Klinische Biochemie, Medizinische Universitatsklinik, Josef Schneider Strasse 2, 8700 Wurzburg, West Germany, and the. I( Centre National de la Recherche ...
Vol . 266, NO.28, Issue of October 5, PP. 19018-19022,1991 Printed in U,S.A.
THEJOURNAL OF BIOLOGICAL CHEMISTRY (0 1991 by The American Society for Biochemistry and Molecular Biology, Inc.
Association of Type I1 CAMP-dependent Protein Kinase with ~ 3 4 ~ ~ ' ' Protein Kinase in Human Fibroblasts* (Received for publication, January 14, 1991)
Sylvie TournierSB, Frangoise RaynaudS, Pascale Gerbaud$, Suzanne M. LohmannlI,Marcel Dorbell, and Daniele Evain-Brio& From the SLaboratoire de Physiopathologie du Developpement, ENS 46 rue d'Ulm 75230 Paris Cedex 05, France, the lLabor fur Klinische Biochemie, Medizinische Universitatsklinik, Josef Schneider Strasse 2, 8700 Wurzburg, West Germany, and the I(Centre National dela Recherche Scientifique and Znstitut National de la Sante et dela Recherche Medicale, BP 5051, ~. 34033 Montpellier, Cedex, France
Previous independent studies suggested that type I1 further examine the role of this and other kinases in fibroblast CAMP-dependent protein kinase and the ~ 3 4 "pro~ " ~ cell growth. tein kinase cell cycle regulator co-localize at centroEvidence for cell cycle and growth control by CAMP-desomes. In order to investigate whether there is an pendent protein kinase and by ~ 3 4 " ~protein "' kinase has been association of type I1 CAMP-dependent protein kinase obtained by studies in yeast and oocytes. In thebudding yeast, with ~ 3 4 " ~in " ' human fibroblasts, we used three dif- Saccharomyces cerevisiae, gene disruption experiments ferent approaches. First, the regulatory subunits RI showed that at least oneof the three TPK genes for Csubunit and RII were photoaffinity-labeled with 8-N3-[32P] was required for a cell to grow normally and that disruption CAMP, and anti-~34"~"' immunoprecipitates were screened for the presence of either RI or RII regulatory of the BCY gene for R subunit caused the constitutive acticells subunits by one- or two-dimensional gel electrophore- vation of the C subunit resulting in, among other things, that displayed a failure to arrest properly in GI (13). In sis. Second, anti-RIIa immunoprecipitates were addition, a certain set of phosphorylated proteins was obscreened for the presence of ~ 3 4 "by ~ Western "~ blot using three different affinity-purified antibodies rec- served to correlate with cell proliferation, whereas another ognizing different domains of human ~ 3 4 ' ~ " Con~ . set correlated with cell cycle arrest (14). Evidence indicated versely, a n t i - ~ 3 4 "immunoprecipitates ~"~ (three differ- that these phosphorylations may be controlled by CAMPent antibodies), as well as the material retained on dependent protein kinase, although involvement of a start pl3"""'-SepharoseBio-Beads, which binds specifically gene regulator with protein kinase activity, CDC28, was also ~ 3 4 " ~ "were , screened for the presence of RIIa. Fi- considered. nally, we have looked for CAMP-dependent protein The fission yeast Saccharomyces pombe, which has a typikinase activity specificallyinhibited by PKI in immu- cally eukaryotic mitoticcycle, has a corresponding cdc2+ gene noprecipitates obtained from extracts treated with dif- that encodes a highly conserved cell cycle regulatory protein ferent anti-p3PdC2 antibodies. kinase, ~ 3 4 ' ~(15). " ~ A human homologue of cdc2' has been All these experiments gave concordant results and cloned (16), and~ 3 4 ' has ~ ' ~beenproposed to regulate the cell demonstrate that at least at Go/GI, human fibroblasts cycle a t late GIat the "start" pointbefore DNA synthesis and contain a complex of active type I1 CAMP-dependent of mitosis (15). Recently, protein kinase associated through its RIIa subunit with at late Gz for determining the onset M-phase promoting factor, a universal factor that controls p34"d"Z. the G,-M phase transition in eukaryotic cells, was shown to be a complex between p34cdc2and cyclin B, a protein whose level follows a sawtooth oscillation during cell cycle (17, 18). Much earlier studies in Xenopus oocytes showedthat microinCyclic AMP has been implicated in regulation the of growth jection of CAMP-dependent protein kinase C subunit inhiband differentiation ina variety of normal and malignantcell ited M-phase promoting factor activation following progestertypes (1,2). With a few exceptions, most biological effects of one addition, whereas R subunit could induce "phase proCAMP are mediated through activation of the CAMP-depende n t protein kinase (3, 4), which in mammalians consists of moting factor activation even in the absence of hormonal several typesof holoenzymes derivedfrom four different genes stimulation (19). In mammalian cells, CAMP-dependent protein kinase (20, (61, RIa for regulatory (R) subunits designated RIIa (5), RIIP 21), and ~ 3 4 " (22, ~ " ~23) have both been independently re(7), and RIP (8) and slightly differing a and P forms of C centrosomes. ~ 3 4 " ~ " subunit (9, 10). We have previously demonstrated a deficiency ported to be associated with mammalian association with centrosomes was suggested to perhaps have of type I1 CAMP-dependent protein kinase in hyperproliferaa function in mitotic spindle formation (22). Our results in tive fibroblasts of psoriatic patients (11, 12) and wanted to fibroblast cells now indicate a specific interaction between , * This study was supported by a grant from le Comiti. de la ligue type I1 CAMP-dependent protein kinase and ~ 3 4 " ~ "further implying that thetwo regulatory systems that they represent Parisienne contre le Cancer and the Deutsche Forschungsgemeinschaft (KO 210/11-1). The costs of publication of this article were may have overlappingor interdependent regulatory effects on defrayed in part by the payment of page charges. This article must cell cycle events. therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. EXPERIMENTALPROCEDURES 3 To whom correspondence should be sent: Laboratoire de PhysioMaterials-Ham's F-12 medium and glutamine were from Flow pathologie du Developpement, ENS 8" etage, 46 rue d'Ulm, 75230 Laboratories,Inc.;fetal calf serum was from Seromed(Munich, Paris Cedex 05, France.
19018
19019
Association of PKII with p34cdc2 F.R.G.); 8-azido["'P]cAMP was from ICN K&K Laboratories Inc.; molecular weight standards were from Bio-Rad; and CAMP, PKI,' and antiproteolytics were from Sigma. All other chemicals used were of analytical grade and purchased from Sigma, Merck (Darmstadt, F.R.G.), or Bio-Rad. Fibroblast Culture-Human fibroblasts were isolated from normal adult patients by enzymatic digestion of small pieces (4-mm diameters) of skin punch biopsies as previously described (11).Cells were grown on 100-mm culture dishes in a 5% CO, humidified atmosphere a t 37 "C in culture medium (Ham's F-12, 20% fetal calf serum, penicillin (50 units/ml),and streptomycin (50 pg/ml)) that was changed every 3 days. The cells reached confluency in 7 days. They were used at subconfluency. More than 70% of the cells were shown by flow cytometry analysis to be in the Go/G1phase. Preparation of Cellular Fractions-All procedures were performed at 0-4 "C. Subconfluent cells (5 X 105-106cells/90-mm diameter dish) were washed 3 times with ice-cold buffer (50 mM Tris-HC1, pH 7.8, 0.33 M sucrose, 1mM MgCl,, and 20 pg/ml each of aprotinin, antipain, leupeptin, soybean trypsininhibitor, and benzamidine), harvested from the dish by scraping, and homogenized using a Dounce homogenizer, For photoaffinity-labeling experiments, the extracts were centrifuged at 800 X g for 10 min, and the pellets were discarded. Supernatants were centrifuged at 100,000 g for 45 min, and the resulting supernatants were used as the cytosolic fractions for labeling. Prior to immunoblot analysis, cells were scraped in lysis Buffer A (50 mM Tris-HCI, pH 7.5, 150 mM NaCl, 1%Triton, 0.1% SDS, and 20 pg/ml each of aprotinin, leupeptin, antipain, soybean trypsin inhibitor, and benzamidine). After centrifugation a t 100,000 X g for 20 min, 100 pg of proteinextract was immunoprecipitated with specific antibodies. Protein concentrations were determined by fluorometric assay using bovine serum albumin as the standard(24). Photoaffinity Labelingwith 8-N:,-f2P]cAMP-RI and RII were photoaffinity-labeled as described by Walter et al. (25) in a reaction mixture (80 pl) of 10 mM Mes, pH 6.2, 10 mM MgCl,, 1.0 p~ 8-Na["'PICAMP containing 100 pg of cytosolic protein. 100 p~ cAMP was included in some samples for the determination of nonspecific labeling. Mixtures were incubated for 60 min in the dark at 4 "C and then irradiated for 10 min with a UV lamp. The samples were then either immunoprecipitated with a specific antibody or heated at 100 "C for 2 min in Laemmli sample buffer. The samples were analyzed by SDSPAGE on a 10%gel according to Laemmli (26). For two-dimensional gel electrophoresis, samples were solubilized in isoelectric focusing lysis buffer. Two-dimensional electrophoresis was performed as described by Rabilloud et al. (27) with a 10% acrylamide gel in the second dimension. The gelswere dried and autoradiographed at -80 "C using Cronex 4 Du Pont medical x-ray film. Antibodies and Affinity Gel-The three polyclonal antibodies against peptide sequences of different domains of ~ 3 4 " ~used " ' in this work were raised in rabbits as described by Lee and Nurse (16). The "PSTAIR antibody was made against the EGVPSTAIREISLLKE peptide sequence perfectly conserved from yeast to man in ~ 3 4 ' ~ ' ' (16). The C8 antibody was raised against the LDNQIKKM peptide of the carboxyl terminus of the human ~ 3 4 ~ protein. ~'' The Nl,% antibody was made against the MEDYAKIEKIGEG peptide of the amino terminusof the starfish ~ 3 4 ' ~ (17), ' ' which differs from human P34'd'2 amino terminus by a single substitution (Ala5 for Thr'). As expected, this antibody cross-reacts perfectly with human ~34'"'. All P3CdCr:! antibodies were purified by means of affinity chromatography on a column of the corresponding peptide coupled to cyanogen bromide-activated Sepharose. Polyclonal anti-cyclin A antibodies were raised in rabbits to a bacterially expressed fusion protein to the 10 amino terminal amino acids of the T7 RNA polymerase and the 386 carboxyl-terminal amino acids of human cyclin A (28). They were purified by affinity chromatography on a column of bacterially synthetized cyclin coupled to Sepharose. The polyclonal antibodies against rat heart RIIa,which cross-reacts with human RIIa,2 and the pl3"""-Sepharose gel were prepared as previously reported (17, 21). Immunoprecipitation, AffinityDepletion, and ImmunoblottingFibroblast Triton extract (100 pg of protein) was incubated for 2 h with the specified antibodies in 1 ml of lysis Buffer A, at 0 "C. In
' The abbreviations used are: PKI, protein kinase inhibitor; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline; BSA, bovine serum albumin; Mes, 4-morpholineethanesulfonic acid; EGTA, [ethylenebis(oxyethylenenitri1o)ltetraacetic acid; Mops, 4-morpholinepropanesulfonic acid Ab, antibodies. ' S. M. Lohmann, unpublished results.
some cases, specificity was determined by using antibodies that had been preabsorbed for 5 min at 20 "C with 10 mg/ml competing antigenic peptide before the immunoprecipitation procedure. After incubation with antibodies, samples were treated with 50 p1 of a 10% protein A-Sepharose suspension (PharmaciaLKB Biotechnology Inc.). In some experiments, extracts were directly treated with 50 pl of plT"'-Sepharose (without antibody). This latter affinity matrix, composed of Sepharose Bio-Beads coupled to a p13 protein, aproduct of the yeast gene sucl+, binds specifically the ~ 3 4 ' ~ protein "' (29). After 1 h incubation at 0 "C on a rotator, samples were centrifuged at 12,000 X g for 5 min, and the pellets were washed 4 times with lysis Buffer A and 2 times with 50 mM Tris-HC1, pH 7.5, containing antiproteolytics (20 pg/ml). Laemmli sample buffer (25 pl) was then added, and the samples were heated at 100 "C for 2 min. SDS-PAGE analysis of each sample was performed on a 10% gel. Proteins were transferred to Immobilon PVDF transfer membrane (Millipore) using a semi-dry blotting system (Bio-Rad). The filters were soaked for 1 h in phosphate-buffered saline (PBS) containing 2 mg/mlbovine serum albumin (BSA). The blots were probed with the specified antibody in PBS containing 0.5 mg/ml BSA, 0.1% Tween 20 for 60 min at 37 "C, washed with PBS containing 1% Tween 20 for 30 min, and incubated for 60 min with "'1-protein A (Amersham Corp.) in PBS containing BSA 0.5 mg/ml. Protein Kinase Assay-Cyclic AMP-dependent protein kinase activities were measured in immunoprecipitates as described hy Ros(Kemptide, a highly koski (30), using Leu-Arg-Arg-Ala-Ser-Leu-Gly specific substrate of CAMP-dependent protein kinases) as the phosphate acceptor. 800 pgof protein was immunoprecipitated as previously described, except that thepellets were washed 4 times with lysis Buffer A and 2 times with 50 mM 3 Mops, pH 7.0, containing antiproteolytics (20 pg/ml each). The pellets were used to catalyze the transfer of'"P from ATP (5000 pmol, 5 X lo6 cpm) to 4 pg of Kemptide in the presence of 50 mM Mops, pH 7.0, 250 pg/ml bovine serum albumin, 10 mM MgCl,, 2.5 mM NaF, antiproteolytics (20 pg/ M cAMP in a total volume of 50 pl. The ml) with or without phosphorylation wasallowed to proceed at 37 "C for 10 min with continuous agitation. The reaction was terminated by spotting 40 p1 of the reaction mixture onto a P-81 paper strip that was dropped in ice-cold 10% phosphoric acid (10 ml/paper strip). This was followed by three more 10% phosphoric acid washes with swirling of the strips. After drying, the paper strips were placed into scintillation solvent and counted. The nonspecific activity determined by the precipitation of the extract with protein A-Sepharose alone was subtracted from all experimental values. Histone HI Kinase Assay-Histone H1 kinase activities were measured in immunoprecipitates made from fibroblast extract using histone H1 as the phosphate acceptor. 800 pg of protein was immunoprecipitated as previously described, except that the pellets were washed 4 times with lysis Buffer A and 2 times with 50 mM TrisHCI, pH 7.3, containing antiproteolytics (20 pg/ml each). The pellets were used to catalyze the transfer of "P from ATP (12.5 nmol, 6.2 X lo6 cpm) to 62 pg of histone H1 in the presence of 50 mM Tris-HC1, pH 7.3,80 mM @-glycerophosphate,20 mM EGTA, 15 mM MgCl,, 2.5 mM NaF, 1 mM dithiothreitol, antiproteolytics (20 pg/ml), 10 p~ CAMP-dependent PKI in atotal volume of 50 pl. The phosphorylation was allowedto proceed a t 37 "C for 15 min with continuous agitation. The reaction was terminated as described previously for the determination of CAMP-dependent protein kinase activit,ies.
RESULTS
In vivo labeling of human fibroblasts with 32Pi,followed by immunoprecipitation with antibody against RII subunit and analysis of the immunoprecipitates by SDS-PAGE and autoradiography, always showedtwo phosphorylated bands of molecular mass 56 (RH) and 34 kDa (data notshown). Further experiments were carried out to assess the possibility that the 34-kDa protein may correspond to the cell cycle regulatory protein kinase ~34'~"'. For this purpose, the CAMP-dependent protein kinase regulatory subunits incytosols of human fibroblasts were specifically photoaffinity-labeled with 8-N3-[32P]cAMP.As shown in Fig. L4, lane a, bands of 51 and 56 kDa, representing RI (more abundant) and RII were labeled. A third band of 46 kDa that was inconsistently observed may correspond to an RII degradation product that has been described (31). The
Association of PKII with ~34'~''
19020 A
based on theirmolecular weights, isolectricpoints, and abundance appeared to correspond to RI (two spotsof 51 kDa, PI = 5.45 and 5.5) and RII (two spots of 51-56 kDa, PI = 5.15.3) as shown inpanel a. Studies are in progress to determine a b c d e f K d a and B isoforms whether the two spots in each case represent or in the case of RII, possibly phospho-forms. Labeling of all - 92 four spots was prevented by the presenceof a 100-fold excess of cAMP in the photoaffinity labeling mixture (Fig. lB, panel - 69 b). Only the twoweakly labeledspots inpanel a that appeared RII * p56 to be RII couldbe immunoprecipitated with PSTAIR Ab RI against ~34'~''(panel c) but notby PSTAIR Ab preabsorbed - 46 with PSTAIR peptide(panel d ) nor by preimmune sera (data not shown). T o confirm association of 8-N3-["PICAMP labeled RII with ~34'~'', we used affinity chromatography on - 30 pl3"""-Sepharose. The yeast p13""" protein has been shown to interact specifically with ~34'~''(29). The two spots correB sponding to 8-N3-[32P]cAMP-labeled RII were also detected E - N . - ( 3 2 ~ ) labeling c ~ ~ ~ by autoradiography after two-dimensional SDS-PAGE analysis of the material retained on this matrix (data shown). not The possibility of a physical interaction between RIIa and p34cdC2 was further ascertained by Western blot detection of (Fig. 3) in the material immunoprecipRIIa (Fig. 2) or ~34'~'' itated from human fibroblast extractsby antibodies directed > against either ~34'~'' or RIIa, respectively. I n agreement with CAMP +CAMP the previous results, ~34'~''immunoprecipitates were found 8 - N 3 - ( 3 *P)cAMP+ PSTAIR Ab to contain a 56-kDa protein immunologically recognized as i RIIa, which co-migrated with authentic RIIa (Fig. 2, lanes b and d). This protein was detected not only in the material immunoprecipitated by antibodies directed against the , 2, lane a ) but also against carboxylthe PSTAIR domain (Fig. C d terminal (Fig. 2, lane c ) or the amino-terminal domain of - PSTAIR peptide +PSTAiR peplido '' not shown). Moreoever, RIIa was also retained FIG. 1. Immunoprecipitation of 8-Ns-[52P]cAMP-labeledR ~ 3 4 ' ~ (data with p34cdc' on pl3-Sepharose (lane e) but not on nonconjusubunits from cytosol of human fibroblasts. Cytosols of human fibroblasts werespecificallylabeledwith8-N3-[RZP]cAMP as de- gated Sepharose Bio-Beads( l a n e f ). We were concerned that scribed under "Experimental Procedures," and samples were analyzed the apparent relativemolecular mass of RIIa is very close to on a 10% one-dimensional SDS-PAGE gel (upper panel) or a two- that reported for human cyclin A (58 kDa), which binds dimensional SDS-PAGE gel (lower panel)prior to autoradiography. Identical samples of 100 pg of fibroblast cytosol were photoaffinity Sample precipitation labeled forlanes a-f (panel A ) . Samples in lanes a and b were labeled in the absence and presence, respectively,of 100 p~ unlabeled cAMP I i Ab t o demonstrate specifically labeled bands (12 h exposure). Samples . . equivalent to those in lane a were immunoprecipitatedby antibodies against either RIIa (lane c) or a conserved regionof p34'"* (PSTAIR Ab) (lanes d-f) as shown (5 daysexposure).In lane e, 100 p~ unlabeled cAMP was present duringthe photoaffinity labeling reaca b c d e t Kd g h tion. In lane f , the PSTAIR Ab had been preabsorbed with PSTAIR peptide. Panel B, analysis of similar experiments in two-dimensional 200 SDS-PAGE. Cytosols(100 pg each) were photoaffinity labeledeither 92 in the absence (panels a, c, and d ) or presence (panel b ) of 100 pM 69 unlabeled CAMP. In panels c and d, photoaffinity-labeled cytosols -p60 p56were immunoprecipitated with PSTAIRAb either in the absence (c) 46 lgGor presence ( d ) of an excess of PSTAIR peptide before two-dimen30 sional gel analysis. Arrows at I , RI; arrows at II, RII; a and b, 2 days exposure; c and d, 6 days exposure. Ab
~
8,.
-
-
21
labeling of all three bandswere specific, since it was inhibited by the presence of excess unlabeled cAMP (lane b). The 8FIG. 2. Autoradiogram of a Western blot using RIIa antiN:#"P]cAMP-labeled RII subunit was immunoprecipitated body or cyclin A antibody. Samples from human fibroblasts were equally well by RIIa Ab (lane c ) or by anti-p3Fdc' PSTAIR subjected to SDS-PAGE on a 10% gel and then transferred to an Immobilon PVDF membrane.The membrane was incubated sequenAb raised against a perfectly conserved 16-amino acid domain tially withantiserumagainst RIIa (lanes a-g) or with antiserum of p34'"' (lane d). RII was not detected in the immunoprecipagainst cyclin A (lane h ) and "'1-protein A and then subjected to itate obtained with PSTAIR Ab if a 100-fold excess of cAMP autoradiography.Extracts (100pg each) were either examined directly (lane e ) . for containing RIIa (lanes d and g) or were precipitated with various was present in the photoaffinity labeling incubation agents beforeSDS-PAGE (lanes a-c, e, f ) . Immunoprecipitations Also, co-immunoprecipitation of RII with ~ 3 4 ' ~ 'was ' not detected when p34'"' precipitation was prevented by preab- with PSTAIR Ab (lane a ) , RIIa Ab ( l a n e b), or C8 AB directed against a peptide sequence corresponding to the first eight amino sorption of PSTAIR Ab with PSTAIR peptide(lane f ). acids from the carboxyl-terminal end of the human ~34'~''(lane c) Two-dimensional SDS-PAGE analysis of similar experiare shown. Extracts were also precipitated withpl3"""-Sepharose ments was performed (Fig. 1B). Photoaffinitylabeling of (lane e ) or nonconjugated Sepharose (lane f ). 100pg of protein extract humanfibroblast cytosol demonstratedfourspots,which was examined directly for containing cyclin A (lane h).
Association of PKII with p3dCdc2 Ab for
A
Imrnunopreclp.
_..
b
a
c
d
K
d
-200 92
-
- 69 - 46
lgGm p34.
30
I-
-
21
-
B
Ab for
Immunoprecip. a
a
a
H H
U U
H H
G r r : G K d a b c 69 p34
9 ,
.-
-
19021
Having demonstrated that part of RIIa exists in human fibroblasts at Go/GI as a complex with ~ 3 4 ' ~ ' it~ ,was of interest to check whether this complex has protein kinase activity. We failed to detect any cdc2 kinase activity in RIIa immunoprecipitates. This is perhaps not surprising, because a cyclin subunit seems to be required for cdc2 kinase activity (17). In contrast, asshown in Fig. 4, CAMP-dependentprotein kinase activity specifically inhibited by the CAMP-dependent PKI is found in the immunoprecipitates made.from the fibroblast extract using different ~ 3 4 ' ~ antibodies '' (PSTAIR, NIX or C8 antibody). This activity is inhibited by preincubation of the extract with cAMP (1mM) before immunoprecipitation (data not shown) indicating that thecatalytic subunit of the type I1 CAMP-dependent protein kinase is not directly associated with ~34'~'' but only via the RIIa regulatory subunit. In order to test the stability of the p34'dc2/PKII complex, the co-immunoprecipitated complex was washed4 times with Buffer A (1M NaCl). After these washes with an elevated salt concentration, more than 50% of PKII remained associated with ~ 3 4 ' ~ 'suggesting ~, a stable association.
46
DISCUSSION
30
In thisstudy we provided evidence for immunoprecipitation of RIIa with a p34 protein cross-reacting with either of three different antibodies directed against different domains of the human p34cdc2,including the PSTAIR domain, the amino terminus, and the carboxyl terminus. Conversely, a p34 protein was detected in anti-RIIa immunoprecipitates, which perfectly cross-reacted with either of the three ~ 3 4 ' ~anti'~ bodies (PSTAIR, NI3,or C8). Moreoevera p56 protein crossreacting with specific antibodies against RIIa and covalently labeled with 8-N3-["PICAMP wasretained on a pl3"""-Sepharose affinity matrix that binds ~34'~''. These results demonstrate that RIIa exists as a complex with ~ 3 4 ' ~in' human ~ fibroblasts. The p34 protein associated with RIIa cannot correspond only to the cdc2-like protein reported to bind cyclin A in HeLa cells, which are not recognized by antibodies against the 8-amino acid carboxyl-terminal sequence of ~ 3 4 ' ~(32). ' ~ The immunoprecipitates containing RIIa and ~ 3 4 ' ~demonstrated '~ CAMP-dependent protein kinase activity with Kemptide as a substrate that could be
FIG. 3. Autoradiogram of a Western blot using PSTAIR Ab, NHzMPF Ab, or C8 Ab against p34cdc2.Panel A, samples from human fibroblasts were subjected to SDS-PAGE on a 10% gel and then transferred to an Immobilon PVDF membrane that was then labeled sequentially with PSTAIR Ab and '*'I-protein A and subjected to autoradiography. Extracts (100 pg each) were either examined directly for containing ~34'"'' (lane d ) or were first immunoprecipitated with PSTAIR Ab (lane c), RIIa Ab (lane b), or a preimmune serum (lane a ) , prior to SDS-PAGE. Panel B , samples from human fibroblasts (100 pg each) were immunoprecipitated with RIIa antibody (lanes a-c), subjected to SDS-PAGE, transferred to an Immobilon PVDF membrane, and examined for containing p34'd'' with Nt3 antibody (lane a ) , PSTAIR antibody (lane b ) or C8 antibody (lane c).
specifically a cdc2-like protein (32). As shown in Fig. 2, lune h, human cyclin A runsmore slowly and was clearly discriminated from RIIa (lune g) under our gel system conditions. Moreover, antibodies against the 8-amino acid carboxyl terminus of ~34'~"' do not recognize the cdc2-like protein associated to cyclin A (32). Therefore, the p56 protein recognized by the RIIa antibody in ~ 3 4 ' ~immunoprecipitates '~ from human fibroblasts is clearly distinct from cyclin A (or cyclin B, which migrates even more slowly). Conversely, RIIa immunoprecipitates were screened by Western blotting for immunoreactivity with antibodies directed against ~ 3 4 ' ~(Fig. ' ~ 3). As shown in panel A , lane d, the ~ 3 4 ' ~protein ~' is recognized as a doublet in the fibroblast extract by the anti-PSTAIR antibody. The significance of this doublet is not shown but may correspond to phospho- or dephospho-forms.A doublet protein (34,36 kDa) co-migrating with authentic ~ 3 4 (lanes ' ~ ~c ~ and d ) was clearly recognized not only by antibodies directed against the PSTAIR domain (panel A , lane b and panel B, lune b) or the amino terminus RII PSTAIR NHZWF u1 (panel B, lune a ) but also against the carboxyl-terminal doFIG. 4. Cyclic AMP-dependent protein kinase activity of main of ~ 3 4 ' ~(panel ' ~ B, lune c). This strengthens the view Extracts from huimmunoprecipitates from fibroblast extract. that thep34 protein associated with RIIa is actually ~ 3 4 ' ~ ' ~ . man fibroblasts (800 pgof protein) were immunoprecipitated with Indeed, the single cdc2-like protein identified so far in human different antibodies ( R h , PSTAIR, Nln,and C8 antibody) and the cells (associated with cyclin A) does not react with antibodies CAMP-dependent protein kinase activity was measured on the Sephagainst the carboxyl terminus of authentic ~ 3 4 (32) ~ ~first ' ~ arose pellet in the presence 0 or absence 0 of cAMP M). The CAMP-dependent protein kinase activity was also measured in the identified by its ability to complement cdc2 mutations in ) cAMP (lo-' M) 8. The activity value fission yeast (17). Interestingly, no cyclin A was detected in presence of PKI (10 p ~ and obtained when the fibroblast extract was precipitated with protein Aassociation with ~ 3 4 ' ~in' ~RIIa immunoprecipitates. No at- Sepharose alone in the presence of cAMP M) was subtracted tempt was made to detect cyclin B, which binds ~ 3 4 ' ~only ' ~ from the activity value obtained in the presence of antibody and at the G2-M phase of the cell cycle (33). M) with protein A-Sepharose. cAMP
t
f
19022
Association of P 'KII with ~ 3 4 " ~ ' ~
7. Lee, D. C., Carmichael, D. F., Krebs, E. G., and McKnight, G. S. inhibited by the PKI inhibitor. The presence of CAMP-de(1983) Proc. Natl. Acad. Sci. U. S. A . 80, 3608-3612 pendent protein kinase activity in immunoprecipitates con8. Clegg, C. H., Cadd, G. G., and McKnight, G. S.(1988) Proc. Natl. taining RII/p34cdc2was prevented by prior treatment of cell Acud. Sci. U. S. A. 85,3703-3707 extracts with CAMP that would dissociate type I1 CAMP- 9. Chrivia, J. C., Uhler, M. D., and McKnight, G. S. (1988) J . Biol. dependent protein kinase catalytic subunit from RII before Chem. 263,5739-5744 immunoprecipitation. Thus, the direct interaction seems to 10. Uhler, M. D., Chrivia, J . C., and McKnight, G. S.(1986) J. Biol. Chem. 261,15360-15363 be between RII and ~ 3 4 " ~ " . 11. Evain-Brion, D., Raynaud, F., Plet, A,, Laurent, P., Leduc, B., In contrast, no H1 histone kinase activity of ~ 3 4 ' ~ 'was ' and Anderson, W. B. (1986) Proc. Natl. Acad. Sci. U. S. A . 83, detected. In this respect, it is important to note that thetype 5272-5276 I1 CAMP-dependent protein kina~e/p34"~"'association was 12. Raynaud, F., Liapi, C., Gerbaud, P., Anderson, W. B., and Evainobserved in a population of fibroblasts of which more than Brion, D. (1988) J . Invest. Dermatol. 8 9 , 105-110 70% were shown by flow cytometry analysis to be in the Go/ 13. Toda, T., Cameron, S., Sass, P., Zoller, M., and Wigler, M. (1987) Cell 50,277-287 GI phase. In Go/G1, there is no association between ~ 3 4 " ~ " ' 14. Tripp, M. L., Pinon, R., Meisenhelder, J., and Hunter, T. (1986) kinase and cyclin B (33), and cyclin A was not detected in Proc. Natl. Acad. Sci. U. S. A . 8 3 , 5973-5977 RIIa immunoprecipitates. A cyclin subunit is most likely 15. Lee, M., and Nurse, P. (1988) Trends Genet. 4 , 287-290 required for expression of ~ 3 4 " ~activity " (17). In this study, 16. Lee, M., and Nurse, P. (1987) Nature 3 2 7 , 31-35 we therefore demonstrated that inhuman fibroblasts the type 17. Labbi., J. C., Capony, J. P., Caput, D., Cavadore, J. C., DeranI1 CAMP-dependent protein kinase is associated with the court, J., Kaghad, M., Lelias, J. M., Picard, A., and Dorie, M. (1989) EMBO J. 8,3053-3058 ~ 3 4 " ~kinase. "' However, due to the nature of the experimental procedure used in this work, an in vitro association between 18. Gautier, J., Minshull, J., Lohka, M., Glotzer, M., Hunt, T., and Maller, J. L. (1990) Cell 60,487-494 type I1 CAMP-dependent protein kinase and ~34'~''kinase 19. Maller, J. L., and Krebs, E. G. (1977) J. Eiol. Chem. 2 5 2 , 1712occurring during the homogenization procedure of the cells 1718 cannot be excluded. 20. N i m E. A., Schafer,. G.,. Hilz,. H.,. and Eppenberger, H. M. (1985) Eli1 4 1 , io39-1051 The possible roleof a complex of p34cdc2/PKIIin regulating the cell cycle is of interest. T h e ~ 3 4 ' ~kinase "' can be directly 21. DeCamilli. P.. Moretti. M.. Denis Donini. S.. Walter.. U... and Lohmann, S.'M. (1986) J.'Cell Biol. 103,'189-203 regulated through physical interaction with other proteins 22. Bailly, E., Dori.e, M., Nurse, P., and Bornens, M. (1989) EMBO such as the p13, product of the sucl+ gene (29, 34-37), or J . 8,3985-3995 cyclins (17, 33, 38, 39). 23. Riabowol, K., Draetta, G., Brizuela, L., Vandre, D., and Beach, The identification of substrates for the p3Pdc2kinase reD. (1989) Cell 57,393-401 mains an important task in the elucidation of mitotic events 24. Bohlen, P., Stain, S., Dairman, W., and Unfriend, S. (1973) Arch. Biochem. Biophys. 155,213-220 (for review, see Ref. 40).Lamins, fibrillar proteins of the inner nuclear membranes, have recently been shown to be in vivo 25. Walter, U., Uno, J., Liu, A. Y. C., and Greengard, P. (1973) J. Biol. Chem. 252,6494-6500 substrates (41). Susuki et al. (42) suggested that the target 26. Laemmli, U. K. (1970) Nature 2 2 7 , 680-685 proteins of ~34'~''may be nuclear proteins involved in tran- 27. Rabilloud, T., Hubert, M., and Tarroux, P. (1986) J. Chromatogr. scriptional regulation. Furthermore, centrosomalp34cdc2may 35 1, 77-89 cause phosphorylations involved in formation of the mitotic 28. Wang, J., Chenivesse, X., Henglein, B., and Brechot, C. (1990) Nature 343,555-557 spindle at mitosis (22,43,44). Our results raise the possibility that RII may be one of the ~ 3 4 " substrates, ~~' since it contains 29. Brizuela, L., Draetta, G., and Beach, D. (1987) EMBO J. 6,35073514 a potential phospho-acceptor site (TPPA)that coincides with 30. Roskoski, R. (1983) Methods Enzymol. 99, 3-6 the consensus motif that has been described (40). In conclu- 31. Walter, U., and Greengard, P. (1982) in Handbook of Experimension, our study suggests a field of investigation intothe tal Pharmacology; Cyclic Nucleotides (Nathanson, J. A., and Kebabian, J. W., eds) pp 479-505, Springer Verlag, Berlin-New interaction of the CAMP-dependentprotein kinase and p3Pdc2 York protein kinase in the regulation of cell cycle, growth, and 32. Pines, J., and Hunter, T. (1990) Nature 3 4 6 , 760-763 differentiation. 33. Pines, J., and Hunter, T. (1989) Cell 58,833-846 34. Dunphy, W. G., Brizuela, L., Beach, D., and Newport, J. (1988) REFERENCES "
1. Pastan, I., Johnson, G. S., and Anderson, W.B. (1975) Annu. Rev. Biochem. 44,491-522 2. Boynton, A. L., and Whitfield, J. F. (1983) Adv. Cyclic Nucleotide Res. 15, 193-294 3. Rubin, C. S.,and Rosen, 0.M. (1975) Annu. Rev. Biochem. 4 4 , 831-887 4. Cohen, P. (1982) Nature 296,613-620 5. Scott, J. D., Glaccum, M. B., Zoller, M. J., Uhler, M. D., Helfman, D.M., McKnight, G. S., and Krebs, E. G. (1987) Proc. Natl. Acud. Sci. U. S. A. 8 4 , 5192-5196 6. Jahnsen, T., Hedin, L., Kidd, V. J., Beattie, W. G., Lohmann, S.
M., Walter, U., Durica, J., Schulz, T. Z., Schiltz, E., Browner, M., Lawrence, C. B., Goldman, D., Ratoosh, S. L., and Richards, J. S. (1986) J . Biol. Chem. 261, 12352-12361
Cell 54,423-431 35. Dunphy, W. G., and Newport, J . (1989) Cell 58, 181-191 36. Jessus, C., Ducommun, B., and Beach, D. (1990) FEES Lett. 266,4-8 37. Moreno, S., Hayles, J., and Nurse, P. (1989) Cell 5 8 , 361-372 38. Murray, A. M., and Kirschner, M. W. (1989) Nature 3 3 9 , 275280 39. Minshull, J., Blow, J. J., and Hunt, T.(1989) Cell 56,947-956 40. Moreno, S.,and Nurse, P. (1990) Cell 6 1 , 549-551 41. Peter, M., Nakagawa, J., Dorke, M., Labb6, J . C., and Nigg, E. A. (1990) Cell 6 1,591-602 42. Suzuki, M. (1989) EMBO J. 8,797-804 43. Suzuki, M. (1989) J . Mol. Biol. 207, 61-84 44. Verde, F., Labbi.,J. C., Dori.e, M., and Karsenti, E. (1990) Nature 343,233-238
