CFTR Gating: Phosphorylation and ATP

The identification of the primary structure of CFTR, in retrospect at least, foreshadowed the rather complex sequence of events that underlies the gating of CFTR channels (Fig. 1). The recognition (by virtue of homology with other proteins) of a regulatory or “R domain” containing consensus sites for phosphorylation by protein kinase A, as well as two nucleotide binding domains likely to bind and hydrolyze ATP, provided a mechanistic link to the previously inferred cAMP-dependence of chloride secretion and implied a novel role for cytosolic ATP in this process. Detailed functional studies²¹²⁸ as well as emerging structural information for NBD homologs,²⁹³⁰ has defined a zeroth order model for the gating process in which phosphorylation of sites on the R domain (and perhaps elsewhere) is necessary, but not sufficient, for channel opening. Once CFTR is primed by phosphorylation, ATP is required for channel opening, and likely channel closing as well.^{23,25,2728,3136} The role of the R domain appears to be in a sense, permissive, in that channels lacking portions of this structure can be opened by exposure to ATP absent any phosphorylation.^22,24,3738 These complex gating phenomena have been recently reviewed and we direct the reader to these sources for details.^25,3942 We will have occasion to address CFTR gating only in as much as some structural changes in the transmembrane segments that modify permeation also impact gating.⁴³ In the long run one may hope that these sorts of observations might lead to an appreciation of the physical nature of the gating events which could, for example, involve movement of one or several of the TMs initiated by ATP hydrolysis at the NBDs.

Permeability and Conductance: Anions Report the Nature of the Pore Interior

The absence of the three-dimensional structure for CFTR places us squarely in the center of the common dilemma of having rather little to go on as regards realistic models for the “pore domain.” Our only guide is the results of functional and biochemical studies viewed through the lens of basic principles that are expected to govern anion-water and anion-protein interactions. The proposals offered here are meant to serve as the basis for discussion and are derived essentially from three sources. A primary source will be the very classic strategy of interpreting selectivity patterns. Here we focus on differences in permeation and binding from one anion to another and ask, “what sort of anion-channel interaction might give rise to the observed patternβ” Although such inferences cannot reveal structure, they might foreshadow aspects of the physical nature of the pore interior that would find a structural basis in later studies. This decidedly classical approach that dates to the earliest studies of ion conduction⁴⁴⁴⁶ has recently been at least partially validated by the crystal structure of the bacterial K channel which revealed the design of a “selectivity filter” based on an array of carbonyl oxygens, much as had been predicted by earlier studies of selectivity patterns in K-selective channels and macro cyclic antibiotics.⁴⁷⁴⁹ Despite the fact that the K channel structure may have no direct implications for anion channels, it brings into focus the value of quantitatively considering the energetics of transferring anion from water into the channel, in particular the central role of hydration energy and the requirement that anion-water and anion-channel interaction energies be balanced so that ion conduction is possible (Fig. 2).⁵⁰⁵¹

Figure 2

Energy landscape representation of anion conduction adapted from Andersen and Procopio. (A) Ion permeation through a channel can be viewed as comprising three steps: leaving water and entering the channel (dehydration and solvation), translocating within (more...)

The second source of inferences about the conduction process will be studies in which the properties of structurally altered channels are compared with those of the wild type. At this point there have been many studies of the functional properties of CFTR constructs bearing point mutations^31,5258 as well as investigation of the properties of fragments of CFTR proteins. ^27,5962 Unfortunately, these constructs have been studied in detail in only a few instances, but we will attempt to provide some overview of the effects of structural changes and their implications where possible. For example, one issue of current interest is the nature of the normal functional unit of CFTR (monomer, dimer, etc.) and the minimal functional unit, i.e., do fragments retain channel function?

We will also discuss the implications of somewhat more subtle modifications of CFTR brought about by covalent and noncovalent modifications of constructs bearing cysteines or histidines engineered into specific locations. Such studies, while they suffer from some of the same limitations as mutagenesis, offer the important advantage of enabling the reversible modification of the protein in real-time. Here we will focus on recent evidence that charged residues may play a role in optimizing the pore environment for anion conduction and propose that reversible charge deposition may provide a crude approach to defining residues that lie within the anion-selective pore.

Additional fuel for speculation is the recently derived structure for the bacterial ClC chloride channel^20,20a that provided, for the first time, a working model for a permeation pathway that favors anions over cations.

Permeability Selectivity

Models for Permeation

Regardless of whether one regards permeation as a simple or complex process, most would agree that it is virtually impossible to think about it in any productive way without resort to some model that provides a physical interpretation of measured parameters like permeability and conductance. In this age of high-speed computing a hierarchy of models is now potentially available that range from consideration of the motions of individual molecules (molecular dynamics) to continuum or kinetic models based on fairly classical macroscopic interpretations of microscopic processes.⁷⁸ It is to the latter, however, that most of us turn for the conceptual framework that is a necessary adjunct to everyday thinking and conversation about the results of permeation studies.

Two types of conceptual models have dominated thinking about permeation mechanisms, and at least a passing familiarity with them is a prerequisite for any attempt to address permeation data. One is the continuum, electrodiffusion approach based on the Nernst-Planck equation and the other is rate theory.⁷⁸⁸³ The central parameter in the continuum model is the permeability, P_i, usually defined as follows,

P_i = β_iD_iA/l      Eqn. 1

Here we envision the channel as an equivalent cylinder of effective cross-sectional area, A, and length, l. Inside the channel anions move by diffusion in accord with the diffusion coefficient, D_i, which may differ from that in free solution. The boundaries of the channel cylinder (the channel-bathing solution interfaces) are envisioned as being near equilibrium and the distribution of the permeant ion between the bulk solution and the channel is described by a partition coefficient, β_i, given by:

β_i = (C_i)_pore/(C_i)_soln       Eqn. 2

where (C_i)_soln and (C_i)_pore are the concentrations of the anion in the bulk solution just outside the channel and just inside the channel, respectively. This expression is central to a solubility-diffusion model of the channel that can be used to operationally define permeability ratios for ions (see below and refs. 46, 57). The central value of this concept lies in the fact that the partition coefficient can be related to the work required to move an anion from the aqueous solution into the channel, ΔG_eq,

β_i = exp(−ΔG_eq/RT)       Eqn. 3

where R and T have their usual significance.

To simplify the interpretation of permeability ratios by means of the solubility diffusion model we will assume that variation in D_i is negligible compared to variations in β_i. There is no empirical justification for this assumption, but the procedure amounts to lumping all of the interaction energy into the term β_i. In the rate theory approach, a symmetrical two-barrier model gives the same result (see below) and this issue disappears because the free energy barriers determine translocation rates (see also ref. 84).

The permeability ratio, P_i/P_j, therefore can be envisioned as:

P_i/P_j = β_i/β_j = exp([-ΔG_eq]_ij /RT)       Eqn. 4

where it is assumed that D_i = D_j and [ΔΔG_eq]_ij is given by

[ΔΔG_eq]_ij = (ΔG_eq)_i −(ΔG_eq)_j       Eqn. 5

so that the permeability ratio, P_i/P_j, is seen as a measure of the difference in the work required to move each of the two ions from the solution into the channel. By making use of a vacuum reference phase^51,85 each of these equilibrium free energy terms can be decomposed into the difference between ion-water interaction energies (hydration energies) and ion-channel interaction energies (referred to herein as solvation energies), i.e.

[ΔG_eq]_i = (ΔG_solv)_i −( ΔG_hyd)_i       Eqn. 6

[ΔG_eq]_j = (ΔG_solv)_j −( ΔG_hyd)_j       Eqn. 7

These equations provide a formal statement of the fact that the ion must trade its stabiliation by bulk water for stabilization by the protein interior, and that it is the difference between these free energies that is a major determinant of permeation.

This notion can be extended to the notion of selectivity by way of equation 5, i.e.,

[ΔΔG_eq]_ij = (ΔΔG_solv)_ij −(ΔΔG_hyd)_ij       Eqn. 8

where:

[ΔΔG_solv]_ij = (ΔG_solv)_i − (ΔG_solv)_j       Eqn. 9

[ΔΔG_hyd]_ij = (ΔG_hyd)_i−(ΔG_hyd)_j       Eqn. 10

The permeability ratio provides an estimate of ΔΔG_eq. Because hydration energies are either known or can be estimated,^51,65 the permeability ratio yields information about [ΔΔG_solv]_ij, the difference in the ion-channel interaction energies from one ion to another. Thus a sequence or pattern of permeabilities can be loosely interpreted in terms of a pattern of solvation energies and compared with known anion interactions in simple physical environments such as those found in synthetic anion-selective electrodes⁸⁶⁸⁷ or ionophores.⁸⁸

As indicated earlier, terms like ion-channel interaction energy or solvation energy are not intended to convey anything about the physical nature of the ion-channel interactions. The ion must be stabilized during its entire progress through the pore and it is likely that the physical nature of the stablization will vary along the length of the channel. For example, if a channel contains a relatively large vestibule, an ion could retain its inner-sphere waters upon entry, but then be forced to discard them as it traverses a narrow region where it is stabilized by ligands contributed by amino acid side chains or the peptide backbone. Precisely, this situation is illustrated clearly by the images of the bacterial K channel, KcsA.^6667,8990

The rate theory approach provides a connection between selectivity and the energetics of permeation via a somewhat different route. Here the permeation path is viewed as a landscape of energy barriers and wells. The ratio of permeabilities for the simplest of such models (two barriers and one well) depends only on the peak energies,^46,83,91 such that we obtain a result identical to that derived from solubility-diffusion, i.e.

P_i/P_j = exp ([−ΔΔG_peak]_ij/RT)       Eqn. 11

where

[ΔΔG_peak]_ij = [ΔG_peak]_i−[ΔG_peak]_j       Eqn. 12

and ΔG_peak is equivalent to ΔG_eq in the Nernst-Planck model.⁵¹

This approach leads to an interpretation of the permeability ratio, P_i/P_j, as being due to the difference in barrier heights, once again taken to reflect the differences in the energies of hydration and solvation experienced by the two anions. As we will see farther along the Rate Theory approach, which in its simplest form is akin to a compartmental analysis of channel function,⁹¹ provides a conceptually attractive approach to the interpretation of anion binding within the pore that is not offered by the simplest version of the electrodiffusion approach (see, however, refs. 82,92). The equilibrium occupancy of a binding site can be related quatitatively to the depth of the energy well so that, again, the relative magnitude of anion-channel interaction energies can be estimated. The simplest model attributes ion binding to a single site, but we note here and discuss below, the possibility of multiple ion binding sites in the CFTR pore.^52,9394

Patterns of Permeability Selectivity and the Mechanism of Anion Solvation: CFTR as a Polarizable Tunnel

Historically, the goal of selectivity studies has been to reveal a pattern of selectivity that can be identified with the behavior of ions in a simple physical system, and therefore used to infer the existence of a particular type of ion-channel interaction in a protein. For example, the cation selectivity of ion channels was compared to that exhibited by ion-selective, macrocyclic antibiotics like valinomycin^48,102 and Smith et al⁵¹ compared the anion selectivity of CFTR to that of synthetic plastic membranes.^8687,103 The consideration of such a broad range of physical systems naturally provokes the question, “what does it mean to say that a membrane or a channel is selective?;” Even water must be regarded as exhibiting selective interactions with ions; small ions (like F^-) interact more strongly with water than large ions (like SCN^-) and this is reflected in their increased energy of hydration. A related question concerns the relation between the magnitude of selectivity (e.g., the ratio P_SCN/P_Cl) and the pattern of selectivity (e.g., permeability increases monotonally with size). The work of Eisenman and colleagues,^50,104107 emphasized the physical basis for selectivity patterns and Lewis and Stevens¹⁰⁸ pointed out the fact that the magnitude of selectivity within a certain pattern could vary.

Moyer and Bonneson¹⁰⁹ in their review of the physical basis for anion separation introduced a conceptual framework that provides a useful basis for organizing the types of selectivity seen in the natural world. They propose that selectivity is best viewed as a continuum. At one end of the scale selectivity is based on simple, physico-chemical “bias” that is a product of fundamental properties like the size of the ion and the local dielectric properties of the medium. Bias-type selectivity is characterized by monotonic patterns (with regard to ion size) and little dependence on local geometry. The other extreme of the scale is ion “recognition”, characterized by a peak in the selectivity-size relation, high sensitivity to local geometry (e.g., protein structure) and increased exclusion of the ion from water. The structure of the bacterial K channel revealed that the physical basis for the recognition of K ions was a structurally specific array of carbonyl oxygen ligands, as foreshadowed by studies of ionophores^48,102 and the selectivity studies of Benzanilla and Armstrong.⁴⁹

Anion permeability selectivity for the CFTR channel follows, with one notable exception (iodide), the so-called lyotropic selectivity pattern in which permeability increases monotonically with decreasing hydration energies.¹¹⁰ Smith et al⁵¹ used an extended series of halide and pseudo halide anions to probe the wild type CFTR pore and found that the permeability ratios (P_x/P_Cl) fell in the order C(CN)₃^- (7.8) >Au(CN)₂^- (6.67) >N(CN)₂^- (4.49) >SCN^- (3.41)>NO₃^- (1.42) >Br^- (1.22)>Cl_- (1.0), also known as the Hofmeister series after the German physical chemist Franz Hofmeister who studied the effects of anions on the stability of proteins¹¹⁰. This sort of pattern (large to small) and the relative non-selectivity of the pore suggest that CFTR exhibits bias-type selectivity that is not likely to depend on a structural specialization, akin to the “selectivity filter” of KcsA. This impression is reinforced by the similarity of the pattern to that exhibited by plastic membranes used in ion-exchange electrodes.^51,111

Smith et al⁵¹ analyzed the selectivity patterns for wild type CFTR by plotting the transfer energies (peak energies in rate theory) for anions vs. the reciprocal anionic radius (Fig. 3). They found that ΔG_eq varied as a linear function of 1/r, a pattern identical to that exhibited by plastic membranes used in anion-sensitive electrodes. Because ΔG_hyd is also a linear function of 1/r, this implied that apparent solvation energy (ΔG_solv), a measure of the anion-channel interaction, also varied linearly with 1/r. The linear relation between ΔG_solv and 1/r suggested that anion solvation could be analyzed by using the Born equation (Fig. 3) to calculate the “effective” dielectric constant seen by the anion in the pore. By fitting the relation with the Born equation, it was possible to model the solvation process as the interaction of the anion with a polarizable space with an effective dielectric constant of about 20. The dielectric constant is a measure of the polarizability of a medium, like water, that possesses a permanent or inducible dipolar character. The notion of an “effective” dielectric constant (see also below) offers a purely phenomenological description of the apparent “polarizability” of the channel interior that may include contributions from the protein side chains and backbone, and water within the channel, as well as from the lipid bilayer and the water at the membrane-solution interface.^63,112122

Figure 3

Analysis of the pattern of the difference in equilibrium free energies of transfer (ΔG_eq) derived from the measurements of permeability ratios. (A) Transfer of an anion from aqueous solution into the channel is viewed as moving the anion from (more...)

The relative insensitivity of permeability ratios to point mutations led Smith et al.⁵¹ to propose that the CFTR pore could be viewed as a polarizable tunnel, in which permeability selectivity was not highly dependent on a structural specialization such as that seen in the selectivity filter of the bacterial K channel^6667,8990 This view is supported by the fact that a number of non-homologous chloride channels (e.g., GABA_AR, GlyR, T84 endogenous Cl^-channel) exhibit qualitatively similar selectivity patterns.

What about iodide? In most CFTR selectivity studies iodide has been reported to exhibit a permeability, relative to that of chloride, that is markedly less than that predicted on the basis of its hydration energy^91,123 The analysis of Smith et al⁵¹ predicted a value of P_I/P_Cl of at least 2, whereas they observed a value of about 0.4. One possible explanation for this anomalous behavior is that iodide interacts with the channel in such a way as to modify its properties beyond that associated with simple anionic blockade (see below). Tabcharani et al,^98,124 in fact, reported observations consistent with this notion. They found that when human CFTR was studied in detached patches from transfected cells, the value of P_I/P_Cl varied with time after exposure to the ion, starting close to the predicted lyotropic value and diminishing with time such that P_I/P_Cl < 1.0. Smith et al⁵¹ reported macroscopic measurements suggesting that exposure of CFTR-expressing oocytes to external iodide produced a long-lasting change in channel properties, although they were unable to discern time-dependent changes in P_I/P_Cl in this experimental setting.

The dielectric stabilization scheme of Smith et al⁵¹ provides a crude explanation for the anion selectivity patterns seen with CFTR, but the importance of the model lies more in the fact that it provides a physically-based reference point from which to view the selectivity of mutant CFTR constructs or other anion-selective channels, like ClC-0. The relatively straightforward analysis of anion partitioning between water and the polarizable tunnel serves to focus attention on several issues that pertain to any analysis of anion selectivity. First, as emphasized by Eisenman and colleagues^50,104107 and Lewis and Stevens,¹⁰⁸ the most relevant physical parameter bearing upon selectivity is likely to be ion size expressed as the reciprocal of the ionic radius. The use of 1/r rather than diameter, for example, as a standard of comparison is the best starting place because hydration energies (as well as channel solvation energies in the simple model) vary roughly linearly with this parameter. A “1/r plot” is likely to be more informative than simply correlating permeability ratios (or log P_X/P_Cl) with ΔG_hyd, because as ΔG_hyd varies so will, in general, ΔG_solv. The model also focuses attention on the simple, but critical fact that it is the difference between ΔG_hyd and ΔG_solv that determines selectivity, so that, as discussed below, differences in anion binding cannot necessarily be equated with differences in anionchannel interaction energies. The dielectric stabilization notion also provides an intuitive guide to the use of terms such as “weak field selectivity.” In the case of CFTR, commonly credited with weak field selectivity,^54,98 the apparent anion-channel interaction energy is hardly weak. Rather it is roughly equal to the anion-water interaction energy, as it must be if anion conduction is to be possible.⁵⁰⁵¹ Finally the model offers an example of the important distinction between patterns of selectivity and the degree of selectivity within a pattern. The model predicts, for example, that reducing the effective dielectric constant will increase lyotropic selectivity, as measured by the slope of the 1/r plot (see for example plots for GABA_AR and GlyR in Smith et al⁵¹). A departure from the lyotropic pattern, such as that seen in ClC-0 where P_SCN is less than P_Cl, should provoke a search for some aspect of the anion-channel interaction to which this deviation from the primitive pattern can be attributed¹²⁵

Binding Selectivity

The Energetics of Anion Binding

To understand anion binding it is useful to begin with the notion that an anion (or a cation for that matter), be it in the bulk solution, within the channel or in some transition state is, in a sense, always “bound.” That is to say the anion is at all times coordinated by ligands that stabilize it by, in one way or another, attenuating or counteracting the strong electric field created by the ion's charge. In bulk solution these ligands are water molecules and it is the coordination of the anion by its surrounding water dipoles that makes the existence of separate anions and cations in a solution energetically feasible. We don't think of anions (or cations) in solution as being "bound" because the anion-water complex is highly mobile and individual H₂O ligands are being continuously replaced on a nanosecond time scale.^48,126 The "affinity" of the binding of the ion for bulk water is measured by the hydration energy (ΔG_hyd), the energy required to remove the ion from bulk water to a vacuum reference state.⁸⁵ Small ions with a more punctate charge distribution are bound to water more tightly than large anions, a result that Born⁸⁵ showed to be predictable on the basis of a simple electrostatic model of the anion-water interaction. In the present context it is worth noting that anions are bound more tightly to water than are cations of equivalent radius, presumably due to the difference in the nature of the interactions of the H₂O molecules with these ions that accrue from the obligatory differences in the orientation of the H₂O dipoles about centers of negative charge and positive charge.^57,63,127

Within a channel, the most unambiguous definition of a “bound” ion is a configuration in which some portion of the inner sphere water molecules are replaced by ligands that are fixed to the “wall” of the channel, as would be most likely in a relatively narrow region of the pore. In this case binding has its more conventional meaning in that, because the wall of the channel is fixed in space, inner-sphere interactions with the channel ligands must impede the movement of the anion through the pore. Smith et al⁵¹ modeled anion binding by CFTR using the highly simplified scheme diagramed in Figure 4 in which the total energy of an anion at any point in the channel was viewed as the sum of two components. One was a “barrier component” reflecting the slight excess of hydration energy over anion solvation energy within the dielectric tunnel. Such a dielectric tunnel would exhibit a lyotropic permeability sequence, but would not bind anions (Fig. 4A). Anion binding was simulated by envisioning a narrow portion of the channel in which elements of the channel lining contact the anion. This “inner sphere” component of the total anion-channel interaction energy was defined conceptually as that energy which would describe the equilibrium association of anions with the channel lining if this interaction occurred in an aqueous medium. In this, albeit hypothetical, situation the background stabilization is identical to that seen in water so that binding depends only on the inner-sphere interactions (Fig. 4B). This conceptual separation of inner sphere interaction energy at the binding site from the “background solvation” due to the polarizable entities beyond the inner-sphere was consonant with the observation that anion binding, as measured by the ability of substitute anions to block Cl^- flow, seems to be more easily disrupted by mutations than permeability selectivity.⁵⁷ Summing the resulting energy barriers and well gives rise to a two barrier, one well model (Fig. 4C)

Figure 4

Modeling the energy landscape of anion permeation. (A) A “barrier component” reflecting the slight excess of hydration energy over anion solvation energy within the dielectric tunnel. Such a dielectric tunnel would exhibit a lyotropic (more...)

So why do anions that exhibit higher permeability ratios than chloride (one may think of these anions as entering the channel more readily than chloride) also bind more tightly than chloride within the pore, so that they have a tendency to block chloride flow? Faced with this observation, it is natural to imagine that a blocking anion, like SCN^-, must interact more strongly than Cl^- with the ligands contributed by the channel wall. Any attempt to quantify the “binding energy” (more appropriately the anion-channel interaction energy), however, immediately reveals a problem with this reasoning. If the “binding energy” is equated with the depth of an energy well in the rate theory model for anion flow, it must be acknowledged that the depth of this well is not solely dependent on the anion-channel interaction energies. Rather, it depends on the difference between this energy and hydration energy, both of which vary from one anion to another. A quantitative analysis of the block of CFTR by permeant anions,⁵¹ shown diagrammatically in Fig. 5, revealed that a blocking anion like SCN^-, because of its larger size, is likely to experience a total anion-channel interaction energy that is, if anything, less than that of Cl^-. In this model the increased well depth for SCN^- is attributable to the fact that as anion size increases the hydration energy is reduced more than is the total anion-channel (solvation) energy. Hence, well depth is greater for the larger anion and we detect that it “binds more tightly,” i.e., that it is a more efficacious blocker of Cl^- flow. Thus, a relatively simple model for the channel as a dielectric tube can account for the fact that both the permeability and the binding selectivity exhibit a lyotropic pattern and that more permeant anions also bind more tightly, largely because they more readily escape water.

Figure 5

(A) Diagrammatic representation of solvation energies estimated from relative permeability ((ΔG_solv)_p) and relative block ((ΔG_solv)_b) plotted vs. 1/r. The total solvation energy seen at the anion binding site (ΔG_solv)_b is depicted (more...)

The diagrams in Fig. 5 are an attempt to illustrate the predictions of the simple model for an energy profile containing two-barriers and one-well in which the peak and well energies are determined by the difference between ΔG_hyd and ΔG_solv. As expected for an energy barrier (as judged by permeability ratios), values of (ΔG_solv)_p (see Fig. 5 and legend) are less than ΔG_hyd for all anions, large or small, but the difference, ΔG_peak, varies with anion size. For an energy well values of (ΔG_solv)_b (see Fig. 5 and legend), on the other hand, are uniformly larger that ΔG_hyd as expected. Here again the difference between (ΔG_solv)_b and ΔG_hyd varies with 1/r in such a way as to produce a deeper well for the larger (and less well hydrated) anions. The different slopes of the respective energies on 1/r produce a reciprocal variation in peak height and well depth like that observed for CFTR. As regards the plots of ΔG_solv for anion entry (ΔG_solv)_p and for anion binding (ΔG_solv)_b, it is important to note that the increase in blocking efficacy with increased permeability, the hallmark of the lyotropic pattern, would not be observed if the slope of these two plots were identical. In this case the barrier to anion exit from the channel (ΔG_peak + ΔG_well) would not change from one anion to another so that the exit rate from an occupied channel would be independent of anion size.

The behavior reported in Smith et al⁵¹ requires the change in background solvation with 1/r at the binding site to be somewhat less than that implicit in the effective dielectric constant derived from permeability measurements. The dielectric constant at the binding site pertains only to the effect of polarizable moieties that lie beyond the inner-sphere ligands contributed by the protein, whereas the “effective” dielectric constant calculated from permeability measurements embraces the entire environment of the anion, including the inner-sphere. By separating out the inner-sphere contribution to the anion-channel interaction energy at the binding site we have made explicit a portion of the energetics that would be implicit in an effective dielectric constant. The assumption of Smith et al⁵¹ that the inner-sphere energies do not vary greatly from one amino acid to another is likely to be a gross oversimplification, in as much as the halides and pseudo halides used to probe the pore varied significantly in their shape, so that it is clearly possible that the preference for SCN^- over Cl^-, for example, could reflect in part increased inner-sphere, anion-channel energy for SCN^-.

Does CFTR Have a Discrete “Selectivity Filter?”

One implication of a polarizable tunnel permeation scheme is that lyotropic permeability selectivity (assayed by means of shifts in reversal potential) could be achieved in the absence of a discrete, structurally identifiable region that serves as a “selectivity filter,” comparable to that visualized within the KcsA channel. Smith et al⁵¹ proposed that the CFTR pore might represent the most primitive sort of channel characterized by what one might describe as a “natural”, bias-based selectivity pattern that reflects simple dielectric stabilization of anions within the pore. Nevertheless, it is difficult to imagine binding selectivity in the absence of some structurally distinct region of the pore (e.g., the narrow region of Smith et al⁵¹) where elements of the protein are in contact with the anion. It might thus be argued that entering the narrow region (say from a vestibule) might constitute the rate limiting barrier to ion flow and that it would be the relative height of these barriers that is measured in anion substitution experiments (see for example, ref. 81). In this case, it would seem perfectly appropriate to refer to this narrow region as the “selectivity filter” as regards permeability and binding selectivity. An extreme example is shown in Figure 5B in which the inner and outer vestibules of the channel are envisioned as containing sufficient water so that the effective dielectric constant is about 80 such that there is no energy barrier to entry into the vestibule from the bathing solution.

It has been argued that CFTR mutations that “disrupt lyotropic (permeability) selectivity” provide evidence for the existence of such a region that controls permeability selectivity and which can be altered by mutations.^55,130,138 For example, Linsdell et al¹³⁸ and Linsdell¹³⁰ reported that the mutations F337A and F337S, both of which reduced side chain volume, “lead to a loss of the characteristic relationship between anion permeability and hydration energies.” The slope of the log P_X/P_Cl on ΔG_hyd for these mutations is indeed flattened. A 1/r plot of the same data (Fig. 6) is not very satisfying due to the scatter of the points but it suggests the possibility that change might be accounted for by an increase in the apparent dielectric constant from about 20 for the wild type pore to between 70 and 80 for the mutants. From the polarized tunnel perspective, therefore, this result might be taken to reflect not the disruption of a filter, but from a change in protein conformation that altered the effective dielectric constant seen by the anions. In addition, one cannot eliminate the alternative that, rather than altering an existing selectivity filter, the mutations may have, in a sense, created one, perhaps by changing the location of the rate-limiting region of the pore. It is important, however to note that several of the mutations studied by Linsdell et al led to an altered single channel conductance. ^130,138 These observations, as well as those of McCarty et al,¹³⁹¹⁴⁰ support the notion that a portion of TM6 may contribute to a narrow, rate-limiting pore region that controls selectivity and conductance.

Figure 6

Peak energies plotted vs. 1/r for wild type and F337A CFTR based on the data of Linsdell et al. The dielectric constants calculated based on the slope of the lines are about 20 for the wild type pore and between 70 and 80 for F337A. Points and solid line (more...)

Linsdell and coworkers^132,141 recently studied permeation and block of CFTR by Au(CN)₂, one of the psuedohalides utilized by Smith et al.⁵¹ They uncovered an additional facet to the interaction of this relatively “hydrophobic” anion, namely that low concentrations on the cytoplasmic side markedly reduced open probability. In addition, block of the channel from the cytoplasmic side exhibited a higher affinity than that seen after exposure to extracellular Au(CN)₂. They also reported that relative permeability and block by this anion were altered by amino acid substitution in TM6.

TM6 Mutants Exhibit Altered Conduction Properties

TM6 has been identified by a number of laboratories as a potential element of the CFTR pore^{5456,98,123,138,140} It contains four basic residues (K335, R334, R347 and R352) (Fig. 1) and McDonough et al⁵³ recognized homology between this TM sequence and the sequence of the pore-lining domain of the ligand-gated chloride channels formed by the GABA_AR and GlyR.

McCarty and Zhang¹⁴⁰ systematically explored the potential role of regions of TM6 in anion permeation and binding by comparing permeability ratios and conductance ratios seen in the presence of substitute anions in the solution bathing Xenopus oocytes for a series of alanine mutants extending from position 335 to 341. They examined the behavior of normalized (to wild type CFTR) values of P_X/P_Cl and g_X/g_Cl as a function of location and compared the behavior of smaller, more permeant anions like Cl^- and SCN^- to that of larger, less permeant anions like acetate and gluconate, reasoning that selectivity between chloride and other small anions might depend upon interactions at sites distinct from those that determine selectivity between chloride and large anions. The results indicated that P_X/P_Cl for the smaller substitute anions (NO₃^-, Br^-, SCN^-, I^-, ClO₄^-) was highly sensitive to mutations to alanine at positions T338 and S341 but not to mutations at position 335, although the latter altered permeability ratios for larger anions. The pattern of the impact of alanine substitutions on gX/gCl was similar, the largest effects being at T338 and S341. To further define regional anion discrimination these authors compared relative affinities for each substitute anion calculated as (P_X/P_Cl)/(gX/gCl). This ratio, used by Halm and Frizzell¹⁴² to identify anions that bound tightly within the pore of the outward-rectifying Cl^- channel in T84 cells, will tend to single out anions like SCN^- that enter the pore more readily than Cl^- (P_SCN/P_Cl > 1.0) but bind more tightly (gSCN/gCl <1.0). Alanine substitutions at K335, T339 and T1134 had no effect on relative affinity, while alanine substitutions at T338 and S341 reduced the ability of CFTR channels to discriminate between anions like Cl^- and SCN^-. These observations are consonant with the findings of Linsdell et al.⁵⁴ and suggest the possibility that the CFTR pore may begin to narrow in a region near T338. Glutamic acid substitutions at 338 and 341 disrupted anion binding as if these residues may lie in the vicinity of, or in some way contribute to, the structure of an anion binding site.¹⁴⁰

Dawson et al⁵⁷ reviewed some early studies that seemed to implicate arginines in the binding of either chloride or thiocyanate by proteins. Arginine 347 had been proposed by Tabcharani et al⁵² to comprise part of such a binding site in CFTR because the anomalous mole fraction effect seen with symmetric SCN^- was absent in R347 mutants. The hydrogen bonding capacity of the guanidinium side chain might favor such a role, but Cotten and Welsh¹⁴³ suggested that this residue might, instead, function to stabilize CFTR structure by forming a salt bridge with D924 in TM8. A cysteine substituted for R347 was not accessible to polar thiol reagents,⁵⁸ however, so we conclude that the nature and location of an anion coordination site in CFTR remains largely a matter of speculation at present.

Several recent studies have suggested that the pore of CFTR may not be conformationally static in that permeation properties are dependent upon gating status. Ishihara and Welsh¹⁴⁴ studied block of the CFTR pore by the pH buffer MOPS. CFTR channels in excised patches with cytoplamic MOPS showed transitions to two open states: one state that was susceptible to flickery block by MOPS and a second state that was not blocked by MOPS. Disruption of ATP-dependent channel gating by exposure to a poorly hydrolyzable ATP analog (AMP-PNP) led to a predominance of the MOPS-insensitive state. Linsdell and Hanrahan¹⁴⁵ reported an asymmetric permeability of large anions (such as gluconate) that was dependent upon ATP hydrolysis, suggesting that the selectivity of the channel was altered when the ATP hydrolysis cycle was disrupted by poorly-hydrolyzable ATP analogs. Bear and colleagues¹⁴⁶ recently reported that inhibitors of permeation in CFTR, such as DPC, affect the rate of ATP hydrolysis, suggesting the possibility of communication between the conduction path and gating machinery. The region of the CFTR pore that experiences a conformational change associated with channel gating has yet to be identified, but recent studies show clearly that TM mutations can produce novel gating phenomena. Zhang et al⁴³ described a point mutation in TM11, at residue 1118, which resulted in the appearance of voltage-jump relaxations. Macroscopic currents showed that the rate of this relaxation in the S1118F mutant was dependent upon the character of the permeating ion: relaxations were slowed in SCN^- compared to those measured when Cl^- was the predominant anion. S1118F also exhibited alterations in mean burst duration and unitary conductance. Finally, selectivity patterns at the beginning of the voltage-jump relaxation differed slightly from selectivity patterns at the end of the relaxation. Similar results at V317 in TM6 also suggest that the conformation of the pore changes during gating.¹⁴⁷

Anion Binding in Model Systems and Other Proteins

Crystallographic results for a number of proteins now provide a variety of examples of anion coordination sites, each comprising an array of full or partial charges. Noteworthy among these is the recently derived structure for the ClC chloride channel,^20a but anion binding sites are found in a variety of other proteins including gecko cone type visual pigment,¹⁴⁸ haloalkane dehydrogenase,¹⁴⁹ adenosine kinase,¹⁵⁰ α-amylase,¹⁵¹¹⁵³ yellow fluorescent protein¹⁵⁴¹⁵⁵ and halorhodopsin.⁷⁶ The latter protein is of particular interest because of the variety of ligands that comprise the chloride binding site and the fact that model calculations provided estimates of the degree to which each ligand contributes to the anion-protein interaction energy, which is more negative than ΔG_hyd by about 1.3 kcal/mole (-2.2 RT). This may be compared to the estimate of -3 to -3.2 RT for the depth of a single well in CFTR derived from a K_1/2 of between 45 and 50 mM for the saturation of chloride conductance with increasing chloride concentration.⁹³

In halorhodopsin the hydration energy of the chloride ion (86.7 kcal/mol or 147 RT) is partially compensated by “background solvation” (i.e., dielectric stabilization excluding solvation due to full and partial charges and buried water) of about 66.4 kcal/mole (112.5 RT) that is attributable to a background dielectric constant of about 4. It is noteworthy that this energy accounts for 77% of the total solvation energy of the bound chloride ion. Compensation of the remaining 20.3 kcal/mol (34 RT) and the ΔG_well (-1.3 kcal/mol, 2.2 RT) is the result of stabilization by the protein microenvironment as follows: in the chloride binding site the guanidinium group of a nearby arginine (paired with a glutamic acid) acting through water molecules stabilizes chloride by -4.6 kcal/mole. A protonated Schiff base (retinal) paired with an aspartic acid contributes -2.4 kcal/mole and dipole interactions with two bound waters and a serine hydroxyl account for -8.9 kcal/mole. The remaining stabilization was attributed to dipole interactions with 10 aliphatic hydrogen atoms (-4.5 kcal/mole, 7.6 RT) and surface charges that create a favorable entry pathway for anions (-1.3 kcal/mole).⁷⁶

Other chloride-binding proteins provide examples of the variety that is seen in chloride coordination. The yellow mealworm β-amylase utilizes one water molecule, two arginines and one asparagine.¹⁵² The chloride binding site of the bacterial homologue differs in that one arginine is substituted by lysine.¹⁵¹ Human pancreatic β-amylase coordinates Cl^- by means of two Arg and one Asp residues and the K_1/2 for Cl^-dependent amylase activity was 0.53 mM, consistent with a well depth of about -4.5 kcal/mol (-7.5 RT). In Toxoplasma gondii adenosine kinase a bound chloride is stabilized by the positive dipole potential due to an alpha helix and contacts with amide nitrogens and a threonine oxygen.¹⁵⁰ Hol¹⁵⁶¹⁵⁷ has reviewed the role of helical dipoles in anion binding.

The iodide-containing cavity in the yellow fluorescent protein is amphiphilic, containing on one side an arginine and polar tyrosine and glutamine and on the other side non polar valines, leucines, isoleucines and phenylalanine.¹⁵⁵ Perhaps the simplest examples of anion “binding sites” are anion inclusion compounds, like the katapinates, that can coordinate anions in aqueous solution.¹⁵⁸ One of these, a diprotonated, in, in-1,11-diazabicyclo[9.9.9]nonacosanebisammonium can host a Cl^- that is stabilized by two hydrogen bonds with amino nitrogen.⁸⁸ The stability constant for this compound is consistent with a well depth of -2.7 kcal/mole (-4.6 RT).

From the perspective of the foregoing discussion the chloride binding site of the ClC chloride channel is particularly interesting in that the single, bound anion is stabilized by the combined influence of helical dipoles and contacts with main chain and side chain nitrogen and oxygen atoms.²⁰ The amino ends of three alpha helices are pointed toward the binding site and other interactions occur with, for example, hydroxyl groups of serine and tyrosine. The authors argue that this cavity, because it is formed by partial charges, creates a lower affinity site that is more compatible with conduction than would be the case if the bound anion were to be near a full positive charge. Indeed, the apparent affinity of the homologous Torpedo ClC-0 channel (K_1/2=75 mM) estimated from changes in conductance with changes in concentration with a correction for surface potential is consistent with a single well depth of about -1.5 kcal/mol (2.5 RT).¹⁵⁹ We also note that a structure has been solved for a soluble form of the CLIC1 chloride channel, but neither the pore nor a bound chloride ion was evident in the monomer.¹⁶⁰

Covalent Labeling

One approach to a more concrete definition of CFTR pore structure is to use covalent labeling of engineered cysteines by highly polar thiol reagents to identify possible pore-lining residues or TMs. The logic of this approach is based on the notion that if the labeling reagents are highly polar, preferably permanently charged, and the reaction can only occur with the ionized, thiolate form of the sulfhydryl moiety,¹⁶¹ then a covalent labeling reaction will be highly probable only in an aqueous environment. In other words, if engineered cysteines can be shown to react with polar thiol reagents applied to the outside of the cell, such reactions will define the outward-facing, water-accessible surface of the protein, some portion of which seems likely to include at least the mouth of the pore where anions transition from bulk solution into the channel. Sorting out which of these residues actually reside “in the pore” is another story and requires the application of some criterion for deciding if modification of an accessible residue actually alters anion conduction and, if so, by what mechanism.

The “cysteine scanning” approach, pioneered by Karlin and colleagues,¹⁶² was first applied to CFTR by Akabas and co-workers¹⁶³¹⁶⁸ and later by Smith et al.⁵⁸ Such studies revealed that, as assayed from functional changes alone, none of the 18 cysteines that inhabit wild type human CFTR are accessible from the outside. It is to be emphasized, however, that cysteines that are reactive, but which reside in locations where the reaction is without a functional consequence, are invisible to this technique. With this caveat in mind it is also noteworthy that endogenous cysteines can be induced to become accessible (or functionally significant) in mutant CFTRs — yet another reminder of the difficulty in defining the extent of structural perturbations resulting from even a single amino acid substitution.

Sites in TMs 1,2,3 and 6 have been reported to be accessible to externally applied thiol reagents, but two issues mitigate against a straightforward interpretation of all of these results in terms of pore location.¹⁶³¹⁶⁸ First, in most of the cited studies no criterion was applied that would permit changes in anion conductance to be distinguished from changes in channel gating. Second, similar experiments conducted in different laboratories did not produce consonant results.⁵⁸ The discussion here will be confined to those results for which an effect of covalent modification on conduction properties seems reasonably well established.

Smith et al⁵⁸ explored the thiol accessibility of cysteines engineered into TM6, a segment implicated in several studies of CFTR anion conduction.⁵⁴⁵⁶ They identified two residues, R334 and K335, which could lie in or near the outer vestibule of the anion-selective pore. Covalent deposition of charges at position 334 by means of methane thiolsulfonate derivatives modified CFTR conductance in charge-dependent fashion, but was without marked effect on open probability. Changes in conductance were accompanied by an alteration in the shape of the macroscopic I-V plot that could be predicted on the basis of models for the pore that included an outer vestibule where fixed positive charge promotes the local accumulation of anions. Similar changes could be achieved by titrating the partial charge on a cysteine or histidine at this position by manipulating the pH of the bathing solution.

Smith et al⁵⁸ suggested that in wild type CFTR, R334 (and to a lesser extent perhaps K335) confer upon the outer vestibule a positive electrostatic potential that promotes anion conduction. The charges do not appear to be required for anion/cation selectivity, however, although they would be expected to promote it. Comparison of covalently modified R334C CFTR with the wild type protein produced yet another example of the perils of mutagenesis: although deposition of a positive charge at position 334 (MTSET, [2-(trimethylammonium)ethyl] methanethiosulfonate bromide) increased the single-channel conductance of R334C CFTR, this modification did not rescue the wild type phenotype.

Preliminary results¹⁶⁹ showed that a cysteine substituted at position 338, one helical turn cytoplasmic to 334, was titratable via changes in bath pH and accessible to externally applied thiols. Anion conduction was highly dependent on the charge at this locus, consistent with the working hypothesis that portions of TM6 may “line the pore.” The apparent pKa for a cysteine at 338 was about 1 pH unit more acidic than expected for a model compound in free solution, as expected if the stability of the thiolate is increased by the presence of a nearby positive charge, like R334. Taken together these results provide evidence that the outward-facing surface of the CFTR pore contains a vestibule that is characterized by a positive electrostatic potential. Such a region has been inferred in the structure of the bacterial ClC channel.²⁰

How Many Pores and How Many Peptides?

Although it is well established that CFTR functions as a chloride channel, neither the number of CFTR peptides required, nor the number of pores per functional channel are known. The crystal structure of the bacterial ClC channel²⁰ confirmed conclusions based on earlier functional,¹⁷⁰¹⁷² biochemical¹⁷³ and X-ray¹⁷⁴ studies that the functional channel is a homodimer and that each of the two monomers contributes a separate pore. This feature contrasts with the bacterial K channel, a homotetramer in which each of four monomers contributes to a single, central pore formed at the interface. Several approaches have been used to address these questions for CFTR, but the results have not produced a consistent picture. Three alternatives have been proposed:

1. one-polypeptide/one-pore,
2. two-polypeptides/one-pore, and
3. one-polypeptide/two-pores.

Marshall et al¹⁷⁵ assayed the oligomeric state of CFTR expressed in C127 cells by comparing the ability of antibodies to immunoprecipitate co-expressed CFTR variants of differing molecular weight. In a series of experiments utilizing different detergents, different antibodies and four CFTR variants, as well as a protein cross-linking reagent, only monomeric CFTR was resolved on polyacrylamide gels. Thus these authors, despite earlier observations of possible multimeric CFTR complexes recovered from solubilized C127 or CHO cells,¹⁷⁶ concluded that the protein is expressed predominantly as a monomer. Bear and coworkers¹⁷⁷ recently studied the quaternary structure of purified, reconstituted CFTR, using chemical cross-linking and polyacrylamide gel electrophoresis conducted under conditions described by the authors as being “non dissociative.” The results suggested that CFTR could exist in monomeric or dimeric forms in both Sf9 and CHO cells, and that reconstitution of monomeric CFTR in phospholipid liposomes led to the appearance of dimers. When equal amounts of monomeric and dimeric proteins, separated by gel filtration, were reconstituted into phospholipid liposomes, no differences were noted in ³⁶ Cl^- transport rate or ATPase activity, suggesting that function was not dependent on higher order structure. Reconstitution of monomeric CFTR into planar lipid bilayers resulted in functional channels that retained characteristics of the native channel, and gated to a single open level. Reconstitution of dimeric CFTRs resulted in the appearance of identical channels that gated between one closed and two open levels. These results led the authors to infer that the monomer is the minimum functional unit for CFTR channel activity. It is important to note that each of these studies relied upon observations of recombinant CFTR heterologously expressed at high density in cells that do not express native CFTR. In addition, the function of the isolated protein and its appearance on gels may have been influenced by the detergents used in such studies in ways that are not evident at this time. Finally, there is some possibility that reconstituted monomers formed dimers spontaneously upon introduction to the bilayer. In a recent study Chen et al¹⁷⁸ reported the results of biochemical characterization and channel recordings in bilayers and could find no evidence for “hybrid” channels when wild type CFTR was coexpressed with several mutatants.

Eskandari et al¹⁷⁹ attempted to estimate the apparent size of several membrane proteins expressed in Xenopus oocytes by analyzing freeze-fracture images and comparing the results to those obtained using proteins of known structure. They concluded that the apparent size of CFTR was consistent with a protein comprising 25 +/- 2 alpha helices, i.e., a dimer. It remains to be determined, however, to what extent this sort of comparison is affected by differences in the folding or packing of CFTR and the reference proteins, aquaporin 1 and opsin.

Ma and coworkers¹⁸⁰¹⁸¹ expressed concatemers of wild type and R-domain-deleted CFTR (CFTRΔR) in HEK-293 cells and these hybrid channels opened to single levels, displaying gating properties between those of WT-CFTR and CFTRΔR. These data were interpreted as showing the formation of a single pore that required the dimerization of two CFTR molecules. However, the possibility remains that the activity of one CFTR molecule may have been modulated by the tether. Indeed, this is consistent with more recent data demonstrating the importance of the N-terminal tail of CFTR in regulating channel gating.¹⁸²¹⁸⁴

Several recent studies reported that CFTR activity can be modulated by association with scaffolding proteins containing bivalent or multivalent PDZ domains, such as CAP70¹⁸⁵ and NHERF,¹⁸⁶ and it has been proposed that PDZ-linked, dimeric forms of CFTR can exhibit an elevated open probability. When detached patches from Calu-3 cells expressing endogenous CFTR were exposed to PDZ-domain-containing peptides (expressed as bacterial fusion proteins), PDZ-mediated interaction of CFTR peptides appeared to result in an increase in the activity (open probability) of a single channel, rather than an increase in the number of open levels. Thus, if PDZ linkage does produce dimerization, we must conclude that both monomers and dimers form channels with the virtually identical conduction properties, as if regardless of the oligomeric state of the protein only a single pore is conducting anions. For this reason, Raghuram et al¹⁸⁶ interpreted their results as suggesting that the functional unit of CFTR is the dimer and intramolecular cross linking via PDZ-NHERF enhances open probability. Similar results were obtained using a bivalent antibody that recognizes the TRL domain in the carboxy terminus of CFTR. It is not clear, however, that it is possible to eliminate the alternative hypothesis that the linker molecules achieve their functional effect by connecting CFTR to some, as yet unidentified, accessory protein.

A complementary approach to defining the functional unit of CFTR is to search for channel activity in CFTR fragments. Unfortunately the most common experiment, the heterologous expression of CFTR fragments coupled with an assay for chloride conductance and/or channel activity, is highly compromised by the difficulty in assigning unambiguously the observed channel activity to the expressed fragments. Nevertheless, some provocative findings have been reported that would lead to the speculation that a chloride-selective channel can be formed by a less than full length portion of the CFTR protein. Sheppard et al⁶⁰ reported activity in an amino terminal fragment truncated after the R domain (D836X) that also exhibited severely reduced expression that might account for the inability of Ostedgaard et al¹⁸⁷ to detect SPQ fluorescence changes in cells transfected with the same construct. The reports of Devidas et al¹⁸⁸ and Yue et al⁶² suggested that CFTR-like activity could be detected in cells expressing either N-terminal or C-terminal portions of CFTR as if the parent molecule might contain two pores. Chan et al⁶¹ carried out a systematic study of the ability of an array of severed CFTRs to give rise to chloride conductance in Xenopus oocytes. They reported that none of the single fragments gave rise to cAMP-induced macroscopic currents, but that certain pairs, when co-expressed, gave rise to activity much like that of wild type CFTR. They proposed that these cut sites define domain boundaries such that “whole” channels can form from fragments coexpressed in the same cell. The issue of whether a partial CFTR can function as a channel is unresolved at present, but is an important area for future investigation.

CFTR is a member of the large, ABC (ATP Binding Cassette) Transporter superfamily of proteins and the oligomeric structure of other ABC Transporters may provide clues to the structure of CFTR. For example, Rosenberg et al¹⁸⁹ reported that lectin-gold labeling of the single glycosylation site in P-glycoprotein resulted in particle size that was consistent with the monomeric form. Loo and Clarke¹⁹⁰¹⁹¹ used site-directed mutagenesis to study the substrate-binding pocket of the same protein and interpreted results as indicating that TM domains in both the front and back halves of the full-length P-glycoprotein peptide contribute to the substrate-binding pocket, and the presence of drug substrate promote their interaction. CFTR and P-glycoprotein are both full-length “ABC Transporters”, comprising two MSDs and two NBDs in one peptide. In contrast, prokaryotic ABC Transporters are more commonly constructed of up to four separate peptides, each comprising, for example, one MSD or one NBD. In the histidine permease, dimerization of HisP, the NBD subunit, is required for function,²⁹¹⁹². This is consistent with the notion that the interaction between the two NBDs of CFTR might be required for function. The crystal structure of MsbA, the transporter of E. coli, was recently solved to 4.5Å by Chang and Roth¹⁶ MsbA is a half-transporter, encoding one MSD and one NBD per peptide. The crystal structure suggests that two half-transporters dimerize to form the functional protein, with a single substrate-binding pocket formed from TM helices from each of the two MSDs. The recently solved structure of E. coli BtuCD, a vitamin B12 transporter, is also consistent with a body plan in which the functional transporter comprises two NBDs and two MSDs.

Channel Blockers Probe Pore Structure

A blocker that enters the pore can, in principle, be used to define a binding site or receptor within the pore. Although a wide variety of organic compounds have been shown to inhibit CFTR only a few have been shown to inhibit CFTR via open-channel block (see ref. ¹³⁹). Two families of compounds (arylaminobenzoates and sulfonylureas) have been extensively studied as blockers of the CFTR pore. The arylaminobenzoates, whose parent compound is diphenylamine- 2-carboxylate (DPC), block CFTR via a simple bimolecular interaction with a site accessible from the cytoplasmic end of the pore.^139,193194 DPC and its congener flufenamic acid (FFA) were shown by McCarty and coworkers to block single CFTR channels with simple kinetics, i.e., application of drug to the cytoplasmic media introduced a single class of closed states into recordings from excised patches.¹⁹⁵ Subsequent studies indicated that a dose-efficacy relationship for DPC block of CFTR macroscopic currents was fit best with a Hill coefficient near unity.¹⁹⁴ Both of these observations suggest that DPC interacts with a single site. Blockade of both macroscopic currents and single-channel currents are sensitive to membrane voltage and to the concentration of permeant anion, consistent with the notion that DPC blocks by entering the pore.^53,196 DPC (as well as related NPPB, 5-nitro-2-(3-phenylpropylamino)- benzoate) blocks CFTR by the classical open-channel block mechanism as evidenced by a drug-induced increase in burst duration¹⁹⁴ (see ref. ¹⁹⁷). The lengthening of the burst duration can be interpreted as an influence of the blocker on the conformational change required for channel closing. Burst durations are increased in direct proportion to the frequency and duration of blocking events. DPC and NPPB exhibit identical voltage-dependence of block of macroscopic currents, suggesting that the two compounds bind at approximately the same position in the pore, accessed by traversing approximately 40% of the voltage field across the membrane, as measured from the cytoplasmic side.¹⁹⁴ Modifications to the structure of NPPB that resulted in increased hydrophobicity increased the potency of block without changing the voltage-dependence.¹⁹⁶

The sulfonylureas are hypoglycemic agents that interact with the sulfonylurea receptor (SUR, another member of the ABC Transporter superfamily¹⁹⁸) and are used clinically to control the release of insulin from pancreatic beta cells. The most potent sulfonylurea for this purpose is glibenclamide.¹⁹⁹ Inhibition of CFTR currents by glibenclamide was first described by Sheppard and Welsh.²⁰⁰ Glibenclamide and its congener, tolbutamide, blocked CFTR whole-cell currents in mammalian cell lines with no apparent voltage-dependence. Subsequent studies attempted to identify the mechanism of interaction by utilizing excised inside-out patches to show that glibenclamide binds to the cytoplasmic side of CFTR and reduces Po in a voltage and concentration-dependent manner.²⁰¹²⁰² Other sulfonylurea compounds, and some nonsulfonylurea hypoglycemic agents, also inhibit CFTR,²⁰³²⁰⁵ and glibenclamide also interacts with several other ABC transporters (see ref. ¹³⁹).

McCarty et al attempted to identify amino acids which, when mutated, altered the interaction of arylaminobenzoate blockers with the pore. Several mutations were identified in TM6 and TM12 that affected affinity and/or voltage dependence of blockade by DPC,⁵³ whereas several other mutations were found that affected neither affinity nor voltage dependence. Both the affinity and voltage-dependence of block were altered dramatically by mutation S341A in TM6. Mutations in TM12 also altered block by DPC,⁵³ as did one mutation in TM11.⁴³ Block of whole-cell currents by NPPB was impacted by the S341A and T1134F mutations in a manner somewhat different from the effects of these mutations on block by DPC.¹⁹⁴ These data indicate that although NPPB and DPC bind at similar positions, their interactions with the pore interior are not identical. Walsh and coworkers have also shown that mutations in TM6 affect block by other arylaminobenzoates.¹⁹⁶

In preliminary experiments, Zhang et al²⁰⁶ tested the effects on glibenclamide block of two mutations which alter block by DPC and NPPB. Overall, the block of CFTR macroscopic currents by glibenclamide was altered in S341A-CFTR and T1134F-CFTR, although not in the same manner or to the same degree as was block by DPC or NPPB. By analyzing macroscopic currents in the presence of glibenclamide, it appeared that mutation S341A more strongly affected binding of glibenclamide to the site with fast microscopic kinetics, while mutation T1134F more strongly affected binding to the site with slow microscopic kinetics. According to these data, the glibenclamide-binding site or sites in the pore of CFTR appear to lie in close proximity to the sites that contribute to block by the arylaminobenzoates. In contrast, Gupta and Linsdell²⁰⁷ recently showed that mutations T338A and F337A had small but significant effects on weakening glibenclamide block. By comparison, the effects of these mutations on glibenclamide block were much smaller than the effects of mutation S341A on block by DPC and NPPB.^53,194 Furthermore, block by DPC, a drug much smaller than glibenclamide, was unaffected by mutation T338A.⁵³ Block of CFTR macroscopic currents by glibenclamide shows considerably weaker voltage-dependence than does that by DPC or NPPB, suggesting that the site(s) of glibenclamide block might lie somewhere closer to the cytoplasmic end of the pore. The general trend is for small blockers of CFTR to exhibit stronger voltage-dependence of blockade than do large blockers, as if the former are able to reach further into the pore (see however, Linsdell).²⁰⁸ Perhaps the most important generalization emerging from blocker studies is that molecules like DPC, glibenclamide, DIDS and gluconate all appear to reach their binding sites from the cytoplamic side.²⁰⁹²¹⁰ The size of these compounds is large when compared to permeant anions, suggesting that the pore may feature a relatively large cytoplasmic vestibule, comparable to that of the K channel where TEA derivatives bind.²¹¹²¹⁴ This basic organizational plan is seen in the crystal structure of the CFTR homologue, MsbA¹⁶ that shows a large cytoplasmic opening formed by the tilt of the transmembrane domains. Zhou et al²¹⁵ recently studied the block of CFTR using detached patches and a CFTR variant, K1250A CFTR, characterized by long open times. The binding characteristics of two organic blockers, glibenclamide and isethionate, particularly, the observation that they block only from the intracellular side of the pore, supports the hypothesis that the CFTR pore has a wide internal vestibule.

A Working Model of Pore Structure

Inner Vestibule

Envisioning the pore of CFTR is a risky business at this point in time in that we lack even the most rudimentary information as to the three dimensional organization of the protein. Smith et al⁵¹ suggested that lyotropic selectivity of CFTR was consistent with a “primitive” structure lacking the sort of structural specialization seen in the KcsA crystal structure, perhaps a structure that betrays the membership of CFTR in a family of transporters, many of which appear to have been designed to carry out tasks much different than anion conduction.¹⁶¹⁹ The selectivity of CFTR, as reported by variation in anion permeability and anion binding, points to at least one narrow region in which visiting anions are transiently coordinated by elements of the peptide backbone or side chains. The action of blockers and a limited amount of covalent labeling results is consistent with the existence of inner and outer vestibules separated by the narrow region, the outer vestibule containing positive charges that serve to enhance conductance by raising local anion concentration, and the inner vestibule large enough to accommodate blocking compounds such as DPC, glibenclamide and gluconate (Fig. 7). As regards specific structural components of the pore, the sheer weight of evidence certainly favors a role for TM6, but the role for other TMs is more speculative.

Figure 7

Cartoon illustration of elements of a structure that might pertain to the CFTR pore. The wall of the pore is assumed to be formed by amino acids contributed by several TM domains. A speculative “location” for TM6 is shown as the rectangle (more...)

Is CFTR an Archetypal “Primitive Pore”?

Smith et al⁵¹ speculated that the lyotropic, bias-based selectivity of CFTR suggested that it was a “primitive” structure that, for one reason or another, had not evolved a mechanism for the highly selective recognition of Cl^- ions. Selectivity for Cl^- is indeed somewhat optimized in CFTR, but in a roundabout way. Smaller anions, like F^-, have difficulty entering the pore due to their relatively high energy of hydration. Large molecules like SCN^- actually appear to enter the pore more readily than Cl^- (due to their lower ΔG_hyd) but are retarded in their passage by a deeper energy well in the conduction path. The end result is a channel that is, in fact, only moderately selective, permitting ions like Br^- and NO₃^- to transit with ease, but which seems nevertheless to serve its function admirably. Perhaps it is the relatively low abundance of these competing ions that reduced the pressure for the evolution of a more selective structure.

Acknowledgements

Research from the authors laboratories described in the text and the writing of this review was supported by grants from the National Institute for Diabetes, Digestive and Kidney Diseases DK 45880 (David C. Dawson), DK 56481 (Nael A. McCarty), and DK 60312 (Xuehong Liu); the Cystic Fibrosis Foundation, DAWSON0210 and the American Heart Association Established Investigator Program, 0140174N (Nael A. McCarty). The authors are grateful to the reviewers and the editor for valuable comments on the text.

References

1.

Riordan JR, Rommens JM, Kerem B. et al. Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science. 1989;245(4922):1066–73. [PubMed: 2475911]

2.

Rommens JM, Iannuzzi MC, Kerem B. et al. Identification of the cystic fibrosis gene: chromosome walking and jumping. Science. 1989;245(4922):1059–65. [PubMed: 2772657]

3.

Kerem B, Rommens JM, Buchanan JA. et al. Identification of the cystic fibrosis gene: genetic analysis. Science. 1989;245(4922):1073–80. [PubMed: 2570460]

4.

Egan M, Flotte T, Afione S. et al. Defective regulation of outwardly rectifying Cl^- channels by protein kinase A corrected by insertion of CFTR. Nature. 1992;358(6387):581–4. [PubMed: 1380129]

5.

Bradbury NA, Jilling T, Berta G. et al. Regulation of plasma membrane recycling by CFTR. Science. 1992;256(5056):530–2. [PubMed: 1373908]

6.

Gabriel SE, Clarke LL, Boucher RC. et al. CFTR and outward rectifying chloride channels are distinct proteins with a regulatory relationship. Nature. 1993;363(6426):263–8. [PubMed: 7683773]

7.

Stutts MJ, Canessa CM, Olsen JC. et al. CFTR as a cAMP-dependent regulator of sodium channels. Science. 1995;269(5225):847–50. [PubMed: 7543698]

8.

Hanaoka K, Devuyst O, Schwiebert EM. et al. A role for CFTR in human autosomal dominant polycystic kidney disease. Am J Physiol. 1996;270(1 Pt 1):C389–99. [PubMed: 8772467]

9.

Barriere H, Poujeol C, Tauc M. et al. CFTR modulates programmed cell death by decreasing intracellular pH in Chinese hamster lung fibroblasts. Am J Physiol Cell Physiol. 2001;281(3):C810–24. [PubMed: 11502558]

10.

Boujaoude LC, Bradshaw-Wilder C, Mao C. et al. Cystic fibrosis transmembrane regulator regulates uptake of sphingoid base phosphates and lysophosphatidic acid: modulation of cellular activity of sphingosine 1-phosphate. J Biol Chem. 2001;276(38):35258–64. [PubMed: 11443135]

11.

Drumm ML, Wilkinson DJ, Smit LS. et al. Chloride conductance expressed by delta F508 and other mutant CFTRs in Xenopus oocytes. Science. 1991;254(5039):1797–9. [PubMed: 1722350]

12.

Sheppard DN, Rich DP, Ostedgaard LS. et al. Mutations in CFTR associated with mild-disease-form Cl^- channels with altered pore properties. Nature. 1993;362(6416):160–4. [PubMed: 7680769]

13.

Hubert D, Bienvenu T, Desmazes-Dufeu N. et al. Genotype-phenotype relationships in a cohort of adult cystic fibrosis patients. Eur Respir J. 1996;9(11):2207–14. [PubMed: 8947061]

14.

Schibler A, Bolt I, Gallati S. et al. High morbidity and mortality in cystic fibrosis patients compound heterozygous for 3905insT and deltaF508. Eur Respir J. 2001;17(6):1181–6. [PubMed: 11491162]

15.

Hyde SC, Emsley P, Hartshorn MJ. et al. Structural model of ATP-binding proteins associated with cystic fibrosis, multidrug resistance and bacterial transport. Nature. 1990;346(6282):362–5. [PubMed: 1973824]

16.

Chang G, Roth CB. Structure of MsbA from E. coli: a homolog of the multidrug resistance ATP binding cassette (ABC) transporters. Science. 2001;293(5536):1793–800. [PubMed: 11546864]

17.

Locher KP, Lee AT, Rees DC. The E.coli BtuCD structure: a framework for ABC transporter architecture and mechanism. Science. 2002;296(5570):1091–8. [PubMed: 12004122]

18.

Lee JY, Urbatsch IL, Senior AE. et al. Projection Structure of P-glycoprotein by Electron Microscopy. Evidence for a closed conformation of the nucleotide binding domains. J Biol Chem. 2002;277(42):40125–31. [PubMed: 12163504]

19.

Murakami S, Nakashima R, Yamashita E. et al. Crystal structure of bacterial multidrug efflux transporter AcrB. Nature. 2002;419(6907):587–593. [PubMed: 12374972]

20.

Dutzler R, Campbell EB, Cadene M. et al. X-ray structure of a ClC chloride channel at 3.0 A reveals the molecular basis of anion selectivity. Nature. 2002;415(6869):287–94. [PubMed: 11796999]

20a.

Dutzler R, Campbell EB, Mackinnon Gating the selectivity filter in ClC chloride channels. Science. 2003;300(5616):108–12. [PubMed: 12649487]

21.

Anderson MP, Berger HA, Rich DP. et al. Nucleoside triphosphates are required to open the CFTR chloride channel. Cell. 1991;67(4):775–84. [PubMed: 1718606]

22.

Rich DP, Gregory RJ, Anderson MP. et al. Effect of deleting the R domain on CFTR-generated chloride channels. Science. 1991;253(5016):205–7. [PubMed: 1712985]

23.

Hwang TC, Nagel G, Nairn AC. et al. Regulation of the gating of cystic fibrosis transmembrane conductance regulator C1 channels by phosphorylation and ATP hydrolysis. Proc Natl Acad Sci U S A. 1994;91(11):4698–702. [PMC free article: PMC43855] [PubMed: 7515176]

24.

Winter MC, Welsh MJ. Stimulation of CFTR activity by its phosphorylated R domain. Nature. 1997;389(6648):294–6. [PubMed: 9305845]

25.

Senior AE, Gadsby DC. ATP hydrolysis cycles and mechanism in P-glycoprotein and CFTR. Seminars in Cancer Biol. 1997;8(3):143–50.

26.

Wilkinson DJ, Strong TV, Mansoura MK. et al. CFTR activation: additive effects of stimulatory and inhibitory phosphorylation sites in the R domain. Am J Physiol. 1997;273(1 Pt 1):L127–33. [PubMed: 9252549]

27.

Csanady LK, Chan W, Seto-Young D. et al. Severed channels probe regulation of gating of cystic fibrosis transmembrane conductance regulator by its cytoplasmic domains. J Gen Physiol. 2000;116(3):477–500. [PMC free article: PMC2233695] [PubMed: 10962022]

28.

Dousmanis AG, Nairn AC, Gadsby DC. Distinct Mg(2+)-dependent Steps Rate Limit Opening and Closing of a Single CFTR Cl(-) Channel. J Gen Physiol. 2002;119(6):545–559. [PMC free article: PMC2233863] [PubMed: 12034762]

29.

Hung LW, Wang IX, Nikaido K. et al. Crystal structure of the ATP-binding subunit of an ABC transporter. Nature. 1998;396(6712):703–7. [PubMed: 9872322]

30.

Yuan YR, Blecker S, Martsinkevich O. et al. The crystal structure of the MJ0796 ATP-binding cassette. Implications for the structural consequences of ATP hydrolysis in the active site of an ABC transporter. J Biol Chem. 2001;276(34):32313–21. [PubMed: 11402022]

31.

Anderson MP, Welsh MJ. Regulation by ATP and ADP of CFTR chloride channels that contain mutant nucleotide-binding domains. Science. 1992;257(5077):1701–4. [PubMed: 1382316]

32.

Baukrowitz T, Hwang TC, Nairn AC. et al. Coupling of CFTR Cl^- channel gating to an ATP hydrolysis cycle. Neuron. 1994;12(3):473–82. [PubMed: 7512348]

33.

Winter MC, Sheppard DN, Carson MR. et al. Effect of ATP concentration on CFTR Cl^- channels: a kinetic analysis of channel regulation. Biophys J. 1994;66(5):1398–403. [PMC free article: PMC1275860] [PubMed: 7520292]

34.

Carson MR, Travis SM, Welsh MJ. The two nucleotide-binding domains of cystic fibrosis transmembrane conductance regulator (CFTR) have distinct functions in controlling channel activity. J Biol Chem. 1995;270(4):1711–7. [PubMed: 7530246]

35.

Gunderson KL, Kopito RR. Conformational states of CFTR associated with channel gating: the role ATP binding and hydrolysis. Cell. 1995;82(2):231–9. [PubMed: 7543023]

36.

Zeltwanger S, Wang F, Wang GT. et al. Gating of cystic fibrosis transmembrane conductance regulator chloride channels by adenosine triphosphate hydrolysis. Quantitative analysis of a cyclic gating scheme. J Gen Physiol. 1999;113(4):541–54. [PMC free article: PMC2217165] [PubMed: 10102935]

37.

Rich DP, Gregory RJ, Cheng SH. et al. Effect of deletion mutations on the function of CFTR chloride channels. Receptors Channels. 1993;1(3):221–32. [PubMed: 7522901]

38.

Baldursson O, Ostedgaard LS, Rokhlina T. et al. Cystic fibrosis transmembrane conductance regulator Cl^- channels with R domain deletions and translocations show phosphorylation-dependent and -independent activity. J Biol Chem. 2001;276(3):1904–10. [PubMed: 11038358]

39.

Gadsby DC, Nairn AC. Control of CFTR channel gating by phosphorylation and nucleotide hydrolysis. Physiol Rev. 1999;79(1 Suppl):S77–S107. [PubMed: 9922377]

40.

Nagel G. Differential function of the two nucleotide binding domains on cystic fibrosis transmembrane conductance regulator. Biochim Biophys Acta. 1999;1461(2):263–74. [PubMed: 10581360]

41.

Kirk KL. New paradigms of CFTR chloride channel regulation. Cell Mol Life Sci. 2000;57(4):623–34. [PubMed: 11130462]

42.

Zou X, Hwang TC. ATP hydrolysis-coupled gating of CFTR chloride channels: structure and function. Biochemistry. 2001;40(19):5579–86. [PubMed: 11341822]

43.

Zhang ZR, McDonough SI, McCarty NA. Interaction between permeation and gating in a putative pore domain mutant in the cystic fibrosis transmembrane conductance regulator. Biophys J. 2000;79(1):298–313. [PMC free article: PMC1300934] [PubMed: 10866956]

44.

Mullins LJ. An analysis of conductance changes in squid axon. J Gen Physiol. 1959;42(5):1013–1035. [PMC free article: PMC2194941] [PubMed: 13654748]

45.

Hodgkin AL, Huxley AF. A quantitative description of membrane current and its application to conduction and excitation in nerve. J Physiol. 1952;117:500–544. [PMC free article: PMC1392413] [PubMed: 12991237]

46.

Hille B. Ionic channels of excitable membranes. 2nd edSunderland MA: Sinauer Associates Inc.2001 . [PMC free article: PMC369639]

47.

Mueller P, Rudin DO. Development of K-Na+ discrimination in experimental bimolecular lipid membranes by macrocyclic antibiotics. Biochem Biophys Res Commun. 1967;26(4):398–404. [PubMed: 6033739]

48.

Eigen M, Winkler R. Alkali-ion Carriers: Dynamics and SelectivityIn: Schmitt FD, ed. The Neurosciences second study programNew York: Rockerfeller Press.1970685–696.

49.

Bezanilla F, Armstrong CM. Negative conductance caused by entry of sodium and cesium ions into the potassium channels of squid axons. J Gen Physiol. 1972;60(5):588–608. [PMC free article: PMC2226091] [PubMed: 4644327]

50.

Eisenman G, Horn R. Ionic selectivity revisited: the role of kinetic and equilibrium processes in ion permeation through channels. J Membr Biol. 1983;76(3):197–225. [PubMed: 6100862]

51.

Smith SS, Steinle ED, Meyerhoff ME. et al. Cystic fibrosis transmembrane conductance regulator. Physical basis for lyotropic anion selectivity patterns. J Gen Physiol. 1999;114(6):799–818. [PMC free article: PMC2230651] [PubMed: 10578016]

52.

Tabcharani JA, Rommens JM, Hou YX. et al. Multi-ion pore behavior in the CFTR chloride channel. Nature. 1993;366(6450):79–82. [PubMed: 7694154]

53.

McDonough S, Davidson N, Lester HA. et al. Novel pore-lining residues in CFTR that govern permeation and open-channel block. Neuron. 1994;13(3):623–34. [PubMed: 7522483]

54.

Linsdell P, Tabcharani JA, Rommens JM. et al. Permeability of wild-type and mutant cystic fibrosis transmembrane conductance regulator chloride channels to polyatomic anions. J Gen Physiol. 1997;110(4):355–64. [PMC free article: PMC2229373] [PubMed: 9379168]

55.

Linsdell P, Zheng SX, Hanrahan JW. Non-pore lining amino acid side chains influence anion selectivity of the human CFTR Cl^- channel expressed in mammalian cell lines. J Physiol. 1998;512(Pt 1):1–16. [PMC free article: PMC2231193] [PubMed: 9729613]

56.

Mansoura MK, Smith SS, Choi AD. et al. CFTR: Anion binding as a probe of the pore. Biophys J. 1998;74(3):1320–1332. [PMC free article: PMC1299479] [PubMed: 9512029]

57.

Dawson DC, Smith SS, Mansoura MK. CFTR: mechanism of anion conduction. Physiol Rev. 1999;79(1):S47–S75. [PubMed: 9922376]

58.

Smith SS, Liu X, Zhang ZR. et al. CFTR. Covalent and noncovalent modification suggests a role for fixed charges in anion conduction. J Gen Physiol. 2001;118(4):407–32. [PMC free article: PMC2233702] [PubMed: 11585852]

59.

Oblatt-Montal M, Reddy GL, Iwamoto T. et al. Identification of an ion channel-forming motif in the primary structure of CFTR, the cystic fibrosis chloride channel. Proc Natl Acad Sci U S A. 1994;91(4):1495–9. [PMC free article: PMC43186] [PubMed: 7509074]

60.

Sheppard DN, Ostedgaard LS, Rich DP. et al. The amino-terminal portion of CFTR forms a regulated Cl^- channel. Cell. 1994;76(6):1091–8. [PubMed: 7511062]

61.

Chan KW, Csanady L, Seto-Young D. et al. Severed molecules functionally define the boundaries of the cystic fibrosis transmembrane conductance regulator's NH(2)-terminal nucleotide binding domain. J Gen Physiol. 2000;116(2):163–80. [PMC free article: PMC2229491] [PubMed: 10919864]

62.

Yue H, Devidas S, Guggino WB. The two halves of CFTR form a dual-pore ion channel. J Biol Chem. 2000;275(14):10030–4. [PubMed: 10744680]

63.

Bockris JOM, Reddy AKN. Modern ElectrochemistryNew York: Plenum Press,1970 . [PMC free article: PMC322739]

64.

Marcus Y. A simple empirical model describing the thermodynamics of hydration of ions of widely varying charges, sizes, and shapes. Biophys Chem. 1994;51:111–127.

65.

Marcus Y. Ion propertiesNew York: Marcel Dekker,1997. [PMC free article: PMC176006]

65a.

Andersen OS, Procopio J. Ion movement through gramicidin A channels. On the importance of the aqueous diffusion resistance and ion-water interactions. Acta Physiol Scand Suppl. 1980;481:27–35. [PubMed: 6159776]

66.

Doyle DA, Cabral JM, Pfuetzner RA. et al. The structure of the potassium channel: molecular basis of K conduction and selectivity. Science. 1998;280:69–77. [PubMed: 9525859]

67.

Zhou Y, Morais-Cabral JH, Kaufman A. et al. Chemistry of ion coordination and hydration revealed by a K channel-Fab complex at 2.0 A resolution. Nature. 2001;414(6859):43–8. [PubMed: 11689936]

68.

Roux B, Yu H-A, Karplus M. Molecular basis for the Born model of ion solvation. J Phys Chem. 1990;94:4683–4688.

69.

Jayaram B, Fine R, Sharp K. et al. Free energy calculations of ion hydration: an analysis of the Born model in terms of microscopic simulations. J Phys Chem. 1989;93:4320–4327.

70.

Rashin AA, Honig B. Reevaluation of the Born model of ion hydration. J Phys Chem. 1985;89:5588–5593.

71.

Sitkoff D, Sharp KA, Honig B. Accurate calculation of hydration free energies using macroscopic solvent models. J Phys Chem. 1994;98:1978–1988.

72.

Tannor DJ, Marten B, Murphy R. et al. Accurate first principles calculation of molecular charge distributions and solvation energies from Ab Initio quantum mechnics and continuum dielectric theory. J Am Chem Soc. 1994;116:11875–11882.

73.

Florian J, Warshel A. Langevin dipoles model for ab Initio calculations of chemical processes in solution: parametrization and application to hydration free energies of neutral and ionic solutes and conformational analysis in aqueous solution. J Phys Chem B. 1997;101:5583–5595.

74.

Florian J. Calculations of hydration entropies of hydrophobic, polar, and ionic solutes in the framework of the Langevin dipoles solvation model. J Phys Chem B. 1999;103:10282–10288.

75.

Dorman V, Partenskii MB, Jordan PC. A semi-microscopic Monte Carlo study of permeation energetics in a gramicidin-like channel: the origin of cation selectivity. Biophys J. 1996;70(1):121–34. [PMC free article: PMC1224914] [PubMed: 8770192]

76.

Kolbe M, Besir H, Essen LO. Structure of the light-driven chloride pump halorhodopsin at 1.8 A resolution. Science. 2000;288(5470):1390–6. [PubMed: 10827943]

77.

Roux B, Berneche S, Im W. Ion channels, permeation, and electrostatics: insight into the function of KcsA. Biochemistry. 2000;39(44):13295–306. [PubMed: 11063565]

78.

Levitt DG. Modeling of ion channels. J Gen Physiol. 1999;113(6):789–94. [PMC free article: PMC2225601] [PubMed: 10352030]

79.

Goldman DE. Potential, impedance, and rectification in membranes. J Gen Physiol. 1943;27:37–60. [PMC free article: PMC2142582] [PubMed: 19873371]

80.

Eyring H, Lumry R, Woodbury JW. Some applications of modern rate theory to physiological systems. Rec Chem Prog. 1949;10:100–14.

81.

Dani JA. Ion-channel entrances influence permeation. Net charge, size, shape, and binding considerations. Biophys J. 1986;49(3):607–18. [PMC free article: PMC1329508] [PubMed: 2421791]

82.

Dani JA, Levitt DG. Diffusion and kinetic approaches to describe permeation in ionic channels. J Theor Biol. 1990;146(3):289–301. [PubMed: 1701840]

83.

Dawson DC. Permeability and conductance in ion channels: A primerIn: Andreoli TF, Hoffman JF, Fanestil DD et al, ed.Molecular Biology of Membrane Transport DisordersNew York: Plenum,199687–109.

84.

Diamond JM, Wright EM. Biological membranes: the physical basis of ion and nonelectrolyte selectivity. Annu Rev Physiol. 1969;31:581–646. [PubMed: 4885777]

85.

Born M. Volumen un hydrationwärme der ionen. Z Phys. 1920;1:45–48.

86.

Morf WE. The principles of ion-selective electrodes and of membrane transportNew York: Elsevier,1981 . [PMC free article: PMC281528]

87.

Yim HS, Kibbey CE, Ma SC. et al. Polymer membrane-based ion-, gas- and bio-selective potentiometric sensors. Biosens Bioelectron. 1993;8(1):1–38. [PubMed: 8499085]

88.

Bell RA, Christoph GC, Fronczek FR. et al. The cation H1306+: a short, symmetric hydrogen bond. Science. 1975;190:151–152.

89.

Morais-Cabral JH, Zhou Y, MacKinnon R. Energetic optimization of ion conduction rate by the K selectivity filter. Nature. 2001;414(6859):37–42. [PubMed: 11689935]

90.

Jiang Y, Lee A, Chen J. et al. Crystal structure and mechanism of a calcium-gated potassium channel. Nature. 2002;417(6888):515–22. [PubMed: 12037559]

91.

Mansoura MK, Dawson DC. 2 barrier-1 site pore: an interactive spreadsheet model for two permeating ions. Comp in Bio and Med. 1998;28:255–273. [PubMed: 9784963]

92.

Levitt DG. Continuum model of voltage-dependent gating. Macroscopic conductance, gating current, and single-channel behavior. Biophys J. 1989;55(3):489–98. [PMC free article: PMC1330502] [PubMed: 2467698]

93.

Linsdell P, Tabcharani JA, Hanrahan JW. Multi-Ion mechanism for ion permeation and block in the cystic fibrosis transmembrane conductance regulator chloride channel. J Gen Physiol. 1997;110(4):365–77. [PMC free article: PMC2229374] [PubMed: 9379169]

94.

Zhou Z, Hu S, Hwang TC. Voltage-dependent flickery block of an open cystic fibrosis transmembrane conductance regulator (CFTR) channel pore. J Physiol. 2001;532(Pt 2):435–48. [PMC free article: PMC2278548] [PubMed: 11306662]

95.

Cahalan M, Begenisich T. Sodium channel selectivity. Dependence on internal permeant ion concentration. J Gen Physiol. 1976;68(2):111–25. [PMC free article: PMC2228421] [PubMed: 956766]

96.

Begenisich TB, Cahalan MD. Sodium channel permeation in squid axons. I: Reversal potential experiments. J Physiol. 1980;307:217–42. [PMC free article: PMC1283042] [PubMed: 6259334]

97.

Dwyer TM, Adams DJ, Hille B. The permeability of the endplate channel to organic cations in frog muscle. J Gen Physiol. 1980;75(5):469–92. [PMC free article: PMC2215262] [PubMed: 6247422]

98.

Tabcharani JA, Linsdell P, Hanrahan JW. Halide permeation in wild-type and mutant cystic fibrosis transmembrane conductance regulator chloride channels. J Gen Physiol. 1997;110(4):341–54. [PMC free article: PMC2229372] [PubMed: 9379167]

99.

Hille B. Ionic selectivity, saturation, and block in sodium channels. A four- barrier model. J Gen Physiol. 1975;66(5):535–60. [PMC free article: PMC2226224] [PubMed: 1194886]

100.

Hille B. Ionic selectivity of Na and K channels of nerve membranesIn: Eisenman G.Membranes: lipid bilayers and biological membranes-dynamic propertiesNew York: Marcel Dekker, Inc.19753255–323. [PubMed: 1202319]

101.

Hille B, Schwarz W. Potassium channels as multi-ion single-file pores. J Gen Physiol. 1978;72(4):409–42. [PMC free article: PMC2228548] [PubMed: 722275]

102.

Eigen M, Winkler R. Carriers and specificity in membranes. II. Characteristics of carriers. Alkali ion carriers: specificity, architecture, and mechanisms: an essay. Neuros Res Prog Bull. 1971;9(3):330–8. [PubMed: 5164651]

103.

Teorell T. Transport processes and electrical phenomena in ionic membranes. Prog Biophysics. 1953;3:305–369.

104.

Eisenman G. On the elementary atomic origin of equilibrium ionic specificityIn: Kleinzeller A, Kotyk A, ed.Symposium on membrane transport and metabolismPrague:Academic Press. New York.1961163–179.

105.

Eisenman G. Cation selective glass electrodes and their mode of operation. Biophys J. 1962;2(2):259–323. [PMC free article: PMC1366487] [PubMed: 13889686]

106.

Eisenman G. Some elementary factors involved in specific ion permeation. Proc. 23rd Intern. Congr. Physiol. Sci. Tokyo, Excerpta Med. Found. Amsterdam. 1965;4:489–506.

107.

Szabo G, Eisenman G, Laprade R. et al. Experimentally observed effects of carriers on the electrical properites of bilayer membrane-equilibrium domainIn: Eisenman, G. Membranes. New York: Marcel Dekker, Inc.19732179–328. [PMC free article: PMC130000] [PubMed: 4585227]

108.

Lewis CA, Stevens CF. Acetylcholine receptor channel ionic selectivity: ions experience an aqueous environment. Proc Natl Acad Sci U S A. 1983;80(19):6110–3. [PMC free article: PMC534370] [PubMed: 6310616]

109.

Moyer BA, Bonnesen PV. Physical factors in anion separationsIn: Bianchi A, Bowman-James K, Garcia-España E, ed(s).Supramolecular chemistry of anionsNew York:Wiley-VCH.19971–44.

110.

Hofmeister F. Lehre von der wirkung der salze. Arch Exp Pathol Pharmakol. 1888;24:247–260.

111.

Armstrong RD, Horvai G. Properties of PVC based membranes used in ion-selective electrodes. Electrochimica Acta. 1990;35:1–7.

112.

Rees DC. Experimental evaluation of the effective dielectric constant of proteins. J Mol Biol. 1980;141(3):323–6. [PubMed: 6253649]

113.

Warshel A, Russell ST. Calculations of electrostatic interactions in biological systems and in solutions. Quarterly Reviews of Biophysics. 1984;17(3):283–422. [PubMed: 6098916]

114.

Pickersgill RW. A rapid method of calculating charge-charge interaction energies in proteins. Protein Eng. 1988;2(3):247–8. [PubMed: 3237687]

115.

Fersht AR, Sternberg MJ. Can a simple function for the dielectric response model electrostatic effects in globular proteins? Protein Eng. 1989;2(7):527–30. [PubMed: 2664761]

116.

Mehler EL. Comparison of dielectric response models for simulating electrostatic effects in proteins. Protein Eng. 1990;3(5):415–7. [PubMed: 2349211]

117.

Mehler EL. Self-consistent free energy based approximation to calculate pH dependent electrostatic effects in proteins. J Phys Chem. 1996;100:16006–16018. [PMC free article: PMC1300308] [PubMed: 10388736]

118.

Mehler EL, Guarnieri F. A self-consistent, microenvironment modulated screened coulomb potential approximation to calculate pH-dependent electrostatic effects in proteins. Biophys J. 1999;77(1):3–22. [PMC free article: PMC1300308] [PubMed: 10388736]

119.

Mehler EL, Solmajer T. Electrostatic effects in proteins: comparison of dielectric and charge models. Prot Eng. 1991;4(8):903–10. [PubMed: 1667878]

120.

Alexov EG, Gunner MR. Incorporating protein conformational flexibility into the calculation of pH-dependent protein properties. Biophys J. 1997;72(5):2075–93. [PMC free article: PMC1184402] [PubMed: 9129810]

121.

Pitera JW, Falta M, van Gunsteren WF. Dielectric properties of proteins from simulation: the effects of solvent, ligands, pH, and temperature. Biophys J. 2001;80(6):2546–55. [PMC free article: PMC1301444] [PubMed: 11371433]

122.

Schutz CN, Warshel A. What are the dielectric “constants” of proteins and how to validate electrostatic models? Proteins. 2001;44(4):400–17. [PubMed: 11484218]

123.

Anderson MP, Gregory RJ, Thompson S. et al. Demonstration that CFTR is a chloride channel by alteration of its anion selectivity. Science. 1991;253(5016):202–5. [PubMed: 1712984]

124.

Tabcharani JA, Chang XB, Riordan JR. et al. The cystic fibrosis transmembrane conductance regulator chloride channel. Iodide block and permeation. Biophys J. 1992;62(1):1–4. [PMC free article: PMC1260465] [PubMed: 1376160]

125.

Fahlke C, Rhodes TH, Desai RR. et al. Pore stoichiometry of a voltage-gated chloride channel. Nature. 1998;394(6694):687–90. [PubMed: 9716133]

126.

Samoilov OY. Residence times of ionic hydrationIn: Horne RA, ed.Water and aqueous solutions: structure, thermodynamics, and transport processesBurlington, MA:Wiley-Interscience.1972597–612.

127.

Kropman MF, Bakker HJ. Dynamics of water molecules in aqueous solvation shells. Science. 2001;291(5511):2118–20. [PubMed: 11251110]

128.

Overholt JL, Hobert ME, Harvey RD. On the mechanism of rectification of the isoproterenol-activated chloride current in guinea-pig ventricular myocytes. J Gen Physiol. 1993;102(5):871–95. [PMC free article: PMC2229181] [PubMed: 8301261]

129.

Overholt JL, Saulino A, Drumm ML. et al. Rectification of whole cell cystic fibrosis transmembrane conductance regulator chloride current. Am J Physiol. 1995;268(3 Pt 1):C636–46. [PubMed: 7534982]

130.

Linsdell P. Relationship between anion binding and anion permeability revealed by mutagenesis within the cystic fibrosis transmembrane conductance regulator chloride channel pore. J Physiol. 2001;531(Pt 1):51–66. [PMC free article: PMC2278455] [PubMed: 11179391]

131.

Linsdell P. Thiocyanate as a probe of the cystic fibrosis transmembrane conductance regulator chloride channel pore. Can J Physiol Pharmacol. 2001;79(7):573–9. [PubMed: 11478590]

132.

Gong X, Burbridge SM, Cowley EA. et al. Molecular determinants of Au(CN)(2)(-) binding and permeability within the cystic fibrosis transmembrane conductance regulator Cl(-) channel pore. J Physiol. 2002;540(Pt 1):39–47. [PMC free article: PMC2290216] [PubMed: 11927667]

133.

Dawson DC, Smith SS. Cystic fibrosis transmembrane conductance regulator. Permeant ions find the pore [editorial] J Gen Physiol. 1997;110(4):337–9. [PMC free article: PMC2229378] [PubMed: 9379166]

134.

Hodgkin AL, Keynes RD. The potassium permeability of a giant nerve fibre. J Physiol (Lond.). 1955;128:61–88. [PMC free article: PMC1365755] [PubMed: 14368575]

135.

Almers W, McCleskey EW. Non-selective conductance in calcium channels of frog muscle: calcium selectivity in a single-file pore. J Physiol. 1984;353:585–608. [PMC free article: PMC1193323] [PubMed: 6090646]

136.

Hess P, Tsien RW. Mechanism of ion permeation through calcium channels. Nature. 1984;309(5967):453–6. [PubMed: 6328315]

137.

Dang TX, McCleskey EW. Ion channel selectivity through stepwise changes in binding affinity. J Gen Physiol. 1998;111(2):185–93. [PMC free article: PMC2222772] [PubMed: 9450938]

138.

Linsdell P, Evagelidis A, Hanrahan JW. Molecular determinants of anion selectivity in the cystic fibrosis transmembrane conductance regulator chloride channel pore. Biophys J. 2000;78(6):2973–82. [PMC free article: PMC1300881] [PubMed: 10827976]

139.

McCarty NA. Permeation through the CFTR chloride channel. J Exp Biol. 2000;203 Pt 13:1947–62. [PubMed: 10851114]

140.

McCarty NA, Zhang ZR. Identification of a region of strong discrimination in the pore of CFTR. Am J Physiol Lung Cell Mol Physiol. 2001;281(4):L852–67. [PubMed: 11557589]

141.

Linsdell P, Gong X. Multiple inhibitory effects of Au(CN)(2-) ions on cystic fibrosis transmembrane conductance regulator Cl(-) channel currents. J Physiol. 2002;540(Pt 1):29–38. [PMC free article: PMC2290227] [PubMed: 11927666]

142.

Halm DR, Frizzell RA. Anion permeation in an apical membrane chloride channel of a secretory epithelial cell. J Gen Physiol. 1992;99(3):339–66. [PMC free article: PMC2216607] [PubMed: 1375274]

143.

Cotten JF, Welsh MJ. Cystic fibrosis-associated mutations at arginine 347 alter the pore architecture of CFTR-Evidence for disruption of a salt bridge. J Biol Chem. 1999;274(9):5429–5435. [PubMed: 10026154]

144.

Ishihara H, Welsh MJ. Block by MOPS reveals a conformation change in the CFTR pore produced by ATP hydrolysis. Am J Physiol. 1997;273(4 Pt 1):C1278–89. [PubMed: 9357772]

145.

Linsdell P, Hanrahan JW. Adenosine triphosphate-dependent asymmetry of anion permeation in the cystic fibrosis transmembrane conductance regulator chloride channel. J Gen Physiol. 1998;111(4):601–614. [PMC free article: PMC2217125] [PubMed: 9524141]

146.

Kogan I, Ramjeesingh M, Huan LJ. et al. Perturbation of the pore of the cystic fibrosis transmembrane conductance regulator (CFTR) inhibits its atpase activity. J Biol Chem. 2001;276(15):11575–81. [PubMed: 11124965]

147.

Zhang ZR, Zeltwanger S, Smith SS. et al. Voltage-sensitive gating induced by a mutation in the fifth transmembrane domain of CFTR. Am J Physiol Lung Cell Mol Physiol. 2002;282(1):L135–45. [PubMed: 11741825]

148.

Yuan C, Kuwata O, Liang J. et al. Chloride binding regulates the Schiff base pK in gecko P521 cone-type visual pigment. Biochemistry. 1999;38(14):4649–54. [PubMed: 10194387]

149.

Pikkemaat MG, Ridder IS, Rozeboom HJ. et al. Crystallographic and kinetic evidence of a collision complex formed during halide import in haloalkane dehalogenase. Biochemistry. 1999;38(37):12052–61. [PubMed: 10508409]

150.

Schumacher MA, Scott DM, Mathews II. et al. Crystal structures of Toxoplasma gondii adenosine kinase reveal a novel catalytic mechanism and prodrug binding. J Mol Biol. 2000;296(2):549–67. [PubMed: 10669608]

151.

Feller G, Bussy O, Houssier C. et al. Structural and functional aspects of chloride binding to Alteromonas haloplanctis alpha-amylase. J Biol Chem. 1996;271(39):23836–41. [PubMed: 8798613]

152.

Strobl S, Maskos K, Betz M. et al. Crystal structure of yellow meal worm alpha-amylase at 1.64 A resolution. J Mol Biol. 1998;278(3):617–28. [PubMed: 9600843]

153.

Numao S, Maurus R, Sidhu G. et al. Probing the role of the chloride ion in the mechanism of human pancreatic alpha-amylase. Biochemistry. 2002;41(1):215–25. [PubMed: 11772019]

154.

Jayaraman S, Haggie P, Wachter RM. et al. Mechanism and cellular applications of a green fluorescent protein-based halide sensor. J Biol Chem. 2000;275(9):6047–50. [PubMed: 10692389]

155.

Wachter RM, Yarbrough D, Kallio K. et al. Crystallographic and energetic analysis of binding of selected anions to the yellow variants of green fluorescent protein. J Mol Biol. 2000;301(1):157–71. [PubMed: 10926499]

156.

Hol WG. The role of the alpha-helix dipole in protein function and structure. Prog Biophys Mol Biol. 1985;45(3):149–95. [PubMed: 3892583]

157.

Hol WG. Effects of the alpha-helix dipole upon the functioning and structure of proteins and peptides. Adv Biophys. 1985;19:133–65. [PubMed: 2424281]

158.

Dietrich B, Hosseini MW. Historical view on the development of anion coordination chemistryIn: Bianchi A, Bowman-James K, Garcia-España E ed(s).Supramolecular chemistry of anionsNew York:Wiley-VCH.199745–62.

159.

White MM, Miller C. Probes of the conduction process of a voltage-gated Cl^- channel from Torpedo electroplax. J Gen Physiol. 1981;78(1):1–18. [PMC free article: PMC2228628] [PubMed: 6265592]

160.

Harrop SJ, DeMaere MZ, Fairlie WD. et al. Crystal structure of a soluble form of the intracellular chloride ion channel CLIC1 (NCC27) at 1.4-A resolution. J Biol Chem. 2001;276(48):44993–5000. [PubMed: 11551966]

161.

Roberts DD, Lewis SD, Ballou DP. et al. Reactivity of small thiolate anions and cysteine-25 in papain toward methyl methanethiosulfonate. Biochemistry. 1986;25(19):5595–601. [PubMed: 3778876]

162.

Stauffer DA, Karlin A. Electrostatic potential of the acetylcholine binding sites in the nicotinic receptor probed by reactions of binding-site cysteines with charged methanethiosulfonates. Biochemistry. 1994;33(22):6840–9. [PubMed: 8204619]

163.

Akabas MH, Kaufmann C, Cook TA. et al. Amino acid residues lining the chloride channel of the cystic fibrosis transmembrane conductance regulator. J Biol Chem. 1994;269(21):14865–8. [PubMed: 7515047]

164.

Cheung M, Akabas MH. Identification of cystic fibrosis transmembrane conductance regulator channel-lining residues in and flanking the M6 membrane-spanning segment. Biophys J. 1996;70(6):2688–95. [PMC free article: PMC1225248] [PubMed: 8744306]

165.

Akabas MH, Cheung M, Guinamard R. Probing the structural and functional domains of the CFTR chloride channel. J Bioenerg Biomembr. 1997;29(5):453–63. [PubMed: 9511930]

166.

Cheung M, Akabas MH. Locating the anion-selectivity filter of the cystic fibrosis transmembrane conductance regulator (CFTR) chloride channel. J Gen Physiol. 1997;109(3):289–299. [PMC free article: PMC2217075] [PubMed: 9089437]

167.

Akabas MH. Channel-Lining Residues In the M3 Membrane-Spanning Segment Of the Cystic Fibrosis Transmembrane Conductance Regulator. Biochemistry. 1998;37(35):12233–12240. [PubMed: 9724537]

168.

Guinamard R, Akabas MH. Arg352 is a major determinant of charge selectivity in the cystic fibrosis transmembrane conductance regulator chloride channel. Biochemistry. 1999;38(17):5528–37. [PubMed: 10220340]

169.

Liu X, Zhang Z-R, Billingsley J. et al. CFTR: pH titration and chemical modification indicate that T338C (TM6) lies on the outward- facing, water-accessible surface of the protein. Ped Pulmonogy Suppl. 2001;22:A1.

170.

Middleton RE, Pheasant DJ, Miller C. Purification, reconstitution, and subunit composition of a voltage-gated chloride channel from Torpedo electroplax. Biochemistry. 1994;33(45):13189–98. [PubMed: 7947726]

171.

Ludewig U, Pusch M, Jentsch TJ. Two physically distinct pores in the dimeric ClC-0 chloride channel. Nature. 1996;383(6598):340–3. [PubMed: 8848047]

172.

Weinreich F, Jentsch TJ. Pores formed by single subunits in mixed dimers of different CLC chloride channels. J Biol Chem. 2001;276(4):2347–53. [PubMed: 11035003]

173.

Middleton RE, Pheasant DJ, Miller C. et al. Homodimeric architecture of a ClC-type chloride ion channel. Nature. 1996;383(6598):337–40. [PubMed: 8848046]

174.

Mindell JA, Maduke M, Miller C. et al. Projection structure of a ClC-type chloride channel at 6.5 A resolution. Nature. 2001;409(6817):219–23. [PubMed: 11196649]

175.

Marshall J, Fang S, Ostedgaard LS. et al. Stoichiometry of recombinant cystic fibrosis transmembrane conductance regulator in epithelial cells and its functional reconstitution into cells in vitro. J Biol Chem. 1994;269(4):2987–95. [PubMed: 7507932]

176.

Ostedgaard LS, Welsh MJ. Partial purification of the cystic fibrosis transmembrane conductance regulator. J Biol Chem. 1992;267(36):26142–9. [PubMed: 1281484]

177.

Ramjeesingh M, Li C, Kogan I. et al. A monomer is the minimum functional unit required for channel and ATPase activity of the cystic fibrosis transmembrane conductance regulator. Biochemistry. 2001;40(35):10700–6. [PubMed: 11524016]

178.

Chen JH, Chang XB, Aleksandrov AA. et al. CFTR is a Monomer: Biochemical and Functional Evidence. J Membr Biol. 2002;188(1):55–71. [PubMed: 12172647]

179.

Eskandari S, Wright EM, Kreman M. et al. Structural analysis of cloned plasma membrane proteins by freeze-fracture electron microscopy. Proc Natl Acad Sci U S A. 1998;95(19):11235–40. [PMC free article: PMC21625] [PubMed: 9736719]

180.

Tao T, Xie J, Drumm ML. et al. Slow conversions among subconductance states of cystic fibrosis transmembrane conductance regulator chloride channel. Biophys J. 1996;70(2):743–53. [PMC free article: PMC1224974] [PubMed: 8789091]

181.

Zerhusen B, Zhao JY, Xie JX. et al. A single conductance pore for chloride ions formed by two cystic fibrosis transmembrane conductance regulator molecules. J Biol Chem. 1999;274(12):7627–7630. [PubMed: 10075649]

182.

Naren AP, Cormet-Boyaka E, Fu J. et al. CFTR chloride channel regulation by an interdomain interaction. Science. 1999;286(5439):544–8. [PubMed: 10521352]

183.

Naren AP, Kirk KL. CFTR Chloride Channels: Binding Partners and Regulatory Networks. News Physiol Sci. 2000;15:57–61. [PubMed: 11390879]

184.

Fu J, Kirk KL. Cysteine substitutions reveal dual functions of the amino-terminal tail in cystic fibrosis transmembrane conductance regulator channel gating. J Biol Chem. 2001;276(38):35660–8. [PubMed: 11468285]

185.

Wang S, Yue H, Derin RB. et al. Accessory protein facilitated CFTR-CFTR interaction, a molecular mechanism to potentiate the chloride channel activity. Cell. 2000;103(1):169–79. [PubMed: 11051556]

186.

Raghuram V, Mak DO, Foskett JK. Regulation of cystic fibrosis transmembrane conductance regulator single-channel gating by bivalent PDZ-domain-mediated interaction. Proc Natl Acad Sci U S A. 2001;98(3):1300–1305. [PMC free article: PMC14749] [PubMed: 11158634]

187.

Ostedgaard LS, Rich DP, DeBerg LG. et al. Association of domains within the cystic fibrosis transmembrane conductance regulator. Biochemistry. 1997;36(6):1287–1294. [PubMed: 9063876]

188.

Devidas S, Yue H, Guggino WB. The second half of the cystic fibrosis transmembrane conductance regulator forms a functional chloride channel. J Biol Chem. 1998;273(45):29373–80. [PubMed: 9792638]

189.

Rosenberg MF, Callaghan R, Ford RC. et al. Structure of the multidrug resistance P-glycoprotein to 2.5 nm resolution determined by electron microscopy and image analysis. J Biol Chem. 1997;272(16):10685–94. [PubMed: 9099718]

190.

Loo TW, Clarke DM. The glycosylation and orientation in the membrane of the third cytoplasmic loop of human P-glycoprotein is affected by mutations and substrates. Biochemistry. 1999;38(16):5124–9. [PubMed: 10213617]

191.

Loo TW, Clarke DM. Identification of residues in the drug-binding domain of human P-glycoprotein. Analysis of transmembrane segment 11 by cysteine-scanning mutagenesis and inhibition by dibromobimane. J Biol Chem. 1999;274(50):35388–92. [PubMed: 10585407]

192.

Nikaido K, Liu PQ, Ames GF. Purification and characterization of HisP, the ATP-binding subunit of a traffic ATPase (ABC transporter), the histidine permease of Salmonella typhimurium. Solubility, dimerization, and ATPase activity. J Biol Chem. 1997;272(44):27745–52. [PubMed: 9346917]

193.

Walsh KB, Wang C. Arylaminobenzoate block of the cardiac cyclic AMP-dependent chloride current. Mol Pharmacol. 1998;53(3):539–46. [PubMed: 9495822]

194.

Zhang ZR, Zeltwanger S, McCarty NA. Direct comparison of NPPB and DPC as probes of CFTR expressed in Xenopus oocytes. J Membr Biol. 2000;175(1):35–52. [PubMed: 10811966]

195.

McCarty NA, McDonough S, Cohen BN. et al. Voltage-dependent block of the cystic fibrosis transmembrane conductance regulator Cl^- channel by two closely related arylaminobenzoates. J Gen Physiol. 1993;102(1):1–23. [PMC free article: PMC2229162] [PubMed: 8397274]

196.

Walsh KB, Long KJ, Shen X. Structural and ionic determinants of 5-nitro-2-(3-phenylprophyl-amino)-benzoic acid block of the CFTR chloride channel. Br J Pharmacol. 1999;127(2):369–76. [PMC free article: PMC1566034] [PubMed: 10385235]

197.

Neher E, Steinbach JH. Local anaesthetics transiently block currents through single acetylcholine- receptor channels. J Physiol. 1978;277:153–176. [PMC free article: PMC1282384] [PubMed: 306437]

198.

Aguilar-Bryan L, Nichols CG, Wechsler SW. et al. Cloning of the beta cell high-affinity sulfonylurea receptor: a regulator of insulin secretion. Science. 1995;268(5209):423–6. [PubMed: 7716547]

199.

Mehnert H, Karg E. Glibenclamide (HB 419): a new oral antidiabetic of sulphonylurea type. Ger Med Mon. 1969;14(8):373–7. [PubMed: 5349665]

200.

Sheppard DN, Welsh MJ. Effect of ATP-sensitive K channel regulators on cystic fibrosis transmembrane conductance regulator chloride currents. J Gen Physiol. 1992;100(4):573–91. [PMC free article: PMC2229110] [PubMed: 1281220]

201.

Schultz BD, DeRoos AD, Venglarik CJ. et al. Glibenclamide blockade of CFTR chloride channels. Am J Physiol. 1996;271(2 Pt 1):L192–200. [PubMed: 8770056]

202.

Sheppard DN, Robinson KA. Mechanism of glibenclamide inhibition of cystic fibrosis transmembrane conductance regulator Cl^- channels expressed in a murine cell line. J Physiol. 1997;503(Pt 2):333–46. [PMC free article: PMC1159866] [PubMed: 9306276]

203.

Venglarik CJ, Schultz BD, DeRoos AD. et al. Tolbutamide causes open channel blockade of cystic fibrosis transmembrane conductance regulator Cl^- channels. Biophys J. 1996;70(6):2696–703. [PMC free article: PMC1225249] [PubMed: 8744307]

204.

Cai Z, Lansdell KA, Sheppard DN. Inhibition of heterologously expressed cystic fibrosis transmembrane conductance regulator Cl^- channels by non-sulphonylurea hypoglycaemic agents. Br J Pharmacol. 1999;128(1):108–18. [PMC free article: PMC1571594] [PubMed: 10498841]

205.

Schultz BD, Singh AK, Devor DC. et al. Pharmacology of CFTR chloride channel activity. Physiol Rev. 1999;79(1 Suppl):S109–44. [PubMed: 9922378]

206.

Zhang ZR, Zeltwanger S, McCarty NA. Interactions between glibenclamide and DPC probe the wide inner vestibule of CFTR. Biophys J. 2000;78:467A.

207.

Gupta J, Linsdell P. Point mutations in the pore region directly or indirectly affect glibenclamide block of the CFTR chloride channel. Pflugers Arch. 2002;443(5):739–47. [PubMed: 11889571]

208.

Linsdell P. Direct block of the cystic fibrosis transmembrane conductance regulator Cl(-) channel by butyrate and phenylbutyrate. Eur J Pharmacol. 2001;411(3):255–60. [PubMed: 11164382]

209.

Linsdell P, Hanrahan JW. Flickery block of single CFTR chloride channels by intracellular anions and osmolytes. Am J Physiol. 1996;271(2 Pt 1):C628–34. [PubMed: 8770004]

210.

Linsdell P, Hanrahan JW. Disulphonic stilbene block of cystic fibrosis transmembrane conductance regulator Cl^- channels expressed in a mammalian cell line and its regulation by a critical pore residue. J Physiol (Lond ). 1996;496(3):687–693. [PMC free article: PMC1160856] [PubMed: 8930836]

211.

Armstrong CM. Anomalous rectification in the Squid Giant Axon injected with Tetraethylammonium chloride. J Gen Physiol. 1965;48:859–872. [PMC free article: PMC2213780] [PubMed: 14324992]

212.

Armstrong CM. Time course of TEA+-induced anomalous rectification in Squid Giant Axons. J Gen Physiol. 1966;50:491–503. [PMC free article: PMC2225642] [PubMed: 11526842]

213.

Armstrong CM. Inactivation of the potassium conductance and related phenomena caused by quaternary ammonium ion injection in Squid Axons. J Gen Physiol. 1969;54:553–575. [PMC free article: PMC2225944] [PubMed: 5346528]

214.

Armstrong CM. Interaction of Tetraethylammonium ion derivatives with the potassium channels of Giant Axons. J Gen Physiol. 1971;58:413–437. [PMC free article: PMC2226036] [PubMed: 5112659]

215.

Zhou Z, Hu S, Hwang TC. Probing an Open CFTR Pore with Organic Anion Blockers. J Gen Physiol. 2002;120(5):647–62. [PMC free article: PMC2229557] [PubMed: 12407077]