Apoptotic surge of potassium currents is mediated
by p38 phosphorylation of Kv2.1
Patrick T. Redman*, Kai He*, Karen A. Hartnett*, Bahiyya S. Jefferson*, Linda Hu†, Paul A. Rosenberg†,
Edwin S. Levitan‡, and Elias Aizenman*§
Departments of *Neurobiology and‡Pharmacology, University of Pittsburgh School of Medicine, Pittsburgh, PA 15261; and†Department of Neurology
and Program in Neuroscience, Children’s Hospital and Harvard Medical School, Boston, MA 02115
Edited by Michael V. L. Bennett, Albert Einstein College of Medicine, Bronx, NY, and approved January 4, 2007 (received for review November 15, 2006)
Kv2.1, the primary delayed rectifying potassium channel in neu-
have described Kv2.1 phosphorylation events affecting channel
gating and the impact of this process on cellular excitability. Kv2.1,
however, also provides the critical exit route for potassium ions
during neuronal apoptosis via p38 MAPK-dependent membrane
insertion, resulting in a pronounced enhancement of K?currents.
mutants identify a p38 phosphorylation site at Ser-800 (S800) that
death. In addition, a phospho-specific antibody for S800 detects a
p38-dependent increase in Kv2.1 phosphorylation in apoptotic
neurons and reveals phosphorylation of S800 in immunopurified
channels incubated with active p38. Consequently, phosphoryla-
tion of Kv2.1 residue S800 by p38 leads to trafficking and mem-
brane insertion during apoptosis, and remarkably, the absence of
apoptosis ? ion channel ? MAPK
Blocking K?channels, or increasing the extracellular K?con-
centration, effectively attenuate cell death in many apoptotic
models (2–6), including oxidant exposure in cortical and mid-
brain dopaminergic neurons (7–10). With the use of dominant
negative mutant subunits, Kv2.1, the major component of the
delayed rectifier K?current in neurons (11, 12), was identified
as the channel responsible for mediating the apoptotic K?
current surge in cortical neurons (13). During oxidant-induced
neuronal apoptosis, the liberation of intracellular Zn2?from
metal binding proteins (7) leads to the activation of p38 MAPK,
which precedes and is necessary for the characteristic K?current
surge (8, 9). This enhancement of K?currents is caused by a
soluble N-ethylmaleimide-sensitive factor attachment protein
receptor-dependent insertion of new Kv2.1-encoded channels,
rather than an alteration in the properties of existing surface
channels, such as a change in activation kinetics (3, 14). Despite
this information, the molecular process connecting p38 activa-
tion and the apoptotic membrane insertion of Kv2.1 K?channels
had heretofore remained undefined.
Here, we combine several experimental approaches to estab-
lish a direct link between active p38 and Kv2.1 during apoptosis
after a sequence-based prediction model (15) was used to
identify a putative phosphorylation site for the MAPK on the C
terminal of the channel. First, alanine substitution of Ser-800
(S800) in Kv2.1 completely abolished the apoptotic enhance-
ment of K?currents. Second, a cysteine-containing mutant of
Kv2.1 and a thiol-reactive covalent inhibitor were used to
demonstrate that S800 is critical for membrane insertion of the
channel during apoptosis. Third, expression of phospho-mimetic
mutant channels resulted in significant increases in basal K?
current densities. Fourth, a phospho-specific antibody directed
at S800 detected a p38-dependent increase in phospho-Kv2.1
nhancement of voltage-gated K?channel activity, resulting
in K?efflux, is an essential step in neuronal apoptosis (1).
levels in apoptotic neurons and revealed phosphorylation of
immunopurified Kv2.1, but not of Kv2.1(S800A), incubated with
active p38. Most significantly, Kv2.1(S800A) did not support
apoptosis mediated by the WT channel in a recombinant ex-
pression system. This study establishes that p38-mediated phos-
phorylation of Kv2.1 is necessary and sufficient for its apoptotic
trafficking and completion of the cell death program.
S800 in Kv2.1 Is Necessary for the Apoptotic Surge in K?Currents.
Scansite (http://scansite.mit.edu), a program that predicts pro-
tein phosphorylation sites based on proteomic and biochemical
data (15), was used to search for potential MAPK targets in
Kv2.1 (GenBank accession no. NM?013186). A medium-
stringency scan revealed a single serine residue at position 800
(Fig. 1A) as a potential phosphorylation target for both p38 and
ERK. However, the Scansite score for this sequence suggested
a substantially better match for p38 than for ERK. S800 resides
in the serine- and threonine-rich intracellular C-terminal tail of
the protein (Fig. 1A), where other phosphorylation sites exist
(16–18). We thus examined the role of S800 in the up-regulation
of K?currents during oxidant-induced apoptosis in a recombi-
nant expression system.
Whole-cell electrophysiological recordings were performed
on CHO cells transiently expressing either WT Kv2.1 channels
or a nonphosphorylatable mutant, Kv2.1(S800A). CHO cells
have no endogenous voltage-gated K?channels (19), but can be
induced to readily undergo oxidant-induced apoptosis after
expressing Kv2.1 (13), and, like neurons, show a pronounced K?
current enhancement during this process (14). Recordings were
obtained under control conditions and after treatment with the
oxidant apoptogen 2,2?-dithiodipyridine (DTDP; refs. 7 and 14).
Electrophysiological measurements were routinely performed
3 h after oxidant exposure, a time when a robust K?current
surge is well established (8, 13, 14). As shown (14), currents
mediated by WT Kv2.1-encoded channels were substantially
enhanced after the apoptotic stimulus (Fig. 1B Left). In contrast,
no current surge was observed in CHO cells expressing
Kv2.1(S800A) after the oxidant treatment (Fig. 1B Right), indi-
cating that S800, a putative target for p38-mediated phosphor-
ylation, is required for the increase in K?currents during
Author contributions: P.T.R. and K.H. contributed equally to this work; P.T.R., K.H., P.A.R.,
E.S.L., and E.A. designed research; P.T.R., K.H., K.A.H., and B.S.J. performed research; L.H.
and P.A.R. contributed new reagents/analytic tools; P.T.R., K.H., K.A.H., B.S.J., and E.A.
analyzed data; and P.T.R., P.A.R., E.S.L., and E.A. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS direct submission.
§To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
This article contains supporting information online at www.pnas.org/cgi/content/full/
© 2007 by The National Academy of Sciences of the USA
February 27, 2007 ?
vol. 104 ?
The apoptotic surge in K?current observed in both cortical
neurons and CHO cells is the result of trafficking and soluble
N-ethylmaleimide-sensitive factor attachment protein receptor-
dependent exocytotic insertion of new Kv2.1 channels into the
plasma membrane (14). New channel membrane insertion was
detected by using the cysteine-containing K?channel mutant
Kv2.1(I379C) (20). The presence of this mutation allows the
thiol reagent (2-trimethylammoniumethyl) methanethiosulfate
(MTSET) (21, 22). Virtually all previously existing surface
Kv2.1(I379C) channels can therefore be silenced after a 5- to
10-min exposure to MTSET. With existing plasma membrane
Kv2.1 channels permanently blocked, the insertion of new
channels is easily detected electrophysiologically by the appear-
ance of new currents under MTSET-free conditions (14). This
assay, which has the distinct advantage of detecting the insertion
of new functional channels, has been validated by parallel
biotinylation studies (14).
To investigate the role of S800 in the apoptotic insertion of
new channels, the double mutant Kv2.1(I379C, S800A) was
generated. Whole-cell recordings first confirmed the preserva-
tion of functional silencing by MTSET in this double mutant
channel (Fig. 2A). Next, CHO cells expressing either
Kv2.1(I379C) or Kv2.1(I379C, S800A) were exposed to MTSET
for 10 min to block all existing surface channels. The cells were
then exposed to oxidant to trigger the apoptotic cascade. Elec-
trophysiological recordings 3 h later revealed the presence of K?
currents in Kv2.1(I379C)-expressing cells (Fig. 2B), similar to
what was reported earlier (14). However, little or no current
could be measured in CHO cells expressing Kv2.1(I379C,
S800A), indicating a lack of apoptotic membrane insertion of
these mutant channels (Fig. 2B). This result not only accounts
for the lack of Kv2.1(S800A)-mediated apoptotic current surge
observed earlier, but demonstrates that phosphorylation of S800
is necessary for Kv2.1 membrane insertion during the cell death
Negatively Charged Amino Acids at Position 800 Mimic the Apoptotic
Surge. Recordings were performed from CHO cells transiently
expressing two additional variants of Kv2.1 to test whether
substitution of S800 with negatively charged residues would
mimic the apoptotic increase in currents observed in WT
channels. Indeed, basal current densities in untreated
Kv2.1(S800D) and Kv2.1(S800E) mutant-expressing cells dis-
played significantly enhanced amplitudes, comparable to those
observed in DTDP-treated Kv2.1-expressing cells (Fig. 3). Ox-
idant treatment of cells expressing these mutations did not
produce any additional current enhancement (data not shown).
Rather, a slight decrease in the currents was observed under
these conditions, likely the result of cellular damage by the
injurious stimulus. These results further support the role of S800
phosphorylation as a key mediator of the K?current surge
site on the cytoplasmic C terminus. (B) (Upper Left) Representative whole-cell
treatment conditions. Currents were obtained 48 h posttransfection and
densities from Kv2.1-expressing CHO cells under control (n ? 25) and DTDP
(n ? 12) treatment conditions (*, P ? 0.05; t test). Currents were induced with
a voltage step to ?5 mV from a holding potential of ?70 mV and normalized
to cell capacitance. (Upper Right) Representative whole-cell K?currents from
Kv2.1(S800A)-expressing CHO cells recorded under control and DTDP treat-
ment conditions. (Lower Right) Mean ? SEM. K?current densities from
Kv2.1(S800A)-expressing CHO cells recorded under control (n ? 24) and DTDP
(n ? 24) treatment conditions.
Ser-800 of Kv2.1 is critical for apoptosis-associated K?current en-
Kv2.1. (A) MTSET blocks Kv2.1(I379C) channel containing the Kv2.1(S800A)
mutation. Calibration was 2 nA, 10 ms. (B) (Upper) Diagram representing the
experimental protocol designed to measure the appearance of new K?chan-
nels during the apoptotic process. Shown are representative whole-cell K?
currents obtained from Kv2.1(I379C)- and Kv2.1(I379C, S800A)- expressing
of ?70 mV. Calibration was 1 nA, 10 ms. (Lower) Mean ? SEM. K?current
densities recorded from Kv2.1(I379C)-expressing CHO cells (n ? 17) and
Kv2.1(I379C, S800A)-expressing CHO cells (n ? 20) 3 h after the sequential
MTSET/DTDP treatments (*, P ? 0.01; t test). These currents represent the
newly inserted channels during apoptosis.
S800A mutation blocks apoptogen-induced membrane insertion of
Redman et al.
February 27, 2007 ?
vol. 104 ?
no. 9 ?
S800 Phosphorylation Is p38-Dependent in Neurons Undergoing Apo-
ptosis. Biochemical studies were performed to detect phospho-
S800 levels in primary cortical neurons after oxidative injury.
Cell lysates were subjected to gel electrophoresis, and the
resulting blots were probed with a commercially available anti-
body directed against amino acids 841–857 of the C terminus of
Kv2.1 (Kv2.1; Alomone Labs, Jerusalem, Israel) or with an
antibody that we generated against the phosphorylated form of
S800 [pKv2.1; supporting information (SI) Fig. 6]. Immunoblots
probed with Kv2.1-directed antibodies revealed the presence of
two major bands in cortical extracts: a diffuse band near 100 kDa
and another, more compact band, at ?80 kDa (SI Fig. 7; see also
refs. 23–25). The diffuse band near 100 kDa likely represents
multiple, non-p38 dependent phosphorylation states of the
predicted full-length channel (18, 26), whereas the 80-kDa band
may reflect the presence of a protein generated by an alternative
mRNA isoform (23, 27) or a proteolytic fragment. Importantly,
preincubation of the two antibodies with their respective immu-
nizing peptides completely blocked both immunoreactive signals
(SI Fig. 7).
Phospho-S800 levels in cortical neuronal lysates were assessed
by using the pKv2.1 antibody under control conditions and after
oxidative injury. We observed a significant increase in pKv2.1
immunoreactivity in cortical neurons undergoing apoptosis,
compared with vehicle-treated controls (Fig. 4 A and B). This
increase was particularly apparent in the compact, 80-kDa band.
Importantly, SB-239063, a selective p38 inhibitor (8, 28), pre-
vented oxidant-mediated increases in pKv2.1 immunoreactivity
(Fig. 4 A and B). These data confirm a p38-dependent increase
in phospho-S800 levels in endogenous Kv2.1 channels from
cortical neurons undergoing apoptosis.
Active p38 Phosphorylates S800 in Cell-Free Assays. To determine
whether p38 could phosphorylate Kv2.1, we performed a cell-
free assay in which recombinantly expressed, immunoprecipi-
tated poly myc-tagged WT Kv2.1 (myc-Kv2.1; refs. 14 and 29)
was used as a substrate for purified, active p38. We observed that
myc-Kv2.1 protein exposed to active p38 produced a phospho-
specific immunoreactive band (Fig. 4C). Similar experiments
performed with an S800A mutant of the myc-tagged Kv2.1
yielded no phospho-specific signal (Fig. 4C). These results
indicate that active p38 can directly induce selective phosphor-
ylation of residue S800 of Kv2.1.
S800A Mutation Disrupts Kv2.1-Mediated Apoptosis. Unlike trun-
cated or pore mutants of the channel (13), Kv2.1(S800A) does
not function as a dominant interfering form in neurons under-
going apoptosis, but simply enhances the overall basal current
amplitude without eliminating the contribution of endogenous
channels (data not shown). As such, the requirement for S800
phosphorylation in apoptosis was evaluated in Kv2.1-expressing
CHO cells under oxidant exposure conditions that are normally
sublethal to vector-expressing control cells (13). A 15-min
exposure to 30 ?M DTDP was sufficient to induce apoptosis in
?50% of WT Kv2.1-expressing cells (Fig. 5). However, the same
oxidative insult was completely innocuous to CHO cells express-
ing Kv2.1(S800A) mutant channels. As a negative control, we
assayed the viability of CHO cells expressing a nonconducting,
double-pore mutant Kv2.1(W365C, Y380T) channel (12). As in
cells expressing Kv2.1(S800A), the oxidant was not toxic to cells
expressing the double-pore mutant channels. These data dem-
onstrate the requirement for functional Kv2.1 channels contain-
ing the phosphorylatable amino acid residue S800 to complete
the apoptotic program.
The results presented in this study reveal an important and
previously unrecognized mechanism of Kv2.1 regulation that has
a critical impact on apoptosis. Kv2.1 is subject to extensive
phosphorylation and dephosphorylation reactions, which alter
from Kv2.1-expressing CHO cells (Left), Kv2.1(S800D)-expressing CHO cells
(Center), and Kv2.1(S800E)expressing CHO cells (Right). Currents were ob-
?35 mV from a holding potential of ?70 mV. Calibration was 5 nA, 15 ms.
(Lower) Mean ? SEM. Current densities from Kv2.1-expressing CHO cells (n ?
13) (Left), Kv2.1(S800D)-expressing CHO cells (n ? 17) (Center), and
Kv2.1(S800E)-expressing CHO cells (n ? 14) (Right) (*, P ? 0.05; t test).
Substitution of S800 with negatively charged amino acids mimics
of Kv2.1-S800 is regulated by p38 MAPK in cultured cortical neurons. Cortical
cultures were pretreated with either vehicle or the specific p38 MAPK inhib-
itor SB293063 (SB; 20 ?M for 10 min) and then incubated with either vehicle
or DTDP (100 ?M for 10 min). Immunoblots were probed with anti-pKv2.1
antibody (1:5,000; Top), a commercially available anti-Kv2.1 antibody
(Alomone Labs; 1:1,000; Middle), and an anti-GAPDH antibody (Novus Bio-
logicals, Littleton, CO; 1:5,000; Bottom) to confirm equal protein loading
between lanes. These data are representative of results from three indepen-
dent experiments. (B) Oxidative injury significantly increases S800 phosphor-
ylation in cortical neurons in vitro. Optical density measurements of the
the 80-kDa band were quantified by using Scion Image software. Values
represent the mean ? SEM normalized pKv2.1 signal (to Kv2.1 immunoreac-
tivity) of both bands as a percentage of vehicle-treated control from three
separate, independent experiments (*, P ? 0.01; ANOVA/Dunnett). Note that
the observed increase in S800 phosphorylation was effectively blocked with a
p38 MAPK inhibitor. (C) S800 of Kv2.1 is phosphorylated by p38? in a cell-free
mM MgCl and 50 ?M ATP), and 50 ?g of activated p38? kinase. Reactions
containing kinase buffer and Mg/ATP and kinase buffer alone were used as
controls. Immunoblots (IB) were probed with the pKv2.1 antibody (1:12,000)
and the Kv2.1 antibody (1:3,000; Alomone Labs).
Phosphorylation of Ku2.1 by p38. (A) DTDP-induced phosphorylation
www.pnas.org?cgi?doi?10.1073?pnas.0610159104Redman et al.
the functional properties of this channel by influencing gating,
and thereby, cell firing properties (18, 30). We have found,
however, a phosphorylation target for p38 on the C-terminal of
Kv2.1 that is vital for the apoptotic surge of K?currents
observed in apoptosis and the completion of the cell death
program. Interestingly, a recent report by Park et al. (18)
reported that S800 is a phosphorylation target, albeit by a
then-unidentified kinase. They further showed that S800 was not
one of several C-terminal sites subject to calcineurin-dependent
dephosphorylation and did not participate in the graded regu-
lation of Kv2.1 gating. We have found here that S800 is a unique
p38 phosphorylation target on Kv2.1 that is required for channel
trafficking and cell death.
Phosphorylation of S800 is not necessary, in and of itself, for
normal Kv2.1 trafficking and plasma membrane insertion, as
Kv2.1(S800A) mutant channels can be functionally expressed.
Indeed, normal neurons overexpressing this construct simply
have larger K?currents, which explains why this mutated form
of Kv2.1 cannot function as a dominant interfering channel
during neuronal apoptosis. Nonetheless, our studies in CHO
cells reveal that under injurious situations this site is critical for
de novo channel insertion, resulting in the current surge neces-
sary for cytoplasmic K?loss and cell death.
Earlier studies revealed that the Kv2.1-mediated current surge
during apoptosis is a result of the insertion of new channels into
the plasma membrane (14), and the p38 MAPK signaling
pathway is instrumental in this process (8, 9, 31). Interestingly,
p38 activation by TNF? has also been implicated in the poten-
tiation of tetrodotoxin-resistant Na?currents in sensory neu-
not been documented (32). The MAPK can phosphorylate
Nav1.6-encoded Na?channels, but in this case the resulting
interaction produces a decrease, rather than an increase, in Na?
currents (33). In addition, another member of the MAPK family,
ERK1/2, phosphorylates Kv4.2, a potassium channel responsible
for A-type K?currents in neurons (34, 35), resulting in a
down-regulation of dendritic K?currents in CA1 hippocampal
pyramidal neurons (36). Collectively, this information indicates
that MAPKs participate in cell signaling events that alter the
contribution of voltage-gated ion channels to overall neuronal
Kv2.1 is required for an entirely different, but critical process:
completion of the apoptotic program.
Several issues remain to be addressed. Primarily among them,
the molecular mechanism linking p38-induced phosphorylation
of Kv2.1 to the membrane insertion of this channel remains to
be determined. We previously observed that cleavage and
inactivation of the exocytotic soluble N-ethylmaleimide-sensitive
factor attachment protein receptor proteins SNAP-25 and syn-
taxin was sufficient to prevent the apoptotic K?current surge
(14). Importantly, both of these proteins are known to directly
interact with Kv2.1 (37): SNAP-25 with the cytoplasmic N-
terminal (38), and syntaxin with two domains of the cytoplasmic
C-terminal tail (39). Overexpression of syntaxin and Kv2.1
inhibits channel surfacing and leads to a decrease in Kv2.1-
mediated current density (39). It is thus tempting to speculate
that p38 phosphorylation of Kv2.1 leads to changes in its
association with syntaxin or other proteins, promoting, in turn,
exocytosis of channel containing vesicles. In addition, Kv2.1
subunits consistently colocalize with cortical neuron subsurface
cisterns, compressed stacks of smooth endoplasmic reticulum
situated ?5–8 nm beneath the plasma membrane (40). Thus, the
possibility exists for cisternae to serve as a nearby holding area
for channels destined for the plasma membrane (41), with p38
phosphorylation serving as the molecular switch necessary to
free Kv2.1 from cisternal retention during apoptosis.
In summary, we have identified a putative p38 phosphoryla-
tion target residue on the C-terminal of Kv2.1. Importantly,
mutation of S800 to a nonphosphorylatable residue is sufficient
to disrupt the apoptotic cascade initiated by oxidative injury and
block cell death. The work described here illustrates a critical
mechanistic link between oxidant-induced Zn2?release (7), p38
MAPK activation (8, 9), the ensuing Kv2.1-mediated apoptotic
K?current surge (13, 14), and cell death. Thus, prevention of
Kv2.1 phosphorylation by p38 is a potential novel therapeutic
Materials and Methods
Plasmids and Site-Directed Mutagenesis. The mammalian expres-
sion vector encoding WT Kv2.1 was the gift of J. Trimmer
(University of California, Davis, CA). The polymyc-tagged
Kv2.1 plasmid was from K. Takimoto (University of Pittsburgh,
Pittsburgh, PA). The FLAG-tagged dominant negative
Kv2.1(W365, Y380T) plasmid was provided by J. Nerbonne
(Washington University, St. Louis, MO). S. Korn (National
Institutes of Health, Bethesda, MD) provided the Kv2.1(I379C)
vector. Mutagenesis of these cDNAs was performed with a
QuikChange XL kit (Stratagene, La Jolla, CA) according to the
manufacturer’s directions. Primers containing the desired mu-
tations (S800A, S800D, S800E) were obtained from Integrated
DNA Technologies (Coralville, IA). Mutations were confirmed
by sequencing. A plasmid encoding EGFP (pCMVIE-eGFP;
Clontech, Palo Alto, CA) was used for the identification of
positively transfected cells.
Tissue Culture. CHO cells were plated at a density of 5.6 ? 104
cells per well on coverslips in 24-well plates 24 h before trans-
fection. Cells were treated for 4 h in serum-free medium (F12
nutrient medium with 10 mM Hepes) with a total of 1.2 ?l of
Lipofectamine (Invitrogen, Carlsbad, CA) and 0.28 ?g of DNA
per well (0.14 ?g of both EGFP and potassium channel cDNA).
Cells were briefly washed in MEM with Earle’s salts containing
25 mM Hepes and 0.01% BSA. After transfection, cells were
maintained in F12 medium containing FBS at 37°C, 5% CO2for
48 h before recordings.
Cortical neurons were prepared from embryonic day 16 rat
embryos and grown in 6-well plates according to McLaughlin et
al. (8). Cultures were exposed to drug treatment procedures at
nonconducting Kv2.1(W365C, Y380T) and 24 h later exposed to 30 ?M DTDP
expressed as a percent of vehicle-treated control. Note that DTDP induced
?50% cell death in Kv2.1-expressing CHO cells but was not lethal to either
the mean ? SEM viability from three separate, independent experiments (*,
P ? 0.05; ANOVA/Dunnett).
S800A mutation blocks Kv2.1-mediated apoptosis. CHO cells were
Redman et al.
February 27, 2007 ?
vol. 104 ?
no. 9 ?
25–29 days in vitro. Cells were briefly washed in MEM with
Earle’s salts containing 25 mM Hepes and 0.01% BSA and
maintained in D2C growth medium until harvesting. Cells were
harvested 3 h posttreatment in lysis buffer [1% Triton X-100/
0.1% SDS/0.25% Na deoxycholate/50 mM Hepes/150 mM NaCl/
protease inhibitor mixture (Roche Diagnostics, Indianapolis,
IN), pH 7.5] after two washes with PBS. Cell lysate samples were
combined in a 1:1 ratio with reducing sample prep buffer and
incubated for 5 min at 100°C to denature proteins before gel
Drug Treatment. The apoptotic stimulus for the electrophysiolog-
ical experiments in CHO cells consisted of a 5-min treatment
with 25 ?M DTDP at 37°C, 5% CO2. The DTDP-containing
solution was then removed and replaced with fresh F12 medium
containing 10 ?M 1–3-boc-aspartyl (Ome)-fluoromethyl-ketone
(BAF), a broad-spectrum cysteine protease inhibitor. BAF was
necessary to maintain cells viable for electrophysiological re-
cordings because Kv2.1-expressing cells are highly susceptible to
DTDP-induced apoptosis (13). Cells were subsequently main-
tained in BAF-containing medium, and electrophysiological
recordings were performed ?3 h after oxidative injury. For
channel insertion experiments, cells were first treated with 4 ?M
MTSET for 10 min to covalently block all Kv2.1(I379C) or
Kv2.1(I379C, S800A) channels present on the plasma membrane
surface before the usual DTDP exposure (14). The apoptotic
stimulus for the biochemical experiments on rat primary cortical
neurons consisted of a 10-min treatment with 100 ?M DTDP at
37°C, 5% CO2 (7, 8). Cortical neurons were pretreated with
either vehicle or the specific p38 MAPK inhibitor SB-293063 (20
?M for 10 min).
Electrophysiological Measurements. Current recordings were per-
formed on EGFP-positive cells by using the whole-cell patch clamp
configuration technique as described (8). We observed that 97% of
GFP-positive CHO cells cotransfected with Kv2.1 had measurable
K-gluconate, 10 mM KCl, 1 mM MgCl2, 1 mM CaCl2? 2H2O, 10
mM Hepes; pH adjusted to 7.2 with concentrated KOH; 0.22 mM
sucrose. The extracellular solution contained 115 mM NaCl, 2.5
mM KCl, 2.0 mM MgCl2, 10 mM Hepes, 0.1 mM 1,2-bis(2-
aminophenoxy)ethane-N,N,N?,N?-tetraacetate acid, 10 mM D-
glucose, and 0.1 mM tetrodotoxin; pH was adjusted to 7.2. Mea-
surements were obtained under voltage clamp conditions with an
Axopatch 1-D amplifier (Axon Instruments, Foster City, CA) and
pClamp software (Axon Instruments) using 2- to 3-M? electrodes.
Recording electrodes were pulled from 1.5-mm borosilicate glass
pipette puller (Sutter Instruments, Novato, CA). Series resistance
at 2 kHz and digitized at 10 kHz with Digidata software (Axon
Instruments). Potassium currents were evoked with incremental
15-mV voltage steps to ?35 from a holding potential of ?70 mV.
To determine current density values, steady-state current ampli-
and normalized to cell capacitance. All data are expressed as
mean ? SEM, and statistical analysis was performed with InStat
software (GraphPad, San Diego, CA).
Generation of Phospho-Specific Antibody. To generate an antibody
specific for the phosphorylated p38 MAPK consensus se-
quence of Kv2.1 (pKv2.1; SI Fig. 6), we synthesized the peptide
C-KNHFESSPLPTS(p)PKFLR (Tufts University Core Facil-
ity, Boston, MA). The HPLC-purified peptide was conjugated
to keyhole limpet hemocyanin (Pierce, Rockford, IL) with
Sulfo-Link (44895; Pierce) following the manufacturer’s pro-
tocol. Antiserum was generated in New Zealand White rabbits
at Covance (Princeton, NJ). After prebleed screening, animals
were initially injected with 1.5 mg of immunogen, followed by
three monthly boosts of 0.75 mg and three monthly boosts of
0.5 mg. The crude serum was affinity-purified by using an
Immunopure (A) IgG Purification Kit (Pierce) according to
the methods described by the manufacturer, and the resulting
product was used in the experiments reported here. A com-
mercially available Kv2.1 polyclonal antibody (Alomone
Labs), which was not targeted to the p38 site, was used as a
control in immunoblots.
Kv2.1(S800A)-expressed protein from transfected CHO cells
were immunoprecipitated by incubating cell lysates with an
anti-myc-tag rabbit polyclonal antibody (40 ?g/?l) (MBL
International Corp., Woburn, MA) followed by a protein A/G
PLUS-Agarose immunoprecipitation reagent (Santa Cruz
Biotechnology, Santa Cruz, CA). The immunopurified sub-
strate was incubated in 15 ?l of kinase buffer (25 mM Hepes,
pH 8.0/2 mM DTT/0.1 mM vanadate), 15 ?l of Mg/ATP (50
mM MgCl and 50 ?M ATP), and 50 ?g activated p38? (Roche
Protein Expression Group, Indianapolis, IN) for 1 h at 30°C.
Reactions containing kinase buffer alone and kinase buffer
plus Mg/ATP were used as controls. The reaction was stopped
with sample preparation buffer (625 mM Tris/25% glycerol/2%
SDS/0.01% bromophenol blue/5% ?-mercaptoethanol) and
incubated at 100°C for 5 min before SDS/PAGE and immu-
noblotting. Immunoblotting was performed with the pKv2.1
antibody (1:12,000) and a commercially available polyclonal
Kv2.1 antibody (1:3,000) (Alomone Labs).
KinaseAssay. Polymyc-Kv2.1- and polymyc-
Electrophoresis and Immunoblotting. SDS/PAGE was carried out
by standard procedures using the Mini Protean 3 System
(Bio-Rad, Hercules, CA). Equal amounts of cell lysate were
separated by reducing 6% or 10% SDS/PAGE gels. Separated
protein bands were transferred onto a 0.2-?m nitrocellulose
membrane (Bio-Rad). The membranes were then blocked with
1% BSA in PBS with 0.05% Tween 20 (PBST) at room
temperature for 1 h and probed with the appropriate primary
antibodies diluted in PBST. Blots were then incubated with
goat secondary antibody conjugated to HRP at room temper-
ature for 1 h. Blots were visualized with a SuperSignal
CL-HRP Substrate System (Pierce) and exposed to BioMax
films (Kodak, New Haven, CT). In the cortical culture exper-
iments, optical density measurements (Scion Image software,
National Institutes of Health) were taken of both pKv2.1
immunoreactive bands, normalized to their respective Kv2.1
immunoreactive bands, and the resulting values were pooled.
Viability Assays. CHO cells were plated and transfected with
EGFP and pRBG4 vector plus Kv2.1, Kv2.1(S800A) or
Kv2.1(W365C/Y380T) according to the protocol described
above, with 0.28 ?g of DNA added per well (0.14:0.13:0.0028
?g; EGFP/empty vector/potassium channel cDNA). Twenty-
four hours after transfection, cells were exposed to either 30
?M DTDP or vehicle for 15 min at 37°C, 5% CO2. Twenty-four
hours after treatment, counts of GFP-positive cells were
obtained by a person blinded to the experimental treatment
groups from 15 fields with a ?20 objective per coverslip; three
coverslips were counted per condition in three independent
We thank D. Jimenez for preliminary electrophysiological measure-
ments, M. Aras for helpful suggestions, and Drs. J. Trimmer, K.
Takimoto, J. Nerbonne, and S. Korn for the gifts of plasmids. This work
was supported by National Institutes of Health Grants NS043277 (to
E.A.) and HL080632 (to E.S.L.) and Mental Retardation Research
Center Grant HD018655 (to P.A.R. and L.H.).
www.pnas.org?cgi?doi?10.1073?pnas.0610159104Redman et al.
1. Yu SP (2003) Prog Neurobiol 70:363–386.
2. Bortner CD, Hughes FM, Jr, Cidlowski JA (1997) J Biol Chem 272:32436–
3. Yu SP, Yeh CH, Sensi SL, Gwag BJ, Canzoniero LM, Farhangrazi ZS, Ying
HS, Tian M, Dugan LL, Choi DW (1997) Science 278:114–117.
4. Hughes FM, Jr, Bortner CD, Purdy GD, Cidlowski JA (1997) J Biol Chem
5. Zaks-Makhina E, Kim Y, Aizenman E, Levitan ES (2004) Mol Pharmacol
6. Grishin A, Ford H, Wang J, Li H, Salvador-Recatala V, Levitan ES, Zaks-
Makhina E (2005) Am J Physiol 289:G815–G821.
7. Aizenman E, Stout AK, Hartnett KA, Dineley KE, McLaughlin BA, Reynolds
IJ (2000) J Neurochem 75:1878–1888.
8. McLaughlin BA, Pal S, Tran MP, Parsons AA, Barone FC, Erhardt JA,
Aizenman E (2001) J Neurosci 21:3303–3311.
9. Bossy-Wetzel E, Talantova MV, Lee WD, Schlozke MN, Harrop A, Matthews
E, Gotz T, Han J, Ellisman MH, Perkins GA, et al. (2004) Neuron 41:351–365.
Aizenman E (2006) Neuroscience 143:1–6.
11. Murakoshi H, Trimmer JS (1999) J Neurosci 19:1728–1735.
12. Malin SA, Nerbonne JM (2002) J Neurosci 22:10094–10105.
13. Pal S, Hartnett KA, Nerbonne JM, Levitan ES, Aizenman E (2003) J Neurosci
14. Pal S, Takimoto K, Aizenman E, Levitan ES (2006) Cell Death Differ
15. Obernauer JC, Cantley LC, Yaffe MB (2003) Nucleic Acids Res 31:3635–3641.
16. Murakoshi H, Shi G, Scannevin RH, Trimmer JS (1997) Mol Pharmacol
17. Mohapatra DP, Trimmer JS (2006) J Neurosci 26:685–695.
18. Park KS, Mohapatra DP, Misonou H, Trimmer JS (2006) Science 313:976–979.
19. Yu SP, Kerchner GA (1998) J Neurosci Res 52:612–617.
20. Kurz LL, Zuhlke RD, Zhang HJ, Joho RH (1995) Biophys J 68:900–905.
21. Zhang HJ, Liu Y, Zuhlke RD, Joho RH (1996) Biophys J 71:3083–3090.
22. Sun H, Shikano S, Xiong Q, Li M (2004) Proc Natl Acad Sci USA 101:16964–
23. Trimmer JS (1993) FEBS 324:205–210.
24. Scannevin RH, Murakoshi H, Rhodes KJ, Trimmer JS (1996) J Cell Biol
25. Chung JJ, Li M (2005) FEBS 272:3743–3755.
26. Misonou H, Mohapatra DP, Park EW, Leung V, Zhen D, Misonou K,
Anderson AE, Trimmer JS (2004) Nat Neurosci 7:711–718.
27. Roder K, Koren G (2006) Biol Chem 387:1237–1246.
RF, McVey MJ, Legos JJ, et al. (2001) J Pharmacol Exp Ther 296:312–321.
29. Ren X, Shand SH, Takimoto K (2003) J Biol Chem 278:43564–43570.
30. Du J, Haak LL, Phillips-Tansey E, Russell JT, McBain CJ (2000) J Physiol
31. Pal S, He K, Aizenman E (2004) Pflu ¨gers Arch 448:296–303.
32. Jin X, Gereau RW, IV (2006) J Neurosci 26:246–255.
33. Wittmack EK, Rush AM, Hudmon A, Waxman SG, Dib-Hajj SD (2005)
J Neurosci 25:6621–6630.
34. Adams JP, Anderson AE, Varga AW, Dineley KT, Cook RG, Pfaffinger PJ,
Sweatt JD (2000) J Neurochem 75:2277–2287.
35. Schrader LA, Birnbaum SG, Nadin BM, Ren Y, Bui D, Anderson AE, Sweatt
JD (2006) Am J Physiol 290:C852–C861.
36. Yuan LL, Adams JP, Swank M, Sweatt JD, Johnston D (2002) J Neurosci
37. Michaelevski I, Chikvashvili D, Tsuk S, Singer-Lahat D, Kang Y, Linial M,
Gaisano HY, Fili O, Lotan I (2003) J Biol Chem 278:34320–34330.
38. MacDonald PE, Wang G, Tsuk S, Dodo C, Kang Y, Tang L, Wheeler MB,
Cattral MS, Lakey JR, Salapatek AM, et al. (2002) Mol Endocrinol 16:2452–
39. Leung YM, Kang Y, Gao X, Xia F, Xie H, Sheu L, Tsuk S, Lotan I, Tshushima
RG, Gaisano HY (2003) J Biol Chem 278:17532–17538.
40. Du J, Tao-Cheng JH, Zerfas P, McBain CJ (1998) Neuroscience 84:37–48.
41. O’Connell KM, Tamkun MM (2005) J Cell Sci 118:2155–2166.
Redman et al.
February 27, 2007 ?
vol. 104 ?
no. 9 ?