POTASSIUM-DEPENDENT MOLECULES 229
channels. The K + channel ’ s pore domain, with the “ inverted tepee ” arrange-
ment of inner helices holding the selectivity fi lter at the extracellular mem-
brane surface, is another mostly conserved feature. However, there are
important differences between prokaryotic and eukaryotic K + channels. For
instance, in eukaryotic Shaker family Kv channels, the S6 helix on the intracel-
lular side of the selectivity fi lter has a highly conserved triplet sequence — Pro -
X - Pro — where X is any amino acid. This sequence is not present in prokaryotic
channels. This sequence has been shown, by mutation, to be important for K +
channel gating but the reasons are unknown at this time. More importantly,
as one looks down from the extracellular side of the membrane, the so - called
T1 domain, inside the cell in Shaker Kv channels, is located directly under the
pore domain entryway. Four T1 domains form a constrictive tetrameric assem-
bly at the intracellular membrane surface. This means that in Shaker Kv chan-
nels the transmembrane pore must communicate with the cytoplasm (the cell
interior) through side portals as described previously for PDB: 1EXB in refer-
ence 16. The side portals allow passage of K + ions as well as inactivation pep-
tides that are part of the K + channel ’ s N - terminal end. These peptides function
as inactivation gates in someShaker Kv channels. The T1 domains also forms
a platform for four intracellularβ subunits arranged in a tetrameric assembly
whose function is unknown at this time. The β subunit was described previ-
ously in the discussion of reference 16 and PDB: 1EXB above. A possible
function for this oxido - reductase, NADPH cofactor - containing subunit may
be to serve as a sensor for the Kv channel and as a mediator of the redox state
of the cell.
Experimentalists agree that voltage - dependent gating is fundamentally
similar for prokaryotic and eukaryotic Kv channel systems. The all - important
voltage sensor domain should be structurally similar as well. As documented
above for the prokaryote KvAP channels, the fl exible, mobile voltage sensor
domain could not be crystallized without the antibody anchoring system, and
this may have affected the orientation of the voltage sensor domain in the
KvAP system (PDB: 1ORQ and 1ORS). In crystallizing the Kv1.2 K + channel
protein reported in theScience 2005 articles (references 37 and 38), MacKin-
non and co - workers used a mixture of detergent and lipids in the crystallizing
liquid and maintained a strongly reducing environment, eliminating the need
for anchoring antibodies. In the crystals that resulted, membrane - spanning
pore and voltage sensor domains alternated with T1 and β subunit domains in
an arrangement that closely mimics a native membrane structure. The T1 and
β subunit domains appeared as rigid and well - packed structures, whereas the
pore appeared somewhat more fl exible and the voltage sensors appeared as
mobile domains making few contacts with neighboring protein domains. Some
parts of the crystalline structure were built as models with reference to previ-
ously determined structures (KvAP for instance) because electron density in
these regions was weak.
The complete Kv1.2 channel – T1 – β subunit complex is a tetramer of appro-
ximate dimensions 135 Å by 95 Å by 95 Å with pore and voltage sensors of