Structure and function of the porin channel - CiteSeerX

Ersch. in: Kidney International ; 48 (1995), 4. - S. 930-940 http://dx.doi.org/10.1038/ki.1995.374

Structure and function of the porin channel WOLFRAM WELTE, UWE NESTEL, THOMAS WACKER,

and KAy

DIEDERICHS

Fakultiit fUr Biologie, Universitiit Konstanz, Konstanz, Germany

The outer membrane bilayer of gram-negative bacteria, which encapsulates completely the plasma membrane, seems to act as a protection system against noxious macromolecules like proteases and bile-salt micelles [1]. The external monolayer is composed of lipopolysaccharides with a variable carbohydrat.e composi~i?n. Long oligosaccharide chains protect the cells agamst recogmtlOn by the complement system [2, 3] and hydrophobic substances [4]. The other monolayer facing the peri plasmic space is mainly composed of phospholipid. This protection system should nevertheless allow free access of small biomolecules like mono- and disaccharides, nucleosides, and amino acids to the cytoplasmic membrane. The proteins allowing for rapid diffusion of these molecules across the outer membrane are called porins [5]. The "general diffusion porins" form aqueous channels with an exclusion limit of typically 600 Da and extremes of 5000 Da. Due to their high copy number they form the major integral protein component of the outer membrane and turn it into a molecular sieve. A single cell of E. coli has been estimated to possess approximately 105 OmpF porin molecules [6]. Purification, reconstitution into black lipid bilayers and electrical measurements

Porins can be isolated and purified by chromatography after solubilization or preferential extraction of the outer membrane with detergents, even with sodium dodecylsulfate [6-11]. For most of the preparations a porin trimer has been reported [8, 12-14]. The trimers insert spontaneously into black phospholipid membranes of the Montal-Mueller or Mueller-Rudin type. Their electrical properties can be studied by measurement with electrodes of the current across these porin-doped black lipid bilayer membranes [7, 15-18]. The bilayer can be made up from asymmetric LPS-phospholipid bilayers mimicking the outer membrane [19]. Patch-clamp techniques have also been used [20]. S~~ll radio labeled substances have been used to study the permeabIlIty of porins in whole cells and porin containing liposomes. Diffusion rates of permeating solutes can be studied by the swelling behavior of liposomes which are initially prepared in presence of a nonpermeable polysaccharide. After dilution into isoosmolar solutions of small permeating solutes an increase of turbidity due to the swelling is observed [21]. Trimers can be dissociated to monomers by heat- and detergent treatment [8]. Usually, the monomers are in some respect denatured, as reassociation of monomers leads to inactive trimers [8]

Konstanzer Online-Publikations-System (KOPS) URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-208907

except in the case of porin from Rhodobacter capsulatus [22] and porins from Pseudomonas aeruginosa [23]. Classification of porins according to their observed functions

The following functional characteristics and classes of porins have emerged from these studies. (1) The majority of porins belong to the class of nonspe~ific so-called "general diffusion" porins [24, 25] and have typical exclusion limits of 600 Da, with extremes of 5000 Da. Each bacterial cell seems to have one of these porins in high copy numbers in its outer membrane. They show slightly different permeabilities for cations (Pc) and anions (Pa), respectively. Pc/Pa ratios vary typically between 0.3 and 4 [25]. Porins show a voltage-induced closing, leading to a macroscopic reduction of current in 1 M KCl by a factor of between 0.85 and 0.15 [7, 14, 16, 26, 27]. The voltage at which this reduction takes place, the "gating voltage," seems to depend upon the orientation of the external field relative to the porin [28] and values of between 5 mV and 200 m V have been reported. Approximately three charges move upon gating [28]. A voltage-dependent anion channel protein (VDAC) in the mitochondrial outer membrane [29] may have evolved from these procaryotic porins. It has a higher exclusion limit (2500) than most general diffusion porins, shows voltage gating at ~ 30 m V [30] and does not seem to form trimers [31-34]. For regulatory porposes, VDAC seems to have evolved the ability to form complexes with hexokinase [35] to regulate uptake of carbohydrate, with creatine kinase [36], with a benzodiazepine receptor [37] to regulate steroidogenesis and with a water-soluble modulating protein [38]. A large-conductance anion channel in astrocytic plasma membranes is highly homologous with VDAC [39]. Porins from chloroplast outer membranes [40] and mycobacteria [41] have also been reported. (2) "Specific porins" possess a binding site which is moderately specific for a certain class of substances (dissociation constants are typically around Kd ~100 /-Lmol [24]) and are usually expressed under certain growth limitation conditions at low copy numbers. They scem to be always part of an energy-driven uptake system like a phosphotransferase (PTS) system [42,43] or a ATP binding cassette (ABC) system [44]. Specific porins have been found for mono- and disaccharides (LamB, ScrY), nucleosides (TsX) [45] and anions (protein P from Pseudomonas aeruginosa). These porins show a lower single channel conductivity than general diffusion porins which can be blocked upon addition of the substrate [46]. The transport rate of specific porins saturates in a Michaelis-Menten like manner when the binding site is occupied.

930

931 Amino acid sequences and conclusions related to them An increasing number of porin sequences is known. General diffusion porins and specific porins typically have masses of 30 to 40 kD and 30 to 50 kD, respectively. Compared with other integral membrane proteins, the amino acid compositions of porin sequences are rather polar and lack hydrophobic segments long enough to span membrane bilayers [47]. In accord with their rather hydrophilic amino acid sequence, porins are exported to the periplasmic space as water soluble precursors [48, 49] with a signal sequence like other periplasmic proteins. The removal of the signal sequence preceeds trimerization [50]. It is conceivable that the removal of the signal sequence allows the formation of a hydrophobic core by N- and C-termini of three porin molecules. The energy of formation of this core may stabilize the trimer with its membrane exposed hydrophobic surface against the watersoluble conformation and thus trigger the insertion into the outer membrane. Other porins or porin-like molecules like OmpA or VDAC, which have N- or C-terminal extensions, thus should not form trimers and might be able to exist both in the water-soluble and the membrane bound conformation. When sequences are compared no global similarity is found, not even within one of the two porin subclasses. Only porins within one subclass and from phylogenetically related organisms show sequence similarities. Several classifications of porins have been proposed. Schiltz et al [51] suggest 10 groups, leanteur, Lakey and Pattus [52] propose a phylogenetic tree for porins. As there is no universal sequence classification for porins yet, general phylogenetic relations of bacteria should be taken into regard as was done by Nikaido [53]. Tommassen [54] compared the rather homologous sequences of the three non-specific porins OmpF, OmpC and PhoE from Escherichia coli and found that there are eight zones within which the sequences are most variable, separated by rather well conserved stretches of some 40 residues. The variable regions coincided with the maxima of hydrophilicity. One of the variable regions was known to be surface exposed [54, 55]. Using monoclonal antibodies against the other variable stretches, Tommassen could show that all except the third form surface exposed epitopes. Infrared studies had previously demonstrated that porins contain a high degree of antiparallel /3-pleated structure with strands of average length of 11 residues oriented roughly parallel to the membrane normal [56, 57]. With these data, Tommassen proposed that the polypeptide forms /3-strands crossing the membrane 16 times in an antiparallel manner which are connected on the surface by the eight variable loop regions. On the peri plasmic side are the N- and C-termini and seven hairpin loops connecting neighboring strands. The same analysis was applied to class 1 to 3 meningococcal and to PIA and PIB gonococcal porins, and indicated a similar structure for these porins as well [58]. This model coincides remarkably well with the structures found lately by X-ray crystallography. It also gives hints how the sequence can be used to predict the folding of porins of unknown structure in absence of homology with porins of known structure [59]. For VDAC, models consisting of a N-terminal helix and 19, 16 or 12 anti parallel transmembrane /3-strands have been proposed [6062]. Structural studies Matrix (OmpF) porin was the first integral membrane protein of which single crystals suitable for X-ray diffraction could be

obtained [63, 64]. Garavito et al also obtained crystals of Maltoporin (LamB) and OmpA from E. coli [65]. The former could be improved by Stauffer et al [66]. Crystals of phosphoporin (PhoE) from E. coli have also been obtained [67]. As the phase determination by multiple isomorphous replacement (MIR) was delayed, electron microscopic work using two-dimensional crystalline sheets [68] was done in parallel. Engel et al [69] and lap, Walian and Gehring [70] used image filtering and three-dimensional reconstruction techniques of stained and unstained, respectively, two dimensional crystals to obtain the structure at a resolution of 28 A a~d 6 A, res\?ectively. The resolution of projection maps was even hIgher (3.2 A) [71]. A large /3-barrel forming the surface of the protein in contact with the membrane lipid phase was inferred from these studies. The appearance of porin trimers in two dimensional crystals is so characteristic that in some bacteria the existence of porins has been inferred from electron micrographs of whole cells [72, 73]. A moderately cation selective porin has been purified from the facultatively phototrophic purple bacterium Rhodobacter capsulatus [74]. The crystals of this porin [75] could be improved to diffract to 1.8 A resolution [7] and were analyzed with X-ray crystallography [76-79]. In the meanwhile, a new trigonal crystal form of OmpF [80] allowed the solution of the structure of this E. coli porin and of the related phosphoporin PhoE as well [81]. Another porin from the purple bacterium Rhodopseudomonas blastica was crystallized [8] and the structure solved to 2 A resolution [82]. The structure analyses of two specific porins, maltoporin (LamB) [66] and sucrose porin (ScrY) [83] are underway. Structure of bacterial porins as determined by X-ray crystallography General description

Figure 1 shows the polypeptide chain fold as a ribbon diagram of porin from Rhodobacter capsulatus, porin from Rhodopseudomonas blastica and OmpF and Phoe from Escherichia coli, respectively. The common feature is a transmembrane /3-barrel of approximately 30 A in diameter. It is formed by 16 antiparallel /3-strands which are connected to their next neighbors. The N- and C-termini are located near the periplasmic side [54, 59, 84]. The strands are inclined to the barrel axis on average by 4SO [76]. The connections between the strands are of the short hair-pin type on the periplasmic side and longer and more irregular on the external side. The latter account for the hydrophilicity peaks and variable sequence zones which Tommassen had observed. The three porin structures show that this sequence variability goes parallel with a structural variability in these loops and in the contour of the upper barrel rim. One extreme is represented by the porin from Rhodobacter capsulatus 37b4. Near the trimer axis and the termini the barrel wall height is about 20 A. At the opposite circumference, which forms the periphery of the trimer, it is ca. 50 A. The external end of the trimer, as a consequence, forms a single channel of about 50 A internal diameter which ends at its bottom in three narrow constriction sites (eyelets) near the bilayer center. Beyond each eyelet one widening channel formed by one porin molecule runs further to the periplasmic surface. In the other porin structures this fusion of three molecular channels to one trimer channel is much less obvious, so a monomeric porin channel would also appear to be conceivable. 0

932 A

B

Fig. 1. Ribbon diagrams showing the polypeptide chain fold and secondary structure of (A) porin from Rhodobactcr capsulatus 37b4, (B) porin from Rhodopseudomonas blastica and (e) OmpF and PhoE from Escherichia

coli. The view is from the trimer axis towards the eyelet. This and the following figures were prepared with the program Molscript [121] on Silicon Graphics workstations.

933

Fig. 2. Ribbon diagram of the trimer of porin from Rhodobacter capsulatus 37b4. Several hydrophobic amino acids and the bound Ca 2 + atom in the molecular interface (black ball) as well as two bound Ca 2 + atoms at aspartate and glutamate residues of loop L3 are shown. The view is from the external medium.

In OmpF, some of the external loops tend to incline towards the trimer axis to constrict the external channel entrance. The differences in the external loop structure between Enterobacteria and Purple bacteria could have evolved due to the different environmental risks to which these cells are exposed in their habitats. In contrast to the other external loops, the third loop L3 of ca 45 residues folds in all porins into the barrel along the wall approximately in the center of the bilayer and then back to the sixth strand. Its interactions with peripheral barrel strands may contribute to the stability and form of the barrel [81]. The molecular inteiface Figure 2 shows the trimer center of porin from Rhodobacter capsulatus. In the trimers of all porins, the N- and C-termini are packed tightly together into the hydrophobic intermolecular contact interface near the trimer axis. Its tight packing is indicated by low temperature factors around 20 A2 [78, 82]. Its energy of formation may stabilize trimerization and membrane incorporation [78]. This could be triggered by the signal sequence cleavage [50]. In case of the porins from Rhodobacter capsulatus and Rhodopseudomonas blastica, N- and C-termini of neighboring porin molecules contribute to the interaction by forming salt bridges. In the former porin a Ca 2 + -binding site, formed by adjacent molecules, also contributes to trimerization (Fig. 2) [77-79]. In the two Escherichia coli porins one of the external loops near the trimer axis folds to the barrel rim of a neighboring molecule and seems to connect both. These features of the interface explain the extraordinary stability of porin trimers which can be dissociated only upon boiling in sodium dodecylsulfate and

is in accord with what was said above about porin export and membrane insertion. The membrane embedded suiface Figure 3 shows a ribbon diagram of the porin from Rhodobacter capsulatus 37b4, as seen from the external surface. All porin structures show two rings of aromatic acids around the trimer peripheral surface, that is, outside the intermolecular contact surfaces. One ring is close to the smooth periplasmic bottom rim. Phenylalanines and occasionally tryptophanes with their aromatic rings always pointing towards the hydrophobic phase and tyrosines with their hydroxyl groups always pointing towards the aqueous phase are found on the first strand position which follows or precedes one of the hair-pin loops. At a distance of ca. 15 A towards the external surface, a second, parallel ring is formed by tyrosines and tryptophanes. In between, only nonpolar residues like Leu, Val, Ala and lIe are found. The resulting hydrophobic zone has an axial height of ca 25 A, approximately sufficient to span the 30 A thick hydrocarbon layer of lipid bilayers [85, 86]. As the thickness of the hydrophobic layer may change due to gel-to-liquid phase transition, which depends upon parameters as temperature, pH, divalent ion concentration and temperature [87] and as the protein may rock upon thermal movements, the position of the hydrophobic-hydrophilic interfaces on the protein surface may change. Porins thus may have to accommodate to such fluctuations. The aromatic rings may confer this adaptability. Two mechanisms have been proposed: the polarizability of the 7T-electron system allowing interaction with polar- and apolar solvents [88, 89] and rotations of the aromatic side chains relative

934

Fig. 3. Ribbon diagram of porin from Rhodobacter capsulatus 37b4 with aromatic amino acid residues. The membrane-embedded surface with the two rings of aromatic residues is shown.

to the backbone [82]. These well-conserved aromatic amino acid residues can be used for structure predictions of porins [59] and are found less clear also in reaction centers [90, 91] and bacteriorhodopsin complexes [92]. Beyond the external aromatic ring the peripheral barrel surface bears some carboxyl groups, which could interact with lipopolysaccharide molecules through divalent cations [77, 81]. The constriction zone In all known porin structures, the space filled by the loop L3 along its path inside the barrel wall effectively reduces the free cross section in the center of the membrane to a constriction site or eyelet of, typically, 8 A by 10 A. Figure 4A shows a 4 A slab from Rhodobacter capsulatus 37b4 porin close to a plane spanned by the trimer (membrane normal) and a line in the membrane plane from the trimer axis to the center of the constriction site. The constriction of the barrel by the loop L3 is obvious. The surface in this channel segment is studded with charged amino acids and tyrosines. In porin from Rhodobacter capsulatus 37b4, there are 7 Asp and 4 Glu contributed by L3 and 2 Lys, 3 Arg and 1 His eontributed by the opposite barrel wall which is adjacent to the trimer center. In the two Escherichia coli porins the external loops L4 and L5 are inclined towards the barrel axis, so that their charged amino acids form a further constriction near the channel entrance as seen in Figure 4C. L3 constricts the barrel cross section to an extent that hydrated

ions cannot pass it freely. Figure 5 shows the constriction site of the Rhodobacter capsulatus 37b4 porin with its charged aminoacid residues superimposed with the diameter of a hydrated K-t ion. The diameter was calculated using the known number of bound water molecules and the volume of a water molecule [93]. As porins allow fast movements of ions across the membrane with the pore conductance beeing proportional to the bulk aqueous conductivity for many electrolytes [15], there must be a compensation of the energy penalty of transiently stripping some of the hydration water. Various evidences exist that interaction of cations with acidic and anions with basic residues provide this energy. The ion selectivity is pH dependent [94, 95]. Furthermore, porins with acetylated or carbamylated lysines show increased cation selectivity [95, 96]. If hybrid porins of a cation- and a anionselective one are made, anion selectivity changes into cation selectivity upon exchange of a lysine by an aspartate [97]. Exchange of severallysines for glutamates in PhoE each resulted in a cation selective channel [98]. When the constriction sites of OmpF (moderately cation selective) and PhoE (moderately anion selective) are compared, the latter is found to possess two additional lysine residues in the central constriction zone [81]. The exact mechanism of ion selectivity, the role which single charged residues play, will probably require further theoretical analyses of the known structures. The constriction zone also determines the size exclusion limit as was concluded [81] from mutants of Ompe with an increased

935 A

B

•

c

Fig. 4. CPK slabs of between 0.6 and 5 A thickness cut along a plane through the trimer axis and the center of the eyelet for (A) porin from Rhodobacter capsulatus 37b4, (B) porin from Rhodopseudomonas blastica and (C) OmpF and PhoE from Escherichia coli. Backbone atoms are drawn as open circles (0), side-chain atoms as filled circles (e). As a visual aid, the shape of the barrel is indicated by a cylinder in A. The cross section together with the barrel walls is seen, the low-height wall near the trimer axis is on the left side. The constriction is caused by the two branches of loop L3 which are attached to the inner side of the peripheral wall. From the constriction towards the external surfaces, the channel is widening, although differently in the four porins: the channels of OmpF and PhoE show a further constriction near the external entrance of the channel.

exclusion limit which were obtained after applying a strong selection pressure by growth on large molecular weight maltodextrins [99] and from site-specific mutations of PhoE [100]. The constriction zone seems to function also as a filter against apolar substances. The arginine and lysine residues are aligned parallel in van der Waals contact with low temperature factors in spite of

their mutual repulsion. This indicates that they are held in position by a strong electric field. The water ordered by this field must be structured and thus cause a high energy barrier for passage of small apolar substances [82]. This reminds of a similar barrier function of the long oligosaccharide chains of lipopolysaccharides [4].

936 constriction and that other loops of the external barrel rim constrict the external channel entrance and alter their conformation upon gating. In VDAC, several amino acids distributed over the whole polypeptide were found to be responsible for voltage gating [103]. Together with the large exclusion limit this also indicates a different folding of the loops as compared to the known bacterial porin structures. The eyelet construction as well as the gating mechanism in these porins thus should not be assumed to be universal to all porins. Bound substances

Detergent binding has been suggested from cylindrical electron density in the inside barrel wall of porin from Rhodobacter capsulatus [78] as well as near the trimer axis at the external end of the barrel of Rhosopseudoman blastica [S2]. Two Ca 2 + -binding sites have been found in the region of loop L3 rich in acidic amino acids [78]. Fig. 5. Ribbon diagram of porin from Rhodobacter capsulatus 37b4, seen from the external side. In addition, the charged residues of the narrow part of the channel are shown: 5 basic residues and 1 His attached to the wall side near th~ trimer axis. and 9 acidic residues, mostly attached to lool? L3. For companson, the dIameter of a hydrated potassium ion (14 A) is indicated by a circle.

Possible relevance of the porin structure for other membrane proteins

Where else could (3-barrels in proteins [104] be used as surfaces of channels? Channels with barrels wider than those of porins

As the arrangement of charged residues around the constriction site seems to be critical for high rates of ion permeation through the porin channel, the voltage-induced partial closing of porins could be due to a field-induced conformational transition disturbing the structural arrangement of charges of these residues. In Figure 6 the positions of these residues are shown when viewing from the trimer axis towards a porin monomer from Rhodobacter capsulatus 37b4. The center-of-mass of negatively charged residues is slightly shifted towards the external surface as compared to the center-of-mass of the positively charged residues. In voltagegating experiments the electric field is applied parallel to the membrane normal and is thus either pushing the opposite charges towards each other or pulling them apart. In the former case, the external field adds to the mutual attraction of the charges. Consequently, a breakdown of the local arrangement is feasible with a concomitant shift of opposite charges towards each other. The shift could be brought about by a swing of the extended side chains of Asp, Glu, Arg and Lys towards each other, the former swinging slightly towards the periplasmic-, the latter towards the external surface. It could also be the case, that L3 is to a certain degree detached by the applied force from the barrel wall and its mobile part swings towards the opposite barrel wall. This view is in accord with the experiments of Brunen and Engelhardt [28] who observed an increase of gating voltage upon titration to high or low pH values. The neutralization of charged amino acid residues would decrease the attraction of the oppositely charged parts, so that a higher external field is needed to cause the breakdown of the arrangement. However, in porin from Haemophilus influenzae a residue in the fourth external loop (Arg 166) was found to be critical for gating [27]. This porin has a significantly larger exclusion limit as compared to the four porins of known structure of 1400 Da [101], reminiscent to protein F of Pseudomonas aeruginosa [102]. This may indicate that loop L3 in this porin does not form an eyelet

A class of outer membrane proteins is formed by the receptors for siderophores (iron chelating complexes), FhuA, FepA and FecA and the vitamin B12 receptor [105]. Recently, for FhuA a 32-stranded l3-barrel has been proposed [106]. One loop forms a binding site for ferrichrome and its deletion renders the protein a large open channel [107]. Similar results were obtained with FepA for which a 29-stranded barrel was proposed [lOS]. A large transmembrane l3-barrel was also proposed for the membrane-inserted oligomer of a bacterial toxin, aerolysin, [109]. Channels with barrels narrower than those of porins

OmpA, an outer membrane protein from Escherichia coli with 325 residues, consists of a 177 amino acids long N-terminal domain for which an S-stranded (3-barrel has been predicted which seems to be linked flexibly to a C-terminal periplasmic domain [110, 111]. OmpA has been shown to form pores of low permeability but similar exclusion limit as general diffusion porins [112]. Barrels in water-soluble binding-proteins of known structure consist of S or 10 strands. The retinol-binding protein [113] is a S stranded barrel, 10 stranded barrels are found in the cellular retinol binding protein and the P2 myelin protein [114]. By their narrow inner diameter the latter are suited to form well-adapted geometrical binding sites for small molecules. These narrow barrels could be used as channel domains of specific ion channels. In addition, a high symmetry of the barrel could be advantageous to achieve specificity [115]. Eight-stranded symmetrical barrels have indeed been proposed for a dimeric Cl- channel [116] and for tetrameric voltage gated ion channels [117, I1S]. Channel geometry and channel properties

The general diffusion porins with their 16-stranded barrels have both a wide entrance and outlet with a constriction site halfway in between. With this geometry they minimize frictional interaction

937

Fig, 6. Simplified ribbon drawing of porin from Rhodobacter capsulatus 37b4 showing the charges of amino acids near the constriction site as seen from the trimer axis. The center of mass of the negative charges is slightly shifted to the external side as compared to the positive charges.

between the solutes and the walls of the pore and allow for high rates of solute permeation, but nevertheless with a restricted exclusion limit and some selectivity due to the cross section and the charged residues of the constriction site [102, 119, 120]. In contrast, 8 stranded barrels allow only for low permeation rates as in OmpA, but may achieve high selectivity due to the numerous interactions of the solutes along the channel wall. Acknowledgments We thank Tilman Schirmer, Hans Jansonius, Jiirg P. Rosenbusch, Andreas Kreusch and Georg E. Schulz for access to the atomic coordinates of OmpF and PhoE porin from Escherichia coli and porin from Rhodopseudomonas blastica. Reprint requests to Wolfram Welte, M.D., Fakultiit for Biologie, Universitiit Konstanz, Universitiitsstr. 10, 78434 Konstanz, Germany. E-mail: [email protected]

References 1. LUGTENBERG B, VAN ALPHEN L: Molecular architecture and functioning of the outer membrane of Escherichia coli and other gramnegative bacteria. Biochim Biophys Acta 737:51-115, 1983

2. ALBERTI S, MARQUES G, CAMPRUBI S, MERINO S, TOMAS IN, VIVANCO F, BENED( VJ: C1q binding and activation of the complement classical pathway by Klebsiella pneumoniae outer membrane proteins. Infect Immun 61:852-860, 1993 3. RIETSCHEL ET, KIRIKAE T, SCHADE U, MAMAT U, SCHMIDT G, LOPPNOW H, ULMER AJ, ZAHRINGER U, SEYDEL U, DIPADOVA F, SCHREIER M, BRADE H: Bacterial endotoxin: Molecular relationships of structure to activity and function. FASEB J 8:217-225, 1994 4. NIKAIDO H, V AARA M: Outer Membrane, in Escherichia Coli and Samonella Typhimurium, Vol.l edited by FC NEIDHARDT, Washington, D.C., American Soc.f.Microbiol., 1987, pp 7-22 5. NAKAE T: Identification of the outer membrane protein of E.coli that produces transmembrane channels in reconstituted vesicle membranes. Biochim Biophys Res Comm 71:877-884, 1976 6. ROSENBUSCH JP: Characterization of the major envelope protein from Escherichia coli. J Bioi Chern 249:8019-8029, 1974 7. SCHINDLER H, ROSEN BUSCH JP: Matrix protein from Escherichia coli outer membranes forms voltage-controlled channels in lipid bilayers. Proc Natl Acad Sci USA 75:3751-3755, 1978 8. NAKAE T, ISHII J, TOKUNAGA M: Subunit structure of functional porin oligomers that form permeability channels in the outer membrane of Escherichia coli. J Bioi Chern 254:1457-1461, 1979 9. KREUSCH A, WEISS MS, WELTE W, WECKESSER J, SCHULZ GE: Crystals of an integral membrane protein diffracting to 1.8 A resolution. J Mol Bioi 217:9-10, 1991

938 10. BuTZ S, BENZ R, WACKER T, WELTE W, LUSTIG A , PLAPP R, WECKESSER J: Biochemical characterization and crystallization of porin from Rhodopseudomonas blastica. Arch MicrobioI159:301-307, 1993 11. VACHON V, LYEW D, COULTON JW: Transmembrane permeability channels across the outer membrane of Haemophilus injluenzae typc b. J BacterioI162:918-924, 1985 12. TOKUNAGA M, TOKUNAGA H , OKAJlMA Y, NAKAE T: Characterization of porins from the outer membrane of Salmonella typhimurium. Eur J Biochem 95:441-44R, 1979 13. NAKAE T, ISHIT IN: Molecular weights and subunit structure of LamB proteins. A nn Microbiol (lnst Pasteur) 133A:21- 25, 1982 14. MAURO A, BLAKE M, LABARCA P: Voltage gating of conductance in lipid bilayers induced by porin from outer membrane of Neisseria gonorrhoeae. Proc Natl Acad Sci USA 85:1071-1075, 19R8 15. BENZ R, JANKO K, Boos W , Li\UOER P: Formation of large, ion-permeable membrane channels by the matrix protein (porin) of Escherichia coli. Biochim Biophys Acta 511:305-319, 1978 16. DARGENT B, HOFMANN W, PATTUS F, ROS ENBUSCH JP: The sel ectivity filter of voltage-dependent channels formed by phosphoprotein (PhoE protein) from E.coli. EMBO J 5:773-778, 1986 17. VACHON V, LAPRADE R, COULTON JW: Propertics of the porin of Haemophilus injluenzae type b in planar lipid bilayer membranes. Biochim Biophys Acta 861 :74 -82, 1986 18. LAKEY LH, PATTUS F: The voltage-dependent activity of Escherichia coli porins in different planar bilayer reconstitutions. Eur J Biochem 186:303-308, 19R9 19. WI ESE A, SCHRODER G , BRANDENBURG K, WELTE W , SEYDEL U: The influence of the lipid matrix on incorporation and function of LPS-free porin from Paracoccus denillificans. Biochim Biophys Acta 1190:231 - 242,1994 20. BUECHNER M, DELCOUR AH, MARTINAC B, ADLER J, KUNG C: Ion channel activities in the Escherichia coli outer membrane. Biochim Biophys Acta 1024:111-121, 1990 21. NIKAIDO H: Proteins forming la rge channels from bacterial and mitochondrial outer membranes: Porins and phage lambda receptor protein. Meth Enzymol 97:85-100, 1983 22. WOITZIK D , WECKESSER J, BENZ R , STEVANOVIC S, JUNG G , ROSENBUSCH JP: Porin of Rhodobacter capsulatus: Biochemical and functional characterization. Z Naturf 45c:576-582, 1990 23. YOSH IHARA E, YONEYAMA H, NAKAE T: In vitro assembly of the functional porin trimer from dissociated monomers in Pseudomonas aeruginosa. J BioI Chem 266:952-957, 1991 24. BENZ R: Structure and function of porins from gram-ncgative bacteria. Ann Rev Microbiol 42:359 -393, 1988 25. BENZ R, BAUER K: Permeation of hydrophilic molecules through the outer membrane of gram-negative bacteria. Eur J Biochem 176: 1- 19, 1988 26. BRUNEN M, ENGELHARDT H: Asymmetry of orientation and voltage gating of the acidivorax delafildii porin Omp34 in lipid bilayers. Eur J Biochem 212:129-135, 1993 27. DAHAN D , VACHON Y, LAPRADE R , COULTON lW: Voltage gating of porins from Haemophilus injluenzae type b. Biochim Biophys Acta 1189:204-211 , ]994 28. BRlJNEN M, ENGELHARDT H: Asymmetry of orientation and voltage gating of the Acidovorax delafildii porin Omp34 in lipid bilayers. Eur J Biochem 212:129-135, 1993 29. ZALMAN LS, NIKAIDO H, KAGAWA Y: Mitochondrial outer membrane contains a protein producing nonspecific diffusion channels. J Bioi Chem 255:1771-1774, 19RO 30. COLOMBINI M: Regulation of the mitochondrial outer membrane channel, VDAC. J Bioenerg Biomemb 19:309-320, 1987 31. MANNELLA CA: Electron microscopy and image analysis of the mitochondrial outer membrane channel, VDACJ Bioenerg Biomemb 19:329-340, 1987 32. MANNELLA C, FRANK J: Effects of divalent metal ions and chelators on the structure of outer mitochondrial membranes from Neurospora crassa. Biophys J 37:3-4, 1982 33. THOMAS L, KOCSIS E, COLOMBINI M, ERBE E, TRUS BL, STEVEN AC: Surface topography and molecular stoichiometry of the mitochondrial channel, VDAC, in crystalline arrays. J Stntct BioI 106: 161-171, 1991

34. PENO S, BLACHLy-DYSON E, COLOMBINI M, FORTE M : Determination of the number of polypeptide subunits in a functional VDAC channel from Saccharomyces cerevisiae. J Bionerget Biomemb 24:2731, 1992 35. BRDICZKA D: Interaction of mitochondrial porin with cytosolic proteins. Experentia 46:161-167, 1990 31i. WAJ.L1MANN T, WYSS M, BRDlCZKA D , NICOLAY K, EpPENB ERGER HM: Intracellular compartmentation , structure and function of cratine kinase isozymes in tissues with high and fluctuating energy demands: The 'phophocreatine circuit' for cellular energy homeostasis. Biochem J 281:21-40, 1992 37. KR UEGER KE, PAPADOPOULOS Y: Peripheral-type benzodiazepine receptors mediate translocation of cholesterol from outer to inner mitochondrial membranes in adrenocortical cells. J Bioi Chem 265: 15015-15022, 1990 38. Llu MY, TORGRIMSON A, COLOMBINI M: Characterization and partial purification of the VDAC-channel-modulating protein from calf liver mitochondria. Biochim Biophys Acta 1185:203- 212, 1994 39. DERMIETZEL R, HWANG TK, BU ETTNER R, HOFER A, DOTZLER E, KREMER M, DEUTZMANN R , THINNES F, FISHMAN GI, SPRAY DC, SIEMEN D: Cloning and in situ localization of a brain-derived porin that constitutes a large-conductance anion channel in astrocytic plasma membranes. Proc Natl Acad Sci USA 91 :499-503, 1994 40. FJ.OGGE UI, BENZ R: Pore-forming activity in the outer membrane of the chloroplast envelope. FF.RS Lett 169:85-89, 1984 41. TRIAS J, JARLIER Y, BENZ R: Porins in the cell wall of mycobacteria. Science 258:1479-1481, 1992 42. SCHMID K, EBNER R , JAHREIS K, LENGELER JW, TITOEM EYER F: A sugar-specific porin, ScrY, is involved in sucrose uptake in enteric bacteria. Mol Microbial 5:941-950, 1991 43. POSTMA PW, LENGELER JW, JACOBSON GR: Phosphoenolpyruvate: carbohydrate phospho transferase systems of bacteria. Microbiol Rev 57:543-594, 1993 44. DASSA E: Sequence-function relationships in MaIG, an inner membrane protein from the ma ltose transport system in Escherichia coli. Mol Microbiol 7:39-47, 1993 45. MAI ER C, BREMER E , SCHMID A , BENZ R. : Pore-forming activity of the Tsx protein from the outer membrane of Escherichia coli. J Bioi Chem 263:2493-2499, 1988 46. SCHOLEIN K, SCHMID K, BENZ R: The sugar specific outer membrane channel ScrY contains functional characteristics of general diffusion pores and substrate-specific porins. Mol MicrobioI5 :2233-2241 , 1991 47. CH EN R, KRAEMER C, SCHMIDTMAYR W, HENNING U: Primary structure of major outer membrane protein I of Escherichia coli Blr. Proc Natl Acad Sci USA 76:5014-5017, 1979 48. NIKAIDO H, REID J: Biogenesis of prokaryotic pores. Experienlill 46:174 -1 80, 1990 49. MOLLER M, MACFARLANE J : Membrane Assembly in Bacteria, in Subcellular Biochemistry (vol 22), edited by AH MADDY, JR HARRIS, New York, Plenum Press, 1994, pp 327-359 50. CAR1.S0N JR, SILHAVY TJ: Signal sequence processing is required for the asse mbly of LamB trimers in the outer membrane of Escherichia coiL J BacterioI175:3327-3334, 1993 51. SCHILTZ E, KREUSCH A, N ESTEL U, SCHU LZ GE: Primary structure of porin from Rhodobacter capsula/us. Eur J Biochem 199:587-594, 1991 52. J EANTEUR D , LAKEY JH, PATTUS F: The bacterial porin superfamily: Sequence alignment and structure prediction. Mol Microbiol 5:21532164,1991 53. NIKAIDO H: POl·ins and specific channels of bacterial outer membranes. Mol Microbiol 6:435-442, 1992 54. TOMMASSEN J: Biogenesis and membrane topology of outer memhrane proteins in Escherichia coli, in Membrane Biogenesis NATOASJ Series, cdited by JAF Op DEN KAMP, Berlin, Springer Verlag, 1988, pp 352-373 55. KORTELAND J, OVERB EEKE N, DEGRAAF P, OVERDUIN P, LUGTENBURG B: Role of the Arg158 residue of the outer membrane PhoE pore protein of Escherichia coli KI2 in bacteriophage TC45 rccognition and in channel characteristics. Eur J Biochem 152:691- 697, 1985 56. KLEI'l'EL B, GARAV ITO RM, BAUMEISTER W, ROSf.NlllJSCH JP: Secondary structure of a channel-forming protein: Porin from E. coli o uter membranes. EMBO J 4:1589-1592, 1985 57. NAREDRYK E, G;\R;\VITO RM, BRETON J: The orientation of J3-sheets

939

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72. 73.

74.

75.

76.

77.

78. 79. 80.

in porin. A polarized fourier transform infrared spectroscopic investigation. Biophys J 53:671-676, 1988 VAN DER LEY P, HECKELS JE, VIRJI M, HOOGERHOUT P, POOLMAN JT: Topology of outer membrane porins in pathogenic Neisseria spp. Infect Immunol 59:2963-2971, 1991 WELTE W, WEISS MS, NESTEL U, WECKESSER J, SCHILTZ E, SCHULZ GE: Prediction of the general structure of porin from Rhodobacter capsulatus. Orientation of porin in the membrane. Biochim Biophys Acta 1080:271-274, 1991 FORTE M, GUY HR, MANNELLA CA: Molecular genetics of the VDAC ion channel: Structural model and sequence analysis. J Bioenerg Biomemb 19:341-350, 1987 BLACHLy-DYSON E, PENG S, COLOMBINI M, FORTE M: Selectivity changes in site-directed mutants of the VDAC ion channel: Structural implications. Science 247:1233-1236, 1990 DE PINTO V, PREZIOSO G, THINNES F, LINK TA, PALMIERI F: Peptide-specific antibodies and proteases as probes of the transmembrane topology of the bovine heart mitochondrial porin transmembrane arrangement of mitochondrial porin. Biochemistry 30:1019110200, 1991 GARAVlTO RM, ROSEN BUSCH JP: Three-dimensional crystals of an integral membrane protein: An initial X-ray analysis. J Cell Bioi 86:327-329, 1980 GARAVITO RM, JENKINS J, JANSONIUS IN, KARLSSON R, ROSENBUSCH JP: X-ray diffraction analysis of matrix porin, an integral membrane protein from Escherichia coli outer membranes. J Mol Bioi 164:313327, 1983 GARAVITO RM, HINZ U, NEUHAUS JM: The crystallization of outer membrane proteins from Escherichia coli. J Bioi Chem 259:42544257, 1984 STAUFFER KA, PAGE MGP, HARD MEYER A, KELLER TA, PAUPTIT RA: Crystallization and preliminary X-ray characterization of maltoporin from Escherichia coli. J Mol Bioi 211:297-299, 1990 TUCKER, AE, JACKMAN S, PARKER MW, TSERNOGLOU D: Crystallization and preliminary X-ray analysis of phosphoporin from the outer membrane of Escherichia coli. J Mol Bioi 222:881-884, 1991 DORSET DL, ENGEL A, HANER M, MASSALSKI A, ROSEN BUSCH JP: Two-dimensional crystal packing of matrix porin. J Mol Bioi 165:701710, 1983 ENGEL A, MASSALSKI A, SCHINDLER H, DORSET DL, ROSENBUSSCH JP: Porin channel triplets merge into single outlets in Escherichia coli outer membranes. Nature 317:643-645, 1985 JAP BK, WALIAN PJ, GEHRING K: Structural architecture of an outer membrane channel as determined by electron crystallography. Nature 350:167-170, 1991 SASS HJ, BOLDT G, BECKMANN E, ZEMLIN F, VAN HEEL M, ZEITLER E, ROSENBUSCH JP, DORSET DL, MASSALSKI A: Densely packed (3-structure at the protein-lipid interface of porin is revealed by high-resolution cryo-electron microscopy. J Mol Bioi 209:171-175, 1989 CHALCROFT JP, ENGELHARDT H, BAUMEISTER W: Structure of the porin from a bacterial stalk. FEBS Lett 211:53-58, 1987 RACHEL R, ENGEL AM, HUBER R, STETTER KO, BAUMEISTER W: A porin-type protein is the main constituent of the cell envelope of the ancestral eubacterium Themotoga maritima. FEBS Lett 262:64-68, 1990 BENZ R, WOlTZIK D, FLAMANN HT, WECKESSER J: Pore forming activity of the major outer membrane protein of Rhodobacter capsulatus in lipid bilayer membranes. Mol MicrobioI148:226-230, 1987 NESTEL U, WACKER T, WOlTZIK D, WECKESSER J, KREUTZ W, WELTE W: Crystallization and preliminary X-ray analysis of porin from Rhodobacter capsulatus. FEBS Lett 242:405-408, 1989 WEISS MS, KREUSCH A, SCHILTZ E, NESTEL U, WELTE W, WECKESSER J, SCHULZ GE: The structure of porin from Rhodobacter capsulatus at 1.8 A resolution. FEBS Lett 280:379-382, 1991 WEISS MS, ABELE U, WECKESSER J, WELTE W, SCHILTZ E, SCHULZ GE: Molecular architecture and electrostatic properties of a bacterial porin. Science 254:1627-1630, 1991 WEISS MS, SCHULZ GE: Structure of porin refined at 1.8 A resolution. J Mol Bioi 227:493-509, 1992 WEISS MS, SCHULZ GE: Porin conformation in the absence of calcium. J Mol Bioi 231:817-824, 1993 PAUPTlT RA, ZHANG H, RUMMEL G, SCHIRMER T, JANSONIUS IN, ROSEN BUSCH JP: Trigonal crystals of porin from Escherichia coli. J Mol Bioi 218:505-507, 1991

81. COWAN SW, SCHIRMER T, RUMMEL G, STEIERT M, GHOSH R, PAUPTIT RA, JANSON IUS JN, ROSENBUSCH JP: Crystal structures explain functional properties of the two E. coli porins. Nature 358:727-733, 1992 82. KREUSCH A, NEUBOSER A, SCHILTZ E, WECKESSER J, SCHULZ G. E.: Structure of the membrane channel porin from Rhodopseudomonas blastica at 2.0 A resolution. Protein Sci 3:58-63, 1994 83. FORST D, SCHOLEIN K, WACKER T, DIEDERICHS K, KREUTZ W, BENZ R, WELTE W: Crystallization and preliminary X-ray diffraction analysis of ScrY, a specific bacterial outer membrane porin. J Mol Bioi 229:258-262, 1993 84. HOENGER A, PAGES JM, FOUREL D, ENGEL A: The orientation of porin OmpF in the outer membrane of Escherichia coli. J Mol Bioi 233:400-413, 1993 85. BOLDT G, GALLY HU, SEELIG A, SEELIG G, ZACCAI G: Neutron diffraction studies on selectively deuterated phospholipid bilayers. Nature 217:182-184, 1978 86. ZACCAI G, BLASIE JK, SCHOENBORN BP: Neutron diffraction studies on the location of water in lecithin bilayer model membranes. Proc Nat! Acad Sci USA 72:376-380, 1975 87. BRANDENBURG K, SEYDEL U: Investigation into the fluidity of lipopolysaccharide and free lipid A membrane systems by Fouriertransform infrared spectroscopy and differential scanning calorimetry. Eur J Biochem 191:229-236, 1990 88. TANFORD C: The Hydrophobic Effect: Formation of Micelles and Biological Membranes (2nd ed). New York, John Wiley & Sons, 1980,

plO 89. SUZUKJ S, GREEN PG, BUMGARNER RE, DASGUPTA S, GOFFARD WA, BLAKE GA: Benzene forms hydrogen bonds with water. Science 257:942-945, 1992 90. DEISENHOFER J, MICHEL H: The photosynthetic reaction centre from the purple bacterium Rhodopseudomonas viridis. EMBO J 8:21492170, 1989 91. SCHIFFER M, CHANG CH, STEVENS FJ: The functions of tryptophan residues in membrane proteins. Protein Engineer 5:213-214, 1992 92. HENDERSON R, BALDWIN JM, CESKA TA, ZEMLIN F, BECKMANN E, DOWNING KH: Model for the structure of bacteriorhodopsin based on high-resolution electron cryo-microscopy. J Mol Bioi 213:899-929, 1990 93. FRIEDMAN HL, KRISHNAN CV: Thermodynamics of ionic hydration, in Water, A Comprehensive Treatise (vol 3, edited by F FRANKS, New York, Plenum Press, 1973, pp 49-50 94. BENZ R, DARVEAU RP, HANCOCK REW: Outer-membrane protein PhoE from Escherichia coli forms anion-selective pores in lipidbilayer membranes. Eur J Biochem 140:319-324, 1984 95. DARVEAU RP, HANCOCK REW, BENZ R: Chemical modification of the anion selectivity of the PhoE porin from Escherichia coli outer membrane. Biochim Biophys Acta 774:67-74, 1984 96. BAUER K, VAN DER LEY P, BENZ R, TOMMASSEN J: The Phocontrolled outer membrane porin PhoE does not contain specific binding sites for phosphate or polyphosphates. J Bioi Chem 263: 13046-13053, 1988 97. BENZ R, SCHMID A, VAN DER LEY P, TOM MASS EN J: Molecular basis of porin selectivity: Membrane experiments with OmpC-PhoE and OmpF-PhoE hybrid proteins of Escherichia coli K-12. Biochim Biophys Acta 981:8-14, 1989 98. BAUER K, STRUYVE M, BOSCH D, BENZ R, TOMMASSEN J: One single lysine residue is responsible for the special interaction between polyphosphate and the outer membrane porin PhoE of Escherichia coli. J Bioi Chem 264:16393-16398, 1989 99. MISRA R, BENSON SA: Genetic identification of the pore domain of the OmpC porin of Escherichia coli K-12. J BacterioI170:3611-3617, 1988 100. STRUYVE, M, VISSER J, AoRlAANSE H, BENZ R, TOMMASSSEN J: Topology of PhoE porin: The 'eyelet' region. Mol Microbiol 7:131140, 1993 101. VACHON V, LYEW DJ, COULTON JW: Transmembrane permeability channels across the outer membrane of Haemophilus inJluenzae type b. J Bacteriol 162:918-924, 1985 102. NIKAIDO H, NIKAIDO K, HARAYAMA S: Identification and characterization of porins in Pseudomonas aeruginosa. J Bioi Chem 266:770779, 1991 103. THOMAS L, BLACHLy-DYSON E, COLOMBINI M, FORTE M: Mapping

940

104. 105. 106.

107. 108.

109.

110. 111.

of residues forming the voltage sensor of the VDAC ion channel. Proc Natl Acad Sci USA 90:5446-5449, 1994 MURZIN AG, LESK AM, CHOTHIA C: Principles determining the structure of (3-sheet barrels in proteins.] Mol BioI 236:1369-1400, 1994 KLEBBA PE, RuTZ JM, LIu J, MURPHY CK: Mechanisms of TonBcatalyzed iron transport through the enteric bacterial cell envelope. ] Bioenerget Biornernb 25:603-611, 1993 KOEBNIK R, BRAUN V: Insertion derivatives containing segments of up to 16 amino acids identify surface- and periplasm-exposed regions of the FhuA outer membrane receptor of Escherichia coli K-12. ] Bacterial 175:826-839, 1993 KILLMANN H, BENZ R, BRAUN V: Conversion of the FhuA transport protein into a diffusion channel through the outer membrane of Escherichia coli. EMBO] 12:3007-3016, 1993 RuTZ J, LIU J, LYONS JA, GORANSON J, ARMSTRONG SK, McINTOSH MA, FElx JB, KLEBBA PE: Formation of a gated channel by a ligand-specific transport protein in the bactcrial outer membrane. Science 258:471-475, 1992 PARKER MW, BUCKLEY JT, POSTMA JPM, TuCKER AD, LEONARD K, PATTUS F, TSERNOGLOU D: Structure of the Aerornonas toxin proaerolysin in its water-soluble and membrane-channel states. Nature 367:292-294, 1994 MORONA R, KLOSE M, HENNING U: Escherichia coli K-12 outer membrane protein (OmpA) as a bacteriophage receptor: Analysis of mutant genes expressing altered proteins.] Bacterial 159:570-578, 1984 VOGEL H, JAHNIG F: Models for the structure of outer-membrane

112. 113. 114. 115. 116. 117. 118. 119. 120. 121.

proteins of Escherichia coli. Derived from Raman spectroscopy and prediction methods.] Mol 8ioI190:191-199, 1986 SUGAWARA E, NIKAIDO H: Pore-forming activity of OmpA protein of Escherichia coli. ] BioI Chern 267:2507-2511, 1992 NEWCOMER ME, JONES TA, AQVIST J, SUNDELlN J, ERIKSSON U, RASK L, PETERSON PA: The three-dimensional structure of retinolbinding protein. EMBO] 3:1451-1454, 1984 COWAN S, NEWCOMER ME, JONES TA: Crystallographic studies on a family of cellular lipophilic transport proteins.] Mol BioI 230:12251246, 1993 YELLEN G: Structure and selectivity. Nature 366:109-110,1993 PAULMICHL M, LI Y, WICKMANN K, ACKERMANN M, PERALTA E, CLAPHAM D: New mammalian chloride channel identified by expression cloning. Nature 356:238-241, 1992 MILLER C: Annus Mirabilis of potassium channels. Science 252: 10921096, 1991 BOGUSZ S, BOXER A, BUSATH DD: An SS1-SS2 (3-barrel structure for the voltage-activated potassium channel. Protein Engineering 5:285293, 1992 NlKAIDO H: Porins and specific diffusion channels in bacterial outer membranes.] BioI Chern 269:3905-3908, 1994 EISENMANN G, DANI JA: An introduction to molecular architecture and permeability of ion channels. Ann Rev Biophys Chern 16:205-226, 1987 KRAULlS PJ: Molscript: A program to produce both detailed and schematic plots of protein structures. ] Appl Cryst 24:946-950, 1991