Review The Origin and Early Evolution of

5429_01_p1-17

1/21/05

11:46 AM

Page 1

ASTROBIOLOGY Volume 5, Number 1, 2005 © Mary Ann Liebert, Inc.

Review The Origin and Early Evolution of Membrane Channels ANDREW POHORILLE,1,2 KARL SCHWEIGHOFER,1,3 and MICHAEL A. WILSON1,2

ABSTRACT The origin and early evolution of ion channels are considered from the point of view that the transmembrane segments of membrane proteins are structurally quite simple and do not require specific sequences to fold. We argue that the transport of solute species, especially ions, required an early evolution of efficient transport mechanisms, and that the emergence of simple ion channels was protobiologically plausible. We also argue that, despite their simple structure, such channels could possess properties that, at the first sight, appear to require markedly greater complexity. These properties can be subtly modulated by local modifications to the sequence rather than global changes in molecular architecture. In order to address the evolution and development of ion channels, we focus on identifying those protein domains that are commonly associated with ion channel proteins and are conserved throughout the three main domains of life (Eukarya, Bacteria, and Archaea). We discuss the potassiumsodium-calcium superfamily of voltage-gated ion channels, mechanosensitive channels, porins, and ABC-transporters and argue that these families of membrane channels have sufficiently universal architectures that they can readily adapt to the diverse functional demands arising during evolution. Key Words: Ion channels–Ion transport–Folding of membrane proteins–Protocells. Astrobiology 5, 1–17.

INTRODUCTION

Even in the simplest prokaryotic cell, membrane proteins are quite common (Stevens and Arkin, 2000). In fact, it is natural to assume that emergence of membrane proteins was essential for evolution from membrane-enclosed structures (vesicles) to the simplest forms of cellular life. Vesicles had the tremendous evolutionary advantage that they could retain their contents as the membrane created a diffusion barrier between the vesicle interior and its environment. However, in the absence of membrane proteins, the ancestors of cells (protocells) would have pos-

M

EMBRANE PROTEINS mediate functions that are essential to all cells. These functions include transport of ions, nutrients, and waste products across cell membranes, capture of energy and its transduction into a form usable in biochemical reactions, transmission of environmental signals to the interior of cells, cellular growth, and volume regulation. In higher organisms, approximately 30% of genes code for membrane proteins (Wallin and von Heijne, 1998).

3Research

Institute for Advanced Computer Science, 1NASA Ames Research Center, Moffett Field, California. of Pharmaceutical Chemistry, University of California, San Francisco, California.

2Department

1

5429_01_p1-17

1/21/05

11:46 AM

Page 2

2

sessed extremely limited capabilities to communicate with their environment and, in particular, to access necessary substrates and to expel waste products. This review is devoted to the origin and the earliest evolution of membrane proteins with a focus on channels. These are macromolecular pores that mediate the transport of ions and small molecules across the cell membrane along their concentration (more correctly, chemical potential) gradient. Channels are lodged in the lipid bilayer that makes up the cell membrane and are positioned in a transmembrane (TM) orientation such that one end is in contact with the environment and the other end is located in the cellular interior. They should be distinguished from other membrane transporters, such as carriers, which bind a solute and move it across the membrane. We also focus only on passive transporters, as opposed to protein pumps, which transport species against the chemical potential gradient, expending energy in the process. Most channels in contemporary cells are highly selective and preferentially conduct only one type of ion, most often Na, K, H, Ca2, Mg2, or Cl, or a specific neutral species, e.g., water. Their selectivity is encoded in the amino acid sequence. Channel activity is regulated by physical or chemical signals. For example, the voltage-gated and mechanosensitive (MS) superfamilies of channels respond, respectively, to electrostatic membrane potentials and to membrane tension caused by mechanical stimuli. The ligand-gated superfamily of channels is activated in response to specific interactions with small molecules, such as -aminobutyrate or glycine. Until recently, most research on channels was focused on the electophysiology of eukaryotic cells. Because of persistent difficulties in cloning and crystallizing these proteins, little was known about their structure or mechanism of action at the molecular level. Fortunately, these difficulties are being overcome, and high-resolution structures of several channels have been recently determined (Walz et al., 1997; Doyle et al., 1998; Bass et al., 2002; Jiang et al., 2002a,b, 2003a; Miller et al., 2003). This, in turn, sheds light on their mechanism of action. This progress was recognized through awarding the 2003 Nobel Prize in Chemistry to Roderick MacKinnon and Peter Agre, who pioneered structural studies on membrane channels. Furthermore, analysis of the rapidly growing databases of sequences revealed that

POHORILLE ET AL.

many eukaryotic channels have homologs in both bacteria and archaea (Lu et al., 2001). In some cases, cellular functions of these proteins are known. For example, MS channels in bacteria are most likely needed to establish osmotic equilibrium between the cell interior and the environment (Martinac, 2001). In their absence, a bacterial cell would take up water and swell, and its fragile cell wall will eventually burst. In other instances, the role of channels is still poorly understood. At any rate, recently acquired knowledge of membrane channels in simple organisms is helping us trace their roots much more reliably than was possible just a few years ago. The departure point for this review is the existence of short protein sequences (peptides) in the protobiological milieu. Starting from this point we discuss how peptides could have partitioned into membranes, self-organized into functional structures, and evolved towards increasing efficiency and specificity. We consider the preceding step—the origin of peptides on the primitive Earth—as falling outside the scope of this review. A wide variety of mechanisms that might have led to the formation of peptides has been recently critically reviewed (Rode, 1999). The author argues that peptides were most likely the first precursors of biopolymers. He also points out interesting connections between preferential sequences generated in model peptide synthesis under prebiotic conditions and found in membrane proteins of simple organisms.

GENERAL CHARACTERISTICS OF MEMBRANE CHANNELS Structurally, integral membrane proteins differ markedly from water-soluble proteins. They consist of one or several TM regions connected by extramembrane segments. TM regions are 15–20 amino acids in length, just enough to span the lipid bilayer. Most of these amino acids are hydrophobic, i.e., poorly soluble in water, but polar or even charged residues also occur and often play a structural or functional role. The number of TM segments is constant in a channel family and offers important clues about the evolutionary history of membrane proteins. Extramembrane segments are considerably more hydrophilic and may markedly vary in length even within a single family. Some of these segments are quite large and participate in regula-

5429_01_p1-17

1/21/05

11:46 AM

Page 3

3

ORIGIN OF MEMBRANE CHANNELS

tion of channel activity, for example, by providing a ligand-binding site. Channels might exist as monomers or, more commonly, as homomeric or heteromeric aggregates. Perhaps the most remarkable feature of membrane proteins is the structural simplicity of their TM segments. Even though many membrane proteins or their functional complexes are among the largest in cells, these segments are arranged as either bundles of -helices or barrels of -strands. Deviations from these structural architectures are small and rather rare (although usually significant functionally) (Popot and Engelman, 2000). In this respect, membrane proteins are quite different from their water-soluble counterparts, which contain a much wider range of structural motifs (Liu et al., 2002). Moreover, some large membrane proteins retain their primary function even if portions are removed (Duff and Ashley, 1992; Montal, 1995). These findings are consistent with the existence of very simple, natural or synthetic proteins capable of forming stable helices that spontaneously aggregate into channels (Lear et al., 1988; Duff and Ashley, 1992; Cafiso, 1994; Bechinger, 1997; Duclohier and Wroblewski, 2001). This is in contrast with the behavior of water-soluble proteins, which typically lose their function and, at least partially, unfold upon even modest truncation of their sequence. The commonality in structural and functional design of membrane channels and the existence of simple models indicate that explaining the origin of membrane proteins might be easier than explaining the origin of water-soluble proteins. Additional concepts emerging from experimental and theoretical studies are very helpful in this task. They are captured in the two-stage model of forming membrane proteins (Popot and Engelman, 1990). In the first stage, individual helices are established across

FIG. 1. Schematic of the three-stage model of interfacial folding. Amphipathic proteins partition to the water– membrane interface, where they fold into -helices. The -helices can then become inserted into the membrane where they associate to form larger protein aggregates such as dimeric or tetrameric bundles.

the hydrophobic region of the membrane. In the second stage, they interact with one another to create functional TM structures, such as channels. Experiments on refolding and the assembly of internal membrane proteins from fragments give strong support to the model (Popot and Engelman, 1990). Recently, the model has been extended in two directions. An additional stage that follows Stage Two was added, which includes such adjustments to the structure as folding and insertion of loops and peripheral domains and ligand binding (Engelman et al., 2003). Perhaps more relevant to this review, it has been argued that the first stage can be considered as consisting of two steps: folding of helices at the water–membrane interface followed by helix insertion into the lipid bilayer (Pohorille et al., 2003). This is shown schematically in Fig. 1. In this paper, we follow a similar framework and consider the self-organization of membrane proteins into functional structures as a series of consecutive, essentially independent steps. We will discuss each step—folding, insertion into the membrane, and aggregation—separately, including its thermodynamic and evolutionary implications. We will argue that each step was quite feasible under protobiological conditions. This will be followed by a discussion of the mechanism of action of simple channels and their early evolution. Before we proceed, however, we will address the question of why assistance is needed in transporting ions across cell membranes.

UNASSISTED ION TRANSPORT ACROSS MEMBRANES Lipids forming membranes of cells or organelles are organized into bilayers such that their

5429_01_p1-17

1/21/05

11:46 AM

Page 4

4

hydrophilic head groups are exposed to water on both sides of the membrane, whereas the bilayer interior consists entirely of hydrophobic tails. In archaea, the same organization is achieved by molecules built of two hydrophilic head groups connected by a long stretch of hydrophobic groups. It is quite plausible that the formation of structures bound by membranes (vesicles) started protocellular evolution. In fact, molecules capable of forming vesicles constitute a large fraction of organic material extracted from the Murchison meteorite (Deamer and Pashley, 1989). They were also obtained in laboratory simulations of interstellar or cometary material (Dworkin et al., 2001; Deamer et al., 2002). Membranes are highly effective barriers to ion transport. Qualitatively, the inability of ions to penetrate the nonpolar interior of the membrane is easy to understand. In an aqueous environment, ions strongly polarize the surrounding water so that dipoles of water molecules tend to align themselves with the electric field of the ion. As the ion moves into the membrane, strong, favorable water–ion interactions are substituted, at a considerable free energy expense, by interactions between ions and oily groups inside the bilayer. In the language of classical, continuum electrostatics, this process can be considered as the transfer of charge from a high to a low dielectric medium. In the simplest treatment of this problem (Parsegian, 1969), both the water and the membrane are represented by continuous dielectric media characterized, respectively, by high and low dielectric constants. The ion is modeled as a point charge in the center of a cavity defined by its effective ionic radius. This treatment yields the activation barrier to transport

POHORILLE ET AL.

(also known as the Born barrier) of Na across a membrane of a typical width equal to 47 kcal/mol. The top of this barrier is located at the center of the membrane. This is shown in Fig. 2. This corresponds to a membrane permeability 14–17 orders of magnitude lower than that measured experimentally (Hauser et al., 1973; Nozaki and Tanford, 1981; Paula et al., 1998). To account for this discrepancy, it was proposed that ions move through transient defects in the membrane rather than through unperturbed hydrophobic interior (Nagle and Scott, 1978; Deamer and Bramhall, 1986). Atomic-level computer simulations of unassisted ion transport across membranes (Wilson and Pohorille, 1996) confirmed this assertion. In these simulations, it was demonstrated that ion permeation is always accompanied by a local, asymmetric deformation of the lipid bilayer that can be well described as a “thinning defect” penetrated by water. The presence of these defects allows the ion to remain partially hydrated during transfer across membranes, which in turn reduces the activation barrier. Indeed, the calculated barrier for Na is 26 kcal/mol, which yields membrane permeability in good agreement with experiment (Wilson and Pohorille, 1996). Even if membrane flexibility is correctly taken into account, we expect only a few ions a day to permeate the lipid bilayer (Deamer and Bramhall, 1986). This is insufficient to support cellular processes requiring ions, such as establishing osmotic equilibrium, chemical catalysis, stabilization of macromolecular structures, and maintaining bioenergetics based on charge separation across cell membranes. For example, potassium ions permeate membranes of modern cells at rates

FIG. 2. Cartoon of unassisted ion transport across membranes. a: The ion loses most or all of its waters of hydration in the membrane interior. b: The ion moves into a thinning defect, which allows the ion to retain much of its hydration shell. While the free energy in (b) is appreciably lower than in (a), both processes require surmounting a large free energy barrier.

5429_01_p1-17

1/21/05

11:46 AM

Page 5

5


approaching 108 ions/s per channel under physiological electrochemical gradients (Hille, 2001). It is possible that walls of vesicles in the protobiological era were more permeable to ions than bilayers made of lipids typical to contemporary membranes. Such membranes, however, would be more fragile and leaky to the organic material inside a vesicle. In addition, mechanisms for regulating ion flow across membranes would be absent.

FORMATION OF SIMPLE, TM CHANNELS Inherent limitations of unassisted ion transport in supporting cellular functions lead to a conclusion that the emergence of molecular structures capable of mediating transfer of ions across membranes in a controlled manner was an imperative in early protocellular evolution. In this section, we discuss the three steps required for the formation of such structures made of TM proteins— folding, insertion into membranes, and self-assembly into channels.

Folding of membrane peptides Most peptides exist in water as disordered structures (random coils), which are not suitable for performing biological functions. Ordered structures, such as -helices, are unstable or, at best, only marginally stable. This lack of stability has a simple, qualitative explanation. Peptide folding to an -helix is accompanied by the formation of intramolecular hydrogen bonds between CÖ and N-H groups separated by three residues along the peptide backbone. If a peptide is considered in isolation these hydrogen bonds strongly favor the helix and more than compensate for the loss of entropy in the folded state. In an aqueous environment, however, the CÖ . . . H-N hydrogen bonds add little to the stability of the helix because both hydrogen-bonding donor and acceptor groups can also form hydrogen bonds with the surrounding water molecules. Peptides that are built of hydrophobic amino acids, such as leucine, valine, or alanine, or a mixture of hydrophobic and hydrophilic residues are attracted to water–membrane interfaces (Kaiser and Kezdy, 1987; Brown and Huestis, 1993; Blondelle et al., 1995). In fact, interfacial activity is characteristic not only of peptides but also, more generally, of many solutes that have both

hydrophobic and hydrophilic groups (Pratt and Pohorille, 2002). The tendency to accumulate at the interface, exhibited by all these solutes, results mostly from two oppositely changing contributions to the free energy (Pohorille and Wilson, 1996). The free energy that must be expended to create a cavity in the solvent sufficiently large to accommodate the solute is larger in water than in the membrane interior. This unfavorable contribution is counterbalanced by the favorable contribution to the free energy associated with electrostatic solute–solvent interactions, which is also larger in water. Commonly, the most favorable balance between these contributions is achieved for the interfacial location of a solute. Upon contact with the membrane, many peptides that are disordered in water fold into -helices or, less often, -sheets or -turns (DeGrado and Lear, 1985; Keire and Fletcher, 1996; Song et al., 1996; Bayley, 1999; Popot and Engelman, 2000). Intramolecular hydrogen bonds are more stable at the water–membrane interface than in aqueous solution because the CÖ and N-H groups in the peptide backbone are at least partially shielded from water (Chipot and Pohorille, 1998). Of particular interest are peptides that have an amino acid sequence such that hydrophobic and hydrophilic residues are located on opposite faces when the peptides fold into -helices. These helices can readily adopt an orientation in which the hydrophilic face is buried in water while the hydrophobic face is exposed to the nonpolar environment formed by the hydrocarbon tails of the lipids. Such peptides are called amphipathic and are quite common among small membrane proteins in contemporary cells (Franklin et al., 1994; Smith et al., 1994; Hristova et al., 2001). The match between the polarities of a peptide and its environment renders the amphipathic helices particularly stable. The specific identity of amino acids appears to be less important. This is a desirable protobiological property because neither a precise protein synthesis mechanism nor the full suite of amino acids was required for the formation of amphipathic helices. Considering that these helices had to fulfill only very modest sequence constraints, their presence in the protocellular environment should not have been rare.

Insertion of -helices into membranes Folded peptides can change their orientations with respect to the interface from parallel to per-

5429_01_p1-17

1/21/05

11:46 AM

Page 6

6

pendicular and incorporate into the membrane such that they span the lipid bilayer. The insertion of largely hydrophobic peptides is most likely because their nonpolar residues interact favorably with the hydrophobic interior of the membrane. This view is supported by statistical analysis of TM segments in contemporary proteins. The most common hydrophilic residue buried in membranes is serine, which constitutes 7% of all amino acids located inside lipid bilayers. All other hydrophilic amino acids total only 1–3% of protein content in membranes and usually play specific structural or functional roles (Engelman et al., 2003). In fact, most successful methods for predicting TM segments from sequences of integral membrane proteins are based on identifying stretches of 15–20 mostly hydrophobic amino acids. These considerations, however, do not imply that single -helices are stable in a TM orientation. In fact, both theoretical and experimental studies suggest that this is not the case. Computer simulations using continuum models (Ben-Tal et al., 1996a,b) and atomic-level representation of both the peptide and its environment (Chipot and Pohorille, 1998; Pohorille et al., 2003) indicate that the TM arrangement of the peptide is only metastable, i.e., it corresponds to a local free energy minimum. The global minimum corresponds to the orientation parallel to the membrane interface. A similar conclusion was reached from nuclear magnetic resonance studies of polyalanine peptides in phospholipid bilayers, in which a few alanine residues were replaced with leucine (Bechinger, 2001). The results pointed to a dynamic equilibrium between orientations, in which the helical peptides were aligned either parallel or perpendicular to the membrane normal. These results can be understood by referring, again, to the hydrophobic effect. Even if all residues are hydrophobic, a peptide still contains hydrophilic groups in its backbone. At the interface these groups are partially exposed to water, but have to undergo dehydration during insertion into the membrane. This process requires expending free energy (Jayasinghe et al., 2001). Conclusions about the instability of single helices in TM orientation are further supported by studies on simple, natural (Lear et al., 1988) or synthetic (Sansom, 1993; Cafiso, 1994) peptides that aggregate into channels upon application of a TM voltage. It has been argued that the voltage-dependent step of channel formation is a

POHORILLE ET AL.

voltage-induced insertion of the helical peptide (Cafiso, 1994). In the absence of voltage the channel-forming helices reside at the membrane surface (Biggin and Sansom, 1996).

Association of TM helices Neither interfacial folding of membrane peptides nor their insertion into lipid bilayers requires specific amino acid sequences. While this lack of sequence selectivity is a desirable property in early evolution, it also poses some difficulties. If all helices capable of insertion into membranes could have self-assembled in all possible combinations, the number of different protein aggregates in the lipid bilayer would have been enormous and, most likely, would have disrupted protocellular membranes. Some degree of sequence specificity was required for helices to form channels, and it emerged at the level of a sequence-modulated interhelical recognition. Helices could self-assemble only if their interactions were sufficiently strong to compensate for the loss of free energy associated with the insertion of individual helices into membranes. Otherwise they were expelled from the membrane. In fact, the ability to associate was probably one of the selection mechanisms operating on protocellular TM peptides. TM helices interact by forming right- or lefthanded associations rather than aggregates of monomers aligned along the bilayer normal. In the left-handed arrangement, helices often wrap around each other to form supercoiled structures. This allows for close contacts between helices across the membrane, which stabilize the helical bundle. Right-handed arrangements of helices allow for the formation of funnel-like structures, which are common to many channels. The recently crystallized potassium channel (Doyle et al., 1998), the MS channel (Chang et al., 1998), and aquaporins (Walz et al., 1997) have this type of molecular architecture. A common feature of helical bundles is that protein surfaces exposed to the lipids are hydrophobic. Hydrophilic residues, if present, are either aligned along the lumen of the channel or involved in interhelical interactions. Several types of specific interactions promote helix association. Recent biophysical and genetic studies on simple models for helix associations in membranes have greatly improved our understanding of these interactions. In general, pack-

5429_01_p1-17

1/21/05

11:46 AM

Page 7


ing of amino acid side chains in interacting helices is well described by the “knobs-into-holes” model (Langosch and Heringa, 1998). Such tight contacts increase the favorable van der Waals interactions between helices. An excellent model to probe the role of these interactions is glycophorin A, which forms a noncovalent dimer of -helices containing a 24-residue TM region (Lemmon et al., 1992; MacKenzie et al., 1997). On the basis of random mutagenesis and nuclear magnetic resonance spectroscopy in detergent micelles, a model of the TM dimer has been proposed, in which helix–helix association results from the specific interaction of seven residues located on one face of each -helix (Mackenzie et al., 1997). Mutations of some of these residues, for example, leucine 75 or isoleucine 76, to alanine, which is also hydrophobic but smaller, destabilize the dimer by 1–2 kcal/mol (Fleming et al., 1997; Fisher et al., 1999), presumably because close van der Waals contacts between the residues in the wild type have been disrupted. Helix–helix recognition may also be promoted by specific electrostatic interactions between polar residues. These residues are not common in the lipid environment because sequestering them from water is associated with a considerable free energy cost. They are, however, necessary to provide the functional diversity of channel as the physical and chemical properties of purely aliphatic side chains of hydrophobic amino acids are very limited. Helpful models for investigating the role of electrostatic interactions are peptides built of 23 hydrophobic leucine residues that contain a single, membrane-buried site for various “guest” amino acids (Zhou et al., 2000, 2001). If the guest site is occupied by hydrophilic glutamate, glutamic acid, aspartic acid, or asparagine, each of which is capable of being simultaneously hydrogen bond donors and acceptors, helix association is promoted in both detergent micelles and biological membranes. In contrast, peptides containing serine and threonine, which occur more frequently than other hydrophilic residues, do not associate better than the purely hydrophobic peptide (Zhou et al., 2001). Very similar results were obtained from studies on another hydrophobic peptide (Choma et al., 2000; Gratkowski et al., 2001). It has been estimated that the presence of a single polar amino acid stabilizes the association of TM peptides by 1–2 kcal/mol (Choma et al., 2000). Serine (S) and threonine (T) can also

7

promote strong helix interactions if they appear in SxxSSxxT and SxxxSSxxT sequence motifs (here x stands for a hydrophobic amino acid) (Adamian and Liang, 2002; Dawson et al., 2002). Most likely this is because they form a cooperative network of interhelical hydrogen bonds. In contrast, it appears that the specific lipid environment plays a lesser role in stabilizing helix associations. Many membrane proteins or complexes of helical peptides remain stable in other nonpolar solvents, such as detergents, or can be reconstituted in non-native lipid environment (Popot and Engelman, 2000). Again, this is a desirable property from the early evolution perspective because the composition of protocellular membranes might have been considerably more heterogeneous than the composition of membranes in contemporary cells. This property, however, does not necessarily extend to complex channels of contemporary cells, which are often sensitive to membrane composition. This may point to coordinated evolution of membranes and integral membrane proteins.

MECHANISM OF CHANNEL-MEDIATED TRANSPORT The central theme of the previous section was that the emergence of TM aggregates that self-assembled into channels from a large “library” of -helical peptides was protobiologically plausible. In this section, the attention shifts to the question of how simple channels could have mediated efficient, selective, and regulated transport of ions and small, neutral molecules across membranes. Some remarkably simple peptides can assemble into structures that exhibit many of the desired characteristics of ion channels. Perhaps the simplest is a synthetic peptide built of only two amino acids, hydrophobic leucine (L) and hydrophilic (S), forming the sequence (LSSLLSL)3. At the water–membrane interface, this peptide folds into an amphipathic -helix, but in the presence of an applied voltage it becomes inserted into the membrane. Then it associates into channel-forming bundles of five to seven helices such that leucines are exposed to the lipids and serine side chains project into the lumen of the channel (Lear et al., 1988). Both cations and anions are transported through the channel with weak selectivity towards positively charged species. The channel is also rectifying, i.e., exhibits asymmetric

5429_01_p1-17

1/21/05

11:46 AM

Page 8

8

dependence of ion current on membrane potential. However, when a negative charge is added at the mouth, the channel becomes completely cation-selective, and rectification is largely eliminated. Oppositely, adding a positive charge at the mouth eliminates cation selectivity and enhances rectification (Lear et al., 1997). All these properties can be readily explained by a primitive electrostatic model indicating that not only the structure but also the mechanism of action of this channel is quite simple. The LS channel is too simple to exhibit high efficiency and selectivity. These two properties appear to be at odds with each other. High selectivity implies that the transported ion has to bind to the channel significantly stronger than its competitors, whereas high efficiency requires that ions move through a channel without encountering strong binding sites or energy barriers. In order to understand how these two properties can be achieved simultaneously we must turn to channels that are considerably more complicated. Although these channels are not truly protobiological, understanding their mechanism of action and its relation to the structure sheds light on the initial steps in the evolution of functions performed by membrane proteins. For our purposes probably the best choice is the family of potassium channels. These channels exist in organisms from all three domains of life, which speaks to their antiquity. Recently, structures of several potassium channels from prokaryotes have been resolved (Doyle et al., 1998; Jiang et al., 2002a,b, 2003a; Kuo et al., 2003), revealing that the ion conduction pore and the mechanism of selectivity are conserved within

POHORILLE ET AL.

the family (Lu et al., 2001). A schematic of an ion channel is shown in Fig. 3. Among potassium channels, the KcsA channel is perhaps the best understood in terms of structure, mechanism of action, and energetics (Doyle et al., 1998; Roux and MacKinnon, 1999; Roux et al., 2000; Jiang et al., 2002a,b; Zhou and MacKinnon, 2003). This channel is built of four identical subunits, each consisting of three -helices. Near the center of the bilayer there is a cavity, which is filled by 20–40 water molecules and is lined by nonpolar residues. Four -helices surrounding the pore are tilted at approximately 45° to the channel axis, and they point to the center of the cavity. On the extracellular (incoming) side of the channel there is a narrow fragment, approximately 12 Å long, which acts as a selectivity filter for K ions. It is formed by the backbone carbonyl oxygen atoms that belong to a highly conserved sequence of five amino acids. The size of the pore matches very well the diameter of potassium ions, ensuring nearly optimal coordination of positively charged K by 20 backbone oxygen atoms (five from each subunit) that carry negative partial charges. The match is not nearly as good for sodium ions, which results in high selectivity for K over Na. The key to high conduction rates of the potassium channel appears to be multiple occupancy of the selectivity filter by potassium ions. Along the filter there are four binding sites for K. At ionic concentrations exceeding 20 mM two of these sites (either 1,3 or 2,4) are occupied by potassium ions separated by a water molecule. Once a third ion enters the filter electrostatic repulsion acting between ions of the same charge

FIG. 3. Cartoon of ion transport through an ion channel. The ion moves through a water-filled pore. The free energy barrier can be quite low and, in some cases, ion transport is diffusion-limited.

5429_01_p1-17

1/21/05

11:46 AM

Page 9


displaces a potassium ion from the opposite side (Zhou and MacKinnon, 2003). If the ion is displaced towards the center of the bilayer it enters the water-filled cavity. As discussed in the section on unassisted ion transport, it should encounter a free energy barrier at this point. However, the channel structure is tuned so precisely that favorable electrostatic interactions with water in the cavity and, probably more importantly, with the channel helices eliminate this barrier, so that ionic motion is diffusion limited (Roux et al., 2000). Other effects are also at play in this and other channels, but this example illustrates our main point—as ion channels evolved they adapted to improve efficiency and selectivity at the same time, but all of it happened without disrupting the helical nature of the TM protein segments. In fact, the same applies to regulation of ion conductance, as demonstrated in the recent studies on a voltage-gated potassium channel, KvAP (Jiang et al., 2003a,b). It appears that the potassium channel has a simple structural analog built of a 16-residue antibiotic peptaibol, antiamoebin (AAM). Its sequence is F-U-U-U-J-G-L-U-U-O-Q-J-O-U-P-F, where U is -aminoisobutyric acid, J is isovaline, and O is hydroxyproline. AAM is capable of inserting into lipid bilayers and associating to form cationic channels across membranes. In contrast to many other simple peptide channels, it exhibits single-level bursts of conductance. This indicates that the channel has a single conducting state that is built of a fixed number of helices. Although its high-resolution structure is not yet known, a detailed, high-quality model has been recently proposed (O’Reilly and Wallace, 2003). It is based on x-ray and nuclear magnetic resonance structures of a single AAM helix (Karle et al., 1998; Snook et al., 1998; Galbraith et al., 2003) and a variety of electrophysiological and mutation data for the channel. The model has the remarkable feature that the structure of AAM exhibits considerable similarities to the structure of the KcsA potassium channel (Doyle et al., 1998). The AAM channel consists of eight TM helices, which form a parallel bundle with a hydrophilic lumen and a hydrophobic exterior in contact with lipids. The KcsA potassium channel, which is tetrameric, is also built of eight helices, two contributed by each monomer. Both channels are narrow at the intracellular side and widen significantly at the extracellular end. The cation-binding region of KcsA is lined by carbonyl groups from the backbone.

9

In the same region of the AAM channel there are carbonyl groups from imino acids exposed to the lumen. Finally, at the center of KcsA there is a water-filled cavity, which lowers the barrier to ion transport. In the AAM channel, the same role is presumably played by the ring of glutamine side chains. An important difference between the two channels is that ion gating in KcsA is complex and involves bending of inner helices, whereas gating in the AAM channel is simply accomplished by helix dissociation. The comparison between the KcsA and AAM channels shows that sequences very different in length and composition can share the same overall structure. Currently, however, it is not known to what extent their mechanism of action is similar. In particular, it is not clear whether AAM yields the same reduction of Born barrier as does the KcsA channel and whether ion transport through the AAM channel also proceeds through a complex mechanism involving a multiple ion occupancy inside the lumen. Of particular protobiological interest, especially in the context of possible nongenomic origins of membrane proteins, is the presence of atypical amino (or imino) acids in AAM. Different lines of evidence indicate that on the early Earth “unnatural” amino and imino acids were at least as common as the amino acids genetically coded in modern organism. Thus it is quite likely that they would have been incorporated into early peptides. The example of the AAM channel indicates that they could have been structurally and functionally useful. This in turn means that they were eliminated from the inventory of biologically common molecules for different reasons. It is less clear whether the same applies to racemic mixtures of amino acids, which existed on the early Earth. Although one contemporary membrane channel, gramicidin A (Wallace, 1990), contains both L- and D-amino acids, its structure is quite unique, and therefore this protein does not appear to be closely related to other channels. It is, however, possible that amino acids of different chirality could have been occasionally incorporated into a membrane protein without destroying channel structure. Alternatively, a mechanism that ensures selectivity between diastereoisomers in peptide synthesis (Saghatelian et al., 2001) might have been at play. Among the different ions that have to be transported across membranes, protons are of special interest. All living systems make use of proton

5429_01_p1-17

1/21/05

11:46 AM

Page 10

10

gradients across cell membranes to convert environmental energy into a high-energy chemical compound, adenosine triphosphate (ATP). ATP, in turn, is used as a source of energy to drive many cellular reactions. The ubiquity of this process in biology suggests that even the earliest cellular systems relied on proton gradients for harvesting the environmental energy needed to support their survival and growth. In contemporary cells, protons are transferred against their concentration gradient in a process assisted by large, complex proteins embedded in membranes. There are, however, simple channels that transport protons, and they may offer valuable clues to the origin of contemporary bioenergetics. Protons can be transferred across membranes through a mechanism that is not available to other ions. It involves translocation of excess protons along hydrogen-bonded chains of water molecules filling the lumen of the channel. This mechanism is more efficient because it requires only small displacements of several protons between consecutive water molecules along the chain rather than a large translocation of a single ion. To achieve sustained proton transport an additional step involving the reorientation of water molecules in the chain is required. The “proton wire” mechanism may involve not only water molecules, but also protein side chains containing protons capable of dissociation. One of the simplest proteins that form proton channels is the M2 protein from the influenza virus (Pinto et al., 1992). This protein is 97 amino acids in length, but a fragment 25 amino acids long, which contains a TM domain of 19 amino acids, is sufficient to transport protons (Duff and Ashley, 1992). Four identical protein fragments, each folded into an -helix, aggregate to form small channels spanning the membrane. Protons are conducted through a narrow pore in the middle of the channel in response to applied voltage. It has been postulated that protons are translocated along the network of water molecules filling the pore of the channel. Selectivity is ensured by the presence of four histidine amino acids, one from each of the helices, which are sufficiently large to occlude the pore. By doing so they block transport of small ions, such as sodium or potassium. Histidines also interrupt the network of water molecules filling the channel, but proton transport can still proceed because each histidine is able to become positively charged by accepting an additional proton. The mechanism by which

POHORILLE ET AL.

the proton is released on the outgoing side of the channel is not yet fully understood and may involve rotation or tautomerization of the histidine that undergoes protonation (Pinto et al., 1997; Zhong et al., 1998; Schweighofer and Pohorille, 2000). The M2 protein is not only a remarkably simple system that efficiently and selectively aids proton transport across membranes, but also may provide clues to the origin of proton pumps, which utilize energy to create a proton gradient across cell membranes (Pohorille et al., 1996; Lanyi and Pohorille, 2001). From the discussion above one might conclude that channels that freely permeate water should also efficiently mediate proton transport. This is, however, not the case, and aquaporin 1 (AQP1) is an interesting counterexample. AQP1 is commonly found in human red blood cells, kidneys, and other tissue, but there is at least one member of the same family, the glycerol uptake facilitator, that is present in bacteria. The main physiological role of AQP1 is to transport water through cell membranes. However, it remains impermeable to protons. This is very fortunate because proton leakage is unproductive and is detrimental to the efficiency of bioenergetic processes. The inability to transport protons by AQP1 implies that the proton wire mechanism through a network of hydrogen-bonded water molecules cannot be at play in this case. On the basis of electron crystallographic data, it has been proposed that hydrogen bonding between water molecules is indeed broken (Murata et al., 2000). This takes place in the constricted space near the center of the bilayer, where water molecules form hydrogen bonds with the aspartic acid side chain protruding into the lumen rather than with each other. If this proposal were correct, it would be another example of a channel in which selectivity is achieved at a single point along the pore. It is interesting to note that other members of the aquaporin family exhibit different substrate specificity. Some of them transport glycerol, but they are not very efficient in facilitating permeation of water. Other aquaporins mediate transport of anions or show broad substrate specificity. Although a molecular-level explanation for these differences is not available, one observation is obvious: Channel specificity may change through small, local modifications in the sequence and structure, a relatively simple evolutionary task compared with global changes in sequence and molecular architecture.

5429_01_p1-17

1/21/05

11:46 AM

Page 11


The apparent simplicity by which peptides can act as TM channels may shed light on the nature of the earliest biopolymers. Attempts to create channels built of nucleic acids, carried out mostly through in vitro evolution, have so far failed, yielding only molecules that attach to and disrupt membranes causing unspecific ionic leakage (Vlassov et al., 2001). In fact, these efforts may never succeed because of poor hydrophobic matching between nucleic acids and lipid tails. This, in turn, suggests that proteins must have been present in protocells and played a functional role as soon as there was a need for communication between the protocellular interior and the environment.

EVOLUTION OF MEMBRANE TRANSPORTERS The main thesis of the previous sections has been that the emergence of simple ion channels was protobiologically plausible. We have also argued that, despite their simple structure, such channels could possess a wide range of interesting properties that, at first sight, appear to require markedly greater complexity. Moreover, these properties can be subtly modulated by local modifications to the sequence rather than global changes in molecular architecture. Because the systems that we considered were either very simple or well studied experimentally, it was possible to provide a molecular-level explanation for their structure and mechanism of action. The central theme of this section is closely related. We will argue that ion channels have a sufficiently universal architecture that they can readily adapt to the diverse functional demands arising during evolution. In other words, there might be a considerable variety of functions within a single family of membrane proteins, characterized not only by similar structure but also, quite often, by sequence homology. In fact, the last point should be understood more broadly than has been implied so far, as we are discovering striking connections between ion channels and active membrane transporters (pumps) or even some water-soluble proteins, which might indicate their common evolutionary origins (Jan and Jan, 1992). For example, porins are present in bacteria, eukaryotes, and viruses. They can form general conductance pores that pass ions up to a certain size, or can serve as spe-

11

cific channels for certain types of molecules (Bateman et al., 2002). Some porins require energy and are active transporters, whereas others are passive in their conductance. Another family of channels that contain multifunctional domains is the cystic fibrosis TM regulator (CFTR). This is a chloride channel that hydrolyzes nucleotide triphosphates, and thus possesses enzymatic activity (Anderson et al., 1992). There are also the multidrug resistance proteins, part of the ABC family (ABC is short for ATP-binding cassette), which export a range of toxins through coupling with proton influx (Sasatsu et al., 1989; Parent and Roy, 1992). Thus, the definitions of an ion channel or an ion channel family is being revised as we find more examples of molecules that fit into multiple roles. Currently, insufficient information exists to develop a detailed scenario that would explain how ion-conducting helical bundles evolved into channels similar to those found in modern cells and how functions of these proteins diversified. However, we can provide some insight into this issue using information that exists in the sequence databases, and on the phylogeny of protein domains. In this section, we will discuss several ancient classes of proteins, each of which could represent one of the earliest forms of transporters, and give some insights as to their potential roles across evolutionary history. A brute force approach to identify a likely ion channel ancestor is to enumerate all known transport proteins, perform exhaustive sequence and structural comparisons, and group sequences by similarity. One can then perform a phylogenetic study to determine which proteins are shared among the most diverse set of species, and which structural motifs, functions, and regulatory mechanisms appear to be conserved throughout evolution. This would be a monumental undertaking. Instead, we decided to identify structural domains most commonly associated with transport, which are shared among the major domains of life (archaea, bacteria, eukaryotes, and viruses). We used the Pfam database (Bateman et al., 2002) of protein domains to construct a set of candidates, and then chose several groups to discuss in this section: the potassium-sodium-calcium superfamily of voltage-gated ion channels, MS channels, porins, and ABC-transporters. The superfamily containing the sodium, potassium, calcium, and cyclic nucleotide-gated ion channels consist of six TM helical domains (S1–

5429_01_p1-17

1/21/05

11:46 AM

Page 12

12

S6), which can be repeated multiple times. The sodium and calcium channel members all contain four homologous six-TM domains, whereas the potassium and cyclic nucleotide-gated channels consist of tetramers of single-domain subunits (Goldin, 2002; Hille, 2001). Although the channels exhibit fundamental differences in size, the primary structural characteristic is a pore region surrounded by the fifth and sixth TM domains. There is also some evidence that the extracellular loop region flanked by S5 and S6 may also be a component of the pore region (Strong et al., 1993). The 1 and the subunits of the calcium and sodium channels, respectively, are approximately four times larger than the equivalent subunit of potassium channels and contain four covalently linked repeats of the six TM domain. It has therefore been widely accepted that these channels form a superfamily, which likely arose by gene duplication (Hille, 2001; Piontkivska and Hughes, 2003; Saier, 2003; Strong et al., 1993). It has also been observed that many bacteria express potassium channels, but do not generally express either sodium or chloride channels, because they do not have a requirement for these ions. Recently, however, a bacterial voltage-gated sodium channel with significant homology to calcium channels (especially in the pore region) has been reported (Ren et al., 2001). It was also shown that this channel is encoded by one six-TM region, similar to that of a potassium channel. This supports the hypothesis that the primordial channel, which gave rise to this superfamily, resembled the potassium channel (Goldin, 2002). It is interesting to note that several attempts to construct functional sodium or calcium channels composed of only six TM domains by artificially dividing the large subunit were unsuccessful (Stuhmer et al., 1989; Ahern et al., 2001). The superfamily of voltage-gated channels discussed above demonstrates how new functions can arise by duplication of a successful motif and subsequent divergence and specialization. The prototypical potassium channel, with moderate changes in key areas, can give rise to a channel with different specificity, but with the same general architecture. Duplication of motifs and subsequent hetero-oligomerization can give rise to a high degree of diversity of channel functions. This diversity is required for adaptation to different nutritional sources, environmental conditions, and the more complex signal transduction mechanisms required for multicellularity.

POHORILLE ET AL.

MS channels are transducers of physical or mechanical stress and, as such, are thought to be one of the oldest signal transduction molecules known (Kloda and Martinac, 2001a). They are believed to mediate many primitive cell processes, such as osmotic regulation and cell growth, as well as the senses of higher organisms (hearing, balance, and touch). There are two known families of these channels: the large-conductance MS channel (MscL) and the small-conductance MS channel (MscS). MS channels are present in the membranes of organisms from the three domains of life: bacteria, archaea, and eukaryotes (Pivetti et al., 2003). This makes them attractive targets for phylogenetic analysis. MS channels can vary in their conductance properties and ion selectivity. Many are nonspecific in their transport properties, but there are specialized versions for potassium, cations, or anions (Sachs, 1992; Kloda and Martinac, 2001a). Thus, it is possible that these channels could have evolved repeatedly over time, or arisen out of successive gene duplication or gene transfer events. MS channels are activated by changes in membrane stress, possibly due to changes in the thickness of the membrane. As such, they are able to transduce signals originating either from osmotic stress (which swells or shrinks the cell volume) or from contact with a foreign object (the classic avoidance mechanism of ciliated protozoa utilizes MS channels and a calcium action potential). Experiments with MscL channels reconstituted into liposomes indicate that membrane thinning decreases the free energy of activation, whereas thickening increases the free energy (Kloda and Martinac, 2001b). It was suggested that the thinner membrane stabilizes the open conformation of the channel because of a better hydrophobic match between the channel open state conformation and the membrane (Kloda and Martinac, 2001b, 2002). Molecular dynamics simulations of this system appear to corroborate this view (Gullingsrud et al., 2001). This type of mechanical regulation could have been one of the first to evolve, since it does not require complicated systems of accessory proteins and cofactors. The channels found in eukaryotes are thus far of the MscS type. They are generally more diverse in their sequence and have a lower free energy of activation than MscL (around 4.5 kcal/mol vs. 9–11.5 kcal/mol for MscL). However, a channel recently isolated from the archaeon Methanococcus jannaschii requires a free energy of activation

5429_01_p1-17

1/21/05

11:46 AM

Page 13


of only 3–3.5 kcal/mol (Kloda and Martinac, 2001b). Presumably, these channels evolved from a common ancestor of bacterial and archaeal MS channels. Porins are another interesting class of ion channels, commonly found in bacterial cell walls, mitochondria, and chloroplasts. Porins are defined as water-filled channels through which molecules of up to 600 Da may diffuse (Nikaido and Rosenberg, 1981; Bolter and Soll, 2001). The general structure of these channels is not an -helical bundle but a -barrel, with the membrane-spanning regions being connected by hydrophilic loops. These loops can function as regulatory or ionselective regions. In Gram-negative bacteria, there are three classes of porins: general diffusion types, solute specific, and ligand gated (Klebba and Newton, 1998; Bolter and Soll, 2001). An example of a solute-specific porin is the Gram-negative protein LamB, also known as maltoporin. Its function is to allow the diffusion of maltodextrin across the outer cell membrane, and its structure is known. This protein associates into a trimeric structure composed of an antiparallel -barrel (Schirmer et al., 1995). The binding affinity of solute-specific porins is low, presumably to allow selectivity without requiring energy for transport. Ligandgated porins bind a substrate and utilize energy to facilitate transport of the target molecule. An example of this is the ferrienterobactin receptor (FepA) protein, which interacts with the tonB proteins to bind the substrate (Locher et al., 1998; Buchanan et al., 1999). This mechanism allows substrates that are either poorly permeable through the porin channels or present at very low concentrations to be transported. Eukaryotes have a different class of porins, which are unrelated by sequence to the bacterial porins but which share a common structural motif. These consist of the mitochondrial porins and chloroplast-associated porins. The former are associated with the outer mitochondrial membrane, and are referred to as voltage-gated anion channels (VDACs). VDACs are generally nonspecific and form a homodimer, each with 16 antiparallel -sheets (Thomas et al., 1991; Benz, 1994; Mannella, 1997; Bolter and Soll, 2001). The channels are highly regulated and are thought to be involved in many cellular functions, implying that these proteins became part of the cellular machinery early on. Chloroplast porins are associated with the outer membrane of the organelle.

13

These porins are thought to share a -barrel architecture with bacterial porins and VDACs, but the number of -strands is different. Porins have not yet been found in any archaeal species, possibly because of a lack of detailed knowledge of these organisms or because of a fundamental difference in the membranes of archaea versus that of bacteria and eukaryotes. Nonetheless, porins represent a clearly ancient function that would have been needed early in evolution. They are good examples of proteins that evolved to perform different functions and fulfill different selectivity and regulation requirements without undergoing major architectural rearrangements. We now turn our attention to one of the largest families of proteins, the ABC transporters. These proteins are involved in a diverse set of functions and are able to translocate a variety of compounds. ABC transporters may be involved in either efflux or uptake, with the efflux systems having been found in all major divisions of life— archaea, eukaryotes, bacteria, and viruses. The vast majority of these proteins are active transporters and use the energy of ATP hydrolysis for transport. A notable exception is the CFTR, which is an ion channel as opposed to a transporter; however, it is still dependent on ATP hydrolysis for activation (Akabas, 2000). ABC transporters are composed of a minimum of four domains, including two integral membrane domains (permeases) and two nucleotidebinding domains (which bind and hydrolyze ATP). Members of this family that are involved in uptake also have an extracytoplasmic solutebinding domain. The integral membrane components are typically composed of six TM segments. The most conserved region is the ATP-binding region, followed by the integral membrane region, and finally the substrate-binding region, which is the most divergent (Tam and Saier, 1993). A crystal structure for the ABC-ATPase of the glucose ABC transporter from the archaeon Sulfolobus is known (Verdon et al., 2003). Members of the efflux category may be associated with drug resistance or antiparasitic activity, and would have been necessary to protect organisms from chemical attack, whereas those specialized for uptake would have been necessary for nutrient absorption. In either case, the presence of a generalized functional molecule that could handle the import and export of a wide variety of molecules would allow an organism to

5429_01_p1-17

1/21/05

11:46 AM

Page 14

14

POHORILLE ET AL.

adapt to a variety of different food sources and potential threats. Also, because the molecule can act as a general transporter or ion channel, it can readily evolve to perform many different specialized functions necessary in higher organisms. Although all of the transport proteins discussed in this section are thought to be ancient in their origins, they are still quite complex compared with what must have existed in the early protocellular systems. In these earliest cells, the first channels were probably simple helical bundles that poorly conducted ions and were relatively nonspecific in their gating properties (Saier, 2003). The primary function of these channels was probably to regulate osmotic pressure and supply ions needed for biochemical processes inside protocells. Combination of different peptides into flexible oligomeric units would have provided the enhanced ion selectivity and conduction properties needed to form the basis of early signal transduction mechanisms. Combination of channel and energy capture functions would have generated the first active transport systems and bestowed the ability to generate TM potential differences needed to store energy and propagate signals over long distances. The superposition of energy-coupling and ion conduction properties likely occurred very early in cellular life, giving protocells endowed with this capability great evolutionary advantage. Further evolution was greatly accelerated by gene duplication and gene fusion events, which in combination with local mutations were remarkably successful in increasing functional diversity of membrane transport proteins.

ACKNOWLEDGMENTS This work was supported by the NASA Exobiology Program and the NASA Astrobiology Institute.

ABBREVIATIONS AAM, antiamoebin; ABC, ATP-binding cassette; AQP1, aquaporin 1; ATP, adenosine triphosphate; CFTR, cystic fibrosis transmembrane regulator; MS, mechanosensitive; MscL, largeconductance mechanosensitive channel; MscS, small-conductance MS channel; TM, transmembrane; VDAC, voltage-gated anion channel.

REFERENCES Adamian, L. and Liang, J. (2002) Interhelical hydrogen bonds and spatial motifs in membrane proteins: polar clamps and serine zippers. Proteins 47, 209–218. Ahern, C.A., Arikkath, J., Vallejo, P., Gurnett, C.A., Powers, P.A., Campbell, K.P., and Coronado, R. (2001) Intramembrane charge movements and excitation-contraction coupling expressed by two-domain fragments of the Ca2 channel. Proc. Natl. Acad. Sci. USA 98, 6935– 6940. Akabas, M.H. (2000) Cystic fibrosis transmembrane conductance regulator. Structure and function of an epithelial chloride channel. J. Biol. Chem. 275, 3729–3732. Anderson, J.A,, Huprikar, S.S., Kochian, L.V., Lucas, W.J., and Gaber, R.F. (1992) Functional expression of a probable Arabidopsis thaliana potassium channel in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 89, 3736– 3740. Bass, R.B., Strop, P., Barclay, M., and Rees, D.C. (2002) Crystal structure of Escherichia coli MscS, a voltagemodulated and mechanosensitive channel. Science 298, 1582–1587. Bateman, A., Birney, E., Cerruti, L., Durbin, R., Etwiller, L., Eddy, S.R., Griffiths-Jones, S., Howe, K.L., Marshall, M., and Sonnhammer, E.L. (2002) The Pfam protein families database. Nucleic Acids Res. 30, 276–280. Bayley, H. (1999) Designed membrane channels and pores. Curr. Opin. Biotechnol. 10, 94–103. Bechinger, B. (1997) Structure and functions of channelforming peptides: Magainins, cecropins, melittin and alamethicin. J. Membr. Biol. 156, 197–211. Bechinger, B. (2001) Membrane insertion and orientation of polyalanine peptides: a 15N solid-state NMR spectroscopy investigation. Biophys. J. 81, 2251–2256. Ben-Tal, N., Ben-Shaul, A., Nicholls, A., and Honig, B. (1996a) Free-energy determinants of -helix insertion into lipid bilayers. Biophys. J. 70, 1803–1812. Ben-Tal, N., Honig, B., Pleitzsch, R.M., Denisov, G., and McLaughlin, S. (1996b) Binding of small basic peptides to membranes containing acidic lipids: Theoretical models and experimental results. Biophys. J. 71, 561–575. Benz, R. (1994) Permeation of hydrophilic solutes through mitochondrial outer membranes: Review on mitochondrial porins. Biochim. Biophys. Acta 1197, 167–196. Biggin, P.C. and Sansom, M.S.P. (1996) Simulation of voltage-dependent interactions of -helical peptides with lipid bilayers. Biophys. Chem. 60, 99–110. Blondelle, S.E., Ostreh, J.M., Houghten, R.A., and PerezPaya, E. (1995) Induced conformational states of amphipathic peptides in aqueous/lipid environments. Biophys. J. 68, 351–359. Bolter, B. and Soll, J. (2001) Ion channels in the outer membranes of chloroplasts and mitochondria: Open doors or regulated gates? EMBO J. 20, 935–940. Brown, J.W. and Huestis, W.H. (1993) Structure and orientation of a bilayer-bound model tripeptide: a 1H NMR study. J. Phys. Chem. 97, 2967–2973. Buchanan, S.K., Smith, B.S., Venkatramani, L., Xia, D., Esser, L., Palnitkar, M., Chakraborty, R., van der Helm,

5429_01_p1-17

1/21/05

11:46 AM

Page 15

ORIGIN OF MEMBRANE CHANNELS D., and Deisenhofer, J. (1999) Crystal structure of the outer membrane active transporter FepA from Escherichia coli. Nat. Struct. Biol. 6, 56–63. Cafiso, D.S. (1994) Alamethicin: A peptide model for voltage gating and protein-membrane interactions. Annu. Rev. Biophys. Biomol. Struct. 23, 141–165. Chang, G., Spencer, R.H., Lee, A.T., Barclay, M.T., and Rees, D.C. (1998) Structure of the MscL homolog from Mycobacterium tuberculosis: A gated mechanosensitive ion channel. Science 282, 2220–2226. Chipot, C. and Pohorille, A. (1998) Folding and translocation of the undecamer of poly-L-leucine across the water-hexane interface. A molecular dynamics study. J. Am. Chem. Soc. 120, 11912–11924. Choma, C., Gratkowski, H., Lear, J.D., and DeGrado, W.F. (2000) Asparagine-mediated self-association of a model transmembrane helix. Nat. Struct. Biol. 7, 161–166. Dawson, J.P., Weinger, J.S., and Engelman, D.M. (2002) Motifs of serine and threonine can drive association of transmembrane helices. J. Mol. Biol. 316, 799–805. Deamer, D.W. and Bramhall, J. (1986) Permeability of lipid bilayers to water and ionic solutes. Chem. Phys. Lipids 40, 167–188. Deamer, D.W. and Pashley, R.M. (1989) Amphiphilic compomenets of the Murchison carbonaceous chondrite: Surface properties and membrane formation. Origins Life Evol. Biosphere 19, 21–38. Deamer, D., Dworkin, J.P., Sandford, S.A., Bernstein, M.P., and Allamandola, L.J. (2002) The first cell membranes. Astrobiology 2, 371–381. DeGrado, W.F. and Lear, J.D. (1985) Induction of peptide conformation at apolar/water interfaces. 1. A study with model peptides of defined hydrophobic periodicity. J. Am. Chem. Soc. 107, 7684–7689. Doyle, D.A., Cabral, J.M., 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 basis of K conduction and selectivity. Science 280, 69–77. Duclohier, H. and Wroblewski H. (2001) Voltage-dependent pore formation and antimicrobial activity by alamethicin and analogues. J. Membr. Biol. 184, 1–12. Duff, K.C. and Ashley, R.H. (1992) The transmembrane domain of influenza A M2 protein forms amantadinesensitive proton channels in planar lipid bilayers. Virology 190, 485–489. Dworkin, J., Deamer, D., Sandford, S., and Allamandola, L. (2001) Self-assembling amphiphilic molecules: Synthesis in simulated interstellar/precometary ices. Proc. Natl. Acad. Sci. USA 98, 815–819. Engelman, D.M., Chen, Y., Chin, C.N., Curran, A.R., Dixon, A.M., Dupuy, A.D., Lee, A.S., Lehnert, U., Matthews, E.E., Reshetnyak, Y.K., Senes, A., and Popot, J.L. (2003) Membrane protein folding: Beyond the two stage model. FEBS Lett. 555, 122–125. Fisher, L.E., Engelman, D.M., and Sturgis, J.N. (1999) Detergents modulate dimerization, but not helicity, of the glycophorin A transmembrane domain. J. Mol. Biol. 293, 639–651. Fleming, K.G., Ackerman, A.L., and Engelman, D.M. (1997) The effect of point mutations on the free energy

15 of transmembrane -helix dimerization. J. Mol. Biol. 272, 266–275. Franklin, J.C., Ellena, J.F., Jayasinghe, S., Kelsh, L.P., and Cafiso, D.S. (1994) Structure of micelle-associated alamethicin from 1H NMR. Evidence for conformational heterogeneity in a voltage-gated peptide. Biochemistry 33, 4036–4045. Galbraith, T.P., Harris, R., Driscoll, P.C., and Wallace, B.A. (2003) Solution NMR studies of antiamoebin, a membrane channel-forming polypeptide. Biophys. J. 84, 185–194. Goldin, A.L. (2002) Evolution of voltage-gated Na channels. J. Exp. Biol. 205, 575–584. Gratkowski, H., Lear, J.D., and DeGrado, W.F. (2001) Polar side chains drive the association of model transmembrane peptides. Proc. Natl. Acad. Sci. USA 98, 880–885. Gullingsrud, J., Kosztin, D., and Schulten, K. (2001) Structural determinants of MscL gating studied by molecular dynamics simulations. Biophys. J. 80, 2074–2081. Hauser, H., Oldani, D., and Philips, M.C. (1973) Mechanism of ion escape from phosphatidylcholine and phosphatidylserine single bilayer vesicles. Biochemistry 12, 4507–4517. Hille, B. (2001) Ion Channels of Excitable Membranes, 3rd ed., Sinauer Associates, Inc., Sunderland, MA. Hristova, K., Dempsey, C., and White, S.H. (2001) Structure, location and lipid perturbations of melittin at the membrane interface. Biophys. J. 80, 801–811. Jan, L.Y. and Jan, Y.N. (1992) Tracing the roots of ion channels. Cell 69, 715–718. Jayasinghe, S., Hristova, K., and White, S.H. (2001) Energetics, stability, and prediction of transmembrane helices. J. Mol. Biol. 312, 927–934. Jiang, Y., Lee, A., Chen, J., Cadene, M., Chait, B.T., and MacKinnon, R. (2002a) Crystal structure and mechanism of a calcium-gated potassium channel. Nature 417, 515–522. Jiang, Y., Lee, A., Chen, J., Cadene, M., Chait, B.T., and MacKinnon, R. (2002b) The open pore conformation of potassium channels. Nature 417, 523–526. Jiang, Y., Lee, A., Chen, J., Ruta, V., Cadene, M., Chait, B.T., and MacKinnon, R. (2003a) X-ray structure of a voltage-dependent K channel. Nature 423, 33–41. Jiang, Y., Ruta, V., Chen, J., Lee, A., and MacKinnon, R. (2003b) The principle of gating charge movement in a voltage-dependent K channel. Nature 423, 42–48. Kaiser, E.T. and Kezdy, F.J. (1987) Peptides with affinity for membranes. Annu. Rev. Biophys. Biophys. Chem. 16, 561–581. Karle, I.L., Perozzo, M.A., Mishra, V.K., and Balaram, P. (1998) Crystal structure of the channel-forming polypeptide antiamoebin in a membrane-mimetic environment. Proc. Natl. Acad. Sci. USA 95, 5501–5504. Keire, D.A. and Fletcher, T.G. (1996) The conformation of substance P in lipid environments. Biophys. J. 70, 1716– 1727. Klebba, P.E. and Newton, S.M. (1998) Mechanisms of solute transport through outer membrane porins: Burning down the house. Curr. Opin. Microbiol. 1, 238–247.

5429_01_p1-17

1/21/05

11:46 AM

Page 16

16 Kloda, A. and Martinac, B. (2001a) Mechanosensitive channels in Archaea. Cell Biochem. Biophys. 34, 349–381. Kloda, A. and Martinac, B. (2001b) Mechanosensitive channels of Thermoplasma, the cell wall-less Archaea: Cloning and molecular characterization. Cell Biochem. Biophys. 34, 321–347. Kloda, A. and Martinac, B. (2002) Common evolutionary origins of mechanosensitive ion channels in Archaea, Bacteria, and cell-walled eukaryotic organisms. Archaea 1, 35–44. Kuo, A., Gulbis, J.M., Antcliff, J.F., Rahman, T., Lowe, E.D., Zimmer, J., Cuthbertson, J., Ashcroft, F.M., Ezaki, T., and Doyle, D.A. (2003) Crystal structure of the potassium channel KirBac1.1 in the closed state. Science 300, 1922–1926. Langosch, D. and Heringa, J. (1998) Interaction of transmembrane helices by a knobs-into-holes packing characteristic of soluble coiled coils. Proteins 31, 150–159. Lanyi, J.K. and Pohorille, A. (2001) Proton pumps: Mechanism of action and applications. Trends Biotechnol. 19, 140–144. Lear, J.D., Wasserman, Z.R., and DeGrado, W.F. (1988) Synthetic amphiphilic peptide models for protein ion channels. Science 240, 1177–1181. Lear, J.D., Schneider, J.P., Kienker, P.K., and DeGrado, W.F. (1997) Electrostatic effects on ion selectivity and rectification in designed ion channel peptides. J. Am. Chem. Soc. 119, 3212–3217. Lemmon, M.A., Flanagan, J.M., Hunt, J.F., Adair, B.D., Bormann, B.J., Dempsey, C.E., and Engelman, D.M. (1992) Glycophorin A dimerization is driven by specific interactions between transmembrane -helices. J. Biol. Chem. 267, 7683–7689. Liu, Y., Engelman, D.M., and Gerstein, M. (2002) Genomic analysis of membrane protein families: abundance and conserved motifs. Genome Biol. 3.research0054.1-0054. 12. Available online at: http://genomebiology.com/ 2002/3/10/research/0054. Locher, K.P., Rees, B., Koebuik, R., Mitschler, A., Moulinier, L., Rosenbusch, J.P., and Moras, D. (1998) Transmembrane signaling across the ligand-gated FhuA receptor: Crystal structures of free and ferrichromebound states reveal allosteric changes. Cell 95, 771–778. Lu, Z., Klem, A.M., and Ramu, Y. (2001) Ion conduction pore is conserved among potassium channels. Nature 413, 809–813. MacKenzie, K.R., Prestegard, J.H., and Engelman, D.M. (1997) A transmembrane helix dimer: Structure and implications. Science 276, 131–133. Mannella, C.A. (1997) Minireview: On the structure and gating mechanism of the mitochondrial channel, VDAC. J. Bioenerg. Biomembr. 29, 525–531. Martinac, B. (2001) Mechanosensitive channels in prokaryotes. Cell. Physiol. Biochem. 11, 61–76. Miller, S., Edwards, M.D., Ozdemir, C., and Booth, I.R. (2003) The closed structure of the MscS mechanosensitive channel. Cross-linking of single cysteine mutants. J. Biol. Chem. 278, 32246–32250. Montal, M. (1995) Design of molecular function: Channels of communication. Annu. Rev. Biophys. Biomol. Struct. 24, 31–57.

POHORILLE ET AL. Murata, K., Mitsuoka, K., Hirai, T., Walz, T., Agre, P., Heymann, J.B., Engel, A., and Fujiyoshi, Y. (2000) Structural determinants of water permeation through aquaporin-1. Nature 407, 599–605. Nagle, J.F. and Scott, H.L. (1978) Lateral compressibility of lipid mono- and bilayers. Theory of membrane permeability. Biochim. Biophys. Acta 513, 236–243. Nikaido, H. and Rosenberg, E.Y. (1981) Effect on solute size on diffusion rates through the transmembrane pores of the outer membrane of Escherichia coli. J. Gen. Physiol. 77, 121–135. Nozaki, Y. and Tanford, C. (1981) Proton and hydroxide ion permeability of phospholipid vesicles. Proc. Natl. Acad. Sci. USA 78, 4324–4328. O’Reilly, A.O. and Wallace, B.A. (2003) The peptaibol antiamoebin as a model ion channel: Similarities to bacterial potassium channels. J. Pept. Sci. 9, 769–775. Parent, R. and Roy, P.H. (1992) The chloramphenicol acetyltransferase gene of Tn2424: A new breed of cat. J. Bacteriol. 174, 2891–2897. Parsegian, V.A. (1969) Energy of an ion crossing a low dielectric membrane: Solutions to four relevant electrostatic problems. Nature 221, 844–846. Paula, S., Volkov, A.G., and Deamer, D.W. (1998) Permeation of halide anions through phospholipid bilayers occurs by the solubility-diffusion mechanism. Biophys. J. 74, 319–327. Pinto, L.H., Holsinger, L.J., and Lamb, R.A. (1992) Influenza virus M2 protein has ion channel activity. Cell 69, 517–528. Pinto, L.H., Dieckmann, G.R., Gandhi, C.S., Papworth, C.G., Braman, J., Shaughnessy, M.A., Lear, J.D., Lamb, R.A., and DeGrado, W.F. (1997) A functionally defined model for the M2 proton channel of influenza A virus suggests a mechanism for its ion selectivity. Proc. Natl. Acad. Sci. USA 94, 11301–11306. Piontkivska, H. and Hughes, A.L. (2003) Evolution of vertebrate voltage-gated ion channel -chains by sequential gene duplication. J. Mol. Evol. 56, 277–285. Pivetti, C.D., Yen, M.R., Miller, S., Busch, W., Tseng, Y.H., Booth, I.R., and Saier, M.H. Jr. (2003) Two families of mechanosensitive channel proteins. Microbiol. Mol. Biol. Rev. 67, 66–85. Pohorille, A. and Wilson, M.A. (1996) Excess chemical potential of small solutes across water-membrane and water-hexane interfaces. J. Chem. Phys. 104, 3760– 3773. Pohorille, A., Chipot, C., New, M.H., and Wilson, M.A. (1996) Molecular modeling of protocellular functions. In Pacific Symposium on Biocomputing ‘96, edited by L. Hunter and T.E. Klein, World Scientific, Singapore, pp. 550–569. Pohorille, A., Wilson, M.A., and Chipot, C. (2003) Membrane peptides and their role in protobiological evolution. Origins Life Evol. Biosphere 33, 173–197. Popot, J.L. and Engelman, D.M. (1990) Membrane protein folding and oligomerization: The two-stage model. Biochemistry 29, 4031–4037. Popot, J.L. and Engelman, D.M. (2000) Helical membrane protein folding, stability, and evolution. Annu. Rev. Biochem. 69, 881–922.

5429_01_p1-17

1/21/05

11:46 AM

Page 17

ORIGIN OF MEMBRANE CHANNELS Pratt, L.R. and Pohorille, A. (2002) Hydrophobic effects and modeling of biophysical aqueous solution interfaces. Chem Rev. 102, 2671–2692. 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. Rode, B.M. (1999) Peptides and the origin of life. Peptides 20, 773–786. Roux, B. and MacKinnon, R. (1999) The cavity and pore helices the KcsA K channel: Electrostatic stabilization of monovalent cations. Science 285, 100–102. Roux, B., Berneche, S., and Im, W. (2000) Ion channels, permeation, and electrostatics: Insight into the function of KcsA. Biochemistry 44, 13295–13306. Sachs, F. (1992) Stretch sensitive ion channels: An update. In Sensory Transduction, edited by D.P. Corey and S.D. Roper, Rockefeller University Press, New York, pp. 241–260. Saghatelian, A., Yokobayashi, Y., Soltani, K., and Ghadiri, M.R. (2001) A chiroselective peptide replicator. Nature 409, 797–801. Saier, M.H. (2003) Tracing pathways of transport protein evolution. Mol. Microbiol. 48, 1145–1156. Sansom, M.S.P. (1993) Structure and function of channelforming peptaibols. Q. Rev. Biophys. 26, 365–421. Sasatsu, M., Shima, K., Shibata, Y., and Kono, M. (1989) Nucleotide sequence of a gene that encodes resistance to ethidium bromide from a transferable plasmid in Staphylococcus aureus. Nucleic Acids Res. 17, 10103– 10103. Schirmer, T., Keller, T.A., Wang, Y.F., and Rosenbusch, J.P. (1995) Structural basis for sugar translocation through maltoporin channels at 3.1 Å resolution. Science 267, 512–514. Schweighofer, K.J. and Pohorille, A. (2000) Computer simulation of ion channel gating: The M2 channel of influenza A virus in a lipid bilayer. Biophys. J. 78, 150–163. Smith, S.O., Jonas, R., Braiman, M., and Bormann, B.J. (1994) Structure and orientation of the transmembrane domain of glycophorin A in lipid bilayers. Biochemistry 33, 6334–6341. Snook, C.F., Woolley, G.A., Oliva, G., Pattabhi, V., Wood, S.P., Blundell, T.L., and Wallace, B.A. (1998) The structure and function of antiamoebin I, a proline-rich membrane-active polypeptide. Structure 6, 783–792. Song, L., Hobaugh, M.R., Shustak, C., Cheley, S., Bayley, H., and Gouaux, J.E. (1996) Structure of staphylococcal -hemolysin, a heptameric transmembrane pore. Science 274, 1859–1866. Stevens, T.J. and Arkin, I.T. (2000) Do more complex organisms have a greater proportion of membrane proteins in their genomes? Proteins 39, 417–420. Strong, M., Chandy, K.G., and Gutman, G.A. (1993) Molecular evolution of voltage-sensitive ion channel genes: On the origins of electrical excitability. Mol. Biol. Evol. 10, 221–242. Stuhmer, W., Conti, F., Suzuki, H., Wang, X.D., Noda, M., Yahagi, N., Kubo, H., and Numa S. (1989) Structural

17 parts involved in activation and inactivation of the sodium channel. Nature 339, 597–603. Tam, R. and Saier, M.H. Jr. (1993) Structural, functional, and evolutionary relationships among extracellular solute-binding receptors of bacteria. Microbiol. Rev. 57, 320–346. Thomas, L., Kocsis, E., Colombini, M., Erbe, E., Trus, B.L., and Steven, A.C. (1991) Surface topography and molecular stoichiometry of the mitochondrial channel, VDAC, in crystalline arrays. J. Struct. Biol. 106, 161–171. Verdon, G., Albers, S.V., Dijkstra, B.W., Driessen, A.J., and Thunnissen, A.M. (2003) Crystal structures of the ATPase subunit of the glucose ABC transporter from Sulfolobus solfataricus: Nucleotide-free and nucleotidebound conformations. J. Mol. Biol. 330, 343–358. Vlassov, A., Khvorova, A., and Yarus, M. (2001) Binding and disruption of phospholipid bilayers by supramolecular RNA complexes. Proc. Natl. Acad. Sci. USA 98, 7706–7711. Wallace, B.A. (1990) Gramicidin channels and pores. Annu. Rev. Biophys. Biophys. Chem. 19, 127–157. Wallin, E. and von Heijne, G. (1998) Genome-wide analysis of integral membrane proteins from eubacterial, archaean, and eukaryotic organisms. Protein Sci. 7, 1029–1038. Walz, T., Hirai, T., Murata, K., Heymann, J.B., Mitsuoka, K., Fujiyoshi, Y., Smith, B.L., Agre, P., and Engel, A. (1997) The three-dimensional structure of aquaporin-1. Nature 387, 624–627. Wilson, M.A. and Pohorille, A. (1996) Mechanism of unassisted ion transport across membrane bilayers. J. Am. Chem. Soc. 118, 6580–6587. Zhong, Q., Husslein, T., Moore, P.B., Newns, D.M., Pattnaik, P., and Klein, M.L. (1998) The M2 channel of influenza A virus: A molecular dynamics study. FEBS Lett. 434, 265–271. Zhou, F.X., Cocco, M.J., Russ, W.P., Brünger, A.T., and Engelman, D.M. (2000) Interhelical hydrogen bonding drives strong interactions in membrane proteins. Nat. Struct. Biol. 7, 154–160. Zhou, F.X., Merianos, H.J., Brünger, A.T., and Engelman, D.M. (2001) Polar residues drive association of polyleucine transmembrane helices. Proc. Natl. Acad. Sci. USA 98, 2250–2255. Zhou, Y. and MacKinnon, R. (2003) The occupancy of ions in the K selectivity filter: Charge balance and coupling of ion binding to a protein conformational change underlie high conduction rates. J. Mol. Biol. 333, 965–975.

Address reprint requests to: Dr. Andrew Pohorille MS 239-4 NASA Ames Research Center Moffett Field, CA 94035 E-mail: [email protected]

Review The Origin and Early Evolution of

Review The Origin and Early Evolution of

Suggest Documents

The Place of RNA in the Origin and Early Evolution

Early evolution of polyisoprenol biosynthesis and the origin of ... - PeerJ

Geochemical constraints on the origin and evolution of early Mesozoic ...

Correction: The Origin and Early Evolution of Sauria ...

Origin and Early Evolution of the Genus Homo

The origin and early evolution of metatherian mammals - CiteSeerX

Origin and early evolution of pterosaurs - University of Birmingham

Origin and early evolution of pterosaurs - University of Birmingham

The Gamburtsev mountains and the origin and early evolution of the ...https://www.researchgate.net/...Antarctic.../The-Gamburtsev-Mountains-and-the-origin...

The origin and evolution of synapses

The origin and evolution of lignin biosynthesis

Origin and Evolution of the Cepheus Bubble

The Origin and Evolution of Photosynthetic Oxygen

The Origin and Evolution of Religious Prosociality

The origin and evolution of human glutaminases

Origin and Evolution of the Sodium -Pumping

THE ORIGIN AND EVOLUTION OF CELESTIAL BODIES ...

Origin and evolution of the peroxisomal proteome

On the origin of early type galaxies and the evolution of the interaction ...

Origin and evolution of mineralizing fluids and

palaeopathology: studying the origin, evolution and ...

The First Humans - Origin and Early Evolution of the Genus Homo

The Place of RNA in the Origin and Early Evolution ... - Semantic Scholar

pdf-1160\the-first-humans-origin-and-early-evolution-of-the ... - Google