THE JOURNAL 0 1992 by The American Society for Biochemistry and

OF BIOLOGICAL CHEMISTRY Molecular Biology, Inc.

Vol. 267, No. 10, Issue of April 5, pp. 7042-7047, 1992 Printed in U. S. A.

The 90-kDa Heat Shock Protein, HSP90, Binds and Protects Casein Kinase I1 from Self-aggregation and Enhances Its Kinase Activity*

(Received for publication, August 12,1991)

Yoshihiko Miyata and Ichiro YaharaS From the Department of Cell Bwbgy, The Tokyo Metropolitan Institute of Medical Science, 3-18-22, Hon-Komagome, Bunkyo-ku, Tokyo 113, Japan

We found that a preparation of the 90-kDa heat shock protein, HSP90, purified to apparent homoge- neity, contains a serinelthreonine kinase which phos- phorylates HSPOO. The protein kinase was identified as casein kinase I1 (CKII) according to its properties. The protein kinase was separable from HSP90 by ad- sorption to heparin-Sepharose or phosphocellulose. CKII was coimmunoprecipitated with HSP9O by anti- HSP9O antibodies from cell extracts. Sucrose density gradient centrifugation analysis revealed that an ad- dition of anti-HSP9O antibodies to cell extracts induces a shift of the sedimentation peak of CKII toward the bottom of a centrifuge tube. These results suggest that CKII is associated with HSP90 in cell lysates at low salt conditions. Furthermore, the CKII. HSP9O com- plex was reconstituted from purified HSP9O-free CKII and CKII-free HSPOO. In a buffer at low ionic strength, CKII forms large aggregates, but HSP9O dissociates the aggregates. Finally, we found that HSP90 activates CKII; an addition of HSP9O to CKII dramatically in- creased phosphorylation of exogenous substrates as well as the CKIID subunit. Taken altogether, these observations suggest that CKII is structurally and functionally active when it forms a complex with HSP90.

~~ ~ ~~

Various environmental stresses induce the expression of heat shock proteins, or stress proteins (for review, see Lind- quist and Craig, 1988; Schlesinger, 1990). Recently, it has been widely accepted that heat shock proteins play important roles in living organisms. The 90-kDa major heat shock pro- tein, HSPSO,’ is synthesized abundantly in unstressed cells and is induced severalfold upon heat shock (Lindquist and Craig, 1988, Schlesinger, 1990). HSP9O is an essential protein for Saccharomyces cereuisiae, even under unstressful condi- tions (Borkovich et al., 1989), and is highly conserved during evolution (Hickey et al., 1989). Thus, HSP9O is likely to function as an essential protein in higher eukaryotic cells under normal conditions as well as stressful conditions. In

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ‘‘aduertkement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$To whom correspondence should be addressed. Tel.: 03-3823- 2101 (ext. 5201).

The abbreviations used are: HSP90, 90-kDa heat shock protein; CKII, casein kinase 11; Hepes, 4-(2-hydroxyethyl)-l-piperazineeth- anesulfonic acid; DTT, dithiothreitok EGTA, [ethylenebis(oxy- ethylenenitri1o)ltetraacetic acid; Mops, 3-[N-morpholino]propane- aulfonic acid; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; Pipes, piperazine-N,N”bis[2-ethanesulfonic acid]; CBB, Coomassie Brilliant Blue R-250; BSA, bovine serum albumin.

fact, various functionally key proteins, such as those belonging to the steroid hormone receptor superfamily proteins (Joab et al., 1984; Catelli et al., 1985; Sanchez et al., 1985) and avian sarcoma virus transforming proteins (Opperman et al., 1981; Brugge et al., 1981; Schuh et al., 1985), have been shown to form complexes with HSP9O. We have shown previously that HSP9O binds to filamentous actin in a Ca2+-calmodulin- regulated manner (Koyasu et al., 1986; Nishida et al., 1986) and, therefore, functions as an anchoring system of the above complexes to cytoskeletal structures (Miyata and Yahara, 1991). In addition, HSP9O may function in maintaining con- formation of the counterpart protein in the complexes so that function(s) of the protein is expressed optimally. For instance, the affinity for glucocorticoids has been shown to be higher in the glucocorticoid receptor-HSP9O complex than in the free receptor (Bresnick et al., 1989). Furthermore, a positive role of HSP9O in signal transduction of glucocorticoid recep- tor has been also demonstrated (Picard et aL, 1990).

In the course of study, we found that an incubation of highly purified HSP9O with [y3’P]ATP resulted in its phos- phorylation. This result seems to suggest that a protein ki- nase(s) might be copurified together with HSP9O and that the copurified kinase(s) phosphorylates HSP9O. Among several protein kinases which have been reported to phosphorylate HSP90, casein kinase type TS (=type 11) has been shown to be coeluted with HSP9O through a gel filtration column (Meggio et al., 1985; Dougherty et al., 1987). These facts led us to examine whether the kinase contained in the preparation of HSP9O is casein kinase I1 (CKII) or not.

CKII was identified as a second messenger-independent protein kinase and shown to be very ubiquitous and conserved during evolution (for review, see Pinna, 1990; Tuazon and Traugh, 1991). Although heparin inhibits while polylysine activates the kinase in uitro, little is known about the regu- latory mechanism of the kinase in intact cells. The self- aggregation of purified CKII has been demonstrated at low ionic strength and also under physiological conditions (Glover, 1986), implicating the negative control mechanism of the protein kinase. A number of both cytosolic and nuclear proteins have been shown to be potential substrates of CKII (Pinna, 1990; Tuazon and Traugh, 1991). Among them, of interest are those concerning the regulation of cell growth that include the myc protein (Luscher et al., 19891, p53 (Meek et al., 1990), SV40 large T antigen (Grasser et al., 1988), and serum response factor (Manak et al., 1990). The CKII activity is stimulated significantly in response to several mitogens (Sommercorn et al., 1987; Klarlund and Czech, 1988; Acker- man and Osheroff, 1989), and microinjection of the kinase to Xenopus oocytes induces their maturation (Mulner-Lorillon et al., 1988). These results altogether suggest that CKII might be one of the key protein kinases involved in signal transduc-

7042

Association and Activation of Casein Kinase 11 by HSPSO 7043

tion of the cell cycle and cell growth regulation (Krebs et al., 1988; Carroll et al., 1988).

Here we report that CKII was copurified with HSP9O and that CKII was associated with HSP90 in cell lysates. Fur- thermore, we found that HSP9O disaggregates large polymeric forms of CKII and converts them into soluble CKII HSP9O complexes. Since an addition of HSP9O to purified CKII prominently stimulated the kinase activity, the complex for- mation is necessary to keep CKII active. These results, to- gether with the previous report that eIF2a kinase is activated by HSPSO (Rose et al., 1987; Matts and Hurst, 1989), provide a new insight into functions of HSP9O and also into the regulation of CKII activity.

EXPERIMENTAL PROCEDURES

Peptide-A synthetic peptide Arg-Arg-Arg-Glu-Glu-Glu-Thr-Glu- Glu-Glu (RRREEETEEE), a specific substrate for CKII (Kuenzel and Krebs, 1985), was synthesized on a Applied Biosystem's solid- phase peptide synthesizer. The peptide was purified on PepRPC column (Pharmacia) using a linear gradient of acetonitrile (0-100%) in the presence of 0.1% trifluoroacetic acid.

Buffers and Cells-HEDG buffer: 25 mM Hepes, 1.5 mM EDTA, 1 mM DTT, and 10% (v/v) glycerol, pH 7.6. Assay buffer: 20 mM Tris- c1, 20 mM KCl, 10 mM MgClz, 60 mM NaCl, 10 mM sodium metabi- sulfite, 20 mM &glycerophosphate, 6 mM EGTA, 6 mM p-nitrophen- ylphosphate, 1 mM DTT, 30 p~ [y3'P]ATP, pH 7.8. KCM buffer: 130 mM KH2P04, 0.1 mM CaClz, 1 mM MgC12, 20 mM Hepes, 10 mM NaCl, pH 7.2. Mouse L cells were cultured in minimal Eagle's medium supplemented with 5% fetal calf serum (GIBCO).

Purification of Porcine Testis Casein Kinase ZZ-CKII was purified from porcine testis by the method described previously (Litchfield et al., 1990) with modifications and further purified using Mono Q fast protein liquid chromatography column (Pharmacia). Briefly, crude lysates of testis containing CKII activity were separated using se- quential column chromatography on DE52 (Whatman), phosphocel- lulose P11 (Whatman), and hydroxylapatite (Calbiochem). Active fractions were collected, and proteins were precipitated with solid ammonium sulfate, solubilized, and applied sequentially to a HiLoad Superdex-200 gel filtration column (Pharmacia) and a heparin-Seph- arose column (Pharmacia) as described (Litchfield et al., 1990). Eluted CKII was further purified with a Mono Q fast protein liquid chro- matography column using a 200-1,000 mM NaCl linear gradient in a buffer containing 15 mM Mops, 1.5 mM EDTA, 0.75 mM DTT, and 25% (v/v) glycerol, pH 7.0. The CKII activity was eluted from the column as a single peak coinciding with the protein peak, and SDS- PAGE analysis of the obtained CKII revealed three polypeptides of a (42 kDa), a' (40 kDa), and fl (27 kDa) subunits.

Purification of HSPSO-Mouse HSP9O was purified to homogene- ity as judged by CBB staining patterns on SDS-PAGE from mouse lymphoma L5178Y cells as described previously (Koyasu et al., 1986; Nishida et al., 1986, Yonezawa et al., 1988).

The expression plasmid of the cloned recombinant S. cerevisiae HSP9O in PETl ld was expressed with the T7 RNA polymerase system in Escherichia coli (DE21 Lys E). HSP9O was induced by the addition of P-D-thiogalactopyranoside (37 "C 2 h) and was purified by the same method as in the case of the purification of mouse HSPSO. Cloning, expression in E. coli, and purification of S. cerevisiae HSP9O were carried out in collaboration with Y. Kimura.2

Preparation of L Cell Lysates-Confluent mouse L cells were washed with ice-cold phosphate-buffered saline and received HEDG buffer containing protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 10 pg/ml aprotinin, and 2 pg/ml leupeptin). The cell mono- layers were immediately frozen with liquid nitrogen and then scraped, thawed, and solubilized on ice. The extracts were centrifuged at 22,000 X g for 20 min at 2 "C, and the supernatants were further centrifuged at 100,000 X g for 60 min at 2 "C, resulting in L cell lysates.

Antibodies-Rabbit antisera against mouse and yeast HSP9O were separately prepared by immunizing the corresponding purified pro- teins as antigens to rabbits. IgG fractions were prepared from the antisera using a protein A-Sepharose (Pharmacia) column.

Immunoprecipitation of HSPSO and Associated Proteins-L cell lysates were incubated with anti-HSP9O IgG (0.5 mg/ml) for 120 min on ice, further mixed with protein A-Sepharose suspension, and

Y. Kimura, unpublished data.

incubated at 4 "C for 60 min. The resin was washed five times with HEDG buffer, and adsorbed proteins including HSP9O and associated proteins were eluted by boiling in a SDS-PAGE sample buffer. The CKII activity in the immunocomplexes was analyzed using the active gel phosphorylation assay as described below.

Assay for the CKZZActivity-CKII was assayed at 30 'C for 10 min in assay buffer containing 0.7 mg/ml CKII-specific synthetic peptide substrate (RRREEETEEE), and incorporation of radioactive ATP into the peptide was determined using P81 phosphocellulose papers as described elsewhere (Kuenzel and Krebs, 1985).

Sucrose Density Gradient Centrifugation-L cell lysates or purified CKII (diluted in HEDG buffer to 10 pg/ml) were incubated with or without anti-HSP9O IgG (0.5 mg/ml) for 120 min on ice. For the reconstitution experiments, purified CKII was preincubated for 60 min on ice with HSP9O (1.0 mg/ml) in HEDG buffer. Samples were layered onto a linear sucrose density gradient (5-20%) prepared in HEDG buffer and were centrifuged at 45,000 rpm for 16 h at 2 "C in a Beckman SW 50.1 rotor. Fractions were successively collected and diluted in assay buffer at start time of the CKII activity assay. Sedimentation coefficients were determined using ovalbumin, bovine serum albumin, aldolase, and catalase as marker proteins.

Detection of CKII Activity after SDS-PAGE in Renatured Gels (Active Gel Phosphorylation Assay)-The method described for cal- modulin-dependent kinase I1 (Kameshita and Fujisawa, 1989) was modified to detect the CKII activity. Immunocomplexes were electro- phoresed on a SDS-polyacrylamide gel containing 2 mg/ml of de- phosphorylated casein (Sigma). After electrophoresis, SDS was re- moved and CKII in the gel was denatured with 6 M guanidine HC1 in a buffer (50 mM Tris-C1, 5 mM 2-mercaptoethanol, pH 8.0) for 1 h and then renatured with the same buffer containing 0.04% Tween 40 at 4 'C for 16 h. Phosphorylation was performed by incubating the gel in a buffer (40 mM Tris-Cl, 50 mM NaCl, 20 mM KC1, 10 mM MgC12, 0.1 mM EGTA, 2 mM DTT, pH 8.0) containing 50 p~ cold ATP and 3 MBq of [y3'P]ATP for 60 min at room temperature. The gel was fixed and washed extensively with 5% (w/v) trichloroacetic acid containing 1% sodium pyrophosphate and dried for autoradiog- raphy.

Effects of HSPSO on the Activity of Purified CKZI-Purified CKII (diluted in HEDG buffer at 4 pg/ml) was mixed with various concen- trations of kinase-free HSP9O (0-0.4 mg/ml) in assay buffer and incubated for 2-30 min (normally 10 min) at 30 "C. The total assay volume (20 pl) and the ionic condition of the samples were kept constant. The mixtures were boiled in SDS sample buffer for 5 min and subjected to SDS-PAGE. When indicated, the synthetic peptide substrate was added to assay buffer as described above. To examine the effect of HSP9O on CKII activity in a physiological buffer, CKII was diluted in HEDG or KCM buffer containing increasing amounts of HSPSO, and the phosphate incorporation into the substrate peptide was determined in KCM buffer at 37 "C for 10 min.

RESULTS

Casein Kinase II Was Copurified with HSPSO-HSPSO was purified from L5178Y cells using sequential column chroma- tography on DE52, hydroxylapatite, gel filtration, and Mono Q (Welch and Feramisco, 1982; Koyasu et al., 1986; Yonezawa et al., 1988). The purified HSP9O appeared to be homogeneous as judged by SDS-PAGE followed by CBB staining (Fig. 1A). We found, however, that HSP9O in the above final prepara- tion was phosphorylated upon incubation with [y-32P]ATP without addition of other materials (Fig. lB, lune 1 ). This implies that a protein kinase(s) might be copurified with HSP9O and phosphorylates HSPSO. The copurified protein kinase(s) was adsorbed to phosphocellulose (Fig. lB, lune 2) and heparin-Sepharose (Fig. lB, lune 3 ) in high salt condi- tions (0.4 and 0.2 M NaC1, respectively). HSPSO itself did not bind to these resins as shown in Fig. l.4 (lane 1-3).

Several characteristics of the copurified protein kinase in the HSPSO preparation were investigated. The protein kinase was identified as CKII according to its properties; inhibition by heparin, stimulation by polylysine, utilization of GTP as well as ATP as phosphate donors, substrate specificity, and molecular mass (43/40 kDa) of its catalytic subunit (data not shown).

7044 Association and Activation of Casein Kinase 11 by HSPSO

30-1 - 22- - 74- -

CBB AR

FIG. 1. Highly purified HSPSO fractions contained a HSPSO-kinase(s) which were adsorbed to phosphocellulose and heparin-Sepharose. 360 pg of highly purified HSPSO fractions were mixed with 30 pl of control Sepharose 4B (lanes I), phospho- cellulose (lanes 2), or heparin-Sepharose (lanes 3 ) in a buffer (40 mM Pipes, 1.2 mM MgC12, 0.8 mM EGTA, 0.4 mM DTT, 3.2 mM @- glycerophosphate, 60 mM sucrose, pH 6.8) containing 0.2 M (for Sepharose 4B or heparin-Sepharose) or 0.4 M (for phosphocellulose) of NaCl a t 4 "C for 120 min, and adsorbed proteins to resins were removed by a centrifugation. Proteins in the supernatants ( 5 pg) were incubated with [y3'P]ATP in the assay buffer for 10 min at 30 "C and analyzed by SDS-PAGE. CBB staining of the gel with marker proteins ( A ) and the autoradiogram ( B ) are shown.

Ill 211 10

Fraction FIG. 2. Copurification of CKII with HSPSO. HSPSO was pu-

rified from mouse lymphoma L5178Y cells as described previously (Koyasu et al., 1986; Yonezawa et al., 1988). The amount of HSPSO contained in each fraction (0-0) was determined by SDS-PAGE followed by CBB staining and densitometry. The CKII activity (0- 0) was determined using a synthetic peptide (RRREEETEEE) as a substrate. These values are expressed as percent of the total. A , DE52; B, hydroxylapatite; C, Sephacryl S-300 gel filtration; D, Mono Q. The column chromatographies were performed in a buffer consisting of: A , 100 mM Pipes, 1 mM MgC12, pH 6.9; B, 20 mM potassium phos- phate, 15 mM 2-mercaptoethanol, pH 7.6; C, 20 mM Tris acetate, 20 mM NaCI, 0.1 mM EDTA, 15 mM 2-mercaptoethanol, pH 7.6; D, 50 mM Tris-C1, pH 7.6. Gradients of NaCl ( A and D ) or potassium phosphate ( B ) are indicated by dashed lines.

We pursued the CKII activity during the purification steps of HSPSO. The results shown in Fig. 2 revealed that CKII coeluted with HSPSO from DE52 (A), hydroxylapatite (B), and Mono Q (D) columns. The CKII activity was eluted slightly faster than HSPSO from a gel filtration column (Fig. 2C). We interpreted these results as an indication that CKII

possibly forms a complex with HSPSO and that the HSPSO. CKII complex is larger than HSPSO dimer. We then critically examined this possibility.

Association of CKII with HSPSO in Cell Lysates-Total lysates of L cells were prepared in HEDG buffer, a low salt buffer. Aliquots were incubated with or without anti-HSPSO IgG and centrifuged in 5-20% sucrose density gradient in HEDG buffer. First, the specificity of the antibody was ex- amined by western blotting (Fig. 3A). The antibody specifi- cally recognized HSPSO in L cell lysate (lane 1 ), reacted with purified HSPSO (lane 2), but did not cross-react with purified CKIIa/a'/P (lane 3). Fig. 3B (M) revealed that the CKII activity exists as a single peak with a sedimentation coeffi- cient of -8 S. It has been reported that purified CKII under high salt conditions exists as a -6 S a2Pz heterooligomer (Glover, 1986; Pinna, 1990; Tuazon and Traugh, 1991), sug- gesting that CKII contained in L cell lysates exists as a larger oligomeric structure than the a& form. To examine the possibility that the 8 S-CKII complex contains HSPSO as a component, we determined an effect of anti-HSPSO antibody on the sedimentation profile of 8 S CKII. The result shown in Fig. 3B (0) clearly indicated that a sedimentation coeffi- cient of CKII was increased to -10 S by anti-HSPSO antibod- ies. The shift of sedimentation peak of CKII by anti-HSPSO antibodies is similar to that of steroid hormone receptor- HSPSO complexes (Joab et aL, 1984). This result strongly suggests that CKII exists in cell lysates as an -8 S complex with HSPSO. Nonimmune bovine IgG at a similar concentra-

2004

1164 97-

664

cHSP9O-

4 5 4

314

Front+ 224

- - I cCKlla/a' I cCKllS

CBB Anti HSP9O

IBI

, Rntm ~ractiar No. *o

FIG. 3. A , specificity of the anti-HSP9O antibody. 50 pg of L cell total lysate (lane I), 0.2 pg of purified HSPSO (lane 2), and 1 pg of purified CKII (lane 3 ) were separated by SDS-PAGE and stained with CBB (left) or anti-HSPSO antibody (right). The positions of molecular weight markers, HSPSO, and CKIIa, a', @ were indicated. B, effect of anti-HSPSO antibodies on the sedimentation profile of CKII in L cell lysates in HEDG buffer. L cell lysates were prepared in HEDG buffer and incubated with (0) or without (W) anti-HSP9O IgG (0.5 mg/ml). The mixtures were subjected to sucrose density gradient (5-20% in HEDG buffer) centrifugation. The CKII activity in each fraction was determined using a synthetic peptide as a substrate. The position of the sedimentation peak of az@2 CKII in a high salt buffer was indicated by an arrow. Sedimentation coefficients were determined using ovalbumin, bovine serum albumin, aldolase, and catalase as marker proteins (indicated by arrowheads).

Association and Activation of Casein Kinase 11 by HSPSO 7045

tion did not result in any shift of the sedimentation peak of CKII (data not shown). Both purified CKII (cy2P2) and CKII in cell extracts in a high salt buffer existed as a -6 S form (Fig. 3B, arrow), and that anti-HSPSO antibody did not affect the sedimentation profile of 6 s CKII (data not shown). This indicates that an interaction of HSPSO with CKII is sensitive to high salt conditions as is the case for the glucocorticoid receptor-HSPSO complex (Joab et al., 1984; Pratt et al., 1989; Denis and Gustafsson, 1989).

Next, we have examined whether CKII is coimmunoprecip- itated with HSPSO. L cell lysates prepared in HEDG buffer were separately incubated with control IgG, anti-mouse HSPSO IgG, and anti-S. cereuisiue HSPSO IgG. The resulting immunocomplexes were subjected to active gel phosphoryla- tion method using dephosphorylated casein as a substrate followed by autoradiography (Fig. 4). The result revealed that radioactive phosphate was incorporated into casein, forming two bands of -43 and -40 kDa. These values of molecular mass are comparable with those reported for a/a’ catalytic subunits of CKII (Tuazon and Traugh, 1991). These results indicated that the casein kinase activity was coimmunopre- cipitated with HSPSO (Fig. 4).

HSPSO Associates and Disaggregates CKZZ in Vitro-It is of interest to examine whether or not CKII.HSPS0 complexes can be re-constructed in vitro from separately purified com- ponents. HSPSO-free CKII was purified from porcine testis. The CKII-free HSPSO was prepared from conventionally pu- rified HSPSO by adsorbing the kinase with heparin-Sepha- rose. Purified CKII was incubated with or without HSPSO at the same ionic condition, and the mixtures were analyzed by sucrose density gradient centrifugation. CKII formed large polymeric structures at low ionic strength in the absence of HSPSO (Fig. 5,O) as reported previously (Glover, 1986). An incubation of aggregated CKII with an excess amount of HSPSO clearly resulted in 8 S form of CKII (Fig. 5, W). Free HSPSO existed in the mixture as a dimeric form with a sedimentation coefficient of -6 S (indicated by an arrow). When the CKII-specific substrate peptide was omitted from the assay mixture, an incorporation of radioactive phosphate was negligible (Fig. 5,0, O), indicating that the 8 S peak did not result from phosphate incorporation into added HSPSO.

An addition of anti-HSPSO antibodies to the mixture of CKII and HSPSO caused a decrease of the 8 S peak and a shift of the 8 S peak toward the bottom (data not shown).

Top-, -2-

43-kDaj I I

(CKWa) I I

FIG. 4. Coimmunoprecipitation of casein kinase I1 with HSPSO. L cell lysates were prepared in HEDG buffer and separately incubated with nonimmune IgG, anti-mouse HSPSO IgG, and a n t i 3 cereuisiae HSPSO IgG (0.5 mg/ml). Resulting immunocomplexes were adsorbed to protein A-Sepharose, washed extensively, and analyzed by active gel phosphorylation assay for CKII activity, followed by autoradiography.

FIG. 5. Reconstitution of CKII.HSPS0 complexes from pu- rified CKII and HSPSO. Purified CKII (10 pg/ml) was incubated with (M, 0) or without (0,O) HSPSO (1 mg/ml) in HEDG buffer and analyzed by sucrose density gradient centrifugation. CKII activity in each fraction was measured with (0, M) a substrate peptide (0.7 mg/ ml). As a control experiment, the peptide was omitted (0,O) in the assay mixture. The position of free HSPSO dimer is indicated by an arrow.

This result suggests that the reconstituted 8 S a CKII contains HSPSO as has been shown for endogenous 8 S . CKII (Fig. 3B) and 8 S-glucocorticoid receptor (Joab et aL, 1984). Taken together, we conclude that HSPSO disaggregated the CKII large polymer observed in a low salt buffer by forming an 8 S CKII.HSPS0 complex that is probably identical to the CKII complex in cell lysates.

HSPSO Activates Casein Kinase ZZ-As is observed in Fig. 5, the total kinase activities of CKII in the absence of HSPSO (large aggregates) were always lower than those of CKII determined in the presence of HSPSO (8 S complex). This implied that HSPSO enhances CKII activity, thus, we pre- cisely determined an effect of HSPSO on the CKII activity. A fixed amount of purified CKII (diluted in HEDG buffer at 4 pg/ml) was mixed in assay buffer containing different amounts of kinase-free yeast recombinant HSPSO. The total assay volume and the ionic condition were kept constant. The mixture was incubated with [y3*PP]ATP. CKII induced very little autophosphorylation of the cy and P subunits in the absence of HSPSO in the condition used (Fig. 6A). An addition of HSPSO to the phosphorylation mixtures dramatically in- creased the autophosphorylation of CKIIB (Fig. 6A). At the same time, HSPSO was also phosphorylated by CKII. CKIIcy was slightly autophosphorylated by the “hyperactive” CKII in the presence of HSPSO. Phosphorylation by CKII of exog- enous substrates such as a synthetic peptide substrate was also dramatically enhanced by yeast HSPSO in an HSPSO dose-dependent manner (Fig. 6B), and the kinase activity of CKII toward exogenous substrate was very well correlated with the amount of autophosphorylation of CKIIP (compare Fig. 6, A and B). Essentially the same results were obtained with kinase-free purified mouse HSPSO. Taken together, we conclude that HSPSO enhances CKII activity toward all of exogenous substrates, HSPSO moiety of CKIIeHSPSO com- plex, and CKII itself.

We next addressed a question of whether the enhancing effect of HSPSO on CKII activity was seen in buffers contain- ing physiological concentrations of salt. Purified CKII was diluted in KCM buffer, a buffer with the physiological salt composition, or HEDG buffer, and CKII activity toward the synthetic substrate peptide was determined in KCM buffer. The data shown in Fig. I clearly indicated that both purified mouse HSPSO (0-0) and recombinant yeast HSPSO (W-W) enhanced the CKII activity in KCM buffer in a dose-depend-

7046 Association and Activation of Casein Kinase II by HSPSO

s 2 ,woo

P x 2 .g Moo

s

2. e - - l o O D 0 Y

u)

0 0 002 0 0 5 010 0 2 0 0 4 0

HSWO (Wlrm)

FIG. 6. HSPSO activates CKII in a dose-dependent manner. A, purified CKII (diluted in HEDG buffer a t 4 pg/ml) was incubated with [y3*P]ATP at 30 "C for 10 min in the presence of 0-0.40 mg/ ml of protein kinase-free recombinant yeast HSPSO produced in E. coli in assay buffer. The mixtures were subjected to SDS-PAGE analysis and an autoradiogram was shown. B, phosphorylation by CKII was performed as described in A with increasing concentrations of HSPSO. The CKII activities toward a synthetic substrate peptide (RRREEETEEE) were determined and shown. The radioactivity of 10,000 cpm corresponds to 660 nmol of phosphate from ATP into the synthetic peptide substrate/min/mg CKII a t 30 "C.

I 0 0.1 0.2 0.3 0.4 0.5

Protein (mglml)

FIG. 7. HSPSO enhanced CKII activity in a physiological buffer. Purified CKII (pre-diluted in HEDG buffer (0, W) or in KCM buffer (0)) was incubated with [y3'P]ATP at 37 "C for 10 min in the presence of 0-0.40 mg/ml of CKII-free mouse HSPSO (0,O) or recombinant yeast HSPSO produced in E. coli (m) in KCM buffer. As controls, the effect of purified BSA and actin on the CKII activity weie shown (X, +), respectively. The CKII activities toward a syn- thetic substrate peptide (RRREEETEEE) were determined and shown.

ent manner. Moreover, Fig. 7 (0-0) indicated that HSP9O activated CKII that had never been exposed to a low salt buffer (CKII was diluted from a high salt buffer directly to KCM buffer). Control proteins, BSA (x) or actin (+) did not enhance the CKII activity at the same concentration. These results indicate that HSP9O may enhance CKII activity under physiological conditions.

DISCUSSION

Purified CKII exists in high salt buffers as a heterotetra- meric structure, a& which has a sedimentation coefficient of 6 S (Glover, 1986; Pinna, 1990; Tuazon and Traugh, 1991). We have shown, in this study, that CKII prepared from cultured L cells in HEDG buffer exists as an 8 S complex with HSPSO. The HEDG buffer is a buffer generally used for preparation of cytoplasmic 8 S steroid hormone receptors

(Joab et al., 1984; Sanchez et al., 1985). Various steroid hormone receptors and arylhydrocarbon receptors have been shown to interact with HSP9O in HEDG buffer. It has been suggested recently that steroid hormone receptors form com- plexes with HSP9O under physiological ionic strength and, therefore, also in intact cells (Howard and Distelhorst, 1988). Analogously, it is likely that HSP9O interacts with CKII under physiological conditions. In fact, we could observe the acti- vation of CKII with HSP9O in a physiological buffer (see Fig. 7). Copurification of HSP9O with CKII would be ascribed to the complex formation of CKII with HSPSO. These results are consistent with the previous report by others which showed that CKII is coeluted from a gel filtration column with HSP9O in rat liver cytosol (Meggio et al., 1985; Dougherty et al., 1987).

Mixing separately prepared CKII-free HSP9O and HSPSO- free CKII, the complex of CKII and HSP9O was re-con- structed. When purified CKII was exposed to relatively low salt conditions, it formed large aggregates as has been previ- ously reported (Glover, 1986). An addition of HSP9O to ag- gregated CKII induced dissociation of the aggregates, forming soluble CKII-HSP9O complexes. This dissociation of CKII aggregates is associated with augmentation of the CKII activ- ity. HSP9O thus appears to function to allow CKII to become soluble and active. It should be noted that a relatively high concentration of HSP9O (0.4 mg/ml) is required for the opti- mum activation of CKII (4 pg/ml). We have preliminary results that HSP9O as low as 50 pg/ml significantly activated CKII in HEDG buffer containing 10 mM MgCl, and 150 mM NaCl at 37 "C? This activation could not be attributed to a protective effect of high concentrations of proteins from de- naturation; high concentrations of BSA or actin did not disaggregate nor activate CKII at all (see Fig. 7). On the other hand, we should consider intracellular concentration of HSP9O in unstressed cells. The content of HSP9O in un- stressed cells was approximately estimated to be -0.3 mg/ml for rat 3Y1 cells and -1.0 mg/ml for mouse L5178Y cells? Local concentrations of HSP9O in particular regions such as membrane ruffles may be higher than these values. Thus, the physiological concentrations of HSP9O are sufficient to form complex with and activate CKII. It would be possible, there- fore, that CKII really interacts with HSP9O in intact cells.

HSP9O interacts with a variety of functionally key proteins, including steroid hormone receptors (Joab et al., 1984; Catelli et al., 1985; Sanchez et al., 1985), dioxin receptors (Perdew, 1988; Wilhelmsson et aZ., 1990), avian sarcoma virus trans- forming proteins such as pp60"'"" (Brugge et al., 1981; Opper- man et al., 1981; Schuh et al., 1985), and eIF-Pa kinase (Rose et al., 1987; Matts and Hurst, 1989). Positive regulatory roles of HSP9O have been suggested for the glucocorticoid receptor complex containing HSPSO. The association of glucocorticoid receptor polypeptide with HSP9O increases the affinity of the receptor to the ligand (Bresnick et al., 1989). In addition, HSP9O is necessary for gene expression in response to gluco- corticoid (Picard et al., 1990). More recently, we have shown that the glucocorticoid receptor complex binds to actin fila- ments through the HSP9O moiety and suggested that HSP9O might function as an anchoring system of the receptor to the cytoskeleton (Miyata and Yahara, 1991). Negative regulatory roles of HSP9O have been also suggested for the steroid hormone receptor complexes; steroid hormone receptors lose the DNA-binding ability in the complexes with HSP9O (for review, see Denis and Gustafsson, 1989; Pratt et al., 1989). The tyrosine-specific protein kinase activity of pp60"-"" has been shown to be suppressed in the complex with HSP9O and

Y. Miyata and I. Yahara, unpublished results.

Association and Activation of Casein Kinase II by HSPSO 7047

pp50, another cellular protein (Courtneidge and Bishop, 1982). In all of these cases, HSP9O appears to regulate func- tions of the target proteins by forming complexes with the proteins. I t should be pointed out that both steroid hormone receptors and pp60""" form complexes with HSP9O soon after synthesis (Dalman et al., 1989; Brugge et al., 1983) and in vitro reconstitution of the complexes from separately isolated molecules have not been succeeded without denaturation or additional factors (Inano et al., 1990; Scherrer et aL, 1990). By contrast, as has been shown above, an addition of HSP9O to CKII prevents aggregation of CKII by readily forming soluble CKII. HSP9O complexes.

The activation of CKII by HSP9O appears to be similar to that of eIF2a kinase by HSP9O (Rose et aZ., 1987; Matts and Hurst, 1989). However, recombinant yeast HSP9O produced in E. coli as well as purified mouse HSP9O were found to similarly activate CKII, whereas phosphorylated HSP9O does but dephosphorylated HSP9O does not activate eIF2a kinase (Szyszka et al., 1989). In addition, it has not been shown whether or not the complex formation of eIF2a kinase with HSP9O correlate with the activation of eIF2a kinase. At present, it is not clear whether similar mechanisms operate in the activation of CKII and eIF2a kinase. As these two protein kinases can phosphorylate the initiation factor com- plexes (for review, see Hershey, 1989), HSP9O may be in- volved in the regulation of protein synthesis.

Finally, we should consider whether or not the activation of CKII by HSP9O is biologically significant. Since the acti- vation requires a relatively high concentration of HSP9O as discussed above, an increase in the concentration of HSP9O in stressed cells may trigger the further activation of CKII. Alternatively, HSP9O is thought to function as scaffold for CKII and perhaps for other protein kinases including eIF2a kinase. This is rather likely because HSP9O is very abundant in most eukaryotic cells. As was shown in this study, CKII is present as a complex with HSP9O in cell lysates and, there- fore, possibly in intact cells. The complex may interact with cytoskeleton and nuclear skeleton through binding to actin filaments as we have recently shown for the glucocorticoid receptor-HSP9O complex (Miyata and Yahara, 1991). If this is the case, HSP9O would be involved not only in activation of CKII but also in intracellular positioning of CKII, as CKII is known to exist in nucleus as well as in cytoplasm (Filhol et al., 1990). As both HSP9O and CKII are well conserved and ubiquitous among species from yeast to man, the interaction between them may play essential and important roles in cells to growth and function.

Acknowledgments-We thank H. Kawasaki and T. Saido for help- ful suggestions in the synthesis and purification of peptide, s. Nakajo, K. Nakaya, and Y. Nakamura for helpful suggestions in the purifi- cation of CKII, and Y. Kimura for yeast HSPSO. We also thank E. Nishida and Y. Minami for helpful discussion, and J. A. Traugh and L. A. Pinna for preprints. We thank H. Izumi and F. Sameshima for their excellent technical assistance.

REFERENCES Ackerman, P., and Osheroff, N. (1989) J. Biol. Chem. 264,11958-11965 Borkovich, K. A,, Farrelly, F. W., Finkelstein, D. B., Taulien, J., and Lindquist,

Bresnick, E. H., Dalman, F. C., Sanchez, E. R., and Pratt, W. B. (1989) J. Biol.

Brugge, J. S., Erikson, E., and Erikson, R. L. (1981) cell 26,363-372 Brugge, J., Yonemoto, W., and Darrow, D. (1983) Mol. Cell. Biol. 3,9-19 Carroll, D., Santoro, N., and Marshak, D. R. (1988) Cofd Spring Harbor Symp.

Catelli, M. G., Binart, N., Jun Testas, I., Renoir, J. M., Baulieu, E. E.,

Courtneidge, S. A,, and Bishop, J. M. (1982) Proc. Natl. Aead. Sei. U. S. A. 79,

Dalman, F. C., Bresnick, E. H., Patel, P. D., Perdew, G. H., Watson, S. J., Jr.,

Denis, M., and Gustafsson, J.-A. (1989) Cancer Res. Suppl. 49,2275s-2281s Dougherty, J. J., Rabideau, D. A,, Iannotti, A. M., Sullivan, W. P., and Toft,

Filhol, O., Cochet, C., and Chamhaz, E. M. (1990) Biochemistry 29,9928-9936 Glover, C. V. C. (1986) J. Biol. Chem. 261, 14349-14354 Grasser, F. A,, Scheidtmann, K. H., Tuazon, P. T., Traugh, J. A., and Walter,

Hershey, J. W. B. (1989) J. Biol. Chem. 264,20823-20826 Hickey, E., Brandon, S. E., Smale, G., Lloyd, D., and Weber, L. A. (1989) Mol.

Howard, K. J., and Distelhorst, C. W. (1988) J. Biol. Chem. 263, 3474-3481 Inano, K., Haino, M., Iwasaki, M., Ono, N., Horigome, T., and Sugano, H.

Joab, I., Radanyl, C., Renoir, M., Buchou, T., Catelli, M.-G., Binart, N., Mester,

Kameshita, I., and Fujisawa, H. (1989) Anal. Biochem. 183,139-143 Klarlund, J. K., and Czech, M. P. (1988) J. Biol. Chem. 263, 15872-15875 Koyasu, S., Nishida, E., Kadowaki, T., Matsuzaki, F., Iida, K., Harada, F.,

Kasuga, M., Sakai, H., and Yahara, I. (1986) Proc. Natl. Acad. Sci. U. S. A. 83,8054-8058

Krebs, E. G., Eisenman, R. N., Kuenzel, E. A., Litchfield, D. W., Lozeman, F. J., Liischer, B., and Sommercorn, J. (1988) Cold Spring Harbor Symp. Quant.

Kuenzel, E. A,, and Krebs, E. G. (1985) Proc. Natl. Acad. Sci. U. S. A. 82,737- Biol. 63,77-84

Lindquist, S., and Craig, E. A. (1988) Annu. Reu. Genet. 22,631-677 741

Litchfield, D. W., Lozeman, F. J., Piening, C. , Sommercorn, J., Takio, K.,

Liischer, B., Kuenzel, E. A,, Krebs, E. G., and Eisenman, R. N. (1989) EMBO

Manak, J. R., de Bisschop, N., Kris, R. M., and Prywes, R. (1990) Genes & Deu.

Matts, R. L., and Hurst, R. (1989) J. Bwl. Chem. 264,15542-15547 Meek, D. W., Simon, S., Kikkawa, U., and Eckhart, W. (1990) EMBO J. 10,

Meggio, F., Agostinis, P., and Pinna, L. A. (1985) Biochim. Biophys. Acta 846,

Miyata, Y., and Yahara, I. (1991) J. Biol. Chem. 266,8779-8783 Mulner-Lorillon, O., Marot, J., Cayla, X., Pouhle, R., and Belle, R. (1988) Eur.

Nishida, E., Koyasu, S., Sakai, H., and Yahara, I. (1986) J. Biol. Chem. 261,

Oppermann, H., Levinson, W., and Bishop, J. M. (1981) Proc. Natl. Acad. Sci.

Perdew, G. H. (1988) J. Biol. Chem. 263,13802-13805 Picard, D., Khursheed, B., Garabedian, M. J., Fortin, M. G., Lindquist, S., and

Pinna, L. A. (1990) Biochim. Biophys. Acta 1054, 267-284 Pratt, W. B., Sanchez, E. R., Bresnick, E. H., Meshinchi, S., Scherrer, L. C.,

Dalman, F. C., and Welsh, M. J. (1989) Cancer Res. Suppl. 49, 2222s-2229s Rose, D. W., Wettenhall, R. E. H., Kudlicki, W., Kramer, G., and Hardesty, B. (1987) Biochemistry 26,6583-6587

Sanchez, E. R., Toft, D. 0.. Schlesinger, M. J., and Pratt, W. B. (1985) J. Bbl . Chem. 260,12398-12401

Scherrer, L. C., Dalman, F. C., Massa E., Meshinchi, S., and Pratt, W. B. (1990) J. Biol. Chem. 266, 21397-21400

Schlesinger, M. J. (1990) J. Biol. Chem. 265, 12111-12114 Schuh, S., Yonemoto, W. Brugge, J., Bauer V. J. Riehl, R. M., Sullivan, W.

Sommercorn, J., Mulligan, J. A,, Lozeman, F. J., and Krebs, E. G. (1987) Proc. P., and Toft, D. 0. (198'5) J. Biol. Chem. 260,13292-14296

Szyszka, R., Kramer, G., and Hardesty, B. (1989) Biochemistry 28, 1435-1438 Natl. Acad. Sci. U. S. A. 84, 8834-8838

Tuazon, P. T., and Traugh, J. A. (1991) Adu. Second Messenger Phosphoprotein

Welch, W. J., and Feramisco, J. R. (1982) J. Biol. Chem. 267,14949-14959 Wilhelmsson, A., Cuthill, S., Denis, M., Wikstrom, A.-C., Gustafsson, J.-A.,

Yonezawa, N., Nishida, E., Sakai, H., Koyasu, S., Matsuzaki, F., Iida, K., and

S. (1989) Mol. Cell. Biol. 9, 3919-3930

Chem. 264,4992-4997

Quunt. Bid. 63,91-95

Feramisco, J. R., and Welch, Vf J. (1985) EMBO J. 4,3131-3135

7117-7121

and Pratt, W. B. (1989) J. Biol. Chem. 264, 19815-19821

D. 0. (1987) Biochim. Biophys. Acta 927, 74-80

G. (1988) Virology 166, 13-22

Cell. Biol. 9, 2615-2626

(1990) FEBS Lett. 267,157-159

J., and Baulieu, E.-E. (1984) Nature 308,850-853

Walsh, K. A,, and Krebs, E. G. (1990) J. Biol. Chem. 266, 7638-7644

J. 8,1111-1119

4,955-967

3253-3260

248-256

J. Biochem. 171, 107-117

16033-16036

U. S. A. 78,1067-1071

Yamamoto, K. R. (1990) Nature 348, 166-168

Res. 23,123-164

and Poellinger, L. (1990) EMBO J. 9, 69-76

Yahara, I. (1988) Eur. J. Biochem. 177, 1-7