Isoform-dependent interaction of voltage-gated sodium channels with protons.

Protons are potent physiological modifiers of voltage-gated Na(+) channels, shifting the voltage range of channel gating and reducing current magnitude (pK(a) approximately 6). We recently showed that proton block of the skeletal muscle isoform (Na(V)1.4) resulted from protonation of the four superficial carboxylates in the outer vestibule of the channel. We concluded that the large local negative electrostatic field shifted the outer vestibule carboxylate pK(a) into the physiological range. However, block was not complete; the best-fit titration curves yielded an acid pH asymptote of 10-15%, suggesting that the selectivity filter carboxylates may not be protonated. Using HEK 293 cells stably expressing different isoforms, each with varying channel density, we demonstrate that a pH-independent current is found in Na(V)1.4, but not in the cardiac isoform (Na(V)1.5). Mutational studies showed that absence of the pH-independent current in Na(V)1.5 could be ascribed to the cysteine in domain I, just above the selectivity filter aspartate (Cys373). We suggest that this cysteine can be protonated in acid solution to produce a positive charge that blocks the pore. Competition between protons and Na(+) did not exist for Na(+) concentrations between 1 and 140 mm. The residual current in acid solution, when the cysteine is absent, confirms that over the range of pH values that can be achieved physiologically, the selectivity filter carboxylates are not protonated. The pH-independent current helps to protect activation of skeletal muscle during the acidosis that occurs during exercise.

Mechanism of local anesthetic drug action on voltage-gated sodium channels.

Local anesthetic drugs interfere with excitation and conduction by action potentials in the nervous system and in the heart by blockade of the voltage-gated Na channel. Drug affinity varies with gating state of the channel. The drugs show low affinity at slow excitation rates, but high affinity when the channels are opened and inactivated during action potentials at high frequency, as they are during pain or during a cardiac arrhythmia. The drugs are thought to access their binding site in the inner pore by passage through the membrane and entry through the inner pore vestibule. There have been three major developments in the last decade that greatly increase our understanding of their mechanism of action. Firstly, amino acid residues critical to drug binding have been located by mutagenesis, and it is possible to develop a molecular model of the drug binding site. Secondly, a path for drug access directly from the outside has been characterized in the cardiac isoform of the channel. Thirdly, the hypothesis that high affinity binding stabilizes the fast inactivated conformation of the channel has been challenged. Rather, the drug may stabilize a slow inactivated state and immobilize the voltage sensor in domain III in its activated outward position. The combination of mutational study of the cloned Na channels and patch clamp offers the opportunity to understand the detailed molecular mechanism of drug action and to resolve drug structure-function.

Modeling of the outer vestibule and selectivity filter of the L-type Ca2+ channel.

Using the KcsA bacterial K+ channel crystal structure [Doyle, D. A., et al. (1998) Science 280, 69-74] and the model of the outer vestibule of the Na+ channel [Lipkind, G. M., and Fozzard, H. A. (2000) Biochemistry 39, 8161-8170] as structural templates, we propose a structural model of the outer vestibule and selectivity filter of the pore of the Ca2+ channel (alpha1C or Ca(v)1.2). The Ca2+ channel P loops were modeled by alpha-helix-turn-beta-strand motifs, with the glutamate residues of the EEEE motif located in the turns. P loops were docked in the extracellular part of the inverted teepee structure formed by S5 and S6 alpha-helices with backbone coordinates from the M1 and M2 helices of the KcsA crystal structure. This construction results in a conical outer vestibule that tapers to the selectivity filter at the bottom. The modeled selectivity ring forms a wide open pore ( approximately 6 A) in the absence of Ca2+. When Ca2+ is present ( approximately 1 microM), all four glutamate side chains move to the center and form a cage around the dehydrated Ca2+ ion, blocking the pore. In the millimolar concentration range, Ca2+ also interacts with two low-affinity sites located externally and internally, which were modeled by the same carboxylate groups of the selectivity filter. Calculation of the resulting electrostatic potentials show that the single Ca2+ ion is located in an electrostatic trap. Only when three Ca2+ ions are bound simultaneously in the high- and low-affinity sites of the selectivity filter is Ca2+ able to overcome electrostatic attraction, permitting Ca2+ flux.

KcsA crystal structure as framework for a molecular model of the Na(+) channel pore.

The crystal structure of the pore-forming part of the KcsA bacterial K(+)-selective channel suggests a possible motif for related voltage-gated channels. We examined the hypothesis that the spacial orientation of the KcsA M1 and M2 alpha-helices also predicts the backbone location of S5 and S6 helices of the voltage-gated Na(+) channel. That channel's P region structure is expected to be different because selectivity is determined by side-chain interactions rather than by main-chain carbonyls, and its outer vestibule accommodates relatively large toxin molecules, tetrodotoxin (TTX) and saxitoxin (STX), which interact with selectivity ring residues. The Na(+) channel P loop was well-modeled by the alpha-helix-turn-beta-strand motif, which preserves the relationships for toxin interaction with the Na(+) channel found experimentally. This outer vestibule was docked into the extracellular part of the inverted teepee structure formed by the S5 and S6 helices that were spacially located by coordinates of the KcsA M1 and M2 helix main chains [Doyle et al. (1998) Science 280, 69-74], but populated with side chains of the respective S5 and S6 structures. van der Waals contacts were optimized with minimal adjustment of the S5, S6, and P loop structures, forming a densely packed pore structure. Nonregular external S5-P and P-S6 segments were not modeled here, except the P-S6 segment of domain II. The resulting selectivity region structure is consistent with Na(+) channel permeation properties, offering suggestions for the molecular processes involved in selectivity. The ability to construct a Na(+) channel pore model consistent with most of the available biophysical and mutational information suggests that the KcsA structural framework may be conserved in voltage-gated channels.

Differences in saxitoxin and tetrodotoxin binding revealed by mutagenesis of the Na+ channel outer vestibule.

The marine guanidinium toxins, saxitoxin (STX) and tetrodotoxin (TTX), have played crucial roles in the study of voltage-gated Na+ channels. Because they have similar actions, sizes, and functional groups, they have been thought to associate with the channel in the same manner, and early mutational studies supported this idea. Recent experiments by. Biophys. J. 67:2305-2315) have suggested that the toxins bind differently to the isoform-specific domain I Phe/Tyr/Cys location. In the adult skeletal muscle Na+ channel isoform (microliter), we compared the effects on both TTX and STX affinities of mutations in eight positions known to influence toxin binding. The results permitted the assignment of energies contributed by each amino acid to the binding reaction. For neutralizing mutations of Asp400, Glu755, and Lys1237, all thought to be part of the selectivity filter of the channel, the loss of binding energy was identical for the two toxins. However, the loss of binding energy was quite different for vestibule residues considered to be more superficial. Specifically, STX affinity was reduced much more by neutralizations of Glu758 and Asp1532. On the other hand, mutation of Tyr401 to Cys reduced TTX binding energy twice as much as it reduced STX binding energy. Kinetic analysis suggested that all outer vestibule residues tested interacted with both toxins early in the binding reaction (consistent with larger changes in the binding than unbinding rates) before the transition state and formation of the final bound complex. We propose a revised model of TTX and STX binding in the Na+ channel outer vestibule in which the toxins have similar interactions at the selectivity filter, TTX has a stronger interaction with Tyr401, and STX interacts more strongly with the more extracellular residues.

A model for the structure of the P domains in the subtilisin-like prohormone convertases.

The proprotein convertases are a family of at least seven calcium-dependent endoproteases that process a wide variety of precursor proteins in the secretory pathway. All members of this family possess an N-terminal proregion, a subtilisin-like catalytic module, and an additional downstream well-conserved region of approximately 150 amino acid residues, the P domain, which is not found in any other subtilase. The pro and catalytic domains cannot be expressed in the absence of the P domains; their thermodynamic instability may be attributable to the presence of large numbers of negatively charged Glu and Asp side chains in the substrate binding region for recognition of multibasic residue cleavage sites. Based on secondary structure predictions, we here propose that the P domains consist of 8-stranded beta-barrels with well-organized inner hydrophobic cores, and therefore are independently folded components of the proprotein convertases. We hypothesize further that the P domains are integrated through strong hydrophobic interactions with the catalytic domains, conferring structural stability and regulating the properties and activity of the convertases. A molecular model of these interdomain interactions is proposed in this report.

Predominant interactions between mu-conotoxin Arg-13 and the skeletal muscle Na+ channel localized by mutant cycle analysis.

High-affinity mu-conotoxin block of skeletal muscle Na+ channels depends on an arginine at position 13 (Arg-13). To understand both the mechanism of toxin interaction and the general structure of its binding site in the channel mouth, we examined by thermodynamic mutant cycle analysis the interaction between the critical Arg-13 and amino acid residues known to be in the channel's outer vestibule. Arg-13 interacts specifically with domain II Glu-758 with energy of about -3.0 kcal/mol, including both electrostatic and nonelectrostatic components, and with Glu-403 with energy of about -2.0 kcal/mol. Interactions with the other charged residues in the outer vestibule were shown to be almost entirely electrostatic, because these interactions were maintained when Arg-13 was replaced by lysine. These results place the bound Arg-13 at the channel mouth adjacent to the P (pore) loops of domains I and II. Distance estimates based on interaction energies suggest that the charged vestibule residues are in relative positions similar to those of the Lipkind-Fozzard vestibule model [Lipkind, G. M., and Fozzard, H. A. (1994) Biophys. J. 66, 1-13]. Kinetic analysis suggests that Arg-13 interactions are partially formed in the ligand-channel transition state.

A model of scorpion toxin binding to voltage-gated K+ channels.

Mutational studies have identified part of the S5-S6 loop of voltage-dependent K+ channels (P region) responsible for tetraethylammonium (TEA) block and permeation properties. Several scorpion peptide toxins-charybdotoxin (ChTX), kaliotoxin (KITX), and agitoxin (AgTX)-also block the channel with high affinity and specificity. Here, we examine the interaction predicted when the toxins are docked onto the molecular model of the K+ channel pore that we recently proposed. Docking with the model of the Kv1.3 channel started by location of Lys-27 side chain into the central axis of the pore, followed by energy minimization. In the optimal arrangement, Arg-24 of KITX or AgTX forms a hydrogen bond with the Asp-386 carboxyl of one subunit, and Asn-30 is in immediate contact with Asp-386 of the opposing subunit in the tetramer. Toxin residues in proximity to the side chain of Lys-27 (Phe-25, Thr-36, Met-29, and Ser-11 in KITX) interact with the four C-end His-404s. For ChTX the interaction with Asp-386 is reduced, but this is compensated by additional nonbonded interactions formed by Tyr-36 and Arg-34. Comparison of calculated energy of interaction of these specific toxin-channel residues with experimental studies reveals good agreement. The similar total calculated energy of interaction is consistent with the similar IC50 for Kv1.3 block by KITX and AgTX. Steric contacts of residues in position 380 of the S5-P linker with residues on the upper part of toxins permit reconstruction of the K+ channel outer vestibule walls, which are about 30 A apart and about 9 A high. Molecular modeling shows complementarity of the pore model to toxin spacial structures, and supports the proposal that the N-terminal borders of the P regions surround residues of their C-terminal halves.

Identification of 14 new glucokinase mutations and description of the clinical profile of 42 MODY-2 families.

Mutations in glucokinase are associated with defects in insulin secretion and hepatic glycogen synthesis resulting in mild chronic hyperglycaemia, impaired glucose tolerance or diabetes mellitus. We screened members of 35 families with features of maturity-onset diabetes of the young for mutations in the glucokinase gene and found 16 different mutations. They included 14 new mutations in the glucokinase gene: 9 missense mutations (A53S, G80A, H137R, T168P, M210T, C213R, V226M, S336L and V367M); 2 nonsense mutations (E248X and S360X); a deletion of one nucleotide resulting in a frameshift (V401del1); a substitution of a conserved nucleotide at a splice acceptor site (L122-1G-->T); and a 10 base pair deletion that removed the GT of the splice donor site and the following eight nucleotides (K161 + 2del10). In addition, we found two previously identified mutations: R186X and G261R. Study of 260 subjects with glucokinase-deficient hyperglycaemia from 42 families with 36 different GCK mutations made it possible to define the clinical profile of this subtype of non-insulin-dependent diabetes mellitus (NIDDM). Hyperglycaemia due to glucokinase deficiency is often mild (fewer than 50% of subjects have overt diabetes) and is evident during the early years of life. Despite the long duration of hyperglycaemia, glucokinase-deficient subjects have a low prevalence of micro- and macro-vascular complications of diabetes. Obesity, arterial hypertension and dyslipidaemia are also uncommon in this form of NIDDM.

A structural motif for the voltage-gated potassium channel pore.

Mutation studies have identified a region of the S5-S6 loop of voltage-gated K+ channels (P region) responsible for teraethylammonium (TEA) block and permeation/selectivity properties. We previously modeled a similar region of the Na+ channel as four beta-hairpins with the C strands from each of the domains forming the external vestibule and with charged residues at the beta-turns forming the selectivity filter. However, the K+ channel P region amino acid composition is much more hydrophobic in this area. Here we propose a structural motif for the K+ channel pore based on the following postulates (Kv2.1 numbering). (i) The external TEA binding site is formed by four Tyr-380 residues; P loop residues participating in the internal TEA binding site are four Met-371 and Thr-372 residues. (ii) P regions form extended hairpins with beta-turns in sequence ITMT. (iii) only C ends of hairpins form the inner walls of the pore. (iv) They are extended nonregular strands with backbone carbonyl oxygens of segment VGYGD facing the pore with the conformation BRLRL. (v) Juxtaposition of P loops of the four subunits forms the pore. Fitting the external and internal TEA sites to TEA molecules predicts an hourglass-like pore with the narrowest point (GYG) as wide as 5.5 A, suggesting that selectivity may be achieved by interactions of carbonyls with partially hydrated K+. Other potential cation binding sites also exist in the pore.

A structural model of the tetrodotoxin and saxitoxin binding site of the Na+ channel.

Biophysical evidence has placed the binding site for the naturally occurring marine toxins tetrodotoxin (TTX) and saxitoxin (STX) in the external mouth of the Na+ channel ion permeation pathway. We developed a molecular model of the binding pocket for TTX and STX, composed of antiparallel beta-hairpins formed from peptide segments of the four S5-S6 loops of the voltage-gated Na+ channel. For TTX the guanidinium moiety formed salt bridges with three carboxyls, while two toxin hydroxyls (C9-OH and C10-OH) interacted with a fourth carboxyl on repeats I and II. This alignment also resulted in a hydrophobic interaction with an aromatic ring of phenylalanine or tyrosine residues for the brainII and skeletal Na+ channel isoforms, but not with the cysteine found in the cardiac isoform. In comparison to TTX, there was an additional interaction site for STX through its second guanidinium group with a carboxyl on repeat IV. This model satisfactorily reproduced the effects of mutations in the S5-S6 regions and the differences in affinity by various toxin analogs. However, this model differed in important ways from previously published models for the outer vestibule and the selectivity region of the Na+ channel pore. Removal of the toxins from the pocket formed by the four beta-hairpins revealed a structure resembling a funnel that terminated in a narrowed region suitable as a candidate for the selectivity filter of the channel. This region contained two carboxyls (Asp384 and Glu942) that substituted for molecules of water from the hydrated Na+ ion. Simulation of mutations in this region that have produced Ca2+ permeation of the Na+ channel created a site with three carboxyls (Asp384, Glu942, and Glu1714) in proximity.

[Computerized structural analysis of O-specific polysaccharides O1A, O1B, and O1C from Escherichia coli]. / Kompiuternyi strukturnyi analiz O-spetsificheskikh polisakharidov Escherichia coli O1A, O1B i O1C.

A computer evaluation of 13C-NMR data for the title polysaccharides based on the monosaccharide and methylation analysis data led to the structure of the repeating unit of the O1A polysaccharide as well as to several probable structures of the O1C polysaccharide, of which the correct one was inferred by means of a single NOE experiment. The analysis of the spectrum of the O1B polysaccharide was unsuccessful, due to the presence in its structure of the fragment alpha-L-Rha-(1-->2)-alpha-D-Gal-(1-->3)-D-GlcNAc with the terminal (1-->2)-linkage, whose spectral data could not be calculated by additive schemes using only glycosylation effects. However in reevaluation of the O1B spectral data by taking into account the deviations from additivities of the chemical shifts values in spectra of the related trisaccharides, to reveal the most probable structure of the O1B's repeating unit. [formula: see text]

Synthesis and 13C NMR spectra of 2,3-di-O-glycosyl derivatives of methyl alpha-L-rhamnopyranoside and methyl alpha-D-mannopyranoside.

The syntheses are described of 2,3-di-O-glycosyl derivatives of methyl alpha-L-rhamnopyranoside (1-5) and alpha-D-mannopyranoside (6-9). [formula: see text] [table: see text] The deviation from additivity in 13C NMR spectra calculated for 1-9 were similar for stereochemically related trisaccharides.

Computer-assisted analysis of the structure of regular branched polysaccharides containing 2,3-disubstituted rhamnopyranose and mannopyranose residues on the basis of 13C NMR data.

A computer-assisted approach to the analysis of the structure of branched polysaccharides that contain 2,3-di-O-glycosylated alpha-rhamnopyranose and alpha-mannopyranose residues is based on evaluation of the 13C NMR spectra, using glycosylation effects and their deviations from additivity (delta delta values) at the branch points. This approach, in combination with monosaccharide and methylation analysis data, has been verified on a series of bacterial polysaccharides of known structure.

[SCAN program for structural analysis of linear polysaccharides based on 13C-NMR spectral data using personal computers]. / Programma SCAN dlia strukturnogo analyiza lineinykh polisakharidov na osnove dannykh 13C-NMR-spectrov s ispol'zovaniem personal'nykh komp'iuterov.

A programme SCAN was elaborated for the 13C NMR--based structural analysis of regular polysaccharides, which represents our earlier programme (Lipkind G.M. et al. Carbohydr. Res. 1988. V. 175. No 1. P. 59-75) modified for IBM-PC-compatible personal computers. SCAN was successfully applied for the structural elucidation of 24 polysaccharides of bacterial origin. Optimal combinations of the computer-assisted method with other approaches for analysis of carbohydrates and scopes of application of the programme are discussed.

Synthesis of di-O-glycosyl derivatives of methyl alpha-L-rhamnopyranoside.

The syntheses are described of 2,3-di-O-glycosyl derivatives (1-12) of methyl alpha-L-rhamnopyranoside where the glycosyl moieties are variously alpha-L-fucopyranose, beta-L-fucopyranose, beta-D-glucopyranose, alpha-D-mannopyranose, and alpha-L-rhamnopyranose. The syntheses involve stereoselective glycosylation of methyl 4-O-benzoyl-3-O-(2,3,4-tri-O-benzoyl-alpha-L- rhamnopyranosyl)-alpha-L-rhamnopyranoside (21), methyl 4-O-benzoyl-3-O-(2,3,4,6-tetra-O-benzoyl-alpha-D-mannopyranosyl)- alpha-L-rhamnopyranoside (25), methyl 4-O-benzoyl-3-O-(2,3,4,6-tetra-O-benzoyl-beta-D-glucopyranosyl)- alpha-L-rhamnopyranoside (29), methyl 4-O-benzoyl-3-O-(2,3,4-tri-O-benzoyl-beta-L-fucopyranosyl)-alpha-L- rhamnopyranoside (35), and methyl 4-O-benzyl-2-O-(2,3,4-tri-O-benzoyl-beta-L-fucopyranosyl)- alpha-L-rhamnopyranoside (59). In the syntheses of compounds 7-9, the alpha-L-fucopyranosyl residues are introduced stereoselectively, using 2,3,4-tri-O-benzoyl-alpha-L-fucopyranosyl bromide (17) and ethyl 2,3,4-tri-O-acetyl-1-thio-beta-L-fucopyranoside (47) as glycosyl donors.

N.m.r. and conformational analysis of some 2,3-disubstituted methyl alpha-L-rhamnopyranosides.

Conformational studies of the branched trisaccharide glycosides X-(1----2)[Y-(1----3)]-alpha-L-Rha-OMe (where X and Y are residues of alpha-L-, beta-L-, alpha-D-, and beta-D-hexopyranoses) were based on 1H- and 13C-n.m.r. data (n.O.e.'s, 13C chemical shifts) and theoretical calculations. In the majority of the trisaccharide glycosides, there is insignificant restriction of rotation around the glycosidic linkages in the disaccharide units as compared to the corresponding disaccharide glycosides X-(1----2)-alpha-L-Rha-OMe and Y-(1----3)-alpha-L-Rha-OMe. Differences in the conformations observed for several compounds resulted in changes of the n.O.e. patterns and in deviations from additivity of glycosylation effects in the 13C-n.m.r. spectra.

The structure of the biological repeating unit of the O-antigen of Hafnia alvei O39.

On mild acid-catalysed degradation of the lipopolysaccharide from Hafnia alvei O39 followed by gel filtration of Sephadex G-50, the O-specific polysaccharide and three oligosaccharides were obtained, which represent the core substituted with 0-2 O-antigen repeating-units. On the basis of sugar and methylation analyses, 13C-n.m.r. data, solvolysis of the polysaccharide with anhydrous hydrogen fluoride, and computer-assisted 13C-n.m.r. analysis of the Smith-degraded polysaccharide, it was concluded that the biological repeating unit of the O39 antigen was Formula; see text

[Conformation analysis of the N-glycosylation site Asn-X-Thr/Ser in glycoproteins]. / Konformatsionnyi analiz saita N-glikozilirovaniia Asn-X-Thr/Ser v glikoproteinakh.

Theoretical conformational analysis of oligopeptides CH3CO-Asn-X-Thr-NHCH3 (X = Gly, Ala, Pro), modelling N-glycosylation site, and their glycosylated derivatives CH3CO-(GlcNAc beta 1-4GlcNAc beta 1) Asn-X-Thr-NHCH3 has been carried out. Active conformations of the site are found, corresponding to structural prerequisities of N-glycosylation: Asn residue's position in beta-turn and hydrogen bond formation between side chains of Asn and Thr/Ser residues. In this case the L conformation of the central residue X is most probable. Since Pro residue does not possess this conformation, sequences with X = Pro are not glycosylated. It is shown that glycosylation of the above-mentioned sites is accompanied by reorientation of the Asn residue's side chains.