Prokaryotic K + channels: From crystal structures to diversity Mario M.-C. Kuo a , W. John Haynes a , Stephen H. Loukin a , Ching Kung a,b, * , Yoshiro Saimi a a Laboratory of Molecular Biology, University of Wisconsin, 1525 Linden Drive, Madison, WI 53706, USA b Department of Genetics, University of Wisconsin, 1525 Linden Drive, Madison, WI 53706, USA Received 31 December 2004; received in revised form 21 March 2005; accepted 23 March 2005 First published online 1 July 2005 Abstract The deep roots and wide branches of the K + -channel family are evident from genome surveys and laboratory experimentation. K + -channel genes are widespread and found in nearly all the free-living bacteria, archaea and eukarya. The conservation of basic structures and mechanisms such as the K + filter, the gate, and some of the gateÕs regulatory domains have allowed general insights on animal K + channels to be gained from crystal structures of prokaryotic channels. Since microbes are the great majority of lifeÕs diversity, it is not surprising that microbial genomes reveal structural motifs beyond those found in animals. There are open-reading frames that encode K + -channel subunits with unconventional filter sequences, or regulatory domains of different sizes and numbers not previously known. Parasitic or symbiotic bacteria tend not to have K + channels, while those showing lifestyle versatility often have more than one K + -channel gene. It is speculated that prokaryotic K + channels function to allow adaptation to environmental and metabolic changes, although the actual roles of these channels in prokaryotes are not yet known. Unlike enzymes in basic metabolism, K + channel, though evolved early, appear to play more diverse roles than revealed by animal research. Finding and sorting out these roles will be the goal and challenge of the near future. Ó 2005 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords: Membrane potential; Potassium channels; Potassium filter; Parasitism; RCK/KTN Contents 1. Introduction ............................................................................ 962 2. K + and the electrochemical gradient ........................................................... 962 3. Ion movements across cell membrane .......................................................... 963 4. The general features of K + channels ........................................................... 964 4.1. The basic structure and activity .......................................................... 964 4.2. The K + filter........................................................................ 965 4.3. The gate and gating................................................................... 965 5. K + channels in prokaryotes ................................................................. 966 6. Structures of prokaryotic K + channels and their eukaryotic counterparts ................................. 969 6.1. The K + filter and its variations........................................................... 969 6.2. The voltage sensor S4 ................................................................. 971 0168-6445/$22.00 Ó 2005 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.femsre.2005.03.003 * Corresponding author. Tel.: +608 262 9472; fax: +608 262 4570. E-mail address: [email protected](Ching Kung). www.fems-microbiology.org FEMS Microbiology Reviews 29 (2005) 961–985
Transcript
www.fems-microbiology.org
FEMS Microbiology Reviews 29 (2005) 961–985
Prokaryotic K+ channels: From crystal structures to diversity
Mario M.-C. Kuo a, W. John Haynes a, Stephen H. Loukin a,Ching Kung a,b,*, Yoshiro Saimi a
a Laboratory of Molecular Biology, University of Wisconsin, 1525 Linden Drive, Madison, WI 53706, USAb Department of Genetics, University of Wisconsin, 1525 Linden Drive, Madison, WI 53706, USA
Received 31 December 2004; received in revised form 21 March 2005; accepted 23 March 2005
First published online 1 July 2005
Abstract
The deep roots and wide branches of the K+-channel family are evident from genome surveys and laboratory experimentation.K+-channel genes are widespread and found in nearly all the free-living bacteria, archaea and eukarya. The conservation of basicstructures and mechanisms such as the K+ filter, the gate, and some of the gate�s regulatory domains have allowed general insightson animal K+ channels to be gained from crystal structures of prokaryotic channels. Since microbes are the great majority of life�sdiversity, it is not surprising that microbial genomes reveal structural motifs beyond those found in animals. There are open-readingframes that encode K+-channel subunits with unconventional filter sequences, or regulatory domains of different sizes and numbersnot previously known. Parasitic or symbiotic bacteria tend not to have K+ channels, while those showing lifestyle versatility oftenhave more than one K+-channel gene. It is speculated that prokaryotic K+ channels function to allow adaptation to environmentaland metabolic changes, although the actual roles of these channels in prokaryotes are not yet known. Unlike enzymes in basicmetabolism, K+ channel, though evolved early, appear to play more diverse roles than revealed by animal research. Finding andsorting out these roles will be the goal and challenge of the near future.� 2005 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.
Ion channels were first studied and are best known inconnection to the nervous system, especially in the gen-eration and propagation of action potentials. However,they are now known to also underlie secretion, endocy-tosis, muscle contraction, synaptic transmission, ciliarycontrol, fertilization, etc. As reviewed below, ion chan-nels are now found in Bacteria, Archaea, as well as Euk-arya, indicating deep evolutionary roots. For historicaland technical reasons, ion channels are understudiedoutside the animal kingdom, especially in prokaryotes.Unexpectedly, recent events have brought microbialion channels into the limelight, because crystal struc-tures of microbial channels at atomic resolutions haveyielded unprecedented knowledge on how this class ofchannel protein works. Hence, the 2003 Nobel Prize inChemistry was awarded for the structural and mechanis-tic studies of a bacterial K+ channel [1,2]. Nonetheless,the schism between neuro- and microbiology remainswide. Neurobiologists see microbial channels as ‘‘mod-els’’ to those in the brain while microbiologists do notunderstand the full breath of ion-channel structure andfunction. Lamentations of this state of affair aside [3],there is a need to communicate between these fields. Thisreview is written for microbiologists and will cover rele-vant electrophysiological, protein chemical, crystallo-graphic, genetic, and genomic literature. Length anddepth considerations limit this review to only one impor-tant class of ion channels, the K+ channels ofprokaryotes.
Long before any molecular identity of ion channelswas known, there had been extensive studies of theiractivities in eukaryotic cells. Because these activitiesare usually registered electrically, such terms as ‘‘con-ductance’’, ‘‘rectification’’ etc. will be used when neces-sary. Brief explanations will be provided but for moredetailed information on ion channel structures and func-tions please consult additional sources, e.g. Hille [4].Readers are also referred to the literature on other types
of ion channels in prokaryotes [5]: Na+ channels [6,7],ClC chloride channels [8,9], and mechanosensitive chan-nels [10–13].
2. K+ and the electrochemical gradient
K+ is the most abundant ion in cytoplasm. Esche-richia coli, for example, has an internal K+ concentrationof �200 mM, while the standard rich Luria–Bertanimedium (LB) contains only �7 mM K+ (from the yeastextract added [14]). This very high internal K+ concen-tration is retained even when bacteria are deprived ofthe ion, i.e. when they are forced to grow in minimalmedia of submicromolar K+ [15]. The enzymatic andmetabolic processes of the cell are generally well adaptedto having this high internal concentration. Unlike mostother cations, hundreds of millimolar internal K+ doesnot significantly interfere with the structures and reac-tions of macromolecules such as DNAs, RNAs, andproteins in aqueous solutions. The consequence of thisabundance of K+ and its counter ions (glutamate, or-ganic phosphates, etc.) is that it serves as one of the ma-jor osmolytes within the cell. These along with neutralmolecules like glycine betaine draw water across the li-pid bilayer into the cytoplasm to sustain a turgor andthis turgor is needed to disrupt existing structures sothat new material can be added during growth [16]. Be-cause of this critical role, it stands to reason that feed-back mechanisms have evolved to regulate the internalconcentration of K+ according to the external totalosmolarity [17]. The K+ content in E. coli rises linearlyfrom some 200 to 1000 mM as external osmolarity risesfrom 100 to 1200 mOsm [18].
The membrane lipid bilayer sustains the different con-centrations of K+, [K+], as well as an electric potentialdifference between the interior and the outside. The elec-trochemical gradient of K+ across the membrane, ex-pressed as energy, is
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 963
DlKþ ¼ FDWþ RT ln½Kþ�in=½Kþ�out;
where DW is the membrane potential (i.e. the electric po-tential of the cytoplasm with respect to the outside,ground). F is Faraday�s constant, R is the gas constantand T is temperature in degrees Kelvin. Expressed as adriving voltage (in millivolts, mV), then
DlKþ=F ¼ DWþ 58 log½Kþ�in=½Kþ�out.
As described above, [K+]in is usually larger than [K+]out,but most cytoplasms are electrically negative (with rareexceptions, below). For example, the membrane poten-tial of aerobically growing E. coli was estimated to bebetween �94 and �157 mV when the external pH raisedfrom pH 6.25 to 8.25 [19,20]. Thus the electric compo-nent of the driving force, DW, tends to draw K+ intothe cell, while the concentration component, 58 log[K+]in/[K+]out, tends to drive it out. K+ can therefore be drivenpassively inward or outward depending on the relativelevels of these two components of the motive force. Thisis different from the electromotive force for H+, the pro-ton motive force (Dp, Dp = DW + 58DpH), where bothof its two components, DW and DpH (pHout � pHin),tend to drive H+ inward (pHout is usually smaller thanpHin) [19].
3. Ion movements across cell membrane
Steady-state gradients of ions, organic or inorganic,can be maintained because hydrocarbons within a lipidbilayer makes the membrane significantly impermeableto ions. Unless neutralized by counter ions (e.g., NH4
acetate), ions traverse membrane through the lumen orpore of various membrane proteins. There are three cat-egories of transport proteins that handle ion traffic.Those familiar to microbiologists are the ion pumpsand exchangers. These proteins move ions against gradi-ents at the expenditure of ATP (pumps) or other iongradients (exchangers, including symporters and anti-porters) such as the proton motive force. For example,E. coli has three K+-uptake systems, Trk, Kdp andKup [17]. The main uptake is by the Trk (transport ofK+) system, a constitutive, low-affinity high-rate systemenergized by the proton motive force and ATP. The sec-ond important K+-uptake system is Kdp, (K+-depen-dent growth), a high-affinity low-rate, inducible P-typeATPase pump. The transmembrane subunits of Trk(TrkH/G) and Kdp (KdpA) are members of proposedsuperfamily of K+ transporters that include homologuesin other prokaryotes (KtrB, formerly NtpJ), fungi(TRK1, TRK2) and plants (HKT1) [21,22]. These trans-porters appear to contain covalently linked TM-P-TMmotifs (TM, transmembrane helix; P, segment for ionselectivity) proposed to have originally evolved from aKcsA-like K+ channel by gene duplication [23–25]. This
paper does not examine the relationship between thetransporter and K+ channel motif (discussed below)since several distinctive sequence differences prevent sig-nificant high scoring matches in BLAST searches. Thethird system Kup is apparently an alternative to TrkHwhen cells are growing at low pH [26]. There is also avarious efflux mechanism including KefB and KefCwhich appear to be more similar to Na+/H+ and K+/H+ antiporters than to K+ channels (see the ConservedDomain Database, NCBI). These K+ uptake and effluxsystems are not to be confused with the topic of this re-view, the K+ channels, which are members of the thirdcategory of conduits.
Ion channels are distinguished from the previous twoclasses of transporters because a channel provides for acompletely ‘‘open pore’’ through which the ions can dif-fuse. To form the conduit, these proteins must enclose ahydrophilic pathway to allow the passage of the solutes(e.g. ions). Unlike the b-barrels that form the porins inthe outer membranes of gram-negative bacteria, ion-channel subunits in the plasma membrane are comprisedof a helices (Fig. 1). These pores are often formed whenhelices from separate identical or similar subunits, cometogether to surround the pathway. In addition to K+
channels, two well-known examples of multi-subunitchannels are the bacterial mechanosensitive channel oflarge conductance (MscL) which is a homopentamer[11] and the mechanosensitive channel of smaller con-ductance (MscS), a homoheptamer [13]. The pore con-formation allows these proteins to function in a waythat is entirely different from those of pumps orexchangers. When a channel opens, at zero membranepotential, ions fall passively down the preexisting con-centration gradients and dissipate the ion gradient.However, as pointed out above, in live cells, the mem-brane potential and concentration gradient constitutethe driving force, DlK+, and can have opposite effectson the direction of K+ flow. Thus, whether K+ flows in-ward or outward through K+ channels will depend onthe balance of these two components. The familiar in-flux of H+ passively down its electrochemical gradientinto the cell is well known to drive myriad importantprocesses such as ATP synthesis, flagella rotation, aswell as the sym- or antiporting of various materialsacross the membrane of prokaryotes. However, thefunction(s) of the passive fluxes of ions through channelsin the prokaryotes are not well understood. In animals,the influx of Ca2+ through Ca2+ channels raises the localcytoplasmic Ca2+ concentration to micromolar levelsthereby regulating actin-myosin and dynein-tubulinmotility, among many other activities. Most K+ chan-nels in animals either maintain the resting membranepotential or rapidly reestablish the potential to the rest-ing level after depolarization. Other animal K+ channelsalter the resting membrane potential upon different stim-uli. K+ channels are usually not for bulk K+ uptake in
Fig. 1. The crystal structure of KcsA K+ channel of St. lividans. Stereoview of the channel from the membrane. The KcsA K+ channel comprisesfour identical subunits, which come together radially to form the hydrophilic pore. Each subunit contains a pore loop flanked by two transmembranehelices, the outer (TM1) and inner (TM2) helices. The four C-termini converge at the bottom of the pore and form the channel gate. The four short Phelices and the TXGYGD loops form the inverted teepee housing the K+-selective filter within. From Doyle et al. [1].
964 M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985
animals, although there is evidence for such a functionin plants [27,28]. In theory, monotypic ion fluxesthrough ion channels are far more effective in affectingthe membrane potential than the bulk concentration ofthat ion. For example, moving K+ out of a hypotheticalmicrobe, with 2 lm in length and 0.75 lm in diameter,to polarize it by 100 mV will only reduce its internal con-centration by mere 66 lM (only �3 · 104 ions).
4. The general features of K+ channels
4.1. The basic structure and activity
K+ channels are tetramers, made up of predomi-nantly identical subunits, although heterotetramers ofsimilar inter- or intra-molecular subunits are found inanimals. In the animal channel literature, this tetrameris also called the a subunit, or the pore-forming subunitof the channel, to distinguish it from other associatedauxiliary subunits (b, c, etc.) that serve to deploy or toregulate the main a subunit. A b subunit of a class ofanimal K+ channels is akin to oxidoreductase [29] andits homologues can also be found in the genomes ofprokaryotes.
The core of a K+-channel a subunit consists of twotransmembrane a helices, traditionally referred to asTM1 and TM2, flanking a short pore helix and theK+-filter sequence (TM, helical transmembrane seg-ment). The simplest TM1-P-TM2 subunit (P, ‘‘pore’’helix and filter loop) comprises only 94 amino-acid res-idues encoded by a virus (PBCV-1) found in a green alga
(Chlorella) that inhabits a paramecium (Paramecium
bursaria) [30]. This TM1-P-TM2 structure (2TM) withno additional peptide domains appears to be the mini-mum needed for permeation, filtration and gating. Thecrystal structures of the Streptomyces lividans K+ chan-nel, KcsA, show that the carboxy (C)-terminal ends ofthe four TM2s converge towards the cytoplasm and oc-clude the ion pathway [1,2]. This convergence is also re-ferred to as the ‘‘gate’’ [31], which is closed in this crystalstructure (Fig. 1). The TM1-P-TM2 core structure alsoexists in a class of K+ channels called inward rectifiers(Kir) found in both pro- and eukaryotes. Kir channelshave a similar architecture as KcsA though with slightlydifferent helical orientation and packing [32]. Althoughnot yet seen in microbial eu- or prokaryotic genomes,a type of subunits with TM1-P1-TM2-TM3-P2-TM4,forming the so-called two-pore-domain K+ channels,are found in animals. Such subunits covalently linktwo different TM1-P-TM2 core structures, presumablyforming dimeric channels. It seems likely that they arosefrom gene duplications to enforce an (a–a 0)2 type ‘‘het-erotetramer’’. Another common structural motif ofK+-channel a subunits is an S1-S2-S3-S4-S5-P-S6 (S,helical transmembrane segment) arrangement (6TM).This is often called the ‘‘Shaker’’ motif, named afterthe Drosophila mutation, the corresponding gene ofwhich was the first K+-channel gene cloned [33,34]. Inthis motif, the S5-P-S6 retains the characteristics seenin TM1-P-TM2 core of the 2TM channels describedabove. The S1-S2-S3-S4 helices function to regulate thiscore as described in Section 6.2. Several additional struc-tural forms occur rarely and have been discovered in
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 965
studies on unicellular eukaryotes such as ciliates andfungi. For example, fungi have K+ channels of an8TM motif (S1-S2-S3-S4-S5-P1-S6-S7-P2-S8) [35,36]and Paramecium has K+ channels of a 12TM (S1-S2-S3-S4-S5-P1-S6-S7-S8-S9-S10-S11-P2-S12) (Haynes,Saimi, and Kung, unpublished results) motif not previ-ously encountered.
As a direct consequence of being an ‘‘open pore’’ theactivity K+ channels are measured directly. Instead oftracing the flux of ions with isotopes a channel will passa species of ions at a rate that can be described in termsof the channel�s conductance (G), defined as the amountof current per unit driving voltage
G ¼ I=V ¼ 1=R;
where the conductance is in the unit of Siemens (S), theinverse of Ohm, the unit of resistance (R) and measuredby changes in current (I) and voltage (V). Techniquessuch as the patch clamp and the planar lipid bilayersare available, with which the conduction through indi-vidual channel proteins can be measured. The conduc-tance of each channel is referred to as its unitaryconductance (g) and usually ranges from 1 to 100 pS(10�12 S). Generally, the current through a channel willshift from zero (closed) to the open level set by its con-ductance and the applied voltage in an all-or-none man-ner as the channel protein migrates between its closedand open conformation (see Section 4.3). The transitionrates between the open and closed states are also molec-ular properties of the channel protein. Because thebehavior of individual molecules is registered, a rareopportunity in biology, the observation is made in amicroscopic world where the behavior of the unit parti-cles is described statistically and stochastically. Outsideof these situations where one contrives to study smallmembrane patches or rare reconstitution events, elec-trodes can be inserted into living cells to measure theensemble activity of the total sum of thousands of ionchannels, often of different types. In this configuration,the activities are observed in the macroscopic world,the world of smooth curves, familiar to most studentsof biology and biochemistry.
4.2. The K+ filter
The term ‘‘channel’’ as previously mentioned refers topathways that allow unobstructed passive flows of sol-utes. Channels are similar to enzymes in that there are‘‘sites’’ that preferentially interact with specific mole-cules or ions. In many cases, the opened gate acts asthe filter so that solutes of certain size and shape are al-lowed through. The acetylcholine receptor/channel inanimals and the MscL and MscS of prokaryotes arechannels of this type. K+ channels, in general, have aseparate filter structure apart from the gate itself. Thiswas anticipated by theoretical considerations [37], and
verified at atomic dimensions by the crystal structureof KcsA K+ channel [1]. The mystery of K+-channel fil-tration was that it is both highly efficient and discrimina-tory at the same time. This appeared to violate theenergetic principles of ion binding. In contrast to en-zymes with turnover rates of some 103 substrates persecond, many types of channels can turn over greaterthan 107 ions per second, yet at the same time, maintainhigh selectivity. A K+ channel can have a permeabilityratio of 300 to 1 in favor of K+ over the smaller Na+.The crystal structure of KcsA has neatly explained awaythis mystery (Fig. 1). (The form of this filter is not uni-versal for ion channels. The crystal structure of a ClCCl� channel from E. coli is constructed in a way thatis entirely different from a K+ channel [9], though thisprokaryotic ClC may not be of a classic channel type[38].)
We learn from KcsA that each K+ channel is madeby helices coming together from four separate subunitsto enclose an aqueous pathway at the center. It narrowsfrom the outside mouth into a selective filter of 2.0 Adiameter near the periplasmic surface [1,2]. The filter islined by horizontally distributed squares of carbonyloxygens of the canonical amino-acid sequenceTXGYGD from each of the four subunits. Pairs of thesequartets of partially negative oxygens surround the K+
ion much like the eight water oxygens surround it inaqueous solution. The hydration shells of ions need tobe removed so that the ions can be discriminated. Re-moval of hydration shells entails a very large energeticcost. Since the K+ filter is structured to precisely mimicthe hydration sphere of a K+ ion, there is little cost for aK+ ion to enter or exit the filter but a substantial cost forcations of different diameter such as Na+ or Ca2+. Thus,the structure neatly explains the channel�s high K+ selec-tivity and near-diffusion rate of throughput. In the 3-Dstructure, the central filter is cradled by the four porehelices in the form of an inverted teepee. Beneath theteepee is a water-filled cavity, which is sealed off at thebottom near the cytoplasmic side by the convergenceof the four TM2s when the channel closes. The TM2stherefore form the gate [31].
4.3. The gate and gating
Because the flow of ions through an open channel isenergetically wasteful, the channel gate is usually closed.Thus, the subunit tetramer must exist in at least two sta-ble conformations: closed and open (although in reality,biophysical analysis invariably reveals multiple closedand open states or substates). In both non-specific MscLand K+-specific KcsA, the conformational changes ofthe subunits during the channel opening expose a path-way similar to the opening of an iris in a camera. Interms of information, the result is an ‘‘on’’ or ‘‘off’’ sig-nal. The currently popular term ‘‘signal transduction’’
966 M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985
derives from ‘‘transduction’’, which means to convertone form of energy to another in its original sense inphysics. Ion channels are considered ‘‘transducers’’ inthe sense that they convert the energy of ligand binding,voltage change, or mechanical work into the opening ofion channels thereby transducing the external chemical,electric, or mechanical signal into electric or ionic signalsin the cytoplasm. These physically different parametersthat open the gates of channels are often referred to asthe ‘‘gating principles’’ or the ‘‘stimuli’’. Specific detailsof the gating mechanisms will be discussed in greater de-tail within each section comparing prokaryotic motifs ordomains to those known to affect gating in similareukaryotic channels (Sections 6.2–6.5).
In general, most proteins feel electric fields by way ofthe charged amino-acid residues. Voltage-gated chan-nels often have distinct patterns of charged residues par-ticularly sensitive to changes in membrane potential.This movement leads to a change in conformation,which tends to open the channel. The mechanism of li-gand gating is much like the activation of an enzyme,i.e. binding of a ligand by a ligand-binding domain in-duces a change from a closed (inactive) conformationto an open (active) conformation. Ion channels in ani-mals can be opened by a diverse array of ligands eitherexternally (glutamate, acetylcholine, H+, etc.) or inter-nally (ATP, cyclic nucleotides, Ca2+, etc.). Finally,stretch forces can be transmitted to embedded channelsthrough the lipid bilayer to open them. This type of gat-ing mechanism is seen in both bacterial channels, likeMscL and MscS, and some animal channels [39,40].
In the physiological literature, in association with thegating, one often describes the rise of the electric currentthrough channels upon a proper stimulus (gating princi-
a Various methods were used to assess channel activities. Direct measuremb Supports the growth of a K+-uptake-system deficient E. coli mutant (kuc Flux of 86Rb+ into liposomes.d Gain-of-function mutations of Kch render cells sensitive to K+.e The glutamate-binding core was crystallized.f The cytoplasmic C-terminus was crystallized.
ple) over time as an ‘‘activation’’ of the channels. Theterm ‘‘activation’’ remains in use today to indicate theincrease in channel opening probability and not the con-tinuous opening of the channel. For many channels, thecurrent subsides after the initial activation, even thoughthe stimulus is sustained. This process is referred to as‘‘inactivation’’. This is not to be confused with ‘‘deacti-vation’’, the process of current shut off after removal ofthe stimulus. Many different additional molecular mech-anisms have been determined behind various activa-tions, inactivations and deactivations (see [4]).
5. K+ channels in prokaryotes
In the realm of prokaryotic K+ channels, Milkman[41] extended the sequencing of the trp operon in E. coli,and encountered an open reading frame (kch) that isconceptually translated into a protein with the canonicalK+-filter sequence similar to Shaker-like K+ channels ofeukaryotes. Although the patch clamp has been success-fully applied on the native membrane of E. coli [42], noK+-specific electric conductance has been recorded todate, not even from kch mutants that likely have higherchannel activities [14]. The breakthrough in the crystal-lography of ion channels using a prokaryotic protein bythe MacKinnon laboratory has attracted an intenseinterest in prokaryote channels [1]. The prokaryoticK+ channels that are fully or partially crystallized and/or examined for their activities are listed in Table 1.The electric activities of prokaryotic K+ channels havebeen successfully recorded in six cases when the channelproteins are expressed heterologously in E. coli, yeast,Xenopus oocyte, cultured mammalian cells, or reconsti-
ents were made with electrodes.p1, DkdpABC5, DtrkA) in low K+ media.
Fig. 2. Types of prokaryotic K+ channels. The prokaryotic K+ channels are sorted by the number of transmembrane helix (TM), the regulatorydomains (highlighted in black), and homology to the studied K+ channels (in parentheses) as described in Section 5. Numbers indicate the totalchannels in that category identified from the 270 prokaryotic genomes. The basic K+-channel cores, TM-P-TM, are highlighted in gray. The channelswith KTN or RCK are the majority (Section 6.3) as well as the voltage-gated channels (Section 6.2).
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 967
tuted into artificial lipid membranes (Table 1). However,prokaryotic K+-channel activities in their native mem-branes have never been reported to date. Thus surveyingK+ channels of prokaryotes using electrodes does notappear to be practical.
A survey of the genome sequence information cur-rently being released provides information on the prev-alence and variation among prokaryotic K+-channelgenes. To identify the prokaryotic K+-channel genes inthe genomes, we used the BLASTP and TBLASTN pro-grams of the NCBI website (http://www.ncbi.nlm.nih.gov/sutils/genom_table.cgi) to blast the 270 pro-karyotic genomes, including 203 completed and 67whole-genome-shotgun finished sequences as of Decem-ber 2004. (We examined only finished genome sequencesto state more definitively than the homologues are pres-ent or absent). First, the K+-filter sequence of Kch (TIT-TVGYGDITP) was used as the initial query. Thetopology of similar sequences containing either GYGor GFG were examined using Kyte-Doolittle hydropa-thy analysis to determine whether the K+-filter sequence
is located as expected towards the C-terminus of a shorthydrophobic patch (the pore-helix) sandwiched betweentwo longer hydrophobic sequences (TM-P-TM). Themajority of the similar sequences fell into three groups:those that have only the two required TMs flankingthe pore, called 2TM, and those having four or occa-sionally two additional TMs immediately in front ofthe core, called 6TM or 4TM, respectively (Fig. 2).The conceptually translated products of open-readingframes that met these criteria were considered K+-chan-nel homologues.
Among these K+-channel homologues, some com-prise only the TM region, and some also containextended tails, beyond the TM region, either at the N-(amino-), C-, or both termini. The 2TM homologueswithout terminal extensions, including KcsA, are sortedinto the ‘‘K+-channel core’’ type (Fig. 2). Most of the6TM homologues without extended termini have recog-nizable evenly spaced arginine (R), lysine (K) orhistidine (H) residues at their S4, indicating a voltage-sensing property (see Section 6.2). To determine the
968 M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985
voltage-sensing feature, we aligned these sequences withClustalX program [43], and arbitrarily sorted the chan-nels that carry two or more R, K or H residues at S4into the ‘‘voltage-gated’’ type and the rests were placedin the category 6TM ‘‘core only’’ (Fig. 2). All of the4TM homologues were found to have no terminal exten-sions (Fig. 2). (Allowing sequencing mistakes, closeexamination of nucleotide sequences did not yield anyadditional TM segments.) This type of channel appearsto be unique to prokaryotic cells since there are cur-rently no known eukaryotic counterparts. Note thatthe K+-channel genes we identified here are based onthe criterion that they contain the K+-channel core(TM-P-TM) motif, and the K+ signature sequence. Ithas been proposed that the core motif or the differentcomponents (e.g. sensor or gate) of a K+ channel�s asubunit could come from two separate peptides (in anoperon) and still makes a function channel [44].
To sort the K+ channels with terminal extensions, weblasted the database again with the previously studiedK+-channel sequences in Table 1, and grouped thembased on their homology to known channels. Thereare homologues of the eukaryotic ‘‘inward rectifier’’,‘‘glutamate receptor’’, and ‘‘cyclic nucleotide-gated’’channels. A large number of 2TM or 6TM K+-channelhomologues carry a domain call KTN or RCK (see Sec-tion 6.3), including Kch and MthK, and are sorted into‘‘RCK/KTN domain’’ category (Fig. 2). A few 2TM
Table 2The distribution of various types of K+ channels among the 17 prokaryotic
a The classification is based on NCBI Taxonomy database.b Numbers of genomes that have be completely or shotgun sequenced as o
channels carry extended tail(s) that are not homologousto any known domain, and are sorted into ‘‘Unknowndomain’’ (Fig. 2). Interestingly, during the blasts usingthese full-length K+-channel sequences as queries, we re-trieved several channels that were not present when onlythe K+-filter sequence was used as the query. These newchannels have the similar topology or carry the same do-main as those above, but their K+-filter sequence are al-tered. Whether these channels still filter K+ remains tobe tested (see Section 6.1). The distribution of the vari-ous types of K+ channels among the 270 prokaryoticgenomes is summarized in Table 2. An extensive tableis provided in supplementary data (Appendix A). Notethat, these 270 sequenced genomes were not sequencedbecause they represent prokaryotic populations or theirdiversity. Most genomes of bacteria and archaea weresequenced, some repeatedly so, because of medical ormilitary interests, e.g. Haemophilus influenzae, Strepto-coccus pyogenes and Bacillus anthracis. Others were se-quenced because they are laboratory model organismsor because they are of special ecological or evolutionaryinterests, e.g. E. coli and Thermotoga maritima. Thoughnot completely representative, these 270 genomes docover the major prokaryotic phyla of both the Archaeaand Bacteria Kingdoms. Also note, while there is noclear consensus on microbial classification, our discus-sion here follows the NCBI Taxonomy (http://www.ncbi.nlm.nih.gov/Taxonomy/) [45].
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 969
The two major phyla of Archaea, Crenarchaeota andEuryarchaeota, contain members that carry K+-channelgenes. For example, Aeropyrum pernix, a hyperthermo-philic aerobic crenarchaeon, has a voltage-gated K+
channel (KvAP) that opens upon membrane depolariza-tion [46] (also see Section 6.2). Methanocaldococcusjannaschii, a hyperthermophilic anaerobic euryarchaeon,apparently makes use of a voltage-gated and two(2TM)-KTN K+ channels. Bacteria encompass the vastmajority of the known microbes and can be sorted phy-logenetically into at least 14 different phyla. Almost allmajor groupings include members that have K+-channelgenes in their genomes. Proteobacteria, the major phy-lum of gram-negative rods and cocci, branches into fiveclasses, alpha through gamma based on 16S rRNA se-quences. Each class includes species that have K+-channel genes in their genomes, e.g. Magnetospirillum
ovibrio vulgaris (d), Helicobacter pylori (e), and E. coli
(c). The majority of gram-positive bacteria belong tothe phyla Actinobacteria and Firmicutes that includesthe high GC and low GC species, respectively, that haveK+-channel genes, e.g. Streptomyces coelicolor, B.
anthracis. The oxygenic photosynthetic bacteria of phy-lum Cyanobacteria are all free living. Each of the over10 species of cyanobacteria whose genomes have beencompletely sequenced contains at least one K+-channelgene. Some have multiple such genes, e.g. Crocosphaerawatsonii (six), Anabaena variabilis (four), Trichodesmium
erythraeum (four). Bacterial phyla with fewer examinedmembers include the Bacteroidetes (e.g. Bacteroides
fragilis), the obligately parasitic Chlamydiae (e.g.Chlamydophila pneumoniae), the green sulfur Chlorobi(e.g. Chlorobium tepidum), the green nonsulfur Chloro-flexi (e.g. Chloroflexus aurantiacus), the Deinococcus-Thermus group (e.g. Deinococcus radiodurans), theFusobacteria (e.g. Fusobacterium nucleatum), the stalkedPlanctomycetes (e.g. Rhodopirellula baltica), the tightlycoiled Spirochetes (e.g. Tre. denticola) and the hyper-thermophilic Aquificae and Thermotogae (e.g. Aquifexaeolicus, T. maritima). In the genomes of the examplesgiven for these phyla each also contains one or moreK+-channel genes except Chlamydiae and Fusobacteria(Table 2).
Since K+-channel genes are found in all major taxa inBacteria and Archaea of the current sample of 270 gen-omes, the parsimonious hypothesis seems to be that K+
channel appeared very early before microbial diver-gence. Potassium channels could have emerged some 3billion years ago before Bacteria and Archaea diverged.This notion is bolstered further by the fact that the ex-tant bacteria considered to have branched off early fromthe root of the tree of life contain K+-channel genes.Aquifex is considered to be among the most deeplybranching bacterial genus on the basis of 16S-rRNAcomparison. Two K+-channel genes can be found in
the A. aeolicus genome [47]. T. maritima is also consid-ered a deeply branching species that evolves slowly[48]. It has one variant K+-channel gene (see Section6.1). These are hyperthermophiles that grow optimallyat 80–90 �C some without oxygen. These are conditionssimilar to what is thought to be those in which life firstappeared. Most studied archaea are extremophiles thatgrow optimally in low pH, reducing environments, hightemperatures and have K+-channel genes.
In addition to the widespread existence of these chan-nel genes, several comparative alignments of similartypes of K+ channels indicate a predominance of a ver-tical inheritance of these genes. While the current litera-ture indicates significant lateral transfer is widespreadamong prokaryotes [49–53], the sequences of the varioustypes of K+ channels seem to consistently cluster withrespect to the class of prokaryotes. Preliminary align-ments, available in supplementary data (Appendix A),indicate that generally the amino acid sequences aremore similar within the various classes of bacteria thanbetween them. There is also consistent clustering amongalignments made with the ‘‘core’’ region (TM-P-TM),the ligand binding domains or both together. The diffi-culty in providing definitive alignments is that the dis-tances between the sequences among the variousclasses are very large, the taxonomic grouping of bacte-ria is not consistent with respect to the distance betweensequences, and the evolution of transmembrane do-mains, particularly their addition, is not very wellunderstood.
6. Structures of prokaryotic K+ channels and their
eukaryotic counterparts
6.1. The K+ filter and its variations
As expected of the precise arrangement of atoms inthe K+ filter, the canonical filter sequence, TXGYGD,exhibits little variations among K+ channels previouslyanalyzed. Among animal channels, although the tyro-sine (Y) is sometimes replaced with a phenylalanine(F), this substitution is conservative since it preservesthe group of aromatic side-chains that surrounds andsupports the filter. We have randomly mutagenized eachof the three residues in the GYG tripeptide of E. coli�sKch and looked for variants that still preserve the K+-specific channel function. Only GYG and GFG werefound, indicating that other variations cannot supportthe filtration function of Kch channel [14]. This substitu-tion also does not alter the function of S. lividans KscA[54]. Viewing evolution as Nature�s vast experimenta-tion, we are not surprised to find GFG in the K+-filtersequence in disparate bacterial species including Ther-moplasma volcanium, Mycoplasma mobile, Magneto-
coccus sp. MC-1, St. coelicolor, to name a few. A
970 M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985
negatively charged aspartate residue, (D), usually trailsthe GYG or GFG sequence forming the external exitof the filter. A common variation is a replacement of thisaspartate with a glutamate (E). This is presumably also aconservative change since the negative charge is pre-served. Among the prokaryotic K+ channels, this D-to-E replacement is found almost exclusively in the(2TM)-KTN/RCK channels in many species of cyano-bacteria, in some species of proteobacteria, and in afew species of other phyla. In a few cases, the D is re-placed with a non-charged residue, including G, A, orS. I, T, N, and Q are also encountered at this position,but these replacements only happen together with otheralterations in the GYG sequence (Table 3). Although we
Table 3The variations of K+-filter sequence (GYGD) among the prokaryotic K+ ch
Organisms 2TM
Channel core RCK/KTN
Archaea
CrenarchaeotaSu. solfataricus P2 GLYS
Su. tokodaii str. 7 GLYA
Bacteria
ChlorobiChl. tepidum TLS GFSE
CyanobacteriaC. watsonii WH 8501 GFSE
Prochlorococcus marinus MIT 9313Synechocystis sp. PCC 6803 GYSD
Thermosynechococcus elongates BP-1 GYSE
Deinococcus-ThermusD. radiodurans R1
FirmicutesB. anthracis (4 strains) GDAN
Bacillus cereus ATCC 10987 GDAN
Bacillus cereus ATCC 14579 GDGN
Bacillus cereus G9241 GDGQ
Bacillus cereus ZK GDAN
GDGQ
Bacillus halodurans C-125 GLGD
Bacillus subtilis str. 168Bacillus thuringiensis str. 97–27 GDAN
Exiguobacterium sp. 255–15 GHPT
Ureaplasma parvum ATCC 700970
ProteobacteriaMagnetococcus sp. MC-1 GYQE
Burkholderia cepacia R18194 GASG
Ralstonia eutropha JMP134 SLGD
Ralstonia metallidurans CH34 SIGD
Desulfotalea psychrophila LSv54 GIAD
D. desulfuricans G20 GFQE
G. metallireducens GS-15 GFKE
Microbulbifer degradans 2–40 GMSE
Azotobacter vinelandii
ThermotogaeT. maritima MSB8 GYSI
The residues that deviate from the consensus, G(Y/F)G(D/E), are underline
do not know the functional significance of these filtervariations, the remaining parts of the sequence are sim-ilar in both putative transmembrane domains and regu-latory domains indicating that these are channel genes.
As stated above, A. aeolicus and T. maritima areextremophilic bacteria considered to be in lineagesbranched off early and slowly evolving based on 16S-rRNA and other molecular analyses. While the twoputative K+ channels of A. aeolicus have relatively con-served GYGE and GYGD filter sequences, that of T.
maritima has a striking GYSI sequence. Originally iso-lated from geothermally heated marine sediment, T.
maritima grows optimally at 80 �C. It has a 1.8 Mb gen-ome, encoding some 1900 genes, about half of which are
annels
6TM
unknown domain Voltage gated RCK core only
GHGD
EPIG
GLGD
GNGD
AYGD
GIGD
d.
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 971
of unknown functions. It has one of the highest percent-age (24%) of genes that are most similar to those of Ar-chaea. Its mosaic genome indicates extensive lateraltransfer of genetic material which is likely to have oc-curred early in the evolution of the two prokaryotic lin-eages [48]. Though it might have been a coincidence, wefind the severe deviation of T. maritima�sK+ filter is sug-gestive of an evolutionary relic. GYSI occupying the fil-ter position is not found elsewhere in our survey of the270 prokaryotic genomes and has not been reportedamong Eukarya. Our experimental results show thatsubstitutions of GYS for the GYG in the Kch K+ chan-nel of E. coli destroys the K+ filter [14]. Assuming it isfunctional, the GYSI filter is not likely an adaptationto high temperatures, since other known hyperthermo-philic archaea or bacteria have conventional filters andthe only other known similar variant (GYSD) is foundin one of the five K+ channels of Synechocystis sp.PCC6803, a fresh water cyanobacterium. Other deviantfilter sequences we encountered include a GFQE in Des-
ulfovibrio desulfuricans and a GFKE in Geobacter metal-
lireducens (Both are underground bacteria associatedwith iron corrosion. They belong to the class deltaprote-obacteria), a GYQE in Magnetococcus sp. MC-1 as wellas those listed in Table 3.
In parallel to the peculiar GYSI filter sequence of T.maritima is the filter sequences found in the channels ofhyperthermophilic and acidophilic genus Sulfolobus
[55,56]. Both of these Crenarchaeota, Sulfolobus solfa-
taricus and Sulfolobus tokodaii, encode a (2TM)-RCKtype channel with a GLYS or GLYA sequence whereone would expect to find the filter (Table 3). Theseorganisms are also considered to have branched offearly near the origin of cellular life. Regardless of evo-lutionary origin, it would be of interest to express theseputative K+-channel genes and directly measure theion-permeability ratios to see whether and how Naturemight have alternative designs from the one commonlydeployed in most organisms. While such a discoverywould be surprising, given the nearly universal filtermotif, there are several eukaryotic examples withintwo-pore domain K+ channels where substitutions ofthe aromatic residues still retains K+ selectivity [57].
6.2. The voltage sensor S4
Fifty years ago, a major break through by Hodgkinand Huxley explained the action potential as a feedbackbetween the membrane potential and the permeabilitiesof specific cations [58]. In today�s words, changes in volt-age across the lipid bilayer can open a class of channels,through which the flow of specific ions (e.g. K+) in turnchanges the voltage. These types of channel proteinsmust have electric dipoles and be able to detect andrespond to the electric field across them [59]. The firstK+-channel gene cloned corresponds to the Drosophila
mutant ‘‘Shaker’’ that lacks a depolarization-activatedK+ current resulting in a set of over-excitable behavioralphenotypes [33,34]. The Shaker protein is a 6TM chan-nel. Shaker can be heterologously expressed in frog oo-cytes conferring them a new channel current that areboth K+ specific and voltage sensitive. This findingmade clear that Shaker gene product alone likely housesboth the K+ filter as well as the voltage sensor. The volt-age sensor is a part of the channel bearing electriccharges that can move when the voltage changes, lead-ing the opening or closing of the channel. Indeed, beforeany channel protein structure was known, movement of‘‘gating charges’’ had been determined to be separateand ahead of the ion fluxes once the channel isopened [59,60]. The S4 of each Shaker subunitcontains a series of five arginines and two lysinesspaced regularly at every third residues, i.e.. . .RxxRxxRxxRxxKxxRxxKxx. . .. This motif of posi-tive charges in S4 helices was then recognized in manyvoltage-activated channels that were subsequentlycloned. A variety of experiments strongly suggest thatit is indeed a major part of the voltage sensor. In theShaker channel of the fly, mutation analyses showedthat the four R�s at the N-terminal end, but not theK�s and R at the C-terminal end, are the important gat-ing charges [61].
As shown below, many prokaryotic genomes containORFs that translate into Shaker-like K+-channel sub-units. Two Shaker-like prokaryotic K+ channels havebeen successfully examined electrophysiologically:KvAP, the depolarization-activated K+-specific channelof Aeropyrum pernix, and MVP, the hyperpolarization-activated K+ channel of Me. jannaschii. Both organismsare hyperthermophilic archea. KvAP, produced fromE. coli, has been purified and reassembled onto planarlipid bilayers and found to express as a K+-specific uni-tary conductance, 170 pS in magnitude [62] (Fig. 3A).The activation of the macroscopic KvAP currents(Fig. 3B), can be described with a Boltzmann distribu-tion (Fig. 3C) in which the transmembrane voltage par-tition the channels between their open and closed state,with �51 mV being the voltage at which half of thechannels are opened. MVP, expressed and examined inthe budding yeast Saccharomyces cerevisiae, is a 37 pSK+-specific channel that activates at voltage negativeto �100 mV. The MVP gene complements the defectsin K+-uptake mutants of yeast and E. coli, indicatingthat it is functionally expressed in both eu- and prokary-otes in vivo. Although the polarity, to which the MVPand its relatives respond, is opposite from that of mostother voltage-gated channels, various experimentsshowed that the MVP protein, including its S4 helix, isnot inserted backward. The current model is that thegate and/or its connection to S4 are arranged differentlysuch that an inward movement of S4 opens MVP butcloses KvAP and Shaker [63].
Fig. 3. The activity and structure of KvAP, the voltage-gated K+ channel from Aeropyrum pernix. (A) Single-channel records of unitary conductanceseen after the shift of the voltage from �100 to +100 mV. (B) After a capacitative artifact (first spike), a macroscopic K+ current is activated throughmultiple channels as the voltage is shifted over time from �100 mV to various less negative or positive levels (inset). The current can be inward(downward) or outward (upward) depending on the driving force. (C) The current/voltage plot of KvAP showing the rise of a macroscopic currentreaching half maximum at �51 mV. (D) A diagram of a Shaker-type voltage-gated K+ channel subunit, including the positive charges on the fourthtransmembrane helix. (E) A diagram of the tetrameric channel emphasizing the distinction of the S5-P-S6 K+-filter core and the S1-S2-S3-S4 voltageregulatory domain in the periphery. Modified from Ruta et al. [62].
972 M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985
As described above, the S5 and S6 of Shaker-typechannels are equivalent to the TM1 and TM2 of KcsA,being the helices that flank the short P helix and theTXGYGD filter sequence. The core structure of S5-P-S6 of the voltage-gated KvAP can be superimposed onthe TM1-P-TM2 of KcsA [64]. In both cases, the innerhelix (TM2 or S6) converges to form the gate. S5, how-ever, is connected to the rest of the helices, S1-S4 inKvAP (Fig. 3D). Shaker-type channel can then beviewed as having two parts: the S5-P-S6 core filter struc-ture, surrounded by a loosely fitting voltage-regulatorydomain composed of several transmembrane segments(S1 through S4; Fig. 3E). That these are indeed two sep-arate domains was shown in an experiment whereby achimera was constructed of the S1-S4 regulator of Sha-ker covalently linked to the TM1-P-TM2 of KcsA, re-sulted in a surprisingly functional voltage-activatedchannel [65]. If the S1-S4 domain was tightly coupledto the S5-P-S6 core through many specific bonds intoone unit, such an insect-bacterium chimeric moleculeswould not be expected to function.
The above findings are not without controversy. Thestructural constancy of the TM-P-TM core that housesthe high throughput, K+ filter is easily understood sincethe precise construction of the K+-specific filter imposes
rigid structural constrains. The S1-S2-S3-S4 regulatorydomain, however, is expected to make sizable conforma-tional changes during gating that include moving thegating charges across the membrane�s electric field.The flexible nature of this domain makes it difficult todetermine the structure, resulting in the current heateddebate. There are at least three major models at present.A popular model describes the S4 helix being completelysurrounded by other helices, as if isolated in its ownaqueous cannel – some called it a ‘‘canaliculus’’ – inwhichS4 can move perpendicular to the membrane in the elec-tric field like a screw [66–68]. The movement of the fourS4s in the channel tetramer is expected to drive the con-formational changes to open or close the gate. A secondmodel assumes a rotation of the S4 segment in an elec-tric field. Based on fluorescence-resonance energy trans-fer experiments and proton-accessibility studies, it isthought that the S4 segment is placed at the center ofa septum and the charged amino acid groups of the S4would travel through the large electric drop upon rota-tion [69,70]. The X-ray structure of KvAP, however,presents an entirely different third picture [64]. The S4with its . . .RxxRxxRxxRxxxxxRxxKxx. . . sequence isshown to divide into the S4 proper containing the firstfour evenly spaced arginines and a C-terminal S4–S5
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 973
linker helix through a loop that includes the last argi-nine. S3 is seen as two short helices S3a and S3b, con-nected through a loop. S3b and S4 proper runantiparallel forming a helix–loop–helix hairpin called‘‘voltage-sensor paddle’’. Most surprisingly, the fourS1s and S2s form the inner ring that encloses the S5-P-S6 K+-filter core, while S3s and S4s are at the outerperimeter. This structural model positions the paddleas a hydrophobic, cationic, helix–turn–helix structurethat appears to be in contact with the lipids and mustmove a large distance during gating. This location issupported by the observation that the tarantula venomtoxin, VSTX1 that attacks specifically the voltage sensorgain access through its partition into lipids [71]. How-ever, to keep the flexible voltage-regulatory domain inplace for crystallization, an antibody was applied, lead-ing to criticisms that the antibody might have pulled thenative arrangement of the helices out of place. A mostrecent study on KvAP with paramagnetic resonancespectroscopy after site-directed spin labeling indeedplaces S4 at the channel–lipid interface. In this model,the original S4 becomes two shorter helices, the first ofwhich contains the four evenly spaced arginines and islinked to the second half with the remaining argininesand lysines through a flexible linker. A twist at the linkeris proposed, placing all positively charged groups facingthe rest of the channel peptides and away from lipidhydrocarbons [72]. Although the X-ray structure ofKvAP is appealing, the movement of the S4 segmentwithin the lipid membrane expected from the theoreticalcalculations [73] is apparently much too slow to accountfor the rapid activation of the channel.
Although features contributed from other parts of thechannel may also be involved in voltage sensing, onlythe regularly spaced charged residues, mostly arginines,provide a reliable searchable feature in genome dat-abases. Examination of the 270 prokaryotic genomes re-vealed many ORFs that can be conceptually translatedinto proteins with six transmembrane helices, the fourthof which has the above feature. As expected, we notethat this feature is never encountered in any of the puta-tive 2TM prokaryotic K+-channel subunits. We also didnot find such feature in the 4TM type K+-channel sub-units. The 6TM Shaker-like channel subunits are wide-spread and found in most of the major prokaryoticdivisions. Thus the parsimonious hypothesis would bethat voltage-regulated K+ channels with the S4 chargeshave evolved early, although the caveat of lateral trans-fer remains.
Similarities of S4 sequences of amino-acid residues,including both the charged and the hydrophobic ones,can be found in related species in many cases. Such sim-ilarities can also be recognized even among genera with-in Alpha-, Beta- or Gamma-proteobacteria, etc. Noobvious sequence homology can be detected across high-er taxonomic levels except the regularly spaced positive
residues. The residues between each pair of positive res-idues, although not conserved, are nearly always hydro-phobic. Thus, the evolutionarily conserved pattern isonly the repeat of (+xx)n where x is hydrophobic, pre-sumably because this regular amphipathic feature aloneis crucial in voltage sensing. The (+xx)n sequence is of-ten disrupted with one three-residue unit in which theleading residue is not charged but is a small hydropho-bic residue. This disruption is found to occur at all posi-tions after the first (+xx) repeat. In a few cases, the(+xx)n series is disrupted with one additional residuebefore returning to the three-residue repeats. A repre-sentative collection of these S4 sequences is shown inTable 3.
As expected from the vast diversity among prokary-otes, there are many more variations in the S4s than pre-viously encountered among animal species. Differentspecies are found to house one, two, three, four, five,six, seven, or even eight regularly spaced positivelycharged residues. Those with only one such repeats arenot confidently assigned as Shaker like in our survey(The 6TM ‘‘core only’’ type in Fig. 2). Among bacteria,we found the first four charges tend to be Rs, i.e. in theseries of . . .RxxRxxRxxRxxR/Kxx. . .. At present, thereare not enough archaeal examples to discern a pattern.Though arginines and lysines are most common, somehistidines are found. Jiang et al. [74] gave three reasonson why the S4s of Shaker and KvAP have predomi-nately R and not K. The survey of prokaryotic K+ chan-nels shows that, in general, R�s are indeed morecommonly found than Ks, e.g. Lactobacillus johnsonii
has a . . .RxxRxxRxxRxxRxxRxxRxx. . . sequence.However, there are numerous exceptions, e.g. Clostrid-ium acetobutylicum has a . . .RxxKxxKxxKxxKxxKxxKxx. . . series in S4. Many varieties between these ex-tremes can be found. No pattern on K placement canbe discerned; K can appear anywhere in the series,including the N-terminal most position (see Table 4).
Genomes of soil- and water-borne free-living bacteriahave received less attention in comparison to pathogenicand extremophilic organisms. Although such free-livingbacteria do not necessarily have Shaker-like K+-channelgenes, the majority of the species that have them tend tobe free living in sand, in soils, in waters (fresh, brackish,or sea), or on animals and plants, environments that aresubjected to variation. Some 6TM-containing speciesare found in the oral cavity, vagina, the lower intestine,or in plant tissues. While these can be commensals oreven opportunistic pathogens, none are intracellularparasites. There is no discernable pattern between thepresence of these likely voltage-gated channels and theorganisms� metabolic habits or growth habitats: Someare photosynthetic, others are heterotrophic; some areaerobes, others are facultative anaerobes; some preferhigh temperatures, others low. In addition, related spe-cies may or may not have Shaker-like K+-channel genes.
Table 4Variations in the S4 motif among selected voltage-gated K+ channel inprokaryotes
a The S4 of Drosophila Shaker is listed for comparison.b Cyanobacteria.c Firmicutes (Bacilli).d Bacteroidetes.e Proteobacteria (gamma).f Firmicutes (Clostridia).g Actinobacteria.h Crenarchaeota.i Euryarchaeota.j Deinococcus-Thermus.
974 M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985
For example, E. coli has no such gene, while the verysimilar Salmonella typhimurium has one.
6.3. The RCK and KTN domains
Unlike the voltage sensor within S4 segment, the do-mains that responsible for ligand gating are present out-side the TM region, and usually trail the last TM (TM2or S6) at the C-terminus [75]. One of these ligand-bind-ing domains, containing the Rossmann fold [76], arefound predominately among the prokaryotic K+ chan-nels we have identified. The first evidence that the Ross-mann fold, or NAD-binding motif, is involved in theregulation of cellular K+ came from a study on the E.
coli TrkA subunit of the Trk K+-uptake system [77].Shortly after, similar structure motifs were recognizedin various K+-transport systems, including KtrA, a dis-tant relative of TrkA [78], KefB and KefC, the K+-effluxsystems of E. coli, Kch K+ channel of E. coli, and theSlo K+ channels of Drosophila and mouse [78,79]. Lateron, the domains that comprises the Rossmann fold andbinds NAD were named KTN or TrkA-N domain in thePfam protein domain family database (Accession num-ber: PF02254) [80]. A characteristic motif, GXGXXG,
in the Rossmann fold serves as an indicator forNAD-binding. The Rossmann-fold is composed of two(b-a-b-a-b) portions linked by a a helix in a (b1-a1-b2-a2-b3)-a3-(b4-a4-b5-a5-b6) arrangement.
The structures of two KTN-domain fragments ofKtrA proteins from Me. jannaschii and B. subtilis wereresolved individually at atomic resolutions [81]. As re-vealed from the X-ray structures, KTN domain com-prises a Rossmann-fold core with a a6 extension. TwoKTN monomers form a dimer as a handshake throughthe a6 extension. In the crystals, NAD+ or NADH actu-ally binds to each monomer. It was proposed that uponbinding to different ligands the two monomers move rel-atively to each other by way of a ‘‘hinge’’, at the shortloop between b6 and a6. This hinged movement wasextrapolated to explain the gating of a K+ channel. Inthe proposed model, two of the KTN dimer form a func-tional tetramer. A ligand-mediated movement of the do-mains is coupled to the bottom gate of K+ channelthrough a short linker, and, thereby, regulates the chan-nel activity [81].
In the case of Kch and Slo K+ channels, there is noevidence that their Rossmann-fold domains bind toNAD+ or NADH, and the GXGXXG motif found inthe KTN domain is absent. Therefore, such a domainis given another name called ‘‘RCK’’ for regulatingthe conductance of K+ to distinguish it from KTN[82]. The RCK structures of Kch in E. coli, and of MthKin M. thermautotrophicus have been resolved at atomicresolutions [82,83]. Both of these X-ray structures reveala common Rossmann-fold conformation and dimericarrangement as the KTN domains. The MthK proteinwas crystallized in its full length with a Ca2+ bound toeach RCK domain, and the channel is shown to be acti-vated by millimolar-range of Ca2+ [83]. However, thephysiological relevance of this finding is questionable,since eukaryotic cytoplasms use micromolar Ca2+ as sig-nals and the role(s) of Ca2+ in prokaryotes are not yetclear. If this Ca2+ binding is a crystallization artifact,then the true or the additional ligand of the MthK�sRossmann fold is still to be found. The MthK genehas an internal start codon, which results in separateexpression of the RCK domain only. Based on the crys-tal structure of MthK and other biochemical evidences,it was concluded that a functional MthK K+ channelcomprises eight RCK domains, and four covalentlylinked to each TM2 though a 17-residue linker, and fourfree RCK peptides co-assembled from cytoplasm [83]. Inthis model, the eight RCKs form a ‘‘gating ring’’. Bind-ing of ligands to the RCKs changes the conformation ofthe gating ring, and in turn pulls open the bottom gateformed by the TM2s through the 17-residue linker [83].
Besides Kch and MthK, the K+ channels carryingRCK or KTN domain are commonly found and wide-spread among the 270 prokaryotic genomes (Fig. 2).The topological presence of these domains in prokary-
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 975
otic K+ channels varies. They are found in both 2TMand 6TM channels but not 4TMs. Mostly there is onlyone such domain attached to the C-terminus of thechannel body, like Kch or MthK, but tandem repeatsat the C-terminus are also encountered in some 2TMchannels. In a few cases, these domains are present atboth N- and C-termini of 2TM channels. Interestingly,all of the domains found in the 6TM channels so farare all RCK (the GXGXXG motif is absent), while bothRCK and KTN are found in the 2TM channels. The(2TM)-KTN/RCK channels are the majority of pro-karyotic K+ channels we have identified (Table 2). Theyare widespread among prokaryotes, but are not found inany eukaryotic genomes to date. (Although the RCKdomain can be found in animal Slo K+ channels, Slochannels are 6TM, not 2TM.) This suggests a commonfunction of these channels in prokaryotes and possibleancient nature of this type of K+ channels.
It is interesting to point out that among the five com-pletely sequenced Mycobacteria (of Actinobacteria),Mycobacteria leprae TN is the only one that does notcarry any K+ channel gene. M. leprae TN is consideredhaving gone through reductive evolution since its gen-ome size (�3.3 Mb) is less that the average (�4.4 Mb)and it carries more than 1100 pseudogenes [84]. In fact,a (2TM)-KTN channel gene, which are homologous tothose carried by other Mycobacteria, is inactivated andbecame a pseudogene (NCBI GI: 15826865) in M. lep-
rae. As the environment changed, the genes that areno longer required for living in certain niches are inacti-vated, and then eliminated by deletions. The fact that aK+ channel gene is one of the >1100 pseudogenes in M.
leprae suggests that those active K+-channel homo-logues, carried by other Mycobacteria, do perform aspecific function that is no longer needed for M. leprae
TN (also see Section 7.1).
6.4. The cyclic-nucleotide binding domain
Cyclic-nucleotides, cAMP in specific, acts as signalingmolecule in E. coli to control gene expressions throughthe well-known CAP, catabolite gene activator protein.In animals, cAMP and cGMP act as secondary messen-gers to regulate numerous cellular signaling pathwaysand enzyme activities. These molecules are also knownto activate or modulate the activity of some cation chan-nels through a cytoplasmic C-terminal cyclic-nucleotidebinding domain, CNBD. In animals, these types of cat-ion channels are classified into two subfamilies, HCN,hyperpolarization-activated, cyclic nucleotide-gated,and CNG, cyclic nucleotide-gated. The CNG channelsare mainly activated by cyclic nucleotides, while theHCN channels� activation depends on the membranepotential and the binding of cyclic nucleotides shifts thisvoltage dependency. The CNG channels do not have the
K+-filter sequence (TxxTxGYGD) at the pore regionand permeate K+ as well as other mono- and di-valentcations. Similarly, the filter sequence of HCN channelsalso deviates from the consensus (e.g. HxxCxGYGRof mouse HCN2), and discriminates K+ poorly fromother alkali cations.
The structure of the cytoplasmic C-terminus of themouse HCN2 channel was solved at 2.0 A resolution[85]. It revealed a two-domain conformationalarrangement, in which the CNBD is connected tothe end of S6 through an 80-residue C-linker domain.Note that the linker between RCK to the TM2 ofMthK is only �17 amino acids, and was not resolvedin the X-ray structure. Similar to those NAD-bindingdomains that have a conserved core structure (theRossmann fold), the domains that bind cyclic nucleo-tides have a jelly-roll b barrel formed by eight b-sheetsin common. It was proposed that, in a functionalchannel, four CNBDs coming from each a-subunitform a tetramer. Binding of cyclic nucleotides changesthe confirmation of the CNBD tetramer and thischange is coupled to the channel gate through theC-linker domain [85].
Recently, a prokaryotic K+ channel, MloK1, con-taining a C-terminal CNBD has been identified fromMesorhizobium loti, a symbiotic N2-fixing soil bacteriumof Alphaproteobacteria. As demonstrated by 86Rb+-up-take assay, this channel is activated by either cAMP orcGMP, and it discriminates K+ and Rb+ from Na+
and Li+ [86]. The structure of the cytosolic C-terminalregion of MloK1 has also been solved at 1.7 A withand without cAMP bound [87]. The CNBD of MloK1is folded in the way similar to that of mouse HCN2 ex-cept for the C-linker domain, which is absent in MloK1.Based on the crystal structures and other evidences, adifferent gating model was proposed. In this model,the four CNBDs form a dimer-of-dimer arrangement in-stead of a tetramer [87].
Unlike the KTN/RCK-containing K+ channels thatare widespread among the 270 genomes, the CNBD-containing K+ channels are rare and restricted. Onlysix cases are found to date. Aside from M. loti, the otheralphaproteobacteria, M. magnetotacticum, and Rhodo-
pseudomonas palustris are each equipped with one suchgene, and Bradyrhizobium japonicum carries two. Theonly other case is the cyanobacterium, Tr. erythraeum,which carries one such gene. Interestingly, besides theCNBD-containing K+ channels, each of these bacteriaalso carries one or more K+-channel genes of othertypes, e.g. T. erythraeum also carries IRK, GluR, andvoltage-gated K+ channels. Moreover, while KTN/RCK domain shows topological variations among theprokaryotic K+ channels, the CNBD is only found inthe 6TM channels, and all at the C-terminus. We havenot encountered CNBD in tandem repeat, in the N-terminus, nor in a 2TM channel so far. Similar to their
976 M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985
eukaryotic HCN homologues, these six CNBD-contain-ing K+ channels all have the (+xx)n voltage-sensing se-quence in their S4, in which the n is varied from 2 to6. While the C-linker domain of MloK1 and the otherfour channels in the same Alpharoteobacteria branchare absent, that of the channel in T. erythraeum is pres-ent as the eukaryotic HCN or CNG channels. Thecanonical K+-filter sequence (TxxTxGYGD) are wellpreserved among these prokaryotic CNBD-containingK+ channels indicating they have higher K+ selectivitythan their eukaryotic homologues. In a protein phylo-gram built with these prokaryotic K+ channels and ani-mal CNG and HCN channels, the five channels ofalphaproteobacteria cluster in one branch, away fromthe one of T. erythraeum with the CNG and HCN chan-nels. These observations suggest that the CNBD-containing K+ channel of T. erythraeum was eitheracquired through horizontal gene transfer or is an ances-tor of the eukaryotic CNG and HCN channels. The evo-lution of this type of K+ channels may become clearwhen more cases are identified in the future.
6.5. The glutamate receptor homologues
Homologues of animal glutamate receptors havebeen identified in prokaryote. The gene for the first pro-karyotic glutamate receptor, GluR0, was discoveredfrom Synechocystis sp. PCC 6803 [88], a model cyano-bacterium isolated from a fresh water lake. The gluta-mate receptors (GluR) in animals are different fromthe classical K+ channels in several respects. They arecomposed of three transmembrane helices with the P-loop located between the TM1 and TM2. They havean inverted membrane topology in which the proteinplaces its N-terminus extracellularly, and the P-loop fac-ing cytoplasm. The P-loop of animal GluRs does notcontain the canonical K+ filter sequence, and as a result,these channels are only cation selective, not K+ specific.The glutamate-binding cores is formed by the extracellu-lar N-terminus (domain 1) and loop between TM2 andTM3 (domain 2), and the ligand-binding site is locatedin the cleft between these two domains as revealed fromthe X-ray structure [89]. Structurally, this core confor-mation is conserved between GluRs and the prokaryoticperiplasmic glutamine-binding protein (GluBP) [90].The glutamate receptors discovered from prokaryotesare different from their eukaryotic homologues bothtopologically and electrophysiologically. In GluR0, theP-loop between TM1 and TM2 preserves the canonicalK+ filter sequence and, therefore, GluR0 favors K+ overother cations [88]. The TM3 found in animal GluRs ismissing in GluR0, but the two domains that form theglutamate-binding core are preserved, and structurallyconserved [91]. Therefore, the prokaryotic glutamatereceptor has the appearance of a fusion of the ligand-binding core from GluBP and a KcsA type K+ channel.
Besides GluR0, there are only four other homologuesencountered in our genome survey so far. The organismsthat carry this type of K+ channel are all multiple K+-channel gene carriers and are all cyanobacteria(C. watsonii, Prochlorococcus marinus str. MIT 9313,Synechococcus sp. WH 8102, Synechocystis sp. PCC6803, and T. erythraeum). Interestingly, among the threeP. marinus strains been completely sequenced (MIT9313, CCMP1375, and CCMP1986), MIT 9313 is theonly one that has the GluR K+ channel. These threestrains represent different oceanic niches with drasticallydifference in light and nutrients supplies [92,93]. Thenumbers and the types of K+ channels they areequipped with and the environments in which they livemay hold clues of the functions of the K+ channels thatthey carry.
7. A natural history of prokaryotic K+ channels
The combination of powerful and incisive genetic andelectrophysiological techniques allows channel biolo-gists to dissect structure–function relationship of indi-vidual channel proteins in exquisite details. Althoughthe knowledge gained from this approach can often beintegrated to our understanding of specific cell physiol-ogy of animals, fixation on a leaf often obscures ourview of the tree, let alone the forest. Here the forest –or is it a jungle? – is enormous. Animals, from mammalsto sponges, are but a small part of the diversity on earth.What can we learn from the opposite approach, i.e. todiscern patterns from the eyes of a bird, or an astronaut?The following is an attempt to gain some insights intothe possible roles of K+ channels in prokaryotes bycomparing the lifestyles of the organisms with the pres-ence, the number, and the kinds of K+ channels theyhave. Given the large variations over the entire rangeof myriad bacteria, variations among those even withina single lifestyle, say, parasitism, can be enormous.There are parasites of mammals, insects, plants, or otherbacteria. There are obligate or occasional parasites, ex-tra- or intracellular parasites, etc. Even with exceptions,however, some general trends can be detected.
7.1. Parasitism, genome downsizing, and the loss of K+
channels
Mitochondria (genome size 0.006–2.5 Mb) and chlo-roplasts (0.09–2 Mb), organelles of endosymbiotic bac-terial origin, do not encode their own K+ channels,nor are K+ channels found in Buchnera aphidicola
(0.62–0.64 Mb) or Wigglesworthia glossinidia (0.7 Mb)that dwell in the cytoplasm of specialized bacteriocytesof aphids or tsetse flies, respectively [94–97]. The gen-omes of these bacteria are so reduced that they seemto only exist to synthesize certain nutrients to supple-
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 977
ment the hosts� poor diet. Bacteria that are obligateintracellular parasites also do not have their own K+
channels. These include the members of Chlamydiaeand Rickettsiales, e.g. Ch. pneumoniae (chronic infec-tions, 1.2 Mb) [98–101], Rickettsia species (rickettsias,1.1–1.3 Mb) [102–104], and the other pathogenic species,e.g. Tropheryma whipplei (Whipple�s disease, 0.93 Mb)[105,106], Coxiella burnetii (Q fever, 2.0 Mb) [107], M.
leprae (leprosy; 3.3 Mb) [84]. Note that these bacteriaare evolutionarily diverse, i.e., obligate associationstook place many times in different bacterial lineages.Even a parasitic K+-channel free archaeon exists (Nano-
archaeum equitans) [108], though no archaeal animalpathogens have been reported. It appears that, as thehost–guest relationship became more and more intimateand the guest genomes became smaller and smaller, theK+ channels of the ancestral free-living forms were jetti-soned. Gene reduction and genome down-sizing appearto take place when the environment is protected orstable [109].
Many extracellular parasitic bacteria also do not en-code recognizable K+ channels. These include the obli-gate extracellular pathogens such as Treponema
pallidum (syphilis, 1.1 Mb) [110], as well as the faculta-tive parasites or commensals such as Campylobacter je-
juni (diarrhea, 1.6 Mb) [111], Haemophilus influenzae
species (gastrointestinal and pulmonary infections, 2.6–2.9 Mb) [113–116], Neisseria meningitidis (meningitis,2.2–2.3 Mb) [117,118], Clostridium species (gas gan-grene, tetanus, 2.9–3.1 Mb), Agrobacterium tumefaciens
(crown galls of plants, 5.7 Mb) [119]. Interestingly, thepredatory bacterium Bdellovibrio bacteriovorus (3.8Mb) [120], which invades and consumes gram-negativepreys one at a time, also does not encode its own K+
channels. Other extracellular pathogens, however, arefound to have retained their own K+-channel genes.These include Tre. denticola (periodontal disease, 2.8Mb) [121], H. pylori (stomach ulcer, 1.6–1.7 Mb)[122,123], Mycobacterium tuberculosis (TB, 4.4 Mb)[124,125], Shigella flexneri (bacillary dysentery, �5Mb) [126,127], Yersinia pestis (the plague, 4.8 Mb)[128], and Xylella fastidiosa (disease of grape-vine andother plants, 2.7 Mb) [129]. These extracellular patho-gens that retain their K+ channels can usually be cul-tured in the laboratory, but whether each of themactually grow naturally in environments outside theirhosts would require further detailed analyses beyondour scope here.
Well-known free-living bacteria have K+-channelgenes. For example, the most studied gram-negativebacteria E. coli (4.6–5.5 Mb) has one such gene (kch)and Salmonella typhimurium (5.0 Mb) has twoSTM4272 and STM1741 [130]. We also found that mul-tiple K+-channel genes in the larger genomes of free-living and metabolically complex and versatile bacteria.
These include St. coelicolor (an antibiotic-producing soilbacterium with five such genes, 9.1 Mb) [131], Brady-rhizobium japonicum (a N2 fixing bacterium with three,9.1 Mb) [132], Methanosarcina acetivorans (a versatilemethanogenic archaeon with five, 5.8 Mb) [133], Pseudo-monas aeruginosa (a ubiquitous environmental bacte-rium with two, 6.3 Mb) [134], Enterococcus faecalis (ingut, soil, etc. with two, 3.4 Mb) [135], and A. aeolicus
(chemolithoautotroph and hyperthermophile with two,1.6 Mb) [47]. The photosynthetic cyanobacteria all havemultiple K+-channel genes except strain CCMP1986 ofProchlorococcus marinus which has only one [93]. Thecase of Rhodopseudomonas palustris is especially clear.This proteobacterium has a 5.5-Mb genome thatcontains three K+-channel genes. The versatility of thisbacterium is truly astounding. It is capable of chemohet-ero-, chemoauto-, photohetero-, and photoautotrophicmetabolism, i.e. able to consume or generate complexcarbohydrate and able to derive energy from chemicalreactions or from sunlight [136].
The above trend is especially clear when certainspecies are compared. In the same genus, the speciesthat are more like ‘‘free-livers’’ have K+ channels,while the species that are more like parasites do not.For example, a K+-channel gene is found in Mycobac-
terium tuberculosis (TB, 4.4 Mb) [124,125], which canbe cultured in the laboratory with ordinary media,while a similar channel gene is inactivated in M. lep-
rae (leprosy, 3.3 Mb) [84], which can only grown ex-tremely slowly in armadillos, besides human patients(also see Section 6.3). Similarly, Clo. acetobutylicum
[137], a free-living bacterium that produces solvents(4.1 Mb) has one K+-channel gene, but C. perfringens(gangrene, 3.1Mb) [138] or C. tetani (tetanus, 2.9Mb)[139] have none.
Not all free-living species have K+-channel genes,however. Thermoanaerobacter tengcongensis (2.7 Mb)[140] and Thermus thermophilus (2.1 Mb) [141], twounrelated hyperthermophiles do not have recognizableK+-channel genes. Nor do Nitrosomonas europaea, achemolithoautotroph that oxidizes ammonia (2.8 Mb)[142] and F. nucleatum, a dominant oral bacterium(2.2 Mb) [143]. Without knowing more about their com-parative ecology, it is not clear whether these organismsare adapted to restricted environments that change little.There are other hyperthermophiles or oral bacteria thatdo have K+-channel genes, although we do not knowwhether these other organisms also thrive freely in otherenvironments. Furthermore, organisms that overlap intheir niches do not necessarily have the samemetabolism.
In short, although it is difficult to draw solid conclu-sions from an overview of the most vastly diverse organ-isms on earth, there does seem to be the trend thatK+-channel genes are found in free-living species and thatmultiple K+-channel genes are found in the genomes
978 M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985
of those that are metabolically complex and versatile.One possible suggestion from this trend therefore is thatK+ channels may function in a large variety of prokary-otes to assist metabolic changes upon environmentalchanges.
It is currently fashionable to define the ‘‘minimal gen-ome’’, i.e. to enumerate the smallest gene set that sup-port growth. Our survey suggests that K+-channelgene is unlikely to be a member of this minimal set.Note, however, that ‘‘growth’’ here means multiplica-tion in a rich medium at ‘‘reasonable’’ laboratory tem-perature, pH, with oxygen, etc. Such a putative‘‘minimal organism’’ will presumably not compete wellin nature. Adaptations to Nature�s vagaries will likelyrequire many other genes, probably including the onesthat encode K+ channels.
7.2. The distribution of K+-uptake pumps vs. K+ channels
K+ is the major cation in the cytoplasms of all cells,bacterial, archaeal, or eukaryotic. Given the vast varia-tions of life forms, this universality is very striking. Thatno Na+-rich cytoplasm is known is certainly not becauseNa+ is rare. Even fresh water is rich in Na+. PresumablyK+ must have been chosen early because of its compat-ibility with the ‘‘Ur Replicator’’, probably ribozymes.Today, K+ is considered a ‘‘compatible solute’’ foundto interfere little with the structures of macromoleculesin the cytoplasms, even at high concentrations comparedto Na+ because of its surface charge density. K+ is a ma-jor osmolite that provides the turgor for growth of allcells, animal cells included. In free-living cells, such asE. coli, which cannot rely on a host or a kidney to bufferits environmental water content, adjusting its internalK+ is a major strategy in long-term survival in face ofosmotic changes as review above. E. coli takes up K+
upon osmotic upshifts by immediately increasing theactivity of its constitutive low-affinity, high-capacityTrk system to pump K+. Upshifts also induce the tran-scription and production of its high-affinity, low-capacityKdp system that is capable of scavenger K+ from K+-depleted environment. E. coli is also equipped with athird K+ pump, Kup.
Kdp, Kup, Trk and its closely related Ktr K+ pumpsare recognizable by their sequences. In the one directionof lifestyle progression above, we distinguish free-livinghabit, commensalism, facultative or obligate extracellu-lar parasitism, intracellular parasitism or symbiosis, andorganelle status. We found that most intracellular para-sites with greatly reduced genomes that do not have K+
channels also do not have K+-uptake mechanisms.Examples are Buchnera aphidicola, Tropheryma whipplei,Chlamydia species, Rickettsia species, Coxiella burnetii,and M. leprae described above. It is interesting to notethat the hypertheromophilic archaeon Su. solfataricus,
which is also an extreme acidophile that grows at pH2–4, has a K+-channel homologue with a peculiar filtersequence also do not have a recognizable K+-uptakesystem. This organism apparently maintains an internalpositive membrane potential presumably by passive dis-tribution of K+ to prevent the acidification of thecytoplasm.
There are also many disparate extracellular patho-gens that no longer have recognizable K+-channel genesnonetheless retain their homologues of Trk, Ktr, Kdp,and Kup. For example, Campylobacter jejuni, Trepo-
nema pallidum, Haemophilus influenzae, Staphylococcusspecies, Neisseria meningitidis, Clostridium perfringens,C. tetani, Agrobacterium tumefaciens, all contain Trkor Ktr homologues and most even have additionalKdp and/or Kup homologues. The exceptional free-living organisms that do not have K+-channel genesnoted above, such as the thermophils Thermoanaerob-
acter tengcongensis and Thermus thermophilus, thechemolithoautotrophic Nitrosomonas europaea, the oralbacterium F. nucleatum, all also have Trk homologues.The predatory Bdellovibrio bacteriovorus has Kup andKdp homologue. Even the extreme endosymbiont Wig-
glesworthia glossinidia in tsetse fly encodes Trk and Kuphomologues in is minute genome (0.7 Mb).
Free-living archaea or bacteria that occupy differentniches, from the well known to the more obscurelisted above, all have recognizable K+-uptake mecha-nisms, usually Trk or Ktr, though some have Kupor Kdp homologue also or instead. Even A. aeolicusand T. maritima, the bacteria judged to have branchedoff very early and evolved slowly, have Ktrhomologues.
The trend that many species retain their K+-uptakemechanisms but forgo their K+ channels suggests thataccumulating a high internal K+ concentration is impor-tant to the identity of the organism and that K+ chan-nels are not important in this accumulation. Perhapsgiving up the K+ gradient across its membrane is a signof final commitment of the parasite (organelle) to be apart of the host. For the purpose here, the trend seemsto indicate the different physiological roles of the uptakesystem and the channels.
7.3. A speculation on the possible function of K+ channels
in prokaryotes
Given the vast diversity of prokaryotes, it is highlyunlikely that the various K+ channels serve only onefunction in all prokaryotes. Much work lies ahead, sincewe do not even know for sure the function of one suchchannel in one bacterium to date!
In the case of Kch, the RCK-containing 6TM K+
channel of E. coli, limited evidence suggests that it isnot for the bulk uptake of K+, but for the adjustment
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 979
of the membrane potential during certain changes in theenvironment yet to be defined. Extensive searches forK+ auxotrophs, at progressively lower and lower exter-nal K+ supply, uncovered the multiple K+ uptakemachineries Trk, Kdp, and Kup [144–146]. If Kch isfor the uptake of K+, it seems unlikely that it wouldnot have turned up in these exhaustive searches. In-stead, as reviewed above, Kch was first discovered byits sequence similarity to the Shaker channel [41], andwhen mutations were engineered into kch, includingdeletion, there is no discernable K+ auxotrophic pheno-type [17].
Certain mutations, judged to be ‘‘gain-of-function’’mutations, led to K+-specific sensitivity instead, likelydue to uncontrolled opening of the Kch channel [14].The K+-sensitivity phenotype of these special ‘‘loose-cannon’’ mutants implies that K+ passively filtered intothe bacterium, an artificial form of uptake under high[K+]out, is in fact toxic. A simple calculation based onthe cell geometry and the specific capacitance of biolog-ical membrane will show that K+ flux alters the mem-brane potential significantly long before it has anysignificant effect on the bulk [K+]in (Section 3). Wefound that, in liquid media, increasing external K+byabout ten times from �0.5 to �5.5mM (approximatelyfrom 10�4 to 10�3 M) had no effect on the wild typebut stopped the growth of the GOF mutant. A concom-itant increase of the external H+ by about ten times from�pH 7.3 to �5.8 (approx. 10�7 to 10�6M) restored themutant�s growth to near normal [14].
It seems likely that the GOF-mutant channels openuncontrollably to increase the K+ permeability, and thusclamp the membrane potential (DW) towards the equi-librium potential of K+ (EK, EK = 58log[K+]out/[K+]in). Raising [K+]out from 0.5 to 5.5 mM loweringEK from � � 150 mV to � � 90 mV so as the DW ofthe mutants. The shallow DW could reduce the protonmotive force (Dp) below the level that needed for a grow-ing cell to drive cellular processes (e.g. ATP synthesis);therefore, stopped cell growth. Supporting this scenariois that a unit lowering in pHout is able to counteract thedegree of DW depolarization, and to keep Dp un-changed. This finding indicates that the K+ sensitivityof the GOF mutants is due to losing membrane poten-tial and thus the proton motive force. It seems likely thatthe wild-type Kch functions to regulate proton motiveforce by changing the membrane potential upon certainenvironmental or metabolic changes.
We are a long way from being certain that mem-brane-potential adjustment is indeed the major functionof K+ channels in prokaryotes. In the event that this canbe proven for both bacteria and archaea, however, thismay become the answer of the fascinating question ofwhy and how cation channels appear, which eventuallyevolved into the entities so important to our brain andpsyche.
8. Concluding remarks
The field of prokaryotic ion channels has developedlargely independent of the rest of microbiology. Theintellectual origin of ion channels from the concerns ofthe nervous system has led to techniques and conceptsforeign to most microbiologists. At present the greatcontributions of prokaryote channel crystallography toour understanding of molecular structures and mecha-nisms are in great contrast with the poverty of knowl-edge on what these channels do for the prokaryotesthemselves. Much can be learned from these organismsgiven the wide distribution and deep evolutionary rootsof K+ channels reviewed here. Readers are also re-minded that there are other types of ion channels amongprokaryotes reviewed elsewhere [3,5,39]. Familiaritywith electrophysiological concepts and literature willbe needed if progress is to be made. We look forwardto the near future in which the power of microbial genet-ics and relative simplicity of bacterial cells allow usdeeper insights on the past evolution, present roles,and future utilities of microbial channels.
Acknowledgments
Work in our laboratory is supported by the VilasTrust of the University of Wisconsin, and NIH GrantsGM47856 to C.K. and GM54867 to Y.S.
Appendix A. Supplementary data
Supplementary data associated with this article canbe found, in the online version, at doi:10.1016/j.femsre.2005.03.003.
References
[1] Doyle, D.A., Morais Cabral, J., Pfuetzner, R.A., Kuo, A.,Gulbis, J.M., Cohen, S.L., Chait, B.T. and MacKinnon, R.(1998) The structure of the potassium channel: molecular basisof K+ conduction and selectivity. Science 280, 69–77.
[2] Zhou, Y., Morais-Cabral, J.H., Kaufman, A. and MacKinnon,R. (2001) Chemistry of ion coordination and hydration revealedby a K+ channel-Fab complex at 2.0 A resolution. Nature 414,43–48.
[3] Kung, C. and Blount, P. (2004) Channels in microbes: So manyholes to fill. Molecular Microbiology 53, 373–380.
[4] Hille, B. (2001) Ion Channels of Excitable Membranes. SinauerAssociates, Inc., Sunderland, MA, p. 814.
[5] Booth, I.R., Edwards, M.D. and Miller, S. (2003) Bacterial ionchannels. Biochemistry 42, 10045–10053.
[6] Ren, D., Navarro, B., Xu, H., Yue, L., Shi, Q. and Clapham,D.E. (2001) A prokaryotic voltage-gated sodium channel.Science 294, 2372–2375.
[7] Koishi, R., Xu, H.X., Ren, D.J., Navarro, B., Spiller, B.W., Shi,Q. and Clapham, D.E. (2004) A superfamily of voltage-gated
980 M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985
sodium channels in bacteria. Journal of Biological Chemistry279, 9532–9538.
[8] Iyer, R., Iverson, T., Accardi, A. and Miller, C. (2002) Abiological role for prokaryotic ClC chloride channels. Nature419, 715–718.
[9] Dutzler, R., Campbell, E.B., Cadene, M., Chait, B.T. andMacKinnon, R. (2002) X-ray structure of a CIC chloridechannel at 3.0 angstrom reveals the molecular basis of anionselectivity. Nature 415, 287–294.
[10] Sukharev, S.I., Blount, P., Martinac, B., Blattner, F.R. andKung, C. (1994) A large-conductance mechanosensitive channelin E. coli encoded by mscL alone. Nature 368, 265–268.
[11] Chang, G., Spencer, R.H., Lee, A.T., Barclay, M.T. and Rees,D.C. (1998) Structure of the MscL homolog from Mycobacte-
rium tuberculosis: A gated mechanosensitive ion channel. Science282, 2220–2226.
[12] Levina, N., Totemeyer, S., Stokes, N.R., Louis, P., Jones, M.A.and Booth, I.R. (1999) Protection of Escherichia coli cells againstextreme turgor by activation of MscS and MscL methanosen-sitive channels: Identification of genes required for MscSactivity. EMBO Journal 18, 1730–1737.
[13] Bass, R.B., Strop, P., Barclay, M. and Rees, D.C. (2002) Crystalstructure of Escherichia coli MscS, a voltage-modulated andmechanosensitive channel. Science 298, 1582–1587.
[14] Kuo, M.M.-C., Saimi, Y. and Kung, C. (2003) Gain-of-functionmutations indicate that Escherichia coli Kch forms a functionalK+ conduit in vivo. EMBO Journal 22, 4049–4058.
[15] Schultz, S.G. and Solomon, A.K. (1961) Cation transport inEscherichia coli I: Intracellular Na and K concentrations and netcation movement. Journal of General Physiology 45, 355–369.
[16] Koch, A.L. (2000) The bacterium�s way for safe enlargement anddivision. Applied and Environmental Microbiology 66, 3657–3663.
[17] Epstein, W. (2003). Progress in Nucleic Acid Research &Molecular Biology, vol. 75, pp. 293–320. Academic Press Inc.,San Diego.
[18] Bakker, E.P. (1993) In: Alkali Cation Transport Systems inProkaryotes (Bakker, E.P., Ed.), pp. 205–224. CRC Press, AnnArbor.
[19] Harold, F.M.a.M.P.C. (1996) In: Escherichia coli and Salmo-nella (Neidhardt, F.C., Ed.), Vol. 1, pp. 283–306. ASM Press,Washington, D.C. USA.
[20] Kashket, E.R. (1982) Stoichiometry of the H+-ATPase ofgrowing and resting, aerobic Escherichia coli. Biochemistry 21,5534–5538.
[22] Durell, S.R., Bakker, E.P. and Guy, H.R. (2000) Does the KdpAsubunit from the high affinity K+-translocating P-type Kdp-ATPase have a structure similar to that of K+ channels?.Biophysics Journal 78, 188–199.
[23] Bertrand, J., Altendorf, K. and Bramkamp, M. (2004)Amino acid substitutions in putative selectivity filterregions III and IV in KdpA alter ion selectivity of theKdpFABC complex from Escherichia coli. Journal ofBacteriology 186, 5519–5522.
[24] van der Laan, M., Gassel, M. and Altendorf, K. (2002)Characterization of amino acid substitutions in KdpA, theK+-binding and -translocating subunit of the KdpFABCcomplex of Escherichia coli. Journal of Bacteriology 184, 5491–5494.
[25] Maser, P., Hosoo, Y., Goshima, S., Horie, T., Eckelman, B.,Yamada, K., Yoshida, K., Bakker, E.P., Shinmyo, A., Oiki, S.,Schroeder, J.I. and Uozumi, N. (2002) Glycine residues inpotassium channel-like selectivity filters determine potassiumselectivity in four-loop-per-subunit HKT transporters from
plants. Proceedings of the National Academy Sciences of theUnited States of America 99, 6428–6433.
[26] Trchounian, A. and Kobayashi, H. (1999) Kup is the major K+
uptake system in Escherichia coli upon hyper-osmotic stress at alow pH. FEBS Letters 447, 144–148.
[27] Kwak, J.M., Murata, Y., Baizabal-Aguirre, V.M., Merrill, J.,Wang, M., Kemper, A., Hawke, S.D., Tallman, G. andSchroeder, J.I. (2001) Dominant negative guard cell K+ channelmutants reduce inward-rectifying K+ currents and light-inducedstomatal opening in Arabidopsis. Plant Physiology 127, 473–485.
[28] Hirsch, R.E., Lewis, B.D., Spalding, E.P. and Sussman, M.R.(1998) A role for the AKT1 potassium channel in plant nutrition.Science 280, 918–921.
[29] McCormack, T. and McCormack, K. (1994) Shaker K+ channelbeta-subunits belong to an NAD(P)H-dependent oxidoreductasesuperfamily. Cell 79, 1133–1135.
[30] Kang, M., Moroni, A., Gazzarrini, S., DiFrancesco, D., Thiel,G., Severino, M. and Van Etten, J.L. (2004) Small potassium ionchannel proteins encoded by chlorella viruses. Proceedings of theNational Academy Sciences of the United States of America 101,5318–5324.
[31] Jiang, Y., Lee, A., Chen, J., Cadene, M., Chait, B.T. andMacKinnon, R. (2002) The open pore conformation of potas-sium channels. Nature 417, 523–526.
[32] Kuo, A., Gulbis, J.M., Antcliff, J.F., Rahman, T., Lowe, E.D.,Zimmer, J., Cuthbertson, J., Ashcroft, F.M., Ezaki, T. andDoyle, D.A. (2003) Crystal structure of the potassium channelKirBac1.1 in the closed state. Science 300, 1922–1926.
[33] Papazian, D.M., Schwarz, T.L., Tempel, B.L., Jan, Y.N. andJan, L.Y. (1987) Cloning of genomic and complementary DNAfrom Shaker, a putative potassium channel gene from Drosoph-
Jan, L.Y. (1987) Sequence of a probable potassium channelcomponent encoded at Shaker locus of Drosophila. Science 237,770–775.
[35] Ketchum, K.A., Joiner, W.J., Sellers, A.J., Kaczmarek, L.K. andGoldstein, S.A. (1995) A new family of outwardly rectifyingpotassium channel proteins with two pore domains in tandem.Nature 376, 690–695.
[36] Zhou, X.L., Vaillant, B., Loukin, S.H., Kung, C. and Saimi, Y.(1995) YKC1 encodes the depolarization-activated K+ channelin the plasma membrane of yeast. FEBS Letters 373, 170–176.
[37] Hille, B. (1973) Potassium channels in myelinated nerve –selective permeability to small cations. Journal of GeneralPhysiology 61, 669–686.
[38] Accardi, A. and Miller, C. (2004) Secondary active transportmediated by a prokaryotic homologue of ClC Cl-channels.Nature 427, 803–807.
[39] Blount, P. (2003) Molecular mechanisms of mechanosensation:Big lessons from small cells. Neuron 37, 731–734.
[40] Patel, A.J., Lazdunski, M. and Honore, E. (2001) Lipid andmechano-gated 2P domain K+ channels. Current Opinion in CellBiology 13, 422–427.
[41] Milkman, R. (1994) An Escherichia coli homologue of eukary-otic potassium channel proteins. Proceedings of the NationalAcademy of Sciences of the United States of America 91, 3510–3514.
[42] Martinac, B., Buechner, M., Delcour, A.H., Adler, J. and Kung,C. (1987) Pressure-sensitive ion channel in Escherichia coli.Proceedings of the National Academy of Sciences of the UnitedStates of America 84, 2297–2301.
[43] Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F.and Higgins, D.G. (1997) The CLUSTAL_X windows inter-face: Flexible strategies for multiple sequence alignment aidedby quality analysis tools. Nucleic Acids Research 25, 4876–4882.
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 981
[44] Kumanovics, A., Levin, G. and Blount, P. (2002) Family ties ofgated pores: Evolution of the sensor module. FASEB Journal 16,1623–1629.
[45] Wheeler, D.L., Chappey, C., Lash, A.E., Leipe, D.D., Madden,T.L., Schuler, G.D., Tatusova, T.A. and Rapp, B.A. (2000)Database resources of the National Center for BiotechnologyInformation. Nucleic Acids Research 28, 10–14.
[46] Ruta, V., Jiang, Y.X., Lee, A., Chen, J.Y. and MacKinnon, R.(2003) Functional analysis of an archaebacterial voltage-depen-dent K+ channel. Nature 422, 180–185.
[47] Deckert, G., Warren, P.V., Gaasterland, T., Young, W.G.,Lenox, A.L., Graham, D.E., Overbeek, R., Snead, M.A., Keller,M., Aujay, M., Huber, R., Feldman, R.A., Short, J.M., Olsen,G.J. and Swanson, R.V. (1998) The complete genome of thehyperthermophilic bacterium Aquifex aeolicus. Nature 392, 353–358.
[48] Nelson, K.E., Clayton, R.A., Gill, S.R., Gwinn, M.L., Dodson,R.J., Haft, D.H., Hickey, E.K., Peterson, L.D., Nelson, W.C.,Ketchum, K.A., McDonald, L., Utterback, T.R., Malek, J.A.,Linher, K.D., Garrett, M.M., Stewart, A.M., Cotton, M.D.,Pratt, M.S., Phillips, C.A., Richardson, D., Heidelberg, J.,Sutton, G.G., Fleischmann, R.D., Eisen, J.A., White, O.,Salzberg, S.L., Smith, H.O., Venter, J.C. and Fraser, C.M.(1999) Evidence for lateral gene transfer between Archaea andBacteria from genome sequence of Thermotoga maritima. Nature399, 323–329.
[49] Kurland, C.G., Canback, B. and Berg, O.G. (2003) Horizontalgene transfer: A critical view. Proceedings of the NationalAcademy of Sciences of the United States of America 100, 9658–9662.
[50] Doolittle, W.F., Boucher, Y., Nesbo, C.L., Douady, C.J.,Andersson, J.O. and Roger, A.J. (2003) How big is the icebergof which organellar genes in nuclear genomes are but the tip?.Philosophical Transactions of the Royal Society of LondonSeries B-Biological Sciences 358, 39–57.
[51] Salzberg, S.L., White, O., Peterson, J. and Eisen, J.A. (2001)Microbial genes in the human genome: Lateral transfer or geneloss?. Science 292, 1903–1906.
[52] Nesbo, C.L., L�Haridon, S., Stetter, K.O. and Doolittle, W.F.(2001) Phylogenetic analyses of two ‘‘Archaeal’’ genes inThermotoga maritima reveal multiple transfers between Archaeaand Bacteria. Molecular Biology and Evolution 18, 362–375.
[54] Splitt, H., Meuser, D., Borovok, I., Betzler, M. and Schrempf,H. (2000) Pore mutations affecting tetrameric assembly andfunctioning of the potassium channel KcsA from Streptomyces
ashi, M., Sekine, M., Baba, S., Ankai, A., Kosugi, H.,Hosoyama, A., Fukui, S., Nagai, Y., Nishijima, K., Otsuka,R., Nakazawa, H., Takamiya, M., Kato, Y., Yoshizawa, T.,Tanaka, T., Kudoh, Y., Yamazaki, J., Kushida, N., Oguchi, A.,Aoki, K., Masuda, S., Yanagii, M., Nishimura, M., Yamagishi,A., Oshima, T. and Kikuchi, H. (2001) Complete genomesequence of an aerobic thermoacidophilic crenarchaeon, Sulfol-obus tokodaii strain7. DNA Research 8, 123–140.
[56] She, Q., Singh, R.K., Confalonieri, F., Zivanovic, Y., Allard, G.,Awayez, M.J., Chan-Weiher, C.C.Y., Clausen, I.G., Curtis,B.A., De Moors, A., Erauso, G., Fletcher, C., Gordon, P.M.K.,Heikamp-de Jong, I., Jeffries, A.C., Kozera, C.J., Medina, N.,Peng, X., Thi-Ngoc, H.P., Redder, P., Schenk, M.E., Theriault,C., Tolstrup, N., Charlebois, R.L., Doolittle, W.F., Duguet, M.,Gaasterland, T., Garrett, R.A., Ragan, M.A., Sensen, C.W. andVan der Oost, J. (2001) The complete genome of the crenarcha-eon Sulfolobus solfataricus P2. Proceedings of the National
Academy of Sciences of the United States of America 98, 7835–7840.
[57] Lesage, F., Guillemare, E., Fink, M., Duprat, F., Lazdunski, M.,Romey, G. and Barhanin, J. (1996) TWIK-1, a ubiquitoushuman weakly inward rectifying K+ channel with a novelstructure. EMBO Journal 15, 1004–1011.
[58] Hodgkin, A.L. and Huxley, A.F. (1952) A quantitative descrip-tion of membrane current and its application to conduction andexcitation in nerve. Journal of Physiology 117, 500–544.
[59] Armstrong, C.D. and Bezanilla, F. (1973) Currents related tomovement of gating particles of sodium channels. Nature 242,459–461.
[60] Armstrong, C.D. and Bezanilla, F. (1974) Charge movementassociated with the opening and closng of the action gates of theNa+ channels. Journal of General Physiology 63, 533–552.
[61] Seoh, S.A., Sigg, D., Papazian, D.M. and Bezanilla, F. (1996)Voltage-sensing residues in the S2 and S4 segments of the ShakerK+ channel. Neuron 16, 1159–1167.
[62] Ruta, V., Jiang, Y., Lee, A., Chen, J. and MacKinnon, R. (2003)Functional analysis of an archaebacterial voltage-dependent K+
channel. Nature 422, 180–185.[63] Sesti, F., Rajan, S., Gonzalez-Colaso, R., Nikolaeva, N. and
Goldstein, S.A. (2003) Hyperpolarization moves S4 sensorsinward to open MVP, a methanococcal voltage-gated potassiumchannel. Nature Neuroscience 6, 353–361.
[64] Jiang, Y., Lee, A., Chen, J., Ruta, V., Cadene, M., Chait, B.T.and MacKinnon, R. (2003) X-ray structure of a voltage-dependent K+ channel. Nature 423, 33–41.
[65] Lu, Z., Klem, A.M. and Ramu, Y. (2001) Ion conduction pore isconserved among potassium channels. Nature 413, 809–813.
[66] Goldstein, S.A. (1996) A structural vignette common to voltagesensors and conduction pores: Canaliculi. Neuron 16, 717–722.
[67] Laine, M., Papazian, D.M. and Roux, B. (2004) Criticalassessment of a proposed model of Shaker. FEBS Letters 564,257–263.
[68] Cohen, B.E., Grabe, M. and Jan, L.Y. (2004) Answers andquestions from the KvAP structures. Neuron 39, 395–400.
[69] Cha, A., Snyder, G.E., Selvin, P.R. and Bezanilla, F. (1999)Atomic scale movement of the voltage-sensing region in apotassium channel measured via spectroscopy. Nature 402, 809–813.
[70] Starace, D.M. and Bezanilla, F. (2004) A proton pore in apotassium channel voltage sensor reveals a focused electric field.Nature 427, 548–553.
[71] Lee, S.Y. and MacKinnon, R. (2004) A membrane-accessmechanism of ion channel inhibition by voltage sensor toxinsfrom spider venom. Nature 430, 232–235.
[72] Cuello, L.G., Cortes, D.M. and Perozo, E. (2004) Moleculararchitecture of the KvAP voltage-dependent K+ channel in alipid bilayer. Science 306, 491–495.
[73] Grabe, M., Lecar, H., Jan, Y.N. and Jan, L.Y. (2004) Aquantitative assessment of models for voltage-dependentgating of ion channels. Proceedings of the National Acad-emy of Sciences of the United States of America 101,17640–17645.
[74] Jiang, Y., Ruta, V., Chen, J., Lee, A. and MacKinnon, R. (2003)The principle of gating charge movement in a voltage-dependentK+ channel. Nature 423, 42–48.
[75] Roosild, T.P., Le, K.-T. and Choe, S. (2004) Cytoplasmicgatekeepers of K+-channel flux: A structural perspective. Trendsin Biochemical Sciences 29, 39–45.
[76] Rossmann, M.G., Moras, D. and Olsen, K.W. (1974) Chemicaland biological evolution of a nucleotide-binding protein. Nature250, 194–199.
[77] Schlosser, A., Hamann, A., Bossemeyer, D., Schneider, E. andBakker, E.P. (1993) NAD+ binding to the Escherichia-Coli K+-uptake protein TrkA and sequence similarity between TrkA and
982 M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985
domains of a family of dehydrogenases suggest a role for NAD+
in bacterial transport. Molecular Microbiology 9, 533–543.[78] Nakamura, T., Yuda, R., Unemoto, T. and Bakker, E.P. (1998)
KtrAB, a new type of bacterial K+-uptake system from Vibrio
alginolyticus. Journal of Bacteriology 180, 3491–3494.[79] Stumpe, S., Schlosser, A., Schleyer, M. and Bakker, E.P. (1996)
In: Transport Processes in Eukaryotic and Prokaryotic Organ-isms (Lolkema, J.S., Ed.), Vol. 2, pp. 473–499. Elsevier ScienceB.V., Amsterdam, The Netherlands.
[80] Bateman, A., Birney, E., Durbin, R., Eddy, S.R., Howe, K.L.and Sonnhammer, E.L.L. (2000) The Pfam protein familiesdatabase. Nucleic Acids Research 28, 263–266.
[81] Roosild, T.P., Miller, S., Booth, I.R. and Choe, S. (2002) Amechanism of regulating transmembrane potassium flux througha ligand-mediated conformational switch. Cell 109, 781–791.
[82] Jiang, Y.X., Pico, A., Cadene, M., Chait, B.T. and MacKinnon,R. (2001) Structure of the RCK domain from the E. coli K+
channel and demonstration of its presence in the human BKchannel. Neuron 29, 593–601.
[83] Jiang, Y.X., Lee, A., Chen, J.Y., Cadene, M., Chait, B.T. andMacKinnon, R. (2002) Crystal structure and mechanism of acalcium-gated potassium channel. Nature 417, 515–522.
[84] Cole, S.T., Eiglmeier, K., Parkhill, J., James, K.D., Thomson,N.R., Wheeler, P.R., Honore, N., Garnier, T., Churcher, C.,Harris, D., Mungall, K., Basham, D., Brown, D., Chillingworth,T., Connor, R., Davies, R.M., Devlin, K., Duthoy, S., Feltwell,T., Fraser, A., Hamlin, N., Holroyd, S., Hornsby, T., Jagels, K.,Lacroix, C., Maclean, J., Moule, S., Murphy, L., Oliver, K.,Quail, M.A., Rajandream, M.A., Rutherford, K.M., Rutter, S.,Seeger, K., Simon, S., Simmonds, M., Skelton, J., Squares, R.,Squares, S., Stevens, K., Taylor, K., Whitehead, S., Woodward,J.R. and Barrell, B.G. (2001) Massive gene decay in the leprosy
R. and Gouaux, E. (2003) Structural basis for modulation andagonist specificity of HCN pacemaker channels. Nature 425,200–205.
[86] Nimigean, C.M., Shane, T. and Miller, C. (2004) A cyclicnucleotide modulated prokaryotic K+ channel. Journal ofGeneral Physiology 124, 203–210.
[87] Clayton, G.M., Silverman, W.R., Heginbotham, L. and Morais-Cabral, J.H. (2004) Structural basis of ligand activation in acyclic nucleotide regulated potassium channel. Cell 119, 615–627.
[88] Chen, G.Q., Cui, C., Mayer, M.L. and Gouaux, E. (1999)Functional characterization of a potassium-selective prokaryoticglutamate receptor. Nature 402, 817–821.
[89] Armstrong, N., Sun, Y., Chen, G.Q. and Gouaux, E. (1998)Structure of a glutamate-receptor ligand-binding core in complexwith kainate. Nature 395, 913–917.
[90] Sun, Y.J., Rose, J., Wang, B.C. and Hsiao, C.D. (1998) Thestructure of glutamine-binding protein complexed with gluta-mine at 1.94 angstrom resolution: Comparisons with otheramino acid binding proteins. Journal of Molecular Biology 278,219–229.
[91] Mayer, M.L., Olson, R. and Gouaux, E. (2001) Mechanisms forligand binding to GluR0 ion channels: Crystal structures of theglutamate and serine complexes and a closed apo state. Journalof Molecular Biology 311, 815–836.
[92] Dufresne, A., Salanoubat, M., Partensky, F., Artiguenave, F.,Axmann, I.M., Barbe, V., Duprat, S., Galperin, M.Y., Koonin,E.V., Le Gall, F., Makarova, K.S., Ostrowski, M., Oztas, S.,Robert, C., Rogozin, I.B., Scanlan, D.J., de Marsac, N.T.,Weissenbach, J., Wincker, P., Wolf, Y.I. and Hess, W.R. (2003)Genome sequence of the cyanobacterium Prochlorococcus mari-
nus SS120, a nearly minimal oxyphototrophic genome. Proceed-ings of the National Academy of Sciences of the United States ofAmerica 100, 10020–10025.
[93] Rocap, G., Larimer, F.W., Lamerdin, J., Malfatti, S., Chain, P.,Ahlgren, N.A., Arellano, A., Coleman, M., Hauser, L., Hess,W.R., Johnson, Z.I., Land, M., Lindell, D., Post, A.F., Regala,W., Shah, M., Shaw, S.L., Steglich, C., Sullivan, M.B., Ting,C.S., Tolonen, A., Webb, E.A., Zinser, E.R. and Chisholm, S.W.(2003) Genome divergence in two Prochlorococcus ecotypesreflects oceanic niche differentiation. Nature 424, 1042–1047.
[94] van Ham, R., Kamerbeek, J., Palacios, C., Rausell, C., Abascal,F., Bastolla, U., Fernandez, J.M., Jimenez, L., Postigo, M.,Silva, F.J., Tamames, J., Viguera, E., Latorre, A., Valencia, A.,Moran, F. and Moya, A. (2003) Reductive genome evolution inBuchnera aphidicola. Proceedings of the National Academy ofSciences of the United States of America 100, 581–586.
[95] Akman, L., Yamashita, A., Watanabe, H., Oshima, K., Shiba,T., Hattori, M. and Aksoy, S. (2002) Genome sequence of theendocellular obligate symbiont of tsetse flies, Wigglesworthia
A.S., Wernegreen, J.J., Sandstrom, J.P., Moran, N.A. andAndersson, S.G.E. (2002) 50 million years of genomic stasis inendosymbiotic bacteria. Science 296, 2376–2379.
[97] Shigenobu, S., Watanabe, H., Hattori, M., Sakaki, Y. andIshikawa, H. (2000) Genome sequence of the endocellularbacterial symbiont of aphids Buchnera sp. APS. Nature 407,81–86.
psittaci GPIC): Examining the role of niche-specific genes in theevolution of the Chlamydiaceae. Nucleic Acids Research 31,2134–2147.
[99] Shirai, M., Hirakawa, H., Kimoto, M., Tabuchi, M., Kishi, F.,Ouchi, K., Shiba, T., Ishii, K., Hattori, M., Kuhara, S. andNakazawa, T. (2000) Comparison of whole genome sequences ofChlamydia pneumoniae J138 from Japan and CWL029 fromUSA. Nucleic Acids Research 28, 2311–2314.
[100] Read, T.D., Brunham, R.C., Shen, C., Gill, S.R., Heidelberg,J.F., White, O., Hickey, E.K., Peterson, J., Utterback, T., Berry,K., Bass, S., Linher, K., Weidman, J., Khouri, H., Craven, B.,Bowman, C., Dodson, R., Gwinn, M., Nelson, W., DeBoy, R.,Kolonay, J., McClarty, G., Salzberg, S.L., Eisen, J. and Fraser,C.M. (2000) Genome sequences of Chlamydia trachomatis MoPnand Chlamydia pneumoniae AR39. Nucleic Acids Research 28,1397–1406.
[101] Kalman, S., Mitchell, W., Marathe, R., Lammel, C., Fan, L.,Hyman, R.W., Olinger, L., Grimwood, L., Davis, R.W. andStephens, R.S. (1999) Comparative genomes of Chlamydia
pneumoniae and C. trachomatis. Nature Genetics 21, 385–389.[102] McLeod, M.P., Qin, X., Karpathy, S.E., Gioia, J., Highlander,
S.K., Fox, G.E., McNeill, T.Z., Jiang, H.Y., Muzny, D., Jacob,L.S., Hawes, A.C., Sodergren, E., Gill, R., Hume, J., Morgan,M., Fan, G.W., Amin, A.G., Gibbs, R.A., Hong, C., Yu, X.J.,Walker, D.H. and Weinstock, G.M. (2004) Complete genomesequence of Rickettsia typhi and comparison with sequences ofother rickettsiae. Journal of Bacteriology 186, 5842–5855.
[103] Ogata, H., Audic, S., Renesto-Audiffren, P., Fournier, P.E.,Barbe, V., Samson, D., Roux, V., Cossart, P., Weissenbach, J.,Claverie, J.M. and Raoult, D. (2001) Mechanisms of evolution inRickettsia conorii and R. prowazekii. Science 293, 2093–2098.
[104] Andersson, S.G.E., Zomorodipour, A., Andersson, J.O., Siche-ritz-Ponten, T., Alsmark, U.C.M., Podowski, R.M., Naslund,A.K., Eriksson, A.S., Winkler, H.H. and Kurland, C.G. (1998)The genome sequence of Rickettsia prowazekii and the origin ofmitochondria. Nature 396, 133–140.
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 983
[105] Raoult, D., Ogata, H., Audic, S., Robert, C., Suhre, K.,Drancourt, M. and Claverie, J.M. (2003) Tropheryma whipplei
twist: A human pathogenic Actinobacteria with a reducedgenome. Genome Research 13, 1800–1809.
[106] Bentley, S.D., Maiwald, M., Murphy, L.D., Pallen, M.J., Yeats,C.A., Dover, L.G., Norbertczak, H.T., Besra, G.S., Quail, M.A.,Harris, D.E., von Herbay, A., Goble, A., Rutter, S., Squares, R.,Squares, S., Barrell, B.G., Parkhill, J. and Relman, D.A. (2003)Sequencing and analysis of the genome of the Whipple�s diseasebacterium Tropheryma whipplei. Lancet 361, 637–644.
[107] Seshadri, R., Paulsen, I.T., Eisen, J.A., Read, T.D., Nelson,K.E., Nelson, W.C., Ward, N.L., Tettelin, H., Davidsen, T.M.,Beanan, M.J., Deboy, R.T., Daugherty, S.C., Brinkac, L.M.,Madupu, R., Dodson, R.J., Khouri, H.M., Lee, K.H., Carty,H.A., Scanlan, D., Heinzen, R.A., Thompson, H.A., Samuel,J.E., Fraser, C.M. and Heidelberg, J.F. (2003) Complete genomesequence of the Q-fever pathogen Coxiella burnetii. Proceedingsof the National Academy of Sciences of the United States ofAmerica 100, 5455–5460.
[108] Waters, E., Hohn, M.J., Ahel, I., Graham, D.E., Adams, M.D.,Barnstead, M., Beeson, K.Y., Bibbs, L., Bolanos, R., Keller, M.,Kretz, K., Lin, X., Mathur, E., Ni, J., Podar, M., Richardson,T., Sutton, G.G., Simon, M., Soll, D., Stetter, K.O., Short, J.M.and Noordewier, M. (2003) The genome of Nanoarchaeum
equitans: Insights into early archaeal evolution and derivedparasitism. Proceedings of the National Academy of Sciences ofthe United States of America 100, 12984–12988.
[109] Bentley, S.D. and Parkhill, J. (2004) Comparative genomicstructure of prokaryotes. Annual Review of Genetics 38, 771–791.
[110] Fraser, C.M., Norris, S.J., Weinstock, C.M., White, O., Sutton,G.G., Dodson, R., Gwinn, M., Hickey, E.K., Clayton, R.,Ketchum, K.A., Sodergren, E., Hardham, J.M., McLeod, M.P.,Salzberg, S., Peterson, J., Khalak, H., Richardson, D., Howell,J.K., Chidambaram, M., Utterback, T., McDonald, L., Artiach,P., Bowman, C., Cotton, M.D., Fujii, C., Garland, S., Hatch, B.,Horst, K., Roberts, K., Sandusky, M., Weidman, J., Smith, H.O.and Venter, J.C. (1998) Complete genome sequence of Trepo-
nema pallidum, the syphilis spirochete. Science 281, 375–388.[111] Parkhill, J., Wren, B.W., Mungall, K., Ketley, J.M., Churcher,
C., Basham, D., Chillingworth, T., Davies, R.M., Feltwell, T.,Holroyd, S., Jagels, K., Karlyshev, A.V., Moule, S., Pallen, M.J.,Penn, C.W., Quail, M.A., Rajandream, M.A., Rutherford,K.M., van Vliet, A.H.M., Whitehead, S. and Barrell, B.G.(2000) The genome sequence of the food-borne pathogenCampylobacter jejuni reveals hypervariable sequences. Nature403, 665–668.
[112] Langen, H., Takacs, B., Evers, S., Berndt, P., Lahm, H.W.,Wipf, B., Gray, C. and Fountoulakis, M. (2000) Two-dimen-sional map of the proteome of Haemophilus influenzae. Electro-phoresis 21, 411–429.
[113] Holden, M.T.G., Feil, E.J., Lindsay, J.A., Peacock, S.J., Day,N.P.J., Enright, M.C., Foster, T.J., Moore, C.E., Hurst, L.,Atkin, R., Barron, A., Bason, N., Bentley, S.D., Chillingworth,C., Chillingworth, T., Churcher, C., Clark, L., Corton, C.,Cronin, A., Doggett, J., Dowd, L., Feltwell, T., Hance, Z.,Harris, B., Hauser, H., Holroyd, S., Jagels, K., James, K.D.,Lennard, N., Line, A., Mayes, R., Moule, S., Mungall, K.,Ormond, D., Quail, M.A., Rabbinowitsch, E., Rutherford, K.,Sanders, M., Sharp, S., Simmonds, M., Stevens, K., Whitehead,S., Barrell, B.G., Spratt, B.G. and Parkhill, J. (2004) Completegenomes of two clinical Staphylococcus aureus strains: Evidencefor the rapid evolution of virulence and drug resistance.Proceedings of the National Academy of Sciences of the UnitedStates of America 101, 9786–9791.
Chen, Y.M., Yuan, Z.H., Zhao, G.P., Qu, D., Danchin, A. andWen, Y.M. (2003) Genome-based analysis of virulence genes in anon-biofilm-forming Staphylococcus epidermidis strain(ATCC12228). Molecular Microbiology 49, 1577–1593.
[115] Baba, T., Takeuchi, F., Kuroda, M., Yuzawa, H., Aoki, K.,Oguchi, A., Nagai, Y., Iwama, N., Asano, K., Naimi, T.,Kuroda, H., Cui, L., Yamamoto, K. and Hiramatsu, K. (2002)Genome and virulence determinants of high virulence commu-nity-acquired MRSA. Lancet 359, 1819–1827.
[116] Kuroda, M., Ohta, T., Uchiyama, I., Baba, T., Yuzawa, H.,Kobayashi, I., Cui, L.Z., Oguchi, A., Aoki, K., Nagai, Y., Lian,J.Q., Ito, T., Kanamori, M., Matsumaru, H., Maruyama, A.,Murakami, H., Hosoyama, A., Mizutani-Ui, Y., Takahashi,N.K., Sawano, T., Inoue, R., Kaito, C., Sekimizu, K., Hirakawa,H., Kuhara, S., Goto, S., Yabuzaki, J., Kanehisa, M., Yamash-ita, A., Oshima, K., Furuya, K., Yoshino, C., Shiba, T., Hattori,M., Ogasawara, N., Hayashi, H. and Hiramatsu, K. (2001)Whole genome sequencing of meticillin-resistant Staphylococcusaureus. Lancet 357, 1225–1240.
[120] Rendulic, S., Jagtap, P., Rosinus, A., Eppinger, M., Baar, C.,Lanz, C., Keller, H., Lambert, C., Evans, K.J., Goesmann, A.,Meyer, F., Sockett, R.E. and Schuster, S.C. (2004) A predatorunmasked: Life cycle of Bdellovibrio bacteriovorus from agenomic perspective. Science 303, 689–692.
[121] Seshadri, G., Myers, G.S.A., Tettelin, H., Eisen, J.A., Heidel-berg, J.F., Dodson, R.J., Davidsen, T.M., DeBoy, R.T., Fouts,D.E., Haft, D.H., Selengut, J., Ren, Q.H., Brinkac, L.M.,Madupu, R., Kolonay, J., Durkin, S.A., Daugherty, S.C.,Shetty, J., Shvartsbeyn, A., Gebregeorgis, E., Geer, K., Tsegaye,G., Malek, J., Ayodeji, B., Shatsman, S., McLeod, M.P., Smajs,D., Howell, J.K., Pal, S., Amin, A., Vashisth, P., McNeill, T.Z.,Xiang, Q., Sodergren, E., Baca, E., Weinstock, G.M., Norris,S.J., Fraser, C.M. and Paulsen, I.T. (2004) Comparison of thegenome Treponema denticola with of the oral pathogen otherspirochete genomes. Proceedings of the National Academy ofSciences of the United States of America 101, 5646–5651.
T., Sasamoto, S., Watanabe, A., Idesawa, K., Iriguchi, M.,Kawashima, K., Kohara, M., Matsumoto, M., Shimpo, S.,Tsuruoka, H., Wada, T., Yamada, M. and Tabata, S. (2002)Complete genomic sequence of nitrogen-fixing symbiotic bacte-rium Bradyrhizobium japonicum USDA110. DNA Research 9,189–197.
[133] Galagan, J.E., Nusbaum, C., Roy, A., Endrizzi, M.G., Mac-donald, P., FitzHugh, W., Calvo, S., Engels, R., Smirnov, S.,Atnoor, D., Brown, A., Allen, N., Naylor, J., Stange-Thomann,N., DeArellano, K., Johnson, R., Linton, L., McEwan, P.,McKernan, K., Talamas, J., Tirrell, A., Ye, W., Zimmer, A.,Barber, R.D., Cann, I., Graham, D.E., Grahame, D.A., Guss,A.M., Hedderich, R., Ingram-Smith, C., Kuettner, H.C.,Krzycki, J.A., Leigh, J.A., Li, W., Liu, J., Mukhopadhyay, B.,Reeve, J.N., Smith, K., Springer, T.A., Umayam, L.A., White,O., White, R.H., de Macario, E.C., Ferry, J.G., Jarrell, K.F.,Jing, H., Macario, A.J.L., Paulsen, I., Pritchett, M., Sowers,K.R., Swanson, R.V., Zinder, S.H., Lander, E., Metcalf, W.W.and Birren, B. (2002) The genome of M. acetivorans revealsextensive metabolic and physiological diversity. GenomeResearch 12, 532–542.
M.M.-C. Kuo et al. / FEMS Microbiology Reviews 29 (2005) 961–985 985
[134] Stover, C.K., Pham, X.Q., Erwin, A.L., Mizoguchi, S.D.,Warrener, P., Hickey, M.J., Brinkman, F.S.L., Hufnagle,W.O., Kowalik, D.J., Lagrou, M., Garber, R.L., Goltry, L.,Tolentino, E., Westbrock-Wadman, S., Yuan, Y., Brody, L.L.,Coulter, S.N., Folger, K.R., Kas, A., Larbig, K., Lim, R., Smith,K., Spencer, D., Wong, G.K.S., Wu, Z., Paulsen, I.T., Reizer, J.,Saier, M.H., Hancock, R.E.W., Lory, S. and Olson, M.V. (2000)Complete genome sequence of Pseudomonas aeruginosa PAO1,an opportunistic pathogen. Nature 406, 959–964.
[135] Paulsen, I.T., Banerjei, L., Myers, G.S.A., Nelson, K.E.,Seshadri, R., Read, T.D., Fouts, D.E., Eisen, J.A., Gill, S.R.,Heidelberg, J.F., Tettelin, H., Dodson, R.J., Umayam, L.,Brinkac, L., Beanan, M., Daugherty, S., DeBoy, R.T., Durkin,S., Kolonay, J., Madupu, R., Nelson, W., Vamathevan, J., Tran,B., Upton, J., Hansen, T., Shetty, J., Khouri, H., Utterback, T.,Radune, D., Ketchum, K.A., Dougherty, B.A. and Fraser, C.M.(2003) Role of mobile DNA in the evolution of vancomycin-resistant Enterococcus faecalis. Science 299, 2071–2074.
[138] Shimizu, T., Ohtani, K., Hirakawa, H., Ohshima, K., Yamash-ita, A., Shiba, T., Ogasawara, N., Hattori, M., Kuhara, S. andHayashi, H. (2002) Complete genome sequence of Clostridium
perfringens, an anaerobic flesh-eater. Proceedings of the NationalAcademy of Sciences of the United States of America 99, 996–1001.
[139] Bruggemann, H., Baumer, S., Fricke, W.F., Wiezer, A., Liese-gang, H., Decker, I., Herzberg, C., Martinez-Arias, R., Merkl,R., Henne, A. and Gottschalk, G. (2003) The genome sequenceof Clostridium tetani, the causative agent of tetanus disease.Proceedings of the National Academy of Sciences of the UnitedStates of America 100, 1316–1321.
[140] Bao, Q.Y., Tian, Y.Q., Li, W., Xu, Z.Y., Xuan, Z.Y., Hu, S.N.,Dong, W., Yang, J., Chen, Y.J., Xue, Y.F., Xu, Y., Lai, X.Q.,Huang, L., Dong, X.Z., Ma, Y.H., Ling, L.J., Tan, H.R., Chen,R.S., Wang, J., Yu, J. and Yang, H.M. (2002) A complete
sequence of the T. tengcongensis genome. Genome Research 12,689–700.
[141] Henne, A., Bruggemann, H., Raasch, C., Wiezer, A., Hartsch,T., Liesegang, H., Johann, A., Lienard, T., Gohl, O., Martinez-Arias, R., Jacobi, C., Starkuviene, V., Schlenczeck, S., Dencker,S., Huber, R., Klenk, H.P., Kramer, W., Merkl, R., Gottschalk,G. and Fritz, H.J. (2004) The genome sequence of the extremethermophile Thermus thermophilus. Nature Biotechnology 22,547–553.
[142] Chain, P., Lamerdin, J., Larimer, F., Regala, W., Lao, V.,Land, M., Hauser, L., Hooper, A., Klotz, M., Norton, J.,Sayavedra-Soto, L., Arciero, D., Hommes, N., Whittaker, M.and Arp, D. (2003) Complete genome sequence of theammonia-oxidizing bacterium and obligate chemolithoauto-troph Nitrosomonas europaea. Journal of Bacteriology 185,2759–2773.
[143] Kapatral, V., Anderson, I., Ivanova, N., Reznik, G., Los, T.,Lykidis, A., Bhattacharyya, A., Bartman, A., Gardner, W.,Grechkin, G., Zhu, L.H., Vasieva, O., Chu, L., Kogan, Y.,Chaga, O., Goltsman, E., Bernal, A., Larsen, N., D�Souza, M.,Walunas, T., Pusch, G., Haselkorn, R., Fonstein, M., Kyrpides,N. and Overbeek, R. (2002) Genome sequence and analysis ofthe oral bacterium Fusobacterium nucleatum strain ATCC 25586.Journal of Bacteriology 184, 2005–2018.
[144] Epstein, W. and Davies, M. (1970) Potassium-dependent mutat-nts of Escherichia coli K-12. Journal of Bacteriology 101, 836–843.
[145] Epstein, W. and Kim, B.S. (1971) Potassium transport loci inEscherichia coli K-12. Journal of Bacteriology 108, 639–644.
[146] Rhoads, D., Waters, F. and Epstein, W. (1976) Cation transportin Escherichia coli VIII: Potassium transport mutants. Jorunal ofGeneral Physiology 67, 325–341.
[147] Schrempf, H., Schmidt, O., Kummerlen, R., Hinnah, S., Muller,D., Betzler, M., Steinkamp, T. and Wagner, R. (1995) Aprokaryotic potassium ion channel with two predicted trans-membrane segments from Streptomyces lividans. EMBO Journal14, 5170–5178.
[148] Wolters, M., Madeja, M., Farrell, A.M. and Pongs, O. (1999)Bacillus stearothermophilus lctB gene gives rise to functional K+
channels in Escherichia coli and in Xenopus oocytes. Receptorsand Channels 6, 477–491.
[149] Hellmer, J. and Zeilinger, C. (2003) MjK1, a K+ channel fromM. jannaschii, mediates K+ uptake and K+ sensitivity in E. coli.FEBS Letters 547, 165–169.
[150] Enkvetchakul, D., Bhattacharyya, J., Jeliazkova, I., Groesbeck,D.K., Cukras, C.A. and Nichols, C.G. (2004) Functionalcharacterization of a prokaryotic Kir channel. Journal ofBiological Chemistry 279, 47076–47080.
