 |
INTRODUCTION |
L-type calcium channels form a subclass of voltage-gated calcium
channels. Three different isoforms of 
1 subunits
(1S, 1C, and 1D) serve as
pore-forming subunits of the three groups of L-type channels: class S,
class C, and class D. Two size forms for each of these isoforms have
been detected in native tissues (1-5), and in each case, the shorter
forms correspond to proteins modified at the C terminus. For the
1D isoform, alternative splicing of mRNA may give
rise to the short and long forms (6). However, this does not appear to
be the case for the 1C and 1S isoforms, and it appears that the shorter forms of the 1C and
1S proteins are generated by a post-translational,
proteolytic processing event. In the case of these two proteins, it
appears that processing would produce C-terminal fragments of 30-50
kDa (1-5). While little is known about the fate of the cleaved
C-terminal domain, results obtained from immunofluorescence studies
have suggested that the C terminus of the 1C subunit is
present in stoichiometric amounts and co-localized with the body of the
1C subunit and the 
subunit (5). This suggests the
possibility that the C-terminal fragments may remain functionally
associated with the channels.
The C-terminal region of the 1C subunit is involved in
many important processes which regulate the class C calcium channel. The major site of
PKA1-mediated phosphorylation
of the 1C subunit has been shown to be
Ser1928 in the C terminus (4, 7, 8) and the C-terminal
truncated 1C subunit lacks Ser1928 and is
not a substrate for phosphorylation by PKA in vitro (9-11). In addition, the 1C C terminus appears to play a role in
inhibiting channel function. Removal of up to 70% of the C terminus
from the full-length 1C protein results in an increase
in open probability when compared with the full-length channel (12).
The C terminus of the 1C subunit also has been shown to
be involved in calcium-dependent inactivation (13-15), and
to bind the calcium-binding proteins calmodulin (14, 15) and sorcin
(16). Thus, the C terminus of 1C is a very important
region for the regulation of the calcium channel, and the proteolytic
processing of the C terminus could conceivably impact many aspects of
channel function.
Since the results from immunocytochemical analysis of the C terminus of
1C in intact cardiac myocytes suggest that this domain is present and co-localizes with the "body" of the
1C subunit and the 2 subunit (5), it is
possible that the 1C subunit is cleaved in
vivo and that the C-terminal fragment of ~50 kDa remains
associated with the 190-kDa body or with other channel subunits or
unknown proteins. In order to begin to study this hypothesis, we have
developed approaches to study the processing of an heterologously
expressed 1C subunit and the properties of the
C-terminal region of this protein.
| |
EXPERIMENTAL PROCEDURES |
Materials--
The catalytic (C)-subunit of PKA was purified to
homogeneity from bovine heart as described previously (17). The various proteases used in these experiments were purchased from Sigma. All
other reagents were purchased from standard sources. The Card I and
Card C antibodies have been previously described (18). The epitopes for
these polyclonal antibodies are shown in Fig. 2A. The
anti-Myc monoclonal mouse antibody, 9E10 (ATCC Cell Lines), was a
generous gift from Dr. Jeffrey Nye.
Isolation of Rabbit Ventricular Myocytes and Preparation of
Lysates and Purified Channels--
Adult rabbit ventricular myocytes
were isolated using standard procedures (19) as described previously
(5). Myocytes were lysed using a Brinkman Polytron homogenizer, and
these lysates were used in the in vitro proteolysis assay as
described below. L-type calcium channels were purified from frozen
rabbit hearts as described (5).
Mutation and Expression of Cardiac Calcium Channel Proteins in
Heterologous Systems--
Mutants of the 1C subunit
were generated with the Transformer Mutagenesis Kit
(CLONTECH) using the rabbit 1C
subunit cDNA. The deletion of amino acids 1966-2004 in the rabbit
1C subunit (1C
PRD) was generated with
three primers: one selection primer and two mutant primers, one which
generated a new NheI site at the C-terminal end of the PRD
and one which created the new restriction site at the N-terminal end of
the PRD. The new NheI sites were digested to remove the PRD
sequence, and the vector was then religated to form the PRD deletion
mutant construct. The wild-type (WT) and mutant cDNAs of the
1C subunit were subcloned into the pCR3 mammalian
expression vector (Invitrogen) and were coexpressed with the rat
2a subunit (and 2
where indicated) in
tsA cells by CaPO4 precipitation (18). Cell membranes were
prepared 48 h post-transfection. Cells were pelleted by
centrifugation at 1000 × g for 10 min. The cell pellet
was resuspended in ice cold "homogenization buffer" (50 mM Tris-HCl, pH 7.4, 2 mM EDTA, and 2 mM EGTA). Cells were lysed using a Brinkman Polytron
homogenizer (power setting 5 or 6, 3 pulses of 10 s each). The
homogenate was then centrifuged at 1,000 × g for 5 min
to remove nuclei and unbroken cells. The supernatant after this
low-speed spin was centrifuged at 100,000 × g for 20 min to separate the membrane/particulate fraction (pellet) from the
soluble/cytosolic (supernatant) fraction. Finally, the pellet was
resuspended in a minimal volume of homogenization buffer and protein
concentration was determined prior to use in the experiments described below.
Baculovirus-mediated expression of the 1C subunit with
or without and coexpression of the 2 and
2a subunits in Sf9 insect cells was performed as
described (20). In vitro translated (IVT) 1C
C-terminal proteins were expressed from cDNAs subcloned into the
pCR3 vector using the TNT-T7 Quick coupled transcription and translation system (Promega). A fusion protein expression vector, pCR3-His/Myc, was derived from pCR3 by insertion of both 6-His and Myc
tags fused at the N terminus. IVT proteins were labeled during
synthesis with [35S]methionine and used in the GST
pull-down experiments detailed below.
In Vitro Proteolysis of the 1C Subunit--
Crude
membrane preparations from tsA cells or Sf9 insect cells
expressing either the 1C subunit alone or with
2a subunits were used for cleavage assays as specified.
Cleavage reactions contained membrane preparations, indicated protease,
and "cleavage buffer" (25 mM Tris-HCl, pH 7.8, and 2.5 mM dithiothreitol). Cleavage reactions with exogenously
added protease (100 µl) were performed at 25 °C with various
concentrations of protease and for various amounts of time as
specified. For proteolysis of the expressed channel with cardiac
myocyte lysates, the expressed 1C subunit was added to
whole cell lysates from isolated cardiac myocytes for either 10 min at
4 °C or overnight at 37 °C prior to the addition of Laemmli
buffer (21) followed by SDS-PAGE and Western blot analysis (18).
Where indicated, cleavage reactions were stopped by addition of 17.4 µg/ml phenylmethylsulfonyl fluoride and 100 µM
tosylamido-2-phenylethyl chloromethyl ketone and centrifuged at
100,000 × g for 20 min to separate the membrane and
cytosolic fractions. Membranes were washed twice by resuspending the
membrane pellet in 100 µl of cleavage buffer followed by
centrifugation to isolate the membrane fraction as described above.
In Vitro Phosphorylation of the C-terminal Fragments of the
1C Subunit with PKA--
In vitro
phosphorylation of the 1C subunit prior to chymotrypsin
cleavage was performed as described previously (20). SDS-polyacrylamide gel electrophoresis (21) and immunoblotting was performed using standard methods (18). Phosphorylated peptides were analyzed using a
Storm PhosphorImager.
GST Pull-down Experiments with the IVT 1C
C-terminal Proteins--
GST pull-down experiments were performed
using various GST-SH3 fusion proteins containing the SH3 domains from
Src, Lyn, Hck, Grb, or amphiphysin (22, 23) or the SH3 domain of the
2a subunit. The 2a-SH3-GST fusion protein
was generated by polymerase chain reaction amplification of the SH3
domain previously identified in the rat 2a subunit (24),
and the polymerase chain reaction product was subcloned into the
pGEX-4T3 GST bacterial expression vector (Amersham Pharmacia Biotech)
and purified as outlined by the manufacturer. In each reaction, the
indicated in vitro translated C-terminal protein from the
1C subunit was diluted in 200 µl of "binding
buffer" (150 mM NaCl, 50 mM Tris-HCl, pH
7.4), added to 20 µl of glutathione-Sepharose beads precoupled to
either GST alone (control) or to one of the various GST-SH3 fusion
proteins, and incubated for 5-16 h at 4 °C with agitation. The
GST-SH3 beads were washed 5 times with 1 ml each of binding buffer
containing 0.1% Triton X-100, and after the last wash, 
volume of Laemmli buffer was added to each. The IVT proteins which bound to the GST-SH3/glutathione-Sepharose beads were separated by
SDS-PAGE, transferred to nitrocellulose, and the
35S-labeled IVT proteins were detected by autoradiography.
Competition of the GST-SH3 binding to the IVT 1C
C-terminal proteins was performed using a non-radiolabeled IVT
1C PRD protein. The non-labeled PRD was immunoblotted
using the anti-Myc antibody (1:50 dilution) as the primary antibody and
anti-mouse horseradish peroxidase-coupled antibody as the secondary
antibody. The amount of the non-labeled PRD was determined by
comparison of the immunoreactivities of the non-labeled and the
35S-labeled 1C PRD. The non-labeled PRD was
added to the GST-SH3 binding reactions at 5 times the concentration of
each of the various IVT 35S-labeled C-terminal proteins,
and the binding reactions were performed as described above.
Electrophysiological Assays--
TsA cells were transiently
co-transfected with the wild-type (WT) rat 2a subunit
and the WT 1C subunit, 1C1905, or
1CPRD. Each construct was in the pCR3 vector and 3 µg of each was used per 10-cm plate, along with the CD8 reporter
vector, 
H3-CD8 (0.5 µg/10 cm plate) (25), using the Effectene
transfection kit following the manufacturer's recommended protocol
(Qiagen). At 38-40 h following transfection, cells were replated on
3-cm collagen-coated plates (Sigma). Transfected cells were visualized by using the anti-CD8 antibody-coated Dynabeads (Dynal) (25).
For the measurement of Ba2+ currents through L-type calcium
channels, the external solution consisted of 10 mM
BaCl2, 105 mM NaCl, 25 mM CsCl, and
10 mM Hepes, pH 7.4. The pipette solution was composed of
100 mM cesium aspartate, 40 mM CsCl, 1 mM MgCl2, 2 mM Mg-ATP, 0.5 mM GTP, 5 mM EGTA, and 5 mM Hepes,
pH 7.4. Membrane currents were measured in the whole cell configuration
of the patch clamp technique using fire-polished borosilicate glass
pipettes (GF-150-10, Warner Instrument Corp.) generated with a
horizontal puller (P-95 Fleming and Poulsen) with a final resistance
between 2-4 megohm. Membrane currents were amplified using a patch
clamp amplifier (Axopatch 200, Axon Instruments) and analog-filtered using a low-pass Bessel filter (1-3 kHz corner frequency). Data were
digitally stored using an IBM compatible PC equipped with a
hardware/software package (ISO2 by MFK) for voltage control, data
acquisition, and data evaluation. Cells were clamped at 
90 mV and
voltage pulses (test pulses) (100-ms duration) to +10 mV were applied
every 10 s in order to activate L-type calcium channels. Current-voltage relationships were measured by varying the potential of
the test pulses from 40 to +30 mV after reaching a steady state of
the current amplitude. Capacitative currents due to recharging the cell
membrane were compensated and leak currents were subtracted. Currents
were normalized to capacitance. The summarized data were pooled from at
least two different transfections and are expressed as mean ± S.E.
| |
RESULTS |
In Vitro Proteolysis of an Heterologously Expressed
1C Subunit--
In order to begin to understand the
processing of the native cardiac 1C subunit from the
full-length, 240-kDa protein to the truncated, 190-kDa form, and to
define the properties of the C-terminal fragments resulting from the
proteolytic cleavage of the 1C subunit, we subjected an
expressed, full-length 1C subunit to a variety of
conditions designed to induce proteolysis. The 1C
subunit was heterologously expressed in Sf9 insect cells by baculovirus infection, which resulted in the expression of only the
full-length, 240-kDa protein in Sf9 cell membranes (Fig.
1) (20).
View larger version (40K):
[in this window]
[in a new window]
|
Fig. 1.
The expressed 1C subunit was
not cleaved into a 190-kDa form upon exposure to lysates of cardiac
myocytes? Crude membrane preparations (100 µg of
protein/reaction) from Sf9 cells expressing either the
1C subunit alone (lanes 1-7) or the
1C, 2, and 2a subunits
(lanes 8-14) were prepared and washed to remove protease
inhibitors and incubated alone (lanes 1, 6, 8, and
13) or with lysates of freshly isolated cardiac myocytes
(300 µg of whole cell protein/reaction) (lanes 2-5, 7, 9-12, and 14). Isoproterenol ("ISO", 5 µM, lanes 3 and 10), forskolin
("FSK", 20 µM, lanes 4 and
11), or CaCl2 (5 mM, lanes
5 and 12) were also included in the reactions. The
reactions were incubated for 10 min on ice (lanes 1-5 and
8-12) or overnight (O/N) at 37 °C
(lanes 6, 7, 13, and 14) prior to the addition of
Laemmli buffer. After SDS-PAGE, proteins were analyzed by Western blot
analysis using the Card I antibody.
|
|
Although the native cardiac and brain 1C protein has
been isolated in the presence of a battery of protease inhibitors (2, 5, 9), it was possible that the truncation was an artifact resulting
from uncontrolled proteolysis during protein isolation. Therefore,
initially, the heterologously expressed, full-length 1C
subunit was exposed to extracts from lysed cardiac myocytes to
determine if the proteolytic event(s) which processed the cardiac 1C subunit could be reproduced in vitro with
the heterologously expressed 1C subunit. The myocyte
lysates were prepared after stimulating the myocytes with several
reagents that might conceivably alter proteolytic activity. The results
showed that under a variety of conditions the heterologously expressed
1C subunit, when expressed either alone or with the
2 and 2 subunits, was resistant to any
proteases which may have been present and active in the myocyte extracts (Fig. 1). Virtually no change in the mobility of the expressed
1C subunit was observed when this protein was incubated with or without myocyte lysates for 10 min at 4 °C (Fig. 1). Only when the reactions were allowed to proceed at 37 °C overnight was
there a marked loss of the 240-kDa full-length 1C
protein when the myocyte lysate was added to the expressed
1C (Fig. 1). However, neither under these circumstances
nor during the shorter incubations was a C-terminal truncated 190-kDa
form of the 1C subunit generated. If this form of the
protein was generated, it would have been readily detected by the Card
I antibody as this antibody recognizes an internal epitope in the
linker between domains II and III (Fig.
2B). These results suggested
that the proteolytic cleavage of the native 1C subunit
isolated from cardiac myocytes was not an artifact resulting from the
isolation process because exposure of the heterologously expressed
1C subunit to the same cardiac, cellular proteases which
the native 1C subunit would have been exposed to upon
cell lysis did not cause cleavage of the expressed 1C
subunit.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 2.
In vitro proteolysis of the
1C subunit with
chymotrypsin. A, crude membranes from Sf9
cells expressing the 1C subunit were cleaved in
vitro by addition of 1 or 10 milliunits of chymotrypsin/100-µl
reaction for 5 min at room temperature. Following proteolysis, the
membrane proteins were separated by SDS-PAGE and detected by
immunoblotting with either the C-terminal specific Card C antibody
(left panel, Sf9) or the internal Card I antibody
(right panel, Sf9). The native 190-kDa
1C subunit isolated from rabbit heart was also analyzed
by SDS-PAGE and immunoblotting with Card I (right panel,
Native). Note that this protein migrated at the same molecular
mass as the chymotrypsin-truncated 190-kDa product of the expressed
1C subunit (right panel, compare lanes
2 and 3). Chymotrypsin cleavage of the expressed
1C subunit also generated several relatively stable
products of the C terminus of approximately 30-56 kDa (left
panel). Note that while the immunoblot with the Card C antibody
(left panel) suggests incomplete digestion with
chymotrypsin, no staining of the full-length 1C subunit
is observed under the same conditions with the Card I antibody due to
the 3-8 times lower sensitivity of this antibody in Western blotting
(5). B, schematic representation of the cardiac
1C subunit showing the full-length 240-kDa protein, the
C-terminal truncated 190-kDa protein and the 30-56-kDa fragments
generated by in vitro proteolysis with exogenously added
chymotrypsin. The location of the Card I and Card C epitopes are also
indicated.
|
|
Since the cardiac myocyte lysates did not induce truncation of the
expressed, full-length 1C subunit, other means were used in an attempt to duplicate the proteolytic processing seen with the
native 1C subunit. The baculovirus-expressed
1C subunit in Sf9 cell membranes was subjected to
limited proteolysis with a variety of exogenous proteases in an attempt
to reproduce the cleavage which occurred on the cardiac
1C subunit, and to potentially enable the development of
a model system with which to study the properties of the cleaved
channel products. Several different proteases were used in these
in vitro proteolysis assays including chymotrypsin, trypsin,
subtilisin, proteinase K, carboxypeptidase, papain, thermolysin, and
calpain. The latter has been suggested to cleave the 1C
subunit in hippocampal neurons (26). Many of these proteases did cleave
the expressed 1C subunit (not shown), but only
chymotrypsin, an extracellular serine protease that most likely is not
responsible for the cleavage of the channel in native systems, produced
the expected 190-kDa C-terminal truncated product of the
1C subunit (Fig. 2). Limited proteolysis with
chymotrypsin generated a 190-kDa protein lacking its C terminus as
evidenced by the reactivity of the 190-kDa fragment to the Card I
antibody but not to the Card C antibody, which recognizes the final 14 amino acids of the C terminus (Fig. 2, A and B).
The 190-kDa product of the 1C subunit generated by
chymotrypsin had the same apparent molecular mass as the truncated,
native 1C subunit isolated from cardiac tissue (Fig.
2A, right). Another cleavage product of ~90 kDa was
recognized by the Card I antibody but not the Card C antibody, suggesting that it represented a cleavage product from an internal domain of 1C. In addition, chymotrypsin generated
C-terminal fragments of approximately 30-56 kDa (Fig. 2A,
left). The C-terminal fragment of ~56 kDa corresponded to the
expected size if the C-terminal domain was cleaved in a single region,
and its immunoreactivity to Card C suggested that it contained an
intact C terminus. The C-terminal 30-48-kDa fragments also contained
intact C termini as evidenced from their immunoreactivity toward Card
C. Since chymotrypsin could produce proteolytic products that were
similar to those expected from the native, purified 1C
subunit, we reasoned that in vitro proteolysis with
chymotrypsin could allow for the development of a model system to study
the properties and potential interactions of the cleavage products.
The C-terminal Fragments Resulting from Proteolysis of the
1C Subunit Remained Associated with the
Membranes--
The solubility of the chymotryptic C-terminal fragments
was determined. The C-terminal fragments should contain no predicted membrane-spanning regions and are hydrophilic (27). However, after
centrifugation of the chymotryptic reaction mixtures, the chymotrypsin-cleaved C-terminal fragments of the 1C
subunit remained associated with the membrane fraction and were not
released into the soluble fraction (Fig.
3).
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 3.
Membrane association of the C-terminal
chymotryptic fragments of
1C. Crude membranes
from Sf9 cells expressing the 1C subunit were
proteolytically cleaved in vitro with chymotrypsin for the
indicated times. Subsequently, the proteins were either analyzed
directly (left panel) or were centrifuged to separate the
membrane and cytosolic fractions and then analyzed by SDS-PAGE and
immunoblotting (right panel). Immunoblot analysis with the
C-terminal Card C antibody demonstrated that the 30-56-kDa C-terminal
fragments of the 1C subunit all remain associated with
the membrane pellets (P) following centrifugation and were
absent from the soluble fractions (S) (right
panel).
|
|
Solubilization studies were performed following limited proteolysis
with chymotrypsin but prior to centrifugation to separate the membrane
and soluble fractions. While the C-terminal fragments remained
associated with the membrane fraction under control conditions, several
of the solubilization conditions that were tested released the
C-terminal fragments from the membrane fraction to the cytosolic fraction. Most notably, addition of 500 mM NaCl to the
cleavage reactions caused the partial release of the C-terminal
fragments from the membrane (Fig. 4). The
solubility of the C-terminal peptides of the 1C subunit
in NaCl suggested that the membrane association was likely to involve a
protein-protein interaction. The C-terminal fragments could be binding
to the truncated 1C subunit present in the membrane, or
with an unknown protein in the crude membrane fractions.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 4.
Solubilization of the membrane-associated
C-terminal fragments. Membrane preparations from
1C-expressing Sf9 cells were subjected to
in vitro proteolysis with chymotrypsin, solubilized in the
presence of 150 or 500 mM NaCl or buffer alone, and
separated into membrane (P) and soluble (S)
fractions by centrifugation, followed by SDS-PAGE and immunoblotting
with Card C.
|
|
Phosphorylation of the C-terminal Fragments of the
1C Subunit Did Not Affect Association with the
Membranes--
An important consideration for understanding the
cleavage of the 1C subunit is that a site that has been
identified to play a role in the regulation of channel activity by PKA
(7, 8) is predicted to reside in the C-terminal domain that is cleaved. In order to test whether this site at Ser1928 was present
in the C-terminal fragments and to test how phosphorylation might
affect interaction of the fragments with the membrane, the chymotryptic
fragments were phosphorylated in vitro with PKA. In the
first of two protocols, crude membrane preparations containing the
baculovirus-expressed 1C subunit were phosphorylated
with PKA and [
-32P]ATP, solubilized,
immunoprecipitated with the Card I antibody, and then subjected to
limited proteolysis with chymotrypsin (Fig. 5A, lanes 1 and 3).
This procedure allowed for the isolation of C-terminal, chymotryptic
fragments of about 48-56 kDa which were indeed phosphorylated by PKA
(Fig. 5A). Both the immunoblot and the autoradiograph from
this procedure also showed a fraction of the 1C subunit
which remained uncleaved by chymotrypsin and was phosphorylated by PKA.
Similar results were obtained when crude membranes containing the
1C subunit were phosphorylated with PKA and then cleaved
with chymotrypsin prior to immunoprecipitation of the C-terminal
fragments with the Card C antibody (Fig. 5A, lanes 2 and
4). A smaller amount of the phosphorylated 30-kDa C-terminal
fragment was also isolated under these conditions (Fig. 5A, lanes
2 and 4). These results demonstrated that a PKA
phosphorylation site, presumably at Ser1928 (4, 7, 8, 28),
was present in the C-terminal fragments isolated through in
vitro cleavage of the 1C subunit.
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 5.
The chymotryptic fragments generated from
the 1C subunit
are phosphorylated by PKA in vitro. A,
crude membranes from Sf9 cells expressing the 1C
subunit were phosphorylated in vitro with PKA, cleaved with
chymotrypsin (10 milliunits/100 µl) either prior to
immunoprecipitation with the C-terminal Card C antibody (lanes
1 and 3) or following immunoprecipitation with the
internal Card I antibody (lanes 2 and 4). The
immunoprecipitated proteins were analyzed by SDS-PAGE and
immunoblotting with the rabbit Card C antibody (left panel)
and by autoradiography (right panel). B,
membranes were phosphorylated in the presence (PKA) or absence of PKA
(no PKA) and cleaved with chymotrypsin as in A. The
reactions were centrifuged to separate membrane-associated
(P) and soluble proteins (S) and analyzed by
Western blotting with Card C.
|
|
We tested whether phosphorylation modified the association of the
fragments with the membranes. Sf9 membranes expressing the 1C subunit were phosphorylated in vitro in
the presence or absence of PKA, cleaved with chymotrypsin, and
centrifuged to separate membrane-associated and soluble proteins.
Immunoblot analysis with Card C showed that the C-terminal 30-56-kDa
fragments of the 1C subunit remained associated with the
membrane fraction either in the presence or absence of PKA-mediated
phosphorylation (Fig. 5B). This demonstrated that
phosphorylation of the C-terminal fragments of the 1C
subunit did not markedly alter the ability of these peptides to
associate with the membranes, however, this does not rule out the
possibility that phosphorylation may play a more subtle role in the
association of the C terminus with the membrane.
Determination of the Region within the C Terminus of the
1C Subunit Responsible for Membrane
Association--
More extensive cleavage of the 1C
subunit with chymotrypsin was performed to identify the smallest
fragment that could remain associated with the membrane. The
1C subunit expressed either by baculovirus infection of
Sf9 insect cells or by transient transfection of mammalian
tsA201 cells was cleaved in vitro using a higher concentration of chymotrypsin. Following separation of the membrane and
cytosolic fractions by centrifugation, immunoblot analysis with the
Card C antibody showed that both the Sf9-expressed (Fig. 6B) and mammalian-expressed (not shown)
1C subunits were cleaved into several C-terminal
fragments including one major peptide which was smaller than the 30-kDa
peptide observed previously. When these reactions were subjected to
centrifugation to separate membrane-bound and soluble fragments, the
majority of the 30-56-kDa fragments remained associated with the
membranes, while the smaller fragment of ~24-25 kDa was present
almost exclusively in the soluble fraction (Fig. 6B). Since
the Card C antibody was directed against the extreme C terminus of the
1C subunit sequence, the approximate amino acid sequence
which encompassed the region necessary for membrane association was
estimated from the apparent molecular masses of the C-terminal
fragments associated with the membrane versus those present
in the soluble fraction. These results suggested that a region
important for membrane association of the C terminus was contained
approximately between amino acids 1900 and 2000 of the
1C subunit sequence.
View larger version (32K):
[in this window]
[in a new window]
|
Fig. 6.
A proline-rich domain in the C terminus of
the 1C subunit
was important for association of the C terminus with the membrane.
A, shown is a diagram depicting the location of the
1C PRD and the alignment between the 1C
PRD and the first PRD from dynamin. A high degree of identity exists
between the two regions including conservation of most of the proline
residues (shown in bold type). Underlined
sequences represent potential class II SH3-binding domains.
B, crude Sf9 membranes expressing the
1C subunit were incubated in the presence (right
lanes) or absence (left lanes) of a high concentration
of chymotrypsin (50 milliunits/100 µl), centrifuged to separate
membrane pellets (P) from soluble peptides (S)
and analyzed by immunoblotting with the Card C antibody (right
lane). C, the WT 1C subunit and a PRD
deletion mutant (1CPRD) were expressed in tsA cells
with the 2a subunit and subjected to proteolysis with
chymotrypsin and membrane fractionation as above. The solubility of the
CT fragments from 1CPRD were analyzed in the presence
and absence of 500 mM NaCl as indicated. The proteins were
separated on an SDS gel containing a gradient of 5-15%
acrylamide.
|
|
A Proline-rich Domain in the C Terminus of the 1C
Subunit Is Important for Membrane Association--
A Blast data base
search using the sequence contained within amino acids 1900-2000
revealed the presence of a sequence with a high degree of identity to
the first proline-rich domain (PRD) of dynamin (Fig. 6A)
(29). The 1C PRD was located between residues 1973 and
2001 of the 1C subunit. This amino acid sequence
contained a large number of conserved proline residues as well as a
class II SH3 binding motif (Fig. 6A). The importance of this
PRD region to the membrane association of the 1C C
terminus was investigated by mutagenesis.
Deletion mutagenesis to remove the entire PRD was used to assess the
significance of the 1C PRD. The WT 1C
subunit or the PRD deletion mutant, 1CPRD, were
transiently expressed with the 2a subunit in tsA cells
and subjected to chymotryptic cleavage in vitro followed by
separation of the membrane and soluble fractions. Immunoblot analysis
using the Card C antibody demonstrated that the C-terminal fragments
from 1CPRD were no longer exclusively located in the
membrane fraction (Fig. 6C). Chymotryptic cleavage of
1CPRD generated the same major fragments of the C
terminus as seen with the WT 1C subunit, although the
size of each fragment was smaller due to the deletion. However, in
contrast to the WT 1C subunit, the C-terminal fragments
of 1CPRD were present in greater quantities in the
cytosolic fraction than the C-terminal peptides generated from the WT
1C subunit (Fig. 6C). The largest C-terminal
fragment of 1CPRD was present in both the membrane and soluble fractions in similar amounts, while the smaller fragments appeared to be located almost completely in the soluble fraction. A
similar pattern of membrane-associated or soluble C-terminal fragments
from 1CPRD was observed in either buffer alone or in
the presence of 500 mM NaCl (Fig. 6C). These
results demonstrated the importance of the 1C PRD for
the association of the C terminus of the 1C subunit with
the membrane fraction. In addition, the solubilization experiments
involving the PRD mutant suggested that one population of C-terminal
peptides generated by chymotrypsin was both soluble in high salt and
dependent upon the PRD for membrane association, while another
population was neither salt-soluble nor dependent upon the PRD for
membrane association.
The 1C PRD Binds GST Fusion Proteins Containing SH3
Domains--
We tested if the 1C PRD could serve as a
ligand for known SH3 domains. Various C-terminal (CT) constructs of the
1C subunit were expressed and labeled with
[35S]methionine using an in vitro
transcription/translation system (Fig.
7B, left). The CT proteins CT6
and CT4 possessed the PRD while CT7 did not (Fig. 7A). The
SH3 domains from Src, Lyn, Hck, Grb, and amphiphysin were expressed as
GST-SH3 domain fusion proteins (Fig. 7B, right), bound to
glutathione-Sepharose beads and used to pull-down the IVT proteins
representing different regions of the 1C C terminus
(Fig. 7, A and B). The nonspecific binding of the
IVT 1C C-terminal proteins to the negative control (GST alone) was negligible (Fig. 7C). CT4 and CT6 bound to the
GST-SH3 fusion proteins while CT7 showed no binding to the GST-SH3
proteins (Fig. 7C). CT6 and CT4 appeared to bind more to the
SH3 domains of Src, Lyn, and Hck than to the SH3 domains of Grb and
amphiphysin. The overlapping region in CT6 and CT4 contained residues
1909-2024 of the C terminus, which included the PRD, while CT7 lacked
the PRD. Therefore, a smaller IVT protein (amino acids 1964-2023) containing the PRD of the 1C subunit (1C
PRD) was generated (Fig. 7B) for use in the pull-down
experiments. The 1C PRD bound the same SH3 domains to
about the same extent as the larger, C-terminal IVT proteins (Fig.
7C). Thus, it appeared that the PRD region of the
1C C terminus was able to serve as a ligand for several known SH3 domains in pull-down experiments.
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 7.
Interaction of the
1C PRD with SH3 domains.
A, schematic representation of the full-length
1C subunit and the in vitro translated
constructs CT4, CT6, and CT7. The PRD domain is present in the CT4 and
CT6 IVT protein, but not in the CT7 protein. B, the IVT
1C C-terminal proteins (left) and GST fusion
proteins containing various SH3 domains (right) were
visualized by autoradiography (IVT) or by Coomassie staining (GST).
C, GST pull-down experiments were performed using either the
GST control (GST) or the GST-SH3 domains from Src, Lyn, Hck,
Grb, or amphiphysin (Amph). D, GST pull-downs
were performed using the GST-SH3 domain from the 2a
subunit (2-SH3), in the presence or absence of a 5-fold
excess of a non-radiolabeled IVT PRD protein to the pull-down reactions
(2-SH3 + PRD).
|
|
We used similar approaches to test if an SH3 domain that has been
identified in the 2 subunit (24) could interact with the
1C PRD. A fusion protein of the SH3 domain of the
2a subunit and GST (GST-2a-SH3) was
generated (Fig. 7B, right), and used in pull-down
experiments with the CT proteins. CT4, CT6, and the 1C
PRD all bound to the GST-2a- SH3 protein while CT7 did
not (Fig. 7D). In order to determine the selectivity of the
interaction between the 1C CT proteins and the
GST-2a-SH3 protein, a non-radiolabeled IVT
1C PRD was added to the binding reactions to compete
with the binding of the IVT proteins to the 2-SH3
domain. Addition of an excess of the non-labeled 1C PRD
to the reactions reduced the binding of CT6, CT4, and 1C
PRD to the GST-2a-SH3 protein (Fig. 7D,
2-SH3 + PRD). These results indicated that the PRD was
responsible for the binding seen between the 1C
C-terminal proteins and the 2a-SH3 domain.
The PRD Participates in the C-terminal Associated Inhibition of
Channel Currents--
Truncation of ~300-500 amino acids from the C
terminus of 1C has been demonstrated to lead to an
increase in calcium channel currents (12), suggesting that the C
terminus of the full-length channel is inhibitory to channel function.
In order to determine if the PRD participates in this inhibitory
function, we measured currents through 1CPRD as well
as through the 1C1905 subunit, which was truncated at
residue 1905 and lacked ~50 kDa of the C terminus. This latter mutant
protein had a similar mobility on SDS gels as that of the ~190-kDa
truncated 1C that is isolated from native systems and to
that produced by chymotrypsin cleavage (Fig. 2). Both mutants were
co-expressed with the rat 2a subunit in tsA201 cells and
barium currents through the channels were compared with those produced
from wild-type 1C and 2a subunits. In
order to control for variations in expression levels of the different
channel subunits, the experiments for all three 1C constructs were performed in parallel transfections, and immunoblotting was performed in parallel using the same batch of transfected cells. In
all experiments shown here the expression of the wild-type 1C was slightly greater than that of the two mutant
1C mutants (Fig.
8A). Interestingly, the
amplitudes of barium currents through channels containing either mutant
1 subunit (1C1905: 27.2 ± 3.3 pA/pF at 0 mV, n = 8 and 1CPRD:
22.4 ± 3.9 pA/pF at 0 mV, n = 9) were similar in
size, but significantly larger than those recorded from channels
containing wild-type 1 subunits (11.5 ± 2.4 pA/pF
at 0 mV, n = 11; Fig. 8, B and
C). These observations confirm the results of Wei et
al. (12) that the C terminus participates in inhibiting channel
function. In addition, they demonstrate that the PRD of the full-length
channel is important for this inhibitory function.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 8.
Effects of C-terminal truncation and deletion
of the 1C PRD on channel
function. A, immunoblot analysis of equivalent amounts
of membrane protein from cells expressing either the WT
1C, 1C1905, or
1CPRD. The proteins were separated on an SDS gel
containing 7% acrylamide. B, voltage-activated barium
currents from TsA cells transiently transfected with the
2a subunit and the wild-type 1C,
1C1905, or 1CPRD subunit were
measured using the whole cell configuration of the patch clamp
technique. Representative current traces in response to a
depolarization from 90 to 0 mV for each of the three different
1C subunits are shown. C, the current-voltage
relationship of voltage-activated barium currents (peak currents) for
each of the indicated conditions were determined by depolarizing the
membrane from the holding potential of 90 mV to test potentials
ranging from 40 to +30 mV.
|
|
| |
DISCUSSION |
Truncation of the C terminus of voltage-dependent
calcium channel 1 subunits occurs in several tissues
including skeletal muscle, heart, and brain (1, 2, 5, 9, 30). An
important question regarding truncation of the 1 subunit
is whether the proteolytic processing is a physiologically relevant
event that occurs in intact cells or an artifact that occurs when the
proteins are isolated. While it remains difficult to definitively prove that truncation of the 1 subunit is not an artifact,
evidence provided both here and previously suggests that the processing of the 1C subunit may not be an artifact. First, the
heterologously expressed 1C subunit is a full-length,
240-kDa protein (18, 20) and does not exhibit the C-terminal truncation
that is observed in native systems. This suggests that the heterologous
systems used to express the 1C subunit lack the
machinery or signals that participate in the proteolysis to the 190-kDa
form. Second, the expressed 1C subunit is resistant to
proteolysis into a 190-kDa, truncated form by whole cell lysates from
cardiac myocytes which presumably contain the potential proteases which
could be responsible for any artifactual cleavage of the native
1C subunit during purification. Additionally, the
1C subunit is resistant to proteolytic processing into a
190-kDa form by a number of exogenously added proteases including
trypsin, subtilisin, proteinase K, carboxypeptidase, papain,
thermolysin, and calpain suggesting that proteolytic processing of this
subunit depends upon specific conditions and/or proteolytic enzymes.
Finally, extensive efforts have been made to isolate a greater amount
of the full-length 1C subunit from cardiac muscle, but
none of the many procedures tested resulted in the isolation of the
full-length 240-kDa subunit as the major form of the protein (38). Even
when isolated cardiac myocytes were lysed in boiling SDS, the
full-length form of 1C only accounted for 5-15% of the total 1C protein. These observations, coupled with the
demonstration of the presence of the C-terminal domain in intact
cardiac myocytes by immunocytochemical methods (5), suggest that
cleavage of the cardiac 1C subunit into its truncated
190-kDa form is not an artifact of isolation, but represents a
post-translational processing event which occurs in intact cardiac myocytes.
The proteolytic enzyme(s) responsible for the processing of the native
cardiac and brain 1C subunits remain unknown. Calpain, a
calcium-dependent, cysteine protease, has been shown to
cleave the neuronal splice form of the 1C subunit into
the short form in hippocampal neurons (26), however, the same did not
appear to occur with the cardiac isoform. Chymotrypsin is an
extracellular serine protease which is probably not expressed in muscle
cells. However, a related protease, chymase, which has similar
substrate selectivity to chymotrypsin, has been suggested to be
involved in the processing and degradation of muscle proteins and may
be present in certain muscle cells (31, 32). Since exposure of the
full-length expressed 1C subunit to lysates of cardiac
myocytes did not result in the truncation of the protein to the 190-kDa form, this suggests that the proteolytic event that is responsible for
the processing is a regulated event. Evidence in support of this
possibility comes from the demonstration that activation of
N-methylaspartate receptors induced truncation of
1C subunits in hippocampal neurons (26). In addition, it
is conceivable that the proteolysis is developmentally regulated,
although this possibility has not been tested in the studies reported here.
The fate of the C terminus of the cardiac 1C subunit
following its proteolytic processing is important to ascertain as this domain is important for many different aspects of channel function including open probability (12), calcium-dependent
inactivation (13-15), and PKA regulation of the channel (7, 8).
Previous results obtained from immunofluorescence studies have shown
that the C terminus of the 1C subunit is present in
intact cardiac myocytes and co-localized with the body of the
1C subunit and the subunit (5). Two possible
explanations would explain these results. First, this could suggest
that all of the 1C protein present in intact myocytes is
a full-length protein, however, this is unlikely (5). The other
possibility is that the C terminus of the 1C subunit is
proteolytically processed in vivo and remains associated
with either the 190-kDa body of the channel or another membrane
protein. Evidence in support of this is that the C-terminal peptides
generated following chymotryptic processing of the 1C subunit were fairly stable fragments that remained associated with the membrane.
The finding that the 1C PRD was important for the
membrane association of the C-terminal peptides provided novel insights into the potential mechanisms by which physiologically cleaved C-terminal 1C fragments might remain functionally
associated with the channel. While it is not yet known what binding
partners define the association of the C-terminal fragments with the
membrane, conceivably the fragments could be binding to the 190-kDa
truncated 1C subunit or to another membrane or
cytoskeletal protein. The demonstration that the 1C PRD
can bind to several known SH3 domains suggests that potential
physiological partners might be SH3 domain containing proteins. That
the SH3 domains from the tyrosine kinases Src, Lyn, and Hck all served
as substrates for the 1C PRD was not too surprising
considering the high degree of homology between the 1C
PRD and the first PRD from dynamin (29). The first PRD from dynamin
binds to GST-SH3 domain fusion proteins from both Src and Fyn, but not
Grb2 or amphiphysin (29). Future experiments will explore the
possibility that the 1C PRD may function to link the
channel physiologically to tyrosine protein kinases that may be
involved in the regulation of channel activity. In this regard,
insulin-like growth factor-1 has been shown to regulate the activity of
neuronal L-type channels (36, 37), and this regulation could
conceivably involve the interaction of the channels with a tyrosine
kinase involved in the insulin-like growth factor-1 signaling cascade.
Other evidence suggests that the smooth muscle splice variant of the
1C subunit may be regulated by Src and that there may be
a direct association of the Src protein with the calcium channel (33).
The PRD is completely conserved in the 1C subunit splice
variant identified in smooth muscle (34), and therefore the PRD from
this 1C subunit may interact with the SH3 domain of Src.
While most of the other 1 isoforms contain some
conserved proline residues within the homologous regions of their C
termini, none of the other 1 subunits contain the class
II SH3-binding domain found in the 1C subunit. However, recent studies have shown that a proline-rich domain in the C terminus
of 1B can bind to an SH3 domain in CASK, an adaptor protein involved in exocytosis (35). In addition, the 1S
and the 1D isoforms have some prolines and other
residues homologous to those found to be important in the dynamin PRD
(29). Further experiments are required to test the possibility that
these other 1 subunits can bind to SH3 domains.
The finding that the 2-SH3 domain bound to the
1C PRD suggested that the association of the
2-SH3 domain with the 1C PRD may be
important for proper channel targeting and function. On the other hand,
an interaction between the 2-SH3 domain and the PRD from
the 1C subunit is probably not necessary for the
association of the 1C C terminus with the membrane as
observed in the in vitro experiments presented here, since
the 1C C-terminal fragments generated by chymotrypsin
were associated with the membrane regardless of whether or not the
2 subunit was coexpressed (Fig. 3). A role for the
2-SH3-1C PRD interaction remains to be elucidated.
Of particular interest was the finding that the deletion of either the
PRD or truncation of the C terminus resulted in increased channel
currents. Previous studies had demonstrated that truncations at
residues 1856, 1733, and 1700 of the C terminus of 1C
gave rise to currents that were significantly larger than currents through wild-type 1C (12). Currents generated by
1CPRD and 1C1905 were similarly
increased, suggesting a critical role for the PRD and the more distal
portions of the C terminus in the inhibitory function of the C
terminus. It will be of interest to determine how the association of
C-terminal domains with truncated 1C subunits, such as
1C1905, affects channel currents. One might predict
that the C-terminal fragments would inhibit currents obtained with a
mutant such as 1C1905. Alternatively it is possible that the C terminus is only inhibitory to channel function in its
non-cleaved form. Furthermore, it will be important to map the C
terminus to identify exactly which regions contribute to the inhibitory
function. The PRD itself may not alone constitute the inhibitory
element but these ~40 amino acids certainly participate in the
inhibition. Since deletion of the PRD region alone resulted in
increased channel currents, it may either comprise part of an
inhibitory domain that restricts channel currents or be important in
targeting the inhibitory domain to the channel via an interaction with
its putative SH3 domain containing partner.
The results presented here provide the first support for the hypothesis
that the C-terminal domain of the 1C subunit may be
cleaved physiologically but remain functionally associated with the
calcium channel. This suggests a novel method of channel regulation.
Since cleavage of the C terminus or deletion of the PRD of the
1C subunit have been demonstrated to allow for increased channel currents, it is possible that the C-terminal cleavage provides
a physiologically important mechanism to regulate the degree of
inhibition imposed by the interaction of the C-terminal domain with the
other components of this highly complex channel. In addition, since the
identified phosphorylation site for PKA resides at Ser1928
in the C terminus, in an area just upstream of the PRD, it is possible
that phosphorylation at Ser1928 serves to regulate the
interaction of the inhibitory C-terminal domain with its binding
partners. The elucidation of the function of the processing of the C
terminus of 1 subunits of voltage-dependent calcium channels will require extensive investigation given the many
roles that this domain is postulated to play in channel function.
1C
Subunit of L-type Calcium Channels and the Role of a
Proline-rich Domain in Membrane Tethering of Proteolytic
Fragments*
,
TOP
ABSTRACT
INTRODUCTION
Supported by a National Research Service Award Training Grant
T32-DK07169.
