POTASSIUM-DEPENDENT MOLECULES 231
Kv1.2) for their job of attracting cations. Inactivation peptides (also called an
inactivation gate) are believed to exist either on the N - terminus of an intracel-
lular domain or on the N - terminal end of an associated β subunit. In either
case, inactivation peptides have (a) a hydrophobic region that can reach into
the inner pore to plug it and (b) a positively charged hydrophilic region that
is attracted to hydrophilic and negatively charged regions on the T1 side
portal. Figure 4B of reference 37 shows a model inactivation peptide reaching
into a Kv1.2 side portal. Questions remain as to the role of the β subunit
domains ofShaker Kv channels. As discussed above, they may confer inactiva-
tion. It has also been observed that β subunits can infl uence levels of channel
expression, leading to the hypothesis that they may be channel chaperones.
Conserved catalytic residues for hydride transfer leads to another hypothesis,
namely, that β subunits have a catalytic function. Perhaps the Kv channel regu-
lates the activity of theβ subunits; conversely, and perhaps more likely, the β
subunits regulate the activity of the Kv channel. In the Kv1.2 structure reported
here (PDB: 2A79), the NADP + cofactor is present in the β subunit active cleft
along with diffuse electron density that possibly represents either a large
organic molecule or polypeptide chain. More defi nitive experiments or struc-
tural detail are necessary before the true purpose of theseβ subunit features
will be known.
In summation for the work of reference 37: (1) Important factors for obtain-
ing crystals and determining structure for Kv channel proteins are the pres-
ence of lipids and detergent during crystallization as well as the maintenance
of a nonoxidizing atmosphere; (2) relationships among integral membrane
components, the intracellular T1 domain, and the β subunit bound to T1 are
consistent with other electrophysiological studies of inactivation processes in
Shaker K + channel proteins; (3) the ion conduction pore structure of the mam-
malian Kv1.2 studied here is similar to that of prokaryotic K + channels; (4)
curved or bent inner, S6, helices result from the amino acid sequence Pro – X –
Pro in Shaker K + channel proteins and from a glycine residue in KvAP and
probably other Kv channels; (5) the Kv1.2 ion conduction pore is in an open
conformation in the PDB: 2A79 structure; (6) 15 - to 20 - Å - wide side portals
between the T1 domain and the ion conduction pore connect the pore to the
cytoplasm allowing passage of K + , other larger cations, and inactivation pep-
tides attracted by negatively charged glutamate and aspartate residues on the
portal ’ s surface; (7) additional experiments are necessary to test the infl uence
of theβ subunit ’ s NADP + cofactor, active site cleft, and probable resident
substrate molecule of unknown character on K + channel protein activity; and
(8) the voltage sensor domain of Kv1.2 resides wholly in the membrane layer
and appears unperturbed by the crystallization process.
In an accompanying article in Science (reference 38) Mackinnon and co -
workers continue their study of the Kv1.2 voltage sensor (PDB: 2A79), trying
to answer the following questions: (1) How do gating charges move within the
membrane electric fi eld? and (2) How do the gating charge movements couple
to opening and closing of the ion conduction pore? Any model must account