Download Fast-Inactivation of Voltage

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Virtual channel wikipedia , lookup

Transcript
Fast Inactivation of Voltage-Gated K+
Channels: From Cartoon to Structure
Christoph Antz and Bernd Fakler
Fast inactivation of voltage-gated potassium (Kv) channels is the best understood
gating transition in ion channels and is brought about by an NH2-terminal
domain (ball domain) of the channel’s α-subunit, which physically blocks the
open pore. Recent analysis by nuclear magnetic resonance spectroscopy
showed that ball domains from various Kv channels exhibit well-defined
but distinct structures in aqueous solution.
E
lectrical signals transferring information in
biological systems are generated by flow of
the inorganic ions Na+, K+, Ca2+, and Cl– through
large transmembrane proteins: the ion channels.
To operate sensibly, these proteins must rapidly
open and close their ion-permeable pores in
response to biological signals such as changes in
transmembrane voltage or concentration of
ligands (6). This process has been referred to as
“gating” and occurs through conformational
changes within the channel protein. “Activation
gating” leads to opening of the channel pore,
whereas “inactivation gating” results in closure
of the ion pathway. Fast inactivation of voltagegated transient outward-rectifier K+ channels (socalled A-type K+ channels) is the best understood
gating transition in ion channels. Together with
an activation in the subthreshold range, inactivation of these channels governs the firing rate
(spiking) of neurons by controlling excitability in
the interspike interval (6). This review focuses on
the fast inactivation gating of A-type K+ channels
as it developed from a phenomenological model
of channel function to a molecular understanding of the process, including the three-dimensional structure of the inactivation gate.
Definition of the inactivation gate
In the first quantitative description of Na+ and
+
K channel gating, Hodgkin and Huxley (7) postulated that multiple gating particles existed and
the time course of conductivity was explained by
the position of these particles. In voltage-dependent ion channels, gating processes are controlled and powered by the membrane potential.
On depolarization, the channels first open by
transition of the activation gate(s) and then enter a
long-living nonconducting state, the inactivated
C. Antz and B. Fakler are in the Dept. of Physiology II, Univ. of
Tübingen,Ob dem Himmelreich 7, 72074 Tübingen, Germany.
0886-1714/98 5.00 @ 1998 Int. Union Physiol. Sci./Am.Physiol. Soc.
state (Fig. 1). The latter results from transition of
the inactivation gate, which occludes the ion pore
as long as the membrane is depolarized. On
repolarization, the inactivation gate removes from
the pore and allows for the next activation of the
channel. In the Hodgkin-Huxley formalism, both
activation and inactivation are driven by the transmembrane voltage, implying that the gating apparatus must carry charges that, by its movement
during gating transitions, should generate a socalled gating current. Later on, such gating-currents were detected by Armstrong (2) in voltageclamp experiments. However, it was only during
activation of the channels that charge movement
(so-called ON-charge movement) was measured.
No component of the gating current could be
associated with the inactivation process, as if
inactivation occurred outside the membrane and
did not involve movement of any charged particle
through the transmembrane electrical field (2).
This view was further supported by the fact that
inactivation could be selectively removed by proteolytic agents (pronase or N-bromoacetamide)
when applied to the cytoplasmic side of the channels. External perfusion with the same agents did
not affect inactivation gating. Additionally, a
number of pharmacological agents [tetraethylammonium (TEA), pancuronium, or N-propylguanidinium] were found to compete with the inactivation process when applied to the intracellular
but not the extracellular face of the membrane
(2).
The findings that inactivation 1) was voltage
independent, 2) removed by proteolytic agents,
and 3) mimicked by various pharmacological
agents led Armstrong (2) and Bezanilla to propose
a novel mechanism for channel inactivation: the
“ball-and-chain” model (Fig. 1). In this model,
inactivation is brought about by an inactivation
particle (ball) that is tethered to the cytoplasmic
face of the channel protein by a protease-cleavable domain (chain). The inactivation ball binds to
its receptor, once exposed by channel activation,
News Physiol. Sci. • Volume 13 • August 1998
“. . .during activation
of the channels that
charge movement . . .
was measured.”
177
FIGURE 1. Fast inactivation in A-type K+ channels: the ball-and-chain mechanism. The cartoon represents functional states of
a channel with respect to position of inactivation particle and activation gate: C, closed or activatable state; O, open or conducting state; I , inactivated state. [Modified from Armstrong (2) and Miller (10).]
and thereby physically occludes the ion pore. Thus
the ball-and-chain model assumes strict coupling
between activation and inactivation in that only
activated or open channels can inactivate. As a
consequence, voltage dependence of inactivation
is not intrinsic to this gating transition but rather
results from the activation process.
Localization of the inactivation gate
“. . .the NH2-terminal
83 amino acids . . .
were involved in
inactivation. . . .”
178
The ball-and-chain model was proposed in the
“premolecular era” and was based exclusively on
electrophysiological and pharmacological data
from “native” channels, mainly Na+ but also K+
channels. The challenge for this model was on
when K+ channels entered the era of molecular
cloning with the isolation of the Shaker gene from
Drosophila melanogaster (14). This gene coded for
a multiplicity of voltage-gated K+ channels
through alternative splicing at the protein’s NH2
terminus. Expression of these splice variations in
Xenopus oocytes resulted in A-type K+ channels
whose time course of inactivation varied significantly and could be removed by intracellular
trypsin (Fig. 2A). These findings, together with the
presence of multiple potential trypsin cleavage
sites in the NH2 terminus of the Shaker B (ShB)
amino acid sequence, led Aldrich and co-workers
(8) to focus on this part of the K+ channel α-subunit. From a number of deletion and insertion
mutations (Fig. 2B), these authors learned that the
NH2-terminal 83 amino acids (aa) were involved
in inactivation and could be divided into two
functional domains. Mutations in the first domain,
which constitutes about the first 20 aa, slowed or
completely removed fast inactivation. Deletions in
the second domain extending from about aa 20 to
at least aa 83 speeded up inactivation, whereas
News Physiol. Sci. • Volume 13 • August 1998
insertions slowed it down (8). Although these findings were very reminiscent of ball-and-chain
domains, they did not tell too much about the
physical nature of the inactivation gate.
However, in a second set of experiments, the
same authors showed that inactivation could be
completely reconstituted in a noninactivating
mutant (ShB∆6–46, Fig. 2B) by cytoplasmic application of a synthetic peptide corresponding to aa
1–20 of ShB (15) (Fig. 2C ). This peptide-induced
inactivation was indeed specific because the rate
of inactivation linearly depended on the concentration of the applied peptide and inactivation
could not be restored by peptides whose
sequence mimicked that of noninactivating NH2terminal point mutations (15).
Moreover, NH2 terminus and the NH2-terminal peptide of ShB behaved like an open-channel
blocker as predicted by the ball-and-chain
model: 1) inactivation was almost voltage independent, 2) recovery from inactivation was
speeded up by increased external K+ concentrations just as blockade can be relieved by permeant ions (4), 3) the pore blocker TEA competes
with and prevents the inactivation gate from closing the channel, and 4) the inactivated channel
reopens on repolarization, indicating that the
inactivation gate leaves an open channel when
released from its receptor (4, 13).
This receptor is constituted at least in part by
aa residues in the internal mouth of the channel.
Isacoff et al. (9) showed that mutations in the S4S5 linker (Fig. 2D) affected both the single-channel conductance and the inactivation process.
All these aa residues, two of which were charged
(the others were hydrophobic), influenced the
stability of the channel-gate complex. This is in
line with the extensive work of Murrel-Lagnado
FIGURE 2. Localization of the inactivation gate in the primary sequence of cloned A-type K+ channels. A: single-channel openings and ensemble currents of Shaker B (ShB) channels before (control) and after application of trypsin to cytoplasmic side of
an inside-out patch. Currents were elicited by a voltage step to 20 mV after a 1-s prepulse to –100 mV; holding potential was
–70 mV. Scale bars are as indicated. B: single-channel openings of deletion mutations in the NH2-terminal region of ShB. Amino
acid (aa) sequence (given in 1-letter code) is shown at top, with the first 20 aa marked by a box; bars indicate deletions. Filled
bars, deletions that disrupted inactivation; open bars, deletions that left inactivation intact. Single-channel openings were elicited
by the same pulse protocol as in A. C: restoration of inactivation in ShB∆6-46 channels induced by cytoplasmic application of
the NH2-terminal inactivation peptides derived from ShB (top) and Raw3 (Kv3.4; bottom). Currents were recorded in response
to voltage steps to 50 mV (ShB) or to –25, 0, 25, and 50 mV (Raw3) from a holding potential of –110 mV. Peptide concentrations and scale bars are as indicated. D: ball-and-chain model after integration of structural data as obtained from site-directed
mutagenesis. [Modified from Hoshi et al. (8), Zagotta et al. (15), and Murrell-Lagnado and Aldrich (11).]
and Aldrich (11) in uncovering that inactivation
induced by the ShB-inactivation peptide is determined by two types of channel-gate interaction:
long-range electrostatic interactions, attracting
the positively charged gate to the receptor, and
hydrophobic interactions, which stabilize the
gate in the receptor-bound state.
Together, all the results obtained from the
cloned K+ channels were entirely consistent with
the ball-and-chain mechanism of inactivation. In
extension to its original version (Fig. 1C ), the
“postcloning” model (Fig. 2D) contains the information about where in the channel protein the
functional domains are localized: 1) the NH2 terminus of Shaker and mammalian voltage-gated
K+ (Kv) channels [Kv3.4 (Raw3) and Kv1.4 (RCK4)]
represents the entity of the inactivation ball
domain (8, 11, 13, 15; Fig. 2C ), 2) the chain is
formed by the rest of the NH2 terminus up to the
beginning of the transmembrane segment S1, and
3) the S4-S5 linker forms part of the ball receptor
(Fig. 2D). As a consequence of localization of the
gate in the NH2 terminus, fast inactivation is currently also known as N-type inactivation.
“. . .fast inactivation
is currently also
known as N-type
inactivation.”
Structure of inactivation gates
Although the ball-and-chain model as it
emerged from the “molecular mutagenesis
machinery” (Fig. 2D) was closer to physical reality, the questions remained, What does an inactivation gate look like? and What do we really
know without a high-resolution molecular structure? (10).
As a first step toward structural understanding
of the inactivation process, our lab analyzed all
News Physiol. Sci. • Volume 13 • August 1998
179
FIGURE 3. Structure of inactivation gates derived from the mammalian Kv channels, Kv1.4 (RCK4) and Kv3.4 (Raw3). Backbone (left) and surface structure (right) of the Kv1.4 peptide (A) and Kv3.4 peptide (B). Backbone representation shows 8th lowest-energy structures for each peptide. N and C, NH2 and COOH terminals, respectively. For Kv1.4 peptide, secondary structural
motifs are highlighted in blue (α-helix, aa 21–34) and green (β-turn type II, aa 19–21). In the surface illustration, colors indicate
the surface charge, which was scaled arbitrarily and ranged from blue (negatively charged) to red (positively charged). Both peptides are arranged with respect to their dipole vectors (red arrows).
“. . .all three peptides
were basically able
to induce
inactivation. . . .”
180
identified inactivation ball domains with highresolution nuclear magnetic resonance (NMR)
spectroscopy in aqueous solution (1). This analysis showed that the mammalian ball peptides
derived from Kv1.4 and Kv3.4 (constituted of aa
1–37 and aa 1–30, respectively) exhibited welldefined and compact structures (Fig. 3), whereas
the Shaker peptide (aa 1–20) revealed a dynamic
equilibrium of locally nonrandom structures
rather than a stable overall structure (1).
Although well defined, the structures of the
two mammalian inactivation peptides, which
share almost no homology in primary sequence,
were rather different in backbone and surface
(Fig. 3). Thus the backbone of the Kv1.4 peptide
consisted of a highly flexible NH2 terminus and
an α-helix capped by a β-turn motif, whereas the
Kv3.4 peptide showed compact folding throughout the molecule but did not exhibit any typical
News Physiol. Sci. • Volume 13 • August 1998
secondary structure (1) (Fig. 3). Despite these differences, however, the spatial distribution of
charged and hydrophobic surface domains as
well as the dipole moment were similar in both
inactivation peptides (Fig. 3, left).
These similarities and differences in structural
characteristics are in line with the functional properties observed for these peptides. On one hand,
all three peptides were basically able to induce
inactivation in noninactivating K+ channels. On
the other hand, when investigated in the same
channel, differentially structured ball peptides
revealed different functional properties: the Kv3.4derived peptide completely blocked noninactivating Kv1.1 channels, whereas the Kv1.4 peptide did
not completely inactivate these channels, even at
a 10-fold higher concentration (Fig. 4, A and B).
Thus binding to the ball receptor is obviously different for ball peptides with different structures.
FIGURE 4. Differential structural and functional properties of Kv3.4 (left) and Kv1.4 (right) inactivation gates. A: surface structure of the peptides with color coding as in Fig. 3. B: inactivation of Kv1.1 channels induced by fast application of the peptides.
Application is indicated by horizontal bar; concentration and scale bars are as indicated. C: ball-and-chain mechanism at its latest stage. Channel core represents latest version of a model of the Shaker channel originally given in Ref. 5; inactivation gate is
Kv3.4 peptide. Only 1 of the 4 ball domains is shown.
Complete understanding of this fact, however,
requires structural determination of both partners,
ball and receptor. As a consequence, the ball-and-
chain model as presented in its latest version in
Fig. 4C must still remain preliminary and awaits
completion by further structural work.
News Physiol. Sci. • Volume 13 • August 1998
181
Concluding remarks
As long as crystals of the entire channel molecule are missing, the goal for the near future is
to continue work on the functionally welldefined ball domains for which, for the first
time, structural data are available. Further goals
on that path are the structural and functional
backgrounds of the inactivation regulation by
protein phosphorylation (3) or that of inactivation transferred to noninactivating α-subunits
by an associated β-subunit (12). These pieces of
work, together with the structure of the channel’s core, will someday allow a complete and
final understanding of the inactivation process
of voltage-dependent K+ channels.
References
1. Antz, C., M. Geyer, B. Fakler, M. K. Schott, H. R. Guy, R.
Frank, J. P. Ruppersberg, and H. R. Kalbitzer. NMR-structure of inactivation gates of mammalian voltage-dependent potassium channels. Nature 385: 272–275, 1997.
2. Armstrong, C. M. Sodium channels and gating currents.
Physiol. Rev. 61: 644–683, 1981.
3. Covarrubias, M., A. Wei, L. Salkoff, and T. B. Vyas. Elimination of rapid potassium channel inactivation by phosphorylation of the inactivation gate. Neuron 13:
1403–1412, 1994.
4. Demo, S. D., and G. Yellen. The inactivation gate of the
Shaker K+ channel behaves like an open-channel
blocker. Neuron 7: 743–753, 1991.
5. Guy, H. R., and S. R. Durell. Atomic scale structure and
functional model of a voltage-gated potassium channel.
Biophys. J. 62: 238–250, 1992.
6. Hille, B. Ionic Channels of Excitable Membranes. Sunderland, MA: Sinauer, 1992.
7. Hodgkin, A. L., and A. F. Huxley. A quantitative description of membrane current and its application to conduction and excitation in nerve. J. Physiol. (Lond.) 117:
500–544, 1952.
8. Hoshi, T., W. N. Zagotta, and R. W. Aldrich. Biophysical
and molecular mechanisms of Shaker potassium channel
inactivation. Science 250: 533–538, 1990.
9. Isacoff, E. Y., Y. N. Jan, and L. Y. Jan. Putative receptor for
the cytoplasmic inactivation gate in the Shaker K+ channel. Nature 353: 86–90, 1991.
10. Miller, C. M. 1990: Annus mirabilis of potassium channels. Science 252: 1092–1096, 1991.
11. Murrell-Lagnado, R. D., and R. W. Aldrich. Interactions of
amino terminal domains of Shaker K channels with a
pore blocking site studied with synthetic peptides. J. Gen.
Physiol. 102: 949–975, 1993.
12. Rettig, J., S. H. Heinemann, F. Wunder, C. Lorra, D. N.
Parcej, J. O. Dolly, and O. Pongs. Inactivation properties
of voltage-gated K+ channels altered by presence of betasubunit. Nature 369: 289–294, 1994.
13. Ruppersberg, J. P., R. Frank, O. Pongs, and M. Stocker.
Cloned neuronal IK(A) channels reopen during recovery
from inactivation. Nature 353: 657–660, 1991.
14. Schwarz, T. L., B. L. Tempel, D. M. Papazian, Y. N. Jan,
and L. Y. Jan. Multiple potassium channel components
are produced by alternative splicing at the Shaker locus
in Drosophila. Nature 331: 137–142, 1988.
15. Zagotta, W. N., T. Hoshi, and R. W. Aldrich. Restoration of
inactivation in mutants of Shaker potassium channels by a
peptide derived from ShB. Science 250: 568–571, 1990.
Computer-Aided Design
of Thrombin Inhibitors
Amedeo Caflisch, Rudolf Wälchli, and Claus Ehrhardt
Computer-aided ligand design is an active, challenging, and multidisciplinary
research field that blends knowledge of biochemistry, physics, and computer
sciences. Whenever it is possible to experimentally determine or to model
the three-dimensional structure of a pharmacologically relevant enzyme
or receptor, computational approaches can be used to design specific
high-affinity ligands. This article describes methods, applications,
and perspectives of computer-assisted ligand design.
C
omputer-based approaches are widespread
and have applications not only in the efficient administration of already existing data but
also in the design and planning of a variety of
objects, from cars and airplanes to the exterior
A. Caflisch is in the Dept. of Biochemistry, University of
Zurich, CH-8057 Zurich, Switzerland; R. Wälchli and C.
Ehrhardt are at the Novartis Pharma, Inc., CH-4002 Basel,
Switzerland.
182
News Physiol. Sci. • Volume 13 • August 1998
and interior of buildings, as well as in the clothing industry and for many projects of the advertisement and entertainment industries. One
important application of computer science methods is the computer-aided design of ligand molecules for a given macromolecular target. For
about the past 15 years, there has been a significant development and application of computer
programs for calculation of optimal conformation(s) of molecules or (macro)molecular assem0886-1714/98 5.00 @ 1998 Int. Union Physiol. Sci./Am.Physiol. Soc.