228 GROUP I AND II METALS IN BIOLOGICAL SYSTEMS
propose that S4 docks in the groove between the tetramer subunits with a tilt
of∼ 60 ° with the membrane plane and interacts with the S5 segment on its
counterclockwise side (viewed from outside the membrane). This view differs
from that of reference 28 and 29 researchers, who say that S4 docks in the
groove between the tetramer subunits with a tilt of∼− 15 ° and interacts with
S5 of an adjacent subunit in a clockwise manner.
The Guy open conformation model docked structure was minimized in
vacuo followed by a 1 - ns molecular dynamics simulation of the complex
embedded in a phosphatidylethanolamine (POPE) lipid bilayer. Adjustments
were made to the model, and simulations were repeated so that very little
movement occurred during the fi nal iterations. Similar methods were used to
dock the two domains in transitional and resting states. However, these results
are more tenuous as little experimental data is available. In particular, the
position of the S4 – S5 linker and its role in opening and closing the pore are
uncertain. The supplemental movie accompanying reference 36 illustrates the
open - to - close - to - open cycle resulting from the simulations.
Conclusions reached by the Guy research group include the following: (1)
A conventional helical screw model can be developed from the two KvAP
crystallographic structures (PDB: 1ORQ, 1ORS) in a manner consistent with
experimental results and theoretical constraints; (2) the model uses the open
conformation KvAP pore - forming domain from PDB: 1ORQ and uses the
voltage sensor domain from PDB: 1ORS, because these appear to be in their
native conformational folds; and (3) energetic, evolutionary, and experimental
criteria were combined to create the models they describe. Energetically, the
root - mean - square deviation (RMSD) of the model structures are low during
simulations, polar atoms (especially the charged atoms of S4 arginines) form
hydrogen bonds or salt bridges with polar atoms of water, protein, or lipid
headgroups, few hydrophobic residues are exposed to water, and most residues
have energetically favorable conformations. The energetic considerations are
in contrast with the MacKinnon group voltage sensor paddle model in which
the helix S4 charged residues are postulated to move through the hydrophobic
lipid phase of the membrane during activation.
New structures published by the MacKinnon team in 2005 eliminated the
need for the antibody attachment (Fabs) to achieve protein crystallization and
show the voltage sensors in the expected more upright position. This research
is described in the following paragraphs.
The Mackinnon group published the crystal structure of a mammalian
voltage - dependent Shaker family K + channel protein (Kv1.2) in Science maga-
zine in 2005 (PDB: 2A79).^37 In an accompanying article, the team discussed
the voltage sensor of Kv1.2 specifi cally attempting to link structural details to
electromechanical coupling in K + channels.^38 First, this study of a eukaryotic
Kv channel from rat brain, PDB: 2A79, may be compared to earlier studies of
prokaryotic cousins — KvAP, in PDB: 1ORQ, 1ORS; and KcsA, in PDB: 1BL8,
1K4C, 1K4D — because there are many similarities in their structures. The
selectivity fi lter domain is so conserved as to be essentially the same for all K +