were incubated at room temperature foranother 4 min and then quenched by addition of 20 mM�ME. The other set was treated with same amount ofMPEOB for 5 min at room temperature, quenched with�ME, and washed free of reagents by two cycles of cen-trifugation and resuspension before being subjected tosonication as above. As shown in Fig. 6, LacY was bio-tinylated only during sonication of PE-containing cells(consistent with cytoplasmic location of C6) whileequal biotinylation occurred at C6 with or withoutsonication of PE-lacking cells (consistent with periplas-mic location). It is important to note that the extent ofbiotinylation was the same before and after sonicationof the PE-lacking sample indicating no alteration of theexposure of the cysteine occurred with sonication. Thismethod does not appear to expose cysteines withinTM segments to modiWcation as indicated by the lack of


labeling of a cysteine in TM segment VII, which shouldnot be accessible to MPEOB. The labeling of membranesduring sonication minimizes many problems for the detec-tion of cysteine in luminal extramembrane domains. Intra-cellular cysteines can be selectively labeled by Wrstblocking external cysteines followed by labeling duringsonication (substituting of sonication for permeabilizationin Fig. 2). This procedure does not require the preparationof oriented membrane vesicles or spheroplasts or the useof chemicals to compromise the membrane barrier.

8. Use of SCAMTM to analyze lipid-dependent topological organization

SCAMTM was used to demonstrate the change intopological organization of LacY when expressed in amutant of E. coli completely lacking PE [4]. A single cys-teine replacement (F208C) in domain C6, which is cyto-plasmic in PE-containing cells, was expressed in either aPE-containing (wild type) or PE-lacking (lipid mutant)strain of E. coli. Whole cells and ISOV from both celltypes were treated with MPB either after or without tolu-ene permeabilization and analyzed for biotinylation ofLacY. As shown in Fig. 7, the cysteine is only labeled ifwhole PE-containing cells are permeabilized, while inISOV made from PE-containing cells the cysteine is mod-iWed to the same extent with or without toluene treat-ment. These results are consistent with a cytoplasmiclocation for the cysteine in domain C6. The cysteine in C6probed in PE-lacking cells and ISOV displayed a mirrorimage labeling pattern when compared to the resultsfrom PE-containing cells consistent with a periplasmic

Fig. 6. Disintegration of the cell membrane by sonication. Cells ofAD93 either containing (+) or lacking (¡) PE were used to expressLacY carrying a single cysteine replacement in extramembranedomain C6 (cytoplasmic in PE-containing cells) or intramembraneTM segment VII. Cells were labeled with 100 �M MPEOB either priorto or during sonication using a Branson sonicator (tapered probe, setat 15% of maximum output). The labeling reaction was quenched withexcess �ME, and the whole cells disrupted by sonication as above. Themembrane fraction was isolated, solubilized with detergent, imm-umnoprecipitated with LacY speciWc antibody, subjected to SDS–PAGE, and biotinylated protein was detected using avidin-HRP andchemiluminescence.

location. Although single cysteine replacements poten-tially could aVect protein structure, the conclusions werebased on comparison of the same protein in two diVerentlipid environments strongly supporting a role for lipids astopological determinants. After the work up of the sam-ples, each was analyzed by Western blotting using anti-LacY antibody to normalize the signal intensity.

9. EVects of secondary structure and determining TM boundaries

Since amino acid side chains are usually packed indeWned structures, replacement with cysteines as speciWctargets may yield further insights regarding elements ofsecondary structure and conformational changes. Lackof labeling can be due to diVerences in local accessibilityrather than by diVerences in topology making SCAManalysis useful to explore the secondary structure inputative extramembrane or transmembrane segments.To conWrm local accessibility more than one residueshould be tested, and the protein should be assayed forreactivity of unreactive replacements after denaturation[128,129]. Thiol reagents that diVer in charge, size, andhydrophilicity can be also utilized for this purpose.

NEM has been shown to react periodically with cys-teine residues engineered into a TM segment and par-tially exposed to an aqueous environment. Cysteinesfacing the hydrophobic core of the bilayer were inacces-sible over a continuous sequence region. This approachhas been used to identify the face of a TM �-helix thatlines a channel or pore through the membrane based ona periodicity of every third to fourth position (one turnof the helix) being reactive [136]. Several single cysteinesin the predicted TM segment 6 of the Lactococcus lactisABC multidrug transporter LmrA could be labeled bythe relatively bulky FM molecule, and the 3-fold period-icity of FM accessibility strongly suggests that the regionspans the membrane as an �-helix with one face of the

Fig. 7. TM orientation of the C6 loop of LacY depends of the presenceof PE. A single cysteine replacement (208C) in the C6 loop of LacYwas expressed in PE-containing or PE-lacking E. coli. Intact cells andISOV were labeled with MPB either after or without permeabilizationwith toluene. After MPB treatment, LacY was immunoprecipitatedand resolved by SDS–PAGE, and biotinylated protein was detectedusing avidin-HRP and chemiluminescence (upper panels). An identicalblot (lower panel) was analyzed with anti-LacY antibody 4B11 todetermine the amount of LacY immunoprecipitated [4].


helix accessible to an aqueous cavity [103]. MPB-reactiveand non-reactive cysteine replacements showed a period-icity of every second residue from Val828 to Leu835,which is consistent with a �-sheet conformation for oneof the TM segments of the human plasma membraneanion exchanger AE1 [98]. In the same study Phe806 toLys814 displayed a 3-fold periodicity that is suggestiveof an �-helix. Local secondary structure for membraneproteins has also been inferred from the periodicity ofcysteine modiWcation by PCMBS [91].

The membrane-aqueous boundaries are generallyinferred indirectly from the transition between accessibleand inaccessible cysteine residues using a large hydro-philic thiol reagent and SCAM over the putative transi-tion from aqueous to hydrocarbon environment. Bulky,hydrophilic, membrane impermeable thiol reagents suchas FM [103,105,122], OGM [96,111,116], or MTSEA-biotin [100,125,126] have been used to map the extracel-lular boundary for TM segments. As indicated in Section3.3, the thiol of cysteine extends 8–10 Å from the peptidebackbone so that these reagents have access, withdecreasing eYciency, to Wve to six residues into thehydrophobic core of the bilayer. MPB at low concentra-tions and under mild reaction conditions was also usedto identify indirectly the ends of TM segments[4,94,97,99,100,102,108–110,125].

10. Use of SCAMTM to monitor topological dynamics

Once inserted TM segments should not oscillate read-ily back and forth across the membrane due to the largefree energy barrier to transfer hydrophobic sequencesout of, and polar sequences back through the bilayer.Most membrane protein topology studies are consistentwith a static and permanent location of extramembranedomains facing either one side or the other of the mem-brane. However, the actual structure of a membrane isvery dynamic. Therefore, is the lipid bilayer really a non-Xipping zone for integral membrane proteins?

The ability to change lipid composition in a regulatedmanner (Section 1.2) coupled with monitoring LacYtopology as a function of changes in membrane phospho-lipid composition has shown that TM organization ofmembrane proteins is potentially more dynamic than pre-viously assumed and responsive to changes in lipid envi-ronment after insertion and assembly. Placing the pssAgene under control of the regulatable araB promoterresults in repressed gene expression in the presence of glu-cose with PE levels of less than 2%. Full gene expression inthe presence of arabinose results in wild type levels of PE.LacY with a single cysteine replacement in either domainC6 (cytoplasmic in PE-containing cells) or in domain P7(periplasmic irrespective of phospholipid composition)was expressed with isopropyl-1-thio-D-galactopyranoside(+IPTG) in the presence of glucose resulting in less than

2% PE in the membranes. Under these conditions (Fig. 8,¡PE, Glu) both C6 and P7 in whole cells were biotinyla-ted by MPB either with (+) or without (¡) toluene treat-ment consistent with exposure of these domains to theperiplasm [4]. New LacY synthesis was stopped (¡IPTG)and glucose was replaced by arabinose to induce synthesisof PE for 60min in the absence of new LacY synthesis.MPB still labeled P7 in whole cells (+PE, Ara) with andwithout toluene treatment to the same extent as observedin the presence of glucose consistent with its periplasmiclocation in PE-containing cells. Remarkably, MPB onlylabeled C6 after cells were permeabilized by toluene con-sistent with the C6 domain returning to the cytoplasmwith the induction of PE synthesis. A similar restorationof topological organization for the mis-assembled N-ter-minal helical hairpin (TM segments I and II) of PheP wasobserved when PE synthesis was turned on after initialinsertion of PheP into membranes lacking PE [5]. Forboth transporters topology and active transport functionwere restored by exposure to PE after membrane insertionand assembly in the absence of PE.

These results challenge the dogma that once topologi-cal organization of a membrane protein is established,TM segments are stably maintained and do not Xip. Thisresult is consistent with the high degree of Xexibility ofLacY and other membrane proteins and suggests thatthe orientation of individual TM segments may not beWxed during assembly, after native structure has beenattained, or during catalytic cycles. The lipid inducedconformational and topological changes observed forLacY and PheP do not occur in wild type cells, but maybe a consequence of their functional properties or reXectintermediates present during folding of these proteins.

Fig. 8. Post-insertional topological reorganization detected bySCAMTM. LacY with a cysteine replacement in either the cytoplasmicdomain C6 or periplasmic domain P7 was expressed in an E. coli strainin which PE synthesis can be repressed by glucose (Glu) and inducedby arabinose (Ara) [4]. LacY expression was under IPTG control.Accessibility of the single cysteine replacements was analyzed by treat-ment of whole cells with MPB either after or without toluene treat-ment. Cells were Wrst grown under conditions of induction of LacYexpression but repression of PE synthesis (+IPTG and ¡PE, Glu).After removing IPTG (+/¡) to stop new synthesis of LacY, PE synthe-sis (+PE, Ara) was induced for 60 min. After reaction with MBP, bio-tinylated LacY was detected as in Fig. 7. This Wgure was reproducedwith permission from [4]. (Copyright 2002 EMBO) where experimen-tal details can be found.


SCAMTM has also been used to monitor topologicalchanges induced by substrate binding and release duringenzyme turnover [51,125]. In order to identify cysteine res-idues located in the extracellular domains of the �-subunitof the Na+/K+-ATPase and to determine whether theseresidues change orientation during the pump cycle, thesodium pump in RSOV was stabilized in two diVerentconformations (the phosphorylated form or cation-occluded form). Topological changes in cysteines exposedto an aqueous or hydrophobic environment were assessedusing AMS or CPM, respectively. Removal of cation frommembrane preparations was accompanied by release(reaction with AMS) from the membrane of the hairpinformed by TM segment 5 and TM segment 6. Under theseconditions some target residues within the C-terminal TMsegment 7 to TM segment 10 fragment became exposed tolipophilic CPM labeling while others relocated outside ofthe membrane and became accessible to membrane imper-meable AMS. Such reorganization occurs only in the C-terminal half of protein, since other changes in AMS orCPM accessibility were not observed.

Plasticity in the membrane location of the TM seg-ment 8 of the Ca2+-ATPase may be the cause of theuncertainty in identifying the organization of this TMsegment. At least two possible arrangements for TM seg-ment 8 were suggested [137]. Competitive bindingbetween thiol speciWc reagents and substrate revealed amobile element within the TM structure of glutamatetransporter GLT-1 [97]. The reactivity of a single cys-teine placed in the loops Xanking TM segment III of theE. coli tetracycline/H+ antiporter was drasticallychanged upon binding tetracycline, indicating that cyto-plasmic and periplasmic loop regions undergo substrate-induced topological changes in opposite directions [52].

11. Topology of �-barrel proteins

SCAMTM can also be used to map the topology ofmembrane proteins that adopt a �-barrel architecturecomposed of anti-parallel TM �-strands. Site-directedXuorescence labeling by OGM was used to assign thetopology of E. coli outer membrane protein PapC thatappears to consist of 32 TM �-strands [121]. OGM read-ily crossed the outer membrane and labeled cysteine resi-dues exposed to both the cell exterior and theperiplasmic space. To diVerentiate cell exterior fromperiplasmic cysteines, Alexa Fluor 594 C5 was used as ablocking reagent. This reagent has a molecular weight of909 Da that exceeds the 600 Da limit of outer membraneOmpF porin thus restricting the reagent access to theperiplasmic space. By varying reaction conditions themodiWcation of cysteines exposed to the cell exterior rel-ative to periplasmic cysteines was reduced 5-fold by theblocking reagent. The single cysteine replacementswithin PapC that were not labeled with OGM in intact

cells were subsequently labeled by OGM after proteinsolubilization with dodecyl-maltoside or denaturationwith SDS suggesting they faced the hydrophobic core ofthe bilayer or were sterically hindered.

12. Use of SCAMTM to assess topology of proteins reconstituted into liposomes

Topological analysis of membrane proteins reconsti-tuted into proteoliposomes is subject to the same con-straints as in whole cells and membrane vesicles. What isthe degree of lipid bilayer permeability to the reagentsand is the protein uniformly oriented in the liposome?The topological organization of LacY, UhpT, and equ-inatoxin II has been successfully determined using MPBafter reconstitution of the proteins with a uniform topol-ogy in proteoliposomes [43,109,138]. The topological ori-entation of TM segments of LacY was found to bedependent on the lipid composition of the proteolipo-somes. Cytoplasmic domain C6 and periplasmic domainP7 were on opposite sides of liposomes containing PE andon the same side in liposomes lacking PE, which is consis-tent with their relative orientation in whole cells [4,43].

In general liposomes are more permeable to the com-monly employed thiol reagents and permeability mayvary with lipid composition. To assess whether thesereagents are indeed membrane impermeable, proteolipo-somes can be encapsulated with a thiol reporter com-pound and then probed with the thiol-speciWc reagentfrom the outside. Thionitrobenzoate (TNB) is an eVec-tive reporter compound since its absorbance at 412 nm issigniWcantly reduced when the thiol group is blocked[139]. Moreover, appropriate thiol scavengers can elimi-nate undesired trans modiWcation. ModiWcation of lumi-nal cysteines within a target protein due to reagentpermeability can be signiWcantly reduced by inclusion ofcharged thiol scavengers inside liposomes. Cysteine is aneVective thiol scavenger, which is strictly membraneimpermeable due to its zwitterionic nature.

Uniform orientation of proteins in the liposome bilayercan usually be concluded from the labeling results. Byusing cysteine replacements in extramembrane domainsthat exhibit opposite orientations in vivo, both accessibil-ity and inaccessibility should be observed by membraneimpermeable reagents in a pattern consistent with in vivoorientation. The inaccessible cysteines should be modiWedafter disruption of the proteoliposomes [43].

13. SCAMTM limitations

As is true of most approaches to determine mem-brane protein topological organization, there arelimitations to using SCAMTM. Caution must be used inassigning an intramembrane location to a cysteine


residue because it is unreactive to hydrophilic thiolreagents in both intact and permeabilized membranes.Lack of or low levels of labeling may result from any ofthe following reasons: (1) steric hindrance due to localsecondary structure; (2) internalization into the compactfold of the protein; (3) lack of ionization of the thiolgroup due to a hydrophobic environment; (4) local envi-ronment with the same charge as the thiol reagent; (5)increased pKa of the thiol due to the high negativecharge density of neighboring residues or anionic lipids.For this reason several membrane-impermeable reagentsshould be tested in order to Wnd the most appropriateone, including ones that are smaller and with diVerentcharge properties.

The secondary structure of extramembrane loops andlocal steric hindrance will inXuence whether or not thecysteine residue is accessible. Periplasmic extramembranedomains tend to be shorter (sometimes only three aminoacids in length) than cytoplasmic domains. Therefore,there may be little or no protrusion of these loops into theextracellular space, thus preventing reaction of the cys-teine residues in these locations with bulky maleimides.Biotinylated reagents appear to react better with cysteinestowards the middle of extended hydrophilic loops thannear the TM segment interfacial domain [4,94,112]. Thelower reactivity of some of the residues in hydrophilicloops may reXect steric hindrance if clearly its neighborsreact quantitatively. Due to the possibility of misidentify-ing residues as intramembrane because they are buried inan extracellular location, it is essential to analyze morethan one position in each region of interest. All of theseproblems can usually be addressed by cysteine scanningreplacement over a putative extramembrane domain thatwill show periodicity of accessibility, a short stretchusually more accessible by smaller thiol reagents, or a gra-dient of accessibility as the TM interface is approached.An extra membrane domain and an aqueous pore are noteasily distinguished by SCAMTM. Substitution at posi-tions crucial to overall protein structure and stability can-not be used, but such substitutions often result in lowlevels of the protein and are informative in themselves.

Local electrostatic and ionic conditions may aVect thepKa of an extramembrane position substituted by cys-teine [140]. A local concentration of negatively chargedamino acids or proximity to an anionic phospholipiddomain would increase the pKa of the thiol due to thehigh negative charge density [130,141]. Negativelycharged side chains proximal to a cysteine inXuence cys-teinyl thiol pKa’s primarily via solvent-mediated cou-lombic interactions [142]. If negatively charged thiolreagents are employed, charge repulsion would furtherreduce the rate of modiWcation of the thiol. Reducedyield due to the reagent properties can be addressed byusing reagents with diVerent chemical properties.

TM topology studies assume that all copies of the tar-get protein have the same orientation. The labeling pat-

terns are the result of end-point titrations and assume arelatively Wxed conformation for extramembrane loops,ignoring the presence of regions with heightened mobil-ity and Xexibility or the possibility of topological inver-sions on the time scale of the labeling. Topologicalmodels derived from accessibility patterns depict a staticTM topology whereas the actual structure in a mem-brane is more likely to be dynamic. Through the applica-tion of SCAMTM some examples exist of dynamicchanges in topological organization induced post-assem-bly of membrane proteins as well as some proteins thatappear to exist with multiple topological organizations.Therefore, the static nature of methods that measuretopological organization may have missed moredynamic properties of membrane proteins. In particularlow yield of modiWcation due to slow rate of reactionmay be due to dynamic movement of a domain into andout of an accessible region. As discussed below there areexamples of cryptic intramembrane regions that becomeexposed to the aqueous phase and extramembranedomains that translocate to the opposite side of themembrane. Therefore, labeling studies in isolated mem-branes should complement topology studies in intactcells where the protein is turning over or metabolic con-ditions inXuence organization.

Even with optimal assays and reagents, membraneproteins that assume multiple conformations eitherwithin the same or between diVerent membranes mayyield confusing and conXicting results. Prenesilin 1 existsin distinctly diVerent conformational states, one witheight TM topology retained within cells and the othertransported to the cell surface in a seven TM organiza-tion. This would explain the contradictory observationsconcerning topological organization made by severalgroups [143]. Several examples of mixed topology ordiVerent topologies for the same protein in diVerentmembranes were noted above. It is not clear whetherthese topological diVerences occurred during or postassembly of these proteins. However, using SCAMTM

topological reorganization was demonstrated for bothLacY [4] and PheP [5] of E. coli post assembly of theproteins. Such results would suggest that diVerences inlipid environment could be a contributing factor todiVerences in organization of the same protein in diVer-ent membranes of eukaryotic cells.

14. SCAMTM advantages

This approach is based on introduction of cysteineresidues one at a time into putative extracellular or intra-cellular loops of a cysteineless membrane protein ofinterest followed by chemical modiWcation with mem-brane impermeable thiol-speciWc probes either before orafter compromising membrane integrity to determinecysteine sidedness. Together, these tools document


residue accessibility, and therefore topology of a mem-brane protein. Cysteine-scanning mutagenesis is themost useful technique thus far developed for topologystudies and has the following advantages: (1) analysiscan be done on the complete protein; (2) cysteinereplacements are well tolerated with retention of topol-ogy and function; (3) analysis is done by chemical modi-Wcation by a broad range of commercially availablereagents with diVerent physical–chemical properties; (4)detection of engineered cysteine modiWcations is simple;(5) only small quantities of the protein are required; (6)no complicated protein puriWcation procedures arerequired; (7) chemical modiWcation can be done usingintact cells, thereby avoiding problems related to theconversion of cells into membrane vesicles with a uni-form orientation.

SCAMTM is not only an alternative approach to deter-mination of membrane protein structure, but also consti-tutes an attractive independent approach to structuraland dynamic studies of membrane proteins. A quantita-tive analysis of the surface accessibility of individualcysteines at various stages of assembly of S. aureus�-hemolysin was carried by cysteine scanning mutagene-sis and targeted chemical modiWcation in order to mapthe structural changes which this polypeptide undergoesduring pore formation [144]. The placement of cysteineresidues in putative extracellular or intracellular domainsof polytopic membrane proteins has proven very usefulin deWning the folding pattern and TM conformationalchanges of the polypeptide in the membrane. Labeling ofsingle cysteine replacements of a major component(SecG) of the translocon with AMS either at rest or dur-ing protein translocation clearly demonstrated that acytoplasmic region of SecG undergoes a change in mem-brane sidedness upon preprotein translocation, indicatingthat SecG undergoes topology inversion [132].

Although the labeling patterns derived fromSCAMTM assay usually reXect end-point topology of amembrane protein, semi-quantitative analysis of the sur-face accessibility of individual cysteines introduced intoextramembrane loops can be carried out at variousstages of protein assembly. Cysteine accessibility duringbacteriorhodopsin translation was monitored by pulse-chase radiolabeling and modiWcation by AMS to deter-mine the order and timing of insertion of TM segmentsinto the membrane of Halobacterium salinarum [145]. Inthis in vivo translocation assay, insertion of TM seg-ments into the H. salinarum cytoplasmic membrane wasmonitored by rapid modiWcation of unique cysteines inextracellular domains of the protein with AMS resultingin a shift in mobility of the protein in SDS–PAGE.SCAMTM also provided topological information at thetime orientation was established during the biosynthesisof the protein and its insertion into the membrane.

Although X-ray crystallography produces highlydetailed structural information of membrane proteins

and even lipids resolved at high resolution within thestructure, the structures only provide a static picture oflipid–protein interactions. The recent determination of ahigh resolution structure for LacY revealed a complexprotein with TM segments of varied length and tilt angle[93]. However, the exact boundaries between TM seg-ments and loop domains remain largely unknown.Where lipids have been resolved in crystal structures,considerable hydrogen bonding between polar residuesat the ends of the TM helices and lipid polar headgroups has been observed [146], indicating that thesehelices extend beyond the hydrophobic core of thebilayer. For most of the highly resolved membrane pro-teins, hydrophobic thicknesses of TM segments do notseem to match the lipid bilayer thickness expected orexperimentally determined from the chain length of thesurrounding lipids. SigniWcant mismatch betweenthe hydrophobic thicknesses of a membrane protein andthe lipid bilayer implies either that the structure of theprotein when crystallized from detergent is diVerentfrom that in the native membrane or that the lipidbilayer is distorted around the protein in the membrane.SCAMTM can be used to deWne the boundaries betweenmembrane-embedded regions and the loop regionsexposed to the aqueous phase of membrane proteins intheir native environment thus supplementing high-reso-lution structural information. The precise boundaries ofTM segments of the tetracycline/H+ antiporter TetA[113] and the human erythrocyte anion exchanger AE1[112] were established by the reactivity of cysteinereplacements with NEM and MPB, respectively.

15. Concluding remarks

Due to the lack of structural information on veryhydrophobic membrane proteins determined by X-raydiVraction or NMR, other lower resolution methodshave been employed to understand topologicalarrangement of proteins in the membrane. Severalmethods for determining TM organization are avail-able and have been extensively reviewed. However,SCAMTM, which has not been reviewed, has emergedover the past few years as the method of choice due toits relative simplicity for determining the TM structureof membrane proteins. SCAMTM has the advantagethat structural perturbation of introduced cysteinemutations is milder than in other methods commonlyused to determine topology. Although most examplesof application have been from bacterial systems, sev-eral examples were outlined for TM mapping ofeukaryotic proteins in their native environment,indicating that the method is generally applicable todiVerent membrane systems.

This method can be used to assess secondary struc-ture, membrane topology, and TM conformational


changes as well as provide topological information dur-ing membrane insertion, folding, and assembly of pro-teins. This approach is capable of distinguishingaccessibility of residues separated by only three or fourresidues and is useful in precise mapping of the ends ofTM segments to deWne membrane aqueous boundariesof membrane proteins. By applying SCAM along extra-membrane or TM segments, information of secondarystructure and patterns of hydrophobic and aqueousexposure of domains can be obtained. Topological andother structural changes can be monitored during thecatalytic cycle of enzymes.

By combining SCAMTM with mutants of E. coli inwhich membrane phospholipid composition can be sys-tematically controlled, the role of phospholipids asdeterminants of membrane protein topological organi-zation was established. In addition, the potential forpolytopic membrane proteins to change their topologi-cal organization after insertion and assembly in themembrane was demonstrated. Combining of these tech-niques provides a system to study the role of lipid–pro-tein interactions in the structure, assembly, and functionof membrane proteins. The ability to regulate lipid com-position temporally provides a powerful means to inves-tigate molecular details of the dynamic topogenesisprocess.

A new algorithm (TMDET) has been developed topredict membrane embedded protein sequences by usingtheir atomic coordinates. The method based on the fre-quency with which given amino acids residues are foundin highly resolved protein transmembrane segments.This algorithm can be accessed at http://tmdet.enzim.hu/.

Acknowledgments

The work from the authors’ laboratory was sup-ported by National Institutes of General Medical Sci-ence Grant GM20478 and the John S. Dunn, Sr.Research Foundation awarded to W.D. Phil Heacockassisted in drawing the Wgures.

References

[1] S.H. White, G. von Heijne, Curr. Opin. Struct. Biol. 14 (2004)397–404.

[2] M. Higy, T. Junne, M. Spiess, Biochemistry 43 (2004) 12716–12722.

[3] D. Levy, Essays Biochem. 31 (1996) 49–60.[4] M. Bogdanov, P.N. Heacock, W. Dowhan, EMBO J. 21 (2002)

2107–2116.[5] W. Zhang, M. Bogdanov, J. Pi, A.J. Pittard, W. Dowhan, J. Biol.

Chem. 278 (2003) 50128–50135.[6] Y. Lu, I.R. Turnbull, A. Bragin, K. Carveth, A.S. Verkman, W.R.

Skach, Mol. Biol. Cell 11 (2000) 2973–2985.[7] W. Dowhan, Biochim. Biophys. Acta 1376 (1998) 455–466.

[8] W. Dowhan, E. Mileykovskaya, M. Bogdanov, Biochim. Bio-phys. Acta 1666 (2004) 19–39.

[9] B.E. Tropp, L. Ragolia, W. Xia, W. Dowhan, R. Milkman, K.E.Rudd, R. Ivanisevic, D.J. Savic, J. Bacteriol. 177 (1995) 5155–5157.

[10] S. Kikuchi, I. Shibuya, K. Matsumoto, J. Bacteriol. 182 (2000)371–376.

[11] A. DeChavigny, P.N. Heacock, W. Dowhan, J. Biol. Chem. 266(1991) 5323–5332.

[12] P.N. Heacock, W. Dowhan, J. Biol. Chem. 264 (1989) 14972–14977.

[13] C. Sohlenkamp, K.E. de Rudder, V. Rohrs, I.M. Lopez-Lara, O.Geiger, J. Biol. Chem. 275 (2000) 18919–18925.

[14] W. Xia, W. Dowhan, J. Bacteriol. 177 (1995) 2926–2928.[15] M. Wikström, J. Xie, M. Bogdanov, E. Mileykovskaya, P. Hea-

cock, Å. Wieslander, W. Dowhan, J. Biol. Chem. 279 (2004)10284–10292.

[16] N.I. Mikhaleva, C.L. Santini, G. Giordano, M.A. Nesmeyanova,L.F. Wu, FEBS Lett. 463 (1999) 331–335.

[17] T. de Vrije, R.L. de Swart, W. Dowhan, J. Tommassen, B. de Kru-ijV, Nature 334 (1988) 173–175.

[18] R. Kusters, W. Dowhan, B. de KruijV, J. Biol. Chem. 266 (1991)8659–8662.

[19] R. Kusters, E. Breukink, A. Gallusser, A. Kuhn, B. de KruijV, J.Biol. Chem. 269 (1994) 1560–1563.

[20] R.J. Cabelli, K.M. Dolan, L.P. Qian, D.B. Oliver, J. Biol. Chem.266 (1991) 24420–24427.

[21] D.A. Phoenix, R. Kusters, C. Hikita, S. Mizushima, B. de KruijV,J. Biol. Chem. 268 (1993) 17069–17073.

[22] N.D. Ulbrandt, E. London, D.B. Oliver, FASEB J. 6 (1992)ABS 481.

[23] W. van Klompenburg, I. Nilsson, G. von Heijne, B. de KruijV,EMBO J. 16 (1997) 4261–4266.

[24] W. van Klompenburg, A.N. Ridder, A.L. van Raalte, A.J. Killian,G. von Heijne, B. de KruijV, FEBS Lett. 413 (1997) 109–114.

[25] E. Crooke, Biochimie 83 (2001) 19–23.[26] W. Zheng, Z. Li, K. Skarstad, E. Crooke, EMBO J. 20 (2001)

1164–1172.[27] W. Xia, W. Dowhan, Proc. Natl. Acad. Sci. USA 92 (1995) 783–

787.[28] M. Aboulwafa, R. Hvorup, M.H. Saier Jr., Arch. Microbiol. 181

(2004) 26–34.[29] M. Aboulwafa, M.H. Saier Jr., Res. Microbiol. 153 (2002) 667–677.[30] A.G. Rietveld, A.J. Verkleij, B. de KruijV, Biochim. Biophys. Acta

1324 (1997) 263–272.[31] P.R. Cullis, B. de KruijV, Biochim. Biophys. Acta 559 (1979) 399–

420.[32] A.G. Rietveld, J.A. Killian, W. Dowhan, B. de KruijV, J. Biol.

Chem. 268 (1993) 12427–12433.[33] A.G. Rietveld, V.V. Chupin, M.C. Koorengevel, H.L. Wienk, W.

Dowhan, B. de KruijV, J. Biol. Chem. 269 (1994) 28670–28675.[34] A.G. Rietveld, M.C. Koorengevel, B. de KruijV, EMBO J. 14

(1995) 5506–5513.[35] E.I. Mileykovskaya, W. Dowhan, J. Biol. Chem. 268 (1993)

24824–24831.[36] E. Mileykovskaya, Q. Sun, W. Margolin, W. Dowhan, J. Bacte-

riol. 180 (1998) 4252–4257.[37] E. Mileykovskaya, I. Fishov, X. Fu, B.D. Corbin, W. Margolin,

W. Dowhan, J. Biol. Chem. 278 (2003) 22193–22198.[38] E. Mileykovskaya, W. Dowhan, J. Bacteriol. 182 (2000) 1172–

1175.[39] E. Mileykovskaya, W. Dowhan, J. Bacteriol. 179 (1997) 1029–

1034.[40] M. Bogdanov, W. Dowhan, J. Biol. Chem. 274 (1999) 36827–

36830.[41] M. Bogdanov, J. Sun, H.R. Kaback, W. Dowhan, J. Biol. Chem.

271 (1996) 11615–11618.

http://tmdet.enzim.hu/

http://tmdet.enzim.hu/


[42] M. Bogdanov, M. Umeda, W. Dowhan, J. Biol. Chem. 274 (1999)12339–12345.

[43] X. Wang, M. Bogdanov, W. Dowhan, EMBO J. 21 (2002) 5673–5681.

[44] W. Dowhan, M. Bogdanov, in: D.E. Vance, J.E. Vance (Eds.),Biochemistry of Lipids, Lipoproteins and Membranes, Elsevier,Amsterdam, 2002, pp. 1–35.

[45] M. Zhang, E. Mileykovskaya, W. Dowhan, J. Biol. Chem. 277(2002) 43553–43556.

[46] D.B. Ostrander, M. Zhang, E. Mileykovskaya, M. Rho, W. Dow-han, J. Biol. Chem. 276 (2001) 25262–25272.

[47] K. PfeiVer, V. Gohil, R.A. Stuart, C. Hunte, U. Brandt, M.L.Greenberg, H. Schagger, J. Biol. Chem. 278 (2003) 52873–52880.

[48] S.C. Chang, P.N. Heacock, C.J. Clancey, W. Dowhan, J. Biol.Chem. 273 (1998) 9829–9836.

[49] D.E. Vance, in: D.E. Vance, J.E. Vance (Eds.), Biochemistry ofLipids, Lipoproteins and Membranes, Elsevier, Amsterdam,2002, pp. 205–231.

[50] V. Goder, M. Spiess, EMBO J. 22 (2003) 3645–3653.[51] S. Lutsenko, S. Daoud, J.H. Kaplan, J. Biol. Chem. 272 (1997)

5249–5255.[52] T. Kimura, Y. Shiina, T. Sawai, A. Yamaguchi, J. Biol. Chem. 273

(1998) 5243–5247.[53] C. Lambert, R. Prange, J. Biol. Chem. 276 (2001) 22265–22272.[54] C. Lambert, R. Prange, Proc. Natl. Acad. Sci. USA 100 (2003)

5199–5204.[55] K. Nishiyama, T. Suzuki, H. Tokuda, Cell 85 (1996) 71–81.[56] K. GouY, F. Gerard, C.L. Santini, L.F. Wu, J. Biol. Chem. 279

(2004) 11608–11615.[57] C. Rutz, W. Rosenthal, R. Schulien, J. Biol. Chem. 274 (1999)

33757–33763.[58] J. Dunlop, P.C. Jones, M.E. Finbow, EMBO J. 14 (1995) 3609–

3616.[59] Q.-S. Zhu, P. von Dippe, W. Xing, D. Levy, J. Biol. Chem. 274

(1999) 27898–27904.[60] S.J. Kim, R.S. Hegde, Mol. Biol. Cell 13 (2002) 3775–3786.[61] R.S. Hegde, P. Tremblay, D. Groth, S.J. DeArmond, S.B. Prus-

iner, V.R. Lingappa, Nature 402 (1999) 822–826.[62] J. Huggett, A. Vaughan-Thomas, D. Mason, FEBS Lett. 485

(2000) 13–18.[63] P.K. Kim, M.G. Annis, P.J. Dlugosz, B. Leber, D.W. Andrews,

Mol. Cell 14 (2004) 523–529.[64] K.J. Linton, C.F. Higgins, Mol. Membr. Biol. 19 (2002) 51–58.[65] M. van Geest, J.S. Lolkema, J. Biol. Chem. 274 (1999) 29705–

29711.[66] J. Nilsson, B. Persson, G. von Heijne, FEBS Lett. 486 (2000) 267–

269.[67] M. Arai, H. Mitsuke, M. Ikeda, J.X. Xia, T. Kikuchi, M. Satake,

T. Shimizu, Nucleic Acids Res. 32 (2004) W390–W393.[68] B. Persson, P. Argos, J. Protein Chem. 16 (1997) 453–457.[69] S. Jayasinghe, K. Hristova, S.H. White, J. Mol. Biol. 312 (2001)

927–934.[70] S. Jayasinghe, K. Hristova, S.H. White, Protein Sci. 10 (2001)

455–458.[71] M. Ikeda, M. Arai, T. Okuno, T. Shimizu, Nucleic Acids Res. 31

(2003) 406–409.[72] C.M. Ott, V.R. Lingappa, J. Cell Sci. 115 (2002) 2003–2009.[73] B. Traxler, D. Boyd, J. Beckwith, J. Membr. Biol. 132 (1993) 1–11.[74] M. van Geest, J.S. Lolkema, Microbiol. Mol. Biol. Rev. 64 (2000)

13–33.[75] M.B. Shimon, R. Goldshleger, S.J. Karlish, J. Biol. Chem. 273

(1998) 34190–34195.[76] T. Toyoda, K. Masunaga, Y. Ohtsu, K. Hara, N. Hamada, T.

Kashiwagi, J. Iwahashi, Curr. Protein Pept. Sci. 1 (2000) 303–308.[77] C. Manoil, D. Boyd, J. Beckwith, Trends Genet. 4 (1988) 223–226.[78] J.K. Broome-Smith, M. Tadayyon, Y. Zhang, Mol. Microbiol. 4

(1990) 1637–1644.

[79] A. Zelazny, E. Bibi, Biochemistry 35 (1996) 10872–10878.[80] M. Ehrmann, D. Boyd, J. Beckwith, Proc. Natl. Acad. Sci. USA

87 (1990) 7574–7578.[81] M.F. Alexeyev, H.H. Winkler, J. Mol. Biol. 285 (1999) 1503–

1513.[82] X.Y. Liu, L.H. Matherly, Biochim. Biophys. Acta 31 (2002) 333–

342.[83] R.A. Schwalbe, A. Rudin, S.L. Xia, C.S. Wingo, J. Biol. Chem.

277 (2002) 24382–24389.[84] L. Olivares, C. Aragon, C. Gimenez, F. Zafra, J. Biol. Chem. 272

(1997) 1211–1217.[85] I.M. Nilsson, G. von Heijne, J. Biol. Chem. 268 (1993) 5798–5801.[86] M. Popov, L.Y. Tam, J. Li, R.A. Reithmeier, J. Biol. Chem. 272

(1997) 18325–18332.[87] M.H. Akabas, D.A. StauVer, M. Xu, A. Karlin, Science 258

(1992) 307–310.[88] J.J. Falke, A.F. Dernburg, D.A. Sternberg, N. Zalkin, D.L. Milli-

gan, D.E. Koshland Jr., J. Biol. Chem. 263 (1988) 14850–14858.[89] A. Karlin, M.H. Akabas, Methods Enzymol. 293 (1998) 123–145.[90] A. Karlin, J. Gen. Physiol. 117 (2001) 235–238.[91] R.T. Yan, P.C. Maloney, Cell 75 (1993) 37–44.[92] S. Frillingos, M. Sahin-Toth, J. Wu, H.R. Kaback, FASEB J. 12

(1998) 1281–1299.[93] J. Abramson, I. Smirnova, V. Kasho, G. Verner, H.R. Kaback, S.

Iwata, Science 301 (2003) 610–615.[94] T.W. Loo, D.M. Clarke, J. Biol. Chem. 270 (1995) 843–848.[95] H. Jung, R. Rubenhagen, S. Tebbe, K. Leifker, N. Tholema, M.

Quick, R. Schmid, J. Biol. Chem. 273 (1998) 26400–26407.[96] L. Ye, Z. Jia, T. Jung, P.C. Maloney, J. Bacteriol. 183 (2001)

2490–2496.[97] M. Grunewald, D. Menaker, B.I. Kanner, J. Biol. Chem. 277

(2002) 26074–26080.[98] Q. Zhu, D.W. Lee, J.R. Casey, J. Biol. Chem. 278 (2003) 3112–

3120.[99] W. Cao, L.H. Matherly, Biochem. J. 378 (2004) 201–206.

[100] Y. Sato, Y.W. Zhang, A. Androutsellis-Theotokis, G. Rudnick, J.Biol. Chem. 279 (2004) 22926–22933.

[101] Y. Olami, A. Rimon, Y. Gerchman, A. Rothman, E. Padan, J.Biol. Chem. 272 (1997) 1761–1768.

[102] T. Wada, J.C. Long, D. Zhang, S.B. Vik, J. Biol. Chem. 274 (1999)17353–17357.

[103] G.J. Poelarends, W.N. Konings, J. Biol. Chem. 277 (2002) 42891–42898.

[104] F.I. Valiyaveetil, R.H. Fillingame, J. Biol. Chem. 273 (1998)16241–16247.

[105] J. Meuller, J. Rydstrom, J. Biol. Chem. 274 (1999) 19072–19080.[106] E. Joseph-Liauzun, P. Delmas, D. Shire, P. Ferrara, J. Biol. Chem.

273 (1998) 2146–2152.[107] X.H. Leng, T. Nishi, M. Forgac, J. Biol. Chem. 274 (1999) 14655–

14661.[108] J. Knol, L. VeenhoV, W.J. Liang, P.J. Henderson, G. Leblanc, B.

Poolman, J. Biol. Chem. 271 (1996) 15358–15366.[109] M.C. Fann, P.C. Maloney, J. Biol. Chem. 273 (1998) 33735–

33740.[110] M. Matos, M.C. Fann, R.T. Yan, P.C. Maloney, J. Biol. Chem.

271 (1996) 18571–18575.[111] D. Fu, P.C. Maloney, J. Biol. Chem. 273 (1998) 17962–17967.[112] J. Fujinaga, X.B. Tang, J.R. Casey, J. Biol. Chem. 274 (1999)

6626–6633.[113] T. Kimura, M. Ohnuma, T. Sawai, A. Yamaguchi, J. Biol. Chem.

272 (1997) 580–585.[114] R.T. Yan, P.C. Maloney, Proc. Natl. Acad. Sci. USA 92 (1995)

5973–5976.[115] J.D. Gregory, J. Am. Chem. Soc. (1955) 3922–3923.[116] D.E. Culham, A. Hillar, J. Henderson, A. Ly, Y.I. Vernikovska,

K.I. Racher, J.M. Boggs, J.M. Wood, Biochemistry 42 (2003)11815–11823.


[117] T.E. Creighton, Proteins: Structures and Molecular Properties,W.H. Freeman and Co, New York, 1983.

[118] T.S. Kumarevel, M.M. Gromiha, M.N. Ponnuswamy, Prep. Bio-chem. Biotechnol. 31 (2001) 163–183.

[119] P.D. Taylor, T.K. Attwood, D.R. Flower, Nucleic Acids Res. 31(2003) 3698–3700.

[120] S.H. White, A.S. Ladokhin, S. Jayasinghe, K. Hristova, J. Biol.Chem. 276 (2001) 32395–32398.

[121] N.S. Henderson, S. Shu Kin So, C. Martin, R. Kulkarni, D.G.Thanassi, J. Biol. Chem. 279 (2004) 53747–53754.

[122] J. Zhou, R.T. Fazzio, D.F. Blair, J. Mol. Biol. 251 (1995) 237–242.[123] P. Jockel, M. Di Berardino, P. Dimroth, Biochemistry 38 (1999)

13461–13472.[124] E. Fujihira, N. Tamura, A. Yamaguchi, J. Biochem. (Tokyo) 131

(2002) 145–151.[125] Y.K. Hu, J.H. Kaplan, J. Biol. Chem. 275 (2000) 19185–19191.[126] W. Cao, L.H. Matherly, Biochem. J. 374 (2003) 27–36.[127] J.C. Long, S. Wang, S.B. Vik, J. Biol. Chem. 273 (1998) 16235–

16240.[128] J.C. Long, J. DeLeon-Rangel, S.B. Vik, J. Biol. Chem. 277 (2002)

27288–27293.[129] S.S. Mordoch, D. Granot, M. Lebendiker, S. Schuldiner, J. Biol.

Chem. 274 (1999) 19480–19486.[130] J.E. Jewell, J. Orwick, J. Liu, K.W. Miller, J. Bacteriol. 181 (1999)

1689–1693.[131] J.G. Chen, S. Liu-Chen, G. Rudnick, J. Biol. Chem. 273 (1998)

12675–12681.[132] S. Nagamori, K. Nishiyama, H. Tokuda, J. Biochem. (Tokyo) 132

(2002) 629–634.

[133] P.C. Jones, A. Sivaprasadarao, D. Wray, J.B. Findlay, Mol.Membr. Biol. 13 (1996) 53–60.

[134] R.E. Dalbey, W. Wickner, J. Biol. Chem. 260 (1985) 15925–15931.[135] R.W. Jackson, J.A. DeMoss, J. Bacteriol. 90 (1965) 1420–1425.[136] T. Kimura-Someya, S. Iwaki, S. Konishi, N. Tamura, Y. Kubo,

A. Yamaguchi, J. Biol. Chem. 275 (2000) 18692–18697.[137] J.V. Moller, G. Ning, A.B. Maunsbach, K. Fujimoto, K. Asai, B.

Juul, Y.J. Lee, A. Gomez de Gracia, P. Falson, M. le Maire, J.Biol. Chem. 272 (1997) 29015–29032.

[138] G. Anderluh, A. Barlic, Z. Podlesek, P. Macek, J. Pungercar, F.Gubensek, M.L. Zecchini, M.D. Serra, G. Menestrina, Eur. J.Biochem. 263 (1999) 128–136.

[139] M. Holmgren, Y. Liu, Y. Xu, G. Yellen, Neuropharmacology 35(1996) 797–804.

[140] M. Krishnasastry, B. Walker, O. Braha, H. Bayley, FEBS Lett.356 (1994) 66–71.

[141] G. Bulaj, T. Kortemme, D.P. Goldenberg, Biochemistry 37 (1998)8965–8972.

[142] G.H. Snyder, M.J. Cennerazzo, A.J. Karalis, D. Field, Biochemis-try 20 (1981) 6509–6519.

[143] J. Kim, R. Schekman, Proc. Natl. Acad. Sci. USA 101 (2004) 905–906.

[144] B. Walker, H. Bayley, J. Biol. Chem. 270 (1995) 23065–23071.[145] H. Dale, C.M. Angevine, M.P. Krebs, Proc. Natl. Acad. Sci. USA

97 (2000) 7847–7852.[146] M.R. Jones, P.K. Fyfe, A.W. Roszak, N.W. Isaacs, R.J. Cogdell,

Biochim. Biophys. Acta 1565 (2002) 206–214.[147] G.E. Tusnady, Z. Dosztanyi, I. Simon, Bioinformatics 21 (2005)

1276–1277.

transmembrane protein topology mapping by the substituted cysteine accessibility method (scamtm):...

Documents