One of the other striking differences between the two enzymes is the selective ..... kinase of other proteins such as phosphorylase kinase, tro- ponin, and the ...
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1986 by The American Society of Biological Chemists, Inc.
Vol. 261, No. 26, Issue of September 15, pp. 1216612171,1986 Printed in U.S.A.
Differential andCommon Recognition of the Catalytic Sitesof the cGMP-dependent and CAMP-dependent Protein Kinases by Inhibitory Peptides Derived from theHeat-stable Inhibitor Protein* (Received for publication, April 17, 1986)
David B. Glass$, Heung-Chin ChengQ,Bruce E.Kempll, and Dona1A. Walshe From the $Department ofPharmacology, Emory University School of Medicine, Atlanta, Georgia 30322, the $Departmentof
Biological Chemistry, School of Medicine, University of Californiu, Davis, California 95616, and the llDepartment of Medicine, University of Melbourne, Repatriation General Hospital, Heidelberg, Victoria 3081, Australia
Synthetic peptides corresponding to the active domain ,of the heat-stable inhibitor protein of CAMPdependent protein kinase (Cheng, H A . , Kemp, B. E., Pearson, R. B., Smith, A. J., Misconi, L., Van Patten, S . M., and Walsh, D. A. (1986) J. Biol. Chem. 261, 989-992) were tested as inhibitors of cGMP-dependent protein kinase. The peptides themselves were not substrates. cGMP-dependent protein kinase activity was assayed using histone H2B and two syntheticpeptide substrates. Consistent with previous observations of other peptide inhibitorsof this enzyme (Glass, D. B. (1983)Biochem. J. 213,159-164), the inhibitorypeptides had no effect on the phosphorylation of histone H2B, but they competitively inhibited cGMP-dependent phosphorylation of the twopeptide substrates. The parent inhibitor peptide, PKI(5-24)amide, and a series of analogs hadK i (or ICa,) values for cGMP-dependent protein kinase in therange of 15-190 PM. In contrast to their effects on the CAMP-dependentprotein kinase, the inhibitory peptides were substantially less potent with cGMP-dependent protein kinase, and potency was reduced by the presence of the NH,-terminal residues (residues 5-13). We conclude that the two protein kinases share a recognition of the basic amino acid cluster within the pseudosubstrate region of the peptide, but that thecGMP-dependent protein kinase does not recognize additional NHz-terminaf determinants that make the inhibitor protein extremely potent toward the CAMP-dependent enzyme. Even- when tested at high concentrations and with peptide substrates, the native inhibitor protein didnot inhibit cGMP-dependent protein kinase underassay conditions in which the peptides derived from it were inhibitory. Thus, the native inhibitor protein appears to havestructural features which block interaction with thecGMP-dependent enzyme and enhance its selectivity for CAMP-dependent protein kinase.
CAMP (5), while activation of the cGMP-dependent protein kinase by cGMP does not involve subunit dissociation (6). One of the other strikingdifferences between the two enzymes is the selective inhibition of the CAMP-dependent protein kinase by the heat-stable inhibitor proteinof that enzyme (7). Several reports (8-13) have indicated that this inhibitor protein does not affect the cGMP-dependent enzyme, at least under the assay conditions used and with the concentration of native inhibitor protein available. The heat-stable inhibitor protein of the CAMP-dependent protein kinase has been purified to homogeneity (14, 15), and the entire sequence of one of the several charge and size isoforms (15-18) has been reported (19). The inhibitor acts by inhibiting the free catalytic subunit of the CAMP-dependent protein kinase (20) in a competitive manner with respect to phosphoryl-accepting substrate (14, 15, 21). Potency of inhibition is extremely high (Ki 0.1 nM) and is independent of the particular protein or peptide used as substrate (20). The catalytic subunit-inhibitor protein interaction is facilitated by MgATP2- (21,22). Recently, an active portion of the inhibitor protein has been defined by obtaining active inhibitory oligopeptides through limited proteolytic digestion with either Staphylococcus aureus Vs protease (23) or mast cell protease I1 (24). The former peptide, termed PKI(5-24),’ is the most potent, having a Kivalue of 0.3 nM2 and an amino acid sequence of Thr-Thr-Tyr-Ala-Asp-Phe-Ile-Ala-Ser-GlyArg-Thr-Gly-Arg-Arg-Asn-Ala-Ile-His-Asp, As one determinant of specificity for interaction with the CAMP-dependent protein kinase, this peptide contains a cluster of arginine residues that, when deleted (24), or when any single arginine residue is substituted by lysine (25), causes a marked loss of inhibitory potency. Studies with synthetic analogs of PKI(524) have also shown that a critical role is played by one or several of the first7 NHz-terminal residues in the PKI(5-24) sequence to dictate the high potency of inhibitor proteinprotein kinase catalytic subunit interaction (25-27). In theinitial studies (24,25), the inhibitory peptides derived
The cGMP-dependent and CAMP-dependent protein kinases are homologous proteins (1)which have similar but not identical substrate specificities (2-4). The subunit structures and mechanisms of activation are different, however, with the CAMP-dependent enzyme being dissociated into subunits by
The nomenclature of the 20-amino-acid peptide derived from the Inhibitor Protein by S. aureus Vs protease digestion has been changed to PKI(5-24) reflecting the now established amino acid sequence of the native inhibitor protein (19). This is the same peptide previously termed by us (25) as “IPZO.” Synthetically, this peptide is prepared as the COOH-terminal amide and is thus termed PKI(5-24)amide. The sequences of this and related peptides are given in Table I. Other abbreviations used are: (AlaU)H2B(29-35), Arg-Lys-Arg-Ser-ArgKempAla-Glu;(Ala3’)H2B(29-35), Arg-Lys-Arg-Ala-Arg-Lys-Glu; tide, Leu-Arg-Arg-Ala-Ser-Leu-Gly; (Ala)Kemptide, Leu-Arg-ArgAla-Ala-Leu-Gly;ICw, concentration producing 50% inhibition. PKI(5-24) and PKI(5-24)amide exhibit Ki values of 0.3 nM (23) and 2.3nM (25), respectively. The reasons for this difference are under current investigation.
* This work was supported by National Institutes of Health Grants GM28144 (to D. B. G.) and AM21019 (to D. A. W) and by the National Health andMedical Council of Australia (to B. E. K.). The costs of publication of this article were defrayed in part by the payment of page charges. This articlemust therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
-
12166
’
Recognition of Inhibitor Peptides
by cGMP-dependent Kinase
12167
from the inhibitor protein appeared to be selective for the CAMP-dependent protein kinase and under the conditions tested had no effect on a number of other protein kinases, including the cGMP-dependent enzyme. Since the inhibitor protein is competitive with respect to protein/peptide substrate, and the CAMP-dependent and cGMP-dependent protein kinases have similar substrate specificities, it has never been clear why the inhibitor protein appears to be absolutely selective for the CAMP-dependent enzyme. Wenow report more detailed studies on the effect of PKI(5-24)amide and other synthetic peptide analogs on the activity of the cGMPdependentprotein kinase. In the 10-300 pM range, these peptidesare able to inhibit the cGMP-dependentprotein kinase, but the interaction is substrate-dependent. The peptides, however, are clearly recognized differentially by the cGMP-dependent and CAMP-dependent protein kinases, and PKI(5-24)amide is several orders of magnitude more potent in inhibiting the CAMP-dependent enzyme. The study shows that the two enzymes do share some of the same recognition sites, but also provides insights into the structural features that contribute to the inhibitor protein's absolute specificity for the CAMP-dependent protein kinase.
used at a final concentration of 50 p ~ and , the cGMP-dependent kinase was used at a concentration of20 pg/ml. Most assays were initiated by the addition of enzyme. In some experiments, as noted, cGMP-dependent protein kinase was preincubated with buffer, cGMP, MgZ+, [-p3'P]ATP, and inhibitor peptide for 2 min at 30 "C, after which the reaction was initiated by the addition of protein or peptide substrate. Reactions were terminatedand 32P-phosphopeptide or -histone was quantitated in 50-p1 aliquots of reaction mixture by the phosphocellulose paper method (35) using a phosphoric acidwashing procedure (36). The highest amounts of peptide substrates and inhibitors used were well over an order of magnitude lower than the capacity of phosphocellulose paper squares for binding peptide (35). For determination of K, values and type of inhibition of selected inhibitor peptides, 8-100 pM (Ala')H2B(29-35) was used as variable substrate with [-p3'P]ATP fixed at a concentration of 200 pM. Alternatively, 10-100 p~ [y3'P]ATP was used as variable substrate at a fixed concentration of 30 pM (Ala')H2B(29-35). In these assays, less than 16% of the least abundant substrate was converted to product. Data from initial enzyme velocity measurements were fitted to the Michaelis-Menten equation by the method of weighted least squares (37) and plotted in double reciprocal form. The Ki values were determined from secondary replots. Material~-[-y-~~P]ATP was synthesized as described by Glynn and Chappel (38) as modified by Reimann et al. (39). P81 phosphocellulose paper was purchased from Whatman.
EXPERIMENTAL PROCEDURES
RESULTS
Synthetic Peptides-The 20-amino-acid peptide from the inhibitor protein of the CAMP-dependent protein kinase and ita analogs were synthesized as COOH-terminal amides by solid-phase synthesis and purified as described by Cheng et al. (25). The purity of these peptides has been previously confvmed by high performance liquid chromatography, amino acid analysis, and sequencing (25,26). The sequences of the parent peptide, PKI(5-24)amide, and the analogs used in this study are given in Table I. (Ala')H2B(29-35), a model substrate of cGMP-dependent protein kinase based upon the phosphorylation site sequence in histone H2B, was synthesized as originally described (28) with the exception that N"-t-amyloxycarbonyl-L-arginine(P-tosyl) was used, and the complete peptide was cleaved from the support with anhydrous HFanisole (9:l) (29). (A1a3')H2B(29-35), an analog of the substrate peptide and an inhibitor of the cGMP-dependent protein kinase, was synthesized as described previously (9). Kemptide, a synthetic peptide based upon the CAMP-dependent protein kinase phosphorylation site of porcine pyruvate kinase (30), was purchased from Peninsula Laboratories (Belmont, CA). Allpeptide concentrations were determined by amino acid analysis. Enzyme, Inhibitor Protein,and Histone Purification-Cyclic GMPdependent protein kinase was purified to homogeneity from bovine lung as described by Glass and Krebs (3). The enzyme had a specific activity of3.5 pmol/min/mg when assayed with histoneH2B as substrate ata concentration of 36 p~ under the conditions described previously (3). Cyclic GMP-dependent protein kinase was quantitated by the protein assay of Lowry et al. (31) using bovine serum albumin as standard. Inhibitor Protein was purified by our recently described modifications (23). Purified histone H2B was prepared from calf thymus by the methods of Johns (32) and Oliver et al. (33). The concentrations of the stock solutions of both were determined by amino acid analysis. Phosphotransferase Assays-Cyclic GMP-dependentprotein kinase activity was measured essentially as described previously (28, 34). Assays were conducted for 2 min at 30 "C in a reaction mixture of final volume 0.08 ml containing 30 mM Tris-HC1 (pH 7.4), 2 mM (215-300 cpm/ magnesium acetate, 1 pM cGMP, 0.2mM [T-~~PP]ATP pmol), substrate protein or peptide as indicated below, the indicated concentrations of the various inhibitor peptides, 3 mM 2-mercaptoethanol, 0.3 mg/ml bovine serum albumin, and 0.2-0.4 pg/ml (1.32.6 nM) cGMP-dependent proteinkinase. Either 1.5 p~ histone H2B, 20 pM (Alau)H2B(29-35), or 150 p~ Kemptide was used as substrate. These concentrations are slightly below the K,,, values of cGMPdependent protein kinasefor the respective substrates (3, 28). Under these assay conditions, the conversion of histone H2B, (Ala')H2B(29-35), and Kemptide to product were approximately 55%, 18%, and 3%. respectively. Utilization of [y-32P]ATPwas approximately 2% or less in each case. When analogs of PKI(5-24)amide were tested as possible substrates of cGMP-dependent protein kinase, they were
Inhibition of cGMP-dependent Protein Kinase by PKI(524)amide and Analogs-In previous studies of the effects of derivatives of PKI(5-24) on the CAMP-dependent protein kinase (25,26), ithas been shown that thedeletion of the two carboxyl-terminal amino acids (i.e. as with peptide PKI(522)amide, see Table I) made essentially no difference on inhibitory activity, whereas either deletion of NH2-terminal residues (i.e. peptides PKI(7-22)amide, PKI(10-24)amide, and PKI( 14-24)amide), or substitution of arginine residues by lysine (peptides (Lys")PKI( 14-24)amide, (Lys")PK1(1424)amide, and (Lys1')PKI(14-24)amide) diminished inhibitory potency. Each of these peptides has now been tested as potential inhibitors of the cGMP-dependent protein kinase. Fig. 1 presents the data showing the titration of cGMPdependent protein kinase activity with increasing concentrations of inhibitory peptides; Table I provides a comparison between IC, (and Ki) values obtained with the cGMP-dependent enzyme, and Ki values for inhibition of the CAMPdependent protein kinase. For theseassays of cGMP-dependent protein kinase, the peptide (Ala3*)H2B(29-35) was used as substrate at near its K,,, concentration. The sequence of this substrate peptide is based upon the site in histone H2B phosphorylated by the cGMP-dependent protein kinase. As detailed previously, this peptide is an effective substrate for this enzyme (28, 40) and is also one with which other inhibitory peptides have been examined (9). For reference, the peptides derived from PKI(5-24) were compared to two other inhibitory peptides that have been previously described, (Ala3')H2B(29-35), in which the cGMP-dependent phosphorylation site serine residue has been replaced by alanine (9, 40), and (Ala)Kemptide, the inhibitory peptide whose sequence is based upon the CAMP-dependent phosphorylation site in porcine hepatic pyruvate kinase (41,42). Several observations are apparent from the data presented in Fig. 1 andTable I. PKI(5-24)amide and four peptides testedwith deletions at theNH2- orcarboxyl-terminals (Table I, peptides 1-5)were effective inhibitors of the cGMP-dependent proteinkinase. Inhibition was concentration-dependent, with high concentrations inhibitingby at least 80%. The IC60 values of peptides 1-5 were in the range of 30-120 p ~ Theseconcentrationsaresubstantially higher thanthe amounts needed to inhibit the CAMP-dependent protein ki-
.
Recognition of Inhibitor Peptides by cGMP-dependent Kinase
12168
TABLEI Comparison of the potency of inhibitory peptides toward the cGMP-dependent protein kinase and the catalytic subunit of the CAMP-dependent protein kinase For assayswith the cGMP-dependent protein kinase,the IC, values were obtained fromthe experiment of Fig. 1 using(Alau)H2B(29-35) as the peptide substrate. Ki values for thecGMP-dependentproteinkinasewere obtainedwith(A1a')H2B(29-35)asvariable substrate either from the experimentpresented in Fig. 2 or as referenced. The Ki values for the catalytic subunit of the CAMP-dependent protein kinase were obtained with Kemptide as the variable substrate. Amino acid residuesof peptides 1-8 are numbered accordingto the sequence of the inhibitor Drotein (19). Peptide
enclature
Number
PKI(14-24)amide PKI(10-24)amide
1
2 3 4 5 6
GRTGRRNAIHD FIASGRTGRRNAIHD YADFIASGRTGRRNAI TTYADFIASGRTGRRNAI TTYADFIASGRTGRRNAIHD GRTGWAIHD GFGRRNAIHD GRTGENAIHD RKRARKE LRRAALG
PKI(7-22)amide PKI(5-22)amide PKI(5-24)amide (Ly~'~)PKI(l4-24)amide (Lys'')PKI( 14-24)amide (Lys1')PKI(14-24)amide (Ala3*)H2B(29-35) (A1a)Kemptide
7 8 9 10
cGMP-dependent protein kinase IC, Ki
CAMP-dependent protein kinase
PM
PM
30 54 62 68 111 42 62 191 52
15 31
86" 800'
Ki
0.057' 0.073" 0.027' 0.003' 0.002. 4.2. 0.37" 36" 550' 376'
Data from Cheng et al. (25). Data from Cheng et al. (26). Data from Glass (9).
I
I
I
I
I
I
I
a $3
m
.-C Y
60
80
--E
1 0 0 4 ,
0
1
I
I
I
1
I
3
10
30
100
300
1000
[Inhibitor Peptide]
I
(pM)
FIG. 1. Titration of cGMP-dependent protein kinase activity by inhibitory peptides. The activityof cGMP-dependent protein kinase was assayed with 20 p~ (Alau)H2B(29-35) as substrate as describedunder"ExperimentalProcedures."In the absence of inhibitor peptide, the velocityof the enzyme was 5.42 pmollminlmg. Inhibitory peptides were PKI(5-24)amide(O),PKI(5-22)amide (0), PKI(7-22)amide (m), PKI(10-24)amide (O), PKI(14-24)amide (A), (Lys16)PKI(14-24)amide(A), (Lys1')PKI(14-24)amide (V),(Lysl9)PKI(14-24)amide (V),and (A1a3')H2B(29-35) (X). nase (Table I) but, nevertheless, these five peptides are quite potent inhibitors of the cGMP-dependent enzyme. The most active of these peptides with the cGMP-dependent protein kinase, PK1(14-24)amide, was slightly more potentthan (Ala32)H2B(29-35),which has been the most potent inhibitory peptide for the cGMP-dependent enzyme so far reported (9, 40). The deletion of the first two, five, or nine NH2-terminal amino acids, which markedly diminishes inhibitory potency with the CAMP-dependent protein kinase, did not decrease inhibitory activity with the cGMP-dependent enzyme. Rather, as the peptide length was shortened, there was a slight increase in inhibitory potency; the shortest peptide, PKI(14-24)amide, was 3-fold more active than PKI(5-
24)amide. A similar pattern of results was also observed for the peptides in which the arginine residues were replaced by an alternate basic residue lysine (Table I, peptides 6-8). These substitutions markedly reduced inhibitory potency with the CAMP-dependent protein kinase (between 8-and 800-fold), but caused only a minimal (peptides 6 and7) or small (peptide 8) change with the cGMP-dependent protein kinase. With both enzymes, substitution ofArg'" caused the biggest decrease in inhibitory activity of PKI(14-24)amide. These peptides were substantially more potent inhibitors of cGMPdependent protein kinase, however, than (A1a)Kemptide. Overall, in comparison of all the peptides tested that were based upon the native inhibitor proteinsequence, but differing in length or type of basic amino acid, while they were good inhibitors for the cGMP-dependent protein kinase, all had similar potencies. For the cGMP-dependent enzyme, the IC6o values only covered a 6.4-fold range from 30 p~ for PKI(14in contrast, 24)amide to 191 p~ for (Ly~'~)PKI(14-24)amide; inhibition of the CAMP-dependent protein kinase was much more dependent on the presence of the full native sequence and, for the same group of peptides, Ki values ranged over 4 orders of magnitude from 2 nM for PKI(5-24)amide to 36 p~ for (Lysl8)PKI(l4-24)amide. None of the peptides based upon the inhibitor proteinsequence was more effective in inhibiting the cGMP-dependent proteinkinase than theCAMP-dependent enzyme. For the cGMP-dependent enzyme,(Ala3')H2B(29-35) remains the most selective inhibitory peptide, being -7-fold more potent in inhibiting it than in blocking the CAMP-dependent enzyme. The kinetic type of inhibition of the cGMP-dependent protein kinase was determined for two representative inhibitor protein peptides. These data are shown in Fig. 2, with respect to protein substrate, and in Fig. 3, with respect to nucleotide substrate. As illustrated, both PKI(14-24)amide and PKI(5-22)amide were competitive inhibitors uersus peptide substrate and exhibited mixed-type noncompetitive inhibition u e r s w MgATP2-. These kinetics are essentially the same as have been previously reported for inhibition by (Ala32)H2B(29-35)(9,40), and they indicate that these inhibitory peptides act by interacting at the peptidelprotein sub-
12169
Recognition of Inhibitor Peptidesby cGMP-dependent Kinase A
c
1.5
1CI
E
I
I n h i b i t o rP e n t i d e /
eo
0
( 2 0 180
l n h l b l t oP r eptlde (JIM)
I
",o , A
/. 5
0
, 0 2. 5 0 5, 0 7, 5 1 0, 0 125
0
-1
-1
[Arg-Lys-Arg-Ser-Arg-Ala-Glu]
(pM-l)
a a
-
2.0
' 0
IJY
1.5 I n h i b i t o rP e p t i d e 0
1 .o (JIM/
0
,025
,050
0
p/d , 0 7,51 0.01 2 5 -1
FIG. 3. Inhibition of cGMP-dependent protein kinase activ-
ity by inhibitorpeptides. A, mixed-type noncompetitive inhibition [Arg-Lys-Arg-Ser-Arg-Ala-Glu] (pM-l) versus MgATP2-by inhibitor peptide PKI(14-24)amide. Inhibitor FIG. 2. Inhibition of cGMP-dependent protein kinase activ- peptide concentrations ( p ~were ) 0 (O), 47 (O),94 (O), and 188 (m). ity by inhibitor peptides.A , competitive inhibition versus peptide Protein kinaseactivity wasassayedwith [T-~~PIATP asvariable
substrate by inhibitor peptide PKI(14-24)amide. Inhibitor peptide concentrations ( p ~were ) 0 (O), 40 (e),80 (o),and 160 (m). Protein kinase activity was assayed with (Alaa)H2B(29-35) as variable substrate as described under "Experimental Procedures." Fromthe secondary replot of the slopes of the lines as a function of inhibitor peptide concentration (inset),the K, value was determined to be 15 p ~ The . apparent K,,, and V,. values of the enzyme for (Ala3')H2B(29-35) in the absence of inhibitor peptide were 25.8 p~ and 13.9 pmol/min/mg, respectively.B, competitive inhibition versus peptide substrate by inhibitor peptidePKI(5-22)amide. Inhibitor peptide concentrations ( p ~were ) 0 (O), 70 (O),140 (O), and 280 (m). Protein kinase activity was assayed with (Alaa)H2B(29-35) as variable substrate. From the secondary replot of the slopes of the lines as a function of inhibitor peptide concentration (inset),the Ki value was determined to be 31 p ~ The . apparent K,,, and V,, values of the enzyme for (Alaa)H2B(29-35) in the absence of inhibitor peptide were 26.0 pM and 13.1 pmol/min/mg, respectively.
substrate as described under "Experimental Procedures." B, mixedtype noncompetitive inhibitionversus MgATPz- by inhibitor peptide PKI(5-22)amide. Inhibitor peptide concentrations ( p ~ were ) 0 (0), 53 (O),106 (O), and 212 (m). Protein kinase activitywas assayed with [y3'P]ATP asvariable substrate. For both A and B, (Ala3')H2B(2935) was at a concentration of 30 p ~ The . insets in both show the secondary replots versus l/V- from which were calculated the K,, (uncompetitive)inhibition constants in accord with Cleland (46). The K k (competitive)inhibition constants were determined similarly from replots versus slope (46). From these, the calculated values were: K, values,PKI(14-24)amide = 35 p ~ PKI(5-22)amide , = 62 p ~ K,; values, PKI(14-24)amide = 260 pM, PKI(5-22)amide = 232 p ~ .
that the interaction of the heat-stable inhibitor protein with the catalytic subunit of CAMP-dependent protein kinase is enhanced by preincubation with ATP (21, 22). Because of this, we examined whether the orderof addition of reagents might affectthe potency of inhibition of the cGMP-dependent strate binding domain in the catalytic site of the cGMPprotein kinase by the inhibitory peptides. The enzyme was dependent protein kinase. Consistent with the titration experiments and calculatedIC60 values (Fig. 1 and Table I), the preincubated at 30 "C with PKI(7-22)amide, MgATP2-, and Ki values determined fromFig. 2, and the K;i values fromFig. allthestandardreactioncomponentsexceptphosphoryl3, show PKI(14-24)amide to bea slightly more potent inhib- accepting substrate, and the reaction was then initiated by itor of the cGMP-dependent protein kinase than was the the addition of peptide substrate. Under these conditions, exactly the same inhibitory potency was obtained as when the longer chain peptide, PKI(5-22)amide. The kinetic mechanisms of the cyclic nucleotide-dependent reaction was initiated by the addition of cGMP-dependent protein kinases are probably ordered Bi-Bi reactions which in protein kinase (data not shown). In addition to these experiMgATP2- binds first (9, 40, 42, 43). It has also been shown ments, PKI(5-22)amide,PKI(10-24)amide, andPKI(l4-
9
Recognition of Inhibitor Peptides
12170
by cGMP-dependent Kinase
24)amide were directly tested as possible substrates of the cGMP-dependent protein kinase. Reaction conditions were as described under “Experimental Procedures.” None of the inhibitory peptides was phosphorylated after 120 min of incubation at substrate concentrations of50 p~ and a final concentration of cGMP-dependent protein kinase that was 50-fold greater than that used in the inhibition experiments. Under these same conditions, (Ala34)H2B(29-35)and Kemptide were stoichiometrically phosphorylated in less than 15 min (data not shown). Substrate-dependent Inhibition of the cGMP-dependent Protein Kinase-With previous studies, it has been shown that inhibition of the cGMP-dependent protein kinase by inhibitory peptides binding at the catalytic sitewas apparently peptide/protein substrate-dependent (9). Thus, (Ala3’)H2B(29-35) and (A1a)Kemptidewere found to be effective inhibitors when a range of peptides were used as the phosphorylacceptor substrate, but notwhen one of several histones were substrate. Similar differences were not observed for inhibition of the CAMP-dependent protein kinase by either of these two peptides (9). Because of these observations, the protein/peptide substrate dependency of inhibition of cGMP-dependent protein kinase by the PKI(5-24) derivatives was tested. For this study (Table11),three different substrates of the enzyme were used at orneartheir K, concentrations. Of these, (A1a3‘)H2B(29-35), the substrate used in the experiments of Figs. 1-3, and histone H2B are kinetically excellent substrates for the cGMP-dependent protein kinase ( K , values = 28 pM and 1.5 p ~ respectively; , Refs. 28 and 3). Kemptide, which has been used extensively as a model peptide substrate for the CAMP-dependent protein kinase, is readily phosphorylated by the cGMP-dependent enzyme, although its K,,, value is much higher for the latter thanfor the former enzyme (K, values = 231 ptM and 5 pM, respectively; Refs. 40 and 42). In Table I1 are presented the results of testing the inhibition of cGMP-dependent protein kinase by (Ala3’)H2B(29-35) and four of the inhibitor proteinpeptides using these three different protein/peptidesubstrates. As shown previously (9), (Ala3’)H2B(29-35) is an effective inhibitor when either (Ala34)H2B(29-35)or Kemptide was used as substrate, but
did not block the phosphorylation of histone H2B. The same result was observed for the four inhibitor protein peptides tested. Each of these peptides inhibited the phosphorylation of (Ala34)H2B(29-35)and Kemptide with similar potencies but none, over the concentrations tested, caused significant inhibition of histone H2B phosphorylation. The same results were also obtained independent of the order of addition of reaction components (not shown). The reasons for this apparent substrate-dependent action of the inhibitor peptides are not yet understood. It is nota consequence of the differences in K,,, values for the peptides and histone, because each was examined at or near its K,,, concentration. For (A1a3’)H2B(29-35),the inability to inhibit substratephosphorylation appears unique to histones since, whereas it does not inhibit the phosphorylation of various histones, it does inhibit cGMP-dependent protein kinase autophosphorylation (40) and the phosphorylation by the cGMP-dependent protein kinase of other proteins such as phosphorylase kinase, troponin,and the Type I regulatory subunit of the CAMPdependent protein kinase? Possibly, these effects are related to the“poly(L-arginine)binding site” of the cGMP-dependent protein kinase that has been reported by Walton and Gill (44, 45); this anionic site binds histones, but is distinct from the catalytic site and may be in the enzyme’s regulatory domain. The use of histones as substrate for the cGMP-dependent protein kinase, however, explains why, when previously reported (23,24),no inhibition of the cGMP-dependent protein kinase by the peptides derived from the inhibitor proteinwas observed. Previously, it has been stated (8-13) that the inhibitor protein of the CAMP-dependent protein kinase does not inhibit the cGMP-dependent proteinkinase and thisconclusion has come from experiments that have used both histone and the peptides, (Ala34)H2B(29-35)and Kemptide, as substrates (9). Since high concentrations of PKI(5-24)amide peptides inhibited the cGMP-dependent protein kinase, we have reevaluated whether very high concentrations of inhibitor protein might be inhibitory. The data on this are presented in Fig. 4; only one peptide, (Ala3”)H2B(29-35),was tested as substrate because of a shortage of available inhibitor protein at the concentrations needed. Even at 100 p~ (i.e. -lo5 x Ki TABLEI1 for the CAMP-dependent protein kinase), the inhibitor proEffect of inhibitor peptides on the phosphorylationof three substrates tein did not inhibit the cGMP-dependent proteinkinase; this by cGMP-dependent protein kinase is a concentration where the peptides derived from it were Cyclic GMP-dependentprotein kinase was assayed in the absence markedly inhibitory. or presence of the indicated inhibitor peptide with either (Ala3)H2B(29-35), Kemptide, or histone H2B as substrates as deDISCUSSION scribed under “Experimental Procedures.” The enzyme activities in the absence of inhibitor peptide (100% of control) were 5.53, 5.08, The data presented in this report provide a comparison and 1.22 pmol/min/mg, respectively, foreach of the substrates. between the protein/peptide binding domains in the catalytic Protein kinase activity sites of the cGMP-dependent and CAMP-dependent protein ConcenInhibitor kinases. As we have recently reported (25,26), thePKI(5-24) peptide tration (Alaa)H2B(29-35) Kemptide peptide derived from the inhibitor protein, and presumably the inhibitor proteinitself, contain at least two critical regions W % of control of amino acids that are essential for the high affinity interPKI(14-24)amide 1 91.5 112.4 98.4 10 70.2 86.4 96.7 action at the catalytic site of the CAMP-dependent protein 100 22.2 23.8 96.7 kinase. These two domains are the arginine cluster of Arg“PKI(7-22)amide 1 91.3 100.0 99.2 Arglg,and some of the first 7 NHz-terminal residues (residues 10 85.4 87.0 93.4 5-11). For the CAMP-dependent protein kinase, deletion or 100 38.7 39.4 104.1 substitution of residues in either of these two domains markPKI(5-22)amide 1 91.7 99.2 108.7 edly modifies inhibitory peptide binding. In contrast to this, 10 82.8 91.5 100.0 100 41.4100.0 44.3 it is clear from the data provided in this report that the 94.0 PKI(5-24)amide 3.5 cGMP-dependent protein kinase contains in common with 35 72.5 77.0 96.7 the CAMP-dependent protein kinase the recognition of the 350 21.9 22.2 70.2 arginine cluster, but does not recognize the NH,-terminal (Ala3*)H2B(29-35) 3 94.2 98.2 95.3 30
300
65.5 19.3
60.1
17.5
95.3 94.5
D. B. Glass, unpublished observations.
Recognition of Inhibitor Peptides I
I
I
I
1
I
by cGMP-dependent Kinase j
12171
derived from it (Fig. 4), and may not bind at all. It would appear that the inhibitor protein has evolved not only specific recognition sequences for the substrate binding site of the CAMP-dependentprotein kinase, but also additional features that have diminished its interaction with other protein kinases such as the cGMP-dependent enzyme. REFERENCES 1. Takio, K.,Wade,R. D., Smith, S. B., Krebs, E. G., Walsh, K. A., and Titani, K. (1984) Biochemistry 23,4207-4218 2. Lincoln, T. M., and Corbin, J. D. (1977) Proc. Natl. A c d . Sei. U. S. A. 7 4 ,
""-
29IIQ-29AR
3. Glass, D. B., and Krebs, E. G. (1979) J.Bwl. Chem. 254,9728-9738 4. Glass, D. B., and Krebs, E. G. (1980) Annu. Rev. Pharmacol. Toxieol. 2 0 ,
"- "-
362-3M
5. Fkimann, E. M., Brostrom, C. O., Corbin, J. D., King, C. A., and Krebs, E. G. (1971) Biochem. Bwphys. Res. Commun. 4 2 , 187-194 6. Gill, G. N., and McCune, R. W. (1979) Curr. Top. Cell. Re&. IS, 1-45 7. Walsh, D.A., Ashby, C. D., Gonzalez, C., Calkins, D., Fischer, E. H., and Krebs, E. G. (1971) J. Biol. Chem. 246,1977-1985 so 8. Demaille, J. G., Peters, K. A., Strandjord, T. P., and Fischer, E. H. (1978) FEES Lett. 86.113-116 9. Glass, D. B. (1983) Biochem. J. 213,159-164 100 10. Gill, G. N., Holdy, K. E., Walton, G. M., and Kanstein, C. B. (1976) Proc. Natl. Acad. Sei. U. S. A. 73,3918-3922 0 20 40 60 80 100 11. Khoo, J. C., Sperry, P. J., Gill, G. N., and Steinberg, D. (1977) Proc. Natl. A c d . Sei. U. S. A. 74.4843-4847 [ I n h i b i t oPr r o t e i n ] or [PK1(14-24)amide] (pM) 12. Takai Y., Nakaya, S., Inoue M. Kishimoto, A., Nishiyama, K., Yamamura, H., b d Nishizuka, Y. (1i76) Chem. 251,1481-1487 FIG.4. Comparison of the effects of inhibitor protein and 13. Inoue, M., Kishimoto, A., Takar, Y., and Nrshrzuka, Y. (1976) J. Biol. Chem. 251,4476-4478 PKI(14-24)amide on activity of the cGMP-dependent protein J. G., Peters, K. A., and Fischer, E. H. (1977) Biochemistry 16, kinase. The reaction was performed as described under"Experimen- 14. Demaille, 3080-3086 tal Procedures" using substrate concentrations of 50 phf ATP and 30 15. McPherion, J. M., Whitehouse, S., and Walsh, D. A. (1979) Biochemistry 18,4835-4845. p~ (Alaa)H2B(29-35) and the indicated concentrations of inhibitor C., Demadle, J. G., and Fischer, E. H. (1979) Biochimie (Paris)61, protein (0)or PKI(14-24)amide (0)in a 20-pl reaction volume. Data 16. Ferraz, fb454.51 ..~ ." are the average of duplicate determinations. 17. Whitehouse, S., McPherson, J. M., and Walsh, D. A. (1980)Arch. Biochem. Bwphys. 2 0 3 , 734-743 18. Whitehouse, S., and Walsh, D. A. (1982) J. BWL Chem. 257,6028-6032 domain amino acid sequence.It is well documented that basic 19. Scott J. D., Fischer, E. H. Takio, K., Demaille J. G., and Krebs, E. G. (19'85) Proc. Natl. A d . Aei. U. S. A. 82,573215736 amino acid residues serve as determinants of peptide substrate 20. Ashby C. D., and Walsh, D.A. (1972) J. Biol. Chem. 247,6637-6642 21. Whitehouse, S., and Walsh, D. A. (1983) J. Biol. Chen. 258,3682-3692 specificity for the cGMP-dependent protein kinase (2-4, 28, 22. Van Patten, S. M., Fletcher, W. H., and Walsh, D. A. (1986) J. Bzol. Chem. 34). Thus, an ionic interaction with the guanidine groups of 261,5514-5523 the arginine cluster of the inhibitory peptides is probably 23. Cheng, H.-C. Van Patten, S. M., Smith, A. J., and Walsh, D. A. (1985) Ewehem. 231,655-661 occurring in a similar pseudosubstrate manner. This recog- 24. Scott, J. D. Fischer, E. H. Demaille, J. G., and Krebs, E. G. (1985) Proc. ~ a t lA&. . Sei. S. A. '~2,4379-4383 nition of the arginine cluster domain is not unexpected since 25. Cheng H.-C. Kemp B. E Pearson R. B. Smith A. J Misconi L Van it reflects the similarity (but notidentity) of protein substrate Pattkn, S. M., and Walsi~,D. A. (i986) d. Biol. khem."261,98&$2 H.-C., Kemp, B. E., Smith, A. J., Pearson, R. B., Van Patten, S. specificity of the cGMP-dependent and CAMP-dependent 26. Cheng, M., Misconi, L., and Walsh, D. A. (1986) Symposium ofAmerican Protein protein kinases (2-4) and is presumably based on the high Chemists (L'Italien, J., ed) Vol. 1, Plenum Publishing Corp. New York,
t
>.,Bioi.
u.
degree of homology betweenthem (1). Nevertheless, there are also clear distinctions between the binding of proteins to the two enzymes, as further evidenced by the results presented here. As noted by the data of Fig. 1 and Table I, when the NH2-terminal region of PKI(5-24) is also present inthe inhibitory peptides, binding of the arginine cluster to the cGMP-dependent protein kinase is partially diminished. Given the difference in theactivation mechanisms of the two cyclic nucleotide-dependent protein kinases and, in consequence, the continued proximity in the cGMP-dependent protein kinase of the regulatory and catalytic regions, then possibly the regulatory region by steric hindrance may diminish binding when the NHz-terminaldomain of the inhibitory peptide is present. Presumably, this difference in recognition of the NHz-terminal domain is one reason why the inhibitor protein is selective for the CAMP-dependent protein kinase. Other features of the native inhibitor protein, however, must also contribute to itshigh specificitysince the native inhibitor protein does not bind to thecGMP-dependent protein kinase with an affinity similar to that of the inhibitory peptides
".~ " 1"
"*OS*
27. Scott, J. D.,Glaccum, M. B., Fischer, E. H., and Krebs, E. G . (1986) P m . Natl. A c d . Sei. U. S. A. 8 3 , 1613-1616 28. Glass, D. B., and Krebs, E . G. (1982) J. Biol. Chem. 2 5 7 , 1196-1200 29. Feinberg, R. S., and Mernfield, R. B. (1975) J. Am. Chem. Soc. 97,34853496
30. Kemp; B. E., Graves, D. J., Benjamini, E., and Krebs, E. G. (1977) J.Biol. Chem. 252,4888-4894 31. Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. BWL Chem. 193,265-275 32. Johns, E. W. (1964) Biochem J.92,55-59 33. Oliver, D., Sommer K. R., Panyim, S., Spikes, S., and Chalkey, R. (1972) Biochem. J. 129: 349-353 34. Glass, D. B., and Smith, S. B. (1983) J. Biol. Chem. 258,14797-14803 35. Glass, D. B., Masaracchia, R. A., Feramisco, J. R., and Kemp, B. E. (1978) Anal. Biochem. 87,566-575 36. Roskoski, R., Jr. (1983) Methods Enzynwl. 99,3-6 37. Cleland, W. W. (1979) Methods Enzynwl. 63,103-138 38. Glynn, I. M., and Chappell, J. B., (1964) Biochem. J. 90,147-149 39. Fkimann, E. M., Walsh, D. A., and Krebs, E. G. (1971) J.Bml. Chem. 246,
"_-
19M-1 w.5 " "
40. Glass D.B. McFann, L. J Miller, M. D and Zeilig,C. E. (1981) Cold Sphng HArbor Conf Cell ?rolif 8,267-2Gl 41. Feramisco J. R., and Krebs, E. G. (1978) J. Biol. Chem. 253,896%3971 42. Whitehow&, S., Feramisco, J. R., Casnellie, J. E., Krebs, E. G., and Walsh, D. A. (1983) J. Biol. Chem. 258,3693-3701 43. Granot, J., Mildvan, A. S., Bramson, H. N., Thomas, N., and Kaiser, E. T. (1981) Biochemist 20 602-610 44. Walton, G . M., and 8 1 1 , N.(1980) J. BioL Chem. 255,1603-1609 45. Walton, G. M and Gill, G. N.(1981) J. Bwl. Chem. 256,1681-1688 46. Cleland, W. I$. (1963) Biochim. Biophys. Acta 6 7 , 173-187
d.