www.sciencemag.org/cgi/content/full/328/5974/67/DC1
Supporting Online Material for
A Gating Charge Transfer Center in Voltage Sensors
Xiao Tao, Alice Lee, Walrati Limapichat
, Dennis A. Dougherty, Roderick MacKinnon*
*To whom correspondence should be addressed. E-mail: mackinn@rockefeller.edu
Published 2 April 2010,
Science
328
, 67 (2010)
DOI: 10.1126/science.1185954
This PDF file includes
Materials and Methods
Figs. S1 to S4
Table S1
References
Supporting Online Materials
Materials and Methods
Mutagenesis
All the mutants were generated using the QuikChange
site-directed mutagenesis kit (Stratagene) and
incorporation of the mutation(s) was verified by sequencing (GeneWiz).
Expression, purification, and crystallization of the F233W mutant of Kv2.1 paddle-Kv1.2
chimera channel
The F233W mutant of Kv2.1 paddle-Kv1.2 chimera channel was expressed and purified the same
way as described for the wild-type chimera channel (
1
). In brief, the channel was co-expressed with
the rat
β
2-core gene in
Pichia pastoris
, extracted with DDM (n-dodecyl-
β
-D-maltopyranoside,
Anatrace), and purified on a cobalt affinity column followed by gel filtration on a Superdex-200
column. Protein was concentrated to 10 mg/ml (Centricon-50, Millipore), mixed (0.4
μ
l + 0.4
μ
l)
with crystallization solution and crystallized us
ing the hanging drop vapor diffusion method over
reservoirs containing 0.1 ml crystallization solution
at 20°C. The crystallization solution contained
25-28% PEG400 and 50 mM TRIS-HCl pH 8.3-8.9. Cr
ystals appeared within 1-3 days and were
approximately 100*100*50
μ
m in size.
Structure determination
Crystals were directly frozen in liquid nitrogen
after overnight equilibration against a reservoir
solution containing 33% PEG400 and 50 mM TRIS-HCl
pH 8.5. Diffraction data were collected to
2.9 Å at beamline X29 (Brookhaven NSLS), images processed with DENZO and intensities merged
with SCALEPACK (
2
). Data were further processed using the CCP4 suite (
3
). The crystals are
isomorphous to the wild-type paddle chimera channel, belonging to the P4212 space group. The wild
type structural model (PDB ID 2R9R) with Phe233
replaced by alanine was used as a starting model.
Iterative refinement with CNS (keeping the same test set as wild type) (
4
) and manual rebuilding
using O (
5
) generated a final model with residual R-free of 24.7%. The final model contains:
β
2.1
residues 36-361, paddle chimera channel residues 29-417. All side chains are included in the model
with the exception of channel residues 133-144 (the link between T1 and the transmembrane region)
which was modeled as poly-glycine. Crystallographic data and refinement statistics are shown in
supplementary information (Table S1). Figures were made using PYMOL (
www.pymol.org
) (
6
).
Shaker K+ channel expression
The Shaker H4 (inactivation removed) construct in a BlueScript vector was used for Shaker K+
channel expression in
Xenopus
oocytes (
7
). The N-type inactivation gate (corresponding to amino
acids 6-46) was not included in the construct (
8
). cRNA was prepared from HindIII linearized
plasmid using T7 RNA polymerase (Promega).
Xenopus
oocytes were harvested from mature female
Xenopus laevis
and defolliculated by
collagenase treatment for 1-2 hours. Oocytes were then rinsed thoroughly and stored in ND96
solution (96 mM NaCl, 2 mM KCl, 1.8 mM
CaCl2, 1.0 mM MgCl2, 5 mM HEPES, 50 μg/ml
gentamycin, pH 7.6 with NaOH). Defolliculated oocytes were selected 2-4 hours after collagenase
treatment and injected with cRNA the next day. The injected oocytes were incubated in ND96
solution before recording. Recordings were usually done 1-2 days post-injection for ionic current
measurements and 3-6 days post-injection for gating current measurements. All oocytes were stored
in an incubator at 18 °C.
Unnatural amino acid incorporation
Unnatural amino acids were incorporated into the Shaker K+ channel using the nonsense codon
suppression method (
9
). THG73 was used as the amber suppressor tRNA (
10
). The preparations of
amino acids coupled to the dinucleotide (dCA) and the ligation of the conjugated dCA-amino acid
have been described previously (
9
). Crude tRNA-amino acid product was used without desalting,
and the product was confirmed by MALDI-TOF MS on 3-hydroxypicolinic acid (3-HPA) matrix.
Deprotection of the NVOC group on tRNA-amino acid was carried out by 10-minute photolysis
immediately prior to injection. Equal volumes of the Shaker cRNA (in which the codon for Phe was
replaced by the amber stop codon) and unprotected tRNA-amino acid were mixed prior to injection.
Approximately 15 ng of tRNA was used per oocyte. As a negative control, 76-nucleotide tRNA
(dCA ligated to 74-nucleotide tRNA) was co-injected with cRNA in the same manner as fully
charged tRNA.
sampled at 10kHz.
Electrophysiological recordings
All recordings were performed at room temperature in two-electrode voltage-clamp configuration
with an oocyte clamp amplifier (OC-725C, Warner Instrument Corp.), Digidata 1440A
analogue-to-digital converter interfaced with a computer, and pClamp10.1 software (Axon
Instruments, Inc) for controlling membrane voltage and data acquisition. The recorded signal was
filtered at 1kHz and
To investigate voltage-dependent channel activation, oocytes were held at -80 mV (Shaker wt and
most of the mutants) or -110 mV [F
→
W i.e. R1K5(W) mutant] with pulse potential starting from
holding potential ending between +30 mV and +180 mV in 10 mV, 5 mV or 2.5 mV increments. The
repolarization potentials were either more negative to the voltage at which channel starts to open (for
most mutants) or slightly positive to that voltage (for mutants with very fast closure rate). To
investigate voltage-dependent channel closure, oocytes were held at -80 mV (Shaker wt and most
other mutants) or -110 mV [R1K5(W) mutant], depolarized to between 0 mV and +80 mV with
repolarization potentials starting from the depolarization potential ending between -100 mV and -140
mV in 10mV decrements. Recording solution for the above experiments contained 98 mM KCl, 0.3
mM CaCl2, 1 mM MgCl2, and 5 mM HEPES pH 7.6.
For capacitive current measurement, immediately prior to recording, the oocytes were incubated in
ND96 recording solution (96 mM NaCl, 2 mM KCl, 0.3 mM CaCl2, 1 mM MgCl2, and 5 mM
HEPES pH 7.6) plus 50 to 100
μ
M Agitoxin2 for 5-10 minutes with gentle rocking. During
recording, holding and repolarization potentials were between -80 mV and -110 mV with pulse
potentials starting between -160 mV and -200 mV ending between +60 mV and +120 mV.
Data analysis
No leak or capacitive current was subtracted from the current traces of voltage-dependent channel
activation and channel closure shown in Figures 2, 3 and 4. For voltage-dependent channel activation
recordings, the amount of current at the repolarization step, typically measured 4-5 ms after the
depolarization step when most of the capacitive
current has relaxed, was normalized against the
maximal current (
I
/
I
max
) and plotted as a function of the depolarization voltage (
I
-
V
plot). This
voltage-dependent activation plot was fitted with the two-state Boltzmann function:
()
max
1
ZF
1exp
RT
m
I
I
VV
=
⎛⎞
+− −
⎜⎟
⎝⎠
where
I
/
I
max
is the fraction of the maximal current,
V
is the depolarization voltage to open the
channels,
V
m
is the voltage at which the channels have reached 50% of their maximal current, F is the
Faraday’s constant, R is the gas constant, T is the absolute temperature, and Z is the apparent valence
of voltage dependence. Note that
I/I
max
does not represent the true open probability (P
o
) of the
channel, given that the maxium P
o
of Shaker wt channel in whole oocytes is less than 1.0 (
11
).
For the transient (capacitive plus gating) current
traces shown in Figure 5A-D, Agitoxin2-insensitive
currents were subtracted separately for the three st
eps of the recording: (i) hyperpolarization, (ii)
depolarization, and (iii) repolarization. For each step
, the leak current was defined as the current
measured near the end of the step. The total gating charges, calculated by integrating the
repolarization-induced transient currents over time and subtracting the linear component due to the
linear capacitance of the cell and the voltage clamp system, are plotted against the depolarization
voltage (
Q
-
V
plot).
All statistical fits and figure plotting were done using Clampfit 10.1 (Axon Instruments, Inc) or Igor
Pro 6.03A2 (WaveMetrics, Inc).
Theoretical analysis of voltage sensor states
The theoretical curves in Figure 6 were produced as solutions to the following expressions describing
the model of voltage sensor states (Fig. 6A). Forward and backward voltage-dependent rate constants
are given by equations (1) and (2), respectively.
(1)
(2)
These equations partition one quarter of the total gating charge per voltage sensor (
z
) to each
transition with one eighth of the total in the forward and one eighth in the backward reaction. The
constants R, T, and F are the gas constant, absolute temperature and Faraday’s constant, respectively.
At membrane voltage
v
the probability
P
i that a voltage sensor is in state i = 1 to 5 was calculated
with equations (3) – (8).
(3)
(4)
(5)
(6)
(7)
(8)
The gating charge curves (
Q
-
V
curves, Fig. 6, C and D solid curves) were calculated with equation
(9)
(9)
and the current curves (
I
V
curves, Fig. 6, C and D dashed
curves) were cal
culated with
quation (10), which approximate
s channel opening as a power fu
nction of voltage sensor state
.
e
5
(10)
he gating current time course curves (Fig. 6B) were calculated
by first solving differential
quations (11) – (15)
T
e
(11)
(12)
(13)
(14)
(15)
with initial values given by equations (3) – (7) and v = membra
ne voltage prior to the voltage
step. The gating current in unit
s of elementary charge units pe
r second as a function of time
as then calculated with equatio
n (16), which describes the tim
e‐dependent flux of voltage
ensor states in the forward direction.
w
s
(16)
The total gating charge per voltage sensor (z) was assigned the value 3.5 elementary charge units
(
12
). For the R1K5(W) channel (Fig. 6, B and C) zero-voltage rate constants were
k
12(0)=
k
23(0)=
k
34(0)=1500 sec-1,
k
21(0)=
k
32(0)=
k
43(0)=100 sec-1,
k
45(0)=150 sec-1,
k
54(0)=10 sec-1. For the
R1R5(W) channel (Fig. 6, B and D) zero-voltage rate constants were the same except
k
54(0)=1000
sec-1. All equations were solved using Maple 12.0 (Maplesoft). The differential equations were
solved numerically using the Runge-Kutta Fehlberg method to fifth order accuracy.
S
5
Fig. S1.
The voltage-dependent chan
nel activation curves are
shown for Phe (Shaker wt)
and Phe to 3,5-F-Phe (F
2
Phe), 4-bromo-Phe (BrPhe), 4-cyano-Phe (CNPhe), 4-methyl-
Phe (MePhe), Trp mutants. Frac
tion of the maximal current (I/I
max
, mean
±
s.e.m.) is
plotted as a function of the depolarization vo
ltage and fitted with the two-state Boltzmann
function (see methods, Phe, n = 11; F
2
Phe, n = 14; BrPhe, n= 10; CNPhe, n = 5; MePhe,
n= 6 and Trp, n =9). The cation-
π
binding energy in kcal/mo
l: Trp -32.6, MePhe -28.5,
BrPhe -27.6, Phe -27.1, F
2
Phe -17.1 and CNPhe -15.7 (
13
). More negative binding
energy means stronger cation-
π
interaction. Chemical structures of the side chains are
shown below the I-V plot.
Fig. S3. Measurement of gating currents.
(A-B)
Transient currents measured from an
uninjected oocyte (A) and an
oocyte expressing the Shaker R1R5(W) mutant channel
after RNA injection (B).
Ionic currents were bloc
ked using at least 50
μ
M Agitoxin2.
(C-
D)
Q-V plots, in which integration of the repolarization-induced transient currents over
time is plotted as a function of the pulse pot
ential, are shown for an uninjected oocyte (C)
and an oocyte expressing the Shak
er R1R5(W) mutant channel (D).
The linear
component (dashed lines) of the Q-V plot reflects the charge required to bring the
membrane voltage to its new value. The non
linear gating charge component reflects the
displacement of charged amino acids in the voltage sensor.
S
6
Fig. S4. Implicated voltage sensor motion.
K5 binds in the occluded binding site when
the voltage sensor is in its depolarized conformation in the crystal structures of Kvchim
wild type and F233W mutant.
The electrophysiological data s
uggest that Lys at position 1
binds in the occluded site in the hyperpolar
ized conformation. This would imply a 21 Å
(
α
-carbon distance between R1 and K5 along the S4 helix, corresponding to 18 Å
perpendicular to the membrane) distance over
which the S4 charged residues move across
the membrane associated with gating. The vo
ltage sensor of Kvchim
(PDB ID 2R9R) is
shown as in Figure 1C.
S
7
S
8
Table S1 Crystallographic data and refinement statistics
Data collection
Space group
P42(1)2
Lattice constants (Å)
a = b = 144.271, c = 284.060
α
=
β
=
γ
= 90
°
Source
BNL X29
Wavelength (Å)
1.0809
Resolution (Å)
50-2.9
Total / unique observations
560,259 / 67,020
I / sigma (I)
a
17.6
(1.7)
Redundancy
8.4 (6.6)
Completeness (%)
99.5 (96.2)
R
sym
(%)
b
12.9 (67.5)
Model refinement
Resolution (Å)
50-2.9
Number of reflections
66,928 (3,228)
R
work
/ R
free
21.1 / 24.7
R.m.s. deviation of bond length (Å)
0.007
R.m.s. deviation of bond angles (
°
)
1.2
R.m.s., root mean-squared.
a
Number in the parentheses represents statistics for data in the highest resolution shell,
same for redundancy, completeness, and R
sym
.
b
R
sym
=
Σ
|
I
i
– <
I
i
>| /
Σ
I
i
, where <
I
i
> is the average intensity of symmetry equivalent
reflections.
S
9
References
1.
S. B. Long, X. Tao, E. B. Campbell, R. Mackinnon,
Nature
450, 376 (Nov 15,
2007).
2.
Z. Otwinowski, W. Minor,
Methods Enzymol.
276, 307 (1997).
3.
N. Collaborative Computational Project,
Acta Cryst.
D50, 760 (1994).
4.
A. T. Brunger
et al.
,
Acta Cryst.
D54, 905 (1998).
5.
T. A. Jones, J. Y. Zou, S. W. Cowan, M. Kjeldgaard,
Acta Cryst.
A47, 110
(1991).
6.
W. L. DeLano,
DeLano Scientific, Palo Alto,
CA, USA. http://www.pymol.org
,
(2002).
7.
A. Kamb, J. Tseng-Crank, M. A. Tanouye,
Neuron
1, 421 (Jul, 1988).
8.
T. Hoshi, W. N. Zagotta, R. W. Aldrich,
Science
250, 533 (Oct 26, 1990).
9.
M. W. Nowak
et al.
,
Methods Enzymol
293, 504 (1998).
10.
M. E. Saks
et al.
,
J Biol Chem
271, 23169 (Sep 20, 1996).
11.
D. Schmidt, R. MacKinnon,
Proc Natl Acad Sci U S A
105, 19276 (Dec 9,
2008).
12.
S. K. Aggarwal, R. MacKinnon,
Neuron
16, 1169 (Jun, 1996).
1
3.
S. Mecozzi, A. P. West, Jr., D. A. Dougherty,
Proc Natl Acad Sci U S A
93,
10566 (Oct 1, 1996).