- Email: [email protected]
Phosphorylation and inactivation of rat heart glycogen synthase by cAMP-dependent and cAMP-independent protein kinases
Int. J. Biochern. Cell Biol. Vol. 27, No. 6, pp. 565 573, 1995
Pergamon
1357-2725(95)00029-1
Copyright @ 1995 Elsevier Science Ltd Printed in Great Britain. All rights reserved 1357-2725/95 $9.50 + 0.00
Phosphorylation and Inactivation of Rat Heart Glycogen Synthase by cAMP-dependent and cAMP-independent Protein Kinases D I M I T R I S G R E K I N I S , I E R W I N M. R E I M A N N , 2 K E I T H K. S C H L E N D E R l* Departments of 1Pharmacology, and :Biochemistry and Molecular Biology, Medical College of Ohio, Toledo, O H 43699, U.S.A. The regulation of cardiac muscle glycogen metabolism is not well understood. Previous studies have indicated that heart glycogen synthase is heavily phosphorylated in vivo on multiple sites. Using purified enzymes, we have investigated the effect of phosphorylation of different sites on the activity of rat heart glycogen synthase. A convenient procedure was developed for the purification of rat heart glycogen synthase. The enzyme was phosphorylated by selected kinases, and glycogen synthase activity, extent of phosphorylation, and phosphopeptide maps were analyzed. Rat heart glycogen synthase, purified to apparent homogeneity (M r 87,000 on SDS-PAGE), had a specific activity of 18 U/mg protein and bad an activity ratio of 0.74 (activity in the absence divided by the activity in the presence of glucose 6-P). cAMP-dependent protein kinase, glycogen synthase kinase 3, Ca2+/calmodulin-dcpendent protein kinase II, protein kinase C, and phosphorylase kinase phosphorylated the enzyme with a concomitant decrease in the activity ratio to values ranging from 0.1 to 0.4. Casein kinase II phosphorylated hut did not inactivate glycogen synthase. Six tryptic phosphopeptides, obtained from heart glycogen synthase phosphorylated by the various kinases, were separated by reverse-phase chromatography. The phosphopeptide(s) obtained with each kinase eluted at the same position(s) as corresponding phosphopeptides obtained from rat skeletal muscle glycogen synthase. The study shows that the pattern of pbosphorylation and effects on activity are very similar for cardiac and skeletal muscle glycogen synthase. It is suggested that the well known differences in heart and glycogen metabolism may be due to the interplay of kinases and phosphatases which could lead to different phosphorylation and activity states of glycogen synthase. Keywords: Glycogen
Heart muscle Phosphorylation Protein kinases
Glycogen
Synthase
Int. J. Biochem. Cell Biol. (1995) 27, 565-573
INTRODUCTION Mammalian glycogenesis is regulated by glycogen synthase (E.C. 2.4.1.11), an enzyme that is subject to both hormonal and non-hormonal control (Stalmans et al., 1987). Most of our knowledge of glycogen synthase subunit structure, kinetic properties, and the phosphorylation~tephosphorylation interplay between the different protein kinases and phosphatases that are involved in the regulation of the enzyme *To whom correspondence should be addressed. Received 15 July 1994; accepted I March 1995.
stem from studies on rabbit skeletal muscle glycogen synthase. The complete amino acid sequence of rabbit skeletal muscle glycogen synthase is known (Zhang et al., 1989) and at least 1 l different phosphorylation sites have been identified. The various sites can be phosphorylated in vitro by cAMP-dependent protein kinase (PKA) and various cAMP-independent protein kinases (reviewed in Roach et al., 1991). Rat heart glycogen synthase appears to be heavily phosphorylated on multiple sites in vivo (McCuliough and Walsh, 1979 and Ramachandran et al., 1983). Using purified bovine heart glycogen synthase preparations, it has been 565
566
Dimitris Grekinis et al.
shown that the enzyme can be phosphorylated in vitro up to 3 mol of [ 32p]phosphate/subunit by PKA and cAMP-independent protein kinases (Mitchell et al., 1980; Mitchell and Thomas, 1981 and Sivaramkrishnan et al., 1982). As in the skeletal muscle, inactivation appears to be correlated with the phosphorylation state of the enzyme. It is known that glycogen metabolism of cardiac muscle is quite different than that of skeletal muscle (Roach, 1986) and the hormonal regulation of heart glycogen synthase remains to be fully understood. It is known from studies using isolated heart preparations that insulin activation of heart glycogen synthase is independent of either Ca 2+ or cAMP, whereas fiadrenergic agonists and glucagon act via cAMP, and ~-agonists most likely act via intracellular Ca 2+ translocation (Ramachandran et al., 1982, 1983). Therefore, it appears that the phosphorylation and inactivation of heart glycogen synthase in vivo might by catalyzed by at least three types of protein kinase: (1) cAMPdependent, (2) Ca2+-dependent, and (3) cAMP/Ca 2+-independent. In the present study, the effects of phosphorylation on the activity of highly purified rat heart glycogen synthase using cAMPdependent, Ca2+-dependent, and cAMP/Ca 2+independent protein kinases was investigated. In addition, the sites labeled in response to the action of the specific kinases were examined. The results suggest that the rat heart and skeletal muscle enzymes share a great degree of similarity. A preliminary account of this work has been reported (Grekinis et al., 1992).
calmodulin-dependent protein kinase II (CAMkinase II) purified from rat forebrain (Hashimoto et al., 1987) was a gift from Dr T. Soderling of Vollum Institute, Oregon Health Science University, Portland, OR. Protein kinase C was partially purified from rat brain (Shearman et al., 1989). Casein kinase II (CK II) was purified from skeletal muscle according to Litchfieid et al. (1990) up to the phosphocellulose step with minor modifications. Purification o f rat heart glycogen synthase
Male Sprague-Dawley rat hearts were excised rapidly, the major blood vessels were removed, the hearts were blotted of excess blood between paper towels, and were rapidly frozen between two dry ice plates. All subsequent steps were carried out at 4°C unless otherwise indicated. Starting with about 70 g of tissue, the procedure of Jaspers et al. (1989) was followed through the first ultracentrifugation step. Briefly, the tissue was thawed in four Vol of ice-cold 50mM Tris-HC1 (pH 7.6), 10mM EDTA, 5mM dithiothreitol (DTT), 10% glycerol, 10 mg oyster glycogen/ml, 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, 1/~g leupeptin/ml, and 1 #g pepstatin A/ml and homogenized in a Waring blender. The homogenate was centrifuged at 13,000 g for 30 min. The supernatant was filtered through glass-wool and then centrifuged at 140,000g for 2hr. The resulting pellets were resuspended at room temperature with the aid of a glass/teflon tissue homogenizer in 150ml of 50mM Tris-HC1, pH 7.6, 1 mM EDTA, 50mg sucrose/ml, 10 mM magnesium acetate, 2 mM DTT, 1 mM each of benzamidine and phenylmethylsulfonyl fluoride, 1 #g leupeptin/ml, 1/~g pepstatin A/ml, and 20/~g soyMATERIALS AND M E T H O D S bean trypsin inhibitor/ml. After adding human Glycogen synthase assay salivary ~-amylase (0.1 mg/ml) the solution was Glycogen synthase was assayed by the filter incubated at 30°C for 45min to digest the paper assay according to Thomas et al. (1968). glycogen and to allow dephosphorylation of One unit of enzyme activity corresponds to the glycogen synthase by endogenous phosphatases. incorporation of 1 #mol of glucose from ~4C- The sample then was recentrifuged at 80,000g UDPglucose into glycogen per min under con- for 90 min. The pellet was discarded and the ditions of the standard assay. The activity ratio supernatant was applied to a 50 ml is the ratio of enzyme activity measured in the DEAE-Sephacel column equilibrated in 50 mM absence and presence of 6.7 mM glucose-6-P. Tris-HC1, 1 mM EDTA, pH 7.6, 5 mM DTT, 10% glycerol, and the protease inhibitors Protein kinases described above (Buffer A). The resin was The catalytic subunit of PKA from bovine washed with 4 Vol of Buffer A and a linear heart (Reimann and Beham, 1983), phosphoryl- gradient (0-500 mM NaC1) was developed. The ase kinase (Cohen, 1973), and glycogen synthase peak activity fractions were combined. One kinase-3 (GSK-3) (Hegazy et al., 1987) were volume of Buffer B (50raM /~-glycerophospurified by published procedures. Ca2+/ phate, pH7.6, 1 mM each of MgC12, MnCI2, and
Rat heart glycogensynthase CaC12, 2raM DTT, 500mM NaC1, and the protease inhibitors of Buffer A) was added to the pooled fractions and the sample was loaded onto a 10ml concanavalin A Sepharose 4B column equilibrated in Buffer B. The flowthrough was recirculated overnight through the resin, and the column was washed with Buffer B. The enzyme was eluted with 200 mM sucrose, 5 mM E D T A in Buffer B. One mg glycogen/ml was added to the active fractions and the solution was incubated for 60 min at 4~C. Samples of 1 ml were layered onto a 24ml sucrose solution (40% w/v in 5 0 r a m Tris-HC1, pH 7.8, 5 mM EDTA, 2 mM EGTA, 1 mM DTT), and subjected to high speed centrifugation at 200,000g for 17 hr (Dickey-Dunkirk and Killilea, 1985). The precipitated enzymeglycogen complex was resuspended in 50raM Tris-HCl, pH 7.0 containing l mM DTT. At this stage, the enzyme could be stored at - 70~'C for at least 12 months without any loss of activity. For some experiments, the enzyme had to be free of glycogen and the procedure of Mellgren (1976) was used to digest the excess glycogen. Phosphorylation of glycogen synthase Unless otherwise noted, the phosphorylation reactions (100~tl) contained 0.1 mg glycogen synthase/ml, 0.05 mM ~,-[32p]ATP (1001000cpm/pmol), and 10 mM magnesium acetate. Phosphorylation by PKA or GSK-3 was carried out in the presence of 50 mM Hepes, pH 7.5, 3 0 m M NaCl, l mM EDTA, and l mM DTT. Phosphorylation by CaM-kinase II was performed in a buffer containing 50 mM Hepes, pH 7.5, l mM DTT, 0.2raM 7-[32p]ATP, 0.5raM CaC12, and l~tM calmodulin. Phosphorylation by protein kinase C was carried out in a buffer containing 27 mM Tris-HCl, pH 7.7, 1 mM EGTA, 0.16 mM calcium acetate, 64 ktM 1 2 - 0 - tetradecanoylphorbol - 13 - acetate, and 26ktg phosphatidylserine/ml. Phosphorylation by phosphorylase kinase was carried out in the presence of 25 mM fl-glycerolphosphate, 25raM Tris-HCl, pH 7.5, l mM DTT, and 1.5mM CaCI> whereas the phosphorylation buffer for CK II was 40 mM Tris-HC1, pH 7.7, 0,8 mM EDTA, 0,2 mM EGTA, 1 mM DTT, and 6% glycerol. In some phosphorylation experiments bovine serum albumin (0.05 mg/ml) was added. When kinases other than PKA were used, the heat-stable protein inhibitor of PKA was included. At various time points 5-101zl aliquots were removed for analysis of
567
[3ZP]phosphate incorporation into the glycogen synthase as described before (Roskoski, 1983). At the same time points, changes in activity ratio were determined using the same assay conditions with non-radioactive ATP instead of [7--~2p]ATP. At predetermined time points, 4/~1 aliquots were diluted about 30-fold with standard glycogen synthase dilution buffer containing 15raM EDTA and glycogen synthase activity was determined in the absence and presence of glucose-6-P under standard synthase assay conditions. H P L C peptide mapping Samples containing [32p]glycogen synthase phosphorylated with the appropriate kinase were precipitated using trichloroacetic acid, and were processed for complete trypsin digestion and HPLC analysis of the obtained phosphopeptides as described by Hegazy et al. (1987). Alternatively, the phosphorylation reaction was stopped by the addition of 0.5 Vol electrophoresis sample buffer (150 mM Tris-HCl, pH 6.8, 0.01 mM EDTA, 30% glycerol, 0.05% bromophenol blue, 3.3% SDS and 7.66% /~mercaptoethanol). After SDS-PAGE, the gel was dried under heat and vacuum onto Whatman 3 MM Chr filter paper. The band corresponding to glycogen synthase was sliced from the dried gel and incubated directly with trypsin (1 mg/ml) at 30~-C for 24hr. Similar results were obtained with either method. A 25 cm Synchropak RP, C~8 column (Synchron Inc., Linden, West LaFayette, Ind.) was used for all experiments together with the Waters Model 510 HPLC System. A gradient from 0.1% trifluoroacetic acid in water (pump A) to 0.08% trifluoroacetic acid in 60% acetonitrile (pump B) was employed. The flow rate was 1 ml/min and fractions of 1 ml were collected, The radioactivity was determined using Cerenkov counting. Other methods and materials Rat skeletal muscle glycogen synthase was purified according to Hegazy et al. (1987). The heat stable inhibitor of PKA was purified from rabbit skeletal muscle (Schlender et al., 1983). The phosphorylated amino acids of [32P]glycogen synthase were identified as described by Cooper et al. (1983). 32P-Labeled glycogen synthase was fragmented with CNBr as described by Hegazy et al. (1987). S D S - P A G E was performed on a Hoefer mini gel apparatus according to Laemmli (1970). For
568
Dimitris Grekinis et al. Table 1. Purification of glycogen synthase from rat heart
Extract 1st Ultrafiltration DEAE pool Concanavalin A pool 2nd Ultrafiltration
Volume (ml)
Protein (mg)
Activity (U)
S.A. (U/mg)
Purification (X)
Yield (%)
230 148 152 4.5 1.15
4487 1957 52.1 4.1 0.587
147 130 47.7 11.5 10.5
0.0328 0.06643 0.916 2.810 17.9
1
2.0 28 85.8 547
100 88 33 7.5 7.2
Glycogen synthase was purified from 70 g rat heart.
autoradiography of the gels, Kodak SB Panoramic dental X-ray film (DF-85) with an intensifying screen was used. Protein concentration .was determined by the method of Bradford (1976) using bovine serum albumin as the standard. Common laboratory reagents were of reagent grade and were obtained from standard suppliers. ~4C-UDPglucose and )~_32p_ 12 --
.=
--
6
-
4
-
J
0.
O 0
Relative
mobility
of a band at 77,000 (Fig. 1). The purified enzyme was virtually free of contaminating glycogen synthase kinase (Table 2) or glycogen synthase phosphatase activity (data not shown).
a ~
b
-~
GS
ofi 4.9
Purification of glycogen synthase
act. was 17.9 U/rag protein. The enzyme exhibited a major band on S D S - P A G E with an apparent molecular weight o f 87,000 and a trace
(B)
5.3 --
RESULTS
A modified procedure was developed for the purification of dephosphorylated rat heart glycogen synthase. The results of the purification are summarized in Table 1. Glycogen synthase, eluted from the concanavalin-A Sepharose column, proved to be only partially purified (spec. act. 2.8U/mg protein) and additional purification was necessary. High speed centrifugation of the glycogen synthase-glycogen complex through 40% sucrose removed all the contaminating proteins present in the previous step and more than 95% of the enzyme could be recovered. The activity ratio of the final preparation was 0.74 and spec.
(A)
10--
8
labeled ATP were obtained from ICN or NEN, CNBr was from Aldrich, acetonitrile from Baker, L-l-tosyl-amino-2-phenylchloromethyl ketone treated trypsin, and soybean trypsin inhibitor from Worthington Biochem Corp.
- -
Table 2. Phosphorylation of rat heart glycogen synthase by cAMP-dependent and cAMP-independent protein kinases
O ,d
Protein kinase
mol [32P]phosphate per mol subunit
Activity ratio
0.0 1.00 0.93 0.15 0.21 0.41 0.80
0.74 0.25 0.10 0.75 0.45 0.40 0.40
e
0 0.1
I
I
I
I
I
I
0.2
0.3
0.4
0.5
0.6
0.7
Relative mobility Fig. 1. SDS-PAGE of the purified glycogen synthase (3 vg). (A). Optical density scan of Coomassie stained SDS-polyacrylamide gel. (B) Relative mobility of the glycogen synthase and protein standards: a, myosin (200,000); b, fl-galactosidase (116,000); c, phosphorylase a (97,000); d, bovine serum albumin (66,000); e, ovalbumin (43,000); and GS, glycogen synthase.
No Kinase PKA GSK-3 CK II Protein Kinase C Phosphorylase Kinase CaM-Kinase II
Heart glycogen synthase was maximally phosphorylated by the indicated kinase and the activity ratios were determined. Conditions for phosphorylation and assay of glycogen synthase were as described in "Materials and Methods".
Rat heart glycogen synthase
(In--O)l!unqnsio,,,/ IO~d~z
:l!unqns lOW / lOtU a~ Z
(0--~)
c~ I
I
I
569
d
I
d
o
6
o
I
I
I
t
'
m
I
I l 0
o
o <-o
~,
•
o
c" o
I~
0
/\ I
0
,
', ~',
I
',
;
o. o! 0q r-: ,q ,,q. , ~
o
o
o
o
o
I
I
I
o
o
o
d
•
/
Io
I
,"! e!. -
\
•
~\
/•
,I
,
0 0
0
,
,
0
I
i,
/
I
0
.o
.~
IZ]
/
Nn
,
,,
0
0
I 0
I 0
l
~d
\
]0
I
0
0
o
~'~
e~
0
~¢ - .g. _ e. N (E]--O)
(o--n) l!unqn~ lOW/lOWclz, Z
%!unqn~;lOW / IOLU dzE
m
~
o
o
'
I
I
c~
c~
I
I
o
<
d
d
I
I
~5 I
d
d o
I
O~
m
0
e~ 0
;-I
o
•
f
0
0
I
• C) if3
\ [3.
.I
\n
[]
0
~
m~
O •
._~ ~0
~0
"~-
o~
0
d
d
d
d
d
e4
~
•
I
0
o
0
0
,c E
0
/
~~ ' ~c
~.
~n
~
,.
e4
Q
~
d
d
d
6
o
(m--m) So ;o o!wJ xl!^!~ov
e~
570
Dimitris Grckinis et al. 1
Phosphorylation o/,zlvcojZen st'nthuse
2
3
4
5
When glycogen synthase was phosphorylated by the different protein kinases, essentially all of the [~2P]phosphate was incorporated onto serine residues associated with an 87,000 Da protein band on SDS P A G E (data not shown). The incorporation of [3:P]phosphate by the different kinases and the corresponding activity ratio changes can be seen in Fig. 2. The maximum amount of ['eP]phosphate incorporated in the glycogen synthase subunit are summarized in Table 2. One tool [~2P]phosphate/mol subunit was introduced by PKA, whereas CK 2 introduced the lowest amount of [32P]phosphate. GSK-3. a major glycogen synthase kinase in skeletal muscle, incorporated a total of about one tool [~:P]phosphate/mol subunit and produced the greatest decrease in activity ratio, whereas incorporation of [~P]phosphate by CK It resulted in no detectable change in the activity ratio of the enzyme. Removal of glycogen by digestion of the puri[ied glycogen synthase with ~,-amylase prior to phosphorylation with the kinases did not change the amount of [3-~P]phosphate incorporated or the activity changes (data not shown).
Fig. 3. CNBr digestion of the rat heart glycogen synthase. The enzyme was phosphorylated with the appropriate kinase at 30 C, the reactions were stopped by adding trichloroacetic acid (10% final), and the precipitated pellets werc washed and resuspended in 70% formic acid. CNBr was added and the digestion was carried out in the dark at room temperature for 8hr. The resulting digest was lyophilized, resuspended in electrophoresis sample buffer and loaded on a 5 18% gradient SDS polyacrylamidegel. A picture of the autoradiogram of the gel is presented. Phosphorylation levels of the glycogen synthase are the same as those described on Table 2. Lane 1, PKA; lane 2, protein kinase C; lane 3, CK lI: lane 4, GSK-3; and lane 5, phosphorylase kinase.
CNBr digestion
Peptide mapping on C~,, recerse-phase H P L C
Rat heart glycogen synthase was phosphorylated by the selected protein kinases to the levels shown in Table 2. Digestion of the enzyme with CNBr followed by SDS P A G E showed that in all cases the [':P]phosphate was incorporated into one or two phosphopeptides (Fig. 3). Similar phosphopeptides, designated CB1 and CB2 were obtained from rat skeletal muscle glycogen synthase (Hegazy el al., 1987). The apparent molecular weights of CBI and CB2 were about 13,000 and 23,000, respectively. Phosphorylation of the enzyme by PKA resulted in the incorporation of [~eP]phosphate predominantly oll the CB2 fragment, but CBI also contained significant radioactivity (about 25% of the total). Phosphorylation by GSK-3 or protein kinase C resulted in incorporation of phosphate predominantly in CB2 with smaller amounts in CB1. Phosphorylation by phosphorylase kinase resulted in incorporation of [~-P]phosphate only in CBI. Casein kinase II incorporated all the ['2P]phosphate into CB2 fragment. These results indicated that the phosphorylation sites in heart glycogen synthase may be similar to skeletal muscle glycogen synthase.
The phosphorylation sites on rat heart glycogen synthase were further defined by H P L C analysis of the ~P-labeled tryptic peptides obtained after phosphorylation by the different protein kinases (Fig. 4). The chromatogram obtained from the phosphorylation of the enzyme by PKA showed that the radioactivity was mainly incorporated in three different peptides with a retention time of 15, 40, and 51 rain. The [~2P]phosphate incorporated into glycogen synthase by GSK-3 was recovered in three peptides with retention times of 31, 37, and 51 rain each containing 52, 37 and 11%, respectively, of the total radioactivity incorporated. All the radioactivity incorporated by phosphorylase kinase was recovered in only one phosphopeptide (51 rain). Protein kinase C incorporated ['-~P]phosphate into sites corresponding to peptides with retention times of 15 and 51 rain, whereas CK II introduced all the [32p]phosphate in only one peptide with a retention time of 18 rain. The tryptic phosphopeptide maps obtained with heart glycogen synthase were very similar to those obtained with rat skeletal muscle glycogen synthase (data not shown and Hegazy et al., 1987): therefore, we used the same
23 kDa
13 kDa
'O 8 ~
Rat heart glycogen synthase
571
(A)
3
r..)
(B)
3
2
1
0
(C/
(D)
~
1
0
;
3
N
2 1
0
10
20
30
40
50
60
Minutes Fig. 4. HPLC chromatograms of phosphorylated rat heart glycogen synthase by different protein kinases. Phosphorylation by PKA, GSK-3, phosphorylase kinase, protein kinase C, and CK 1I (A, B, C, D, and E, respectively). Phosphorylation levels of the glycogen synthase are the same as those described in Table 2.
nomenclature. Peptides with a retention time of 15, 18, 40, and 51 min were designated as peptides la, 5, lb and 2, respectively. The skeletal muscle glycogen synthase phosphorylation sites known collectively as sites 3 are three closely
spaced serines which are in a single tryptic peptide (Roach, 1990). This phosphopeptide, presumably because of different states of phosphorylation, eluted as two peaks (Hegazy et al., 1987); thus, the two peaks obtained from syn-
572
Dimitris Grekinis et al.
thase phosphorylated by GSK 3 which eluted at 31 and 37 min were designated as the T-3 peptide. DISCUSSION
We developed a modified procedure for the purification of rat heart glycogen synthase which combines strategies used for the purification of rat heart (Jaspers et al., 1989) and bovine heart (Dickey-Dunkirk and Killilea, 1985) glycogen synthase. This convenient procedure resulted in a practically homogeneous enzyme preparation, as judged by SDS PAGE, with an apparent molecular weight of 87,000 (Fig. 1). A minor protein band that was present at about 77,000 in some preparations most likely represented a proteolytic cleavage product of the enzyme as reported for the bovine heart enzyme (Mitchell et al., 1980; Sivaramakrishnan et al., 1982, and Dickey-Dunkirk and Killilea, 1985). The [32P]phosphate incorporated by the kinases was located in the 87,000 Da species, and virtually no [32P]phosphate was detected when no protein kinase was added to the reaction indicating that the preparation was free of contaminating glycogen synthase kinase(s). Purified rat heart glycogen synthase was phosphorylated by PKA and a number of cAMP-independent protein kinases and calcium-dependent protein kinases. The phosphorylation levels obtained with rat heart glycogen synthase are similar to those reported for the phosphorylation of the rat skeletal muscle enzyme by the same kinases (Hiken and Lawrence, 1985 and Hegazy et al., 1987). Higher phosphorylation levels have been reported for bovine heart enzyme catalyzed by PKA, as well as a mixture of non-characterized cAMPindependent protein kinases (Mitchell and Thomas, 1981, Mitchell et al., 1980 and Sivaramakrishnan et al., 1982). Except for CK II, the incorporation of [32P]phosphate into glycogen synthase was associated with inactivation of the enzyme. Although the inactivating kinases introduced similar maximum amounts of phosphate into the enzyme, different changes in activity ratios were observed suggesting that different sites were being phosphorylated (DePaoli-Roach et al., 1983). McCullough and Walsh (1979) reported that the CNBr phosphopeptide maps from bovine heart glycogen synthase are similar to those obtained from rabbit skeletal muscle, whereas others (Mitchell et al., 1980) reported that differences might exist between bovine heart and
rabbit skeletal muscle enzymes. In the present study, the SDS-PAGE phosphopeptide maps obtained from CNBr digestion of rat heart enzyme phosphorylated by the different protein kinases, showed that [32p]phosphate was located mainly on the phosphopeptides termed CB1 and CB2 (Fig. 3). Similar CNBr profiles for the incorporation of phosphate into rat skeletal muscle by PKA, GSK-3, CK II, and phosphorylase kinase have been reported (Hiken and Lawrence, 1985 and Hegazy et al., 1987). Knowledge of the distribution of the phosphate on the enzyme molecule, incorporated by a given kinase, could provide valuable information as to which kinases are involved under defined in vivo conditions, e.g. under hormonal stimulation. Clearly, the separation of the phosphopeptides by reverse-phase HPLC chromatography after complete trypsin digestion can give more complete characterization of the phosphorylation sites. Reversed phase HPLC of rat heart glycogen synthase phosphopeptides was attempted by others without a clear separation (Wolleben et al., 1987). In this study we were able to separate the phosphopeptides resuiting from the tryptic digests of rat heart glycogen synthase phosphorylated by the various kinases. Since the phosphopeptides co-eluted with phosphopeptides obtained from rat skeletal enzyme the heart phosphopeptides were also designated Tla, Tlb, T2, T3, and T5. Of particular interest was the elution position of T2, the N-terminal phosphopeptide. Because the N-terminal sequence of rabbit and rat skeletal muscle glycogen synthase are not identical (Jaspers et al., 1989) the N-terminal phosphopeptides are easily separated by C~8 reversephase chromatography (Hegazy et al., 1987). The N-terminal sequence of rat skeletal muscle glycogen synthase and rat heart glycogen synthase are identical (Jaspers et al., 1989) and as expected rat heart T2 eluted at the same position as rat skeletal muscle T2. Heart glycogen metabolism is quite different than skeletal muscle metabolism (Roach, 1986). Although the complete amino acid sequence of heart glycogen synthase has not been determined it is evident from this and previous studies that structure of thee heart enzyme and the enzymatic properties are very similar to those of the skeletal enzyme. It seems likely that differences in the synthesis of heart and skeletal glycogen are determined by the interplay of the kinases and phosphatases which could lead to a difference in the phosphorylation state and thus
Rat heart glycogen synthase t h e a c t i v i t y s t a t e o f g l y c o g e n s y n t h a s e in t h e t w o tissues. Acknowledgements--We thank Susan Wilson for her technical assistance and Martha Heck for her secretarial assistance. This work was supported by Grant HL 36576 from the National Institutes of Health and by Grant 1881165 from the Juvenile Diabetes Foundation. REFERENCES
Bradford M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analvt. Biochem. 72, 248 254. Cohen P. (1973) The subunit structure of rabbit-skeletalmuscle phosphorylase kinase, and the molecular basis of its activation reactions. Eur. J. biochem. 34, I 14. Cooper J. A., Sefton B. M. and Hunter T. (1983) Detection and quantification of phosphotyrosine in proteins. In Methods Enzvmology (Edited by Corbin S. and Hardman S.), Vol. 99, pp. 387-402. Academic Press, New York. DePaoli-Roach A. A., Aham Z., Camici M., Lawrence J. C. Jr and Roach P. J. (1983) Multiple phosphorylation of rabbit skeletal muscle glycogen synthase. Evidence for interactions among phosphorylation sites and the resolution of electrophoretically distinct forms of the subunit. J. biol. Chem. 258, 10702-10709. Dickey-Dunkirk S. and Killilea S. D. (1985) Purification of bovine heart glycogen synthase. Analyt. Biochem. 146, 199-205. Grekinis D., Reimann E. M. and Schlender K. K. (1992) Purification and phosphorylation of rat heart glycogen synthase. FASEB J. 6, A340. Hashimoto Y., Schworer C. M.~ Colbran R. J. and Soderling T. R. (1987) Autophosphorylation of Ca2~/ calmodulin-dependent protein kinase II. Effects on total and Ca2+-independent activities and kinetic parameters. J. biol. Chem. 262, 8051-8055. Hegazy M. G., Schlender K. K. and Reimann E. M. (1987) Differences between glycogen synthases from rat and rabbit skeletal muscle as indicated by phosphopeptide maps. Biochim. biophys. Acta 927, 269 279. Hiken J. F. and Lawrence J. C. Jr (1985) Rat skeletal muscle glycogen synthase: phosphorylation of the purified enzyme by cAMP-dependent and -independent protein kinases. Arch. biochem. Biopho's, 236, 59 71. Jaspers S. R.. Rulfs J., Johnson G. L., Mole J. E. and Miller T. B. Jr (1989) Amino-terminal sequence analysis of rat heart and muscle glycogen synthase: homology to the rabbit enzyme and the implications for hormonal control. Arch. biochem. Biophys. 268, 630-636. Laemmli U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. Litchfield D. W., Lozeman F. J., Piening C., Sommercorn J., Yakio K., Walsh K. A. and Krebs E. G. (1990) Subunit structure of casein kinase II from bovine testis. Demonstration that the alpha and alpha" subunits arc distinct polypeptides. J. biol. Chem. 265, 7638-7644. McCullough T. E. and Walsh D. A. (1979) Phosphorylation of glycogen synthase in the perfused rat heart. J. biol. Chem. 254, 7336-7344. Mellgren R. L. (1976) Control of dephosphorylation of bovine heart glycogen synthase D. Ph.D. Dissertation, Iowa State University.
573
Mitchell J. W., Mellgren R. L. and Thomas J. A. (1980) Phosphorylation of heart glycogen synthase by cAMPdependent protein kinase. Regulatory effects of ATP. J. biol. Chem. 255, 10368-10374. Mitchell J. W. and Thomas J. A. (198l) Phosphorylation of bovine heart glycogen synthase by two protein kinases. Kinetic properties of phosphorylated forms of the enzyme. J. biol. Chem. 256, 6160~169. Ramachandran C., Angelos K. L. and Walsh D. A. (1982) Cyclic AMP-dependent and cyclic AMP-independent antagonism of insulin activation of cardiac glycogen synthase. J. biol. Chem. 257, 1448-1457. Ramachandran C., Angelos K. L. and Walsh D. A. (1983) Hormonal regulation of the phosphorylation of glycogen synthase in perfused rat heart. Effects of insulin, catecholamines, and glucagon. J. biol. Chem. 258, 13377 13383. Reimann E. M. and Beham R. A. (1983) Catalytic subunit of cAMP-dependent protein kinase. In Methods Enzymology (Edited by Corbin S. and Hardman S.), Vol. 99, pp. 51 55. Academic Press, New York. Roach P. J. (1986) Hormonal Control of Glycogen Metabolism. In Regulation of Heart Function: Basic Concepts and Clinical Appfcations (Edited by Rupp H.), pp. 201 -222. Thieme, New York. Roach P. J. (1990) Control of glycogen synthase by hierarchal protein phosphorylation. FASEB J. 4, 2961-2967. Roach P. J., Cao Y., Corbett C. A., DePaoli-Roach A. A., Farkas I., Fiol C. J., Flotow H., Graves P. R., Hardy T. A., Hrubey T. W., Viskupic E. and Zhang W. (1991) Glycogen metabolism and signal transduction in mammals and yeast. Adv. Enzyme Reg. 31, 101 - 120. Roskoski R. Jr (1983) Assays of protein kinase. In Methods Enzymology (Edited by Corbin S. and Hardman S.), Vol. 99, pp. 3~6. Academic Press, New York. Schlender K. K., Tyma J. L. and Reimann E. M. (1983) Preparation of a partially purified protein kinase inhibitor. In Methods Enzymology (Edited by Corbin S. and Hardman S.), Vol. 99, pp. 77-80. Academic Press, New York. Shearman M. S., Ogita K., Kikkawa U. and Nishizuka Y. (1989) A rapid method for the resolution of protein kinase C subspecies from rat brain tissue. In Methods Enzymology (edited by Cohn P.), Vol. 168, pp. 347 35l. Academic Press, New York. Sivaramakrishnan S., High C. W. and Walsh D. A. (1982) Regulation of cardiac glycogen synthase by phosphorylation: catalysis of inactivation by cAMP-dependent and cAMP-independent protein kinases and comparison with the phosphorylation of the skeletal muscle enzyme. Arch, biochem. Biophys. 214, 311-325. Stalmans W., Bollen M. and Mvumbi L. (1987) Control of glycogen synthesis in health and disease. Diab./Metab. Rez. 3, 127-161. Thomas J. A., Schlender K. K. and Larner J. (1968) A rapid filter paper assay for UDPglucose-glycogen glucosyltransferase, including an improved biosynthesis of UDPJ4C-glucose. Analyt. Biochem. 25, 486~-99. Wolleben C. D., McPherson R. K., Rulfs J., Johnson G. L. and Miller T. B. Jr (1987) Phosphorylation of rat heart glycogen synthase: studies in cardiomyocytes and in vitro phosphorylations with cAMP-dependent kinase and protein phosphatase-l. Biochim. biophys. Acta 928, 98 106. Zhang W., Browner M. F., Fletterick R. J., DePaoli-Roach A. A. and Roach P. J. (1989) Primary structure of rabbit skeletal muscle glycogen synthase deduced from cDNA clones. FASEB J. 3, 2532-2536.
Pergamon
1357-2725(95)00029-1
Copyright @ 1995 Elsevier Science Ltd Printed in Great Britain. All rights reserved 1357-2725/95 $9.50 + 0.00
Phosphorylation and Inactivation of Rat Heart Glycogen Synthase by cAMP-dependent and cAMP-independent Protein Kinases D I M I T R I S G R E K I N I S , I E R W I N M. R E I M A N N , 2 K E I T H K. S C H L E N D E R l* Departments of 1Pharmacology, and :Biochemistry and Molecular Biology, Medical College of Ohio, Toledo, O H 43699, U.S.A. The regulation of cardiac muscle glycogen metabolism is not well understood. Previous studies have indicated that heart glycogen synthase is heavily phosphorylated in vivo on multiple sites. Using purified enzymes, we have investigated the effect of phosphorylation of different sites on the activity of rat heart glycogen synthase. A convenient procedure was developed for the purification of rat heart glycogen synthase. The enzyme was phosphorylated by selected kinases, and glycogen synthase activity, extent of phosphorylation, and phosphopeptide maps were analyzed. Rat heart glycogen synthase, purified to apparent homogeneity (M r 87,000 on SDS-PAGE), had a specific activity of 18 U/mg protein and bad an activity ratio of 0.74 (activity in the absence divided by the activity in the presence of glucose 6-P). cAMP-dependent protein kinase, glycogen synthase kinase 3, Ca2+/calmodulin-dcpendent protein kinase II, protein kinase C, and phosphorylase kinase phosphorylated the enzyme with a concomitant decrease in the activity ratio to values ranging from 0.1 to 0.4. Casein kinase II phosphorylated hut did not inactivate glycogen synthase. Six tryptic phosphopeptides, obtained from heart glycogen synthase phosphorylated by the various kinases, were separated by reverse-phase chromatography. The phosphopeptide(s) obtained with each kinase eluted at the same position(s) as corresponding phosphopeptides obtained from rat skeletal muscle glycogen synthase. The study shows that the pattern of pbosphorylation and effects on activity are very similar for cardiac and skeletal muscle glycogen synthase. It is suggested that the well known differences in heart and glycogen metabolism may be due to the interplay of kinases and phosphatases which could lead to different phosphorylation and activity states of glycogen synthase. Keywords: Glycogen
Heart muscle Phosphorylation Protein kinases
Glycogen
Synthase
Int. J. Biochem. Cell Biol. (1995) 27, 565-573
INTRODUCTION Mammalian glycogenesis is regulated by glycogen synthase (E.C. 2.4.1.11), an enzyme that is subject to both hormonal and non-hormonal control (Stalmans et al., 1987). Most of our knowledge of glycogen synthase subunit structure, kinetic properties, and the phosphorylation~tephosphorylation interplay between the different protein kinases and phosphatases that are involved in the regulation of the enzyme *To whom correspondence should be addressed. Received 15 July 1994; accepted I March 1995.
stem from studies on rabbit skeletal muscle glycogen synthase. The complete amino acid sequence of rabbit skeletal muscle glycogen synthase is known (Zhang et al., 1989) and at least 1 l different phosphorylation sites have been identified. The various sites can be phosphorylated in vitro by cAMP-dependent protein kinase (PKA) and various cAMP-independent protein kinases (reviewed in Roach et al., 1991). Rat heart glycogen synthase appears to be heavily phosphorylated on multiple sites in vivo (McCuliough and Walsh, 1979 and Ramachandran et al., 1983). Using purified bovine heart glycogen synthase preparations, it has been 565
566
Dimitris Grekinis et al.
shown that the enzyme can be phosphorylated in vitro up to 3 mol of [ 32p]phosphate/subunit by PKA and cAMP-independent protein kinases (Mitchell et al., 1980; Mitchell and Thomas, 1981 and Sivaramkrishnan et al., 1982). As in the skeletal muscle, inactivation appears to be correlated with the phosphorylation state of the enzyme. It is known that glycogen metabolism of cardiac muscle is quite different than that of skeletal muscle (Roach, 1986) and the hormonal regulation of heart glycogen synthase remains to be fully understood. It is known from studies using isolated heart preparations that insulin activation of heart glycogen synthase is independent of either Ca 2+ or cAMP, whereas fiadrenergic agonists and glucagon act via cAMP, and ~-agonists most likely act via intracellular Ca 2+ translocation (Ramachandran et al., 1982, 1983). Therefore, it appears that the phosphorylation and inactivation of heart glycogen synthase in vivo might by catalyzed by at least three types of protein kinase: (1) cAMPdependent, (2) Ca2+-dependent, and (3) cAMP/Ca 2+-independent. In the present study, the effects of phosphorylation on the activity of highly purified rat heart glycogen synthase using cAMPdependent, Ca2+-dependent, and cAMP/Ca 2+independent protein kinases was investigated. In addition, the sites labeled in response to the action of the specific kinases were examined. The results suggest that the rat heart and skeletal muscle enzymes share a great degree of similarity. A preliminary account of this work has been reported (Grekinis et al., 1992).
calmodulin-dependent protein kinase II (CAMkinase II) purified from rat forebrain (Hashimoto et al., 1987) was a gift from Dr T. Soderling of Vollum Institute, Oregon Health Science University, Portland, OR. Protein kinase C was partially purified from rat brain (Shearman et al., 1989). Casein kinase II (CK II) was purified from skeletal muscle according to Litchfieid et al. (1990) up to the phosphocellulose step with minor modifications. Purification o f rat heart glycogen synthase
Male Sprague-Dawley rat hearts were excised rapidly, the major blood vessels were removed, the hearts were blotted of excess blood between paper towels, and were rapidly frozen between two dry ice plates. All subsequent steps were carried out at 4°C unless otherwise indicated. Starting with about 70 g of tissue, the procedure of Jaspers et al. (1989) was followed through the first ultracentrifugation step. Briefly, the tissue was thawed in four Vol of ice-cold 50mM Tris-HC1 (pH 7.6), 10mM EDTA, 5mM dithiothreitol (DTT), 10% glycerol, 10 mg oyster glycogen/ml, 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, 1/~g leupeptin/ml, and 1 #g pepstatin A/ml and homogenized in a Waring blender. The homogenate was centrifuged at 13,000 g for 30 min. The supernatant was filtered through glass-wool and then centrifuged at 140,000g for 2hr. The resulting pellets were resuspended at room temperature with the aid of a glass/teflon tissue homogenizer in 150ml of 50mM Tris-HC1, pH 7.6, 1 mM EDTA, 50mg sucrose/ml, 10 mM magnesium acetate, 2 mM DTT, 1 mM each of benzamidine and phenylmethylsulfonyl fluoride, 1 #g leupeptin/ml, 1/~g pepstatin A/ml, and 20/~g soyMATERIALS AND M E T H O D S bean trypsin inhibitor/ml. After adding human Glycogen synthase assay salivary ~-amylase (0.1 mg/ml) the solution was Glycogen synthase was assayed by the filter incubated at 30°C for 45min to digest the paper assay according to Thomas et al. (1968). glycogen and to allow dephosphorylation of One unit of enzyme activity corresponds to the glycogen synthase by endogenous phosphatases. incorporation of 1 #mol of glucose from ~4C- The sample then was recentrifuged at 80,000g UDPglucose into glycogen per min under con- for 90 min. The pellet was discarded and the ditions of the standard assay. The activity ratio supernatant was applied to a 50 ml is the ratio of enzyme activity measured in the DEAE-Sephacel column equilibrated in 50 mM absence and presence of 6.7 mM glucose-6-P. Tris-HC1, 1 mM EDTA, pH 7.6, 5 mM DTT, 10% glycerol, and the protease inhibitors Protein kinases described above (Buffer A). The resin was The catalytic subunit of PKA from bovine washed with 4 Vol of Buffer A and a linear heart (Reimann and Beham, 1983), phosphoryl- gradient (0-500 mM NaC1) was developed. The ase kinase (Cohen, 1973), and glycogen synthase peak activity fractions were combined. One kinase-3 (GSK-3) (Hegazy et al., 1987) were volume of Buffer B (50raM /~-glycerophospurified by published procedures. Ca2+/ phate, pH7.6, 1 mM each of MgC12, MnCI2, and
Rat heart glycogensynthase CaC12, 2raM DTT, 500mM NaC1, and the protease inhibitors of Buffer A) was added to the pooled fractions and the sample was loaded onto a 10ml concanavalin A Sepharose 4B column equilibrated in Buffer B. The flowthrough was recirculated overnight through the resin, and the column was washed with Buffer B. The enzyme was eluted with 200 mM sucrose, 5 mM E D T A in Buffer B. One mg glycogen/ml was added to the active fractions and the solution was incubated for 60 min at 4~C. Samples of 1 ml were layered onto a 24ml sucrose solution (40% w/v in 5 0 r a m Tris-HC1, pH 7.8, 5 mM EDTA, 2 mM EGTA, 1 mM DTT), and subjected to high speed centrifugation at 200,000g for 17 hr (Dickey-Dunkirk and Killilea, 1985). The precipitated enzymeglycogen complex was resuspended in 50raM Tris-HCl, pH 7.0 containing l mM DTT. At this stage, the enzyme could be stored at - 70~'C for at least 12 months without any loss of activity. For some experiments, the enzyme had to be free of glycogen and the procedure of Mellgren (1976) was used to digest the excess glycogen. Phosphorylation of glycogen synthase Unless otherwise noted, the phosphorylation reactions (100~tl) contained 0.1 mg glycogen synthase/ml, 0.05 mM ~,-[32p]ATP (1001000cpm/pmol), and 10 mM magnesium acetate. Phosphorylation by PKA or GSK-3 was carried out in the presence of 50 mM Hepes, pH 7.5, 3 0 m M NaCl, l mM EDTA, and l mM DTT. Phosphorylation by CaM-kinase II was performed in a buffer containing 50 mM Hepes, pH 7.5, l mM DTT, 0.2raM 7-[32p]ATP, 0.5raM CaC12, and l~tM calmodulin. Phosphorylation by protein kinase C was carried out in a buffer containing 27 mM Tris-HCl, pH 7.7, 1 mM EGTA, 0.16 mM calcium acetate, 64 ktM 1 2 - 0 - tetradecanoylphorbol - 13 - acetate, and 26ktg phosphatidylserine/ml. Phosphorylation by phosphorylase kinase was carried out in the presence of 25 mM fl-glycerolphosphate, 25raM Tris-HCl, pH 7.5, l mM DTT, and 1.5mM CaCI> whereas the phosphorylation buffer for CK II was 40 mM Tris-HC1, pH 7.7, 0,8 mM EDTA, 0,2 mM EGTA, 1 mM DTT, and 6% glycerol. In some phosphorylation experiments bovine serum albumin (0.05 mg/ml) was added. When kinases other than PKA were used, the heat-stable protein inhibitor of PKA was included. At various time points 5-101zl aliquots were removed for analysis of
567
[3ZP]phosphate incorporation into the glycogen synthase as described before (Roskoski, 1983). At the same time points, changes in activity ratio were determined using the same assay conditions with non-radioactive ATP instead of [7--~2p]ATP. At predetermined time points, 4/~1 aliquots were diluted about 30-fold with standard glycogen synthase dilution buffer containing 15raM EDTA and glycogen synthase activity was determined in the absence and presence of glucose-6-P under standard synthase assay conditions. H P L C peptide mapping Samples containing [32p]glycogen synthase phosphorylated with the appropriate kinase were precipitated using trichloroacetic acid, and were processed for complete trypsin digestion and HPLC analysis of the obtained phosphopeptides as described by Hegazy et al. (1987). Alternatively, the phosphorylation reaction was stopped by the addition of 0.5 Vol electrophoresis sample buffer (150 mM Tris-HCl, pH 6.8, 0.01 mM EDTA, 30% glycerol, 0.05% bromophenol blue, 3.3% SDS and 7.66% /~mercaptoethanol). After SDS-PAGE, the gel was dried under heat and vacuum onto Whatman 3 MM Chr filter paper. The band corresponding to glycogen synthase was sliced from the dried gel and incubated directly with trypsin (1 mg/ml) at 30~-C for 24hr. Similar results were obtained with either method. A 25 cm Synchropak RP, C~8 column (Synchron Inc., Linden, West LaFayette, Ind.) was used for all experiments together with the Waters Model 510 HPLC System. A gradient from 0.1% trifluoroacetic acid in water (pump A) to 0.08% trifluoroacetic acid in 60% acetonitrile (pump B) was employed. The flow rate was 1 ml/min and fractions of 1 ml were collected, The radioactivity was determined using Cerenkov counting. Other methods and materials Rat skeletal muscle glycogen synthase was purified according to Hegazy et al. (1987). The heat stable inhibitor of PKA was purified from rabbit skeletal muscle (Schlender et al., 1983). The phosphorylated amino acids of [32P]glycogen synthase were identified as described by Cooper et al. (1983). 32P-Labeled glycogen synthase was fragmented with CNBr as described by Hegazy et al. (1987). S D S - P A G E was performed on a Hoefer mini gel apparatus according to Laemmli (1970). For
568
Dimitris Grekinis et al. Table 1. Purification of glycogen synthase from rat heart
Extract 1st Ultrafiltration DEAE pool Concanavalin A pool 2nd Ultrafiltration
Volume (ml)
Protein (mg)
Activity (U)
S.A. (U/mg)
Purification (X)
Yield (%)
230 148 152 4.5 1.15
4487 1957 52.1 4.1 0.587
147 130 47.7 11.5 10.5
0.0328 0.06643 0.916 2.810 17.9
1
2.0 28 85.8 547
100 88 33 7.5 7.2
Glycogen synthase was purified from 70 g rat heart.
autoradiography of the gels, Kodak SB Panoramic dental X-ray film (DF-85) with an intensifying screen was used. Protein concentration .was determined by the method of Bradford (1976) using bovine serum albumin as the standard. Common laboratory reagents were of reagent grade and were obtained from standard suppliers. ~4C-UDPglucose and )~_32p_ 12 --
.=
--
6
-
4
-
J
0.
O 0
Relative
mobility
of a band at 77,000 (Fig. 1). The purified enzyme was virtually free of contaminating glycogen synthase kinase (Table 2) or glycogen synthase phosphatase activity (data not shown).
a ~
b
-~
GS
ofi 4.9
Purification of glycogen synthase
act. was 17.9 U/rag protein. The enzyme exhibited a major band on S D S - P A G E with an apparent molecular weight o f 87,000 and a trace
(B)
5.3 --
RESULTS
A modified procedure was developed for the purification of dephosphorylated rat heart glycogen synthase. The results of the purification are summarized in Table 1. Glycogen synthase, eluted from the concanavalin-A Sepharose column, proved to be only partially purified (spec. act. 2.8U/mg protein) and additional purification was necessary. High speed centrifugation of the glycogen synthase-glycogen complex through 40% sucrose removed all the contaminating proteins present in the previous step and more than 95% of the enzyme could be recovered. The activity ratio of the final preparation was 0.74 and spec.
(A)
10--
8
labeled ATP were obtained from ICN or NEN, CNBr was from Aldrich, acetonitrile from Baker, L-l-tosyl-amino-2-phenylchloromethyl ketone treated trypsin, and soybean trypsin inhibitor from Worthington Biochem Corp.
- -
Table 2. Phosphorylation of rat heart glycogen synthase by cAMP-dependent and cAMP-independent protein kinases
O ,d
Protein kinase
mol [32P]phosphate per mol subunit
Activity ratio
0.0 1.00 0.93 0.15 0.21 0.41 0.80
0.74 0.25 0.10 0.75 0.45 0.40 0.40
e
0 0.1
I
I
I
I
I
I
0.2
0.3
0.4
0.5
0.6
0.7
Relative mobility Fig. 1. SDS-PAGE of the purified glycogen synthase (3 vg). (A). Optical density scan of Coomassie stained SDS-polyacrylamide gel. (B) Relative mobility of the glycogen synthase and protein standards: a, myosin (200,000); b, fl-galactosidase (116,000); c, phosphorylase a (97,000); d, bovine serum albumin (66,000); e, ovalbumin (43,000); and GS, glycogen synthase.
No Kinase PKA GSK-3 CK II Protein Kinase C Phosphorylase Kinase CaM-Kinase II
Heart glycogen synthase was maximally phosphorylated by the indicated kinase and the activity ratios were determined. Conditions for phosphorylation and assay of glycogen synthase were as described in "Materials and Methods".
Rat heart glycogen synthase
(In--O)l!unqnsio,,,/ IO~d~z
:l!unqns lOW / lOtU a~ Z
(0--~)
c~ I
I
I
569
d
I
d
o
6
o
I
I
I
t
'
m
I
I l 0
o
o <-o
~,
•
o
c" o
I~
0
/\ I
0
,
', ~',
I
',
;
o. o! 0q r-: ,q ,,q. , ~
o
o
o
o
o
I
I
I
o
o
o
d
•
/
Io
I
,"! e!. -
\
•
~\
/•
,I
,
0 0
0
,
,
0
I
i,
/
I
0
.o
.~
IZ]
/
Nn
,
,,
0
0
I 0
I 0
l
~d
\
]0
I
0
0
o
~'~
e~
0
~¢ - .g. _ e. N (E]--O)
(o--n) l!unqn~ lOW/lOWclz, Z
%!unqn~;lOW / IOLU dzE
m
~
o
o
'
I
I
c~
c~
I
I
o
<
d
d
I
I
~5 I
d
d o
I
O~
m
0
e~ 0
;-I
o
•
f
0
0
I
• C) if3
\ [3.
.I
\n
[]
0
~
m~
O •
._~ ~0
~0
"~-
o~
0
d
d
d
d
d
e4
~
•
I
0
o
0
0
,c E
0
/
~~ ' ~c
~.
~n
~
,.
e4
Q
~
d
d
d
6
o
(m--m) So ;o o!wJ xl!^!~ov
e~
570
Dimitris Grckinis et al. 1
Phosphorylation o/,zlvcojZen st'nthuse
2
3
4
5
When glycogen synthase was phosphorylated by the different protein kinases, essentially all of the [~2P]phosphate was incorporated onto serine residues associated with an 87,000 Da protein band on SDS P A G E (data not shown). The incorporation of [3:P]phosphate by the different kinases and the corresponding activity ratio changes can be seen in Fig. 2. The maximum amount of ['eP]phosphate incorporated in the glycogen synthase subunit are summarized in Table 2. One tool [~2P]phosphate/mol subunit was introduced by PKA, whereas CK 2 introduced the lowest amount of [32P]phosphate. GSK-3. a major glycogen synthase kinase in skeletal muscle, incorporated a total of about one tool [~:P]phosphate/mol subunit and produced the greatest decrease in activity ratio, whereas incorporation of [~P]phosphate by CK It resulted in no detectable change in the activity ratio of the enzyme. Removal of glycogen by digestion of the puri[ied glycogen synthase with ~,-amylase prior to phosphorylation with the kinases did not change the amount of [3-~P]phosphate incorporated or the activity changes (data not shown).
Fig. 3. CNBr digestion of the rat heart glycogen synthase. The enzyme was phosphorylated with the appropriate kinase at 30 C, the reactions were stopped by adding trichloroacetic acid (10% final), and the precipitated pellets werc washed and resuspended in 70% formic acid. CNBr was added and the digestion was carried out in the dark at room temperature for 8hr. The resulting digest was lyophilized, resuspended in electrophoresis sample buffer and loaded on a 5 18% gradient SDS polyacrylamidegel. A picture of the autoradiogram of the gel is presented. Phosphorylation levels of the glycogen synthase are the same as those described on Table 2. Lane 1, PKA; lane 2, protein kinase C; lane 3, CK lI: lane 4, GSK-3; and lane 5, phosphorylase kinase.
CNBr digestion
Peptide mapping on C~,, recerse-phase H P L C
Rat heart glycogen synthase was phosphorylated by the selected protein kinases to the levels shown in Table 2. Digestion of the enzyme with CNBr followed by SDS P A G E showed that in all cases the [':P]phosphate was incorporated into one or two phosphopeptides (Fig. 3). Similar phosphopeptides, designated CB1 and CB2 were obtained from rat skeletal muscle glycogen synthase (Hegazy el al., 1987). The apparent molecular weights of CBI and CB2 were about 13,000 and 23,000, respectively. Phosphorylation of the enzyme by PKA resulted in the incorporation of [~eP]phosphate predominantly oll the CB2 fragment, but CBI also contained significant radioactivity (about 25% of the total). Phosphorylation by GSK-3 or protein kinase C resulted in incorporation of phosphate predominantly in CB2 with smaller amounts in CB1. Phosphorylation by phosphorylase kinase resulted in incorporation of [~-P]phosphate only in CBI. Casein kinase II incorporated all the ['2P]phosphate into CB2 fragment. These results indicated that the phosphorylation sites in heart glycogen synthase may be similar to skeletal muscle glycogen synthase.
The phosphorylation sites on rat heart glycogen synthase were further defined by H P L C analysis of the ~P-labeled tryptic peptides obtained after phosphorylation by the different protein kinases (Fig. 4). The chromatogram obtained from the phosphorylation of the enzyme by PKA showed that the radioactivity was mainly incorporated in three different peptides with a retention time of 15, 40, and 51 rain. The [~2P]phosphate incorporated into glycogen synthase by GSK-3 was recovered in three peptides with retention times of 31, 37, and 51 rain each containing 52, 37 and 11%, respectively, of the total radioactivity incorporated. All the radioactivity incorporated by phosphorylase kinase was recovered in only one phosphopeptide (51 rain). Protein kinase C incorporated ['-~P]phosphate into sites corresponding to peptides with retention times of 15 and 51 rain, whereas CK II introduced all the [32p]phosphate in only one peptide with a retention time of 18 rain. The tryptic phosphopeptide maps obtained with heart glycogen synthase were very similar to those obtained with rat skeletal muscle glycogen synthase (data not shown and Hegazy et al., 1987): therefore, we used the same
23 kDa
13 kDa
'O 8 ~
Rat heart glycogen synthase
571
(A)
3
r..)
(B)
3
2
1
0
(C/
(D)
~
1
0
;
3
N
2 1
0
10
20
30
40
50
60
Minutes Fig. 4. HPLC chromatograms of phosphorylated rat heart glycogen synthase by different protein kinases. Phosphorylation by PKA, GSK-3, phosphorylase kinase, protein kinase C, and CK 1I (A, B, C, D, and E, respectively). Phosphorylation levels of the glycogen synthase are the same as those described in Table 2.
nomenclature. Peptides with a retention time of 15, 18, 40, and 51 min were designated as peptides la, 5, lb and 2, respectively. The skeletal muscle glycogen synthase phosphorylation sites known collectively as sites 3 are three closely
spaced serines which are in a single tryptic peptide (Roach, 1990). This phosphopeptide, presumably because of different states of phosphorylation, eluted as two peaks (Hegazy et al., 1987); thus, the two peaks obtained from syn-
572
Dimitris Grekinis et al.
thase phosphorylated by GSK 3 which eluted at 31 and 37 min were designated as the T-3 peptide. DISCUSSION
We developed a modified procedure for the purification of rat heart glycogen synthase which combines strategies used for the purification of rat heart (Jaspers et al., 1989) and bovine heart (Dickey-Dunkirk and Killilea, 1985) glycogen synthase. This convenient procedure resulted in a practically homogeneous enzyme preparation, as judged by SDS PAGE, with an apparent molecular weight of 87,000 (Fig. 1). A minor protein band that was present at about 77,000 in some preparations most likely represented a proteolytic cleavage product of the enzyme as reported for the bovine heart enzyme (Mitchell et al., 1980; Sivaramakrishnan et al., 1982, and Dickey-Dunkirk and Killilea, 1985). The [32P]phosphate incorporated by the kinases was located in the 87,000 Da species, and virtually no [32P]phosphate was detected when no protein kinase was added to the reaction indicating that the preparation was free of contaminating glycogen synthase kinase(s). Purified rat heart glycogen synthase was phosphorylated by PKA and a number of cAMP-independent protein kinases and calcium-dependent protein kinases. The phosphorylation levels obtained with rat heart glycogen synthase are similar to those reported for the phosphorylation of the rat skeletal muscle enzyme by the same kinases (Hiken and Lawrence, 1985 and Hegazy et al., 1987). Higher phosphorylation levels have been reported for bovine heart enzyme catalyzed by PKA, as well as a mixture of non-characterized cAMPindependent protein kinases (Mitchell and Thomas, 1981, Mitchell et al., 1980 and Sivaramakrishnan et al., 1982). Except for CK II, the incorporation of [32P]phosphate into glycogen synthase was associated with inactivation of the enzyme. Although the inactivating kinases introduced similar maximum amounts of phosphate into the enzyme, different changes in activity ratios were observed suggesting that different sites were being phosphorylated (DePaoli-Roach et al., 1983). McCullough and Walsh (1979) reported that the CNBr phosphopeptide maps from bovine heart glycogen synthase are similar to those obtained from rabbit skeletal muscle, whereas others (Mitchell et al., 1980) reported that differences might exist between bovine heart and
rabbit skeletal muscle enzymes. In the present study, the SDS-PAGE phosphopeptide maps obtained from CNBr digestion of rat heart enzyme phosphorylated by the different protein kinases, showed that [32p]phosphate was located mainly on the phosphopeptides termed CB1 and CB2 (Fig. 3). Similar CNBr profiles for the incorporation of phosphate into rat skeletal muscle by PKA, GSK-3, CK II, and phosphorylase kinase have been reported (Hiken and Lawrence, 1985 and Hegazy et al., 1987). Knowledge of the distribution of the phosphate on the enzyme molecule, incorporated by a given kinase, could provide valuable information as to which kinases are involved under defined in vivo conditions, e.g. under hormonal stimulation. Clearly, the separation of the phosphopeptides by reverse-phase HPLC chromatography after complete trypsin digestion can give more complete characterization of the phosphorylation sites. Reversed phase HPLC of rat heart glycogen synthase phosphopeptides was attempted by others without a clear separation (Wolleben et al., 1987). In this study we were able to separate the phosphopeptides resuiting from the tryptic digests of rat heart glycogen synthase phosphorylated by the various kinases. Since the phosphopeptides co-eluted with phosphopeptides obtained from rat skeletal enzyme the heart phosphopeptides were also designated Tla, Tlb, T2, T3, and T5. Of particular interest was the elution position of T2, the N-terminal phosphopeptide. Because the N-terminal sequence of rabbit and rat skeletal muscle glycogen synthase are not identical (Jaspers et al., 1989) the N-terminal phosphopeptides are easily separated by C~8 reversephase chromatography (Hegazy et al., 1987). The N-terminal sequence of rat skeletal muscle glycogen synthase and rat heart glycogen synthase are identical (Jaspers et al., 1989) and as expected rat heart T2 eluted at the same position as rat skeletal muscle T2. Heart glycogen metabolism is quite different than skeletal muscle metabolism (Roach, 1986). Although the complete amino acid sequence of heart glycogen synthase has not been determined it is evident from this and previous studies that structure of thee heart enzyme and the enzymatic properties are very similar to those of the skeletal enzyme. It seems likely that differences in the synthesis of heart and skeletal glycogen are determined by the interplay of the kinases and phosphatases which could lead to a difference in the phosphorylation state and thus
Rat heart glycogen synthase t h e a c t i v i t y s t a t e o f g l y c o g e n s y n t h a s e in t h e t w o tissues. Acknowledgements--We thank Susan Wilson for her technical assistance and Martha Heck for her secretarial assistance. This work was supported by Grant HL 36576 from the National Institutes of Health and by Grant 1881165 from the Juvenile Diabetes Foundation. REFERENCES
Bradford M. M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analvt. Biochem. 72, 248 254. Cohen P. (1973) The subunit structure of rabbit-skeletalmuscle phosphorylase kinase, and the molecular basis of its activation reactions. Eur. J. biochem. 34, I 14. Cooper J. A., Sefton B. M. and Hunter T. (1983) Detection and quantification of phosphotyrosine in proteins. In Methods Enzvmology (Edited by Corbin S. and Hardman S.), Vol. 99, pp. 387-402. Academic Press, New York. DePaoli-Roach A. A., Aham Z., Camici M., Lawrence J. C. Jr and Roach P. J. (1983) Multiple phosphorylation of rabbit skeletal muscle glycogen synthase. Evidence for interactions among phosphorylation sites and the resolution of electrophoretically distinct forms of the subunit. J. biol. Chem. 258, 10702-10709. Dickey-Dunkirk S. and Killilea S. D. (1985) Purification of bovine heart glycogen synthase. Analyt. Biochem. 146, 199-205. Grekinis D., Reimann E. M. and Schlender K. K. (1992) Purification and phosphorylation of rat heart glycogen synthase. FASEB J. 6, A340. Hashimoto Y., Schworer C. M.~ Colbran R. J. and Soderling T. R. (1987) Autophosphorylation of Ca2~/ calmodulin-dependent protein kinase II. Effects on total and Ca2+-independent activities and kinetic parameters. J. biol. Chem. 262, 8051-8055. Hegazy M. G., Schlender K. K. and Reimann E. M. (1987) Differences between glycogen synthases from rat and rabbit skeletal muscle as indicated by phosphopeptide maps. Biochim. biophys. Acta 927, 269 279. Hiken J. F. and Lawrence J. C. Jr (1985) Rat skeletal muscle glycogen synthase: phosphorylation of the purified enzyme by cAMP-dependent and -independent protein kinases. Arch. biochem. Biopho's, 236, 59 71. Jaspers S. R.. Rulfs J., Johnson G. L., Mole J. E. and Miller T. B. Jr (1989) Amino-terminal sequence analysis of rat heart and muscle glycogen synthase: homology to the rabbit enzyme and the implications for hormonal control. Arch. biochem. Biophys. 268, 630-636. Laemmli U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. Litchfield D. W., Lozeman F. J., Piening C., Sommercorn J., Yakio K., Walsh K. A. and Krebs E. G. (1990) Subunit structure of casein kinase II from bovine testis. Demonstration that the alpha and alpha" subunits arc distinct polypeptides. J. biol. Chem. 265, 7638-7644. McCullough T. E. and Walsh D. A. (1979) Phosphorylation of glycogen synthase in the perfused rat heart. J. biol. Chem. 254, 7336-7344. Mellgren R. L. (1976) Control of dephosphorylation of bovine heart glycogen synthase D. Ph.D. Dissertation, Iowa State University.
573
Mitchell J. W., Mellgren R. L. and Thomas J. A. (1980) Phosphorylation of heart glycogen synthase by cAMPdependent protein kinase. Regulatory effects of ATP. J. biol. Chem. 255, 10368-10374. Mitchell J. W. and Thomas J. A. (198l) Phosphorylation of bovine heart glycogen synthase by two protein kinases. Kinetic properties of phosphorylated forms of the enzyme. J. biol. Chem. 256, 6160~169. Ramachandran C., Angelos K. L. and Walsh D. A. (1982) Cyclic AMP-dependent and cyclic AMP-independent antagonism of insulin activation of cardiac glycogen synthase. J. biol. Chem. 257, 1448-1457. Ramachandran C., Angelos K. L. and Walsh D. A. (1983) Hormonal regulation of the phosphorylation of glycogen synthase in perfused rat heart. Effects of insulin, catecholamines, and glucagon. J. biol. Chem. 258, 13377 13383. Reimann E. M. and Beham R. A. (1983) Catalytic subunit of cAMP-dependent protein kinase. In Methods Enzymology (Edited by Corbin S. and Hardman S.), Vol. 99, pp. 51 55. Academic Press, New York. Roach P. J. (1986) Hormonal Control of Glycogen Metabolism. In Regulation of Heart Function: Basic Concepts and Clinical Appfcations (Edited by Rupp H.), pp. 201 -222. Thieme, New York. Roach P. J. (1990) Control of glycogen synthase by hierarchal protein phosphorylation. FASEB J. 4, 2961-2967. Roach P. J., Cao Y., Corbett C. A., DePaoli-Roach A. A., Farkas I., Fiol C. J., Flotow H., Graves P. R., Hardy T. A., Hrubey T. W., Viskupic E. and Zhang W. (1991) Glycogen metabolism and signal transduction in mammals and yeast. Adv. Enzyme Reg. 31, 101 - 120. Roskoski R. Jr (1983) Assays of protein kinase. In Methods Enzymology (Edited by Corbin S. and Hardman S.), Vol. 99, pp. 3~6. Academic Press, New York. Schlender K. K., Tyma J. L. and Reimann E. M. (1983) Preparation of a partially purified protein kinase inhibitor. In Methods Enzymology (Edited by Corbin S. and Hardman S.), Vol. 99, pp. 77-80. Academic Press, New York. Shearman M. S., Ogita K., Kikkawa U. and Nishizuka Y. (1989) A rapid method for the resolution of protein kinase C subspecies from rat brain tissue. In Methods Enzymology (edited by Cohn P.), Vol. 168, pp. 347 35l. Academic Press, New York. Sivaramakrishnan S., High C. W. and Walsh D. A. (1982) Regulation of cardiac glycogen synthase by phosphorylation: catalysis of inactivation by cAMP-dependent and cAMP-independent protein kinases and comparison with the phosphorylation of the skeletal muscle enzyme. Arch, biochem. Biophys. 214, 311-325. Stalmans W., Bollen M. and Mvumbi L. (1987) Control of glycogen synthesis in health and disease. Diab./Metab. Rez. 3, 127-161. Thomas J. A., Schlender K. K. and Larner J. (1968) A rapid filter paper assay for UDPglucose-glycogen glucosyltransferase, including an improved biosynthesis of UDPJ4C-glucose. Analyt. Biochem. 25, 486~-99. Wolleben C. D., McPherson R. K., Rulfs J., Johnson G. L. and Miller T. B. Jr (1987) Phosphorylation of rat heart glycogen synthase: studies in cardiomyocytes and in vitro phosphorylations with cAMP-dependent kinase and protein phosphatase-l. Biochim. biophys. Acta 928, 98 106. Zhang W., Browner M. F., Fletterick R. J., DePaoli-Roach A. A. and Roach P. J. (1989) Primary structure of rabbit skeletal muscle glycogen synthase deduced from cDNA clones. FASEB J. 3, 2532-2536.