Toxin binding reveals two open state structures for one acid-sensing ion channel

Of the three principal conformations of acid-sensing ion channels (ASICs)—closed, open and desensitized—only the atomic structure of the desensitized conformation had been known. Two recent papers report the crystal structure of chicken ASIC1 in complex with the spider toxin psalmotoxin 1, and one of these studies finds that, depending on the pH, channels are in two different open conformations. Compared with the desensitized conformation, toxin binding induces only subtle structural changes in the lower part of the large extracellular domain but a complete rearrangement of the two transmembrane domains (TMDs), suggesting that desensitization gating (the transition from open to desensitized) is mainly associated with conformational rearrangements of the TMDs. Moreover, the study reveals how two different arrangements of the TMDs in the open state give rise to ion pores with different selectivity for monovalent cations.     

Identification of the Ca2+ blocking site of acid-sensing ion channel (ASIC) 1: implications for channel gating

Acid-sensing ion channels ASIC1a and ASIC1b are ligand-gated ion channels that are activated by H+ in the physiological range of pH. The apparent affinity for H+ of ASIC1a and 1b is modulated by extracellular Ca2+ through a competition between Ca2+ and H+. Here we show that, in addition to modulating the apparent H+ affinity, Ca2+ blocks ASIC1a in the open state (IC50 approximately 3.9 mM at pH 5.5), whereas ASIC1b is blocked with reduced affinity (IC50 > 10 mM at pH 4.7). Moreover, we report the identification of the site that mediates this open channel block by Ca2+. ASICs have two transmembrane domains. The second transmembrane domain M2 has been shown to form the ion pore of the related epithelial Na+ channel. Conserved topology and high homology in M2 suggests that M2 forms the ion pore also of ASICs. Combined substitution of an aspartate and a glutamate residue at the beginning of M2 completely abolished block by Ca2+ of ASIC1a, showing that these two amino acids (E425 and D432) are crucial for Ca2+ block. It has previously been suggested that relief of Ca2+ block opens ASIC3 channels. However, substitutions of E425 or D432 individually or in combination did not open channels constitutively and did not abolish gating by H+ and modulation of H+ affinity by Ca2+. These results show that channel block by Ca2+ and H+ gating are not intrinsically linked.     

On the molecular basis of ion permeation in the epithelial Na+ channel.

The epithelial Na+ channel (ENaC) is highly selective for Na+ and Li+ over K+ and is blocked by the diuretic amiloride. ENaC is a heterotetramer made of two alpha, one beta, and one gamma homologous subunits, each subunit comprising two transmembrane segments. Amino acid residues involved in binding of the pore blocker amiloride are located in the pre-M2 segment of beta and gamma subunits, which precedes the second putative transmembrane alpha helix (M2). A residue in the alpha subunit (alphaS589) at the NH2 terminus of M2 is critical for the molecular sieving properties of ENaC. ENaC is more permeable to Li+ than Na+ ions. The concentration of half-maximal unitary conductance is 38 mM for Na+ and 118 mM for Li+, a kinetic property that can account for the differences in Li+ and Na+ permeability. We show here that mutation of amino acid residues at homologous positions in the pre-M2 segment of alpha, beta, and gamma subunits (alphaG587, betaG529, gammaS541) decreases the Li+/Na+ selectivity by changing the apparent channel affinity for Li+ and Na+. Fitting single-channel data of the Li+ permeation to a discrete-state model including three barriers and two binding sites revealed that these mutations increased the energy needed for the translocation of Li+ from an outer ion binding site through the selectivity filter. Mutation of betaG529 to Ser, Cys, or Asp made ENaC partially permeable to K+ and larger ions, similar to the previously reported alphaS589 mutations. We conclude that the residues alphaG587 to alphaS589 and homologous residues in the beta and gamma subunits form the selectivity filter, which tightly accommodates Na+ and Li+ ions and excludes larger ions like K+.     

Strong activation of bile acid-sensitive ion channel (BASIC) by ursodeoxycholic acid

Bile acid-sensitive ion channel (BASIC) is a member of the DEG/ENaC gene family of unknown function. Rat BASIC (rBASIC) is inactive at rest. We have recently shown that cholangiocytes, the epithelial cells lining the bile ducts, are the main site of BASIC expression in the liver and identified bile acids, in particular hyo- and chenodeoxycholic acid, as agonists of rBASIC. Moreover, it seems that extracellular divalent cations stabilize the resting state of rBASIC, because removal of extracellular divalent cations opens the channel. In this addendum, we demonstrate that removal of extracellular divalent cations potentiates the activation of rBASIC by bile acids, suggesting an allosteric mechanism. Furthermore, we show that rBASIC is strongly activated by the anticholestatic bile acid ursodeoxycholic acid (UDCA), suggesting that BASIC might mediate part of the therapeutic effects of UDCA.     

The bile acid-sensitive ion channel (BASIC), the ignored cousin of ASICs and ENaC

The DEG/ENaC gene family of ion channels is characterized by a high degree of structural similarity and an equally high degree of diversity concerning the physiological function. In humans and rodents, the DEG/ENaC family comprises 2 main subgroups: the subunits of the epithelial Na⁺ channel (ENaC) and the subunits of the acid sensing ion channels (ASICs). The bile acid-sensitive channel (BASIC), previously known as BLINaC or INaC, represents a third subgroup within the DEG/ENaC family. Although BASIC was identified more than a decade ago, very little is known about its physiological function. Recent progress in the characterization of this neglected member of the DEG/ENaC family, which is summarized in this focused review, includes the discovery of surprising species differences, its pharmacological characterization, and the identification of bile acids as putative natural activators.     

The bile acid-sensitive ion channel (BASIC), the ignored cousin of ASICs and ENaC

The DEG/ENaC gene family of ion channels is characterized by a high degree of structural similarity and an equally high degree of diversity concerning the physiological function. In humans and rodents, the DEG/ENaC family comprises 2 main subgroups: the subunits of the epithelial Na⁺ channel (ENaC) and the subunits of the acid sensing ion channels (ASICs). The bile acid-sensitive channel (BASIC), previously known as BLINaC or INaC, represents a third subgroup within the DEG/ENaC family. Although BASIC was identified more than a decade ago, very little is known about its physiological function. Recent progress in the characterization of this neglected member of the DEG/ENaC family, which is summarized in this focused review, includes the discovery of surprising species differences, its pharmacological characterization, and the identification of bile acids as putative natural activators.     

Control of ion selectivity in LeuT: two Na+ binding sites with two different mechanisms

The x-ray structure of LeuT, a bacterial homologue of Na⁺/Cl⁻-dependent neurotransmitter transporters, provides a great opportunity to better understand the molecular basis of monovalent cation selectivity in ion-coupled transporters. LeuT possesses two ion binding sites, NA1 and NA2, which are highly selective for Na⁺. Extensive all-atom free-energy molecular dynamics simulations of LeuT embedded in an explicit membrane are performed at different temperatures and various occupancy states of the binding sites to dissect the molecular mechanism of ion selectivity. The results show that the two binding sites display robust selectivity for Na⁺ over K⁺ or Li⁺, the competing ions of most similar radii. Of particular interest, the mechanism primarily responsible for selectivity for each of the two binding sites appears to be different. In NA1, selectivity for Na⁺ over K⁺ arises predominantly from the strong electrostatic field arising from the negatively charged carboxylate group of the leucine substrate coordinating the ion directly. In NA2, which comprises only neutral ligands, selectivity for Na⁺ is enforced by the local structural restraints arising from the hydrogen-bonding network and the covalent connectivity of the polypeptide chain surrounding the ion according to a “snug-fit” mechanism.     

Differences in open state of NBA-modified cardiac Na+ channels

Patch clamp recordings from neonatal cardiac Na⁺ channels treated with N-bromoacetamide (NBA, 5–50 x 10^-mol/l) showed modified Na⁺ channel activity. By chemical removal of inactivation, repetitive openings with an increased life time and burst-like activity occurred. NBA-modified Na⁺ channels differ in life time and may attain either a slightly (mean open time 3.1±0.2 ms) or a strongly (mean open time 15.2±1.4 ms) prolonged open state. This strongly suggests a heterogeneous population of NBA-modified Na⁺ channels in newborn rat cardiocytes.     

Permeability properties of ENaC selectivity filter mutants.

The epithelial Na(+) channel (ENaC), located in the apical membrane of tight epithelia, allows vectorial Na(+) absorption. The amiloride-sensitive ENaC is highly selective for Na(+) and Li(+) ions. There is growing evidence that the short stretch of amino acid residues (preM2) preceding the putative second transmembrane domain M2 forms the outer channel pore with the amiloride binding site and the narrow ion-selective region of the pore. We have shown previously that mutations of the alphaS589 residue in the preM2 segment change the ion selectivity, making the channel permeant to K(+) ions. To understand the molecular basis of this important change in ionic selectivity, we have substituted alphaS589 with amino acids of different sizes and physicochemical properties. Here, we show that the molecular cutoff of the channel pore for inorganic and organic cations increases with the size of the amino acid residue at position alpha589, indicating that alphaS589 mutations enlarge the pore at the selectivity filter. Mutants with an increased permeability to large cations show a decrease in the ENaC unitary conductance of small cations such as Na(+) and Li(+). These findings demonstrate the critical role of the pore size at the alphaS589 residue for the selectivity properties of ENaC. Our data are consistent with the main chain carbonyl oxygens of the alphaS589 residues lining the channel pore at the selectivity filter with their side chain pointing away from the pore lumen. We propose that the alphaS589 side chain is oriented toward the subunit-subunit interface and that substitution of alphaS589 by larger residues increases the pore diameter by adding extra volume at the subunit-subunit interface.     

Characterization of a partially degraded Na+ channel from urinary tract epithelium

Summary The mammalian urinary bladder contains in its apical membrane and cytoplasmic vesicles, a cation-selective channel or activating fragment which seems to partition between the apical membrane and the luminal (or vesicular space). To determine whether it is an activating fragment or whole channel, we first demonstrate that solution known to contain this moiety can be concentrated and when added back to the bladder causes a conductance increase, with a percent recovery of 139±25%. Next, we show that using tip-dip bilayer techniques (at 21°C) and a patch-clamp recorder, the addition of concentrated solution resulted in the appearance of discrete current shots, consistent with the incorporation of a channel (as opposed to an activating fragment) into the bilayer. The residency time of the channel in the bilayer was best described by the sum of two exponentials, suggesting that the appearance of the channel involves an association of the channel with the membrane before insertion. The channel is cation selective and more conductive to K⁺ than Na⁺ (by a factor of 1.6). It has a linearI–V relationship, but a singlechannel conductance that saturates as KCl concentration is raised. This saturation is best described by the Michaelis-Menten equation with aK _m of 160mm KCl and aG _max of 20 pS. The kinetics of the channel are complex, showing at least two open and two closed states.Since the characteristics of this channel are similar to a channel produced by the degradation of amiloride-sensitive Na⁺ channels by the proteolytic enzyme kallikrein (which is released by the cortical collecting duct of the kidney), we suggest that this channel then is not synthesized by the cell but is rather a degraded form of the epithelial Na⁺ channel.     

Quaternary ammonium ion blockade of IK in nerve axons revisited. Open channel block vs. state independent block

The mechanism of blockade of the delayed rectifier potassium ion channel in squid giant axons by intracellular quaternary ammonium ions (QA) appears to be remarkably sensitive to the structure of the blocker. TEA, propyltriethyl-ammonium (C₃), and propyltetraethylammonium (TAA-C₃) all fail to alter the deactivation, or tail current time course following membrane depolarization, even with relatively large concentrations of the blockers, whereas butyltriethylammonium (C₄), butyltetraethylammonium (TAA-C₄), and pentytriethyammonium (C₅) clearly do have such an effect. The relative electrical distance of blockade for all of these ions is 0.25–0.3 from the inner surface of the membrane. The observations concerning TEA, C₃, and TAAC₃ suggest that these ions can block the channel in either its open or its closed state. The results with C₄, TAA-C₄, and C₅ are consistent with the open channel block model. Moreover, the sensitivity of block mechanism to the structure of the blocker suggests that the gate is located close to the QA ion binding site and that TEA, C₃, and TAA-C₃ do not interfere with channel gating, whereas C₄, TAA-C₄, C₅, and ions having a longer hydrophobic tail than C₅ do have such an effect. The parameters of block obtained for all QA ions investigated were unaffected by changes in the extracellular potassium ion concentration.The author gratefully acknowledges Paul Guth of Cambridge Chemical for custom synthesis of the C _n compounds used in this study, Michael Rogawski for helpful discussion of this work, and Adam Sherman of Alembic Software and Vijay Kowtha for technical assistance in data acquisition and analysis.     

Structural basis for the voltage-gated Na+ channel selectivity of the scorpion alpha-like toxin BmK M1

Scorpion alpha-like toxins are proteins that act on mammalian and insect voltage-gated Na+ channels. Therefore, these toxins constitute an excellent target for examining the foundations that underlie their target specificity. With this motive we dissected the role of six critical amino acids located in the five-residue reverse turn (RT) and C-tail (CT) of the scorpion alpha-like toxin BmK M1. These residues were individually substituted resulting in 11 mutants and were subjected to a bioassay on mice, an electrophysiological characterization on three cloned voltage-gated Na+ channels (Nav1.2, Nav1.5 and para), a CD analysis and X-ray crystallography. The results reveal two molecular sites, a couplet of residues (8-9) in the RT and a hydrophobic surface consisting of residues 57 and 59-61 in the CT, where the substitution with specific residues can redirect the alpha-like characteristics of BmK M1 to either total insect or much higher mammal specificity. Crystal structures reveal that the pharmacological ramification of these mutants is accompanied by the reshaping of the 3D structure surrounding position 8. Furthermore, our results also reveal that residues 57 and 59-61, located at the CT, enclose the critical residue 58 in order to form a hydrophobic "gasket". Mutants of BmK M1 that interrupt this hydrophobic surface significantly gain insect selectivity.     

Interaction of pyrethroids with the Na channel in mammalian neuronal cells in culture

The interaction of a series of pyrethroids with the Na⁺ channel of mouse neuroblastoma cells has been followed using both an electrophysiological and a ²²Na⁺ influx approach. By themselves, pyrethroids do not stimulate ²²Na⁺ entry through the Na⁺ channel (or the stimulation they give is too small to be analyzed). However, they stimulate ²²Na⁺ entry when used in conjuction with other toxins specific for the gating system of the channel. These include batrachotoxin, veratridine, dihydrograyanotoxin II or polypeptide toxins like sea anemone and scorpion toxins. This stimulatory effect is fully inhibited by tetrodotoxin with a dissociation constant of 1.6 nM for the tetrodotoxin-receptor complex. Half-maximum saturation of the pyrethroid receptor on the Na⁺ channel is observed in the micromolar range for the most active pyrethroids, Decis and RU 15525. The synergism observed between the effect of pyrethroids on ²²Na⁺ influx on the one hand, and the effects of sea anemone toxin II, Androctonus scorpion toxin II, batrachotoxin, veratridine and dihydrograyanotoxin II on the other, indicates that the binding component for pyrethroids on the Na⁺ channel is distinct from the other toxin receptors. It is also distinct from the tetrodotoxin receptor.Some of the pyrethroids used in this study bind to the Na⁺ channel but are unable to stimulate ²²Na⁺ entry. These inactive compounds behave as antagonists of the active pyrethroids.An electrophysiological approach has shown that pyrethroids by themselves are active on the Na⁺ channel of mammalian neurones, and essentially confirm the conclusions made from ²²Na⁺ flux measurements.Pyrethroids are also active on C9 cells in which Na⁺ channels are ‘silent’, that is, not activatable by electrical stimulation. Pyrethroids chemically activate the silent Na⁺ channel in a manner similar to that with veratridine, batrachotoxin, or polypeptide toxins, which are known to slow down the inactivation process of a functional Na⁺ channel.     

The characterization of a monovalent cation-selective channel of mammalian cardiac muscle sarcoplasmic reticulum

Summary Rabbit cardiac muscle sarcoplasmic reticulum (SR) was isolated and separated into ryanodine-sensitive and-insensitive fractions (L.R. Jones and S.E. Cala,J. Biol. Chem. 256:11809–11818, 1981). Vesicles of cardiac SR were incorporated into planar phospholipid bilayers by fusion and the channel activity of the membrane studied under voltage-clamp conditions (C. Miller,J. Membrane Biol. 40: 1–23, 1978). Both fractions contain a monovalent cation-selective three-state channel. In the presence of 75mm K₂SO₄, the fully open state () conductance of this channel is 157.2±30 pS and the sub-state () conductance is 100.7±21 pS. Both open states display the same selectivity sequence for monovalent cations, i.e. K⁺>NH ₄ ⁺ >Rb⁺>Na⁺>Li⁺ and may be blocked by the skeletal muscle relaxants decamethonium and hexamethonium. Block occurs when the compounds are added to either side of the membrane. The properties of the cardiac SR cation channel are compared with those of the previously reported monovalent cation-selective channels of mammalian and amphibian skeletal muscle SR.     

Uncoupling of gating charge movement and closure of the ion pore during recovery from inactivation in the Kv1.5 channel

Both wild-type (WT) and nonconducting W472F mutant (NCM) Kv1.5 channels are able to conduct Na(+) in their inactivated states when K(+) is absent. Replacement of K(+) with Na(+) or NMG(+) allows rapid and complete inactivation in both WT and W472F mutant channels upon depolarization, and on return to negative potentials, transition of inactivated channels to closed-inactivated states is the first step in the recovery of the channels from inactivation. The time constant for immobilized gating charge recovery at -100 mV was 11.1 +/- 0.4 ms (n = 10) and increased to 19.0 +/- 1.6 ms (n = 3) when NMG(+)(o) was replaced by Na(+)(o). However, the decay of the Na(+) tail currents through inactivated channels at -100 mV had a time constant of 129 +/- 26 ms (n = 18), much slower than the time required for gating charge recovery. Further experiments revealed that the voltage-dependence of gating charge recovery and of the decay of Na(+) tail currents did not match over a 60 mV range of repolarization potentials. A faster recovery of gating charge than pore closure was also observed in WT Kv1.5 channels. These results provide evidence that the recovery of the gating elements is uncoupled from that of the pore in Na(+)-conducting inactivated channels. The dissociation of the gating charge movements and the pore closure could also be observed in the presence of symmetrical Na(+) but not symmetrical Cs(+). This difference probably stems from the difference in the respective abilities of the two ions to limit inactivation to the P-type state or prevent it altogether.     

Regulation of neuronal ion channels via P2Y receptors

Within the last 15 years, at least 8 different G protein-coupled P2Y receptors have been characterized. These mediate slow metabotropic effects of nucleotides in neurons as well as non-neural cells, as opposed to the fast ionotropic effects which are mediated by P2X receptors. One class of effector systems regulated by various G protein-coupled receptors are voltage-gated and ligand-gated ion channels. This review summarizes the current knowledge about the modulation of such neuronal ion channels via P2Y receptors. The regulated proteins include voltage-gated Ca²⁺ and K⁺ channels, as well as N-methyl-d-aspartate, vanilloid, and P2X receptors, and the regulating entities include most of the known P2Y receptor subtypes. The functional consequences of the modulation of ion channels by nucleotides acting at pre- or postsynaptic P2Y receptors are changes in the strength of synaptic transmission. Accordingly, ATP and related nucleotides may act not only as fast transmitters (via P2X receptors) in the nervous system, but also as neuromodulators (via P2Y receptors). Hence, nucleotides are as universal transmitters as, for instance, acetylcholine, glutamate, or -aminobutyric acid.     

Accessibility of ENaC extracellular domain central core residues

The epithelial Na⁺ channel (ENaC)/degenerin family has a similar extracellular architecture, where specific regulatory factors interact and alter channel gating behavior. The extracellular palm domain serves as a key link to the channel pore. In this study, we used cysteine-scanning mutagenesis to assess the functional effects of Cys-modifying reagents on palm domain β10 strand residues in mouse ENaC. Of the 13 ENaC α subunit mutants with Cys substitutions examined, only mutants at sites in the proximal region of β10 exhibited changes in channel activity in response to methanethiosulfonate reagents. Additionally, Cys substitutions at three proximal sites of β and γ subunit β10 strands also rendered mutant channels methanethiosulfonate-responsive. Moreover, multiple Cys mutants were activated by low concentrations of thiophilic Cd²⁺. Using the Na⁺ self-inhibition response to assess ENaC gating behavior, we identified four α, two β, and two γ subunit β10 strand mutations that changed the Na⁺ self-inhibition response. Our results suggest that the proximal regions of β10 strands in all three subunits are accessible to small aqueous compounds and Cd²⁺ and have a role in modulating ENaC gating. These results are consistent with a structural model of mouse ENaC that predicts the presence of aqueous tunnels adjacent to the proximal part of β10 and with previously resolved structures of a related family member where palm domain structural transitions were observed with channels in an open or closed state.     

Protons and Psalmotoxin-1 reveal nonproton ligand stimulatory sites in chicken acid-sensing ion channel: Implication for simultaneous modulation in ASICs

Acid-sensing ion channels (ASICs) are proton-sensitive, sodium-selective channels expressed in the nervous system that sense changes in extracellular pH. These ion channels are sensitive to an increasing number of nonproton ligands that include natural venom peptides and guanidine compounds. In the case of chicken ASIC1, the spider toxin Psalmotoxin-1 (PcTx1) activates the channel, resulting in an inward current. Furthermore, a growing class of ligands containing a guanidine group has been identified that stimulate peripheral ASICs (ASIC3), but exert subtle influence on other ASIC subtypes. The effects of the guanidine compounds on cASIC1 have not been the focus of previous study. Here, we investigated the interaction of the guanidine compound 2-guanidine-4-methylquinazoline (GMQ) on cASIC1 proton activation and PcTx1 stimulation. Exposure of expressed cASIC1 to PcTx1 resulted in biphasic currents consisting of a transient peak followed by an irreversible cASIC1 PcTx1 persistent current. This cASIC1 PcTx1 persistent current may be the result of locking the cASIC1 protein into a desensitized transition state. The guanidine compound GMQ increased the apparent affinity of protons on cASIC1 and decreased the half-maximal constant of the cASIC1 steady-state desensitization profile. Furthermore, GMQ stimulated the cASIC1 PcTx1 persistent current in a concentration-dependent manner, which resulted in a non-desensitizing inward current. Our data suggests that GMQ may have multiple sites within cASIC1 and may act as a “molecular wedge” that forces the PcTx1-desensitized ASIC into an open state. Our findings indicate that guanidine compounds, such as GMQ, may alter acid-sensing ion channel activity in combination with other stimuli, and that additional ASIC subtypes (along with ASIC3) may serve to sense and mediate signals from multiple stimuli.     

Interactions of K+ ATP channel blockers with Na+/K+-ATPase

Two K⁺ _ATP channel blockers, 5-hydroxydecanoate (5-HD) and glyburide, are often used to study cross-talk between Na⁺/K⁺-ATPase and these channels. The aim of this work was to characterize the effects of these blockers on purified Na⁺/K⁺-ATPase as an aid to appropriate use of these drugs in studies on this cross-talk. In contrast to known dual effects (activating and inhibitory) of other fatty acids on Na⁺/K⁺-ATPase, 5-HD only inhibited the enzyme at concentrations exceeding those that block mitochondrial K⁺ _ATP channels. 5-HD did not affect the ouabain sensitivity of Na⁺/K⁺-ATPase. Glyburide had both activating and inhibitory effects on Na⁺/K⁺-ATPase at concentrations used to block plasma membrane K⁺ _ATP channels. The findings justify the use of 5-HD as specific mitochondrial channel blocker in studies on the relation of this channel to Na⁺/K⁺-ATPase, but question the use of glyburide as a specific blocker of plasma membrane K⁺ _ATP channels, when the relation of this channel to Na⁺/K⁺-ATPase is being studied.     

Tuning the voltage dependence of tetraethylammonium block with permeant ions in an inward-rectifier K+ channel.

To understand the role of permeating ions in determining blocking ion-induced rectification, we examined block of the ROMK1 inward-rectifier K+ channel by intracellular tetraethylammonium in the presence of various alkali metal ions in both the extra- and intracellular solutions. We found that the channel exhibits different degrees of rectification when different alkali metal ions (all at 100 mM) are present in the extra- and intracellular solution. A quantitative analysis shows that an external ion site in the ROMK1 pore binds various alkali metal ions (Na+, K+, Rb+, and Cs+) with different affinities, which can in turn be altered by the binding of different permeating ions at an internal site through a nonelectrostatic mechanism. Consequently, the external site is saturated to a different level under the various ionic conditions. Since rectification is determined by the movement of all energetically coupled ions in the transmembrane electrical field along the pore, different degrees of rectification are observed in various combinations of extra- and intracellular permeant ions. Furthermore, the external and internal ion-binding sites in the ROMK1 pore appear to have different ion selectivity: the external site selects strongly against the smaller Na+, but only modestly among the three larger ions, whereas the internal site interacts quite differently with the larger K+ and Rb+ ions.