(Received for publication, July 10, 1996, and in revised form, October 11, 1996)
From the Phosphorylation of myristoylated alanine-rich
protein kinase C substrate (MARCKS) in intact cells has been employed
as an indicator for activation of protein kinase C (PKC). Specific PKC isoenzymes responsible for MARCKS phosphorylation under physiological conditions, however, remained to be identified. In our present study
using stably transfected NIH 3T3 cell clones we demonstrate that
expression of constitutively active mutants of either conventional cPKC- Molecular cloning and biochemical studies identified the
myristoylated alanine-rich protein kinase C substrate
(MARCKS)1 as the major in vivo
substrate of protein kinase C (PKC) (1-9). The ability to
phosphorylate this substrate is not restricted to members of the PKC
family. MARCKS is predominantly phosphorylated on serine (S) residues
(in an order Ser-152 > Ser-156 > Ser-163) (10) in a
PKC-dependent fashion, but can also be phosphorylated on
serine and threonine residues by proline-directed protein kinases cdc2
and tau protein kinase II (11, 12). MARCKS is an acidic filamentous
actin cross-linking protein which is targeted to the plasma membrane by
its amino-terminal, myristoylated membrane-binding domain. This
specific interaction positions the substrate close to PKC, facilitating
its efficient phosphorylation. One of the striking features of MARCKS
is its phosphorylation-dependent translocation from the
membrane to the cytosol (8). Consequently, cytosolic MARCKS is not
further cross-linking actin filaments. It has also been proposed that
non-phosphorylated MARCKS complexes calmodulin resulting in a reduction
of Ca2+/calmodulin-dependent signaling
mechanisms and thereby blocking the entry of cells into the cell cycle
(13). In murine macrophages, immunoreactive MARCKS protein was found in
clusters at the interface of the substratum with pseudopodia and
filopodia, where it is colocalized with other PKC substrates of actin
cytoskeleton such as vinculin and talin (14). MARCKS is also highly
concentrated in presynaptic junctions and is phosphorylated in a
PKC-dependent manner when synaptosomes are depolarized (15,
16). Recently, it was demonstrated that MARCKS cycles between the
plasma membrane and Lamp-1 positive lysosomes in fibroblasts in a
PKC-dependent manner (17). In vitro studies
demonstrate that conventional PKC- Dulbecco's modified Eagle's medium
(DMEM), geneticin (G-418), and gentamycin were obtained from Boehringer
Mannheim Biochemicals (Mannheim, Germany). Fetal calf serum and
L-glutamine were purchased from Schoeller Pharma (Vienna,
Austria). Phorbol 12,13-dibutyrate (PDBu), phorbol 12-myristate
13-acetate (PMA), phosphatidyl-L- COS-1 cells
(1 × 106/100-mm dish) were transfected with 15 µg
of circular PKC-plasmid DNA/dish by Lipofectin reagents, as described by the manufacturer. Forty-eight hours post-transfection, cells were
lysed in 1 ml of buffer A (150 mM NaCl, 20 mM
HEPES, pH 7.5, 1% Nonidet P-40, 50 µg/ml each aprotinin and
leupeptin, and 1 mM phenylmethylsulfonyl fluoride). Lysates
from these transfectants were purified by using a
Ni2+-resin batch procedure, and equal amounts of
recombinant PKC isoenzymes were subjected to an enzymatic PKC assay as
described elsewhere (19, 24, 25). Enzyme activities of PKC- In contrast, to the PKC
plasmids described above, the pRc-CMV PKC- NIH 3T3
fibroblasts were kept at logarithmic growth phase in DMEM supplemented
with 10% heat-inactivated fetal calf serum and 2 mM
L-glutamine. One day after plating, semiconfluent cell cultures were refed with low serum medium (Opti-Mem I, Life
Technologies, Inc.) and transfected with 15 µg of circular
PKC-isoform plasmid DNA/dish by Lipofectin reagents, as described by
the manufacturer. 16 h post-transfection selection for G-418
resistant colonies was performed in 10% DMEM containing 700 µg/ml
G-418. After 12 days of selection antibiotic-resistant cells were
isolated as mass cultures and expression of recombinant PKC isoenzymes
was checked employing a standard Western blotting technique as
described elsewhere (19).
Pulse labeling and MARCKS phosphorylation of NIH 3T3
wild-type fibroblasts or stable transfectants were done by minor
modification of the method described by Herget and Rozengurt (27).
Briefly, NIH 3T3 wild-type cells or clones stably transfected with
distinct PKC isoenzyme expression constructs (1 × 106/dish) were incubated in phosphate-free medium (in the
presence of 20 nM to 6 µM GF109203X or
solvent, Me2SO, final concentration 0.15%) for 4 h,
followed by 50 µCi/ml carrier-free
[ Phosphorylation of MARCKS in intact cells has been
employed as an indicator for PKC activation but the specific PKC
isoenzyme responsible for MARCKS phosphorylation in intact cells,
however, remained to be identified. Significant reduction of
PDBu-induced MARCKS phosphorylation could be demonstrated by using a
PKC-specific inhibitor. At concentrations which do not exert a
significant effect on cell growth (20 nM to 6 µM), the PKC selective inhibitor GF109203X, a
bisindolylmaleimide, that inhibits PKC isoenzymes by competing with
enzyme-bound ATP (28, 29), cause a dose-dependent inhibition of PDBu-induced MARCKS phosphorylation (Fig. 1,
A and B), demonstrating that
MARCKS phosphorylation in intact cells depends in part on enzymatic
activity of PKC.
The fact that the NIH 3T3 fibroblast clone used in
this study express several different species of PKC isoenzymes
(predominantly PKC-
Kinase activities in vitro of recombinant PKC wild-type proteins in the
absence or presence of various concentrations of GF109203X
Institute of Medical Chemistry and
Biochemistry,
German Cancer
Research Center,
or novel nPKC-
increased phosphorylation of endogenous MARCKS in the absence of phorbol 12,13-dibutyrate in intact mouse fibroblasts, implicating that each of these PKC isoforms itself is
sufficient to induce enhanced MARCKS phosphorylation. Similarly, ectopic expression of a constitutively active mutant of PKC-
significantly increased MARCKS phosphorylation compared to vector controls, identifying PKC-
as a MARCKS kinase. The PKC-specific inhibitor GF 109203X (bisindolylmaleimide I) reduced MARCKS
phosphorylation in intact cells at a similar dose-response as enzymatic
activity of recombinant isoenzymes cPKC-
, nPKC-
, and nPKC-
in vitro. Consistently, phorbol
12,13-dibutyrate-dependent MARCKS phosphorylation was
significantly reduced in cell lines expressing dominant negative mutants of either PKC-
K368R or (dominant negative) PKC-
K436R. The fact, that the constitutively active PKC-
A119E mutant did not
alter the MARCKS phosphorylation underscores the assumption that
atypical PKC isoforms are not involved in this process. We conclude
that under physiological conditions, conventional cPKC-
and novel
nPKC-
, but not atypical aPKC-
are responsible for MARCKS
phosphorylation in intact NIH 3T3 fibroblasts.
, novel PKC-
and nPKC-
, but
not atypical PKC-
phosphorylate partially purified recombinant
MARCKS protein (10, 18). Phosphorylation of MARCKS in intact cells has
been employed as an indicator for activation of PKC, however, which PKC
isoform phosphorylates MARCKS in intact cells is still unknown.
Importantly, MARCKS gene knockout experiments (9) have been
demonstrated a dramatic genetic deficiency in mouse forebrain
development concerning the importance of this protein in perinatal
signal transduction. Therefore, PKC isoenzyme-specific MARCKS
phosphorylation in living mouse cells was tested. Conventional cPKC-
, novel nPKC-
, novel nPKC-
, and atypical aPKC-
isoforms have been selected as representatives of the three PKC
subfamilies. In order to identify the PKC isoenzyme-specific functions,
we investigated MARCKS phosphorylation following: 1) expression of transdominant negative (DN) PKC mutants (19, 20); 2) expression of
constitutively active (CA) PKC mutants in resting cells (19, 20); and
3) overexpression of wild-type isoenzymes. For comparative purposes,
the PKC-selective bisindolylmaleimide GF109203X was used for in
vitro and in vivo PKC inhibition studies.
Reagents and Plasmids
-serine (PtdSer),
leupeptin, and aprotinin were purchased from Sigma
(Vienna, Austria). Protein A-Sepharose was obtained from Pharmacia
(Vienna, Austria). GF109203X I is a product of Calbiochem (Lucerne,
Switzerland), Lipofectin transfection reagents and Opti-Mem I medium
were purchased from Life Technologies, Inc. (Vienna, Austria).
[
-32P]Orthophosphate (10 mCi/ml, 8500-9120 Ci/mmol),
[
-32P]ATP (10 mCi/ml, 3000 Ci/mmol), and Hyperfilm-MP
were obtained from Amersham (Amersham, Little Chalfont, UK). The mouse
polyclonal MARCKS antibody was raised in rabbits against the synthetic
oligopeptide EAAEPEQPEQPEQPAA with the amino acid sequence
corresponding to the sequence 222-237 of murine MARCKS and was
purified as described previously (21). Concerning the fact that this
antibody does not interact with one of the described critical
phosphorylation sites it should be suitable for immunoprecipitation
experiments of all kinds of phosphorylated MARCKS proteins. The plasmid
pRc-CMV-PKC-
, as well as pRc-CMV-PKC-
K275W encoding DN PKC-
was a kind gift from J. Moscat (22, 23). Human PKC-
, bovine PKC-
,
and rat PKC-
wild-type cDNAs were subcloned into the
cytomegalovirus (CMV) expression vector pRc-CMV (Invitrogen)
essentially as described (24). Site-directed mutagenesis of PKC
cDNA was performed using the Transformer SystemTM
(Clontech, Palo Alto, CA) as described by the manufacturer. Subcloning strategy, mutagenic primers, as well as selection primers for CA
PKC-
A25E, CA PKC-
A159E, and CA PKC-
A148E, as well as DN
PKC-
K409R have been described elsewhere (19). The PvuI selection primer (5
-CGG-TCC-TCC-GTT-CGT-TGT-CAG-3
) and the mutagenic primers (5
-CTA-TGC-TGT-GA
-GGT-CTT-AAA-GAA-3
and
5
-CGA-CGT-GGA-G
A-AGA-TGG-AG-3
) have been created to
construct the mutated version of (DN) PKC-
K436R and (CA) PKC-
A119E, respectively.
and -
are expressed as cofactor-dependent phosphorylation of the
[A25S] synthetic PKC peptide (RFARKG
LRQKNVY; presenting
the pseudosubstrate sequence of PKC-
with an alanine to serine
substitution) by recombinant PKC or control preparations. Enzyme
activity of PKC-
is expressed employing the PKC-
-specific
substrate peptide (153-[Ser-159]PKC-
-164)-NH2 corresponding to amino acid residues PRKRQG
VRRV (Upstate
Biotechnology Inc., Lake Placid, NY). The concentrations of substrate
peptides and cofactors used are: 50 µg/ml [A25S] and
(153-[Ser-159]PKC-
-164)-NH2, 280 µg/ml PtdSer, 10 µM PMA, and 1 mM CaCl2. To
measure PKC activity in the absence of Ca2+, EGTA (1 mM, final concentration) was added instead of
CaCl2. Expression of the fusion tag-peptide COOH-terminal
of PKC-
, PKC-
, or PKC-
did not affect the kinase activity
in vitro (24).
Immunocomplex Kinase Assay
plasmids do not carry a
6xHis tag, therefore kinase activity of PKC-
wild-type (in the
absence and presence of various concentrations of GF109203X), as well
as DN PKC-
K275W and constitutively active (CA) PKC-
A119E
mutants was performed as described by Müller and co-workers (26),
employing an immunocomplex kinase assay. Briefly, transiently
transfected COS-1 cells were washed with cold phosphate-buffered saline
and lysed on ice in 500 µl of lysis buffer A (50 mM
Tris-HCl, pH 7.3, 50 mM NaCl, 5 mM
Na4P2O7·10H2O (NaPP),
5 mM EDTA, 2% Nonidet P-40, 25 µg/ml leupeptin, 25 µg/ml aprotinin, 50 mM sodium fluoride, and 100 mM Na3VO4) for 10 min and lysates
clarified by centrifugation at 10,000 × g, 5 min, 4 °C. Aliquots of 1.5 × 106 NIH cell equivalents,
containing equal amounts of protein (500 µl, corresponding to
approximately 1.5 mg/ml; protein concentrations were determined by the
Bradford assay; Bio-Rad), were subjected to an immunoprecipitation (IP)
procedure employing a corresponding PKC-
antibody (Transduction
Laboratories, Lexington, KY). IPs were recovered by using protein
A-Sepharose beads (Pharmacia, Vienna). PKC-
molecules bound to 45 µl of protein A-Sepharose were resuspended in 20 µl of kinase
buffer and mixed with 9 µg of myelin basic protein (Sigma M-1891,
Sigmas, Vienna), 10 mM PtdSer and
±various concentrations of GF109203X. The kinase reactions were
initiated by the addition of 0.4 µCi of [
-32P]ATP
(10 mCi/ml, 3000 Ci/mmol) and incubation of the tubes by frequent
vortexing at 30 °C for 10 min. Phosphorylation of myelin basic
protein was terminated by the addition of 5 µl of 5 × SDS sample buffer and boiling the samples for 5 min. Probes were analyzed by SDS-PAGE (10%) and transferred to PVDF membranes (Millipore, Vienna). Determination of PKC-
enzyme activities was done by PhosphorImaging of the corresponding PVDF membranes. The putative potential of the immunocomplexes to alter, either directly or indirectly, the inhibitory activity of GF109203X was addressed by
in vitro PKC kinase assays of PKC immunoprecipitates in the presence or absence of various concentrations of GF109203X. Concerning the fact that there was no significant alteration in the dose-response curve (e.g. for recPKC-
) under these conditions our
results favor no direct or indirect influence of the immunocomplexes on
the inhibitory potency of GF109203X.
-32P]orthophosphate (Amersham) pulse labeling of the
endogenous ATP pool for an additional 2-h period. During the last 5 min
of the labeling period, cells were stimulated with 300 nM
PDBu. After harvesting the cells in 500 µl of lysis buffer (20 mM Tris/HCl, pH 7.5, 10 mM EDTA, 2% Nonidet
P-40, 1 mM phenylmethylsulfonyl fluoride, 100 µM Na3VO4, 50 µg/ml each of
leupeptin and aprotinin) the lysates were transferred to new tubes.
Lysates were clarified by centrifugation at 13,000 × g
for 10 min and cleared by incubation with protein A-Sepharose for
1 h at 4 °C. After removal of protein A-Sepharose by brief
centrifugation (30 s, Eppendorf microcentrifuge) the supernatants were
transferred to fresh tubes and boiled for 5 min at 90 °C. After this
procedure to destroy heat-instabile proteins the lysates were clarified
once more by centrifugation at 13,000 × g for 10 min.
Equal amounts of protein (500 µl, corresponding to approximately 1.5 mg/ml) were subjected to immunoprecipitation with a polyclonal
anti-MARCKS peptide antiserum overnight at 4 °C. Protein A-Sepharose
(100 µl) was added to the tubes for 1 h at 4 °C. Protein
A-Sepharose-coated immunocomplexes were collected by brief
centrifugation (1 min, 10,000 × g), washed five times with lysis buffer, mixed with 5 × Laemmli sample buffer, analyzed by SDS-PAGE (10%), and transferred to PVDF membrane (Millipore, Vienna). MARCKS phosphorylation analysis was done by PhosphorImaging of
the corresponding PVDF membranes.
PDBu-induced MARCKS Phosphorylation in Intact Cells Is Sensitive in
a Dose-dependent Fashion to the PKC-specific Inhibitor
GF109203X
Fig. 1.
PDBu-induced endogenous MARCKS
phosphorylation is sensitive to the PKC-specific inhibitor GF109203X in
a dose- dependent fashion. NIH 3T3 wild-type fibroblasts (1 × 106/dish) were incubated in phosphate-free medium (DMEM)
for 4 h followed by 50 µCi/ml
[-32P]orthophosphate pulse labeling for an additional
2-h period. During the last 5 min of the labeling procedure, cells were
pretreated 24 h with various concentrations of GF109203X (20 nM to 6 µM; lanes 3-6) or solvent
(Me2SO, final concentration 0.15%, lane 1) were
stimulated with 300 nM PDBu (lanes 2-6).
Heat-labile proteins were eliminated by centrifugation (13,000 × g, 10 min, 4 °C) after boiling the lysates for 5 min at
90 °C. Equal amounts of supernatants were subjected to
immunoprecipitation with a polyclonal MARCKS antiserum overnight at
4 °C. Protein A-Sepharose-coated immunocomplexes were washed five
times with lysis buffer, separated on PAGE (10%), and transferred to
PVDF membranes. A, PDBu-induced MARCKS phosphorylation in
the absence or presence of various concentrations of GF109203X as
indicated (shown is a representative autoradiogram out of three experiments done in triplicates). MARCKS phosphorylation in the presence of solvent (Me2SO, final concentration 0.15%;
lane 1), 5 min PDBu stimulation (lane 2), 5 min
PDBu stimulation of cells 24 h preincubated with various
concentrations of GF109203X (lanes 3-6). B,
statistical analysis of experiments described under A. MARCKS phosphorylation was determined by PhosphorImaging of PVDF membranes and data are expressed as the means (±S.E.) of at least three independent experiments done in triplicates.
[View Larger Version of this Image (18K GIF file)]
, -
, -
, and -
) make it difficult to
define exactly the isoenzyme(s) affected. Therefore, purified
recombinant PKC isoforms were tested in the presence of GF109203X.
COS-1 cells were transiently transfected with the appropriate PKC
expression constructs or a vector control, and the recombinant PKC was
purified by exploiting the COOH-terminal six-histidine
(His6)-fusion tag and analyzed in standard PKC kinase
assays (19) against substrate peptides ([A25S]peptide for PKC-
and
-
, and [Ser-159]peptide for PKC-
) in the absence or presence of
known PKC cofactors including PtdSer, PMA, and Ca2+ as
described previously (24). In the case of PKC-
an immunocomplex kinase assay with myelin basic protein as synthetic substrate was done
as described under "Materials and Methods." The results are
summarized in Table I. At lower concentrations than in
intact cells, addition of GF109203X to purified recombinant PKC
isoenzymes resulted in a significant reduction of protein kinase
activity of cPKC-
, nPKC-
, and nPKC-
. Atypical aPKC-
,
however, did not demonstrate significant inhibition up to GF109203X
concentrations of 6 µM (Table I). This dose-response is
in agreement with IC50 concentrations determined for
inhibition of the atypical PKC-
isoform by GF109203X
(IC50 of PKC-
5.8 µM) (28), an isoenzyme which exhibits 72% sequence homology to PKC-
on the amino acid level (30). These results implicate that under physiological conditions
conventional and novel, but not atypical PKC isoforms are involved in
MARCKS phosphorylation.
Recombinant PKC
isoenzymes
Cofactorsa
Inhibitor or solvent
Enzyme
activityb
%
None
EGTA
2
± 1
None
PtdSer/PMA/Ca2+
Me2SO
3
± 2
PKC-
wt
PtdSer/PMA/Ca2+
+
Me2SO
100 ± 11
PKC-
wt
PtdSer/PMA/Ca2+
20 nM
38 ± 8
PKC-
wt
PtdSer/PMA/Ca2+
200 nM
11
± 2
PKC-
wt
PtdSer/PMA/Ca2+
2
µM
2 ± 1
PKC-
wt
PtdSer/PMA/Ca2+
6 µM
1 ± 1
PKC-
wt
PtdSer/PMA
+Me2SO
100 ± 5
PKC-
wt
PtdSer/PMA
20 nM
53 ± 9
PKC-
wt
PtdSer/PMA
200 nM
22 ± 4
PKC-
wt
PtdSer/PMA
2 µM
18 ± 5
PKC-
wt
PtdSer/PMA
6 µM
15 ± 3
PKC-
wt
PtdSer/PMA
+Me2SO
100 ± 23
PKC-
wt
PtdSer/PMA
20 nM
8 ± 3
PKC-
wt
PtdSer/PMA
200 nM
3 ± 2
PKC-
wt
PtdSer/PMA
2 µM
2 ± 1
PKC-
wt
PtdSer/PMA
6 µM
1 ± 1
None
3 ± 2
None
PtdSer
+Me2SO
11 ± 2
PKC-
wt
PtdSer
+Me2SO
100 ± 11
PKC-
wt
PtdSer
20 nM
123 ± 8
PKC-
wt
PtdSer
200 nM
119 ± 23
PKC-
wt
PtdSer
2 µM
91 ± 12
PKC-
wt
PtdSer
6 µM
14 ± 6
a
The concentrations of synthetic substrates and
cofactors used are: 50 µg/ml [A25S] and
(153-[Ser-159]PKC- -164)-NH2, 3 µg/µl MBP, 280 µg/ml
PtdSer, 10 µM PMA, and 1 mM CaCl2. To
measure PKC activity in the absence of Ca2+, EGTA (1 mM final concentration) was added instead of CaCl2. Expression of the fusion tag-peptide COOH-terminal of the recombinant PKC isoenzymes thereby was found not to affect the kinase activity in vitro (24).
b
To correct for differences in transfection efficiencies,
enzyme activities are expressed as a percentage of
cofactor-dependent phosphorylation of the [A25S]PKC
peptide, which was determined separately in each experiment. Data
expressed as the means (±S.E.) of at least three independent
experiments done in triplicates.
In order to
eliminate pleiotropic GF109203X effect(s), we employed DN PKC mutants.
In these constructs the critical lysine at the ATP-binding site is
replaced by an arginine to produce transdominant-negative phenotypes.
Such PKC mutant proteins have been shown to compete with endogenous PKC
in an isoenzyme specific manner (19, 20, 22, 31-34). Employing
transient transfection assays in COS-1 cells (DN) PKC- K368R, (DN)
PKC-
K436R, and (DN) PKC-
K409R have been found to lack kinase
activity as recently described (19). To confirm the biological
relevance of these DN PKC mutants in NIH 3T3 cells, the mutant proteins
were tested in two independent biological systems. The biological
function of PKC-
K368R was confirmed on thrombin-induced release of
Ca2+ from intracellular stores. Briefly, expression of
PKC-
K368R has been shown to overcome a PKC-
-mediated feedback
inhibition of thrombin-induced intracellular Ca2+ release
(data not shown). Furthermore, transient expression of DN PKC-
K436R, not, however, of PKC-
K368R blocks the transcriptional activation of c-fos by oncogenic
Ras.2 The expression of transfected PKC
mutants and wild-type isoenzymes was assessed by immunoblotting of
Ni2+-chelating resin precipitates, employing PKC
isoform-specific antibodies (data not shown). Comparable levels of
recombinant PKC mutants and wild-type isoforms were detected in
different NIH 3T3 clones, representing approximately 2.5-fold
overexpression relative to the levels of endogenous PKC isoforms (see
Fig. 4 for wild-type). Taken together, these data indicate that all
dead kinase mutants are expressed at comparable levels and exert
dominant negative effects in an isoenzyme-specific fashion. Consistent with the above findings concerning the biological relevance of the DN
mutants, MARCKS phosphorylation induced by 300 nM PDBu is
significantly reduced in cell lines constitutively expressing DN
PKC-
K368R or DN PKC-
K436R mutants (Fig. 2,
A and B). Importantly, due to the lack of
endogenous PKC-
expression in our NIH clone, the catalytically
inactive DN PKC-
K409R had no effect on PDBu-induced MARCKS
phosphorylation, providing the isoform specificity of the dominant
negative kinase approach used in this study.
Expression of CA PKC-
Circumstantial evidence suggests that point
mutation in the pseudosubstrate motif of the regulatory domain of PKC
disrupt the interaction between the catalytic site and the
pseudosubstrate sequence, generating an individual PKC isotype mutant
independent of the stimulatory effects of phorbol esters or
diacylglycerols, as shown by our standard in vitro PKC assay
(19). To document that PKC-, and PKC-
by itself, are sufficient
to phosphorylate MARCKS, cell lines stably expressing CA PKC-
A25E,
CA PKC-
A159E, or PKC-
A119E were analyzed. In agreement with the
results obtained so far, expression of CA PKC-
A25E and CA PKC-
A159E resulted in a significant increase in MARCKS phosphorylation in
the absence of PDBu when compared with vector transfected control cells
(Fig. 3, A and B). Expression of
PKC-
A119E, however, did not demonstrate any significant increase in
MARCKS phosphorylation. Interestingly, NIH 3T3 cells ectopically
expressing a CA PKC-
A148E mutant was found to significantly enhance
MARCKS phosphorylation in the absence of PDBu, identifying PKC-
as a
new MARCKS kinase. In experiments where CA PKC-
, -
and -
were
expressed, addition of 300 nM PDBu further enhanced the
level of MARCKS phosphorylation (data not shown), suggesting a
submaximal activation status. As expected, the addition of PDBu to
transfectants where CA PKC-
A119E was expressed lead to a complete
phosphorylation status of MARCKS, presumably based on the activation of
all PDBu-responsive MARCKS phosphorylating PKC isoenzymes endogenously
expressed in NIH 3T3 fibroblasts.
Kinase Activity of Recombinant PKC-
In order to characterize the enzymatic properties of
PKC- A119E, an immunocomplex kinase assay was performed as described under "Materials and Methods." COS-1 cells were transiently
transfected with the appropriate PKC-
expression constructs or a
vector control, and the recombinant PKCs were collected by
immunoprecipitation and analyzed in standard PKC kinase assays (19)
against MBP as a substrate peptide in the absence or presence of
PtdSer. To correct for differences in transfection efficiencies, enzyme
activities are expressed as a percentage of
PtdSer-dependent phosphorylation of MBP by PKC-
wild-type. The results are summarized in Table II.
Importantly, in comparison with PKC-
wild-type enzyme activity, the
PKC-
A119E mutant was capable of phosphorylating synthetic myelin
basic protein (MBP) substrate in the absence of PtdSer (Table II). The
activation level, however, was suboptimal, indicating either a
submaximal activation status of PKC-
A119E in vitro, or
the requirement of additional cofactors such as
-interacting protein
(23).
|
The NIH 3T3 fibroblast clone used in
our study was shown to express predominantly PKC-, -
, -
, and
-
isoenzymes (data not shown). Under our experimental conditions
used, conventional cPKC-
, nPKC-
, and novel nPKC-
accept MARCKS
as substrate in intact cells. To further characterize the biological
relevance of these MARCKS kinases we have examined the potential role
of cell lines overexpressing theses particular PKC isoenzymes.
Conventional cPKC-
and novel nPKC-
were compared with ectopically
expressed nPKC-
on MARCKS phosphorylation. The isoenzyme-specific
overexpression was confirmed by Western analysis (Fig.
4) and found to be similar for PKC-
(2.2-fold),
PKC-
(2.6-fold), and PKC-
(2.0-fold). PKC-
protein, as
expected, was only expressed in cells ectopically transfected with a
plasmid encoding PKC-
.
Consistent with our findings, in the presence of 300 nM
PDBu, MARCKS phosphorylation was found to be enhanced in cells
overexpressing cPKC- and nPKC-
(Fig. 5).
Interestingly, ectopically expressed nPKC-
showed the highest levels
of MARCKS phosphorylation implicating nPKC-
as a potent MARCKS
kinase (Fig. 5). The fact that the PDBu-dependent MARCKS
phosphorylation was only slightly enhanced by overexpression of
cPKC-
, nPKC-
, and nPKC-
could reflect that the total amount of
MARCKS and/or the accessibility of MARCKS per cell is a limiting factor
in PKC-overexpressing clones. In non-PDBu-stimulated cells, overexpression of PKC-
, -
, and -
did not cause a significant change in basal MARCKS phosphorylation (data not shown).
PKC isoenzyme-specific phosphorylation of MARCKS had been studied
previously employing cell-free extracts (10, 18). The fact that the
molecular mechanisms of the processes leading to PKC activation are
still insufficiently understood and that cell lines are expressing
several distinct PKC isoenzymes have made it difficult to extrapolate
from in vitro studies to the situation in vivo.
Therefore, the involvement of four PKC subfamilies were tested under
physiological conditions. It is demonstrated that in intact cells
conventional cPKC-, novel nPKCs
, and nPKC-
, not, however,
atypical aPKC-
accept MARCKS as a substrate. These conclusions are
based on the following data: 1) MARCKS phosphorylation is significantly
enhanced following a brief exposure to the phorbol ester PDBu and the
PDBu-induced MARCKS hyperphosphorylation is depressed by concentrations
of the specific PKC inhibitor GF109203X which have been shown to
inhibit the phorbol ester-responsive PKC isoforms
,
, and
. At
2 µM GF109203X, MARCKS phosphorylation is reduced to
background levels. At this concentration the atypical PKCs-
and -
(28) are only partially inhibited. Although PKC-
and
are phorbol
ester non-responsive (30, 35), an implication of these isoenzymes after
phorbol ester treatment cannot a priori be excluded. It has
been suggested that c- or n-type PKC isoforms upon activation by
phorbol esters may stimulate phospholipase D, phosphoinositol 3-kinase,
or phospholipase A2 which in turn could activate PKC-
or
(26, 36-38). The dose-response relationship shown in Fig. 1 argues
against an implication of atypical PKC isoenzymes in PDBu-induced
MARCKS phosphorylation. 2) The conclusion that in intact NIH 3T3
fibroblasts, PKC-
and -
are capable of phosphorylating MARCKS is
further supported by the results obtained with kinase-dead DN mutants.
Both, the DN PKC-
K368R as well as DN PKC-
K436R mutants
significantly reduce the PDBu-induced MARCKS phosphorylation. As
PKC-
is not expressed in these cells, the DN PKC-
K409R mutant
should not be able to affect PDBu-induced MARCKS phosphorylation which
is indeed the case.
PKC isoenzymes are located in different subcellular locations and/or
compartments in a given cell, therefore prepositioning of PKCs in
resting cells may be the key determinant in substrate phosphorylation,
e.g. that the total amount and/or the accessibility of
MARCKS molecules in a PKC isoenzyme relevant subcellular location, but
not simple substrate competition, could be the prerequisite of a DN PKC
isoenzyme-specific inhibition of MARCKS phosphorylation. 3) In order to
eliminate pleiotropic effects of the phorbol ester, we employed CA PKC
mutants. In accordance with the conclusions drawn so far, expression of
the constitutively active mutants of PKC-, -
, and -
,
respectively, were found to enhance MARCKS phosphorylation also in the
absence of PDBu. Addition of PDBu, however, further enhances the level
of MARCKS phosphorylation (data not shown) which may be due to a
submaximal activation status of type proteins (19). Alternatively and
perhaps more likely, this phenomenon may simply be due to the fact that
exposure to PDBu leads to an activation of all PDBu-responsive
MARCKS-phosphorylating PKC isoenzymes. Consequently, only part of the
total activity can be obtained by expressing one of these isoenzymes as
a constitutively active form. 4) These first hints suggesting an
implication of PKC-
, -
, and -
in MARCKS phosphorylation are
further substantiated by studies with cell lines overexpressing the
individual PKC isoforms. Ectopically expressed PKC-
, the closest
relative to PKC-
, which was found to be predominantly expressed in
hematopoietic cell lines and skeletal muscle (39) proved to be an
especially active MARCKS kinase, although as judged from Western blots,
the expression level of PKC-
was lower to the levels obtained for
PKC-
and -
overexpressing cells. The total increase in MARCKS
phosphorylation obtained by the overexpression of PKC-
, -
, and
-
is relatively small. It should be considered, however, that MARCKS
is sequentially phosphorylated on serine residues in the order serine,
Ser-156 > Ser-163 > Ser-153 (10). No isoenzyme-specific
major differences have been reported with regard to the sequential
phosphorylation of the MARCKS protein. 5) In contrast to all the data
demonstrating that the
,
, and (ectopically expressed)
isoforms of PKC are implicated in MARCKS phosphorylation in intact
fibroblasts, there is so far no evidence for MARCKS as a substrate of
PKC-
. As a matter of fact, all studies conducted to reveal MARCKS
phosphorylation by PKC-
yielded negative results. Studies with
PKC-
are hampered by the fact that so far no exogenous stimulating
agonist of this isotype has been described. Intracellularly, PKC-
has been reported to be regulated by a
-interacting protein (23).
The mechanisms by which
-interacting protein is regulated are,
however, still obscure. Diaz-Meco and co-workers (22) recently
demonstrated that overexpression of PKC-
in COS-1 cells or NIH 3T3
fibroblasts leads to a transcriptional activation of a NF
B-driven
reporter plasmid. It is shown here that even the expression of a CA
PKC-
mutant does not significantly alter the phosphorylation of
MARCKS. The data on MARCKS phosphorylation in intact cells described
here are in excellent agreement with studies obtained with isolated PKC
isoenzymes in cell-free assays (10, 18). In vitro, cPKC-
, cPKC-
1, nPKC-
, and nPKC-
but not a PKC-
were identified as enzymes that accept MARCKS as a substrate. PKC-
which exhibited the
highest rate of MARCKS phosphorylation in vitro shares 75% amino acid homology with PKC-
(39), an isoform so far not
investigated with regard to its ability to phosphorylate MARCKS.
PKC-
which has been shown to be expressed in NIH 3T3 cells (30),
exhibits 72% amino acid sequence homology with PKC-
(30). Within
the kinase (C3) region the identities between PKC-
and PKC-
are even 86% (30). In view of this homology it is not surprising that both
PKC isoenzymes show overlapping substrate specificities. Indeed, both
isoenzymes have been implicated to stimulate the transcription of a
NF
B-driven reporter plasmid, and are obviously unable to
phosphorylate MARCKS. The major differences between PKC-
and PKC-
are to be found in the regulatory zinc finger domain explaining
differential mechanisms of activation of the two enzymes (23).
The functional divergence of PKC isoenzymes provides a rational to further explain the presence of multiple PKC family members in a given cell. Furthermore, it will permit detailed functional dissection of the complex signal transduction cascades involving distinct PKC family members. Undoubtedly more work is necessary to determine the precise mechanism utilized by PKC isoenzymes to phosphorylate the myristoylated alanine-rich protein kinase C substrate MARCKS, but our data represent an important step toward the identification of PKC isoenzymes involved in MARCKS regulation in vivo.
We are grateful to Dr. Jorge Moscat for
providing the pRc-CMV-PKC- wild-type and dominant negative kinase
dead PKC-
K275W constructs and Gerd Utermann for helpful comments,
stimulating discussions, and critical reading of the manuscript.
Furthermore, we thank Eugen Preuss (Presentation, Documentation, and
Learning Systems) for illustration.