Download Channel Protein From Rat Heart Using Subtype

735
Partial Characterization of the rHl Sodium
Channel Protein From Rat Heart Using
Subtype-Specific Antibodies
Sidney A. Cohen, Linda K. Levitt
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Three subtype-specific antisera were generated against peptides corresponding to portions of the amino
terminus, interdomain 1-2, and carboxy terminus of the rHl sodium channel primary sequence to confirm
the expression of this protein in the adult rat heart and to determine selected biochemical properties of
this protein that might contribute to its subtype-specific characteristics. All three antisera identify a
240-kD band on Western blots of partially purified cardiac membrane proteins and by immunoprecipitation of iodinated partially purified membrane proteins. Unlike other characterized mammalian sodium
channels, no P3 subunit is detected in association with the rHl subunit. The rHl a subunit is a complex
sialoglycoprotein as evidenced by its interaction with wheat germ agglutinin-Sepharose and by reduction
in its apparent molecular weight after treatment with neuraminidase; deglycosylation with N-glycanase
confirms that the rHl protein contains significantly less carbohydrate than other sodium channel proteins
characterized to date (5% versus 25% to 39%). Consistent with electrophysiological studies indicating a
role of phosphorylation in channel regulation, the rHl a subunit can be phosphorylated by the catalytic
subunit of cAMP-dependent protein kinase A. The possible functional significance of these findings is
discussed. (Circ Res. 1993;73:735-742.)
KEY WoRDs * ion channels * cardiovasculature * phosphorylation * glycosylation * J3 subunits
a
lectrophysiological, toxin binding, and molecular
studies provide evidence for the existence of
multiple sodium channel subtypes in cardiac
tissues.1 Multiple subtypes are thought to provide
functional diversity to excitable cells. Although the
predominant sodium current in the heart is tetrodotoxin insensitive and is responsible for the rapid
upstroke phase of the action potential, electrophysiological studies suggest the presence of at least two
additional sodium currents,2 6 whereas radiolabeled
toxin-binding studies identify two distinct toxin-binding phenotypes in cardiac tissues.78 Molecular studies
using Northern blotting, RNase protections assays,
and the sequencing of clones have identified five
distinct sodium channel transcripts in rat cardiac
tissues and, to date, two related sodium channel
transcripts in human cardiac tissues. The predominant
transcript in rat (rHl) and human (hHl) cardiac
tissues appear to be homologues of the same subtype9J10; both possess electrophysiological characteristics virtually identical to the dominant channel seen
electrophysiologically.11'2 mRNA corresponding to a
second sodium channel subtype originally cloned from
rat glial cells has also been shown to be abundantly
expressed in rat cardiac tissues13; a related channel
has been cloned and sequenced from human heart and
E
Received August 10, 1992; accepted June 22, 1993.
From the Cardiology Division, Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia.
Reprint requests to Sidney A. Cohen, MD, PhD, 522A Johnson
Pavilion, University of Pennsylvania, School of Medicine, 36th and
Hamilton Walk, Philadelphia, PA 19104-6060.
myometrium.14 A third subtype (CSC-1) has only been
partially sequenced but appears to be rat heart specific.15 Finally, transcripts corresponding to two sodium
channel subtypes that are identical to channels originally cloned from rat brain (rBl and rB3) have also
been identified in rat cardiac tissues.10,16
Sodium channel subtypes differ in their flux rates,
channel kinetics, voltage sensitivity, and drug and toxin
sensitivity.17-19 Although the common features of sodium channel function are thought to reside in portions
of the primary sequence that demonstrate high homology among subtypes, it is not clear whether the unique
characteristics of each subtype are due to differences in
channel primary sequence, posttranslational modification, presence of associated f, and/or /32 subunits, or a
combination of the above factors.
Two previous studies provide insight into the biochemical/biophysical characteristics of sodium channels
in the rat heart. In the first, an antibody against a
portion of the conserved interdomain 3-4 region was
used to identify a 230-kD a subunit that contained
approximately 8 kD of sialic acid and that could be
phosphorylated by the catalytic subunit of protein kinase A.20 In the second study, the same authors used a
rat brain 6,1-specific antibody to demonstrate the presence of ,1 subunits in cardiac tissues.2' Because the
antibody used in the first study was generated against a
region that is nearly 100% conserved in all sodium
channels, because the second study used a rat brain
13,-specific antibody, and because multiple sodium channel subtypes are expressed in the rat heart, it is unclear
which sodium channel subtype(s) was identified in these
previous studies.
736
D-28
Circulation Research Vol 73, No 4 October 1993
D-492
D-1956
NH
LOCATION
NAME
AMINO ACIDS
D-28
28-48
N-TERMINAL
D-492
D-1956
492-511
ID 1-2
1956-1971
C-TERMINAL
-m
m_*C00
SEQUENCE
MAEKQARGGSATSQESREGLOC
DRLPKSDSEDGPRALNQLSLC
GNFSRRSRPLSSSSISC
FIG 1. Diagram showing location in the rH1 primary sequence ofpeptides used to generate subtype-specific antisera.
ID 1-2 indicates interdomain 1-2.
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Because of this uncertainty and in order to further
characterize each of the subtypes whose messages have
been identified in cardiac tissues, we have begun to
develop subtype-specific antipeptide antibodies for use
in biochemical and immunocytochemical studies. The
present study details our efforts with the rHl subtype.
These studies confirm the expression of the rHl protein
in the adult rat heart and elucidate several unique
characteristics that may be important in regulating the
synthesis, membrane insertion, and half-life of this
channel and in modulating its electrophysiological
properties.
Materials and Methods
Materials
Materials for the preparation of oligopeptides and
antibodies, isolation of crude membranes, and partial
purification of membrane proteins were obtained from
sources previously identified.22 G-25 Sephadex, DEAE
Sephadex (A 25-120), wheat germ agglutinin-agarose,
immobilized protein A, the catalytic subunit of cAMPdependent protein kinase from bovine heart (PKA),
protease inhibitors, neuraminidase, and one set of prestained molecular weight standards (26 200 to 180 000
D) were obtained from Sigma Chemical Co, St Louis,
Mo. All electrophoresis materials including a second set
of prestained molecular weight standards (14 400 to
97 400 D) were from Bio-Rad Laboratories, Richmond,
Calif. Agarose-immobilized diaminodipropylamine
(DADP-agarose), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), and Amino Link coupling gel
were obtained from Pierce Chemical Co, Rockford, Ill.
1251-Protein A, 25I-Bolton Hunter reagent, and
[y-32P]ATP were from ICN Radiochemicals, Irvine,
Calif. Recombinant N-glycanase was obtained from
Genzyme Corp, Boston, Mass.
Preparation of Site-Directed Antisera
Polyclonal antibodies were prepared to synthetic oligopeptides as described previously.22 All oligopeptides
correspond to portions of the rHl cardiac sodium
channel sequence as given in Fig 1. A carboxy-terminal
cysteine residue was added to assist in coupling peptides
to the carrier protein.
Preparation of Crude Surface Membrane
Freshly frozen rat heart (100 g) was pulverized with a
mortar and pestle and extracted five times (to remove
blood) with 0.25 mol/L sucrose in 0.1 mol/L Tris at pH
7.4 (4°C) containing the following protease inhibitors:
10 mmol/L EDTA, 10 mmol/L EGTA, 0.1 mmol/L
phenylmethylsulfonyl fluoride (PMSF), 1 mmol/L iodoacetamide, and 0.1 ,ug/mL pepstatin A. The procedure was otherwise as previously described22 except that
the final sucrose gradient was omitted. Membrane fragments were kept frozen at -70°C until use.
Partial Membrane Protein Purification
Membranes were solubilized in buffer containing
180 mmol/L KCI, 20 mmol/L potassium phosphate
(pH 6.5), 0.5 mmol/L MgCI2, 1 mmol/L EGTA, 0.1
mol/L PMSF, and 0.1 ,ug/mL pepstatin A (buffer A) by
the addition of 10% Nonidet P-40 (NP-40) to achieve
a final concentration of 1% detergent and then placed
on a rotating platform at 4°C for 15 minutes. After
centrifugation at 125 OOOxg for 30 minutes, the supernatant was batch-adsorbed onto DEAE Sephadex (A
25-120), which was preequilibrated with buffer A
containing 0.1% NP-40 (buffer B). The adsorbed resin
was washed extensively with buffer B and eluted with
buffer containing 400 mmol/L KCI, 20 mmol/L potassium phosphate (pH 6.5), 0.5 mmol/L MgC12, 1
mmol/L EGTA, 0.1 mmol/L PMSF, 0.1 ,ug/mL pepstatin A, and 0.1% NP-40 (buffer C). The eluate from
the DEAE Sephadex was applied to a column of wheat
germ agglutinin immobilized on agarose. The wheat
germ agglutinin column was extensively washed with
buffer C, equilibrated with buffer containing 100
mmol/L NaCl, 50 mmol/L sodium phosphate (pH 7.5),
0.5 mmol/L MgCl, 1 mmol/L EGTA, and 0.05% NP-40
(buffer D), and eluted with buffer D containing 100
mmol/L N-acetylglucosamine. Two-step purified membrane proteins were kept at 4°C in 0.03% sodium azide
and used within 1 week.
Sodium Dodecyl Sulfate-Polyacrylamide
Gel Electrophoresis
Sodium dodecyl sulfate (SDS)-polyacrylamide gel
electrophoresis (PAGE) sample buffer (5 x) contained
5% SDS, 0.43 mol!L dithiothreitol, 0.28 mol/L Tris (pH
6.8), and 44% glycerol. Proteins were denatured with
1 x SDS-PAGE sample buffer at 65°C for 20 minutes
and resolved by electrophoresis using the buffer system
of Cleveland et al.23 The stock solution contained 30%
acrylamide and 0.2% bisacrylamide. The separating
gels were 7% to 20% linear acrylamide gradients, and
the stacking gels were 5% acrylamide made from a stock
solution of 30% acrylamide and 0.8% bisacrylamide.
Gels containing labeled protein were dried and subjected to autoradiography using Kodak XAR-5 film
with a Cronex (DuPont) intensifying screen.
Two sets of molecular weight standards were run on
all gels. The first set contained ca2-macroglobulin (subunit Mr, 180 000), ,B-galactosidase (subunit Mr, 1 16 000),
fructose-6-phosphate kinase (Mr, 84 000), pyruvate kinase (Mr, 58 000), fumarase (Mr, 48 500), lactate dehydrogenase (Mr, 36 500), and triosephosphate isomerase
(Mr, 26 600). The second set contained phosphorylase B
(Mr, 97 400), bovine serum albumin (BSA; Mr, 66 000),
ovalbumin (Mr, 45 000), carbonic anhydrase (Mr,
29 000), soybean trypsin inhibitor (Mrs 21 500), and
lysozyme (Mr, 14 400). Adult rat skeletal muscle sodium
channel (M, 276 000) was also used for the phosphor-
Cohen and Levitt rHl Sodium Channel Characterization
737
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
ylation and iodination studies and served both as a
control as well as a molecular weight standard.
7.0, at 37°C for 4 hours before being subjected to
SDS-PAGE.
Transblots
Transblots were performed as described previously.22
For detection of immunoreactivity, the nitrocellulose
transblots were incubated in antiserum diluted 1:100 in
2% BSA, 50 mmol/L sodium phosphate (pH 7.4), and
150 mmol/L NaCl (buffer E) for 2 to 3 hours at room
temperature. The blots were washed five times with
buffer containing 50 mmol/L sodium phosphate (pH
7.4), 150 mmol/L NaCI, and 0.25% Tween 20 (buffer F).
All blots were finally incubated with 125`-labeled protein
A (250 ,Ci/mL, ICN Radiochemicals) diluted 1:200 in
buffer D for 1 hour at room temperature. Blots were
washed extensively in buffer F and wrapped in saran
wrap to prevent drying. Bound antibody was visualized
by autoradiography using Kodak XAR-5 film with a
Cronex (DuPont) intensifying screen.
Results
Antibodies directed against synthetic peptides corresponding to three sites distributed along the rHl channel primary sequence were used to identify the rHl
protein (Fig 1). Each peptide generated antisera that
was reactive with its respective peptide in radioimmunoassays (data not shown) and that specifically labeled
surface, t-tubular, and intercalated disc membranes in
immunocytochemistry of the rat heart.24
A crude membrane preparation was used as the
starting material for these studies because purified
membrane preparations did not retain sufficient
amounts of channel protein for biochemical characterization. Partial membrane protein purification was also
necessary because of the low density of this channel in
cardiac membranes.
All three antisera specifically identify an identical
240-kD band on Western blots of partially purified rat
heart membrane proteins (Fig 2, left). Occasionally, and
especially when protease inhibitors were omitted, additional lower molecular mass bands were visible on blots
probed with each of the rHl-specific antisera (Fig 2,
right). The number and migration of these lower molecular mass bands differ with the antibody used for
visualization; the pattern obtained with each antibody
resembles that produced by endogenous proteolysis of
the rat skeletal muscle sodium channel protein.25 Thus,
the 240-kD band represents the intact rHl a subunit,
whereas the lower molecular mass bands correspond to
proteolytic fragments of the a subunit.
Studies of mammalian sodium channels have consistently demonstrated that they contain noncovalently
associated P,B subunits and, in brain tissue, covalently
associated P2 subunits.26 To determine whether / subunits are associated with the rHl a subunit, partially
purified membrane proteins were first iodinated using
Bolton Hunter reagent and then immunoprecipitated
with each of the rHl-specific antisera. Each antiserum
immunoprecipitated a 240-kD band identical to the
a-subunit band obtained on Western blots; no lower
molecular mass bands were visualized on the autoradiogram (Fig 3). The absence of lower molecular mass
bands strongly suggests that the rHl a subunit is not
associated with either 3,B or 2 subunits. However, the
possibility that [3 subunits might not be sufficiently
labeled to be visualized on autoradiograms was further
investigated.
Scanning densitometry was used to determine the
relative intensity of the a and /8 subunit bands of the rat
skeletal muscle sodium channel positive controls; a ratio
of 6:1 (a: /) was obtained. This ratio more closely
reflects the relative quantity of carbohydrate than of
protein in these subunits (6:1 versus 8:1, respectively).
Confirming the suggestion that radiolabeled iodine is
preferentially incorporated into carbohydrate rather
than protein is the observation that neuraminidase
treatment results in the release of most of the radioactivity associated with iodinated skeletal muscle sodium
channel protein (data not shown).
Because the rHl sodium channel contains significantly less carbohydrate than is present on adult skeletal
muscle channels (see below), it is possible that any
Iodination With I251-Bolton Hunter Reagent
Partially purified membrane proteins were concentrated 40-fold to a final volume of 25 ,uL, adjusted to pH
8.7 with 5 ,uL of 1 mol/L sodium bicarbonate, and added
to a reaction vessel containing 50 ,uCi of `25I-Bolton
Hunter reagent. The reaction was allowed to proceed
for 3 hours at room temperature before being terminated by the addition of 10 ,L of 0.1 mol/L glycine.
lodinated protein was separated from reactants by
chromatography using G-25 Sephadex and immunoprecipitated using affinity-purified antibody bound to protein A-Sephadex (see below).
Phosphorylation
Partially purified membrane proteins were dialyzed
against deionized water, lyophilized, and resuspended
in buffer containing 25 mmol/L HEPES (pH 7.4), 5
mmol/L MgCI2, 5 mmol/L EGTA, and 0.05% NP-40
and phosphorylated by the addition of 0.4 ,ug PKA and
20 ,Ci [y-32P]ATP (3000 Ci/mmol) at 37°C. Reactions
were terminated by the addition of 10 mmol/L EDTA in
25 mmol/L sodium phosphate buffer (pH 7.4). The
labeled channel was then immunoprecipitated with purified antiserum bound to protein A-Sepharose (see
below).
Immunoprecipitation
Antibodies were affinity-purified on columns containing peptide immobilized to either DADP-agarose or
Amino Link resin (Pierce). Affinity-purified antibody
was bound to protein A-Sepharose and incubated overnight at 4°C with either iodinated or phosphorylated
partially purified membrane proteins and 1 mg/mL
BSA. Resin was recovered by gentle centrifugation,
washed extensively with phosphate-buffered saline, and
either used directly for deglycosylation or for SDSPAGE as described above.
Deglycosylation With Neuraminidase or N-Glycanase
Phosphorylated partially purified membrane proteins
were immunoprecipitated with affinity-purified D-492
antibody immobilized on protein A-Sepharose. Immunoprecipitated protein was then deglycosylated by treatment with either 20 U/mL neuraminidase or 28 U/mL
N-glycanase in 0.1 mol/L sodium phosphate buffer, pH
738
Circulation Research Vol 73, No 4 October 1993
D-28
D-28
240 kDa
D-492
D-1 956
D-492
IgG
240 kDa
-
142 kDa
-
240 kDa
~0
136 kDa
~88 kDa
48.5 kDa
-_
4I
14.8 kDa
-
FIG 2. On the left is a Western blot of partially purified cardiac membrane proteins. A crude rat heart membrane preparation
solubilized with Nonidet P-40 and partially purified by sequential ion exchange and lectin affinity chromatography.
Concentrated partially purified protein was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PA GE)
on 7% to 20% gradient gels. Protein was transferred to nitrocellulose, blocked with 4% bovine serum albumin/phosphate-buffered
saline (BSA/PBS), incubated with the indicated affinity-purified antibody in 2% BSA/PBS, developed with '24-protein A, and
exposed to Kodak XAR film. Immunoglobulin G (IgG) was affinity-purified using immobilized protein A and used at the same
final concentration as the other antibodies. All three antibodies identify an identical compact 240-kD band representing the rHi
subunit. On the right is a Western blot of endogenously proteolyzed rHi. Crude rat cardiac membrane proteins prepared without
protease inhibitors were solubilized with Nonidet P-40 and subjected to sequential ion exchange and lectin affinity chromatography.
Partially purified membrane proteins were then subjected to SDS-PAGE on 7% to 20% gels, transferred to nitrocellulose, probed
with antibody, developed with "5I-protein A, and exposed to Kodak XAR film. The pattern ofproteolysis resembles that described
by Kraner et a12-5 in their studies of endogenously proteolyzed skeletal muscle sodium channel protein.
was
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
a
rHI-associated f8 subunits might also contain less carbohydrate than their nerve and skeletal muscle counterparts and would thus be difficult to discern on
autoradiograms. To test for this, autoradiograms were
prepared in which the intensity of iodinated and immunoprecipitated skeletal muscle sodium channel bracketed the intensity of the rHl a-subunit bands. M subunits were easily visualized on these autoradiograms
(see lanes 6 and 7 in Fig 3). Scanning densitometry of
both these and the rHI autoradiograms demonstrates
both the ability to visualize fl-subunit bands when less
radioactivity is present and the absence of a discernible
fl-subunit band in the rHl-containing lanes (see bottom
of Fig 3). In addition, autoradiograms were systematically overexposed to ensure that /3 subunits would be
visualized even if they were not efficiently iodinated by
Bolton Hunter reagent. In each of these experiments,
no specific immunoprecipitated radiolabeled bands
were visible in the region where P3 subunits would be
expected to be located (=20 to 40 kD). These data
support the conclusion that the rHl sodium channel
does not contain associated : subunits.
Sodium channel phosphorylation by cAMP-dependent PKA has been reported to both regulate the
expression of functional channels and modulate the
activity of voltage-dependent sodium channels.27-31 To
determine whether the rHl protein possesses sites
capable of being phosphorylated by PKA, partially
purified membrane proteins were phosphorylated using
the catalytic subunit of PKA and then immunoprecipitated using each of the subtype-specific antisera. Again,
an identical 240-kD rHi protein was specifically immu-
noprecipitated (Fig 4), confirming the availability of
sites for PKA phosphorylation on the rHI a subunit.
The rHI subunit appears as a relatively compact
band on SDS-PAGE (Figs 3 and 4). This is in contrast
to previously characterized sodium channel a subunits,
which have molecular masses of 260 to 280 kD and
appear as diffuse bands on SDS-PAGE. All sodium
channel a subunits characterized to date contain approximately 25% to 30% (wt/wt) complex carbohydrate,
of which approximately half is sialic acid.32,33 The diffuse
character of the a-subunit band
Western blots
reflects the known microheterogeneity of glycosylation
found in complex glycoproteins.34 Based on a core
protein molecular weight of 227 417 and the relatively
compact appearance of the 240-kD band, it appears
that the rHI channel contains only 12.6 kD of carbohydrate compared with 50 to 60 kD of carbohydrate for
subunits from eel electroplax, rat
sodium channel
brain, and rat skeletal muscle.
The carbohydrate of the rHl channel contains either
N-acetylglucosamine or terminal sialic acid, as evidenced by the specific interaction of this glycoprotein
with immobilized wheat germ. To determine whether
this sugar is simple or complex and to determine the
core protein molecular mass of the rHl protein, partially purified, phosphorylated, and immunoprecipitated rHl protein was treated with neuraminidase or
N-glycanase. Treatment with neuraminidase resulted in
an :6-kD decrease in electrophoretic mobility, whereas
treatment with N-glycanase produced an t12-kD decrease in electrophoretic mobility (Fig 5).
a
on
a
Cohen and Levitt rHi Sodium Channel Characterization
1
276 kDa-
i
38 kDa-e
2
00
3
4
1
5
-240 kDa
2
3
4
739
5
ikDa
_F
_
240 kDa
.
FIG 4. Immunoprecipitation ofphosphorylated partially pu-
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
FIG 3. Immunoprecipitation of iodinated partially purified
cardiac membrane proteins. Partially purified membrane proteins were iodinated using 12[I-Bolton Hunter reagent and
immunoprecipitated with antibody bound to protein A-Sephadex. The lanes are identified as follows: lanes 1, 6, and 7,
adult rat skeletal muscle sodium channel (SkMJ) immunoprecipitated with I-467, an SkM1 subtype-specific antibody;
lanes 2 through 5, iodinated cardiac membrane proteins
specifically immunoprecipitated with antibody D-28 (lane 2),
D-492 (lane 3), D-1956 (lane 4), and immunoglobulin G
(lane 5). Note the characteristic diffuse SkM] a subunit band
with a mean molecular mass of 276 kD; lower molecular mass
bands in lane 1 are proteolytic fragments of the 276-kD
subunit with the exception of the 38-kD band, which represents
the SkM1 P subunit. Antibodies D-28 and D-492 immunoprecipitate a much tighter band with a mean molecular mass
of 240 kD. The 240-kD band immunoprecipitated by antibody
D-1956 is very faint and is not easily seen on these photographs. Because of the intense labeling present in the S/M1
subunit band in lane 1, lesser amounts of radiolabeled SkM1
were immunoprecipitated in lanes 6 and 7 to demonstrate the
ability to see channel /3 subunits when less radioactivity is
present. At the bottom of this figure are scanning densitometry
tracings of lanes 3, 6, and 7 performed under identical
conditions. These tracings demonstrate the relative levels of
125j incorporation into the ScM1 a and P subunits, the ability
to visualize channel ,B subunits under conditions of reduced
label incorporation, and the absence of label in the region
expected for /8 subunits in the rHi samples.
a
a
Discussion
The predominant sodium channel identified in cardiac tissues in vivo is tetrodotoxin and conotoxin insensitive, is sensitive to divalent cations, demonstrates
use-dependent behavior with local anestheties and antiarrhythmic agents, and is located diffusely throughout
atrial and ventricular muscle.' The rHl rat cardiac
channel was initially cloned from both a newborn rat
heart cDNA library10 and a denervated rat skeletal
muscle cDNA library.35 The rHl channel includes an
open reading frame of 6058 nucleotides that encodes for
a protein of 2019 amino acids with a predicted M, of
rified cardiac membrane proteins. Partially purified membrane
proteins were phosphorylated using the catalytic subunit of
protein kinaseA and (y-_32P)A TP and immunoprecipitated with
antibody bound to protein A-Sephadex. The lanes are identified as follows: lane 1, reaction mixture in the absence of
membrane proteins (ie, autophosphorylation); lane 2, phosphorylated total membrane proteins; and lanes 3 through 5,
phosphorylated cardiac membrane proteins specifically immunoprecipitated with antibody D-492 (lane 3), D-28 (lane 4),
and immunoglobulin G (lane 5). All sodium channel antibodies specifically immunoprecipitate the 240-kD rHi
subunit.
a
227 417.11i12 Northern blot analysis and RNase protection assays confirm that message for this channel is
abundantly expressed in heart and denervated skeletal
muscle but is not detectable in brain, kidney, spleen,
liver, uterus, or adult skeletal muscle. Electrophysiological examination of the rHi channel in a heterologous
expression system both demonstrates kinetics closely
resembling the predominant channel identified in vivo
and confirms the tetrodotoxin and conotoxin insensitivity of this channel, its use-dependent behavior with
type-I antiarrhythmic agents, and its sensitivity to divalent cations.11,36-38 In the present study, we confirm the
expression of the protein product of this gene in the
adult rat heart and biochemically characterize selected
properties of the rHl protein that may be important in
regulating both channel expression and function. In a
related abstract, we describe the immunocytochemical
localization of the rHl channel to the surface and
t-tubular membranes of atrial and ventricular muscle
cells and to the intercalated discs between adjacent
ventricular muscle cells.36
rHi Molecular Mass
The rHl subunit appears as a compact band on
SDS-PAGE with a molecular mass of 240 kD compared
with 260 to 280 kD for the eel electroplax, rat brain, and
rat skeletal muscle sodium channels. With a core protein molecular mass of 227.4 kD, this implies the
presence of only 12.6 kD of carbohydrate compared
with 50 to 60 kD of carbohydrate in other characterized
sodium channel proteins. The finding that the molecular
mass of the deglycosylated protein approximates the
calculated core protein molecular mass indicates that
minimal, if any, posttranslational proteolysis of the
a
740
Circulation Research Vol 73, No 4 October 1993
FIG 5. Deglycosylation of the ri-f channel with
neuraminidase and N-glycanase. The autoradiogram shows partially purified membrane proteins
phosphorylated using the catalytic subunit ofprotein
kinase A and (y-32P)A TP, immunoprecipitated with
antibody D- 492 bound to protein A-Sephadex, and
subjected to sodium dodecyl sulfate-polyacrylamide
gel electrophoresis (SDS-PA GE) on 7% to 20%
acrylamide gradient gel The lanes are identified as
follows: lane 1, control immunoprecipitated phosphorylated cardiac sodium channel protein; lane 2,
immunoprecipitated phosphorylated cardiac sodium channel protein treated with 20 U/mL
neuraminidase for 4 hours at 25°C before being
subjected to SDS-PAGE; and lane 3, immunoprecipitated phosphorylated cardiac sodium channel
protein treated with 28 U/mL N-glycanase for 4
hours at 25°C before being subjected to SDS-PAGE.
2
240 kDa
236 kDa
-4-
228 kDa
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
channel a subunit has occurred. This is confirmed by the
ability of both amino- and carboxy-terminal antibodies
to immunoprecipitate the 240-kD protein.
rHl has 14 potential N-linked glycosylation sites
compared with up to 18 in other sodium channels; 5 of
these sites are common to all sodium channels sequenced to date. The reduced quantity of carbohydrate
associated with the rHI subunit implies either that
less carbohydrate is attached per potential glycosylation
site or that not all potential sites are glycosylated.
Assuming 2 to 6 kD of sugar per potential glycosylation
site,22 then anywhere from 2 to 6 of the 14 potential
glycosylation sites may be posttranslationally modified
by carbohydrate. This can be compared with the adult
skeletal muscle sodium channel protein, in which nearly
every potential glycosylation site contains carbohydrate
(S.A. Cohen, S.J. Zwerling, and R.L. Barchi, unpublished observations).
Core glycosylation appears to be required for the
normal synthesis, processing, and membrane insertion
of sodium channels32; terminal sialic acid groups are
thought to play a role both in determining the magnitude of the electric field affecting sodium channel
activation and in stabilizing the normal conducting
conformation of the ionic pore.37 The present study
demonstrates that, similar to other sodium channels,
carbohydrate attached to the rHI protein is complex, as
evidenced by an 56-kD reduction in molecular mass
after neuraminidase treatment. However, although the
rHl channel resembles other sodium channel subunits
in being a complex sialated glycoprotein, it is unique in
that it contains 5% rather than 25% to 30% carbohydrate by weight.
Finally, rHI channel cDNA expressed in oocytes has
a linearly shaped current-voltage curve and a reduced
single-channel current amplitude. These unique electrophysiological characteristics are thought to result
from decreased negative charge near the extracellular
opening of the rHI channel. Although differences in
local amino acid residue charge have been suggested as
being responsible for this effect,938 it is also possible
that the reduced glycosylation and resulting fewer negatively charged sialic acid residues present on the rHl
channel might contribute to this finding. The exact role
of the reduced levels of glycosylation in the expression
a
a
and function of this channel will need to be determined
by further studies.
Absence of ,B Subunits
Although our autoradiograms clearly demonstrate no
radiolabeled bands in the region where one would
expect to see channel ,B subunits, it is possible that /3
subunits might have become dissociated from a subunits
during biochemical manipulation. We think this unlikely
since ,3 subunits are not easily dissociated from either
subunits, the only
rat brain or rat skeletal muscle
other /-associated subunits that have been extensively
biochemically characterized. Rat brain fl3 subunits require treatment with 1.0 mol/L MgCl2 followed by
sedimentation through sucrose gradients to be separated from rat brain a subunits, whereas treatment with
disulfide reducing agents is required to separate rat
subunits.27 Rat
brain /32 subunits from rat brain
skeletal muscle /3 subunits require ionic detergents to be
separated from subunits.28 a: P stoichiometry in adult
rat skeletal muscle was determined to be 1 :1,28 whereas
a:/3: 32 stoichiometry in rat brain was shown to be
1:1.39 In no case in the literature2' 39 or in our
experience with skeletal muscle sodium channels has
solubilization with nonionic detergents or immunoprecipitation resulted in dissociation of /3 from a subunits.
Therefore, it is unlikely that unrecognized loss of P
subunits occurred during any of our biochemical
a
a
a
a
manipulations.
An alternate possibility for our results is that P3
subunits might be minimally labeled, thus making it
difficult or impossible to be visualized on autoradiograms. Fig 3 demonstrates that simply reducing the
extent of radioactive labeling still provides an easily
discernable /-subunit band on autoradiograms. This
analysis and the systematic overexposure of autoradiograms make it extremely unlikely that minimally labeled
P3 subunits were present but were undetected.
Various roles have been assigned to channel /3 subunits, including those of stabilizing a-subunit structure39,4"1 and modulating channel kinetiCs.41A42 Although
initial reconstitution studies of purified eel electroplax
sodium channels (which lack /3 subunits) demonstrated
that subunits alone were sufficient to mediate ion
fluxes with the voltage dependence, ion selectivity, and
a
Cohen and Levitt rHI Sodium Channel Characterization
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
pharmacological properties expected of native sodium
channels,43 the rat brain and adult rat skeletal muscle
channels could only be successfully reconstituted as
complexes of a and ,l3 subunits.4044
Recent heterologous expression studies of a-subunit
cRNA in oocytes support the notion that these small
subunits may not be required for the production of fully
functional channels but indicate that they may play a
modulating role in both channel expression and function. Heterologous expression of rat brain and rat
skeletal muscle sodium channels produces channels
with abnormal inactivation kinetics. Coinjection of either low molecular weight mRNA from rat brain or of
cRNA prepared from a subgroup of cDNAs that does
not itself encode for sodium channel a subunits reduces
the time course of channel inactivation to values comparable to those seen in vivo and increases the cell
surface expression of functional sodium channels almost
fourfold.41l42
The recent elucidation of the primary structure and
the functional expression of the f3i subunit of the rat
brain sodium channel has confirmed most of these
suggestions.45 Coexpression of ,13 subunits with a subunits increased the size of the peak inward sodium
current, accelerated its inactivation, and shifted the
voltage dependence of inactivation to more negative
membrane potentials. These results indicate that f1
subunits play a role in sodium channel assembly, expression, and functional modulation. It is tempting to speculate that the normal channel kinetics of rHl cRNA
expressed alone in oocytes is both consistent with the
lack of need for 13 subunits by rHl and with their
absence as demonstrated in this study.
Channel Phosphorylation
Each sodium channel studied to date is capable of
being rapidly and specifically phosphorylated by cAMPdependent PKA. All sodium channels have high levels
of basal phosphorylation, and activation of endogenous
PKA induces phosphorylation of the same residues that
are phosphorylated in vitro.4647 Electrophysiological
effects of PKA activation include regulation of the
surface expression of functional channels,29,48 reduction
in channel conductance,27'49 and shifting of the voltage
dependence of channel inactivation to more negative
potentials.50 The rHl protein contains five potential
sites for PKA phosphorylation: one in the amino terminus, two in the interdomain 1-2 region, and two in the
interdomain 2-3 region. Our results demonstrate that
the rHl channel protein can be phosphorylated by PKA
in at least one of these sites.
Two previous studies address the question of the
biochemical characteristics of sodium channels in mammalian cardiac membranes. In the first study, antibody
SP-19, directed against an 18-residue sequence in the
100% conserved interdomain 3-4 region, was used to
survey the biochemical characteristics of sodium channels in several tissues including rat heart.20 A 230-kD
phosphorylated a subunit containing -"'8 kD of sialic
acid and lacking a 132 subunit was immunoprecipitated
by this antiserum from rat heart. In the second study,
the same investigators specifically immunoprecipitated
a 1, subunit from solubilized cardiac membranes using
antisera directed against purified nerve B13 subunits21;
this finding was recently corroborated by Northern blot
741
analysis of total RNA from rat heart using a rat brain
13,-subunit-specific probe.45 One difference between
these and the present study is that the former studies
used a single antibody against a highly conserved sodium channel sequence, whereas the present study uses
three sequence- and subtype-specific antisera directed
against different regions of the rHi protein. Our subtype-specific antisera confirm that the biochemical characteristics observed are specific for the rHi channel. In
addition, the ability to immunoprecipitate a 81 subunit
from solubilized cardiac membranes or to identify P,subunit mRNA on Northern blots does not identify the
a subunit with which it is associated; the 18 subunit could
be associated either with sodium channels present on
autonomic nerves that innervate the heart or with any
of the other sodium channels whose messages have been
identified in cardiac tissues. Antisera specific for each of
the other transcripts identified in cardiac preparations
will need to be generated to confirm the presence,
determine the tissue localization, and elucidate the
characteristics of each of these distinct channels.
Acknowledgments
This study was supported by a grant to Dr Cohen from the
W.W. Smith Charitable Trust. Dr Cohen is a scholar of the
Pfizer Scholars Program for New Faculty.
References
1. Cohen SA, Barchi RL. Cardiac sodium channel structure and
function. Trends Cardiovasc Med. 1992;2:133-140.
2. Kiyosue T, Arita M. Late sodium current and its contribution to
action potential configuration in guinea pig ventricular myocytes.
Circ Res. 1989;64:389-397.
3. Liu YM, Defelice LJ, Mazzanti M. Na channels that remain open
throughout the cardiac action potential plateau. Biophys J. 1992;
63:654-662.
4. Saint DA, Ju YK, Gage PW. A persistent sodium current in rat
ventricular myocytes. J Physiol (Lond). 1992;453:219-231.
5. Coraboeuf E, Deroubaix E, Coulombe A. Effect of tetrodotoxin on
action potentials of the conducting system in the dog heart. Am J
Physiol. 1979;236:H561-H567.
6. Attwell D, Cohen I, Eisner D, Ohba M, Ojeda C. The steady state
TTX-sensitive ("window") sodium current in cardiac Purkinje
fibres. Pflugers Arch. 1979;379:137-142.
7. Renaud JF, Kazazoglou T, Lombet A, Chicheportiche R, Jaimovich E, Romey G, Lazdunski M. The sodium channel in mammalian cardiac cells: 2 kinds of tetrodotoxin receptors in rat heart
membranes. J Biol Chem. 1983;258:8799-8805.
8. Lombet A, Renaud J, Chicheportiche R, Lazdunski M. A cardiac
tetrodotoxin binding component: biochemical identification, characterization, and properties. Biochemistry. 1981;20:1279-1285.
9. Gellens ME, George AL, Chen LQ, Chahine M, Horn R, Barchi
RL, Kallen RG. Primary structure and functional expression of the
human cardiac tetrodotoxin-insensitive voltage-dependent sodium
channel. Proc Natl Acad Sci U S A. 1992;89:554-558.
10. Rogart RB, Cribbs LL, Muglia LK, Kephart DD, Kaiser MW.
Molecular cloning of a putative tetrodotoxin-resistant rat heart
Na+ channel isoform. Proc Natl Acad Sci U S A. 1989;86:
8170-8174.
11. Cribbs LL, Satin J, Fozzard HA, Rogart RB. Functional expression
of the rat heart I Na+ channel isoform: demonstration of properties
characteristic of native cardiac Nal channels. FEBS Lett. 1990;275:
195-200.
12. White MM, Chen LQ, Kleinfield R, Kallen R, Barchi R. SkM2, a
Na4 channel cDNA clone from denervated skeletal muscle,
encodes a tetrodotoxin-insensitive Nal channel. Mol Pharmacol.
1991;39:605-608.
13. Gautron S, Dossantos G, Pintohenrique D, Koulakoff A, Gros F,
Berwaldnetter Y. The glial voltage-gated sodium channel: cellspecific and tissue-specific messenger RNA expression. Proc Natl
Acad Sci U S A. 1992;89:7272-7276.
14. George AL, Knittle TJ, Tamkun MM. Molecular cloning of an
atypical voltage-gated sodium channel expressed in human heart
742
15.
16.
17.
18.
19.
20.
21.
22.
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
Circulation Research Vol 73, No 4 October 1993
and uterus: evidence for a distinct gene family. Proc Natl Acad Sci
U S A. 1992;89:4893-4897.
Sills MN, Xu YC, Baracchini E, Goodman RH, Cooperman SS,
Mandel G, Chien KR. Expression of diverse sodium channel messenger RNA in rat myocardium: evidence for a cardiac-specific
sodium channel. J Clin Invest. 1989;84:331-336.
Schaller KL, Krzemien DM, Mckenna NM, Caldwell JH. Alternatively spliced sodium channel transcripts in brain and muscle.
J Neurosci. 1992;12:1370-1381.
Rogart R. Sodium channels in nerve and muscle membrane. Annu
Rev Physiol. 1981;51:711-726.
Trimmer JS, Agnew WS. Molecular diversity of voltage-sensitive
sodium channels. Annu Rev Physiol. 1989;51:401-418.
Grant AO. The electrophysiology of the cardiac sodium channel.
Trends Cardiovasc Med. 1991;1:321-330.
Gordon D, Merrick D, Wollner DA, Catterall WA. Biochemical
properties of sodium channels in a wide range of excitable tissues
studied with site-directed antibodies. Biochemistry. 1988:27:
7032-7038.
Sutkowski EM, Catterall WA. Beta-1 subunits of sodium channels
studies with subunit-specific antibodies. J Biol Chem. 1990;265:
12393-12399.
Zwerling SJ, Cohen SA, Barchi RL. Analysis of protease-sensitive
regions in the skeletal muscle sodium channel in vitro and implications for channel tertiary structure. J Biol Chem. 1991;266:
4574-4580.
Cleveland DW, Fischer SG, Kirschner MW, Laemmli UK. Peptide
mapping by limited proteolysis in sodium dodecyl sulfate and
analysis by gel electrophoresis. J Biol Chem. 1977;252:1102-1106.
Cohen SA. Immunocytochemical characterization of the rat
cardiac sodium channel. Circulation. 1991;84(suppl 2):I1-182.
Abstract.
Kraner S, Yang J, Barchi R. Structural inferences for the native
skeletal muscle sodium channel as derived from patterns of
endogenous proteolysis. J Biol Chem. 1989;264:13273-13280.
Cohen SA, Barchi RL. Voltage-dependent sodium channels. In:
Friedlander M, Meckler M, eds. International Review of Cytology,
Volume 137C, Molecular Biology of Receptors and Transporters:
Transporters, Receptors, and Channels. New York, NY: Academic
Press, Inc; 1993:55-103.
Smith RD, Goldin AL. Protein kinase-A phosphorylation
enhances sodium channel currents in Xenopus oocytes. Am J
Physiol. 1992;263(suppl 3, pt 1):C660-C666.
Li M, West JW, Lai Y, Scheuer T, Catterall WA. Functional
modulation of brain sodium channels by cAMP-dependent phosphorylation. Neuron. 1992;8:1151-1159.
Ginty DD, Fanger GR, Wagner JA, Maue RA. The activity of
cAMP-dependent protein kinase is required at a posttranslational
level for induction of voltage-dependent sodium channels by
peptide growth factors in PC12 cells. J Cell Biol. 1992;116:
1465-1473.
Matsuda JJ, Lee H, Shibata EF. Enhancement of rabbit cardiac
sodium channels by beta-adrenergic stimulation. Circ Res. 1992;70:
199-207.
Sorbera LA, Morad M. Modulation of cardiac sodium channels by
cAMP receptors on the myocyte surface. Science. 1991;253:
1286-1289.
Schmidt JW, Catterall WA. Palmitylation, sulfation, and glycosylation of the alpha subunit of the sodium channel: role of posttranslational modifications in channel assembly. J Biol Chem. 1987;
262:13713-13723.
33. Roberts RH, Barchi RL. The voltage-sensitive sodium channel
from rabbit skeletal muscle: chemical characterization of subunits.
JBiol Chem. 1987;262:2298-2303.
34. Cohen SA, Barchi RL. Glycoprotein characteristics of the sodium
channel saxitoxin binding component from mammalian sarcolemma. Biochim Biophys Acta. 1981;645:253-261.
35. Kallen RG, Sheng ZH, Yang J, Chen LQ, Rogart RB, Barchi RL.
Primary structure and expression of a sodium channel characteristic of denervated and immature rat skeletal muscle. Neuron.
1990;4:233 -242.
36. Cohen SA. Immunocytochemical characterization of the rat
cardiac sodium channel. Circulation. 1991;84(suppl 2):I1-182.
Abstract.
37. Recio-Pinto E, Thornhill WB, Duch DS, Levinson SR, Urban BW.
Neuraminidase treatment modifies the function of electroplax
sodium channels in planar lipid bilayers. Neuron. 1990;5:675-684.
38. White MM, Chen LQ, Kleinfield R, Kallen RG, Barchi RL. SkM2,
a Na` channel cDNA clone from denervated skeletal muscle,
encodes a tetrodotoxin-insensitive Na+ channel. Mol Pharmacol.
1991;39:604-608.
39. Tejedor FJ, Mchugh E, Catterall WA. Stabilization of a sodium
channel state with high affinity for saxitoxin by intramolecular
cross-linking evidence for allosteric effects of saxitoxin binding.
Biochemistry. 1988;27:2389-2397.
40. Messner DJ, Feller DJ, Scheuer T, Catterall WA. Functional properties of rat brain sodium channels lacking the beta-1 or beta-2
subunit. JBiol Chem. 1986;261:14882-14890.
41. Auld VJ, Goldin AL, Krafte DS, Marshall J, Dunn JM, Catterall
WA, Lester HA, Davidson N, Dunn RJ. A rat brain Nal channel
alpha subunit with novel gating properties. Neuron. 1988:1:
449-461.
42. Zhou J, Potts FF, Trimmer JS, Agnew WA, Sigworth FJ. Multiple
gating modes and the effect of modulating factors on the g1
sodium channel. Neuron. 1992;7:775-785.
43. Duch S, Levinson SR. Neurotoxin-modulated uptake of sodium by
highly purified preparations of the electroplax tetrodotoxinbinding glycopeptide reconstituted into lipid vesicles. JMembr Biol.
1987;98:43-50.
44. Weigele J, Barchi R. Functional reconstitution of the purified
sodium channel from rat sarcolemma. Proc Natl Acad Sci U S A.
1982;79:3651-3655.
45. Isom LL, Dejongh KS, Patton DE, Reber BFX, Offord J, Charbonneau H, Walsh K, Goldin AL, Catterall WA. Primary structure
and functional expression of the beta 1-subunit of the rat brain
sodium channel. Science. 1992;256:839-842.
46. Rossie S, Gordon D, Catterall WA. Identification of an intracellular
domain of the sodium channel having multiple cyclic AMPdependent phosphorylation sites. JBiol Chem. 1987;262:17530-17535.
47. Rossie S, Catterall WA. Phosphorylation of the alpha subunit of
rat brain sodium channels by cAMP-dependent protein kinase at a
new site containing Ser686 and Ser687. J Biol Chetn. 1989;264:
14220-14224.
48. Taouis M, Sheldon RS, Hill RJ, Duff HJ. Cyclic AMP-dependent
regulation of the number of [3H] batrachotoxinin benzoate binding
sites on rat cardiac myocytes. J Biol Chem. 1991;266:10300-10304.
49. Seelig TL, Kendig JJ. Cyclic nucleotide modulation of sodium and
potassium currents in the isolated node of ranvier. Brain Res.
1982;245:144-147.
50. Coombs J, Scheuer T, Rossie S, Catterall W. Evidence that cyclic
AMP-dependent phosphorylation promotes inactivation in
embryonic rat brain cells in primary culture. Biophys J. 1988;
53:542a. Abstract.
Partial characterization of the rH1 sodium channel protein from rat heart using
subtype-specific antibodies.
S A Cohen and L K Levitt
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Circ Res. 1993;73:735-742
doi: 10.1161/01.RES.73.4.735
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1993 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/73/4/735
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/