Limits...
Caution is required in interpretation of mutations in the voltage sensing domain of voltage gated channels as evidence for gating mechanisms.

Kariev AM, Green ME - Int J Mol Sci (2015)

Bottom Line: The cavity created by the mutation has space for up to seven more water molecules than were present in wild type, which could be displaced irreversibly by the MTS reagent.These could produce the results of the experiments that have been interpreted as evidence for physical motion of the S4 segment, without physical motion of the S4 backbone.The computations strongly suggest that the interpretation of cysteine substitution reaction experiments be re-examined in the light of these considerations.

View Article: PubMed Central - PubMed

Affiliation: Department of Chemistry, City College of New York, 160 Convent Avenue, New York, NY 10031, USA. alisher@sci.ccny.cuny.edu.

ABSTRACT
The gating mechanism of voltage sensitive ion channels is generally considered to be the motion of the S4 transmembrane segment of the voltage sensing domains (VSD). The primary supporting evidence came from R → C mutations on the S4 transmembrane segment of the VSD, followed by reaction with a methanethiosulfonate (MTS) reagent. The cys side chain is -SH (reactive form -S-); the arginine side chain is much larger, leaving space big enough to accommodate the MTS sulfonate head group. The cavity created by the mutation has space for up to seven more water molecules than were present in wild type, which could be displaced irreversibly by the MTS reagent. Our quantum calculations show there is major reorientation of three aromatic residues that face into the cavity in response to proton displacement within the VSD. Two phenylalanines reorient sufficiently to shield/unshield the cysteine from the intracellular and extracellular ends, depending on the proton positions, and a tyrosine forms a hydrogen bond to the cysteine sulfur with its side chain -OH. These could produce the results of the experiments that have been interpreted as evidence for physical motion of the S4 segment, without physical motion of the S4 backbone. The computations strongly suggest that the interpretation of cysteine substitution reaction experiments be re-examined in the light of these considerations.

Show MeSH

Related in: MedlinePlus

The central part of the voltage sensing domain, side view from R297 (top) to R303 (bottom) on the S4 TM segment, and nearby residues on S1, S2 and S3. Most of the protein is shown as dark blue (N) and gray (C). Arginines 297 and 303 are light blue, tyrosine (Y266) is orange, phenylalanine F233 green; in (B,C), C300 has the sulfur as a large yellow sphere. Glutamate side chains are magenta. The water molecules are red (oxygen) and white (hydrogen). Hydrogen bonds are dashed lines. A preliminary version of this figure was posted on arXiv [38]). Parts (B,C) are optimized with HF/6-31G** from a starting position that allowed salt bridges to be maintained where possible; part (A) was similarly optimized directly from the 3Lut structure, and the changes are small.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4307324&req=5

ijms-16-01627-f001: The central part of the voltage sensing domain, side view from R297 (top) to R303 (bottom) on the S4 TM segment, and nearby residues on S1, S2 and S3. Most of the protein is shown as dark blue (N) and gray (C). Arginines 297 and 303 are light blue, tyrosine (Y266) is orange, phenylalanine F233 green; in (B,C), C300 has the sulfur as a large yellow sphere. Glutamate side chains are magenta. The water molecules are red (oxygen) and white (hydrogen). Hydrogen bonds are dashed lines. A preliminary version of this figure was posted on arXiv [38]). Parts (B,C) are optimized with HF/6-31G** from a starting position that allowed salt bridges to be maintained where possible; part (A) was similarly optimized directly from the 3Lut structure, and the changes are small.

Mentions: There has been experimental work reported that is directly relevant to those amino acids in the VSD with aromatic side chains, which are the ones that prove especially important in our results. The most work has appeared on the F233 residue of the Kv1.2/2.1 chimera (F290 of Shaker) [2,3,4]. The most important experimental evidence for the standard models comes from experiments in which an arginine is replaced with a cysteine. The cysteine can ionize, leaving an ion with a side chain consisting of just –S−; unlike the neutral –SH species, the S− can react with MTS reagents (un-ionized SH is about nine orders of magnitude slower, as the MTS reactive group is electrophilic) [5]. The R→C mutation leaves a cavity with a size essentially the difference between the single atom cysteine side chain and the very large guanidinium side chain of arginine. The cavity suffices to hold several water molecules, which in turn may be displaced if a reagent—such as MTS that can react irreversibly with the cysteine—is introduced. The MTS reagents may have various groups, positive, neutral, or negative, attached to the thiosulfonate head group. The method, originally used to study the lactose lacY transporter by Kaback and coworkers (van Iwaarden et al. [6]), was applied to ion channels (albeit ligand gated, not voltage gated, channels) by Akabas et al., who labeled it the substituted cysteine accessibility method (SCAM) [7]. They made three explicit assumptions: (1) The cysteine residue is either at the water accessible membrane surface, the lipid accessible surface, or the protein interior; (2) Hydrophilic reagents react faster at the water accessible surface; (3) Because the reaction with electrophiles like MTS reagents is so much faster with the ionized form than with the un-ionized form, and ionization is much more probable in a high dielectric medium, reaction only occurs at one of the water accessible membrane surfaces. They did not consider the possibility that an R→C, or other substitution of cysteine for a large side chain, would produce a cavity of up to roughly 200 Å3 (see Figure 1 in results), large enough for two, up to possibly seven, water molecules to enter, very possibly invalidating the third assumption. It is this point that we address here. In the original application by Karlin, Akabas, and coworkers, the amino acids for which the substitution was made were not as large as arginine, the largest being threonine, so that this issue was less significant. Horn and coworkers applied the method to voltage gated channels [8], and it has been used by multiple groups since. With voltage gated channels, the principal substitution has been for the S4 arginines, which, in the standard models of gating, are presumed to carry the gating charges as they move across the electric field of the membrane. Horn recognized that the third assumption might not be valid, but his interpretation of the results essentially requires that assumption; if the assumption is not made, the modeling of the S4 motion no longer holds. If the reaction need not be at the membrane surface, the S4 need not move during gating. So far as we are aware, none of the literature has considered the consequences of the discrepancy in size of the cysteine and arginine side chains. Not all cysteine substitution mutations leave a cavity. For example, substitution of cysteine for alanine or glycine would leave no cavity, but none of the definitive experiments involve these; as we noted, some substitutions leave only a small cavity, as in the original work of Akabas et al. [7]. Some experiments from the MacKinnon laboratory [9,10] were not limited to arginine substitutions, but their primary evidence still requires similar cautions. We will consider a specific R→C mutation, in the pdb 2A79 and 3Lut structures [11,12]. This will allow us to evaluate the consequences of an R→C mutation for the availability of the cysteine side chain to MTS reagents. A second consequence of a cysteine substitution is that salt bridges, as well as hydrogen bond networks, could be disrupted. This is very likely with an R→C mutation in the VSD, as the arginines are salt bridged to aspartates and glutamates on the other transmembrane segments, with the wild type guanidinium side chain stretching across the space that becomes the cavity in the mutant. Other amino acid substitutions that alter the arrangement of water molecules would disrupt hydrogen bonds, with uncertain consequences; other cysteine mutations may also be difficult to interpret, but we do not consider these mutations here. Several of the standard (S4 moves up) models of gating require the S4 arginines to exchange partners with the acidic residues on S2 and S3; with these salt bridges disrupted by substitutions, it introduces a further difficulty in understanding the applicability of the mutant results to the models.


Caution is required in interpretation of mutations in the voltage sensing domain of voltage gated channels as evidence for gating mechanisms.

Kariev AM, Green ME - Int J Mol Sci (2015)

The central part of the voltage sensing domain, side view from R297 (top) to R303 (bottom) on the S4 TM segment, and nearby residues on S1, S2 and S3. Most of the protein is shown as dark blue (N) and gray (C). Arginines 297 and 303 are light blue, tyrosine (Y266) is orange, phenylalanine F233 green; in (B,C), C300 has the sulfur as a large yellow sphere. Glutamate side chains are magenta. The water molecules are red (oxygen) and white (hydrogen). Hydrogen bonds are dashed lines. A preliminary version of this figure was posted on arXiv [38]). Parts (B,C) are optimized with HF/6-31G** from a starting position that allowed salt bridges to be maintained where possible; part (A) was similarly optimized directly from the 3Lut structure, and the changes are small.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4307324&req=5

ijms-16-01627-f001: The central part of the voltage sensing domain, side view from R297 (top) to R303 (bottom) on the S4 TM segment, and nearby residues on S1, S2 and S3. Most of the protein is shown as dark blue (N) and gray (C). Arginines 297 and 303 are light blue, tyrosine (Y266) is orange, phenylalanine F233 green; in (B,C), C300 has the sulfur as a large yellow sphere. Glutamate side chains are magenta. The water molecules are red (oxygen) and white (hydrogen). Hydrogen bonds are dashed lines. A preliminary version of this figure was posted on arXiv [38]). Parts (B,C) are optimized with HF/6-31G** from a starting position that allowed salt bridges to be maintained where possible; part (A) was similarly optimized directly from the 3Lut structure, and the changes are small.
Mentions: There has been experimental work reported that is directly relevant to those amino acids in the VSD with aromatic side chains, which are the ones that prove especially important in our results. The most work has appeared on the F233 residue of the Kv1.2/2.1 chimera (F290 of Shaker) [2,3,4]. The most important experimental evidence for the standard models comes from experiments in which an arginine is replaced with a cysteine. The cysteine can ionize, leaving an ion with a side chain consisting of just –S−; unlike the neutral –SH species, the S− can react with MTS reagents (un-ionized SH is about nine orders of magnitude slower, as the MTS reactive group is electrophilic) [5]. The R→C mutation leaves a cavity with a size essentially the difference between the single atom cysteine side chain and the very large guanidinium side chain of arginine. The cavity suffices to hold several water molecules, which in turn may be displaced if a reagent—such as MTS that can react irreversibly with the cysteine—is introduced. The MTS reagents may have various groups, positive, neutral, or negative, attached to the thiosulfonate head group. The method, originally used to study the lactose lacY transporter by Kaback and coworkers (van Iwaarden et al. [6]), was applied to ion channels (albeit ligand gated, not voltage gated, channels) by Akabas et al., who labeled it the substituted cysteine accessibility method (SCAM) [7]. They made three explicit assumptions: (1) The cysteine residue is either at the water accessible membrane surface, the lipid accessible surface, or the protein interior; (2) Hydrophilic reagents react faster at the water accessible surface; (3) Because the reaction with electrophiles like MTS reagents is so much faster with the ionized form than with the un-ionized form, and ionization is much more probable in a high dielectric medium, reaction only occurs at one of the water accessible membrane surfaces. They did not consider the possibility that an R→C, or other substitution of cysteine for a large side chain, would produce a cavity of up to roughly 200 Å3 (see Figure 1 in results), large enough for two, up to possibly seven, water molecules to enter, very possibly invalidating the third assumption. It is this point that we address here. In the original application by Karlin, Akabas, and coworkers, the amino acids for which the substitution was made were not as large as arginine, the largest being threonine, so that this issue was less significant. Horn and coworkers applied the method to voltage gated channels [8], and it has been used by multiple groups since. With voltage gated channels, the principal substitution has been for the S4 arginines, which, in the standard models of gating, are presumed to carry the gating charges as they move across the electric field of the membrane. Horn recognized that the third assumption might not be valid, but his interpretation of the results essentially requires that assumption; if the assumption is not made, the modeling of the S4 motion no longer holds. If the reaction need not be at the membrane surface, the S4 need not move during gating. So far as we are aware, none of the literature has considered the consequences of the discrepancy in size of the cysteine and arginine side chains. Not all cysteine substitution mutations leave a cavity. For example, substitution of cysteine for alanine or glycine would leave no cavity, but none of the definitive experiments involve these; as we noted, some substitutions leave only a small cavity, as in the original work of Akabas et al. [7]. Some experiments from the MacKinnon laboratory [9,10] were not limited to arginine substitutions, but their primary evidence still requires similar cautions. We will consider a specific R→C mutation, in the pdb 2A79 and 3Lut structures [11,12]. This will allow us to evaluate the consequences of an R→C mutation for the availability of the cysteine side chain to MTS reagents. A second consequence of a cysteine substitution is that salt bridges, as well as hydrogen bond networks, could be disrupted. This is very likely with an R→C mutation in the VSD, as the arginines are salt bridged to aspartates and glutamates on the other transmembrane segments, with the wild type guanidinium side chain stretching across the space that becomes the cavity in the mutant. Other amino acid substitutions that alter the arrangement of water molecules would disrupt hydrogen bonds, with uncertain consequences; other cysteine mutations may also be difficult to interpret, but we do not consider these mutations here. Several of the standard (S4 moves up) models of gating require the S4 arginines to exchange partners with the acidic residues on S2 and S3; with these salt bridges disrupted by substitutions, it introduces a further difficulty in understanding the applicability of the mutant results to the models.

Bottom Line: The cavity created by the mutation has space for up to seven more water molecules than were present in wild type, which could be displaced irreversibly by the MTS reagent.These could produce the results of the experiments that have been interpreted as evidence for physical motion of the S4 segment, without physical motion of the S4 backbone.The computations strongly suggest that the interpretation of cysteine substitution reaction experiments be re-examined in the light of these considerations.

View Article: PubMed Central - PubMed

Affiliation: Department of Chemistry, City College of New York, 160 Convent Avenue, New York, NY 10031, USA. alisher@sci.ccny.cuny.edu.

ABSTRACT
The gating mechanism of voltage sensitive ion channels is generally considered to be the motion of the S4 transmembrane segment of the voltage sensing domains (VSD). The primary supporting evidence came from R → C mutations on the S4 transmembrane segment of the VSD, followed by reaction with a methanethiosulfonate (MTS) reagent. The cys side chain is -SH (reactive form -S-); the arginine side chain is much larger, leaving space big enough to accommodate the MTS sulfonate head group. The cavity created by the mutation has space for up to seven more water molecules than were present in wild type, which could be displaced irreversibly by the MTS reagent. Our quantum calculations show there is major reorientation of three aromatic residues that face into the cavity in response to proton displacement within the VSD. Two phenylalanines reorient sufficiently to shield/unshield the cysteine from the intracellular and extracellular ends, depending on the proton positions, and a tyrosine forms a hydrogen bond to the cysteine sulfur with its side chain -OH. These could produce the results of the experiments that have been interpreted as evidence for physical motion of the S4 segment, without physical motion of the S4 backbone. The computations strongly suggest that the interpretation of cysteine substitution reaction experiments be re-examined in the light of these considerations.

Show MeSH
Related in: MedlinePlus