(Received for publication, July 25, 1996, and in revised form, November 26, 1996)
From the Institute of Biomedical Sciences, Academia
Sinica, Taipei, Taiwan, § Institute of Neuroscience,
National Yang-Ming Medical College, Taipei, Taiwan, Republic of China,
and ¶ Department of Neurology, E. Gallo Clinic and Research
Center, University of California,
San Francisco, California 94110
We have previously reported that phosphorylation
of adenylyl cyclase type VI (AC6) may result in the suppression of
adenylyl cyclase activity during desensitization of the A2a-adenosine
receptor-mediated cAMP response (A2a desensitization) in rat
pheochromocytoma PC12 cells. In the present study, we demonstrate that
protein kinase C (PKC) is responsible for the phosphorylation and
inhibition of AC6 during A2a desensitization. Inhibition of PKC by
several independent methods markedly blocked the suppression of AC6
during A2a desensitization. Purified PKC from rat brain directly
phosphorylated and inhibited recombinant AC6 expressed in Sf21 cells.
Substantially lower AC6 activities were also observed in PC12 cells
overexpressing PKC or PKC
. Stimulation of A2a-R in PC12 cells
under the same conditions as those required for A2a desensitization
resulted in an increase in Ca2+-independent PKC activity.
Most importantly, exogenous PKC did not further suppress AC6 activity
in A2a-desensitized membranes. In vitro PKC phosphorylation
of AC6 isolated from A2a-desensitized cells was also profoundly lower
than that from control cells, suggesting a specific role for PKC in
regulating AC6 during A2a desensitization in PC12 cells. Taken
together, our data demonstrate that a calcium-independent, novel PKC
inhibits AC6 activity during A2a desensitization in PC12 cells.
Independent regulation of AC6 by calcium-independent PKC and by
Ca2+ provides an exquisite mechanism for integrating
signaling pathways to fine-tune cAMP synthesis.
We have recently reported that prolonged activation of A2a adenosine receptor (A2a-R)1 in PC12 cells significantly inhibits the activity of adenylyl cyclase type VI (AC6), which in turn causes a lower response to subsequent stimulation of A2a-R (A2a desensitization). In addition, there is evidence that protein phosphorylation may mediate the suppression of AC6 activity during A2a desensitization (1, 2). However, the exact molecular events underlying the phosphorylation and inhibition of AC6 during prolonged activation of A2a-R in PC12 cells remain largely uncharacterized.
Genes of at least 10 distinct mammalian adenylyl cyclases (AC), which
can be further divided into five subfamilies, have been reported (3,
4). The specific tissue distribution of these different adenylyl
cyclases supports biochemical evidence for distinct modes of regulation
of cAMP levels (5). Although these enzymes can all be activated by the
a subunit of Gs protein, each is under very distinct
regulation. For example, AC6 can be inhibited by physiologically
relevant concentrations of Ca2+, while other AC isozymes
(adenylyl cyclase type I (AC1) and adenylyl cyclase type VIII (AC8))
are activated by Ca2+ in the presence of calmodulin (6). In
addition, Ca2+- mediated inhibition has been clearly
demonstrated for adenylyl cyclase type III (AC3) due to direct
phosphorylation of AC3 by a calmodulin-kinase II (7, 8). The
susceptibility of each individual adenylyl cyclase to inhibition by
Gi also differs (9). Most interestingly, protein kinase
C (PKC) has been implicated in stimulation of several ACs (AC1,
adenylyl cyclase type II (AC2), and adenylyl cyclase type V (AC5))
(10-12), but markedly inhibits the activity of AC6 as demonstrated in
the present study. Regulation of the Ca2+-inhibitable AC6
by PKC provides a very useful mechanism to detect and integrate
coincidental signals mediated by Ca2+, cAMP, and PKC.
PKC is a family of serine/threonine protein kinases which is composed
of at least 11 different members from 10 distinct genes. These PKC
isozymes are divided into three subfamilies based on their amino acid
sequences and their specific modes of activation. The "conventional
PKCs" including ,
, and
are dependent on calcium and
diacylglycerols (or phorbol esters) for activation. Members of another
subfamily of PKC, novel PKCs (
,
,
, and
), require
diacylglycerols (or phorbol esters) but not calcium for their
activation. Activation of the "atypical PKCs," including
and
, is independent of either calcium or diacylglycerols (13). Recently, two related, phospholipid-dependent kinases,
PKCµ and PKD, have been identified that are phorbol ester-sensitive
and calcium-independent. They are unique in having a hydrophic domain near the amino terminus, an internal pleckstrin homology domain, and a
substrate specificity different from other PKCs, suggesting that they
constitute a new PKC subgroup (14, 15).
In the present study, we present evidence to demonstrate that stimulation of A2a-R activates calcium-independent PKC, which phosphorylates and inhibits AC6. Suppression of AC6 by protein phosphorylation thus yields a lower response of AC6 to subsequent stimulation of A2a adenosine receptor in PC12 cells.
CGS21680 were obtained from Research
Biochemicals, Inc. (Natick, MA). Cyclic AMP and ATP were obtained from
Sigma. Okadaic acid (OKA) and protein kinase C were from Boehringer
Mannheim Biochemica (Mannheim, Germany). Western blot analysis revealed that this mixture of PKCs from rat brain contains at least PKC,
I,
II,
,
,
, and µ (data not shown). The monoclonal
anti-FLAG antibody, M2, was obtained from Eastman Kodak Co. Purified
human protein phosphatase type 2A, and the
-PKC substrate were
purchased from Upstate Biotechnology Inc. (Lake Placid, NY).
Phorbol-12,13-didecanoate (PDD) was from Calbiochem-Novabiochem Int.
(San Diego, CA).
PC12 cells were originally obtained from
American Type Culture Collection (Rockville, MD). The pRc/RSV, PKC,
and PKC
-transfected clones (designated C1,
1, and
1,
respectively) were developed and characterized as described elsewhere
(16). These PC12 variants were maintained in Dulbecco's modified
Eagle's medium (Life Technologies, Inc.) supplemented with 5% fetal
bovine serum (Life Technologies, Inc.) plus 10% horse serum (Life
Technologies, Inc.) in an incubation chamber gassed with 10%
CO2, 90% air at 37 °C. A123, the PKA-deficient variant
of PC12 cells (17), was kindly provided by Dr. J. A. Wagner (Cornell
University Medical College). A123 cells were grown in Dulbecco's
modified Eagle's medium supplemented with 10% fetal bovine serum plus
5% horse serum in the presence of 10% CO2, 90% air at
37 °C. Sf21 cells were obtained from Pharmingen (San Diego, CA) and
were grown in Grace's insect medium (Life Technologies, Inc.)
supplemented with Yeastolate (0.33%), lactalbunin hydrolysate (0.33%), and 10% fetal bovine serum at 27 °C.
Adenylyl cyclase activity was assayed as described previously (1). Briefly, cells were resuspended in the lysis buffer (10 mM EDTA, 20 mM Tris-HCl, 250 mM sucrose, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride (PMSF), and 40 µM leupeptin, pH 7.4), and sonicated using a W-380 sonicator (Ultrasonics Inc.) at a setting of 20% output power for a total of 45 s. The homogenate was centrifuged at 50,000 × g for 30 min to collect the P1 membrane fractions. The adenylyl cyclase activity assay was performed at 37 °C for 10 min in a 400-µl reaction mixture containing 1 mM ATP, 100 mM NaCl, 0.4 unit of adenosine deaminase, 50 mM Hepes, 0.5 mM 3-isobutyl-1-methylxanthine, 6 mM MgCl2, 1 µM GTP, and 20 µg of membrane protein. Reactions were stopped by 0.6 ml of 10% trichloroacetic acid. The cAMP formed was isolated by Dowex chromatography (Sigma) and assayed by radioimmunoassay using the 125I-labeled cAMP assay system (Amersham). No significant difference was found by adding 20 mM creatine phosphate (Sigma), or 100 units/ml creatine phosphokinase (Sigma) to the cyclase reaction. In experiments with elevated calcium concentrations, membrane preparations were washed two more times with ice-cold HDLP buffer (20 mM Hepes, 0.5 mM dithiothreitol, 0.1 mM leupeptin, 40 µM PMSF, pH 7) containing 1 mM EGTA to remove the endogenous calcium and calmodulin (10). Membrane fractions were then resuspended in a EGTA/calcium buffer to control the free calcium concentration at zero or 30 µM for adenylyl cyclase assay. Since only one calcium-inhibitable adenylyl cyclase (AC6) exists in PC12 cells (2), the AC6 activity in PC12 cells was determined as the difference between the forskolin-evoked AC activity assayed in the absence and the presence of 30 µM free calcium. Addition of calcium in the cyclase assay did not evoke any detectable phosphodiesterase activity in control or desensitized membrane. The enzyme activity was linear for up to 20 min with membrane protein up to 100 µg. All samples were assayed in triplicate.
A2a Desensitization in a Cell-free SystemP1 membrane fractions were collected from PC12 cells as described above, and resuspended in DB buffer (0.4 mM EDTA, 1 mM EGTA, 0.1 mM leupeptin, 40 µM PMSF, 100 µM GTP, 1 mM ATP, 30 nM OKA, 0.2 mM sodium vanadate, 5 mM MgCl2, and 25 mM Tris, pH 7.2). A2a desensitization was triggered by incubating the P1 membrane fractions with CGS21680 (100 µM) for 30 min at 37 °C. The membranes were washed twice with ice-cold HDLP buffer to remove the excess agonist. Adenylyl cyclase activities of these treated membranes were determined as described above.
Expression of AC6 in Sf21 CellsThe rat AC6 cDNA was kindly provided by Dr. Iyengar (18). A synthetic 8-amino acid peptide, FLAG, was ligated to the N terminus of AC6 to encode a chimeric FLAG-AC6 recombinant protein (designated as F-AC6). Monoclonal antibody against FLAG was obtained from Kodak. Expression of AC6 and F-AC6 was carried out in a recombinant baculovirus-driven Sf21 cell system following the manufacturer's protocol (Pharmingen). Membrane fractions were collected as described above from Sf21 cells infected with the desired virus 68-72 h postinfection.
Preparation of AntibodyAn oligopeptide corresponding to amino acids 1162 to 1181 of rat AC6 was purchased (Genosys, Woodlands, TX) and conjugated to bovine serum albumin using m-maleimidobenzoyl-N-hydroxysuccinimide ester as described by Harlow and Lane (19). To produce anti-AC6 antibodies, male New Zealand White rabbits (~2.5 kg) were immunized by an intramuscular injection of the bovine serum albumin-conjugated AC6 peptide using standard procedures (19). This polyclonal AC6-specific anti-peptide antibody is designated as AC6C. A recombinant protein containing amino acid 987 ~1187 of AC6 was expressed and purified from Escherichia coli BL21-DE3, and used to raise polyclonal antibody in rabbits using standard procedures (19). The resulting anti-serum was designated as AC6D.
SDS-Polyacrylamide Gel Electrophoresis and Western BlottingWe determined protein concentration by a simple
colorimetric assay based on the Bradford dye binding procedure (20)
using the Bio-Rad protein assay dye reagent concentrate (Bio-Rad). For SDS-polyacrylamide gel electrophoresis, membrane fractions were combined with 2× sample buffer containing 125 mM Tris-HCl
(pH 6.8), 20% glycerol, 1% SDS, 15% 2-mercaptoethanol, 200 mM dithiothreitol, and 0.01% bromphenol blue, boiled for 5 min, centrifuged to remove the insoluble material, and then separated
on 8% separating gels according to the method of Laemmli (21).
Following electrophoresis, the gel was transferred to a polyvinylidene
difluoride membrane, blocked with 5% skim milk in PBS, then incubated
with the desired antiserum at 4 °C overnight. Typically, we used a
1:1000 dilution for PKC antiserum, 1:500 for PKC
antiserum, 1:250
for PKC
antiserum, 1:1000 for AC6C, and 1:5000 for AC6D,
respectively, unless stated otherwise. After three washes of 5 min in
PBS, the membranes were incubated with peroxidase-conjugated donkey
anti-rabbit IgG (Amersham, UK) at 1:5000 dilution for 1 h at room
temperature. The membranes were washed three times with PBS, and the
immunoreactive bands were stained using a light emitting nonradioactive
method (ECL, Amersham, UK).
To carry
out the protein phosphorylation study, AC6 and F-AC6 expressed in Sf21
cells were purified by immunoprecipitation using AC6C. In brief, 1 mg
of P1 membrane fractions collected from Sf21 cells infected with the
indicated virus were solubilized in 0.4 ml of ice-cold radioimmune
precipitation buffer (50 mM NaCl, 1% Nonidet P-40, 0.5%
sodium deoxycholate, 0.1% SDS, 50 mM Tris, pH 8) at
4 °C for 30 min. Immunoprecipitation was initiated by addition of
AC6C (1:50 dilution). The reaction mixture was incubated with gentle
agitation at 4 °C overnight. Immunocomplexes were purified using
Sephadex conjugated protein A (Sigma) and washed three times with
ice-cold radioimmune precipitation buffer. Phosphorylation by PKC was
carried out in a final volume of 0.2 ml reaction mixture containing 10 mM MgCl2, 1 mM CaCl2,
0.25% bovine serum albumin, 10 µM H89, 0.5 mM [-32P]ATP (2 Ci/mole), 0.1 mM leupeptin, 40 µM PMSF, 30 nM
OKA, 0.2 mM sodium vanadate, and 20 mM Tris (pH
7.5). The phosphorylation reaction was initiated by addition of protein
kinase C (0.1 milliunit) purified from rat brain (Boehringer Mannheim)
for 30 min at 4 °C, and terminated by the addition of 2 × SDS
sample treatment buffer. The samples were then boiled for 5 min,
analyzed by SDS-polyacrylamide gel electrophoresis (8%) and Western
blot analysis using AC6D. To visualize the phosphorylation of AC6 by
PKC, immunoblots were rinsed twice with PBS, air-dried, and
autoradiographed.
To phosphorylate AC6 purified immunologically from PC12 cells using
exogenous PKC, 17 mg of P1 membrane fractions collected from control or
A2a desensitized cells (CGS21680, 2.5 µM, 16 h) were
solubilized in 3.2 ml of ice-cold radioimmune precipitation buffer at
4 °C for 30 min. Immunological purification of AC6 using AC6C was
conducted as described above. Phosphorylation by PKC was carried out in
a final volume of 0.1 ml of reaction mixture containing 10 mM MgCl2, 0.25% bovine serum albumin, 11 µM [-32P]ATP (35 Ci/mole), 0.1 mM leupeptin, 40 µM PMSF, 1 µM
OKA, 0.2 mM sodium vanadate, 5 mM
-glycerolphosphate, 1 µM PDD, and 25 mM
Tris (pH 7.5). The phosphorylation reaction was initiated by addition
of PKC (24 microunits) purified from rat brain (Boehringer Mannheim)
for 5 min at 37 °C, and terminated by the addition of 2 × SDS
sample treatment buffer. The samples were then boiled for 5 min and
analyzed by SDS-polyacrylamide gel electrophoresis (10%) and Western
blot using AC6D. To visualize the phosphorylation of AC6 by PKC,
immunoblots were rinsed twice with PBS, air-dried, and
autoradiographed.
PKC activities were measured under
the same conditions as those for the A2a desensitization in
vitro. In brief, PKC was activated by incubating the P1 membrane
fractions with or without CGS21680 (100 µM) for 30 min at
37 °C in a 25-µl reaction mixture containing 0.2 µM
of protein kinase A inhibitor peptide (Upstate Biotechnology Inc.), 2 µM of compound R24571 (Upstate Biotechnology Inc.), 10 µM of peptide (Upstate Biotechnology Inc.), 0.4 mM EDTA, 1 mM EGTA, 0.1 mM
leupeptin, 40 µM PMSF, 100 µM GTP, 1 mM [
-32P]ATP (0.2 Ci/mmol), 30 nM OKA, 0.2 mM sodium vanadate, 5 mM MgCl2, and 25 mM Tris (pH 7.2).
Reactions were stopped by adding 10 µl of 30% trichloroacetic acid.
A 30-µl aliquot was spotted on phosphocellulose paper (Whatman,
P-81). The papers were washed five times with 0.75% phosphoric acid,
once with acetone, and added to scintillation vials for counting. PKC
activity was determined by subtracting the background activity measured
in the absence of sample from that measured in its presence. The
PKC-
peptide was chosen as the substrate in our PKC assays since
this 12-residue synthetic peptide (PRKRQGSVRRRV-NH2) is an
effective substrate of the calcium-insensitive protein kinase C-
and
other isoforms (22).
To trigger A2a desensitization in intact cells, PC12 cells at 70% confluency were treated with CGS21680 (CGS, 2.5 µM), an A2a-selective agonist, for 30 min. Membrane fractions were then collected from control and desensitized cells, and assayed for CGS21680- or forskolin-evoked AC activity. As we have reported earlier (1, 2), AC activities evoked by activation of A2a-R or by direct stimulation using forskolin were both significantly decreased in desensitized cells (Table I), confirming that AC is one of the major sites for the suppression of A2a-R-mediated cAMP signal during A2a desensitization. To dissect and identify the molecular events underlying the suppression of AC activity during A2a desensitization, we examined whether this phenomenon can be reproduced in a cell-free system as in intact cells. A 50,000 × g membrane fraction collected from PC12 cells was incubated with or without CGS21680 for 30 min to stimulate A2a-R and trigger A2a desensitization. As shown in Table I, significant inhibition of the A2a-R-stimulated and of the forskolin-stimulated AC activity was observed in membrane fractions pretreated with CGS21680. In addition, similar extents of AC inhibition during A2a desensitization were observed in the P1 membrane fractions and in intact cells (Table I). Therefore, molecules that are responsible for the suppression of AC activity during A2a desensitization in intact PC12 cells must exist in the P1 membrane fractions. We designated such inhibition of AC activity during prolonged exposure of PC12 membrane fractions to an A2a-selective agonist as "A2a desensitization in vitro."
|
We have previously reported that AC6 appears to be the major AC involved in A2a desensitization in PC12 cells (2). AC6 activity in PC12 cells was determined as the difference between the forskolin (5 µM)-evoked AC activity assayed in the absence and the presence of 30 µM free calcium (2). Similar to what we have observed in intact PC12 cells, AC6 activity was inhibited by 44 ± 4% (mean ± S.E. of 12 determinations from four independent experiments) in P1 membrane fractions pretreated with CGS21680 to trigger A2a desensitization in vitro. Therefore, AC6 is also inhibited during A2a desensitization in vitro as that observed in intact PC12 cells.
Nucleotide Dependence of A2a Desensitization in VitroTo
further characterize A2a desensitization in vitro, we
examined the nucleotide dependence of this phenomenon. As shown in Fig.
1A, removal of GTP from the desensitization
reaction significantly reduced inhibition of AC activity, indicating
that A2a desensitization is dependent upon GTP and may be mediated
through a GTP-binding protein.
Since our previous study (2) suggested that protein phosphorylation might mediate the suppression of AC6 activity during A2a desensitization, we examined whether A2a desensitization in vitro required ATP. As shown in Fig. 1B, lowering of ATP to 0.1 mM almost completely abolished the suppression of AC6 activity during A2a desensitization in vitro, indicating that the suppression of AC6 activity may be mediated through an ATP-dependent pathway. This observation is consistent with our previous hypothesis that AC6 is regulated through protein phosphorylation during A2a desensitization. Exclusion of GTP or ATP during A2a desensitization consistently and significantly reduced the forskolin-evoked AC activity (data not shown). The physiological significance of this observation however remains unclear.
Protein Phosphatase 2A Blocked the Inhibition of AC6 during A2a Desensitization in VitroTo further investigate the role of protein phosphorylation in suppression of AC6 during A2a desensitization in vitro, we examined the effect of protein phosphatase 2A (PP2A) on the regulation of AC6 during A2a desensitization in vitro. As shown in Table II, incubation of the P1 membrane with PP2A (0.1 unit) completely blocks the inhibition of AC6 during A2a desensitization in vitro. This observation is consistent with our previous finding that PP2A in intact cells plays a very important role in the regulation of AC6 activity during A2a desensitization (2). Furthermore, inhibition of AC6 activity may be mediated through a protein phosphorylation mechanism during A2a desensitization.
|
To
investigate the potential role of protein kinases, inhibitors of PKC or
PKA were included during A2a desensitization in vitro. As
shown in Table II, an H-series compound, H8, which inhibits both PKA
and PKC at micromolar concentrations (23) blocked the inhibition of AC6
activity during A2a desensitization. By contrast, a PKA-selective
inhibitor (H89) which is 30 times more potent than H8 in inhibiting PKA
(23), exerted no significant effect on the inhibition of AC6 during A2a
desensitization. Therefore, PKA is unlikely to inhibit AC6 activity
during A2a desensitization. Alternatively, PKC might mediate the
reduction of AC6 activity during prolonged activation of A2a-R. To test
the above hypotheses, the effect of a pseudosubstrate of PKC
(PKC19-36) was examined. At 1.5 µM,
PKC
19-36 almost completely blocked the suppression of
AC6 activity, suggesting that PKC might play a cardinal role in A2a
desensitization (Table II). This peptide inhibitor of PKC has no
significant effect on smooth muscle myosin light chain kinase,
calcium/calmodulin-dependent protein kinase II and PKA (24)
at the concentration tested. To further verify the role of PKC, PC12
cells were treated with a PKC-stimulating phorbol ester (PDD, 100 nM), for 17 h to induce proteolysis and "down-regulation" of PKC. In these PKC down-regulated cells, no inhibition of AC6 activity during A2a desensitization in
vitro was observed. These results strongly suggest that PKC, but
not PKA, mediates the suppression of AC6 during A2a desensitization resulting from prolonged activation of A2a-R.
To further
confirm the role of PKC in suppression of AC6 during A2a
desensitization, we examined whether addition of purified PKC which
contains a mixture of PKC isozymes (,
I,
II,
,
,
and
µ; data not shown) from rat brain suppressed AC6 activity. As shown
in Table III, PKC caused a 54 ± 3% inhibition of
AC6 activity in the P1 membranes of PC12 cells. To facilitate the study
of biochemical properties of AC6, we overexpressed AC6 and F-AC6 in
insect cells (Sf21) using the baculovirus expression system. F-AC6 is a
chimeric AC6 with an N-terminal peptide containing an 8-amino acid
epitope (DYKDDDDK, designated as FLAG). This epitope can be recognized
by a commercially available antibody (M2) and has been fused to various
proteins, including AC2 (11), with no conspicuous effect on their
activity (25). Likewise, no difference in the enzymatic properties of
AC6 and F-AC6 was detectable (data not shown). Exogenous PKC
dramatically inhibited AC6 activity in membrane fractions collected
from Sf21 cells overexpressing AC6 or F-AC6 as indicated (Table
III).
|
In order to determine whether
PKC phosphorylates AC6, we prepared two polyclonal antibodies
(designated AC6C and AC6D) which recognize amino acid 1162 ~ 1181 of AC6 and amino acid 987 ~ 1187 of rat AC6, respectively.
Both antibodies are capable of immunoprecipitating F-AC6 expressed in
Sf21 cells as shown in Fig. 2. We then purified AC6 and
F-AC6 from Sf21 cells expressing the indicated cDNA using AC6C by
immunoprecipitation. Phosphorylation of AC6 or F-AC6 by PKC (0.1 milliunit in a 200-µl reaction) was then carried out at 4 °C for
30 min in the presence of [-32P]ATP. As shown in Fig.
3, PKC phosphorylated both AC6 and F-AC6. These data
illustrate that PKC not only inhibits AC6 activity, also directly
phosphorylates this calcium-inhibitable AC.
CGS21680 Pretreatment in Intact PC12 Cells Prevents Further Inhibition of AC6 Activity and Leads to Less Phosphorylation of AC6 by Exogenous PKC
To determine that the regulation of AC6 during A2a
desensitization in PC12 cells was indeed due to a PKC, PC12 cells were pretreated with CGS21680 (2.5 µM) for 16 h to obtain
maximal inhibition of AC6 (1). P1 membrane fractions of control (cells
pretreated with no agonists) or CGS21680-pretreated cells were then
incubated with exogenous PKC for 30 min at 4 °C. Consistent with our
previous observation, exogenous PKC inhibited AC6 activity in control
cells by 65 ± 8%. As predicted, pretreatment of PC12 cells with
CGS21680 for 16 h caused a typical 68 ± 5% reduction in AC6
activity compared to that in control cells. Most importantly,
incubation of CGS21680-pretreated membrane with PKC did not further
suppress the AC6 activity (Fig. 4). To demonstrate that
in vitro phosphorylation by PKC is truly related to the
desensitization of AC6, exogenous PKC phosphorylation of AC6 purified
from control or CGS21680-pretreated cells were examined. As shown in
Fig. 5, A2a desensitization profoundly reduced in
vitro phosphorylation of AC6 by PKC. Taken together, pretreatment of PC12 cells with CGS21680 which leads to A2a desensitization might
induce phosphorylation of AC6 at the same residues which are
phosphorylated by exogenous PKC in vitro.
Novel PKCs Might Play a Very Critical Role in the Reduction of AC6 Activity during A2a Desensitization
Since inhibition of AC6
during A2a desensitization was well demonstrated in the "cell-free"
P1 membrane preparation of PC12 cells with all of the characteristics
tested identical to those observed in intact cells, enzymes necessary
to suppress AC6 activity during prolonged activation of A2a-R must
exist in the P1 membrane fractions. Western blot analysis demonstrated
the existence of at least seven different PKC isozymes (,
,
,
,
, µ, and
) in the P1 membrane fractions of PC12 cells
(data not shown). Among these seven PKC isozymes, we are most
interested in the novel PKCs which are insensitive to calcium, but
require diacylglycerols (or phorbol esters) for their activation based
on the following two observations. First, A2a desensitization in
vitro occurs in the absence of calcium, suggesting that the PKC
responsible for the suppression of AC6 during A2a desensitization is
likely to be calcium-independent. Second, down-regulation of PKC by
pretreating PC12 cells with a PKC-stimulating phorbol ester (PDD)
prevented the reduction of AC6 during A2a desensitization (Table II).
Thus, the PKC involved in the regulation of AC6 must be sensitive to phorbol esters. We therefore examined the expression of novel PKCs in
P1 membrane fractions after 17-h exposure of PC12 cells to PDD. As
shown in Fig. 6, P1 membrane fractions contained
significant amounts of PKC
and PKC
, but not PKC
. As predicted,
treating PC12 cells with PDD (100 nM) for 17 h almost
completely eliminated the immunoreactive bands of PKC
and PKC
in
both the P1 membrane and the supernatant fractions. Under the same
conditions, no suppression of AC6 activity was observed during A2a
desensitization in vitro. These data imply that PKC
and
PKC
are the two most likely candidates for the reduction of AC6
activity during A2a desensitization.
To further strengthen the above hypothesis, regulation of AC6 activity
during A2a desensitization was examined in stably transfected PC12 cell
lines overexpressing PKC or PKC
characterized elsewhere (16). A
significantly lower AC6 activity was observed in both cell lines
transfected with PKC
(
1) or PKC
(
1) compared to that
observed in control cells (cells transfected with the pRc/RSV empty
vector, designated C1). Given that these PKC-overexpressing cells
exhibit significantly higher PKC activities and in turn might lead to
higher PKC-mediated phosphorylation, it is reasonable to hypothesize
that the lower AC6 activity in these cells might be due to a higher
level of AC6 phosphorylation. Intriguingly, prolonged activation of
A2a-R in these PKC-overexpressing cells did not further suppress AC6
activity whereas the same treatment resulted in a 67 ± 3%
inhibition in the control C1 cells (Fig. 7A),
a 50 ± 7% inhibition in the parental PC12 cells and a 45 ± 11% inhibition in a PKA-deficient PC12 variant (17), A123 cells (Fig.
7B). Since higher PKC activity in novel PKC-overexpressing cells (
1 and
1) and prolonged activation of A2a-R both inhibited AC6 activity in a nonadditive fashion, AC6 is very likely to be regulated through a PKC-dependent pathway. These
observations strongly suggest that an increase in novel PKC-mediated
phosphorylation during prolonged activation of A2a-R inhibits AC6
activity in PC12 cells. In addition, PKA is not required for the
regulation of AC6.
Activation of A2a-R Leads to an Increase in the Calcium-insensitive PKC Activity
The question arises as to whether PKC is activated
during A2a desensitization. Since inhibition of AC6 during A2a
desensitization in vitro occurs in the absence of calcium,
the PKC responsible for the regulation of AC6 is likely to be
calcium-independent. Using a PKC-epsilon substrate peptide, we found
that the calcium-independent PKC activity was considerably enhanced by
treating the P1 membrane fractions with CGS21680 under the identical
conditions as those for A2a desensitization in vitro. The
PKC inhibitor, PKC19-36, almost completely abolished
the enhanced calcium-independent PKC activity by CGS21680 (Table
IV). Activation of A2a-R therefore indeed activated a
calcium-insensitive PKC which phosphorylated AC6 and caused a
subsequent inhibition of AC6 activity.
To
elucidate the signaling pathway underlying the suppression of AC6 by
PKC, we examined the effect of PTX on A2a desensitization. As shown in
Fig. 8, PTX treatment almost entirely blocked the suppression of AC activity resulting from prolonged activation of
A2a-R. These data indicate that a PTX-sensitive G protein is
involved in the inhibition of AC activity during A2a
desensitization.
Our previous studies indicated that activity of a Ca2+-inhibitable AC (AC6) during A2a desensitization may be regulated through a phosphorylation-dependent pathway (2). In the present study, we demonstrate that suppression of AC6 activity can be reproduced in 50,000 × g membrane fractions (P1) following prolonged stimulation of A2a-R. Thus, enzymes that regulate AC6 activity during A2a desensitization must exist in the P1 membrane fractions. Various lines of evidence strongly suggest that novel PKC isozymes in the P1 membrane fractions phosphorylate and inhibit AC6 activity after prolonged stimulation of A2a-R.
The cell-free membrane system provides a very useful and convenient
model to elucidate the molecular event(s) underlying the regulation of
AC6 during A2a desensitization. We have previously shown that A2a
desensitization in PC12 cells is heterologous (1). AC activities
stimulated by adenosine agonists, GTPS, and forskolin were equally
impaired, suggesting that the regulation occurred at the effector
enzyme itself, but not at the A2a-R nor at the Gs
protein. The suppression of AC6 during A2a desensitization remained
largely unchanged when GDP
S and MnCl2 were used to
replace GTP and MgCl2 in some cyclase assays. Thus, the
potential influence of Gs
protein in the inhibition of
AC6 was further ruled out (2). Although two adenosine receptors (A2a
and A2b) exist in PC12 cells, we have previously demonstrated that
desensitization of the A2a response in PC12 cells is due to prolonged
activation of A2a-R based on the following evidence (1). First, the
relative potency of three adenosine agonists to desensitize the A2a
response in PC12 cells is in the same order as their potencies for
activating A2a-R. Second, an adenosine antagonist
(8-cyclopentyl-1,3-dipropylxanthine) is able to block the
CGS21680-evoked inhibition of AC activity (1). CGS21680 is a very
selective A2a adenosine agonist (26), which does not activate A2b-R in
PC12 cells up to 300 µM (1). In the present study, we
have shown that the degree of AC inhibition during A2a desensitization
in intact cells and in the P1 membrane fractions is very similar (Table
I). Marked suppression of both CGS21680- and forskolin-evoked AC
activities were observed, suggesting that the reduction of A2a
signaling in P1 membrane fractions by prolonged activation of A2a-R is
caused largely by inhibition of AC itself as in intact cells. In
addition, inhibition of AC activity during prolonged activation of
A2a-R is GTP-dependent (Fig. 1A), further
confirming that this phenomenon is mediated through a GTP-binding
protein coupled pathway.
Stimulation of A2a-R has long been established to cause activation of
AC activity via Gs protein (27). However, the fact that
A2a desensitization requires GTP (Fig. 1A) and was not observed in the PTX-treated PC12 cells (Fig. 8), strongly suggests that
this phenomenon is mediated via a PTX-sensitive G
protein. Koizumi et al. (28) have recently reported that an
A2-selective agonist (CGS22492), but not an A1-selective agonist
(CCHA), potentiates ATP-evoked dopamine release via a pertussis
toxin-sensitive mechanism in PC12 cells in a cAMP-independent manner
(28). Their finding is consistent with our hypothesis that A2a-R may
transduce the cellular signal not only via Gs
, but also
through a PTX-sensitive G protein in PC12 cells.
We have previously hypothesized that protein phosphorylation of AC6 may account for the inhibition of its activity during prolonged exposure of PC12 cells to adenosine agonists. However, it was unclear whether regulation of the phosphorylation state of AC6 resulted from activation of an unknown kinase or inactivation of PP2A during A2a desensitization (2). Using the cell-free system, we now show that the inhibition of AC6 during A2a desensitization is ATP-dependent (Fig. 2B), suggesting that activation of a kinase is required. Since stimulation of A2a-R leads to an increase in intracellular cAMP in PC12 cells, PKA is an attractive candidate for the regulation of AC6 activity during A2a desensitization. However, treatment with dibutyryl-cAMP or forskolin did not reproduce the inhibition of AC activity during A2a desensitization in PC12 cells (1). Hence, PKA by itself is not sufficient to trigger inhibition of AC6. Furthermore, a very selective and potent PKA inhibitor (H89) at 10 µM did not exert any detectable effect on the inhibition of AC6 during A2a desensitization. This PKA-selective inhibitor is an H-series kinase inhibitor which blocks PKA by competing with ATP (23). At the same concentration (10 µM) as that tested in A2a desensitization (Table II), H89 almost entirely abolished the activation of phosphodiesterase activity during prolonged activation of A2a-R in PC12 cells.2 In addition, the extent of AC6 suppression during A2a desensitization in a PKA-deficient PC12 variant (A123) is comparable to that observed in the parental PC12 cells (Fig. 7B). Taken together, elevation of cellular cAMP upon stimulation of A2a-R, that subsequently leads to activation of PKA, is not responsible for the inhibition of AC6.
In contrast to PKA, PKC appears to play a very critical role in the
regulation of AC6. Inhibiting PKC through two independent means
(down-regulation of PKC with PDD and use of a PKC-specific inhibitor)
effectively blocked the suppression of AC6 during A2a desensitization
in the cell-free membrane fractions (Table II). In addition, incubation
of membrane fractions expressing AC6 with purified PKC caused a marked
reduction of AC6 activity (Table III). PKC also directly phosphorylated
purified AC6 (Fig. 3). A specific role for PKC in regulating AC6 during
A2a desensitization in PC12 cells is further suggested by the lack of
effect of exogenous PKC on the activity of AC6 isolated from
desensitized cells (Fig. 4). Furthermore, in vitro PKC
phosphorylation of AC6 obtained from desensitized cells was markedly
reduced (Fig. 5), indicating that AC6 is indeed a substrate for PKC
during A2a desensitization. Most importantly, activation of A2a-R
resulted in a significant increase in the calcium-independent PKC
activity (Table IV). These data strongly support PKC as the kinase
responsible for the suppression of AC6 during A2a desensitization in
PC12 cells. This hypothesis is further strengthened by examining the
regulation of AC6 activity in PC12 clones overexpressing PKC or
PKC
. When compared to the control clone, the
Ca2+-independent PKC activities in clones overexpressing
PKC
(
1) and PKC
(
1) were increased by 2.2- and 1.6-fold,
respectively (14). Since the levels of the phosphorylation state of AC6
were expected to be higher in these PKC-overexpressing cells, it was not surprising to find that the AC6 activity in these cells was significantly lower than in the control cells (Fig. 7A).
Most strikingly, no further inhibition of AC6 activity was observed in
the PKC overexpressing cells after prolonged activation of A2a-R by
CGS21680, whereas the same treatment resulted in a 67 ± 3%
inhibition in control cells (Fig. 7A). Prolonged activation of A2a-R may therefore regulate AC6 activity through the very same
mechanism used in the PKC-overexpressing cells. In addition, AC6
appears to be very sensitive to small changes in PKC activity (~2-fold). All of the strategies utilized above provide convincing evidence to support our hypothesis that PKC inhibits the activity of
AC6 in PC12 cells during prolonged activation of A2a-R.
Modulation of AC isozymes by PKC has been actively examined in the past
few years. In brief, PKC has been shown to stimulate AC1, AC2, and AC5
(10-12). Phosphorylation of these AC isozymes, including AC2 and AC5,
affects mainly catalytic activities (11, 12). Most intriguingly,
PKC-evoked phosphorylation allows AC2 to integrate multiple signals
produced by Gq- and Gi
-coupled receptors
in a Gs
-independent manner (29). An inhibitory effect of
PKC on a specific, unidentified AC subtype which is expressed mainly in
hepatocytes has also been reported (30). We have shown in the present
study that PKC specifically inhibits AC6 activity through protein
phosphorylation. Such negative influences of PKC on a
Ca2+-inhibitable AC (AC6) provides an exquisite mechanism
for the fine-tuning of cAMP synthesis by Ca2+ and PKC.
Activation of PKC upon stimulation by extracellular signals may
therefore lead to negative regulation of AC6 and a suppression of
signals transduced by receptors coupled to AC6, such as A2a-R. The
effect of the PKC-induced phosphorylation on the biochemical properties
of AC6 is currently under investigation.
As early as 1987, Yoshimasa et al. (31) reported that the
effect of a PKC-stimulating phorbol ester
(12-O-tetradecanoyl phorbol-13-acetate) on AC activity was
inhibitory in some cells, but stimulatory in others. Overexpressing
various PKC isozymes did not convey the desired inhibitory or
stimulatory effect of phorbol esters on AC activity, suggesting that
the cellular selectivity of PKC on AC results from the existence of
specific AC subtypes in different cells (31). In addition, results
obtained from cells transiently expressing the desired AC subtype were
not always consistent with those acquired from purified AC and PKC
isozymes. For example, phorbol esters were found to exert "little, if
any, change" in AC activity of cells transfected with AC4, AC5 and AC6 (32, 33). However, Kawabe et al. (12) have reported that
PKC and PKC
both directly phosphorylate AC5 and markedly stimulate AC5 activity using purified PKCs and AC5. In the present study, we demonstrate that purified PKC directly phosphorylated and
inhibited AC6 (Table III and Fig. 3). The discrepancy between results
obtained from the two systems (in vivo versus in vitro) may
reflect, at least in part, the existence of distinct ACs and PKCs in
different cells.
In summary, our data suggest that PKC directly phosphorylates and suppresses AC6 activity in PC12 cells during prolonged activation of A2a-R. In addition, stimulation of A2a-R results in activation of both adenylyl cyclase and calcium-insensitive, novel PKC in PC12 cells.
We thank Dr. J. A. Wagner for kindly providing the PKA-deficient variant of PC12 cells. We are also grateful to Drs. L. Y. Chau, C.-M. Chiang, L.-S. Kao, and C. V. Weaver for their helpful suggestions and comments.