Dual-topology membrane proteins

in Escherichia coli

Susanna Seppälä

ii

©Susanna Seppälä, Stockholm 2011 ISBN 978-91-7447-351-3, pp. 1-66 Printed in Sweden by US-AB, Stockholm 2011 Distributor: Department of Biochemistry and Biophysics, Stockholm University

iii

Vanhemmilleni

iv

List of publications

Primary publications

I Rapp M*, Granseth E*, Seppälä S, von Heijne G (2006): Identification and evolution of dual-topology membrane proteins. Nature Structural and Molecular Biology 13, 112-116

II Rapp M*, Seppälä S*, Granseth E, von Heijne G (2007): Emulating membrane evolution by rational design. Science 315, 1282-1284

III Seppälä S, Slusky JS, Lloris-Garcerá P, Rapp M, von Heijne G (2010): Control of membrane topology by a single C-terminal residue. Science 328, 1698-1700

IV Lloris-Garcerá P, Bianchi F, Slusky JSG, Seppälä S, Daley DO, von Heijne G (201x): Antiparallel dimers of the small multidrug-resistance protein EmrE are more stable than parallel dimers. Manuscript in preparation

(* these authors contributed equally)

Additional publications

Granseth E, Seppälä S, Rapp M, Daley DO, von Heijne G (2007): Membrane protein structural biology – how far can the bugs take us? (Review) Molecular Membrane Biology 24, 329-332

Xie K, Hessa T, Seppälä S, Rapp M, von Heijne G, Dalbey R (2007): Features of transmembrane segments that promote the lateral release from the translocase into the lipid phase. Biochemistry 46, 15153-15161

Cassel M, Seppälä S, von Heijne G (2008): Confronting fusion-protein based membrane protein topology mapping with reality: the Escherichia coli ClcA H+/Cl- exchange transporter. Journal of Molecular Biology 381, 860-866

v

Abstract

Cellular life, as we know it, is absolutely dependent on biological membranes; remarkable superstructures made of lipids and proteins. For example, all living cells are surrounded by at least one membrane that protects the cell and holds it together. The proteins that are embedded in the membranes carry out a wide variety of key functions, from nutrient uptake and waste disposal to cellular respiration and communication. In order to function accurately, any integral membrane protein needs to be inserted into the cellular membrane where it belongs, and in that particular membrane it has to attain its proper structure and find partners that might be required for proper function. All membrane proteins have evolved to be inserted in a specific overall orientation, so that e.g. substrate-binding parts are exhibited on the ‘right side’ of the membrane. So, what determines in which way a membrane protein is inserted? Are all membrane proteins inserted just so?

The focus of this thesis is on these fundamental questions: how, and when, is the overall orientation of a membrane protein established? A closer look at the inner membrane proteome of the familiar gram-negative bacterium Escherichia coli revealed a small group of proteins that, oddly enough, seemed to be able to insert into the membrane in two opposite orientations. We could show that these dual-topology membrane proteins are delicately balanced, and that even the slightest manipulations make them adopt a fixed orientation in the membrane. Further, we show that these proteins are topologically malleable until the very last residue has been synthesized, implying interesting questions about the topogenesis of membrane proteins in general. In addition, by looking at the distribution of homologous proteins in other organisms, we got some ideas about how membrane proteins might evolve in size and complexity. Structural data has revealed that many membrane bound transporters have internal, inverted symmetries, and we propose that perhaps some of these proteins derive from dual-topology ancestors.

vi

Table of Contents

List of publications ...................................................................................... iv

Abstract ......................................................................................................... v

Abbreviations ............................................................................................. viii

Introduction................................................................................................... 9

The model organism........................................................................... 10

Biological membranes ................................................................................ 12

Membrane lipids................................................................................. 13

The lipid bilayers of E. coli ................................................................ 14

The outer lipid bilayer........................................................... 14

The inner lipid bilayer........................................................... 15

Membrane proteins............................................................................. 15

Membrane protein topology.................................................. 16

β-barrel membrane proteins ................................................. 17

α-helical membrane proteins ................................................ 18

E. coli membrane proteins.................................................................. 18

Outer membrane proteins ..................................................... 18

Inner membrane proteins ...................................................... 19

α-helical membrane proteins: topology, structure and evolution.......... 20

Topology and structure of α-helical bundles ..................................... 20

Structural repeats and evolution........................................... 21

Unusual topologies ............................................................................. 23

Bona fide dual-topology membrane proteins........................ 23

Other modes of dual topology ............................................... 27

Biogenesis of α-helical membrane proteins.............................................. 28

Targeting and integration of membrane proteins in E. coli................ 29

vii

Targeting of exported proteins.............................................. 29

Targeting of cytoplasmic membrane proteins....................... 29

The Sec translocon ................................................................ 30

YidC....................................................................................... 31

Topogenesis........................................................................................ 32

The nature of the polypeptide chain ................................................... 32

Hydrophobicity and aromatic amino acid residues .............. 32

Charged amino acid residues ............................................... 33

The importance of context: neighbouring helices................. 35

Role of the translocon/insertase.......................................................... 36

The size of the protein-conducting channel .......................... 36

The surrounding membrane................................................................ 37

The effect of lipids ................................................................. 37

Protein content of the membrane .......................................... 37

Methods and publications .......................................................................... 39

The model protein............................................................................... 39

Major experimental methods.............................................................. 40

Topology mapping using reporter proteins .......................... 40

Protein expression and selective radiolabelling ................... 41

In vivo ethidium toxicity assays ............................................ 42

Blue-Native PAGE ................................................................ 43

Cysteine labelling and crosslinking ...................................... 43

Summary of papers............................................................................. 44

Paper I................................................................................... 44

Paper II ................................................................................. 44

Paper III ................................................................................ 45

Paper IV ................................................................................ 45

Conclusions and perspectives .................................................................... 47

Populärvetenskaplig sammanfattning på svenska................................... 49

Acknowledgements ..................................................................................... 50

References.................................................................................................... 52

viii

Abbreviations

CL cardiolipin cryo-EM cryo-electron microscopy C-terminus carboxy-terminus GFP green fluorescent protein IPTG isopropyl β-D-thiogalactopyranoside N-terminus amino-terminus SMR small multidrug resistance PCC protein-conducting channel PE phosphatidylethanolamine PG phosphatidylglycerol PhoA alkaline phosphatase RNC ribosome:nascent chain complex SRP signal recognition particle

Amino acids

A Ala Alanine C Cys Cysteine D Asp Aspartate E Glu Glutamate F Phe Phenylalanine G Gly Glycine H His Histidine I Ile Isoleucine K Lys Lysine L Leu Leucine M Met Methionine N Asn Asparagine P Pro Proline Q Gln Glutamine R Arg Arginine S Ser Serine T Thr Threonine V Val Valine W Trp Tryptophan Y Tyr Tyrosine

9

Introduction

Cellular identity relies on the existence of the cellular membrane, a semipermeable barrier that encloses any cell and defines its boundary. In many cells, the interior is further divided into membrane enclosed compartments with specialized functions (organelles), and multicellular organisms are, simply put, large conglomerates of specialized, yet discrete, cells. The generation and maintenance of intracellular and organellar disparity is largely managed by membrane proteins that permit a controlled, continuous, transmembrane flow of material and information. Figuratively speaking, membrane proteins are the windows and doors of the cell: acting as receptors, connectors, channels and pumps, they are involved in innumerable translocations and signalling pathways; they are crucial to cell division and communication, as well as for energy harvesting processes such as photosynthesis and cellular respiration. Their importance is reflected by the fact that about a quarter of the genes in a typical organism encode integral membrane proteins (1-3); and that the majority of marketed drugs are, in one way or another, targeted towards membrane bound transporters, receptors and enzymes (4, 5).

In any living cell, at any given time, a wide variety of proteins are targeted to, and integrated into the membranes where they belong, and in those membranes the proteins assemble into functional units. Importantly, membrane proteins are believed to contain inherent information that is decoded by the membrane integration machineries, and that ensures that the protein is inserted in a correct overall orientation relative to the membrane. This thesis examines how, and when, the correct overall orientation of membrane proteins are established.

Most of the research for this thesis has been done using Escherichia coli EmrE as a model protein. Therefore, much of what is said below is focused on processes that have been observed in that particular bacterium. Naturally, examples from, and comparisons to, other organisms are drawn and made when deemed necessary. Here follows a short presentation of E. coli, and an introduction to biological membranes and membrane proteins. Then follows

10

more detailed discussions on the topology and biogenesis of α-helical membrane proteins. In an attempt to put this thesis work in a context, major results are presented and discussed throughout. Specific comments on methods, and a summary of the papers, are given in a separate chapter, followed by some conclusions and perspectives.

The model organism

E. coli is a rod-shaped, ~0.5 µm wide and ~2 µm long, gram-negative bacterium. It is part of the normal gut flora in humans and other animals, although there are pathogenic variants that may cause disease. Since its discovery by Theodor Escherich in 1885, E. coli has become a popular workhorse in laboratories worldwide, the main reasons being that the cells are easy to cultivate, and that their genetic content is easily manipulated. Many fundamental cellular processes and pathways were first illuminated with the aid of this bacterium, and it has proven central to our current understanding of basic biochemistry, molecular genetics and structural biology. It is exceptionally useful as an expression host for heterologous proteins, not the least evidenced by the successful expression of membrane proteins for structural studies (6, 7).

With this in mind, it is important to realize that as a taxon, E. coli comprises strains that may differ in as much as 80% of their genes (8, 9). Typically, E. coli cells have one circular chromosome of ~4.5 Mbp, encoding ~4500 proteins. Some strains, such as the enterotoxigenic variants, carry extra-chromosomal genes on plasmids (10). Strikingly, the pan genome of 61 sequenced E. coli strains was shown to encompass ∼15 000 gene families, and only about 1000 of these were found in every genome (9). This diversity may be explained by frequent horizontal gene transfer events, accompanied by a reasonably conserved chromosome size, meaning that apart from some core genes that seem to be indispensable, the genomic material of E. coli is highly dynamic and exchangeable (11).

The common laboratory strains derive from so called K- and B-strains that were in use already in the early twentieth century (12-14). Most of the work for this thesis was done using B-strain BL21(DE3) (15). The first complete E. coli genome to be published was that of K-12 MG1655, and the recent sequencing of the complete genome of BL21(DE3) explained interesting differences between K- and B-strains (14, 16, 17). For example, due to

11

deletions in genes that encode flagellar components, BL21(DE3) cells do not have flagella, and insertions in other genes cause the lack of a capsule polysaccharide, and a truncated core oligosaccharide in the cell wall (17, 18).

Henceforth, E. coli, refers to a ‘typical’ laboratory strain, however differences between strains are pointed out whenever necessary. As any gram-negative bacterium, E. coli has two membranes enveloping the cytoplasm. In between the outer and inner membranes is the periplasmic space that contains a layer of peptidoglycan: a netlike structure made of glycan strands that are cross-linked by peptides. In essence, the peptidoglycan functions as an elastic net bag that protects the cells against turgor pressure and rupture (19). The membranes and membrane constituents are described in more detail below.

12

Biological membranes

Biological membranes have two major components: lipids and proteins. In a seminal paper published in 1972, Singer and Nicholson conceptualized a biological membrane as an essentially two-dimensional fluid mosaic, with membrane proteins embedded in a lipid matrix (figure 1a) (20). Importantly, the model recognized that proteins can move laterally in the membrane, and associate with other membrane proteins. Indeed, it has become increasingly clear that biological membranes are rather organized structures, characterized by extensive interactions between a wide variety of lipids and proteins, the latter often found in large homo- and hetero-oligomeric complexes (figure 1b) (21).

Figure1.Abiologicalmembraneconsistsoflipidsandproteins.1a)Theclassicalmosaicmodel, from1972 (20). 1b)Amore recent takeon themosaicmodel,from2005(21).Bothreprintedwithpermission.

Most cellular membranes contain approximately equal amounts of lipid and protein by mass, although the ratios vary between different membranes, as does of course the lipid and protein species. For example, the cytoplasmic membrane of E. coli is densely packed with hundreds of different transporters, receptors and enzymes, while specialized human myelin sheaths are mostly lipidic and accommodates only a few different proteins

13

(22). Further, as is revealed by the variable shapes of cells and organelles, biological membranes are highly dynamic and exceptionally flexible. Depending on circumstances, cellular membranes can undergo considerable qualitative changes, and they can fuse and disconnect, disassemble and reassemble, on cue (23).

Cellular membranes function as semipermeable barriers that protect cells and allow a controlled separation of ions, molecules and biochemical reactions into confined spaces. The selective permeability is largely accounted for by membrane proteins, whereas the lipids represent an efficient barrier to most hydrophilic substances. It is however clear that the lipids are an active part of the functional membrane; they play pivotal roles in vesicular transport and signalling and, as is discussed below, they influence the topology, stability and function of many membrane proteins (23, 24). Here is a brief description of the lipid framework of a typical biological membrane, followed by an introduction to the proteins that reside therein.

Membrane lipids

With the exception of archaeal monolayers, a typical biological membrane is based on two layers of lipid molecules. Common membrane lipids are phospholipids, galactolipids, sphingolipids and sterols; their most conspicuous feature being an extended hydrophobic part carrying a hydrophilic headgroup (25, 26). Membrane lipids are orientated so that the hydrophobic parts make up the bilayer core, while the polar headgroups face the aqueous environment, forming the interfacial region. The membranes are generally in a fluid state, but in spite of the high thermal disorder, some structural information has been obtained, e.g. by the study of liquid crystals of dioleoylphosphatidylcholine (27). Biological membranes are approximately 50 Å thick, with the fatty acid core accounting for the central ∼30 Å. The polarity of the bilayer varies dramatically along the membrane normal, from the polar interfaces to the highly hydrophobic interior, and this creates an efficient barrier to the flow of ions and other polar molecules.

Many membrane lipids, such as the lipids in the cytoplasmic membrane of E. coli, have a backbone of glycerol with two ester-linked acyl chains and a phosphodiester-linked polar group. Depending on length and degree of

14

saturation, the acyl chains contribute to membrane thickness and viscosity, and it has been shown that this trait can be conditionally altered: for example, E. coli cells growing at room temperature have a higher degree of unsaturated fatty acids compared to fellow cells growing at 37 °C (28). The polar headgroups vary with respect to size and charge, and taken together, the acyl chains and the headgroup define the overall shape of the lipid molecule. Cylindrical lipids have similar cross-sectional areas for the headgroups and the acyl chains, and form bilayers in aqueous solutions. In contrast, lipids with comparatively small, or large, headgroups are conical and prefer hexagonal phases. Both bilayer and nonbilayer prone lipids are found in all biological membranes, and it is conceivable that just the right mixture of lipids is necessary for maintaining membrane shape and integrity (23, 29). In particular, the lipid composition is important for the stability and function of many membrane proteins, for reviews see (24, 30, 31).

The lipid bilayers of E. coli

E. coli cells have two membranes: an outer membrane and an inner, cytoplasmic, membrane. Here is a description of the lipid bilayers of E. coli, while the proteins that are embedded in the membranes are described later.

The outer lipid bilayer

The lipid distribution in the outer membrane is asymmetric. The outer leaflet consists of lipopolysaccharide, typically made of lipid-A molecules that are linked, via core oligosaccharides, to highly variable O-antigen polysaccharides (32, 33). O-antigen polysaccharides often cover the outer surface of bacteria, however many laboratory strains lack O-antigen, and some strains, such as BL21(DE3), have a truncated core oligosaccharide (12, 17, 18). This additional defect may explain the increased permeability of B-strains to some antibiotics and toxins (34-36 and our unpublished observations). The lipid composition of the inner leaflet resembles that of the inner membrane.

15

The inner lipid bilayer

Both leaflets of the inner membrane contain phosphatidylethanolamine (PE, ~75%), phosphatidylglycerol (PG, 20-25%), and cardiolipin (CL, ~5%) (37, 38). PE is zwitterionic and nonbilayer prone due to a small headgroup size, while the anionic PG and CL readily form bilayers. Remarkably, E. coli can survive without PE, as is evident by the engineered AD93 strain that lacks one of the enzymes in the PE synthesis pathway (39). The strain is not entirely healthy, it requires divalent cations to grow in rich medium and the cell division seems impaired – but the cells are nonetheless viable. Lipid analyses give at hand that the membranes of AD93 are enriched in PG and CL, and thus both the surface charge and the lateral pressure profile of the membrane are altered. It has been proposed that the cations required for growth make up for the lack of nonbilayer prone lipids, by altering the lipid packing and perhaps also by neutralizing the negative charge (40). Strikingly, monoglucosyldiacylglycerol, a nonbilayer prone lipid that is not normally found in E. coli, can largely substitute for the lack of PE, and it has been shown that the size of the lipid headgroup is critical for the function of membrane proteins (29, 41). The altered lipid composition of AD93 has major effects on the structural organisation of several membrane proteins, as is discussed further in the chapter on α-helical membrane protein topology below.

Membrane proteins

As mentioned above, membrane proteins intersperse all cellular membranes, and carry out a salmagundi of key processes, from transport of nutrients and waste products to signalling and cell division. It is well known, that proteins often function in homo- or hetero-oligomeric complexes, and that they may require cofactors such as metal ions or nucleotides for function. This is also true for membrane proteins, as shown e.g. by recent analyses of membrane protein complexes in E. coli (42, 43), for reviews see e.g. (44, 45). Here, for simplicity, the word ‘protein’ is used to denote a ‘single polypeptide chain’, and while allowing this, it is of course important to bear in mind that the functional unit may comprise several polypeptide chains and other molecules, as well as e.g. metal ions.

16

Membrane proteins are, by definition, more or less tightly associated with membranes. Depending on the strength of the attachment, the proteins are either peripheral or integral. Peripheral membrane proteins are engaged in relatively shallow interactions, such as hydrogen bonding with lipid headgroups, and they are readily extracted with carbonate (46). Many of these interactions are transient and difficult to predict theoretically. Integral membrane proteins, on the other hand, have hydrophobic parts that are firmly anchored in the membrane core, and they can only be extracted by the use of detergents or organic solvents. The hydrophobic characteristic of these proteins make them fairly easy to find by analyzing sequence data only, and genomic analyses suggest that about a quarter of the genes in any organism encode integral membrane proteins (1-3).

Membrane protein topology

The topology of a membrane protein describes how the polypeptide chain traverses the membrane and, importantly, gives the relative orientations of the transmembrane segments and the overall orientation of the protein relative to the membrane (47). Monotopic membrane proteins do not cross the membrane in its entirety; bitopic membrane proteins have one transmembrane segment; and proteins with more than one transmembrane segment are called polytopic (47). Dual-topology membrane proteins, at the focus of this thesis, have the unusual ability to adopt both opposite overall orientations in the membrane.

Reflecting the organization of the lipids in the bilayer, all integral membrane proteins are amphipathic. The hydrophobic parts anchor the proteins in the membrane core, while hydrophilic portions interact with lipid headgroups and polar surroundings. Any protein that resides in a lipid bilayer has to adapt to a highly anisotropic and energetically complex environment (48, 49). Pulling hydrophobic parts out of, and pushing hydrophilic parts into, the membrane core is energetically costly, and once properly inserted, it is thus not likely that integral membrane proteins undergo spontaneous major reorientations. The membrane milieu favours the formation of secondary structure, and based on the prevailing fold, membrane proteins are either categorized as α-helical bundles, or β-barrels (figure 2). While β-barrels have only been found in outer membranes of gram-negative bacteria, chloroplasts and mitochondria (50, 51), α-helical bundles constitute the

17

dominating class in virtually all other cellular membranes. Here follows a presentation of the two protein classes.

Figure2.Twoclassesofmembraneproteins.2a)Aβ‐barrelmembraneprotein(PDB 1BXW). 2b) An α‐helical membrane protein (PDB 1FQY). These, andfollowing,molecular graphics imageswere produced using theUCSF ChimerapackagefromtheResourceforBiocomputing,Visualization,andInformaticsattheUniversity of California, San Francisco (supported byNIH P41 RR001081),seehttp://www.cgl.ucsf.edu/chimera/.

β-barrel membrane proteins

A membrane β-barrel is made of an even number of antiparallel β-strands that are folded into a can-like structure, so that backbone hydrogen bonds are satisfied between the β-strands (figure 2a). Known membrane β-barrels have between 8 and 24 strands, and typically, although not always, the primary sequence is composed of alternating hydrophilic and hydrophobic amino acid residues, resulting in a barrel with a hydrophilic interior and hydrophobic exterior. Many β-barrel membrane proteins function as more or less selective transporters, while others have been implicated in membrane anchoring and stability (52-55). Although important for the function and integrity of gram-negative bacteria and the organelles of endosymbiotic origin, without further ado we will now turn our focus to α-helical membrane proteins.

18

α-helical membrane proteins

Analyses of predicted membrane proteomes give at hand that α-helical integral membrane proteins constitute the dominating protein class in most cellular membranes (1, 2). α-helix formation satisfies the hydrogen bonding potential of the polypeptide backbone, and although helices can be distorted and interrupted, they often traverse the hydrophobic membrane core (figure 2b). α-helical membrane proteins carry out a wide range of highly specialized functions and come in different sizes. Hypothetically, polytopic membrane proteins can have any number of transmembrane segments. In nature, the upper limit for an individual protein seems to be around 20 transmembrane helices, however one should not forget that functional membrane protein complexes often contain many more helices, see e.g. (56). Interestingly, experimental mapping of the cytoplasmic membrane proteomes of E. coli and Saccharomyces cerevisiae, revealed that polytopic membrane proteins often have an even number of transmembrane segments, with both N- and C-termini in the cytoplasm (57, 58). A well-known exception to this is the heptahelical members of the G-protein coupled receptor superfamily. These proteins are not prevalent in either E. coli or yeast, but are predicted to make up as much as ∼5% of the human proteome (3, 59).

The topology, structure and biogenesis of α-helical membrane proteins are discussed in more detail in the remaining chapters of this thesis. First, however, is a brief look at the membrane proteomes of E. coli.

E. coli membrane proteins

Outer membrane proteins

E. coli outer membrane proteins are either lipoproteins, or β-barrels. β-barrels make up approximately 2% of the entire E. coli proteome (60). Most β-barrel membrane proteins form porins, passive transporters that typically permit transmembrane diffusion of molecules up to 600 Da, while others are important for membrane stability and adhesion (52-55). Some β-barrel proteins interact with cytoplasmic membrane proteins and play important roles in the active extrusion of toxins, as exemplified by the AcrAB-TolC system (61, 62). A novel, and so far unusual, structural

19

arrangement was discovered in the Wza polysaccharide transporter that forms an α-helical barrel in the outer membrane (63). Notably, all outer membrane proteins are synthesized in the cytoplasm and transported through the inner membrane, further through the peptidoglycan layer, and into the outer membrane. Several components of the complex translocation/insertion systems have been identified, although many details remain to be described (64).

Inner membrane proteins

About a quarter of the genes in a typical E. coli cell are predicted to encode inner membrane proteins, and they are all of the α-helical bundle type. About 40% of the polytopic inner membrane proteins have been predicted to function as transporters and channels, and roughly 5% are involved in metabolic processes, signalling, and biogenesis of the cell envelope, respectively (57, 65). The remaining ∼35% have no annotated function. Especially smaller membrane proteins have eluded functional characterization, although some have been shown to be involved in signalling and stabilization of membrane protein complexes (66, 67). A global topology analysis of the inner membrane proteome revealed that the majority of the polytopic proteins have an even number of transmembrane helices and both N- and C-termini in the cytoplasm (57). As is discussed further below, a small subset of proteins in the inner membrane of E. coli are able to insert into the membrane in two opposite orientations (Paper I, 57, 68). As described in the chapter on membrane protein biogenesis, most proteins in the inner membrane of E. coli are inserted via the Sec pathway, and some of them require the insertase/chaperone YidC.

20

α-helical membrane proteins: topology, structure and evolution

Topology and structure of α-helical bundles

α-helical, integral membrane proteins have one or more helical transmembrane segment(s), and while the architecture of an individual membrane protein can be quite complex, some topological features are generally applicable. As is described in the next chapter of this thesis, all integral α-helical membrane proteins have hydrophobic stretches, and especially the distribution of positively charged amino acid residues has been shown to be an important topology-determining factor. In accordance with the positive-inside rule, cytoplasmic loops of membrane proteins tend to be enriched in lysines and arginines, and it has been shown that the orientation of membrane proteins can be altered by manipulation of this charge bias (69, 70). These, and other, general characteristics of membrane proteins have been used to successfully develop topology prediction algorithms (2, 71, 72).

In 1975, Henderson and Unwin managed to determine the three-dimensional structure of bacteriorhodopsin by cryo-electron microscopy (cryo-EM) (73). Ten years later, the first atomic-resolution structure of a bacterial photosynthetic reaction centre was solved by x-ray crystallography (74). As of 17 August 2011, 302 unique high-resolution membrane protein structures have been solved (http://blanco.biomol.uci.edu/mpstruc). Even if one takes into account structural variants, and low-resolution structures, merely 2% of the ~75 000 protein structures that have been deposited in the RCSB Protein Data Bank are of membrane proteins (75) (http://www.rcsb.org/pdb, http://pdbtm.enzim.hu). Nevertheless, intense research is continuously improving the methods to over-express, purify, and crystallize membrane proteins, and the number of atomic-resolution structures is exponentially increasing (7, 76-78).

High-resolution crystal structures have made it clear that the helices of polytopic membrane proteins are often rather elaborately organized

21

(http://blanco.biomol.uci.edu/mpstruc), for reviews see (79, 80). Although some proteins, such as rhodopsins, form canonical helix bundles, in many cases the helices are distorted, not all helices span the entire membrane, and re-entrant loops, formed by a polypeptide segment that enters and exits the membrane on the same side, are common. Interestingly, many membrane proteins are built of structural repeats, as described below.

Structural repeats and evolution

In light of the scarce structural data, careful sequence analyses and hydropathy profiling have been invaluable for the classification of membrane proteins into different families; see e.g. (81, 82). Membrane proteins are believed to evolve through gene duplication and fusion events (83-87). High-resolution structures have confirmed that most membrane proteins contain structural repeats that are arranged around an approximate symmetry axis either perpendicular to, or in the plane of, the membrane (figure 3) (88).

Figure3. LacY isoneexampleofamembraneproteinwithstructural repeats.3a)Thethree‐dimensionalstructureofLacY,withtheN‐andC‐terminalhalvesindifferentshadesofgrey(PDB1PV7).3b)AtopologymapofLacY, indicatingthe parallel organisation of the N‐ and C‐terminal halves (see text). Thetransmembranesegmentsarerepresentedasgreyandwhitesausages,andtheN‐andC‐terminiareindicated.

Generally, protein structure is more conserved than primary sequence. Two proteins with similar structures are not necessarily evolutionarily related, but at the same time this implies that proteins, or protein domains, that are related may share the same fold, yet the relationship is not necessarily apparent at sequence level (89). This is for example the case of the E. coli

22

AmtB ammonia channel, where the relationship between the repeated structural units is no longer visible in the primary sequence (90).

One example where the repeated elements are apparent at sequence level is the family of the hexahelical ADP/ATP carriers. Structural data show that these proteins are composed of three hairpin repeats that are arranged around a pseudo-threefold axis in the membrane (91), and the tripartite architecture was predicted by sequence analysis suggesting a triplication of the basic hairpin element (81). Another example is the arrangement of the 12 transmembrane helices of the E. coli multidrug transporter AcrB (belonging to the Root Nodulation and Division family). AcrB is folded in such a way that the six N-terminal and six C-terminal helices form domains that are related by twofold pseudo-symmetry (61). The 12 helices of E. coli Lactose permease, LacY (Major Facilitator Transporter superfamily), are correspondingly organized (figure 3). Again, the two halves of the protein are folded with twofold pseudo-symmetry around the central, substrate-binding cavity (92). Interestingly, each half of LacY is in turn composed of a three-helix repeat with internal, inverted symmetry (93). Similar inverted-repeats are seen in e.g. aquaporins, the H+/Cl- exchange transporters, and the aforementioned E. coli AmtB ammonia channel (figure 4) (90, 94, 95). The inverted-repeats are arranged around an approximate twofold axis in the membrane plane, and in contrast to the previous examples where the homologous units are parallel in the membrane, the inverted domains are oppositely orientated (96).

Figure4.AQP1isoneexampleofamembraneproteinwithinvertedstructuralrepeats. 3a) The three‐dimensional structure of AQP1, with the N‐ and C‐terminalhalves indifferentshadesofgrey(PDB1FQY).3b)AtopologymapofAQP1, indicating theantiparallelorganisationof theN‐andC‐terminalhalves.Thetransmembranesegmentsarerepresentedasgreyandwhitesausages,andtheN‐andC‐terminiareindicated.

23

Among the structures are many more examples of internal structural symmetries, especially in the class of transporters, and it is believed that the symmetrical architecture is necessary for function.

A particularly interesting structure for this thesis is that of the small multidrug resistance (SMR) transporter EmrE from E. coli, showing an antiparallel dimer composed of two identical, yet oppositely orientated subunits (97). This unusual arrangement was first suggested by projection structures determined by cryo-EM and image reconstruction of two-dimensional crystals (98, 99). As is discussed below, the work presented in this thesis further support the notion that EmrE is a genuine dual-topology membrane protein, and as such it adopts two opposite orientations in the membrane.

Unusual topologies

The topology of a membrane protein is sometimes dynamic and can be conditionally altered. As previously mentioned, E. coli strain AD93 lacks the zwitterionic membrane lipid PE, and has a negatively charged cytoplasmic membrane owing to elevated levels of lipids PG and CL (39). The altered lipid composition of AD93 has a fascinating effect on the topology of LacY that exhibits a partial and reversible topological inversion (100, 101). Similar lipid-dependent reorganisations have been shown to occur in a number of membrane bound transporters expressed in AD93, and it seems clear that these rearrangements are the result of a complex interplay between charges in the lipids and in the proteins (102-104). Further, there are proteins that exhibit somewhat unusual topological arrangements in their native membrane, and some of them are described here.

Bona fide dual-topology membrane proteins

A dual-topology membrane protein adopts two opposite orientations in its native membrane. A global topology analysis of the inner membrane proteome of E. coli disclosed a group of five proteins that seemed to have dual topology (57). A closer look at these proteins revealed that they are small, consisting of around 110 amino acid residues folded into four transmembrane segments, and, importantly, they have very few positively charged amino acid residues that are evenly distributed between cytoplasmic

24

and periplasmic loops (figure 5). In other words, these proteins have a very weak charge bias, and we could show that in compliance with the positive-inside rule, they are very sensitive to the addition or removal of positively charged residues (Paper I). One of these proteins was the aforementioned SMR transporter EmrE, the others were the SMR transporter SugE, camphor-resistance protein CrcB, and two were proteins of unknown function. Subsequent topology analyses have confirmed the dual topology of several other SMRs (105).

Figure5.ThedualtopologyofE.coliEmrE.5a)Thestructureoftheantiparallelhomodimer,withtheidenticalsubunitsindifferentshadesofgrey(PDB3B5D).5b) The topology of EmrE, indicating the dual topology and theweak chargebias.Thetransmembranehelicesarerepresentedasgreysausages,andtheN‐andC‐terminiareindicated.Thefilledblackcirclesarepositivelychargedaminoacidresidues(K22,R29,R82,R106).

As mentioned above, the dual topology of EmrE is supported by structural data (97-99). The high-resolution crystal structure is readily superimposed onto the cryo-EM density map (97). Importantly, it has been shown that the two-dimensional crystals bind substrate in a site between two monomers, with affinities that are indicative of a functional protein (99, 106-108). Cysteine labelling has further validated the dual orientation of EmrE monomers (109, 110). Taking into account evolutionary constraints and the position of essential residues (111-113), a model was based on the cryo-EM structure that supports antiparallel organisation of the subunits in the dimer (114). However while it seems clear that monomeric EmrE is a genuine dual-topology membrane protein, the relative orientation of the subunits in the functional dimer is under debate (Paper IV, 109, 115-118).

Other SMRs have been shown to consist of obligate heterodimers, such as E. coli YdgEF (MdtJI), and YkkCD and EbrAB from Bacillus subtilis (119-

25

121). The subunits of the dimeric EbrAB have opposite orientations in the membrane, and normally both are required for function (122). By exchanging and shortening the loops of the proteins, Kikukawa and co-workers managed to generate solely functional EbrA and EbrB (123, 124). The changes made, were in fact a manipulation of the distribution of positively charged amino acid residues, and seeing that the wild type protein forms an antiparallel dimer it is feasible that the solely functional proteins have dual topology, so that the antiparallel organisation is maintained.

A case of putative dual topology was found by sequence analyses of DUF606, a bacterial membrane protein family of unknown function. This family comprises i) genes that are predicted to encode dual-topology membrane proteins, ii) paired genes that encode homologous proteins that have a fixed orientation in the membrane, and iii) genes that encode large membrane proteins with homologous, oppositely orientated halves (Paper I, 85). The presence of these three topological variants in the same family suggests underlying evolutionary pathways, where a gene encoding a dual-topology protein can undergo duplication and/or fusion to generate homologous, oppositely orientated proteins, and proteins with structural, inverted, repeats (figure 6) (85).

The E. coli glutamate transporter GltS is a ten-transmembrane helix protein that is predicted to have an inverted structural repeat (125). In a study of the evolution of antiparallel two-domain membrane proteins, GltS was split, whereafter the two halves of the protein were fused in the reverse order, without loss of function in any case (126). On the same note, we duplicated the gene encoding EmrE, and made changes in the genes so that they encoded proteins with fixed, opposite, orientations in the membrane (Paper II). Importantly, the two oppositely orientated EmrE variants had to be co-expressed in order to get a functional transporter.

26

Figure 6. Proposed evolutionary relationships between dual‐topology andinverse‐topology membrane proteins. Gene X encodes a dual‐topologymembraneprotein,andundergoesduplication/fusionevents.GenesXaandXbencodehomologous,oppositelyorientatedmembraneproteins,whilegeneXa‐Xb encodes a twice as large membrane protein with an inverted structuralrepeat.Reprintedwithpermissionfrom(127).

Recently, experimental topology mapping of a number of small proteins in the cytoplasmic membrane of E. coli suggested that some of these one-transmembrane segment proteins exhibit dual topology (68). The function of these proteins is not known. Another one-transmembrane segment protein is MRAP, the Melanocortin 2 receptor accessory protein that is involved in the trafficking of the G-protein coupled receptor MC2 to the plasma membrane in adrenal glands. The protein functions as a homodimer and the two subunits have opposite orientations in the membrane (128). As such, it is the first eukaryotic antiparallel homodimer that has been reported to date. Further analyses indicated that antiparallel dimerisation is a prerequisite for the function of MRAP, and that the proteins are synthesized in two opposite orientations, i.e. the final antiparallel topology is acquired in the endoplasmic reticulum (129).

27

Other modes of dual topology

Ductin is a four-transmembrane helix protein that does not only exhibit dual topology, but also dual function (130). However it is worth to notice, that in this case, the two opposite topologies - and functions - exist in different cellular membranes. A hexamer of ductins make up the vacuolar Vo, which is a part of the vacuolar V-ATPase. In this scenario, ductin adopts a topology where both its N- and C-termini are in the extracytoplasmic lumen. However, ductin is also a part of a connexon channel in gap junctions, and there it has the opposite topology, i.e. both N- and C-termini are in the cytoplasm. How this dual topology has evolved is not clear.

Yet another example of dual topology is seen in the Hepatitis B virus large envelope protein (131). At the endoplasmic reticulum, the protein is inserted with a three-transmembrane helix topology. During maturation, in approximately 50% of the molecules, an N-terminal segment is inserted to generate a four-transmembrane segment topology. This mixed arrangement is preserved in the viral envelope, and presumably the two topologies represent two different functions.

The structures and topologies that are described here are just a few examples of what α-helical membrane proteins look like. Further experimentation and the increasing number of high-resolution crystal structures are likely to reveal additional cases of ‘odd’ structural arrangements that may be difficult to predict from sequence. These, and the examples mentioned previously, show that while topology can be complicated and dynamic, there are a few basic features that seem all-important for topogenesis, such as hydrophobicity and charge distribution. So, how are membrane proteins inserted into the membranes? What are the topology determinants? These questions are addressed in the next chapter of this thesis.

28

Biogenesis of α-helical membrane proteins

Most membrane proteins are integrated in a co-translational manner, and cellular insertion machineries have evolved to handle a number of challenging tasks: how are proteins targeted to the membrane? How are transmembrane segments recognized and inserted? How do transmembrane segments interact with each other, and how is the correct composition and stoichiometry of protein complexes controlled? Central to this thesis is the fundamental question of orientation: how, and when, is the correct topology of a membrane protein established?

With the exception of a few proteins encoded by mitochondrial and chloroplast genomes, cellular proteins are synthesized by cytoplasmic ribosomes. Proteins that exhibit their function in specific compartments, or outside the cell, are targeted to their respective locations either co- or post-translationally, as suggested by Blobel and coworkers some 40 years ago (132, 133). Bacterial and archaeal cells have machineries that allow targeting of proteins to, into, and through the cellular envelope; and eukaryotic cells have additional elaborate systems that ensure that nuclear, mitochondrial, chloroplast, and other organellar proteins end up in their respective compartment, see e.g. (134-138).

The cytoplasmic membrane integration process in E. coli is similar to the processes taking place at the cytoplasmic membrane of archaea, and at the endoplasmic reticulum of eukaryotes, and many of the participating molecules are homologous and universally conserved (139). Here is a presentation of the cytoplasmic membrane protein insertion systems in E. coli, followed by more general discussions on membrane protein topogenesis.

29

Targeting and integration of membrane proteins in E. coli

Targeting of exported proteins

In E. coli, proteins are targeted to the cytoplasmic membrane either for insertion or for translocation, for recent reviews see (135, 139, 140). Most exported proteins contain a well-characterized N-terminal cleavable signal sequence (141), and are targeted to the membrane via the post-translational SecB-pathway (142). SecB is a cytoplasmic chaperone that interacts with the exported protein, keeping it in a translocation compatible, unfolded, state. The substrate is delivered to the membrane, and facilitated by the SecA ATPase, the protein is fed through the membrane-embedded and universally conserved Sec translocon (143). Proteins that are translocated in a folded form, perhaps because cofactors have to be inserted previous to export, generally use the Twin-Arginine Translocation (TAT) system that exclusively handles folded proteins (144, 145).

Targeting of cytoplasmic membrane proteins

Most cytoplasmic membrane proteins in E. coli are targeted to the membrane by the universal signal recognition particle (SRP) pathway and inserted by the Sec translocon; in addition some proteins require the insertase/chaperone YidC for proper integration and folding, see e.g. (45). Compared to its substantially larger mammalian homolog, the bacterial SRP is minimalistic, comprising only a protein called Ffh and a 4.5S RNA (146, 147). When the ribosome starts translating an mRNA that encodes a membrane protein, perhaps already when the nascent chain is in the exit tunnel of the ribosome, SRP binds to the ribosome:nascent chain complex (RNC) (148, 149). In eukaryotes, it is believed that SRP binding stalls translation, but whether this is the case in E. coli is not entirely clear (150, 151). In any case, the RNC is guided to the membrane by SRP and its receptor, E. coli FtsY (152, 153). The GTPase activities of SRP and FtsY now leads to the disassembly of RNC:SRP, and subsequent docking of RNC onto the membrane embedded protein-conducting channel. Powered by the ribosome, the nascent chain is fed through - or into - the membrane, while it is being made.

Whether, and how, a transmembrane segment is recognized and inserted into the membrane is the result of the interplay between several factors that will be described below. First, however, follows a description of the major

30

known membrane protein insertion machineries in E. coli, the Sec translocon and the YidC insertase/chaperone.

The Sec translocon

One of the major protein translocation systems, not only in E. coli but in any cell, is the Sec machinery. At the core of the Sec machinery is the membrane embedded Sec translocon that allows both export of proteins through the membrane, and integration of proteins into the membrane. The central protein-conducting/inserting channel (PCC) is found in the cytoplasmic membranes of bacteria (SecYEG) and archaea (SecYEβ), and in the endoplasmic reticulum of eukaryotic cells (Sec61αβγ) (139). SecYE/αγ are homologous and form the core of the channel, while the G/β subunits are neither homologous nor essential. Parts of the PCC are also found in thylakoid membranes of chloroplasts, and while most mitochondria seem to have lost the genes, the mitochondrial genome of the unicellular fungus Reclinomonas americana encodes a SecY homolog, corroborating that the Sec machinery is truly universal (154, 155).

The high-resolution crystal structure of SecYEβ from Methanocaldococcus jannaschii provided clues as to how the PCC handles the dual task of translocation/insertion (figure 7) (156). SecY forms the actual channel, SecE seems to form a stabilizing clamp that holds the channel together, and the nonessential Secβ is peripherally located. SecY has 10 transmembrane helices and viewed from the plane of the membrane, it has an hourglass shape (figure 7a). At the most constricted point, the channel is lined by a ring of hydrophobic amino acid residues, and in the non-translocating structure, a re-entrant loop on the periplasmic side seems to act as a plug that prevents uncontrolled leakage of ions from one side of the membrane to the other. It has been suggested that when the Sec translocon is activated, the plug moves out of the way so that polypeptide chains can pass (157). Viewed along the membrane normal, SecY has an approximate clamshell shape: the two halves of the protein, that incidentally are related by an inverted twofold pseudo-symmetry, seem to be hinged around transmembrane helices 5 and 6; and between transmembrane helices 7 and 8 there is an opening that has been suggested to function as a lateral gate through which the polypeptide chain is released into the membrane bilayer (figure 7b) (156).

31

Figure 7.M. jannaschii SecYEβ heterotrimer (PDB 1RHZ). 7a) View from themembraneplane.SecYandSecE/βareindifferentshadesofgrey.Theasteriskindicatesthelateralgate.7b)Cytoplasmicview.

It has been proposed that during translocation, the channel gate opens towards the membrane lipids in such a way that a passing polypeptide chain has access to the lipid environment (158-160). If the passing polypeptide is hydrophobic enough, it will partition into the hydrophobic bilayer; otherwise it will pass by and exit on the other side of the membrane.

In E. coli, SecY has ten, SecE has three, and SecG has two transmembrane helices. The overall architecture of the core SecYE is conserved (161). Interestingly, E. coli can be depleted of many components of the Sec machinery and still survive (162-164). In addition to the components mentioned above, the E. coli Sec machinery also includes the heterotrimeric membrane bound SecDF-YajC complex (165). The exact role of these proteins is not known, however it has been suggested that SecDF aids in proton-motive force driven translocation, regulates the membrane cycling of SecA, and mediates contact between SecYEG and YidC (166-168).

YidC

Some proteins in the cytoplasmic membrane of E. coli require YidC for proper insertion and/or assembly (45, 169-172). YidC is an integral membrane protein belonging to the YidC/Oxa1/Alb3 family that has members in bacteria, mitochondria, chloroplasts and some archaea (173, 174). E. coli YidC has six transmembrane helices, however most family members are lacking the N-terminal helix and the following periplasmic

32

loop (174). Apart from assisting in the assembly of some Sec-dependent membrane proteins, YidC has been shown to function as an independent insertase for a number of small, hydrophobic membrane proteins such as M13 and Pf3 phage coat proteins (169, 175, 176). A recent global proteomic study of YidC-depleted E. coli cells revealed that in particular, proteins with comparatively small soluble domains (<100 amino acid residues) are sensitive to YidC depletion (177).

The exact details of how YidC recognizes and integrates membrane proteins are not clear. Cross-linking studies have shown that the conserved third transmembrane helix in E. coli YidC contacts nascent membrane proteins (178). A cryo-EM structure of YidC bound to a translating ribosome suggests that YidC functions as a dimer (179), and a recent study indicated that YidC-mediated insertion can be facilitated by additional membrane-associated proteins, such as YidD (180).

Topogenesis

The polypeptide chains of membrane proteins are believed to contain inherent information that is decoded by the membrane integration machinery, and that ensures correct insertion and overall orientation of the protein in the membrane. How a membrane protein is inserted into the membrane depends on several factors. Primarily, the nature of the polyptide segment is important, with respect to hydrophobicity and distribution of charged amino acid residues. Second, the actual insertase and associated proteins may provide interaction surfaces that facilitate insertion and topogenesis, and third, it is conceivable that the surrounding lipids and the protein content of the membrane are likewise important for insertion, assembly and oligomerisation.

The nature of the polypeptide chain

Hydrophobicity and aromatic amino acid residues

It has long been recognized, that a firm anchoring of proteins in biological membranes requires hydrophobicity (181, 182). More recently, the free energy of transfer from the polar interior of the translocon into the bilayer

33

membrane was systematically determined for each of the twenty naturally occurring amino acids, and the experimental data was subsequently used to develop and improve algorithms for membrane protein topology prediction (72, 183, 184). The ‘molecular code’ for insertion into the endoplasmic reticulum is similar regardless of the orientation of the helix (185). As expected, it was shown that hydrophobic amino acid residues promote membrane insertion of a model helix, while charged and other polar residues oppose it (184). In a recent study, non-proteinogenic amino acids with increasingly large hydrophobic side chains were incorporated into a similar model helix, and it was shown that membrane partitioning is directly proportional to the hydrophobic area of the amino acid side chain (186).

The distribution of amino acids in transmembrane helices is restricted by the nature of the side chains, with respect to hydrophobicity, bulkiness, and opportunities for intra- and interhelical interactions. Leu, Ile, Val, and Ala are good helix-formers and often found in transmembrane segments, while Pro is known to induce kinks (187-189). Gly, the smallest of the amino acids with a side chain consisting of only one hydrogen atom, is considered a helix breaker in soluble proteins but is often found in transmembrane helices where it is involved in helix-helix interactions (190-192). For some cases the position of the amino acids in the chain is important, as in the case of tryptophans and tyrosines, and charged amino acid residues (see below). Centrally placed Trp and Tyr oppose the insertion of a model helix, while the same residues are permitted, or even enhance insertion, when moved toward the ends of the helix (184). This is entirely in accordance with what has been observed for natural proteins, where Trp and Tyr are often found in the interfacial regions, an occurrence referred to as the aromatic belt. The aromatic residues interact favourably with the polar headgroups of the lipids, and are believed to anchor and fix tilt angles of the helices relative to the bilayer (193-195).

Charged amino acid residues

Flanking charges have been shown to be important topology determinants; with the most conspicuous example being the distribution of positively charged amino acid residues Lys and Arg. Statistical sequence analyses revealed that positively charged residues are especially prominent in the cytoplasmic loops of membrane proteins, as stated by the positive-inside rule (69). The distribution of positively charged amino acid residues has been

34

shown to act as a strong topological determinant, and seems generally applicable (70, 187). The topology of membrane proteins can be inverted by the addition of positively charged amino acid residues to extramembraneous loops (Paper II, 70). Similarly, ‘frustrated’ membrane proteins where marginally hydrophobic helices are forced to insert, or hydrophobic helices are forced out of the membrane, have been generated by simple manipulation of the positive charge bias (196, 197). We have shown that for the dual-topology membrane protein EmrE, the effect of a single positive charge is equally strong regardless of position: an N-terminally placed positive charge fixes the orientation of the protein as well as a C-terminally placed positive charge (Paper III).

Lysines and arginines have a similar and equally strong effect on the topology of transmembrane proteins (198, 199). Histidines are partially positively charged during cellular conditions and exhibit a similar topological effect as Lys and Arg, providing that three or more His residues are present (Paper III, 199). As topology determinants, negatively charged residues are much less potent than positively charged residues, unless they are present in large quantities (198). However there are some indications that in some cases the negatively charged amino acids are important, and while there is no conclusive evidence for a ‘negative-outside rule’, sometimes the difference between positively and negatively charged amino acid residues might be a more suitable determining factor, especially for eukaryotic membrane proteins (104, 200).

Importantly, transmembrane helices are not all hydrophobic, but rather often contain polar and even charged amino acid residues that act as functional groups (201). One example is the multiple Arg in the S4 segment of the KvAP voltage-dependent potassium channel; another is the highly conserved Glu in the first transmembrane helix of SMR transporters (202-204). Neighbouring residues and helices conceivably aid the insertion of positively and negatively charged amino acid residues into the membrane. In addition, Lys and Arg are able to exhibit ‘snorkelling’ with the side chains being so long that the charged end groups may be in contact with the polar lipid-water interface region, while the aliphatic parts of the residues remain in the hydrocarbon region, maximizing energetically favourable interactions (205).

35

The importance of context: neighbouring helices

It is becoming increasingly clear that some polytopic membrane proteins are inserted correctly only when the whole protein is present, and that the individual transmembrane helices can not always be inserted on their own. In some cases, the insertion efficiency of a transmembrane segment is heavily affected by neighbouring helices and flanking residues, see e.g. (206-209). The eukaryotic ABC-transporters P-glycoprotein and CFTR have been shown to exhibit multiple topologies, likely caused by inefficient insertion of marginally hydrophobic helices (206, 210, 211). Native aquaporins have six transmembrane helices and while each subunit forms a functional channel, they all form homo-tetramers in membranes. Despite these structural, and functional, similarities aquaporins seem to display different modes of membrane integration. The six helices of AQP4 are inserted in an orderly fashion, while AQP1 is initially inserted in a four-transmembrane helix topology (212, 213). During maturation, the third transmembrane helix of AQP1 undergoes a 180° rotation, so that the loops between helices 3-4 and 4-5 are pulled through the membrane, resulting in the final six-membrane spanning topology. A similar case is found in the archaeal glutamate transporter homolog GltPh from Pyrococcus horikoshii (214). This protein functions as a homotrimer and each subunit has eight transmembrane helices. Transmembrane helix 4 is marginally hydrophobic and kinked, and is initially not inserted, but rather pulled into the membrane at a later stage. The mechanisms for these rearrangements is not known, however it is likely that they are coupled to, and stabilised by, assembly and oligomerisation. Selective retention of transmembrane helices at the translocon has also been seen for e.g. the heptahelical transmembrane protein Opsin, where the three last transmembrane helices remain in the vicinity of the translocon until the whole protein has been synthesized (215, 216). Also, as mentioned above, we could show that a C-terminally placed positively charged amino acid residue had an effect on the overall topology of the small multidrug transporter EmrE (Paper III). This is telling of a remarkable topological malleability, however the underlying mechanisms are not clear.

36

Role of the translocon/insertase

It is conceivable that the insertion machinery affects the way a membrane protein is integrated into the membrane. A polypeptide chain that is fed into the PCC is either laterally released into the lipid bilayer, or it remains in the interior of the channel and undergoes subsequent translocation (156). At the centre of the translocation channel of SecY/61α, there is a ring of six hydrophobic residues believed to function as a seal against the flow of ions during translocation. It has been shown in yeast, that if these residues are replaced with polar or charged residues, insertion efficiency of a moderately hydrophobic test segment was increased (217). This indicates that the protein-conducting channel adjusts the hydrophobicity threshold for membrane partitioning. Also other residues, e.g. in the plug-region, of the SecY/61α translocon have been shown to affect the integration efficiency and orientation of transmembrane segments (218).

How membrane proteins are fed into the PCC is not entirely clear. Experimental data indicates that regardless of the final orientation, the first transmembrane segment of a membrane protein inserts head-first into the Sec61α translocon pore (219, 220). This implies that a segment that has its N-terminus in the cytoplasm has to reorient, and it has been suggested that the segment has a limited amount of time to do so (219). An alternative mechanism for insertion of individual helices is the “hairpin model”, where the polypeptide chain is looped into the protein-conducting channel, and once sufficient length of downstream sequence is translated, the segment is integrated into the membrane, see e.g. (221). It is not clear if all transmembrane segments follow the same mode of insertion, and the molecular details of reorientation and looping of transmembrane segments remain to be described.

The size of the protein-conducting channel

In any case, the size of the actual PCC seems important. In the M. jannaschii structure of SecY, which is believed to represent the closed state, the central pore is narrow, and the ~10 Å diameter provides space for one transmembrane helix at a time (156). Recently, SecYEG from E. coli was challenged with sizable rigid spherical molecules fused to known Sec-dependent preproteins, and it was shown that the translocon could handle molecules up to 22-24 Å (222). However there is fluorescence quenching

37

data indicating that the size of the eukaryotic Sec61 translocon pore is substantially larger (40-60 Å) during co-translational translocation (223). The oligomeric state of active SecY/61 complexes has been under debate, however crosslinking- and structural data agree that while the Sec translocon is capable of forming higher oligomers, a single copy of the heterotrimer is able to form an active PCC (224-226). Lastly, the role of Sec-translocon associated proteins should not be underestimated, as they may provide interaction surfaces that enhance insertion and topogenesis. Examples of such proteins are the aforementioned E. coli SecDF-YajC, and YidC; and although not discussed in this thesis, eukaryotic cells have their own setup of additional translocon-associated proteins, such as TRAM, that have been implied in the insertion and assembly of membrane proteins, see e.g. (221) and references therein.

The surrounding membrane

The effect of lipids

As discussed above, membrane lipids are important for proper insertion and stability of many membrane proteins, for reviews see (24, 30, 31). The aforementioned lipid-dependent topological inversions of transporters in E. coli are striking examples (100-104). Also, anionic phospholipids have been shown to be important for topogenesis of E. coli Leader peptidase, Lep, in vitro (227). In eukaryotic cells, many membrane proteins are inserted into the endoplasmic reticulum and subsequently targeted to e.g. the plasma membrane by vesicular transport. The lipid composition of the different membranes in a eukaryotic cell varies, and it is not unlikely that this affects the topology, and function, of the membrane-embedded proteins.

Protein content of the membrane

Most biological membranes are quite densely packed with many different kinds of proteins (22). It is perhaps difficult to test in vivo, but a recent computational simulation suggests that the protein content of the membrane is important for efficient insertion of especially polar amino acid residues (228). In contrast to models assuming that the membrane consists of pure lipid, this model incorporated protein segments into the bilayer. The result

38

was that the free energy cost of inserting a positively charged amino acid residue into the membrane was dramatically lower, compared to the pure lipid-models, and in fact similar to experimentally determined values (183, 184).

39

Methods and publications

Here is a presentation of the major experimental methods used throughout this work, with comments. Then follows a summary of the papers that are included in this thesis.

The model protein

A lot of the work for this thesis was done on E. coli EmrE, a secondary transporter belonging to the SMR family (figure 8) (229). EmrE is expressed in the cytoplasmic membrane of E. coli, and it uses the proton gradient to extrude hydrophobic, cationic toxins such as ethidium, acriflavine, and methyl viologen, see e.g. (230). EmrE is 110 amino acids long and has four transmembrane helices. As indicated in figure 8, it has four positively charged amino acid residues (K22, R29, R82, R106), one histidine (H110), and three negatively charged amino acid residues (E14, E25, D84). The mechanism of EmrE has been carefully studied. It has been shown that the minimal functional unit is a homodimer, although the presence and functional importance of higher oligomers in vivo cannot be excluded (106, 231, 232). EmrE binds toxins and protons in a site between two monomers, and the binding is coordinated by a glutamate from each subunit (111-113). The glutamate in this position (E14) is highly conserved within the SMR family and cannot be replaced even by an aspartate in the wild-type protein, however we have shown that an Asp is allowed in one of the subunits in the functional dimer, although the function is slightly impaired (Paper II, III, 111, 112). While structural, biochemical and phylogenetic data support a dual-topology for EmrE (Paper I, II, III, 97-99, 110), the organization of the subunits in the dimer is still under debate (Paper IV, 109, 115-118).

40

Figure8.ThetopologyandfunctionofE.coliEmrE.AnantiparallelEmrEdimerisshown.Transmembranesegmentsarerepresentedasgreysausages,andtheN‐ and C‐termini are indicated. The filled black circles are positively chargedaminoacidresidues(K22,R29,R82,R106),thefilledgreycirclesarehistidines(H110),andtheemptycirclesarenegativelychargedaminoacidresidues(E14,E25,D84).Theantiporterfunctionisindicated.

Major experimental methods

Topology mapping using reporter proteins

One way to experimentally probe the topology of a membrane protein is to use reporter proteins such as Green Fluorescent Protein (GFP) and Alkaline Phosphatase (PhoA). These two reporters are particularly useful in E. coli, because their activities are disparately localized. Most GFP-variants only fold and become fluorescent in the cytoplasm of E. coli; while PhoA, owing to the requirement of a couple of disulfide bonds, is active only when localized to the periplasm (233, 234). Therefore, if GFP or PhoA is fused to the membrane protein of interest, the localization of the fusion point (cytoplasm/periplasm) is easily assayed. This way, GFP and PhoA have been successfully used to determine the location of the C-terminus of virtually every single polytopic protein in the inner membrane of E. coli, and GFP has further proven useful for monitoring the expression and folding of

41

membrane proteins e.g. for structural studies (57, 235-237). Here, we have used GFP and PhoA to monitor changes in the overall orientation of e.g. EmrE-variants (Paper I, II). However adding a tag to a topologically sensitive membrane protein may in itself affect the overall orientation, and therefore tags should be used with care for topological studies (Paper II). This is particularly true for tags that contain charged residues and/or histidines (Paper III, 109). In some cases it can be desirable to do complementary topology determinations e.g. by selective cysteine labelling, see below.

Protein expression and selective radiolabelling

Throughout this work we have expressed and selectively radiolabelled EmrE with 35S-methionine, using E. coli BL21(DE3) cells, pET-vectors, and the rifampicin blocking technique (15, 238). E. coli (DE3)-strains have a gene encoding T7 RNA polymerase in their chromosome (15). The T7 RNA polymerase originates from bacteriophage T7, and recognizes a promoter sequence that is not naturally present in the E. coli genome, but that is present in commercially available pET-vectors. In the cells, the expression of the T7 RNA polymerase is under control of the lacUV5 promoter, and inducible by isopropyl β-D-thiogalactoside (IPTG). IPTG induction leads to the production of T7 RNA polymerase, and to the subsequent expression of any genes that are under control of a T7 promoter. Rifampicin is an antibiotic that inhibits bacterial RNA polymerases, while the T7 RNA polymerase is left unaffected (238). Therefore addition of rifampicin results in the selective expression of T7 promoter-controlled genes, and this can be used for selective radiolabelling: E. coli (DE3) cells are treated with IPTG and rifampicin, and any gene(s) under the T7 promoter are subsequently selectively labelled by the addition of 35S-methionine. The gene products are easily analyzed by SDS-PAGE. This is particularly convenient if no antibody is available against the protein of interest, as in the case of EmrE.

The pET Duet-1 vector (Novagen) used in this work has two multiple cloning sites, each preceded by a T7 promoter/lac operator and a ribosome binding site. This allows for co-expression of two genes, however while the expression from the two cloning sites is strong, it is not equal (Paper IV). It is worth to note that the DE3/T7-system described here is ‘leaky’ and that some expression occurs even in the absence of inducer (239). This is of

42

course of special importance if the gene product is somehow toxic to the cells, as is often the case with membrane proteins (77).

In vivo ethidium toxicity assays

As described above, E. coli EmrE renders the cells resistant to e.g. ethidium. Throughout the work for this thesis, we have introduced many changes to the primary sequence of EmrE, e.g regarding the distribution of positively charged amino acid residues. Of course, when one changes the primary sequence of a protein, it is important to test whether the protein is still functional. For this purpose, we used an in vivo ethidium toxicity assay, with liquid cell cultures or plates. In short, E. coli BL21(DE3) cells were transformed with the relevant plasmids and grown in the presence of varying concentrations ethidium bromide. To avoid toxic EmrE levels due to over-expression, no inducer was used in the assays. It is worth to note, that the expression system that we used allows simultaneous co-expression of two genes (see above). A comparison between two different strains, BL21(DE3) and MC4100(DE3), revealed that MC4100(DE3) cells are far more resistant to ethidium than are BL21(DE3) cells (unpublished observation). One explanation could be that BL21(DE3) cells are more permeable to the toxin, due to a truncated core oligosaccharide in the cell wall (17, 18, 34-36). Importantly, we could not observe any qualitative differences between the strains, meaning that different EmrE-variants rendered the cells equally resistant, providing that enough ethidium was present to kill ‘background growth’ (=cells carrying empty plasmid vectors). Further, comparing sessile and liquid growth of BL21(DE3) cells, we saw that the growth conditions are important: BL21(DE3) cells growing on a plate are much more sensitive to ethidium, compared to fellow cells growing in liquid medium. One possible explanation is the fact that BL21(DE3) cells are not able to form proper biofilms, and therefore sessile growth makes them more sensitive (17). Again, we have not observed any qualitative differences between the assays.

In the toxicity assays, we often took use of the fact that the activity of EmrE is abolished if the conserved glutamate (E14) in the first transmembrane helix is replaced by any other amino acid. However, as we first show in Paper II, the glutamate can be replaced by an aspartate in one of the subunits in a heterodimer formed by oppositely orientated EmrE-variants. Assuming that an antiparallel dimer is a prerequisite for function, this allowed us to

43

indirectly assay the orientation of charge-altered and otherwise mutated EmrE versions, by co-expression with EmrE variants with a known, fixed, orientation in the membrane (generated in Paper II and coined EmrE-Nin/Cin and EmrE-Nout/Cout, see below).

Blue-Native PAGE

Blue-native polyacrylamide gel electrophoresis (BN-PAGE) allows for the separation of membrane protein complexes, see e.g. (42). In Paper IV, we solubilised radiolabelled EmrE in the mild detergent n-dodecyl-β-D-maltoside (DDM), and analysed the presence of dimers and higher oligomers using BN-PAGE. The oligomeric state in the micelles generally reflects the organisation in the native membrane, however we could show that upon heating, proteins were able to reassemble into the preferred, most stable arrangement, regardless of starting point (Paper IV).

Cysteine labelling and crosslinking

Selective labelling and crosslinking of cysteines can be used to probe the topology and organisation of transmembrane helices, see e.g. (109, 240). By introducing single cysteines into a cysteine-less membrane protein, one can determine the localization (cytoplasmic/extracytoplasmic) of the unique residue. One can for example use membrane-impermeable 2-(trimethylammonium) ethyl methanethiosulfonate (MTSET) to block any cysteines that are exposed on the outside of the cytoplasmic membrane. Following lysis, any unreacted cysteines can be detected by use of MalPEG5000, a cysteine-specific reagent that causes a size shift that is easily detected by conventional SDS-PAGE.

Cysteine crosslinking can be catalyzed by copper phenantroline. Again, unique cysteines are introduced to the membrane protein of interest. If two cysteines are sufficiently close, addition of copper phenantroline will induce the formation of disulfide bridges, and dimer formation can be analyzed by SDS-PAGE.

44

Summary of papers

Paper I

Here, we investigated the nature and occurrence of putative dual-topology membrane proteins in bacteria. Five candidate dual-topology membrane proteins in the cytoplasmic membrane of E. coli had in common a small size and a weak charge bias; i.e. they have very few positively charged amino acid residues that are evenly distributed between cytoplasmic and periplasmic loops (57). Using C-terminal topology reporters GFP and PhoA, we could show that in compliance with the positive-inside rule, these proteins are topologically hypersensitive to the addition and removal of single positively charged amino acid residues, as opposed to homologous proteins with an unmistakable natural charge bias.

Looking for dual-topology membrane proteins in other bacteria, we found that genes encoding putative dual-topology proteins occur as isolated ‘singleton genes’. In contrast, we also found closely spaced gene pairs encoding homologous proteins with predicted opposite orientations in the membrane (such as E. coli YdgE/F). Further, we came across one bacterial membrane protein family (DUF606) that contained i) genes encoding putative dual-topology membrane proteins, ii) paired genes encoding homologous proteins with predicted opposite orientations in the membrane, and iii) genes that are twice as long and that code for large membrane proteins with homologous, oppositely orientated halves. This finding suggested an underlying evolutionary pathway, where a gene encoding a dual-topology membrane protein can be duplicated/fused and undergo divergent evolution to form a family of homologous, yet topologically mixed, proteins.

Paper II

Here we wanted to look closer at the evolutionary pathway outlined above. We duplicated the gene encoding the dual-topology membrane protein EmrE, and changed the duplicated genes so that they encoded proteins with fixed, opposite, orientations in the membrane. The topologically fixed variants were generated by the following charge manipulations: EmrE-Nin/Cin (R29G, R82S, S107K), and EmrE-Nout/Cout (T28R, L85R, R106A). The overall orientations of the proteins were probed using C-

45

terminal reporters, PhoA and GFP. Using the in vivo ethidium toxicity assay described above, we could show that co-expression of the oppositely orientated proteins was absolutely required for function. Also, it became apparent that the heterodimer allows for changes that are not permitted in the homodimer. One example is the glutamate in position 14, which can be replaced by an aspartate in one of the subunits in the dimer, but not both. This could be one of the reasons why paired genes encoding heterodimers seem to be more common in nature, than single genes encoding dual-topology homodimers.

Paper III

In this paper, we wanted to look closer at the effect of positive charges on the overall orientation of EmrE. Using mutagenesis, we systematically added single positive charges to each of the loops of EmrE. Using the ethidium in vivo toxicity assay described above, we could determine the topological effect of the introduced charges. For example, we could show that EmrE with an additional positive charge in the loop between transmembrane helices 1-2 is only functional if it is co-expressed with EmrE-Nin/Cin, indicating that the positive charge makes the protein insert with its N- and C-termini on the periplasmic side of the membrane. A positive charge in the next loop, between transmembrane helices 2-3, had the opposite topological effect, and this particular protein was only functional if it was co-expressed with EmrE-Nout/Cout. Importanly, we could also show that EmrE(E14D), which of course is not functional by itself due to the replacement of E14 by D, forms a functional dimer if it is co-expressed either with EmrE-Nin/Cin or with EmrE-Nout/Cout, as would be expected if the wild-type protein has dual topology.

Remarkably, most positions where we placed a positive charge had an equally strong, and predictable, effect on the overall orientation of the protein, from the N-terminus to the very C-terminal end. This is telling of an exceptional topological malleability, where the whole protein remains undecided until the very last residue has been synthesized.

Paper IV

Here we have investigated the dimeric organisation of EmrE. Although it seems clear that monomeric EmrE can insert into the membrane in two

46

opposite orientations, the relative orientation of the subunits in the dimer has been under debate (109). Using EmrE variants with fixed, opposite orientations in the membrane (EmrE-Nin/Cin and EmrE-Nout/Cout), we show that although the proteins can form parallel dimers, an antiparallel organization of the subunits in the dimer is preferred. Cysteine crosslinking and Blue-Native PAGE analyses of intact oligomers reveal that in membranes, the proteins form parallel dimers only if no oppositely orientated partner is present. Co-expression of oppositely orientated proteins almost exclusively yields antiparallel dimers. Further, parallel dimers can be disrupted and converted into antiparallel dimers by heating of detergent solubilized proteins. As we have seen before, in vivo function is clearly correlated to the presence of antiparallel dimers, and taken together our results strongly suggest that an antiparallel arrangement of the subunits in the dimer is more stable than a parallel organization.

47

Conclusions and perspectives

The subject of this thesis is dual-topology membrane proteins in E. coli. Dual-topology membrane proteins can insert into their native membrane in two opposite orientations. One prerequisite for dual topology seems to be a weak positive charge bias, and we have shown that in compliance with the positive-inside rule, these finely balanced proteins are very sensitive to changes in the distribution of positively charged amino acid residues. Further, we have shown that the dual-topology protein EmrE is topologically undecided until the very last residue has been synthesized. This addresses interesting questions about the insertion of these proteins. Currently, very little is known about the mechanisms for membrane targeting and integration of dual-topology membrane proteins. We are in the process of investigating the SecY and/or YidC dependency of E. coli EmrE, and the future will show whether these proteins require one, both, or neither of these insertion machineries. Apart from the (lack of) positively charged amino acid residues, it is not entirely clear how the dual topology of EmrE and similar proteins is ensured. It is conceivable that the lipid composition of the native membrane is important.

While it seems clear that monomeric EmrE is inserted in two opposite orientations, the organisation of the subunits in the EmrE dimer is under some debate. Our results suggest that an antiparallel organisation is a prerequisite for function, and the preferred arrangement, however the proteins are able to form parallel dimers if no antiparallel partner is provided. Also, possible functional importance of higher oligomers in vivo cannot be ruled out. It will be interesting to compare the intermolecular arrangements of the antiparallel and parallel dimers, and investigate the exact function of the different conformations and assemblies. This could be done by cysteine crosslinking, mutagenesis and subsequent analyses by Blue-Native PAGE, and computational modelling.

In nature, dual-topology membrane proteins seem to be quite rare. We hypothesize that at least some ancestors of homologous proteins with opposite orientations in the membrane, and of membrane proteins with

48

homologous, oppositely orientated halves, had dual topology. For EmrE, we have shown that substitutions that abolish the function of the dual-topology homodimer are allowed in an EmrE-derived heterodimer, providing that the ‘harmful’ substitution is present in only one of the subunits. This implies that the heterodimer is in some sense more robust than the homodimer, and at the same time that its evolutionary space has expanded. It is conceivable that a series of hypothetical mutations may lead to altered specificity and an improved function of the heterodimer that cannot occur in the homodimer without loss of function. Our system with topologically fixed heterodimers makes it possible to further investigate the roles of individual amino acid residues in the two subunits of the dimer.

49

Populärvetenskaplig sammanfattning på svenska

Levande organismer består av en eller flera celler. Cellerna innehåller bland annat arvsmassa och proteiner som katalyserar de kemiska reaktioner som cellerna behöver för att växa, frodas och föröka sig. Alla levande celler är omgivna av åtminstone ett cellmembran, ett oljigt hölje som håller ihop cellen och skyddar den. I membranet sitter en stor mängd olika membranproteiner som bildligt talat fungerar som cellernas fönster, dörrar, antenner och gripklor: tack vare membranproteinerna kan cellerna ta upp näringsämnen, utsöndra restprodukter och överhuvudtaget kommunicera med sin omvärld. Många membranproteiner är väldigt specifika och släpper bara in en typ av ämnen, t.ex. socker, medan andra skickar ut ämnen som inte behövs längre, eller som rentav är giftiga för cellen. Membranproteinerna sitter i membranen just så, att de kan släppa in specifika ämnen från utsidan, och skicka ut andra ämnen från insidan.

Vi har undersökt hur membranproteiner sätts in i membranen och vad det är som bestämmer åt vilket håll ett protein hamnar. Som alla proteiner består även membranproteiner av aminosyrakedjor, och proteinernas struktur bestäms av aminosyrakedjornas sammansättning. En liten grupp membranproteiner i den vanliga E. coli-bakterien har egenheten att de kan sättas in i cellmembranet åt två motsatta håll. Vi har visat att dessa så kallade dualtopologi-proteiner är väldigt känsliga för ändringar i aminosyrakedjan, särskilt med avseende på positivt laddade aminosyror. Det är väl känt att membranproteiner ofta sitter i membranet på det sättet att positiva laddningar hamnar inne i cellen, och genom att ändra var de positiva laddningarna sitter i proteinet kan man bestämma åt vilket håll proteinet sätts in i membranet.

De flesta membranproteiner sätts in i membranen samtidigt som de syntetiseras av ribosomer. Vi har visat att dual-topologiproteiner får sin slutgiltiga orientering i membranet först när hela proteinet har syntetiseras. Vidare föreslår vi att genom genduplikation och evolution kan dualtopologi-proteiner ge upphov till större proteiner som innehåller två motsatt orienterade, besläktade halvor.

50

Acknowledgements

I owe thanks to so many people that I cannot possibly name all of you here, simply because there is not enough space to do so. However I hope you know who you are, and how much you mean to me! Tack, alla vänner!

Here are just a few ‘special thanks’ to some of you, without whom this thesis would not exist. Först och främst vill jag såklart tacka Gunnar von Heijne - den bästa handledaren man kan tänka sig. Det har, kort sagt, varit toppen! Din schyssta inställning till forskning är klart föredömlig Stort tack till Mikaela Rapp för att jag fick haka på dina spännande projekt. Jag kommer -aldrig - att glömma alla roliga stunder på labbet (frysen är fortfarande full av konstrukt som har mycket sofistikerade och väl genomtänkta namn…;)). Jag hoppas att vi får möjlighet att jobba ihop igen! Tack Pilar Lloris-Garcerá, Joanna SG Slusky, Daniel O Daley, Erik Granseth och Frans Bianchi för gott samarbete. Many thanks for fruitful collaborations - it’s been great fun! Let’s stay in touch!

Many thanks to other present and past members of the GvH/IMN-cluster: Karin Ö, Florian, Salomé, Patricia and Nina in the ’best office’ - thanks for nice discussions, good advice, and a lot of laughs! I wish you all the best! Carmen, Nurzian, Rickard and now also Johannes - you all contribute to the great atmosphere of the lab! Thank you for all good times! Bill and Patrik, keep up the good work! Thanks to all former office/lab mates: Marie (det var kul att dela kontor med någon som fick en att skratta innan man ens hunnit dricka upp sitt morgonkaffe ), Morten a.k.a. Dr Clone (särskilt tack för att du pekade ut den gamle mossen), Roger (it was great having you in the lab… it went a bit quiet when you left…), Carolina (jag glömmer nog aldrig våra intressanta kulinariska erfarenheter i Bilbao…), Joy (always so kind and helpful!), and Katrin (hör av dig om du tar båten över nån gång! ), Yoko, Mirjam, Nadja, Tara, Andreas, Marika - many thanks to all of you for such good times Extra stort tack till IngMarie Nilsson - jag vet inte vad vi

51

skulle göra utan dig! Du förtjänar en stor guldstjärna! Thank you Jan-Willem (for scientific input), David D (great that you’re back! Good luck!), David V (m. familj för barnstolen som står i pentryt - den är väl använd ), Louise, Anna, Sam, Dimitra, Susan, Mirjam - you guys in JWdG’s group know how to throw a pub/fika - we’ve shared many good cakes Kalle (tack inte minst för avhandlings-mallen…), Filippa (för allt du lärt mig om membranprotein-komplex!), Stephen, Jörg, Isolde (som tillsammans med Ann-Louise och Erika delar plats som bästa studenten nånsin ), Rob (have fun in the ’new’ lab! ), Johan, Mili, Minttu, and everyone else in the ’GvH/deGier/IMN/Daley/Daniels-community’. It’s been a blast! Thanks and good luck to all of you!

Many thanks to everyone else at DBB for making the department such an enjoyable place! You are an inspiration. Särskilt tack till Åke, Inger, Elzbieta, Peter B, Pia Ä, Pia H, Mikael, Agneta, Andreas, Astrid, Lena och alla i era grupper, inte minst för att ni tar er tid att svara på mina frågor Tusen tack till Stefan Nordlund för ditt enorma engagemang i doktorandernas vardag. Tack till Ann, Maria och Lotta på sekretariatet - ni rockar! Tack till Håkan för att du fixar med praktiska saker och Torbjörn och Peter N för all hjälp med e-post/datorer/skrivare. Bogos R.I.P.

Sist men inte minst: tack Martin för hjälp med molekylgrafiken, och för allt annat också. Du och Eunike är det bästa jag vet! Puss!

52

References

1. Wallin E, von Heijne G (1998): Genome-wide analysis of integral membrane proteins from eubacterial, archaean, and eukaryotic organisms. Protein Sci 7, 1029-1038

2. Krogh A, Larsson B, von Heijne G, Sonnhammer EL (2001): Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J Mol Biol 305, 567-580

3. Fagerberg L, Jonasson K, von Heijne G, Uhlen M, Berglund L (2010): Prediction of the human membrane proteome. Proteomics 10, 1141-1149

4. Hopkins AL, Groom CR (2002): The druggable genome. Nat Rev Drug Discov 1, 727-730

5. Overington JP, Al-Lazikani B, Hopkins AL (2006): How many drug targets are there? Nat Rev Drug Discov 5, 993-996

6. Granseth E, Seppala S, Rapp M, Daley DO, von Heijne G (2007): Membrane protein structural biology--how far can the bugs take us? Mol Membr Biol 24, 329-332

7. Schlegel S, Klepsch M, Gialama D, Wickstrom D, Slotboom DJ, de Gier JW (2010): Revolutionizing membrane protein overexpression in bacteria. Microb Biotechnol 3, 403-411

8. Touchon M, Hoede C, Tenaillon O, Barbe V, Baeriswyl S, Bidet P, Bingen E, Bonacorsi S, Bouchier C, Bouvet O, Calteau A, Chiapello H, Clermont O, Cruveiller S, Danchin A, Diard M, Dossat C, Karoui ME, Frapy E, Garry L, Ghigo JM, Gilles AM, Johnson J, Le Bouguenec C, Lescat M, Mangenot S, Martinez-Jehanne V, Matic I, Nassif X, Oztas S, Petit MA, Pichon C, Rouy Z, Ruf CS, Schneider D, Tourret J, Vacherie B, Vallenet D, Medigue C, Rocha EP, Denamur E (2009): Organised genome dynamics in the Escherichia coli species results in highly diverse adaptive paths. PLoS Genet 5, e1000344

9. Lukjancenko O, Wassenaar TM, Ussery DW (2010): Comparison of 61 sequenced Escherichia coli genomes. Microb Ecol 60, 708-720

10. Johnson TJ, Nolan LK (2009): Pathogenomics of the virulence plasmids of Escherichia coli. Microbiol Mol Biol Rev 73, 750-774

11. Lawrence JG, Ochman H (1998): Molecular archaeology of the Escherichia coli genome. Proc Natl Acad Sci U S A 95, 9413-9417

12. Lederberg J (2004): E. coli K-12. Microbiology Today 31, 116 13. Daegelen P, Studier FW, Lenski RE, Cure S, Kim JF (2009): Tracing

ancestors and relatives of Escherichia coli B, and the derivation of B strains REL606 and BL21(DE3). J Mol Biol 394, 634-643

14. Studier FW, Daegelen P, Lenski RE, Maslov S, Kim JF (2009): Understanding the differences between genome sequences of Escherichia coli B strains REL606 and BL21(DE3) and comparison of the E. coli B and K-12 genomes. J Mol Biol 394, 653-680

53

15. Studier FW, Moffatt BA (1986): Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J Mol Biol 189, 113-130

16. Blattner FR, Plunkett G, Bloch CA, Perna NT, Burland V, Riley M, Collado-Vides J, Glasner JD, Rode CK, Mayhew GF, Gregor J, Davis NW, Kirkpatrick HA, Goeden MA, Rose DJ, Mau B, Shao Y (1997): The complete genome sequence of Escherichia coli K-12. Science 277, 1453-1462

17. Jeong H, Barbe V, Lee CH, Vallenet D, Yu DS, Choi SH, Couloux A, Lee SW, Yoon SH, Cattolico L, Hur CG, Park HS, Segurens B, Kim SC, Oh TK, Lenski RE, Studier FW, Daegelen P, Kim JF (2009): Genome sequences of Escherichia coli B strains REL606 and BL21(DE3). J Mol Biol 394, 644-652

18. Jansson PE, Lindberg AA, Lindberg B, Wollin R (1981): Structural studies on the hexose region of the core in lipopolysaccharides from Enterobacteriaceae. Eur J Biochem 115, 571-577

19. Vollmer W, Bertsche U (2008): Murein (peptidoglycan) structure, architecture and biosynthesis in Escherichia coli. Biochim Biophys Acta 1778, 1714-1734

20. Singer SJ, Nicolson GL (1972): The fluid mosaic model of the structure of cell membranes. Science 175, 720-731

21. Engelman DM (2005): Membranes are more mosaic than fluid. Nature 438, 578-580

22. Guidotti G (1972): Membrane proteins. Annu Rev Biochem 41, 731-752 23. Zimmerberg J, Kozlov MM (2006): How proteins produce cellular

membrane curvature. Nat Rev Mol Cell Biol 7, 9-19 24. Dowhan W, Bogdanov M (2009): Lipid-dependent membrane protein

topogenesis. Annu Rev Biochem 78, 515-540 25. Dowhan W (1997): Molecular basis for membrane phospholipid diversity:

why are there so many lipids? Annu Rev Biochem 66, 199-232 26. van Meer G, Voelker DR, Feigenson GW (2008): Membrane lipids: where

they are and how they behave. Nat Rev Mol Cell Biol 9, 112-124 27. Wiener MC, White SH (1992): Structure of a fluid

dioleoylphosphatidylcholine bilayer determined by joint refinement of x-ray and neutron diffraction data. III. Complete structure. Biophys J 61, 434-447

28. Marr AG, Ingraham JL (1962): Effect of temperature on the composition of fatty acids in Escherichia coli. J Bacteriol 84, 1260-1267

29. Wikstrom M, Kelly AA, Georgiev A, Eriksson HM, Klement MR, Bogdanov M, Dowhan W, Wieslander A (2009): Lipid-engineered Escherichia coli membranes reveal critical lipid headgroup size for protein function. J Biol Chem 284, 954-965

30. Lee AG (2011): Lipid-protein interactions. Biochem Soc Trans 39, 761-766 31. Dowhan W, Bogdanov M (2011): Lipid-protein interactions as

determinants of membrane protein structure and function. Biochem Soc Trans 39, 767-774

32. Wang X, Quinn PJ (2010): Lipopolysaccharide: Biosynthetic pathway and structure modification. Prog Lipid Res 49, 97-107

54

33. Milkman R, Jaeger E, McBride RD (2003): Molecular evolution of the Escherichia coli chromosome. VI. Two regions of high effective recombination. Genetics 163, 475-483

34. Herrera G, Martinez A, Blanco M, O'Connor JE (2002): Assessment of Escherichia coli B with enhanced permeability to fluorochromes for flow cytometric assays of bacterial cell function. Cytometry 49, 62-69

35. Herrera G, Urios A, Aleixandre V, Blanco M (1993): Mutability by polycyclic hydrocarbons is improved in derivatives of Escherichia coli WP2 uvrA with increased permeability. Mutat Res 301, 1-5

36. Jernaes MW, Steen HB (1994): Staining of Escherichia coli for flow cytometry: influx and efflux of ethidium bromide. Cytometry 17, 302-309

37. Raetz CR, Dowhan W (1990): Biosynthesis and function of phospholipids in Escherichia coli. J Biol Chem 265, 1235-1238

38. Shibuya I (1992): Metabolic regulations and biological functions of phospholipids in Escherichia coli. Prog Lipid Res 31, 245-299

39. DeChavigny A, Heacock PN, Dowhan W (1991): Sequence and inactivation of the pss gene of Escherichia coli. Phosphatidylethanolamine may not be essential for cell viability. J Biol Chem 266, 5323-5332

40. Killian JA, Koorengevel MC, Bouwstra JA, Gooris G, Dowhan W, de Kruijff B (1994): Effect of divalent cations on lipid organization of cardiolipin isolated from Escherichia coli strain AH930. Biochim Biophys Acta 1189, 225-232

41. Wikstrom M, Xie J, Bogdanov M, Mileykovskaya E, Heacock P, Wieslander A, Dowhan W (2004): Monoglucosyldiacylglycerol, a foreign lipid, can substitute for phosphatidylethanolamine in essential membrane-associated functions in Escherichia coli. J Biol Chem 279, 10484-10493

42. Stenberg F, Chovanec P, Maslen SL, Robinson CV, Ilag LL, von Heijne G, Daley DO (2005): Protein complexes of the Escherichia coli cell envelope. J Biol Chem 280, 34409-34419

43. Maddalo G, Stenberg-Bruzell F, Gotzke H, Toddo S, Bjorkholm P, Eriksson H, Chovanec P, Genevaux P, Lehtio J, Ilag LL, Daley DO (2011): Systematic analysis of native membrane protein complexes in Escherichia coli. J Proteome Res 10, 1848-1859

44. Daley DO (2008): The assembly of membrane proteins into complexes. Curr Opin Struct Biol 18, 420-424

45. Dalbey RE, Wang P, Kuhn A (2011): Assembly of bacterial inner membrane proteins. Annu Rev Biochem 80, 161-187

46. Fujiki Y, Hubbard AL, Fowler S, Lazarow PB (1982): Isolation of intracellular membranes by means of sodium carbonate treatment: application to endoplasmic reticulum. J Cell Biol 93, 97-102

47. Blobel G (1980): Intracellular protein topogenesis. Proc Natl Acad Sci U S A 77, 1496-1500

48. White SH, Wimley WC (1999): Membrane protein folding and stability: physical principles. Annu Rev Biophys Biomol Struct 28, 319-365

49. Popot JL, Engelman DM (2000): Helical membrane protein folding, stability, and evolution. Annu Rev Biochem 69, 881-922

50. Bigelow HR, Petrey DS, Liu J, Przybylski D, Rost B (2004): Predicting transmembrane beta-barrels in proteomes. Nucleic Acids Res 32, 2566-2577

55

51. Wimley WC (2002): Toward genomic identification of beta-barrel membrane proteins: composition and architecture of known structures. Protein Sci 11, 301-312

52. Cowan SW, Schirmer T, Rummel G, Steiert M, Ghosh R, Pauptit RA, Jansonius JN, Rosenbusch JP (1992): Crystal structures explain functional properties of two E. coli porins. Nature 358, 727-733

53. Basle A, Rummel G, Storici P, Rosenbusch JP, Schirmer T (2006): Crystal structure of osmoporin OmpC from E. coli at 2.0 A. J Mol Biol 362, 933-942

54. Vogt J, Schulz GE (1999): The structure of the outer membrane protein OmpX from Escherichia coli reveals possible mechanisms of virulence. Structure 7, 1301-1309

55. Pautsch A, Schulz GE (1998): Structure of the outer membrane protein A transmembrane domain. Nat Struct Biol 5, 1013-1017

56. Schagger H (2002): Respiratory chain supercomplexes of mitochondria and bacteria. Biochim Biophys Acta 1555, 154-159

57. Daley DO, Rapp M, Granseth E, Melen K, Drew D, von Heijne G (2005): Global topology analysis of the Escherichia coli inner membrane proteome. Science 308, 1321-1323

58. Kim H, Melen K, Osterberg M, von Heijne G (2006): A global topology map of the Saccharomyces cerevisiae membrane proteome. Proc Natl Acad Sci U S A 103, 11142-11147

59. Fredriksson R, Lagerstrom MC, Lundin LG, Schioth HB (2003): The G-protein-coupled receptors in the human genome form five main families. Phylogenetic analysis, paralogon groups, and fingerprints. Mol Pharmacol 63, 1256-1272

60. Wimley WC (2003): The versatile beta-barrel membrane protein. Curr Opin Struct Biol 13, 404-411

61. Murakami S, Nakashima R, Yamashita E, Yamaguchi A (2002): Crystal structure of bacterial multidrug efflux transporter AcrB. Nature 419, 587-593

62. Seeger MA, Schiefner A, Eicher T, Verrey F, Diederichs K, Pos KM (2006): Structural asymmetry of AcrB trimer suggests a peristaltic pump mechanism. Science 313, 1295-1298

63. Dong C, Beis K, Nesper J, Brunkan-Lamontagne AL, Clarke BR, Whitfield C, Naismith JH (2006): Wza the translocon for E. coli capsular polysaccharides defines a new class of membrane protein. Nature 444, 226-229

64. Knowles TJ, Scott-Tucker A, Overduin M, Henderson IR (2009): Membrane protein architects: the role of the BAM complex in outer membrane protein assembly. Nat Rev Microbiol 7, 206-214

65. Bernsel A, Daley DO (2009): Exploring the inner membrane proteome of Escherichia coli: which proteins are eluding detection and why? Trends Microbiol 17, 444-449

66. Lippa AM, Goulian M (2009): Feedback inhibition in the PhoQ/PhoP signaling system by a membrane peptide. PLoS Genet 5, e1000788

67. Gassel M, Mollenkamp T, Puppe W, Altendorf K (1999): The KdpF subunit is part of the K(+)-translocating Kdp complex of Escherichia coli

56

and is responsible for stabilization of the complex in vitro. J Biol Chem 274, 37901-37907

68. Fontaine F, Fuchs RT, Storz G (2011): Membrane localization of small proteins in Escherichia coli. J Biol Chem 286, 32464-32474

69. Heijne G (1986): The distribution of positively charged residues in bacterial inner membrane proteins correlates with the trans-membrane topology. EMBO J 5, 3021-3027

70. von Heijne G (1989): Control of topology and mode of assembly of a polytopic membrane protein by positively charged residues. Nature 341, 456-458

71. Melen K, Krogh A, von Heijne G (2003): Reliability measures for membrane protein topology prediction algorithms. J Mol Biol 327, 735-744

72. Bernsel A, Viklund H, Falk J, Lindahl E, von Heijne G, Elofsson A (2008): Prediction of membrane-protein topology from first principles. Proc Natl Acad Sci U S A 105, 7177-7181

73. Henderson R, Unwin PN (1975): Three-dimensional model of purple membrane obtained by electron microscopy. Nature 257, 28-32

74. Deisenhofer J, Epp O, Miki R, Huber R, Michel H (1985): Structure of the protein subunits in the photosynthetic reaction centre of Rhodopseudomonas viridis at 3Å resolution. Nature 318, 618-624

75. Berman HM, Battistuz T, Bhat TN, Bluhm WF, Bourne PE, Burkhardt K, Feng Z, Gilliland GL, Iype L, Jain S, Fagan P, Marvin J, Padilla D, Ravichandran V, Schneider B, Thanki N, Weissig H, Westbrook JD, Zardecki C (2002): The Protein Data Bank. Acta Crystallogr D Biol Crystallogr 58, 899-907

76. White SH (2009): Biophysical dissection of membrane proteins. Nature 459, 344-346

77. Wagner S, Klepsch MM, Schlegel S, Appel A, Draheim R, Tarry M, Hogbom M, van Wijk KJ, Slotboom DJ, Persson JO, de Gier JW (2008): Tuning Escherichia coli for membrane protein overexpression. Proc Natl Acad Sci U S A 105, 14371-14376

78. Bill RM, Henderson PJ, Iwata S, Kunji ER, Michel H, Neutze R, Newstead S, Poolman B, Tate CG, Vogel H (2011): Overcoming barriers to membrane protein structure determination. Nat Biotechnol 29, 335-340

79. von Heijne G (2006): Membrane-protein topology. Nat Rev Mol Cell Biol 7, 909-918

80. Vinothkumar KR, Henderson R (2010): Structures of membrane proteins. Q Rev Biophys 43, 65-158

81. Kuan J, Saier MH, Jr. (1993): The mitochondrial carrier family of transport proteins: structural, functional, and evolutionary relationships. Crit Rev Biochem Mol Biol 28, 209-233

82. Lolkema JS, Slotboom DJ (2003): Classification of 29 families of secondary transport proteins into a single structural class using hydropathy profile analysis. J Mol Biol 327, 901-909

83. Saier MH, Jr. (2003): Tracing pathways of transport protein evolution. Mol Microbiol 48, 1145-1156

84. Shimizu T, Mitsuke H, Noto K, Arai M (2004): Internal gene duplication in the evolution of prokaryotic transmembrane proteins. J Mol Biol 339, 1-15

57

85. Lolkema JS, Dobrowolski A, Slotboom DJ (2008): Evolution of antiparallel two-domain membrane proteins: tracing multiple gene duplication events in the DUF606 family. J Mol Biol 378, 596-606

86. Hennerdal A, Falk J, Lindahl E, Elofsson A (2010): Internal duplications in alpha-helical membrane protein topologies are common but the nonduplicated forms are rare. Protein Sci 19, 2305-2318

87. Lam VH, Lee JH, Silverio A, Chan H, Gomolplitinant KM, Povolotsky TL, Orlova E, Sun EI, Welliver CH, Saier MH, Jr. (2011): Pathways of transport protein evolution: recent advances. Biol Chem 392, 5-12

88. Choi S, Jeon J, Yang JS, Kim S (2008): Common occurrence of internal repeat symmetry in membrane proteins. Proteins 71, 68-80

89. Oberai A, Ihm Y, Kim S, Bowie JU (2006): A limited universe of membrane protein families and folds. Protein Sci 15, 1723-1734

90. Khademi S, O'Connell J, Remis J, Robles-Colmenares Y, Miercke LJ, Stroud RM (2004): Mechanism of ammonia transport by Amt/MEP/Rh: structure of AmtB at 1.35 A. Science 305, 1587-1594

91. Pebay-Peyroula E, Dahout-Gonzalez C, Kahn R, Trezeguet V, Lauquin GJ, Brandolin G (2003): Structure of mitochondrial ADP/ATP carrier in complex with carboxyatractyloside. Nature 426, 39-44

92. Abramson J, Smirnova I, Kasho V, Verner G, Kaback HR, Iwata S (2003): Structure and mechanism of the lactose permease of Escherichia coli. Science 301, 610-615

93. Radestock S, Forrest LR (2011): The alternating-access mechanism of MFS transporters arises from inverted-topology repeats. J Mol Biol 407, 698-715

94. Murata K, Mitsuoka K, Hirai T, Walz T, Agre P, Heymann JB, Engel A, Fujiyoshi Y (2000): Structural determinants of water permeation through aquaporin-1. Nature 407, 599-605

95. Dutzler R, Campbell EB, Cadene M, Chait BT, MacKinnon R (2002): X-ray structure of a ClC chloride channel at 3.0 A reveals the molecular basis of anion selectivity. Nature 415, 287-294

96. Pornillos O, Chang G (2006): Inverted repeat domains in membrane proteins. FEBS Lett 580, 358-362

97. Chen YJ, Pornillos O, Lieu S, Ma C, Chen AP, Chang G (2007): X-ray structure of EmrE supports dual topology model. Proc Natl Acad Sci U S A 104, 18999-19004

98. Tate CG, Kunji ER, Lebendiker M, Schuldiner S (2001): The projection structure of EmrE, a proton-linked multidrug transporter from Escherichia coli, at 7 A resolution. EMBO J 20, 77-81

99. Ubarretxena-Belandia I, Baldwin JM, Schuldiner S, Tate CG (2003): Three-dimensional structure of the bacterial multidrug transporter EmrE shows it is an asymmetric homodimer. EMBO J 22, 6175-6181

100. Wang X, Bogdanov M, Dowhan W (2002): Topology of polytopic membrane protein subdomains is dictated by membrane phospholipid composition. EMBO J 21, 5673-5681

101. Bogdanov M, Heacock PN, Dowhan W (2002): A polytopic membrane protein displays a reversible topology dependent on membrane lipid composition. EMBO J 21, 2107-2116

102. Zhang W, Bogdanov M, Pi J, Pittard AJ, Dowhan W (2003): Reversible topological organization within a polytopic membrane protein is governed

58

by a change in membrane phospholipid composition. J Biol Chem 278, 50128-50135

103. Zhang W, Campbell HA, King SC, Dowhan W (2005): Phospholipids as determinants of membrane protein topology. Phosphatidylethanolamine is required for the proper topological organization of the gamma-aminobutyric acid permease (GabP) of Escherichia coli. J Biol Chem 280, 26032-26038

104. Vitrac H, Bogdanov M, Heacock P, Dowhan W (2011): Lipids and topological rules of membrane protein assembly: balance between long and short range lipid-protein interactions. J Biol Chem 286, 15182-15194

105. Kolbusz MA, ter Horst R, Slotboom DJ, Lolkema JS (2010): Orientation of small multidrug resistance transporter subunits in the membrane: correlation with the positive-inside rule. J Mol Biol 402, 127-138

106. Tate CG, Ubarretxena-Belandia I, Baldwin JM (2003): Conformational changes in the multidrug transporter EmrE associated with substrate binding. J Mol Biol 332, 229-242

107. Ubarretxena-Belandia I, Tate CG (2004): New insights into the structure and oligomeric state of the bacterial multidrug transporter EmrE: an unusual asymmetric homo-dimer. FEBS Lett 564, 234-238

108. Korkhov VM, Tate CG (2008): Electron crystallography reveals plasticity within the drug binding site of the small multidrug transporter EmrE. J Mol Biol 377, 1094-1103

109. Nasie I, Steiner-Mordoch S, Gold A, Schuldiner S (2010): Topologically random insertion of EmrE supports a pathway for evolution of inverted repeats in ion-coupled transporters. J Biol Chem 285, 15234-15244

110. Nara T, Kouyama T, Kurata Y, Kikukawa T, Miyauchi S, Kamo N (2007): Anti-parallel membrane topology of a homo-dimeric multidrug transporter, EmrE. J Biochem 142, 621-625

111. Yerushalmi H, Schuldiner S (2000): A common binding site for substrates and protons in EmrE, an ion-coupled multidrug transporter. FEBS Lett 476, 93-97

112. Muth TR, Schuldiner S (2000): A membrane-embedded glutamate is required for ligand binding to the multidrug transporter EmrE. EMBO J 19, 234-240

113. Gutman N, Steiner-Mordoch S, Schuldiner S (2003): An amino acid cluster around the essential Glu-14 is part of the substrate- and proton-binding domain of EmrE, a multidrug transporter from Escherichia coli. J Biol Chem 278, 16082-16087

114. Fleishman SJ, Harrington SE, Enosh A, Halperin D, Tate CG, Ben-Tal N (2006): Quasi-symmetry in the cryo-EM structure of EmrE provides the key to modeling its transmembrane domain. J Mol Biol 364, 54-67

115. Korkhov VM, Tate CG (2009): An emerging consensus for the structure of EmrE. Acta Crystallogr D Biol Crystallogr 65, 186-192

116. Soskine M, Mark S, Tayer N, Mizrachi R, Schuldiner S (2006): On parallel and antiparallel topology of a homodimeric multidrug transporter. J Biol Chem 281, 36205-36212

117. Steiner-Mordoch S, Soskine M, Solomon D, Rotem D, Gold A, Yechieli M, Adam Y, Schuldiner S (2008): Parallel topology of genetically fused EmrE homodimers. EMBO J 27, 17-26

59

118. Schuldiner S (2007): When biochemistry meets structural biology: the cautionary tale of EmrE. Trends Biochem Sci 32, 252-258

119. Higashi K, Ishigure H, Demizu R, Uemura T, Nishino K, Yamaguchi A, Kashiwagi K, Igarashi K (2008): Identification of a spermidine excretion protein complex (MdtJI) in Escherichia coli. J Bacteriol 190, 872-878

120. Jack DL, Storms ML, Tchieu JH, Paulsen IT, Saier MH, Jr. (2000): A broad-specificity multidrug efflux pump requiring a pair of homologous SMR-type proteins. J Bacteriol 182, 2311-2313

121. Masaoka Y, Ueno Y, Morita Y, Kuroda T, Mizushima T, Tsuchiya T (2000): A two-component multidrug efflux pump, EbrAB, in Bacillus subtilis. J Bacteriol 182, 2307-2310

122. Zhang Z, Ma C, Pornillos O, Xiu X, Chang G, Saier MH, Jr. (2007): Functional characterization of the heterooligomeric EbrAB multidrug efflux transporter of Bacillus subtilis. Biochemistry 46, 5218-5225

123. Kikukawa T, Miyauchi S, Araiso T, Kamo N, Nara T (2007): Anti-parallel membrane topology of two components of EbrAB, a multidrug transporter. Biochem Biophys Res Commun 358, 1071-1075

124. Kikukawa T, Nara T, Araiso T, Miyauchi S, Kamo N (2006): Two-component bacterial multidrug transporter, EbrAB: Mutations making each component solely functional. Biochim Biophys Acta 1758, 673-679

125. Dobrowolski A, Sobczak-Elbourne I, Lolkema JS (2007): Membrane topology prediction by hydropathy profile alignment: membrane topology of the Na(+)-glutamate transporter GltS. Biochemistry 46, 2326-2332

126. Dobrowolski A, Lolkema JS (2010): Evolution of antiparallel two-domain membrane proteins. Swapping domains in the glutamate transporter GltS. Biochemistry 49, 5972-5974

127. Bowie JU (2006): Flip-flopping membrane proteins. Nat Struct Mol Biol 13, 94-96

128. Sebag JA, Hinkle PM (2007): Melanocortin-2 receptor accessory protein MRAP forms antiparallel homodimers. Proc Natl Acad Sci U S A 104, 20244-20249

129. Sebag JA, Hinkle PM (2009): Regions of melanocortin 2 (MC2) receptor accessory protein necessary for dual topology and MC2 receptor trafficking and signaling. J Biol Chem 284, 610-618

130. Dunlop J, Jones PC, Finbow ME (1995): Membrane insertion and assembly of ductin: a polytopic channel with dual orientations. EMBO J 14, 3609-3616

131. Lambert C, Prange R (2001): Dual topology of the hepatitis B virus large envelope protein: determinants influencing post-translational pre-S translocation. J Biol Chem 276, 22265-22272

132. Blobel G, Sabatini DD (1971) Ribosome-membrane interaction in eukaryotic cells. In L.A. Manson (ed.): Biomembranes. Plenum Publishing Corporation, New York. Volume 2, 193-195

133. Blobel G, Dobberstein B (1975): Transfer of proteins across membranes. I. Presence of proteolytically processed and unprocessed nascent immunoglobulin light chains on membrane-bound ribosomes of murine myeloma. J Cell Biol 67, 835-851

60

134. Emanuelsson O, Brunak S, von Heijne G, Nielsen H (2007): Locating proteins in the cell using TargetP, SignalP and related tools. Nat Protoc 2, 953-971

135. Cross BC, Sinning I, Luirink J, High S (2009): Delivering proteins for export from the cytosol. Nat Rev Mol Cell Biol 10, 255-264

136. Becker T, Gebert M, Pfanner N, van der Laan M (2009): Biogenesis of mitochondrial membrane proteins. Curr Opin Cell Biol 21, 484-493

137. Strittmatter P, Soll J, Bolter B (2010): The chloroplast protein import machinery: a review. Methods Mol Biol 619, 307-321

138. Marfori M, Mynott A, Ellis JJ, Mehdi AM, Saunders NF, Curmi PM, Forwood JK, Boden M, Kobe B (2011): Molecular basis for specificity of nuclear import and prediction of nuclear localization. Biochim Biophys Acta 1813, 1562-1577

139. Rapoport TA (2007): Protein translocation across the eukaryotic endoplasmic reticulum and bacterial plasma membranes. Nature 450, 663-669

140. Driessen AJ, Nouwen N (2008): Protein translocation across the bacterial cytoplasmic membrane. Annu Rev Biochem 77, 643-667

141. von Heijne G (1990): The signal peptide. J Membr Biol 115, 195-201 142. Hartl FU, Lecker S, Schiebel E, Hendrick JP, Wickner W (1990): The

binding cascade of SecB to SecA to SecY/E mediates preprotein targeting to the E. coli plasma membrane. Cell 63, 269-279

143. Mothes W, Prehn S, Rapoport TA (1994): Systematic probing of the environment of a translocating secretory protein during translocation through the ER membrane. EMBO J 13, 3973-3982

144. Sargent F (2007): The twin-arginine transport system: moving folded proteins across membranes. Biochem Soc Trans 35, 835-847

145. Robinson C, Matos CF, Beck D, Ren C, Lawrence J, Vasisht N, Mendel S (2011): Transport and proofreading of proteins by the twin-arginine translocation (Tat) system in bacteria. Biochim Biophys Acta 1808, 876-884

146. Poritz MA, Bernstein HD, Strub K, Zopf D, Wilhelm H, Walter P (1990): An E. coli ribonucleoprotein containing 4.5S RNA resembles mammalian signal recognition particle. Science 250, 1111-1117

147. Keenan RJ, Freymann DM, Stroud RM, Walter P (2001): The signal recognition particle. Annu Rev Biochem 70, 755-775

148. Bornemann T, Jockel J, Rodnina MV, Wintermeyer W (2008): Signal sequence-independent membrane targeting of ribosomes containing short nascent peptides within the exit tunnel. Nat Struct Mol Biol 15, 494-499

149. Berndt U, Oellerer S, Zhang Y, Johnson AE, Rospert S (2009): A signal-anchor sequence stimulates signal recognition particle binding to ribosomes from inside the exit tunnel. Proc Natl Acad Sci U S A 106, 1398-1403

150. Avdeeva ON, Myasnikov AG, Sergiev PV, Bogdanov AA, Brimacombe R, Dontsova OA (2002): Construction of the 'minimal' SRP that interacts with the translating ribosome but not with specific membrane receptors in Escherichia coli. FEBS Lett 514, 70-73

151. Raine A, Ullers R, Pavlov M, Luirink J, Wikberg JE, Ehrenberg M (2003): Targeting and insertion of heterologous membrane proteins in E. coli. Biochimie 85, 659-668

61

152. Luirink J, ten Hagen-Jongman CM, van der Weijden CC, Oudega B, High S, Dobberstein B, Kusters R (1994): An alternative protein targeting pathway in Escherichia coli: studies on the role of FtsY. EMBO J 13, 2289-2296

153. de Gier JW, Scotti PA, Saaf A, Valent QA, Kuhn A, Luirink J, von Heijne G (1998): Differential use of the signal recognition particle translocase targeting pathway for inner membrane protein assembly in Escherichia coli. Proc Natl Acad Sci U S A 95, 14646-14651

154. Aldridge C, Cain P, Robinson C (2009): Protein transport in organelles: Protein transport into and across the thylakoid membrane. FEBS J 276, 1177-1186

155. Tong J, Dolezal P, Selkrig J, Crawford S, Simpson AG, Noinaj N, Buchanan SK, Gabriel K, Lithgow T (2011): Ancestral and derived protein import pathways in the mitochondrion of Reclinomonas americana. Mol Biol Evol 28, 1581-1591

156. Van den Berg B, Clemons WM, Jr., Collinson I, Modis Y, Hartmann E, Harrison SC, Rapoport TA (2004): X-ray structure of a protein-conducting channel. Nature 427, 36-44

157. Tam PC, Maillard AP, Chan KK, Duong F (2005): Investigating the SecY plug movement at the SecYEG translocation channel. EMBO J 24, 3380-3388

158. Egea PF, Stroud RM (2010): Lateral opening of a translocon upon entry of protein suggests the mechanism of insertion into membranes. Proc Natl Acad Sci U S A 107, 17182-17187

159. Zimmer J, Nam Y, Rapoport TA (2008): Structure of a complex of the ATPase SecA and the protein-translocation channel. Nature 455, 936-943

160. du Plessis DJ, Berrelkamp G, Nouwen N, Driessen AJ (2009): The lateral gate of SecYEG opens during protein translocation. J Biol Chem 284, 15805-15814

161. Breyton C, Haase W, Rapoport TA, Kuhlbrandt W, Collinson I (2002): Three-dimensional structure of the bacterial protein-translocation complex SecYEG. Nature 418, 662-665

162. Wickstrom D, Wagner S, Baars L, Ytterberg AJ, Klepsch M, van Wijk KJ, Luirink J, de Gier JW (2011): Consequences of depletion of the signal recognition particle in Escherichia coli. J Biol Chem 286, 4598-4609

163. Baars L, Ytterberg AJ, Drew D, Wagner S, Thilo C, van Wijk KJ, de Gier JW (2006): Defining the role of the Escherichia coli chaperone SecB using comparative proteomics. J Biol Chem 281, 10024-10034

164. Baars L, Wagner S, Wickstrom D, Klepsch M, Ytterberg AJ, van Wijk KJ, de Gier JW (2008): Effects of SecE depletion on the inner and outer membrane proteomes of Escherichia coli. J Bacteriol 190, 3505-3525

165. Duong F, Wickner W (1997): Distinct catalytic roles of the SecYE, SecG and SecDFyajC subunits of preprotein translocase holoenzyme. EMBO J 16, 2756-2768

166. Economou A, Pogliano JA, Beckwith J, Oliver DB, Wickner W (1995): SecA membrane cycling at SecYEG is driven by distinct ATP binding and hydrolysis events and is regulated by SecD and SecF. Cell 83, 1171-1181

167. Nouwen N, Driessen AJ (2002): SecDFyajC forms a heterotetrameric complex with YidC. Mol Microbiol 44, 1397-1405

62

168. Tsukazaki T, Mori H, Echizen Y, Ishitani R, Fukai S, Tanaka T, Perederina A, Vassylyev DG, Kohno T, Maturana AD, Ito K, Nureki O (2011): Structure and function of a membrane component SecDF that enhances protein export. Nature 474, 235-238

169. Samuelson JC, Chen M, Jiang F, Moller I, Wiedmann M, Kuhn A, Phillips GJ, Dalbey RE (2000): YidC mediates membrane protein insertion in bacteria. Nature 406, 637-641

170. Scotti PA, Urbanus ML, Brunner J, de Gier JW, von Heijne G, van der Does C, Driessen AJ, Oudega B, Luirink J (2000): YidC, the Escherichia coli homologue of mitochondrial Oxa1p, is a component of the Sec translocase. EMBO J 19, 542-549

171. Pop OI, Soprova Z, Koningstein G, Scheffers DJ, van Ulsen P, Wickstrom D, de Gier JW, Luirink J (2009): YidC is required for the assembly of the MscL homopentameric pore. FEBS J 276, 4891-4899

172. Xie K, Dalbey RE (2008): Inserting proteins into the bacterial cytoplasmic membrane using the Sec and YidC translocases. Nat Rev Microbiol 6, 234-244

173. Yuan J, Zweers JC, van Dijl JM, Dalbey RE (2010): Protein transport across and into cell membranes in bacteria and archaea. Cell Mol Life Sci 67, 179-199

174. Funes S, Kauff F, van der Sluis EO, Ott M, Herrmann JM (2011): Evolution of YidC/Oxa1/Alb3 insertases: three independent gene duplications followed by functional specialization in bacteria, mitochondria and chloroplasts. Biol Chem 392, 13-19

175. Chen M, Samuelson JC, Jiang F, Muller M, Kuhn A, Dalbey RE (2002): Direct interaction of YidC with the Sec-independent Pf3 coat protein during its membrane protein insertion. J Biol Chem 277, 7670-7675

176. Xie K, Hessa T, Seppala S, Rapp M, von Heijne G, Dalbey RE (2007): Features of transmembrane segments that promote the lateral release from the translocase into the lipid phase. Biochemistry 46, 15153-15161

177. Wickstrom D, Wagner S, Simonsson P, Pop O, Baars L, Ytterberg AJ, van Wijk KJ, Luirink J, de Gier JW (2011): Characterization of the consequences of YidC depletion on the inner membrane proteome of E. coli using 2D blue native/SDS-PAGE. J Mol Biol 409, 124-135

178. Yu Z, Koningstein G, Pop A, Luirink J (2008): The conserved third transmembrane segment of YidC contacts nascent Escherichia coli inner membrane proteins. J Biol Chem 283, 34635-34642

179. Kohler R, Boehringer D, Greber B, Bingel-Erlenmeyer R, Collinson I, Schaffitzel C, Ban N (2009): YidC and Oxa1 form dimeric insertion pores on the translating ribosome. Mol Cell 34, 344-353

180. Yu Z, Laven M, Klepsch M, de Gier JW, Bitter W, van Ulsen P, Luirink J (2011): A role for Escherichia coli YidD in membrane protein insertion. J Bacteriol

181. Davis NG, Model P (1985): An artificial anchor domain: hydrophobicity suffices to stop transfer. Cell 41, 607-614

182. Heinrich SU, Mothes W, Brunner J, Rapoport TA (2000): The Sec61p complex mediates the integration of a membrane protein by allowing lipid partitioning of the transmembrane domain. Cell 102, 233-244

63

183. Hessa T, Meindl-Beinker NM, Bernsel A, Kim H, Sato Y, Lerch-Bader M, Nilsson I, White SH, von Heijne G (2007): Molecular code for transmembrane-helix recognition by the Sec61 translocon. Nature 450, 1026-1030

184. Hessa T, Kim H, Bihlmaier K, Lundin C, Boekel J, Andersson H, Nilsson I, White SH, von Heijne G (2005): Recognition of transmembrane helices by the endoplasmic reticulum translocon. Nature 433, 377-381

185. Lundin C, Kim H, Nilsson I, White SH, von Heijne G (2008): Molecular code for protein insertion in the endoplasmic reticulum membrane is similar for N(in)-C(out) and N(out)-C(in) transmembrane helices. Proc Natl Acad Sci U S A 105, 15702-15707

186. Ojemalm K, Higuchi T, Jiang Y, Langel U, Nilsson I, White SH, Suga H, von Heijne G (2011): Apolar surface area determines the efficiency of translocon-mediated membrane-protein integration into the endoplasmic reticulum. Proc Natl Acad Sci U S A 108, E359-364

187. Nilsson J, Persson B, von Heijne G (2005): Comparative analysis of amino acid distributions in integral membrane proteins from 107 genomes. Proteins 60, 606-616

188. Nilsson I, von Heijne G (1998): Breaking the camel's back: proline-induced turns in a model transmembrane helix. J Mol Biol 284, 1185-1189

189. Yohannan S, Faham S, Yang D, Whitelegge JP, Bowie JU (2004): The evolution of transmembrane helix kinks and the structural diversity of G protein-coupled receptors. Proc Natl Acad Sci U S A 101, 959-963

190. Russ WP, Engelman DM (2000): The GxxxG motif: a framework for transmembrane helix-helix association. J Mol Biol 296, 911-919

191. Senes A, Gerstein M, Engelman DM (2000): Statistical analysis of amino acid patterns in transmembrane helices: the GxxxG motif occurs frequently and in association with beta-branched residues at neighboring positions. J Mol Biol 296, 921-936

192. Walters RF, DeGrado WF (2006): Helix-packing motifs in membrane proteins. Proc Natl Acad Sci U S A 103, 13658-13663

193. Braun P, von Heijne G (1999): The aromatic residues Trp and Phe have different effects on the positioning of a transmembrane helix in the microsomal membrane. Biochemistry 38, 9778-9782

194. Killian JA, von Heijne G (2000): How proteins adapt to a membrane-water interface. Trends Biochem Sci 25, 429-434

195. White SH, Wimley WC (1998): Hydrophobic interactions of peptides with membrane interfaces. Biochim Biophys Acta 1376, 339-352

196. Ota K, Sakaguchi M, von Heijne G, Hamasaki N, Mihara K (1998): Forced transmembrane orientation of hydrophilic polypeptide segments in multispanning membrane proteins. Mol Cell 2, 495-503

197. Gafvelin G, von Heijne G (1994): Topological "frustration" in multispanning E. coli inner membrane proteins. Cell 77, 401-412

198. Nilsson I, von Heijne G (1990): Fine-tuning the topology of a polytopic membrane protein: role of positively and negatively charged amino acids. Cell 62, 1135-1141

199. Andersson H, Bakker E, von Heijne G (1992): Different positively charged amino acids have similar effects on the topology of a polytopic transmembrane protein in Escherichia coli. J Biol Chem 267, 1491-1495

64

200. Hartmann E, Rapoport TA, Lodish HF (1989): Predicting the orientation of eukaryotic membrane-spanning proteins. Proc Natl Acad Sci U S A 86, 5786-5790

201. Illergard K, Kauko A, Elofsson A (2011): Why are polar residues within the membrane core evolutionary conserved? Proteins 79, 79-91

202. Hessa T, White SH, von Heijne G (2005): Membrane insertion of a potassium-channel voltage sensor. Science 307, 1427

203. Jiang Y, Lee A, Chen J, Ruta V, Cadene M, Chait BT, MacKinnon R (2003): X-ray structure of a voltage-dependent K+ channel. Nature 423, 33-41

204. Nishino K, Yamaguchi A (2001): Analysis of a complete library of putative drug transporter genes in Escherichia coli. J Bacteriol 183, 5803-5812

205. Chamberlain AK, Lee Y, Kim S, Bowie JU (2004): Snorkeling preferences foster an amino acid composition bias in transmembrane helices. J Mol Biol 339, 471-479

206. Enquist K, Fransson M, Boekel C, Bengtsson I, Geiger K, Lang L, Pettersson A, Johansson S, von Heijne G, Nilsson I (2009): Membrane-integration characteristics of two ABC transporters, CFTR and P-glycoprotein. J Mol Biol 387, 1153-1164

207. Meindl-Beinker NM, Lundin C, Nilsson I, White SH, von Heijne G (2006): Asn- and Asp-mediated interactions between transmembrane helices during translocon-mediated membrane protein assembly. EMBO Rep 7, 1111-1116

208. Hedin LE, Ojemalm K, Bernsel A, Hennerdal A, Illergard K, Enquist K, Kauko A, Cristobal S, von Heijne G, Lerch-Bader M, Nilsson I, Elofsson A (2010): Membrane insertion of marginally hydrophobic transmembrane helices depends on sequence context. J Mol Biol 396, 221-229

209. Norholm MH, Shulga YV, Aoki S, Epand RM, von Heijne G (2011): Flanking residues help determine whether a hydrophobic segment adopts a monotopic or bitopic topology in the endoplasmic reticulum membrane. J Biol Chem 286, 25284-25290

210. Zhang JT, Ling V (1991): Study of membrane orientation and glycosylated extracellular loops of mouse P-glycoprotein by in vitro translation. J Biol Chem 266, 18224-18232

211. Sadlish H, Skach WR (2004): Biogenesis of CFTR and other polytopic membrane proteins: new roles for the ribosome-translocon complex. J Membr Biol 202, 115-126

212. Skach WR, Shi LB, Calayag MC, Frigeri A, Lingappa VR, Verkman AS (1994): Biogenesis and transmembrane topology of the CHIP28 water channel at the endoplasmic reticulum. J Cell Biol 125, 803-815

213. Shi LB, Skach WR, Ma T, Verkman AS (1995): Distinct biogenesis mechanisms for the water channels MIWC and CHIP28 at the endoplasmic reticulum. Biochemistry 34, 8250-8256

214. Kauko A, Hedin LE, Thebaud E, Cristobal S, Elofsson A, von Heijne G (2010): Repositioning of transmembrane alpha-helices during membrane protein folding. J Mol Biol 397, 190-201

215. Ismail N, Crawshaw SG, High S (2006): Active and passive displacement of transmembrane domains both occur during opsin biogenesis at the Sec61 translocon. J Cell Sci 119, 2826-2836

65

216. Ismail N, Crawshaw SG, Cross BC, Haagsma AC, High S (2008): Specific transmembrane segments are selectively delayed at the ER translocon during opsin biogenesis. Biochem J 411, 495-506

217. Junne T, Kocik L, Spiess M (2010): The hydrophobic core of the Sec61 translocon defines the hydrophobicity threshold for membrane integration. Mol Biol Cell 21, 1662-1670

218. Junne T, Schwede T, Goder V, Spiess M (2007): Mutations in the Sec61p channel affecting signal sequence recognition and membrane protein topology. J Biol Chem 282, 33201-33209

219. Goder V, Spiess M (2003): Molecular mechanism of signal sequence orientation in the endoplasmic reticulum. EMBO J 22, 3645-3653

220. Devaraneni PK, Conti B, Matsumura Y, Yang Z, Johnson AE, Skach WR (2011): Stepwise insertion and inversion of a Type II signal anchor sequence in the ribosome-Sec61 translocon complex. Cell 146, 134-147

221. Skach WR (2009): Cellular mechanisms of membrane protein folding. Nat Struct Mol Biol 16, 606-612

222. Bonardi F, Halza E, Walko M, Du Plessis F, Nouwen N, Feringa BL, Driessen AJ (2011): Probing the SecYEG translocation pore size with preproteins conjugated with sizable rigid spherical molecules. Proc Natl Acad Sci U S A 108, 7775-7780

223. Hamman BD, Chen JC, Johnson EE, Johnson AE (1997): The aqueous pore through the translocon has a diameter of 40-60 A during cotranslational protein translocation at the ER membrane. Cell 89, 535-544

224. Cannon KS, Or E, Clemons WM, Jr., Shibata Y, Rapoport TA (2005): Disulfide bridge formation between SecY and a translocating polypeptide localizes the translocation pore to the center of SecY. J Cell Biol 169, 219-225

225. Osborne AR, Rapoport TA (2007): Protein translocation is mediated by oligomers of the SecY complex with one SecY copy forming the channel. Cell 129, 97-110

226. Becker T, Bhushan S, Jarasch A, Armache JP, Funes S, Jossinet F, Gumbart J, Mielke T, Berninghausen O, Schulten K, Westhof E, Gilmore R, Mandon EC, Beckmann R (2009): Structure of monomeric yeast and mammalian Sec61 complexes interacting with the translating ribosome. Science 326, 1369-1373

227. van Klompenburg W, Nilsson I, von Heijne G, de Kruijff B (1997): Anionic phospholipids are determinants of membrane protein topology. EMBO J 16, 4261-4266

228. Johansson AC, Lindahl E (2009): Protein contents in biological membranes can explain abnormal solvation of charged and polar residues. Proc Natl Acad Sci U S A 106, 15684-15689

229. Bay DC, Rommens KL, Turner RJ (2008): Small multidrug resistance proteins: a multidrug transporter family that continues to grow. Biochim Biophys Acta 1778, 1814-1838

230. Schuldiner S (2009): EmrE, a model for studying evolution and mechanism of ion-coupled transporters. Biochim Biophys Acta 1794, 748-762

231. Butler PJ, Ubarretxena-Belandia I, Warne T, Tate CG (2004): The Escherichia coli multidrug transporter EmrE is a dimer in the detergent-solubilised state. J Mol Biol 340, 797-808

66

232. Rotem D, Schuldiner S (2004): EmrE, a multidrug transporter from Escherichia coli, transports monovalent and divalent substrates with the same stoichiometry. J Biol Chem 279, 48787-48793

233. Manoil C, Beckwith J (1986): A genetic approach to analyzing membrane protein topology. Science 233, 1403-1408

234. Feilmeier BJ, Iseminger G, Schroeder D, Webber H, Phillips GJ (2000): Green fluorescent protein functions as a reporter for protein localization in Escherichia coli. J Bacteriol 182, 4068-4076

235. Rapp M, Drew D, Daley DO, Nilsson J, Carvalho T, Melen K, De Gier JW, Von Heijne G (2004): Experimentally based topology models for E. coli inner membrane proteins. Protein Sci 13, 937-945

236. Drew D, Slotboom DJ, Friso G, Reda T, Genevaux P, Rapp M, Meindl-Beinker NM, Lambert W, Lerch M, Daley DO, Van Wijk KJ, Hirst J, Kunji E, De Gier JW (2005): A scalable, GFP-based pipeline for membrane protein overexpression screening and purification. Protein Sci 14, 2011-2017

237. Drew D, Lerch M, Kunji E, Slotboom DJ, de Gier JW (2006): Optimization of membrane protein overexpression and purification using GFP fusions. Nat Methods 3, 303-313

238. Studier FW, Rosenberg AH, Dunn JJ, Dubendorff JW (1990): Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol 185, 60-89

239. Studier FW (1991): Use of bacteriophage T7 lysozyme to improve an inducible T7 expression system. J Mol Biol 219, 37-44

240. Bogdanov M, Zhang W, Xie J, Dowhan W (2005): Transmembrane protein topology mapping by the substituted cysteine accessibility method (SCAM(TM)): application to lipid-specific membrane protein topogenesis. Methods 36, 148-171