POTASSIUM-DEPENDENT MOLECULES 223
family sequences indicated that while trp - tolerant sites varied among the
species, trp - intolerant sites were highly conserved. These observations rein-
forced the suggestion that trp - tolerant sites in the Shaker K + channel are
exposed to the lipid bilayer while trp - intolerant residues make the protein –
protein contacts. In the Laine model being discussed here, the trp - tolerant
residues all point toward the hydrophobic lipid bilayer arguing for S3 ’ s posi-
tion outside of S4. In the Mackinnon KvAP model (1ORQ, 1ORS), the S3
helix is broken down into S3a and S3b sectors connected by a short loop while
S1 and S2 are modeled asα - helical segments. The secondary structure ele-
ments are in agreement for both models, it is their position within the K +
channel that differs.
Charge reversal mutations have been shown to affect maturation of the
protein and indicate positions of strong electrostatic interactions.^31 Particular
Shaker residues involved are glu283 in S2 with arg368 in S4 in an intermediate
closed pore conformation and glu283 in S2 with arg371 in S4 in the activated
(open pore) conformation. Other interactions involve lys374 in S4 with glu293
in S2 and asp316 in S3. These interactions are postulated to constrain the S2,
S3, and S4 helices into the positions that this theory advocates. Conserved
basic and acidic residues were found to form the same sort of salt bridges in
the isolated S1 – S4 voltage - sensor paddle structure of KvAP (1ORS), but the
same sort of interactions were not found in the full - length channel structure
(1ORQ). Additionally, KvAP interactions do not occur in precisely the same
register as those seen for the Shaker closed or open pore interactions. As
stated above in the discussion of the KvAP X - ray crystallographic structures,
in the 1ORS isolated voltage sensor paddle, arg133 in S4 (Shaker position
arg365) interacts with asp62 in S2 (Shaker position glu283) to bring the S2 and
S4 helices together. KvAP residue lys136 in S4 (comparable to Shaker lys374)
is within a long hydrogen bonding distance (4.1 Å ) of asp72 (comparable to
Shaker glu293), again placing S2 and S4 in proximity. KvAP residue lys136
also is within hydrogen bonding distance of glu93 in S3a (comparable to
Shaker asp316), bringing the S3a and S4 helices into close proximity as seen
in the Shaker experiments. It does not appear that these interactions per se
will argue for one S1 – S4 conformation versus the other.
Various experiments have attempted to place the S4 helix in relation to the
ion conduction pore ’ s position. Using lanthanide - based resonance energy
transfer (LRET), Bezanilla and co - workers measured a distance of 45 Å
between val363 residues in S4 helices across the tetrameric structure from each
other.^32 This places the N - terminal end of S4 near the groove between adjacent
subunits. Other experiments using tethered quaternary ammonium pore block-
ers localized S1 – S4 segments relative to the central pore.^33 These experiments
placed the extracellular ends of S1 and S3 approximately 30 Å from the extra-
cellular pore opening and the S3 – S4 linker at approximately 17 – 18 Å from the
pore opening. Formation of disulfi de bonds between cysteines substituted for
Shaker arg362 (in S4) and phe416 and ala419 (in S5) and other confi rming
experiments for the two residues ’ proximity place the S4 and S5 helices close