Casein kinase II associates with Egr-1 and acts as a negative modulator of its DNA binding and transcription activities in NIH 3T3 cells.

Although the activation domains within early growth response gene protein 1 (Egr-1) have been mapped, little is known of the kinases which phosphorylate Egr-1 and how phosphorylation correlates with the transcriptional activity of Egr-1. In this study we report that casein kinase II (CKII) co-immunoprecipitates with Egr-1 from NIH 3T3 cell lysates. The association of Egr-1 and CKII requires the C terminus of Egr-1 and CKII phosphorylates Egr-1 in vitro. The in vitro phosphorylation of Egr-1 by CKII and that induced by serum in vivo was compared by examining the CNBr-digested fragments of the phosphorylated Egr-1. CKII strongly phosphorylates fragments 7 and 10 which cover part of the activation/nuclear localization and DNA binding domains of Egr-1. CKII also phosphorylates, albeit weakly, fragments 5 and 8 which cover part of activation domain and the entire repression domain of Egr-1, respectively. Strong phosphorylation on fragment 10 as well as fragment 5 was also observed in Egr-1 immunoprecipitated from serum-induced, 32P-labeled cells. CKII phosphorylation of Egr-1 resulted in a decrease of its DNA binding as well as its transcriptional activities.

Although the activation domains within early growth response gene protein 1 (Egr-1) have been mapped, little is known of the kinases which phosphorylate Egr-1 and how phosphorylation correlates with the transcriptional activity of Egr-1. In this study we report that casein kinase II (CKII) co-immunoprecipitates with Egr-1 from NIH 3T3 cell lysates. The association of Egr-1 and CKII requires the C terminus of Egr-1 and CKII phosphorylates Egr-1 in vitro. The in vitro phosphorylation of Egr-1 by CKII and that induced by serum in vivo was compared by examining the CNBr-digested fragments of the phosphorylated Egr-1. CKII strongly phosphorylates fragments 7 and 10 which cover part of the activation/nuclear localization and DNA binding domains of Egr-1. CKII also phosphorylates, albeit weakly, fragments 5 and 8 which cover part of activation domain and the entire repression domain of Egr-1, respectively. Strong phosphorylation on fragment 10 as well as fragment 5 was also observed in Egr-1 immunoprecipitated from serum-induced, 32 P-labeled cells. CKII phosphorylation of Egr-1 resulted in a decrease of its DNA binding as well as its transcriptional activities.
It is likely that the different effects mediated by Egr-1 in various cells result from its phosphorylation and dephosphorylation (10). Reports that cells treated with kinase and phosphatase inhibitors (11,12) show differences in their Egr-1 DNA binding ability support this hypothesis. However, little is known about which kinase(s) phosphorylates Egr-1 and if its functional domain(s) is phosphorylated. In addition, the identification of a repressor protein that associates with Egr-1 implies that protein-protein interactions can also regulate its transcriptional activity (13,14).
The protein kinase CKII is an inducible kinase found in both the cytoplasmic and nuclear compartments, although recently CKII was reported to be a predominantly nuclear protein (15). It is a tetrameric Ser/Thr-specific protein kinase complex containing two catalytic (␣, 44 kDa, ␣Ј, 42 kDa) subunits and two regulatory (␤, 25 kDa) subunits (15). CKII is known to associate with a number of nuclear proteins such as DNA topoisomerase II (16) and FKBP25, the 25-kDa FK506-binding protein (17). CKII also phosphorylates a number of transcription factors and nuclear proteins such as the serum response factor (18), c-Jun (19), and p53 (20). Little is known of the physiological role of CKII or how it regulates the activity of transcription factors in vivo.
In this report, we demonstrate that CKII associates with cellular Egr-1 both in vivo and in vitro and phosphorylates it in vitro. We have also identified the CKII phosphorylated CNBr peptides of Egr-1 and have studied the effect of CKII phosphorylation on Egr-1 DNA binding activity in vitro and its transcriptional activity in vivo.

EXPERIMENTAL PROCEDURES
Reagents-588, an antibody to the last 14 amino acids at the C terminus of Egr-1, was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). 5232 (10) and 1133 are polyclonal antibodies generated against full-length Egr-1 fusion proteins. Anti-Egr-1 monoclonal antibody 4F18 was a gift from Amgen Inc. (Thousand Oaks, CA). Antimouse and anti-rabbit IgG conjugated to horseradish peroxidase were purchased from Sigma. A monoclonal antibody to the CKII ␣-subunit was purchased from Boehringer Mannheim (Mannheim, Germany) and the CKII polyclonal antibody against its ␣-subunit was purchased from Upstate Biotechnology Inc. (Lake Placid, NY). The pCMVCKII ␣-subunit plasmid was a gift from Dr. Steven Sloan (Howard Hughes Medical Institute, University of Pennsylvania School of Medicine, Philadelphia).
Construction of Expression Vector for GST-Egr-1 Fusion Proteins-A near full-length GST-Egr-1 fusion construct was obtained by digesting pCMV-Egr-1 (21) with NcoI to give a 1.92-kb fragment that was cloned into the pGEX-KG (22) NcoI site. The resulting plasmid was named pGEX-NCO 1.9 and its protein product designated F-1. The Egr-1 deletion N-1 was produced by digesting pCMV-Egr-1-TTL (21), which contains a stop codon at nucleotide position 768 of the egr-1 cDNA, with NcoI and inserting it into the NcoI site of pGEX-KG such that a protein encoding only the first 171 amino acids was expressed. This plasmid was called pGEX-Egr-1-TTL. The Egr-1 deletions C-2 and C-1 were made by digesting pCMV-Egr-1-TTL with RsaI (nucleic acids 1221-2253) and PvuII (779 -2060) and cloning the corresponding DNA fragments into the blunted XbaI site in pGEX-KG to produce pGEX-RSA 1.0 and pGEX-PVU 1.2, respectively.
Expression and Purification of Recombinant Egr-1-Recombinant Egr-1 was expressed in Escherichia coli (DH5␣) and bound to glutathione-Sepharose 4B beads (Pharmacia Biotech, Uppsala, Sweden) as described by Smith and Johnson (23) except that bacteria were grown to * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
an A 600 of 0.5 before inducing the fusion protein production with 0.2 mM of isopropyl ␤-D-thiogalactopyranoside for 3 h. GST fusion proteins thus expressed were bound to glutathione-Sepharose beads and stored in phosphate-buffered saline (PBS) with 0.2% sodium azide at 4°C and were used for phosphorylation and binding experiments.
Cytoplasmic and Nuclear Extract Preparation-Confluent NIH 3T3 cells were induced with 20% serum for 1 h. Cytoplasmic and nuclear extracts were prepared as described elsewhere (24,25) with phosphatase and protease inhibitors added to all buffers at the following concentrations: 50 mM NaF, 100 M sodium orthovanadate, 2 g/ml leupeptin, 2 g/ml aprotinin, 1 g/ml pepstatin A, and 1 mM PMSF. Cytoplasmic extracts were prepared by low speed centrifugation of homogenized cells, and nuclear extracts were produced by extracting nuclear proteins with a high salt buffer.
Immunoprecipitation and Western Blots-Immunoprecipitation was performed with 0.5 mg of nuclear protein in a buffer containing 1% Triton X-100, 1% sodium deoxycholate, 150 mM NaCl, 50 mM Tris-HCl, pH 7.4, 1 mM PMSF, and the protease and phosphatase inhibitors as listed above. Samples were precleared with protein A-agarose beads prior to immunoprecipitation with the appropriate antibody. Immunocomplexes bound to protein A-agarose beads were washed in immunoprecipitation buffer four times and once in PBS. The protein samples were separated by 7.5% or 10% SDS-polyacrylamide gel electrophoresis (PAGE) after boiling in Laemmli (26) buffer. Western blots were performed as described by Towbin et al. (27). After the transfer, nitrocellulose membranes were blocked overnight at 4°C with 5% skim milk in PBST (PBS and 0.1% Tween 20) and incubated with the appropriate primary and secondary antibodies in PBST containing 1% skim milk for 1 h each. The bands were visualized either by using a chromogenic reaction (0.6 mg/ml 3Ј3Ј-diaminobenzidine, 0.03% hydrogen peroxide, 50 mM Tris-HCl, pH 7.5, and 0.03% cobalt chloride) or by using the Amersham ECL kit (Amersham International Plc, UK).
Cloning and Expression of CNBr Fragments-The DNA fragments corresponding to the CNBr peptides of Egr-1 were amplified from pUC-Egr-1 by the polymerase chain reaction (28), using oligonucleotides complementary to 5Ј and 3Ј ends of each fragment as primers (Table I) and cloned downstream of the T3, T7 promoters in pCRScript (Stratagene, La Jolla, CA) or the SP6 promoter in pSP65 (Promega, Madison, WI). The oligonucleotides had an EcoRI site at the 5Ј end and a SalI site at the 3Ј end to facilitate cloning into the vectors mentioned, although blunt end ligation was also used. The oligonucleotides were designed with an extra ATG codon at the 5Ј oligonucleotide to ensure ease of visualization of the corresponding peptides on synthesis in the presence of [ 35 S]methionine using the TNT reticulocyte lysate kit (Promega). The peptides produced were analyzed by 16.5% SDS-PAGE by the procedure of Schä gger and Von Jagow (29) with a minor modification as described in the Sigma Technical Bulletin no. MWM-100.
[ 32 P]Orthophosphate Labeling and CNBr Mapping-Cells in 150-mm dishes were labeled for 5 h as described previously by Cao et al. (10) except that [ 32 P]orthophosphate was used at a final concentration of 1 mCi/ml. Cells were lysed in radioimmune precipitation buffer (1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 50 mM Tris-HCl, pH 7.4, and 150 mM NaCl) with appropriate protease and phosphatase inhibitors as listed above. Egr-1 was immunoprecipitated from the lysates and the immunoprecipitates were washed five times with radioimmune precipitation buffer. The proteins were separated by 7.5% SDS-PAGE and transferred to a nitrocellulose membrane. The Egr-1 band, located after an overnight exposure to x-ray film, was excised from the membrane, cut into small pieces, and used for CNBr digestion. CNBr (70 mg/ml in 70% formic acid) digestion was carried out overnight in the dark at room temperature. The supernatant containing digested peptides was transferred to a new tube and lyophilized. The dried peptides were washed with distilled H 2 O twice and then resuspended in sample buffer and analyzed by 16.5% SDS-PAGE as described above.
In Vitro Kinase Assay and Phosphoamino Acid Analysis-The glutathione-Sepharose beads containing GST-Egr-1 fusion proteins (50 l) were resuspended in 100 l of CKII assay buffer (200 mM NaCl, 10 mM MgCl 2 , and 25 mM Tris-HCl, pH 7.4, 10 M ATP) in the presence of 2-5 Ci of [␥-32 P]ATP. The kinase reaction was started by adding 1 unit of CKII (Promega) and incubating samples at 37°C for 30 min. The beads were washed with the Triton buffer (1% Triton X-100, 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM PMSF) four times and separated by 10% SDS-PAGE, dried, and exposed to x-ray film. For phosphoamino acid analysis the gel was blotted onto polyvinylidene difluoride membrane, and the Egr-1 band was excised. Phosphoamino acid analysis was performed as described by Boyle et al. (30) and Kamps (31).
In Vitro Binding Assay-The assay was essentially similar to that described by Jin and Burakoff (17). CKII (50 units) was allowed to autophosphorylate in the presence of [␥-32 P]ATP, and the reaction was stopped by adding 50 mM EDTA. The volume was increased to 1.0 ml with Triton buffer. The labeled kinase was precleared with Sepharose beads containing GST alone (50 l) for 1 h at 4°C. The supernatant was removed after centrifugation at 5000 rpm for 5 min and divided into five tubes containing the three Egr-1 deletions, an almost full-length Egr-1, and GST alone (50 l each). The volume was again increased to 1.0 ml with Triton buffer, and the samples were mixed for 1 h at 4°C. The beads were then washed three times with Triton buffer and once with TBS (10 mM Tris-HCl, pH 8.0, 150 mM NaCl), resuspended in Laemmli buffer, boiled for 5 min, and analyzed by 12.5% SDS-PAGE and exposure to x-ray film.
CKII Activity Assay-Cellular extracts were prepared and assayed for CKII activity as described by Heller-Harrison and Czech (32). 10 g of protein extracts were assayed at 37°C in a total reaction volume of 30 l in CKII assay buffer containing 1 mM synthetic substrate peptide RRREEETEEE (33), 100 M ATP, and 5 Ci of [␥-32 P]ATP. Controls consisted of assays done in the absence of substrate peptide and were subtracted as background. Reactions were terminated by adding an equal volume of 0.01 mM ATP and 0.4 N HCl prior to trichloroacetic acid precipitation. The samples (15 l) were spotted on Whatman P81 paper and washed in 0.5% H 2 PO 4 three times for 10 min each, and the samples were counted.

CKII Co-precipitates with Cellular Egr-1 and Associates with
It in Vitro-A kinase activity was found to be associated with Egr-1 immunoprecipitated from nuclear extracts of serum induced NIH 3T3 and MRC5 cells. This activity was determined to be similar to that of CKII based on its ability to utilize GTP during phosphorylation and its inhibition by heparin (data not shown). To further investigate this, Egr-1 was immunoprecipitated from nuclear extracts by a monoclonal Egr-1 antibody. Conversely, CKII was also precipitated from the same nuclear extracts by a monoclonal CKII antibody. The immunoprecipitated samples were resolved by SDS-PAGE, blotted onto nitrocellulose and probed with a polyclonal antibody to the CKII ␣-subunit. The polyclonal antibody recognizes the CKII ␣-subunit immunoprecipitated by the monoclonal CKII antibody (Fig. 1A, lane 1) as well as that co-immunoprecipitated with Egr-1 (lane 2). CKII was also co-immunoprecipitated with other Egr-1 polyclonal antibodies 588 and 5232, but not with preimmune serum (data not shown). The converse was also true when CKII was immunoprecipitated from cytoplasmic and nuclear extracts with a monoclonal CKII antibody, and the immunoblot was probed with an anti Egr-1 polyclonal antibody. Egr-1 was found to be associated with CKII immunoprecipitated from nuclear extract (Fig. 1B, lane 2), but not from the cytoplasmic sample (Fig. 1B, lane 1).
The association of Egr-1 and CKII in vitro and the localization of CKII binding domain on Egr-1 was further investigated. A near full-length GST-Egr-1 fusion protein (amino acids 29 -533) as well as three Egr-1 deletions coupled to GST were made (Fig. 1C). All the fusion proteins appeared as multiple bands on SDS-PAGE stained with Coomassie Blue (Fig. 1D). This could be due to the degradation of Egr-1 products or could also be due to the multiple initiation sites. The near full-length GST-Egr-1 fusion protein, F-1, migrates as an about 100-kDa protein on SDS-PAGE, although theoretically it is only 82 kDa. This anomalous behavior is observed with cellular Egr-1 as well which runs at 80 kDa instead of its predicted molecular mass of 56 kDa (11). The three Egr-1 deletions coupled to GST, C-2 (shorter C terminus fragment from amino acids 322-533), C-1 (longer C terminus Egr-1 fragment from amino acids 175-533), and N-1 (N terminus fragment of Egr-1 from amino acids 29 -171) each had apparent molecular masses that matched their calculated molecular masses of 51, 66, and 43 kDa, respectively, as shown in Fig. 1, C and D.
Purified CKII can be autophosphorylated on its ␤-subunit, and [␥-32 P]ATP was used to tag CKII by autophosphorylation. The labeled CKII was incubated with the various GST-Egr-1 fusion proteins attached to glutathione-Sepharose beads. As shown in Fig. 1E, the association of CKII with Egr-1 was indicated by the presence of the 25-kDa phosphorylated ␤-subunit band with F-1, but not with GST alone. It was also found that CKII associated with C-2 as well as with C-1, but to a lesser extent. However, no association occurred with N-1. These results indicated that the CKII association domain within Egr-1 lies between amino acids 322 and 519. This region contains the three zinc fingers as well as the 5Ј and 3Ј nuclear localization signals as identified by Gashler et al. (13) and Russo et al. (14). The last 14 amino acids at the C terminus of Egr-1 are unlikely to be involved in CKII binding as immunoprecipitation with 588 (an anti-Egr-1 antibody raised against the last 14 amino acids) did not disrupt the interaction between the two proteins (results not shown).
CKII Phosphorylates Egr-1 in Vitro-In order to ascertain whether CKII phosphorylates Egr-1 in vitro, F-1 was incubated with CKII and 32 P-labeled ATP. As shown in Fig. 2A, F-1 (lane 1) but not GST alone (lane 2) was phosphorylated by CKII. Phosphoamino acid analysis was performed on the phosphorylated F-1, and as shown in Fig. 2B the phosphorylation occurs more on serine than on threonine residues. To test this in vivo, Egr-1 was immunoprecipitated from platelet-derived growth factor-induced, [ 32 P]orthophosphate-labeled, NIH 3T3 cells, and in this case strong phosphorylation on serine and weak phosphorylation on threonine residues were observed (Fig. 2C).
The specificity of Egr-1 phosphorylation by CKII was investigated. Phosphorylation of the F-1 fusion protein by CKII was performed with different concentrations of the kinase. Increas-ing the amount of CKII resulted in a corresponding increase in the phosphorylation of Egr-1 (Fig. 3A, lanes 1-5). It has been reported that CKII phosphorylates a specific substrate, RRREEETEEE (33) and can also phosphorylate consensus sequences SXD/E, SXXD/E, or SXXXD/E (34). The phosphoryla-FIG. 1. CKII co-immunoprecipitates with Egr-1. A, confluent NIH 3T3 cells were induced with serum for 1 h, and nuclear extracts were prepared. The nuclear extracts were immunoprecipitated by a monoclonal CKII antibody to the ␣-subunit (lane 1) or by a monoclonal Egr-1 antibody (lane 2). The immunoprecipitates were resolved by 10% SDS-PAGE, transferred to a nitrocellulose membrane, and probed with an anti-CKII antibody. The bands were visualized with a chromogenic reaction as described under "Experimental Procedures." The 44-kDa CKII ␣-subunit is indicated by ␣. B, the cytoplasmic (lane 1) and nuclear extracts (lane 2) were immunoprecipitated by a monoclonal CKII antibody, and the immunoprecipitates were separated by 7.5% SDS-PAGE, blotted onto a nitrocellulose membrane, and probed with an anti-Egr-1 antibody. The bands were visualized by enhanced chemiluminescence. The location of Egr-1 band is indicated. C, schematic diagram shows the various constructs of GST-Egr-1 fusion proteins. F1, C-2, C-1, and N-1 are the designated names of the fusion proteins with their predicted molecular masses indicated on the right. The shaded area indicates the DNA binding region in Egr-1. The amino acid positions (29, 171, 175, 322, and 533) in the protein fragments are also indicated below these constructs. D, GST-Egr-1 fusion proteins were expressed and purified from E. coli by using glutathione-Sepharose beads. The fusion proteins were separated by 10% SDS-PAGE, and the gel was stained with Coomassie Blue R-250 and dried. The designated name of each fusion protein is indicated on top of each lane and the molecular mass markers are shown on left of the gel. E, autoradiograph shows the autophosphorylated ␤-subunit of CKII bound to various GST-Egr-1 fusion proteins. Various constructs of GST-Egr-1 fusion proteins (F-1, N-1, C-1, and C-2) and GST alone attached to glutathione-Sepharose beads were mixed with [␥-32 P]ATP-labeled CKII. The beads were washed, and the 32 P-labeled ␤-subunit bound to the fusion proteins on beads was eluted by boiling the beads in Laemmli buffer and resolved by 12.5% SDS-PAGE as described under "Experimental Procedures." The gel was fixed, dried, and exposed to x-ray film. The lane labeled CKII represents the autophosphorylated CKII ␤-subunit (as indicated by ␤) without incubating with any fusion proteins. Lanes labeled GST, F-1, N-1, C-1, and C-2 indicate the 32 P-labeled CKII that binds to the respective fusion proteins. tion on Egr-1 is competitively inhibited by the CKII specific substrate peptide (lanes 6 -9) as well as by FX, a peptide from Egr-1 containing an SXD consensus sequence (lanes 10 -13) suggesting that Egr-1 phosphorylation by CKII is specific.
Mapping the CKII Phosphorylation Sites-To determine the CKII phosphorylation site(s) on Egr-1, a CNBr peptide map was generated. Theoretically, CNBr treatment of Egr-1 is predicted to produce six large peptides, separable on a polyacrylamide gel, as well as five small peptides lacking in kinase phosphorylation sites (Fig. 4A). The DNA fragments corresponding to the six large CNBr peptides (numbered 5-10) were first amplified by polymerase chain reaction. The amplified DNA was then cloned in vectors containing a SP6, T7, or T3 promoter. The [ 35 S]methionine-labeled CNBr peptides were then synthesized in vitro with a rabbit reticulocyte lysate system and resolved by 16.5% SDS-PAGE. The peptides thus produced were in close agreement to their predicted molecular mass indicated in Fig. 4A (data not shown).
In parallel to the above experiment, F-1 was phosphorylated by CKII, digested by CNBr, and resolved by 16.5% SDS-PAGE. Fragment 10 was strongly phosphorylated and fragment 7 moderately phosphorylated as compared to the other fragments (Fig. 4B, lane 1, 2-h exposure). The phosphorylation on fragments 5 and 8 was also observed on longer exposure of the autoradiograph (Fig. 4B, lane 2, 6-h exposure), whereas fragment 6 is detected only on overnight exposure (not shown). Some of the high molecular mass bands as shown by asterisks in Fig. 4B are due to incomplete CNBr digestion of Egr-1. CKII phosphorylation sites on Egr-1 in vitro were compared with those found in vivo from serum-induced, 32 P-labeled NIH 3T3 cells. Egr-1 was immunoprecipitated from these cells and digested with CNBr. Fragments 5 and 10 were strongly phosphorylated, whereas fragment 7 was moderately phosphorylated in vivo (Fig. 4B, lane 3). Phosphorylation of fragments 6 and 8 was also observed only on longer exposure of the gel. Hence, seruminduced phosphorylation of fragments 5, 7, and 10 could be due to endogenous CKII, but the differential levels of phosphorylation on these fragments suggest that CKII is not the only kinase that phosphorylates Egr-1 in vivo.
To further delineate the phosphorylation sites, CKII in vitro phosphorylated Egr-1 was digested with trypsin. Small peptides were separated on HPLC, and the fractions containing 32 P-labeled peptides were subjected to microsequencing. The results indicate that these trypsin-generated peptides are parts of fragments 5, 6, 7, and 8 (see Fig. 4A (c)). The labeled peptides contain CKII sites (34) which are S 101 XXD in fragment 5, T 145 XXXE in fragment 6, S 194 XD in fragment 7, and S 299 XD in fragment 8, all of which are potential sites for CKII phosphorylation. There are three CKII recognition sites (S 376 XXD, T 389 XE, and T 516 XXXD) in fragment 10. T 389 and T 516 are unlikely to be responsible for the strong phosphorylation of fragment 10 as phosphorylation on threonine is much weaker than that on serine. Thus S 376 is likely to be the major phosphorylation site by CKII within fragment 10. The reason S 376 XXD could not be detected by this method, despite heavy phosphorylation in fragment 10, is because trypsin digestion   FIG. 2. CKII phosphorylation of Egr-1 in vitro. A, the GST-Egr-1 fusion protein F-1 (lane 1) or GST alone (lane 2) attached to glutathione-Sepharose beads (50 l) was phosphorylated with CKII in the presence of [␥-32 P]ATP. Samples were resolved by 10% SDS-PAGE, dried, and exposed to x-ray film, and the resulting autoradiograph is shown. The GST-Egr-1 protein band is indicated as EGR-1. B, phosphoamino acid analysis of CKII-phosphorylated Egr-1. Egr-1 was phosphorylated with CKII and separated by a 7.5% SDS gel which was blotted onto a PVDF membrane and exposed to x-ray film. The 32 P-labeled Egr-1 band was excised from the membrane and subjected to two-dimensional phosphoamino acid analysis on thin layer chromatography plates as described under "Experimental Procedures." The plates were exposed to x-ray film, and in the resulting autoradiograph the position of amino acid standards are indicated by dotted lines and labeled by the one-letter amino acid code. C, phosphoamino acid analysis of in vivo phosphorylated Egr-1. Confluent NIH 3T3 cells were labeled with 32 P-orthophosphate for 4 h and induced with platelet-derived growth factor (75 ng/ml) for 2 h. The cell extract was immunoprecipitated by an anti-Egr-1 polyclonal antibody, and the labeled Egr-1 was subjected to phosphoamino acid analysis as described in B.

FIG. 3. Phosphorylation of Egr-1 by CKII in the presence of CKII substrate peptides.
A, Egr-1 fusion protein (F-1) attached to glutathione-Sepharose beads was incubated with increasing amounts of CKII in the absence of peptide (lanes 1-5), or with CKII (2 units) and increasing amounts of CKII specific substrate (RRREEETEEE) peptide (lanes 6 -9) or with FX (IKAFATQSGSQDLKALNT) peptide (lanes 10 -13) as shown on top of autoradiograph. The samples were separated by 7.5% SDS-PAGE and the gel was fixed, dried, and exposed to x-ray film. B, the autoradiograph obtained from the above experiment was scanned on a densitometer and the relative densities were plotted. The numbers at the x axis corresponds to the lanes in panel A. resulted in a peptide around S 376 of only four amino acids (NFS 376 R) which are not resolved by HPLC. We further confirmed phosphorylation of fragments 6, 8, and 10 by sequencing these peptides from excised gel sections (data not shown). Identification of the exact sites of CKII phosphorylation would require a site-directed mutational approach.
Effect of CKII Phosphorylation on Egr-1 DNA Binding Activity-The three zinc fingers of Egr-1 are located from amino acids 332-416. As fragment 10 contains the second and third zinc fingers, it is likely that phosphorylation on this fragment alters its DNA binding activity. To study the effect of CKII phosphorylation on Egr-1 DNA binding activity, GST-Egr-1 fusion protein (F-1) was phosphorylated with CKII prior to carrying out the DNA binding reaction in a gel mobility shift assay. The phosphorylation of Egr-1 by CKII resulted in a decrease in its DNA binding activity in a stoichiometric manner (Fig. 5, lanes 2-6). This decrease in DNA binding activity can be blocked competitively in a dose-dependent manner by the increasing concentrations of either CKII specific substrate peptide (Fig. 5, lanes 7-10) or by the FX peptide which has SXD as the CKII phosphorylation site (lanes 11-14).
CKII Affects the Transcriptional Activity of Egr-1 in NIH 3T3 Cells-In order to study the effect of CKII phosphorylation on Egr-1 transcriptional activity in vivo, NIH 3T3 cells were cotransfected with a chloramphenicol acetyltransferase reporter plasmid comprised of three Egr-1 binding sites, an Egr-1 expression plasmid, pCMV-Egr-1, and a CKII expression plasmid, pCMVCKII (containing the CKII ␣-subunit).
The CKII activity in untransfected and transfected cells was determined to ensure that the pCMVCKII plasmid produced active CKII in NIH 3T3 cells. A 1.6-fold increase in CKII activity was observed on transfecting cells with 2 g of the pCMVCKII ␣-subunit (data not shown), which was in close agreement with previously reported results in COS-1 cells (32).
Transfection of cells with increasing concentrations of pCMV-Egr-1 results in an increase in transcriptional activity reflected by an increase in chloramphenicol acetyltransferase activity (Fig. 6A, top panel). However, when the increasing concentrations (1, 2, and 5 g) of CKII plasmids were cotransfected with egr-1, a corresponding decrease in egr-1 transcriptional activity was observed (Fig. 6A, lower panels).
Nuclear extracts were prepared from cells transfected with or without pCMVCKII in the presence of pCMV-Egr-1 and gel mobility shift analysis was performed using the Egr-1 binding site as a probe. A decrease in DNA binding activity was found when CKII was co-transfected with pCMV-Egr-1 (Fig. 6B,  panel i, lane 2). Western analysis of these extracts showed that there was no change in the amount of Egr-1 protein present in each sample (Fig. 6B, panel ii). Thus, in the presence of the CKII plasmid there is a decrease in the DNA binding ability of Egr-1 in accordance with the in vitro mobility shift data (Fig. 5).
To show that this decrease is a result of phosphorylation of the Egr-1 protein by CKII, transfected cells were labeled with [ 32 P]orthophosphate. Egr-1 was immunoprecipitated with Egr-1 antibody, resolved by 7.5% SDS-PAGE, and transferred to a nitrocellulose membrane, and autoradiography was performed. A 2-fold increase in Egr-1 phosphorylation was observed when cells were co-transfected with pCMV-Egr-1 in the presence of pCMVCKII as compared to that in the absence of pCMVCKII (Fig. 6C, panel i). Western blot analysis was performed on the same nitrocellulose membrane to confirm that there was no difference in the amount of Egr-1 protein (Fig. 6C, panel ii).

DISCUSSION
To activate or repress transcription, transcription factors must locate to the nucleus, bind DNA, and interact with the basal transcription apparatus (35). Accordingly, extracellular signals that regulate transcription factor activity may affect one or more of these processes. Most commonly regulation is achieved by phosphorylation of the transcription factor(s) which in turn modulate either its trans-activating activity or DNA binding activity (36). Although Egr-1, a transcription factor, has been shown to play an important role in proliferating and differentiating cells, little is known of its function in relation to its phosphorylation status. In this report we show that Egr-1 is phosphorylated by the protein kinase CKII and that this has a negative effect on its DNA binding and transcription activities.
Although the CKII phosphorylation pattern of Egr-1 does not exactly coincide with that of serum-induced Egr-1, it is possible that CKII is one of several kinases which phosphorylate and regulate Egr-1 function. As shown in Fig. 4A, the CKII phosphorylation sites on Egr-1 cover the regions of the protein important for its transcriptional activation (fragment 7), repressor binding (fragment 8), nuclear import, and DNA binding (fragment 10). The CKII sites present on these fragments are of S/TXD/E, S/TXXD/E, and S/TXXXD/E types (34). The observation that CKII attenuates Egr-1 DNA binding activity appears to be contradictory to earlier reports that Egr-1 DNA binding activity (25) is elevated after cell stimulation with growth factors. This could be due to a change in conformation arising from phosphorylation by other kinases beside CKII after serum induction. Furthermore, strong phosphorylation in the activation domain (fragment 5) of Egr-1 in vivo after serum induction might lead to an increase in its transcriptional activity.
It has been reported that the CKII ␣-subunit alone is sufficient to increase CKII activity in vivo for several reasons. First, the ␣-subunit alone is moderately active without the ␤ subunit (37,38). Second, the ␣-subunit in nuclei is not bound to ␤ but is complexed to other nuclear components (39). Third, the ␤-subunit is synthesized in large excess over the ␣-subunit and only a small fraction of it contributes to the tetrameric holoenzyme complex (40). In our experiments, co-transfection of pCMV-CKII␣ and pCMV-Egr-1 results in a decrease in transcriptional activity in vivo which is possibly due to a decrease in DNA binding activity caused by CKII phosphorylation of the DNA binding fragment.
Besides phosphorylating Egr-1, the association of CKII and Egr-1 could also induce further protein-protein interactions. As CKII is a tetrameric protein, it is possible that by anchoring different proteins to its four subunits it could promote interactions between diverse proteins. DNA topoisomerase II, which associates with CKII, is believed to participate in DNA transcription (as well as replication and recombination) as part of a multienzyme complex (41). CKII could thus be the building block for such a multienzyme complex which might possibly include a transcription factor. On the other hand, CKII has also been found to associate with the proteasome complex resulting in its degradation (42). Hence, CKII association with Egr-1 could be a prelude to degradation. This would correlate well with the fact that CKII phosphorylation of Egr-1 leads to a decrease in DNA binding and transcriptional activities. It is thus possible that Egr-1 transcriptional activity might be affected via its link with CKII by other CKII associated proteins.
Our data show that CKII associates with Egr-1 and regulates its DNA binding and transcriptional activities by phosphorylation. Egr-1 is not the only DNA-binding protein that is inhibited by CKII. Other such proteins are c-Jun (19), c-Myb (43), and serum response factor (44). It is now necessary to study Egr-1 regulation in the context of CKII. A map of CNBr fragments of Egr-1 has been generated and the fragments phosphorylated by CKII have been identified. This will facilitate not only the further study of Egr-1 phosphorylation by CKII as well as other kinases but also aid in determining the phosphorylated residues on Egr-1 and lead to a better understanding of Egr-1 function in regulating cell growth and differentiation.
Acknowledgments-We are grateful to Dr. Steven Sloan for the gift of the pCMVCKII plasmid and to Amgen Inc. for the Egr-1 monoclonal antibody. We thank Drs. Catherine J. Pallen and Kee Chuan Goh for critical reading of the manuscript and Francis Leong and Sock Yng Oh for photography.
FIG. 6. CKII phosphorylation of Egr-1 affects its transcriptional activity. A, expression of the CKII ␣-subunit reduces Egr-1 transcriptional activity. Chloramphenicol acetyltransferase assays were performed on NIH 3T3 cells transfected with increasing concentrations (shown at the bottom) of pCMV-Egr-1 alone (top panel) or with increasing concentrations of pCMV-Egr-1 in the presence of pCMVCKII ␣-subunit (1, 2, and 5 g, shown at the left of the panels). TLC plates were exposed to x-ray film and the resulting autoradiograph is shown. B, a decrease in transcription is related to a decrease in DNA binding activity. Cells were transfected with (10 g) of pCMV-Egr-1 with (5 g) or without pCMVCKII. Nuclear extracts were prepared from the transfected cells, subjected to gel mobility shift analysis as described under "Experimental Procedures," and are shown in an autoradiograph (panel i). The same amount (10 g) of nuclear protein used for gel mobility shift analysis was separated by SDS-PAGE and subjected to Western blot analysis with an anti-Egr-1 antibody, and the signal was detected by enhanced chemiluminescence (panel ii). C, increased phosphorylation of transfected egr-1 gene product by co-transfection with CKII expression plasmid. Panel i, NIH 3T3 cells were transfected with pCMV-Egr-1 in the presence (5 g) or absence (0 g) of pCMVCKII and labeled with [ 32 P]orthophosphate. Cells were lysed with radioimmune precipitation buffer and immunoprecipitated by an anti-Egr-1 antibody (1133). The precipitates were separated by 7.5% SDS-PAGE, transferred to a nitrocellulose membrane, and exposed to x-ray film, and the resulting autoradiograph is shown. Panel ii, Western blot of the same membrane probed with an anti-Egr-1 antibody, the location of which was revealed by enhanced chemiluminescence.