(Received for publication, January 2, 1997, and in revised form, May 6, 1997)
From the Departments of Veterinary Pathobiology and
§ Biochemistry, Purdue University, West Lafayette,
Indiana 47907-1153
Depolarization of rat brain synaptosomes causes an increase in phosphorylation of serine residues 573, 610, 623, and 687 on voltage-sensitive sodium channels. Although these sites have been shown to be phosphorylated by cAMP-dependent protein kinase in vitro and in situ, the depolarization-induced increase in their state of phosphorylation is not due to increased cAMP-dependent protein kinase activity, but requires calcium influx and protein kinase C. Since phosphorylation at this cluster of sites inhibits sodium current and would decrease neuronal excitability, this may be an important negative feedback mechanism whereby calcium influx during prolonged or repetitive depolarization can attenuate neuronal excitability and prevent further calcium accumulation. Phosphorylation of purified channels by protein kinase C decreases dephosphorylation of cAMP-dependent phosphorylation sites by purified calcineurin or protein phosphatase 2A. This suggests that one mechanism by which protein kinase C may increase phosphorylation of cAMP-dependent phosphorylation sites in sodium channels is to inhibit their dephosphorylation. This represents an important new mechanism for convergent regulation of an ion channel by two distinct signal transduction pathways.
Sodium influx through voltage-sensitive sodium channels is responsible for the initiation and propagation of action potentials in most neurons, and ultimately promotes the opening of voltage-sensitive calcium channels at the synapse, leading to calcium influx and exocytosis. Because of their central role in these processes, the response properties of sodium channels can control neuronal excitation and secretion. Phosphorylation of brain sodium channels by PKA1 and PKC dramatically alters their electrophysiological responses. Cyclic AMP-dependent phosphorylation decreases sodium current in response to membrane depolarization (1, 2) due to a decrease in the probability of channel opening (1). Phosphorylation by PKC slows channel inactivation (3), decreases sodium current (3-5), and is required for the electrophysiological effect of cAMP-dependent phosphorylation to be observed (6). Two electrophysiological studies have demonstrated that neurotransmitters modulate sodium channel function via second messenger-induced phosphorylation. Stimulation of D1 dopamine receptors in dissociated rat nigrostriatal neurons reduces sodium current in a cAMP-dependent manner (7). Stimulation of muscarinic receptors in cultured hippocampal neurons, via activation of PKC, also reduces sodium current (8).
The subunit of brain sodium channels is phosphorylated by PKA on
multiple serine residues clustered in a single intracellular loop
between homologous domains I and II of the channel protein (9-11).
Replacement of this loop with the corresponding loop from a skeletal
muscle sodium channel lacking these phosphorylation sites abrogates the
cAMP-mediated inhibition of channels expressed in Xenopus
oocytes (12). Thus, this intracellular segment contains all
phosphorylation sites necessary for cAMP-dependent
inhibition of brain sodium channels. Li and co-workers (6) found that the effect of 8-bromo-cAMP on sodium current could not be observed in
Chinese hamster ovary cells expressing sodium channels without prior
activation of PKC. Ser-1506, located in the intracellular loop between
homologous domains III and IV, was the critical phosphorylation site
for PKC in causing this effect (6). Thus, sodium channels are a target
for convergent regulation by two distinct second messenger-regulated
pathways causing phosphorylation in two different regions of the
channel protein. Although PKC-mediated channel phosphorylation appears
to be a prerequisite for cAMP-dependent reduction in sodium
current, it is not known whether PKC phosphorylation alters the sodium
channel's response to cAMP-dependent phosphorylation or
whether it alters the channel's susceptibility to phosphorylation or
dephosphorylation at cAMP-dependent phosphorylation
sites.
In the present study, we observed that depolarization of rat brain synaptosomes caused a calcium-dependent increase in phosphorylation of cAMP-dependent phosphorylation sites on sodium channels. This response was not due to increased PKA activity, but required activation of PKC. These results suggest that accumulation of calcium and activation of PKC in nerve terminals may lead to increased phosphorylation and consequent decreased responsiveness of sodium channels. We also found that purified reconstituted sodium channels phosphorylated by both PKC and PKA were poorly dephosphorylated by purified calcineurin or PP2A, in contrast to sodium channels phosphorylated by PKA alone. This observation suggests one potential mechanism by which PKC activation during synaptosome depolarization could lead to an increase in phosphorylation of cAMP-dependent phosphorylation sites on sodium channels. The ability of one phosphorylation site on a given substrate to control dephosphorylation of a separate site represents a potentially important molecular mechanism for conditional or convergent regulation of the responses of ion channels and other phosphoproteins by distinct signal transduction pathways.
Sodium
channels were purified through the wheat germ-agglutinin chromatography
step according to Hartshorne and Catterall (13). Heterotrimeric PP2A
containing a mixture of PP2A1 and PP2A0 was
purified from rat brain as described previously (14). The catalytic
subunit of PKA was purified from bovine heart (15). Monoclonal antibody
1G11 generated toward purified rat brain sodium channel was kindly
provided by Dr. William Catterall (University of Washington).
Recombinant mouse PKC (16), expressed in Sf9 cells (17) and
partially purified from Sf9 cell lysates by anion exchange
chromatography was kindly provided by Dr. Curt Ashendel (Purdue University). Other biochemicals and reagents were purchased from the following sources: purified calcineurin, Upstate Biotechnology Inc.; N-tosyl-L-phenylalanine chloromethyl
ketone-treated trypsin, Worthington; phosphocellulose paper, Whatman;
[
-32P]ATP, NEN Life Science Products; protein kinase
inhibitor, which specifically inhibits PKA, Sigma; forskolin, PMA,
MPMA, BIM, (RP)-cAMPS, okadaic acid, lipids, and
cyclosporin A, Calbiochem.
All procedures were performed at 0-4 °C unless otherwise noted. Synaptosomes were freshly prepared from adult male rat brains as described previously (18) and resuspended in (20 mM HEPES, pH 7.4, 10 mM glucose, 132 mM NaCl, 4.8 mM KCl, 2.4 mM MgSO4, 0.1 mM EGTA, 1.2 mM CaCl2) to a final protein concentration of approximately 1.5 mg/ml. Samples (50-150 µg protein) were preincubated for 5 min at 37 °C, then treated with drugs or appropriate vehicle for various times before depolarization. To depolarize, synaptosomes were diluted 2-fold with prewarmed (20 mM HEPES, pH 7.4, 10 mM glucose, 22 mM NaCl, 115 mM KCl, 2.4 mM MgSO4, 0.1 mM EGTA, 1.2 mM CaCl2) for a final concentration of 60 mM potassium and incubated for 1 min at 37 °C. When examining the requirement for calcium, CaCl2 was omitted from each buffer.
Isolation and Back Phosphorylation of Sodium ChannelsAfter
treatment samples were solubilized, and sodium channels were
immunoprecipitated, back-phosphorylated with the catalytic subunit of
PKA and [-32P]ATP, then subjected to SDS-PAGE and
autoradiography as described previously (18). Gel slices containing
sodium channel
subunits were excised and dissolved, and their
32P content was quantified by liquid scintillation
spectrometry.
Two-dimensional tryptic
phosphopeptide maps were generated from SDS-polyacrylamide gel
electrophoresis gel bands containing 32P-labeled sodium
channel subunits as described previously (19, 20) or, in some
cases, by digesting after electrophoretic transfer to nitrocellulose as
described by Luo and colleagues (21).
Protein kinase A activity was
measured according to the protocol of Roskoski (22) with some
modifications. Briefly, after treatment synaptosomes were lysed with an
equal volume of (100 mM Tris, pH 7.4, 0.2 mM
EDTA, 0.2 mM EGTA, 9% Triton X-100). An aliquot of the
lysate was mixed with assay buffer in a final volume of 30 µl
containing 50 mM Tris, pH 7.4, 4 mM magnesium
acetate, 50 mM -glycerophosphate, 0.1 mM
3-isobutyl-1-methylxanthine, 2 mM EGTA, 1.5% Triton X-100,
1 mg/ml bovine serum albumin, 200 µM
[
-32P]ATP (2000 cpm/pmol), 1 mg/ml histone 2A, and
incubated for 5 min at 30 °C. An aliquot of each sample was spotted
onto phosphocellulose paper, washed in 75 mM
H3PO4, and air-dried, and 32P was
quantified by liquid scintillation spectrometry.
Purified rat brain sodium channels were reconstituted
into phospholipid vesicles and phosphorylated with the purified
catalytic subunit of PKA, with PKC, or with both kinases as described
by Murphy and Catterall (20) with minor modifications. To label cAMP-dependent phosphorylation sites, reconstituted
channels (1-4 pmol) were incubated at 30 °C for 2 h with the
catalytic subunit of PKA (0.5 µg/ml) in 295 mM sucrose,
50 mM Tris-HCl, pH 7.4, 50 mM
Na2SO4, 10 mM MgCl2, 1 mM EGTA, and 160 µM
[-32P]ATP (1-2 Ci/mmol). To label PKC phosphorylation
sites, reconstituted channels were treated at 30 °C for 2 h
with recombinant PKC (1 µg/ml) in 295 mM sucrose, 50 mM Tris-HCl, pH 7.4, 50 mM
Na2SO4, 10 mM MgCl2,
1.5 mM CaCl2, 60 µM
phosphatidylserine, 1.5 µM diolein, and 1 µg of protein
kinase inhibitor, to inhibit endogenous PKA activity present in wheat
germ-agglutinin purified channel preparations (20). To phosphorylate
with both kinases, reconstituted channels were treated first with PKC
without the addition of protein kinase inhibitor for 2 h, then the
catalytic subunit of PKA and EGTA (to a final concentration of 10 mM) were added, and the incubation was continued for an
additional 2 h. Phosphorylated reconstituted channels were passed
through Sephadex G-50 equilibrated in 221 mM sucrose, 50 mM Tris-HCl, pH 7.4, 75 mM
Na2SO4, 0.15 mM MgCl2, then used in phosphatase assays. Phosphopeptide maps of phosphorylated reconstituted channels indicated that the patterns of phosphorylation by PKC and PKA were similar to those previously reported (19, 20) (data
not shown). The average stoichiometry of channel phosphorylation (expressed as moles of phosphate/mol of channel) was 3, 2, and 4.2 for
phosphorylation by PKA, PKC, or both kinases, respectively.
Phosphorylated reconstituted sodium
channels were mixed in a 30-µl final volume containing either PP2A
(50 ng) in the presence of (0.1 mM EDTA, 0.1 mM
EGTA, 0.1% -mercaptoethanol) or with calcineurin (100 ng) in the
presence of (0.75 mM dithiothreitol, 10 mM
MgCl2, 0.1 mM CaCl2, 100 nM calmodulin, 500 nM okadaic acid). Okadaic
acid was included during treatment with calcineurin to inhibit low
levels of okadaic acid-sensitive phosphatase activity present in wheat
germ-agglutinin purified sodium channel
preparations.2 Samples were incubated at
37 °C for varying amounts of time, then the reaction was terminated
with 100 µl of 10% trichloroacetic acid and 10 µl of 10 mg/ml
bovine serum albumin. Protein was then sedimented, and acid-soluble
[32P]Pi was quantified.
Back phosphorylation (23) was used to measure changes in
phosphorylation of cAMP-dependent phosphorylation sites on
sodium channels. In this technique, endogenous phosphate is
incorporated into channels in synaptosomes during experimental
treatments, then channels are isolated under conditions preventing
dephosphorylation and treated with the catalytic subunit of PKA and
[-32P]ATP to incorporate radioactive phosphate into
sites that were not phosphorylated in situ. Thus, a decrease
in back phosphorylation reflects an increase in endogenous
phosphorylation. Back phosphorylation of immunopurified sodium channels
with the catalytic subunit of PKA specifically labels serine residues
573, 610, 623, and 687 of rat brain type IIA sodium channels (11).
Treatments that elevate cAMP in synaptosomes (24) or cultured rat brain
cells (19) block subsequent back phosphorylation of these sites,
indicating that these sites are phosphorylated by PKA in
situ. Serine residues 573, 610, 623, and 687 will be collectively
referred to as cAMP-dependent phosphorylation sites in this
study. It is not known whether other kinases also phosphorylate these
sites in vivo.
Depolarization of synaptosomes by high potassium in the presence of
calcium decreased back phosphorylation of channel
cAMP-dependent phosphorylation sites, indicating an
increase in endogenous phosphorylation (Fig. 1). The
magnitude of this response varied from experiment to experiment, but
ranged from 25 to 50% in 30 independent experiments. In the absence of
calcium no change in phosphorylation occurred, suggesting that the
increase in phosphorylation during depolarization requires calcium
influx.
The Depolarization-induced Increase in Channel Phosphorylation Is Not Due to an Increased PKA Activity
The role of PKA in this
response was examined. Depolarization of synaptosomes did not increase
PKA activity (Fig. 2). In addition, the increase in
channel phosphorylation caused by depolarization was not blocked by
pretreatment with the inhibitory cAMP derivative, (RP)-cAMPS (25) (Fig. 3). In
contrast, forskolin increased PKA activity in synaptosomes and
decreased back phosphorylation of sodium channels; both of these
responses were blocked by (RP)-cAMPS (Figs. 2
and 3). Treatment with (RP)-cAMPS alone had no
effect on control kinase activity or channel phosphorylation. These
results indicate that the depolarization-induced increase in
phosphorylation of cAMP-dependent phosphorylation sites on
sodium channels is not caused by increased PKA activity. However, since
(RP)-cAMPS may not completely block basal PKA
activity (26), basal PKA activity may still be responsible for the
accumulation of phosphate in cAMP-dependent phosphorylation
sites that occurs during depolarization.
The Response to Depolarization Requires PKC and Calcium
Because depolarization of synaptosomes increases
PKC-mediated phosphorylation (27, 28) and sodium channels are a
potential substrate for PKC phosphorylation in synaptosomes (29), the involvement of PKC in this response was examined. Treatment of synaptosomes with PMA (100 ng/ml) mimicked the effect of depolarization on back phosphorylation of sodium channels (Fig.
4A). The response to depolarization or PMA
treatment was blocked by BIM (1 µM), a specific inhibitor
of PKC (30, 31), suggesting that the effect of depolarization on sodium
channel phosphorylation is mediated by PKC. The response to PMA, like
that to depolarization, was prevented by removal of buffer calcium.
Treatment of synaptosomes with BIM alone (Fig. 4A) or with
the inactive phorbol ester MPMA (Fig. 4B) had no effect on
channel phosphorylation.
Protein kinase C and PKA both phosphorylate Ser-610 in
vitro, but other phosphorylation sites are not shared (20). To
determine whether the increase in endogenous phosphorylation at
cAMP-dependent phosphorylation sites was due solely to
phosphorylation of Ser-610, tryptic phosphopeptide maps of back
phosphorylated sodium channel subunits were prepared and the
changes in phosphorylation for each of the four
cAMP-dependent phosphorylation sites were quantified. As
shown in Fig. 5, depolarization affected all four
cAMP-dependent phosphorylation sites to a similar extent.
This demonstrates that the depolarization-induced increase in
phosphorylation of cAMP-dependent phosphorylation sites is
not solely due to the direct phosphorylation of Ser-610 by PKC.
However, we cannot rule out the possibility that PKC phosphorylates
additional cAMP-dependent phosphorylation sites in
vivo, since each of these sites are also potential phosphorylation sites for PKC (32). Circumstances in vivo, such as the
presence of a particular subtype of PKC (33) or its strategic
localization by A kinase anchoring proteins (34), may permit
PKC-mediated phosphorylation of cAMP-dependent
phosphorylation sites in addition to Ser-610.
Phosphorylation of Purified Channels by PKC Inhibits Dephosphorylation of cAMP-dependent Phosphorylation Sites by PP2A and Calcineurin
Since the state of phosphorylation of a
protein in situ reflects the competing activities of kinases
and phosphatases, the effect of phosphatase inhibitors on sodium
channel phosphorylation in synaptosomes was evaluated. Previous studies
have shown that okadaic acid, which inhibits PP2A and related
phosphatases, and cyclosporin A, which can inhibit calcineurin, both
increase endogenous phosphorylation of sodium channels in synaptosomes
(18). The effect of treatment with each of these agents was not
additive with the effect of depolarization (Fig. 6),
suggesting that the response to depolarization may be equivalent to
blocking dephosphorylation.
A decrease in dephosphorylation may arise from direct inhibition of
phosphatases or from a decrease in channel susceptibility to the action
of phosphatases. To test this second possibility, the ability of
phosphatases to dephosphorylate sodium channels phosphorylated by both
PKA and PKC was examined. Purified sodium channels were reconstituted
and phosphorylated with each kinase alone, or with both kinases, then
used as substrates for dephosphorylation. As in the case of soluble
sodium channels (11), reconstituted sodium channels phosphorylated with
PKA alone could be dephosphorylated by both calcineurin and PP2A (Fig.
7). In contrast, channels phosphorylated with PKC alone,
or with both PKC and PKA, were poorly dephosphorylated by either
phosphatase. This suggests that PKC phosphorylation of sodium channels
decreases the ability of both calcineurin and PP2A to attack
cAMP-dependent phosphorylation sites.
This study suggests that an important physiologic stimulus, neuronal depolarization, influences the responsiveness of voltage-dependent brain sodium channels by controlling the state of phosphorylation of cAMP-dependent phosphorylation sites. In addition, two novel aspects of the role of PKC in controlling sodium channel phosphorylation are revealed; PKC is required for the effect of synaptosomal depolarization on sodium channel phosphorylation, and the direct phosphorylation of purified sodium channels by PKC reduces the ability of phosphatases to attack cAMP-dependent phosphorylation sites in vitro. Thus, activation of PKC influences the extent of phosphorylation at cAMP-dependent phosphorylation sites on voltage-sensitive sodium channels in brain.
The mechanism by which depolarization increases sodium channel phosphorylation is not known. Our results with purified sodium channels suggest that a PKC-induced decrease in channels' susceptibility to dephosphorylation could play a role in elevating phosphorylation of cAMP-dependent phosphorylation sites on sodium channels in response to depolarization. However, the circumstances in synaptosomes are complex due to the presence of multiple Ser/Thr phosphatases and second messenger-regulated kinases, and other mechanisms may also lead to increased sodium channel phosphorylation. These include the possibility that PKC itself or a kinase activated by PKC directly phosphorylates cAMP-dependent phosphorylation sites on sodium channels in situ, and the possibility that activation of PKC may decrease phosphatase activity in synaptosomes. For example, PKC has been shown to phosphorylate and inhibit calcineurin in vitro (35), and a recent study suggests that activation of PKC may modulate PP2A activity in cardiac myocytes (36). These potential mechanisms remain to be investigated in synaptosomes.
Although it is not clear whether the role of PKC in the depolarization-induced elevation of sodium channel phosphorylation is due to a direct phosphorylation of sodium channels or is indirect, it is unlikely that PKC acts solely by promoting calcium influx or membrane depolarization. Although PKC inhibition during synaptosome depolarization decreases calcium influx and membrane depolarization to a moderate extent (37), treatment of synaptosomes with phorbol ester has little or no effect on calcium levels (38, 39) or membrane depolarization (39) and does not enhance either of these responses to treatment with high potassium (37). In the present study, PMA increased sodium channel phosphorylation in the absence of a depolarizing stimulus, conditions under which calcium elevation is not expected (38, 39). Although the effect of PMA was prevented by incubating synaptosomes in a calcium-free medium, the loss of responsiveness may be explained by a depletion of calcium in synaptosomes during treatment with calcium-free buffer containing EGTA. Others have demonstrated that PKC can phosphorylate different synaptosome substrates depending on the means of activation (40). To determine whether PKC directly phosphorylates sodium channels during depolarization, the phosphorylation of Ser-1506 should be measured. Although phosphorylation of this site in vivo has not yet been examined, it is phosphorylated by PKC but not PKA in vitro (20) and this site plays a key role in the electrophysiological effects of PKC on sodium channels (6).
In addition to its essential role in initiating exocytosis, depolarization-induced calcium influx can modulate the subsequent excitability of the synaptic terminal by activating a host of calcium-dependent enzymes which can alter ion channel function (41, 42). Because phosphorylation at cAMP-dependent phosphorylation sites inhibits brain sodium channels, the enhanced state of channel phosphorylation after depolarization would be expected to inhibit subsequent synaptic excitation, decreasing further calcium influx and secretion. However, a number of studies suggest that activation of PKC enhances neurosecretion (reviewed in Refs. 39 and 43). One possible explanation reconciling these two views may be that rapid or early PKC-mediated phosphorylation events enhance secretion, whereas the increase in channel phosphorylation at cAMP-dependent sites that we observe may require periods of repetitive or sustained depolarization. Early PKC-mediated events may include the phosphorylation of sodium channels at Ser-1506. Phosphorylation at this site slows channel inactivation and would be expected to prolong depolarization and promote calcium influx. Thus, one possible scenario may be that depolarization and calcium influx activates PKC, which phosphorylates a number of target proteins, including sodium channels on Ser-1506, enhancing neurosecretion. During repetitive or prolonged depolarization, the phosphorylation of cAMP-dependent phosphorylation sites is gradually enhanced due to increased phosphorylation, decreased dephosphorylation, or both. This leads to inhibition of sodium channels during subsequent depolarizing stimuli. This may represent a negative feedback mechanism by which neurons are protected from accumulating dangerously high levels of calcium during periods of prolonged depolarization or hyperexcitation.
Several cases have been described in which phosphorylation of one site
on a given substrate controls the phosphorylation of another site (44).
Although fewer examples of one phosphorylation site controlling the
dephosphorylation of a separate phosphorylation site have been
reported, several important cases have recently been described. These
include an enhanced rate of phosphotyrosine dephosphorylation in the
presence of a nearby phosphothreonine residue on mitogen-activated
protein kinase by dual-specificity phosphatases (45, 46), the
inhibition of calcineurin-mediated dephosphorylation of neuronal
DARPP-32 at Thr-34 by casein kinase I-mediated phosphorylation at
Ser-137 (47), and the resistance to dephosphorylation of PKC induced
by phosphorylation of Thr-638 (48). Such interaction between
phosphorylation sites represents an important molecular mechanism for
the conditional regulation of a given response. Our results with
purified sodium channels suggest that the ability of phosphatases to
dephosphorylate cAMP-dependent phosphorylation sites on
sodium channels is modulated by PKC phosphorylation of the channel
itself. This effect may contribute to the convergent regulation of
sodium channel function by PKC and PKA, observed in
electrophysiological studies (6), and to the increased phosphorylation of cAMP-dependent phosphorylation sites during synaptosomal
depolarization. Since PKC phosphorylation of Ser-1506 is critical for
cAMP-dependent inhibition of sodium channel function (6),
it will be important to learn if phosphorylation of Ser-1506 is
responsible for the effect on dephosphorylation of
cAMP-dependent phosphorylation sites observed in
vitro.
This is journal paper no. 15276 from the Purdue University Agricultural Experiment Station.
We thank Dr. William Catterall for providing sodium channel antibodies, Dr. Curt Ashendel for providing recombinant PKC, Anna Kempa for purifying sodium channels, Laura Lofland for preparing this manuscript, and Drs. David Armstrong and Stan Rane for critically reading the manuscript.