- Email: [email protected]
Phosphorylation of two eukaryotic transcription factors, Jun dimerization protein 2 and activation transcription factor 2, in Escherichia coli by Jun N-terminal kinase 1
Available online at www.sciencedirect.com
ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 376 (2008) 115–121 www.elsevier.com/locate/yabio
Phosphorylation of two eukaryotic transcription factors, Jun dimerization protein 2 and activation transcription factor 2, in Escherichia coli by Jun N-terminal kinase 1 Takehide Murata a,*, Yoriko Shinozuka a,1, Yuichi Obata b, Kazunari K. Yokoyama a b
a Gene Engineering Division, RIKEN BioResource Center, 3-1-1 Koyadai, Tsukuba, Ibaraki 305-0074, Japan Department of Biological Systems, RIKEN BioResource Center, 3-1-1 Koyadai, Tsukuba, Ibaraki 305-0074, Japan
Received 30 November 2007 Available online 6 February 2008
Abstract Recombinant eukaryotic proteins are frequently produced in Escherichia coli and such proteins are often used for biochemical studies in vitro. However, proteins produced in this way are not modified chemically, for example, by phosphorylation, acetylation, methylation, sumoylation, or ubiquitination, during their synthesis in bacterial cells. We constructed vectors for expression in E. coli of human Jun Nterminal kinase 1 (JNK1), mouse Aurora kinase B (Aurkb), and the histone acetyltransferase (HAT) domain of P/CAF. These expression vectors included the origin of replication of p15A and the origin of replication of pBR322 or ColE1. Using these expression vectors in E. coli, we were able to phosphorylate mouse and human Jun dimerization protein 2 (JDP2) and human activation transcription factor 2 (ATF-2) by the action of human JNK1 that was expressed simultaneously. Moreover, the tail region of mouse histone H3 was phosphorylated and acetylated, respectively, by Aurkb and by the HAT domain of P/CAF. We also observed that the interaction of ATF-2 with JDP2 was prevented when ATF-2 was phosphorylated. Our expression systems for production of enzyme-modified proteins in E. coli should be widely applicable and useful for biochemical studies of chemically modified eukaryotic proteins in vitro. Ó 2008 Elsevier Inc. All rights reserved. Keywords: Protein phosphorylation; Protein modification; Recombinant protein
Production of large amounts of recombinant proteins in Escherichia coli is a powerful tool for biochemical studies of the functions of such proteins and eukaryotic proteins are frequently produced in E. coli. However, the recombinant proteins produced in E. coli are not modified chemically by intracellular reactions, such as phosphorylation, acetylation, methylation, sumoylation, or ubiquitination, because of, in general, the absence of the machinery for the posttranslational modification of proteins in E. coli. Unmodified proteins allow easy and straightforward purification but experiments with unmodified proteins cannot *
Corresponding author. Fax: +81 29 836 9120. E-mail address: [email protected] (T. Murata). 1 Present address: Department of Immunology and Cell Biology, Graduate School of Medicine, Kyoto University, Yoshidakonoe-cho, Sakyo-ku, Kyoto 606-8501, Japan. 0003-2697/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2008.01.038
be expected to reveal full details of the native functions of these proteins in vivo. Production of eukaryotic proteins at large amounts has been achieved with viral vectors, such as baculovirus, retrovirus, and adenovirus [1,2]. However, production of eukaryotic proteins in E. coli is simpler in terms of both construction of the vector and purification of the protein of interest. It would, thus, be useful to develop expression systems for the large-scale production of eukaryotic proteins in E. coli in their chemically modified forms. The phosphorylation of transcription factors plays key roles in the regulation of their transcriptional activities [3]. Several groups have demonstrated the roles of the phosphorylation of transcription factors, in particular of ATF-2, in the regulation of transcriptional activity [4–6]. Phosphorylation of ATF-2 was reported after treatment of cells with stress-inducing reagents and UV irradiation.
116
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
A mutant from of ATF-2, with alanine residues instead of residues Thr69 and Thr71, is stable and is able to bind to the cyclic AMP response element (CRE).2 Moreover, appropriate homodimerization or heterodimerization with other leucine zipper proteins appears to occur normally. However, the mutant protein is no longer able to participate in the activation of CRE-dependent transcription and in the normal development of liver and heart when the homolog of ATF-2, ATF-7, is simultaneously deleted [7]. Furthermore, the interaction between ATF-2 and p300 (or related) proteins in vivo does not require phosphorylation of ATF-2 at Thr69 and Thr71 [8]. In addition to sites for phosphorylation, there are acetylation sites in ATF-2 at positions 357 and 374 that might potentially contribute to the ATF-2-mediated activation of transcription [9]. Significant enhancement of acetylation of ATF-2 does, however, not occur upon treatment of cells with so-called stress reagents, such as tumor necrosis factor-a, phorbol 12-myristate 13-acetate, or cycloheximide. By contrast, phosphorylation of ATF-2 is enhanced by similar reagents, suggesting that the phosphorylation reactions catalyzed by MAP kinases are not correlated with acetylation of lysine residues [9]. The protein kinase ATM also phosphorylates the serine residues of ATF-2 at positions 490 and 498 during the response of cells to DNA damage [10]. In a previous report, we demonstrated the interaction between ATF-2 and JDP2, both in vitro and in vivo, via their bZIP domain [11]. We also demonstrated the repression of transcription by JDP2, showing that JDP2 bound to the CRE to repress the CRE-dependent transcription that is mediated by ATF-2 in embryonal carcinoma (EC) F9 cells. ATF-2 and JDP2 bound to the differentiation response element (DRE) in the promoter of the gene for JUN in untreated EC F9 cells. After treatment of EC F9 cells with retinoic acid, JDP2 was released from the DRE and, thus, transcription of JUN was activated [12]. JDP2 has also been reported to be a repressor of the function of the JUN homodimer and the JUN/FOS heterodimer [13]. JDP2 recruits histone deacetylase 3 to the promoter of the gene for JUN and to the gene that encodes the CCAAT/enhancer-binding protein d and transcription of both gene is inhibited [12,14]. We and others have observed the immediate phosphorylation on JDP2 on Thr148 after exposure of cells to UV light (unpublished results) [15,16]. However, it is not yet clear whether JDP2 is activated by such phosphorylation. The details of the physiological roles of the phosphorylation of ATF-2 and JDP2 remain unknown and the mechanisms of transcriptional activation by phosphorylated ATF-2 and JDP2 remain unclarified. 2 Abbreviations used: CRE, cyclic AMP response element; EC, embryonal carcinoma; DRE, differentiation response element; PMSF, phenylmethylsulfonyl fluoride; GST, glutathione S-transferase; IPTG, isopropylb-D-thiogalactopyroside; PBS, phosphate-buffered saline; CIAP, calf intestinal alkaline phosphatase; MALDI, matrix-assited laser desorption ionization.
In this report, we describe an expression system for the production of phosphorylated proteins in E. coli, which should be very useful for studies of the biochemical functions of modified proteins in vitro. We also provide results of an initial study that demonstrate that binding of ATF-2 and JDP2 is affected by the phosophorylation of ATF-2. Materials and methods Plasmids Mouse JDP2-expression plasmids [12,17] derived from pGEX (Amersham Pharmacia Biotech Co., Piscataway, NJ, USA) were obtained from the RIKEN DNA Bank, RIKEN BioResource Center (Tsukuba, Ibaraki, Japan). pGST-mmJDP2_full (RDB 4757; RIKEN DNA Bank), including a gene for full-length murine JDP2, and pGSTmmJDP2_T148A (RDB 4778; RIKEN DNA Bank), with a mutation of Thr to Ala at position 148 of the encoded mouse JDP2, were prepared as described [17]. Expression vectors for His-tagged human JDP2 (pET_His-hsJDP2) and His-FLAG-tagged human ATF-2 (pET_HisFLAGhsATF-2) were also obtained from the RIKEN DNA Bank. The cDNA for human Jun N-terminal kinase 1 (JNK1; NCBI Accession No. NM_002750) and for the HAT domain of human P/CAF cDNA (NCBI Accession No. NM_003884) were amplified by reverse transcription polymerase chain reaction (RT-PCR) from first-strand cDNA prepared from HEK293 cells (RIKEN Cell Bank, Ibaraki, Japan) using JNK1-specific primers (forward 50 - AGC AGA AGC AAG CGT GAC AAC-30 ; reverse 50 - TCA CTG CTG CAC CTG TGC TAA-30 ) and P/CAF-specific primers (forward 50 -CTA GAA GAA GAA GTA TAT AGT C-30 ; reverse 50 -TCA CTT GTC AAT TAA TCC AGC-30 ). The amplified fragments of JNK1- and P/CAF-encoding DNA were inserted into the pET15b plasmid (Novagen, Merck KGaA, Darmstadt, Germany) to yield pEThsJNK1 and pEThsPCAF-HAT, respectively. T7 expression units of pEThsJNK1 and pEThsPCAF-HAT were isolated with BglII and HindIII and then inserted into the BamHI-HindIII site of pACYC184 [18] (from the Cloning Vector Collection, National Institute of Genetics, Mishima, Japan) to yield pT7Cam_hsJNK1 and pT7Cam-hsPCAF_HAT, respectively. The cDNA for mouse Aurora kinase B (Aurkb; NCBI Accession No. NM_011496) was prepared by PCR from the NIA/NIH mouse 15K cDNA clone H3053G07 [19] with the following primers: forward 50 -CTA AAG GGC AGA GGG AGG CAG A-30 ; reverse 50 -CCA TGG CTC AGA AGG AGA ACG CCT-30 . The resultant product of PCR was cloned into the SmaI site of pGEM3zf(+) (Promega, Madison, WI, USA). The NcoISalI DNA fragment [1040-basepair (bp)] that encoded the entire coding region of Aurkb was then introduced into the NcoI-XhoI site of pT7Tet1 (RDB 6211, RIKEN DNA Bank), which is a derivative of pACYC184 [18], to generate
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
pT7Tet-mmAurkb. pACYC184 contains the origin of replication of p15A [18], which allows pACYC184 to coexist in E. coli cells with plasmids in the ColE1 compatibility group (e.g., pBR322 and pUC19). Thus, we were able to coexpress proteins from two vectors in E. coli [20–22]. The cDNA for mouse histone H3.2 (Genbank Accession No. BC106177) was prepared by RT-PCR from a cDNA library of the mouse embryonic stem cell line E14TG2a, with the following primers: forward 50 -AGA ATT CGG CTC GTA CTA AGC AGA CCG CT-30 ; reverse 50 -ATT ACG CCC TCT CCC CGC GGA T-30 . The product of PCR was cloned into the SmaI site of pGEX-4T-3 (Amersham Pharmacia Biotech). The resultant plasmid was digested with SalI and then self-ligated to yield pGEXmmH3_tail, which encoded amino acids 2 through 57. All the plasmids constructed in this study will be available from the RIKEN DNA Bank. Preparation of a lysate of E. coli and purification of GSTtagged JDP2 Transformed E. coli BL21(DE3) cells (Novagen) were cultured in 40 ml of Luria-Bertani medium that contained 8 mM MgSO4 and appropriate antibiotics (50 lg/ml ampicillin, 12.5 lg/ml tetracyclin, and 12.5 lg/ml chloramphenicol) with shaking at 160 rpm until the absorbance at 600 nm (OD600) reached 1.0. Then production of recombinant proteins was induced by addition of isopropyl-b-Dthiogalactopyranoside (IPTG) to a final concentration of 1.0 mM. After incubation of cells for another 4 h, cells were pelleted by centrifugation and stored at 80 °C until use. Thawed cells were lyzed on ice by sonication in 600 ll of phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 1.5 mM KH2PO4, 8.0 mM Na2HPO4) that contained 1 mM phenylmethylsulfonyl fluoride (PMSF; Sigma Biochemical Co., St. Louis, MO, USA), 1 lg/ml pepstatin A (Sigma Biochemical), and 1 lg/ml leupeptin (Peptide Institute Inc., Minoh, Osaka, Japan). Glutathione–S-transferase-tagged (GST-tagged) proteins were isolated from the lysate by affinity chromatography on glutathione-Sepharose 4B (Amersham Pharmacia Biotech) according to the manufacturer’s instructions. Phosphatase treatment Purified GST-tagged JDP2 (0.5 lg) was incubated in 17 ll of reaction buffer (2 mM MgCl2, 1 mM dithiothreitol, 1 mM PMSF, 1 lg/ml pepstatin A, 1 lg/ml leupeptin, 20 mM Tris–Cl, pH 7.5) that contained 3.0 units of calf intestinal alkaline phosphatase (CIAP; Takara Bio Inc., Otsu, Shiga, Japan) at 37 °C for 30 min. For the analysis of modified JDP2 in vitro, purified modified GST-JDP2 proteins were separated by SDS–PAGE and eluted from the gel. Further analysis, including matrixassisted laser desorption ionization (MALDI) and mass spectrometry were performed by Shimadzu Biotech (Kyoto, Japan).
117
Antibodies Rabbit polyclonal antiserum specific for phosphorylated JDP2 was raised by Medical & Biological Laboratories Co., Ltd. (Nagoya, Aichi, Japan) against the phosphorylated oligopeptide NH2-CIVRTDSVR(pT)PESEG-COOH in which the Thr residue corresponding to residue 148 of mouse JDP2 [15] was phosphorylated. Excess amounts of unphosphorylated JDP2 are also detected by Western blotting with the phospho JDP2-specific antiserum. Monoclonal antibodies mAb 176 and 249, specific for mouse JDP2, were described previously [12]. Antibodies against GST conjugated with horseradish peroxidase (RPN-1236; Amersham Pharmacia Biotech), against His (H-15; sc803; Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), against ATF-2 (N-96; sc-6233; Santa Cruz Biotechnology), against phosphorylated ATF-2 (Thr71-R; sc7982R; Santa Cruz Biotechnology), against phosphorylated histone H3 (Ser10; 06-570; Upstate, Charlottesville, VA, USA), against acetylated histone H3 (06-599; Upstate), and against FLAG (M2; F-3165; Sigma) were purchased as indicated. Binding assays Two micrograms (as protein) of E. coli lysate (approx. 50 ll) from cells that expressed either His-JDP2 or HisFLAG-ATF-2 were mixed with 800 ll of binding buffer (PBS that contained 0.1% Nonidet-p40, 20% glycerol, and 1 mM PMSF; pH 7.4) and incubated at 4 °C for 4 h and then with 0.4 lg of antibody at 4 °C for 2 h. Immunocomplexes were recovered by the addition of protein GSepharose beads (Amersham Pharmacia Biotech) and incubation at 4 °C for 1 h. After the beads had been washed extensively with the binding buffer, bound proteins were eluted by the addition of 50 ll of 1 Laemmli sample buffer and incubation at 95 °C for 5 min. Results and discussion Phosphorylation of JDP2 It has been reported that JDP2 is phosphorylated by JNK1 and by p38 MAP kinase in vitro [15]. Therefore, we introduced pT7Cam_hsJNK1 and/or pGSTmmJDP2_full into E. coli BL21(DE3) cells. To examine the level of phosphorylation of a recombinant protein in E. coli, we subjected cell lysates, before and after induction by IPTG, to SDS–PAGE and then stained proteins with Coomassie brilliant blue to examine the total complement of soluble proteins (Fig. 1). We detected obvious bands of proteins with apparent molecular masses of 44.5 kDa (corresponding to human JNK1) and 47.0 kDa (GSTJDP2) in E. coli transformed with pT7Cam_hsJNK1 and with pGST-mmJDP2_full, respectively, after induction by IPTG (Fig. 1; compare lanes 1 and 2 and lanes 6 and 7). The lysates prepared from E. coli cells that had been trans-
118
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
Fig. 1. Production in E. coli of JNK1 and GST-JDP2 simultaneously. E. coli cells harboring pT7Cam_hsJNK1 alone (lanes 1 and 6), pGST-mmJDP2 alone (lanes 2, 4, 7, and 9), and both plasmids together (lanes 3, 5, 8, and 10) were cultured in the absence (lanes 1–5) and presence (lanes 6–10) of 1 mM IPTG for 4 h at 30 °C. The total lysate from each bacterial culture was analyzed by SDS–PAGE (8% acrylamide) and subsequent staining with Coomassie brilliant blue. Positions of GST-JDP2 and JNK1 are indicated.
formed with plasmids that encoded JNK1 and GST-JDP2, together yielded bands of proteins with apparent molecular masses of 44.5 kDa (human JNK1) and 47.0 kDa (GSTJDP2) (Fig. 1, lane 8). These results indicated that both recombinant proteins were produced simultaneously and efficiently in E. coli under our experimental conditions. To examine the extent of phosphorylation of JDP2 by JNK1 in E. coli, we performed Western blotting after treatment of affinity-purified GST-fused JDP2 with CIAP. A 47.0-kDa protein appeared on the gel when bacteria was transformed with pGST-mmJDP2 alone, and we then assumed that this protein was the unphosphorylated form of GST-JDP2 (Fig. 2, lane 1). We observed two
bands of proteins of approximately 47 kDa when bacteria had been transformed with pGST-mmJDP2_full plus pT7Cam_hsJNK1 (Fig. 2, lane 3). Treatment with CIAP reduced the intensity of the more slowly migrating band (lane 4). Furthermore, no slowly migrating form of JDP2 was detected in the lysate of bacteria that had been transformed with both pT7Cam_hsJNK1 and pGSTmmJDP2_T148A together (Fig. 3, lanes 9–12). Because the Thr residue at position 148 of JDP2 is known to be a site of phosphorylation by either JNK1 or p38 MAP kinase [15], we postulated that the more slowly migrating band corresponded to the phosphorylated form of JDP2.
Fig. 2. E. coli cells harboring pGST-mmJDP2_full alone (lanes 1, 2, 5, and 6) and with pT7Cam_hsJNK1 (lanes 3, 4, 7, and 8) were cultured in the presence of 1 mM IPTG for 4 h at 30 °C. The total lysate from each bacterial culture was incubated with glutathione-Sepharose beads. Proteins that had bound to the beads were treated with CIAP and then analyzed by SDS–PAGE (8% acrylamide) and subsequent Western blotting with monoclonal JDP2specific antibody (mAb249; see [12]) (A) and polyclonal antiserum against phosphorylated JDP2 (see text for details) (B). Positions of phosphorylated (GST-pJDP2) and unphosphorylated (GST-JDP2) JDP2 are indicated.
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
119
position that corresponded to residue 148 of JDP2 (designated here as antiserum against phosphorylated JDP2), as described under Materials and methods. With this antiserum, a strong signal was generated by the more slowly migrating band of JDP2 (Fig. 2, lane 7). Treatment of JDP2 protein with CIAP reduced the intensity of the reaction of the more slowly migrating band with the antiserum against phosphorylated JDP2 (Fig. 2, lane 8). Recombinant JDP2 accumulated in E. coli during a 4-h incubation. Similarly, phosphorylated GST-JDP2 accumulated over the same period of time (Fig. 3). Modification of various proteins Fig. 3. Accumulation in E. coli of GST-JDP2 and its phosphorylated form. E. coli cells harboring pGST-mmJDP2_full alone (lanes 1–4), pGST-mmJDP2_full plus pT7Cam_hsJNK1 (lanes 5–8), and pGSTmmJDP2_T148A plus pT7Cam_hsJNK1 (lanes 9–12) were cultured in the presence of 1 mM IPTG for the indicated times. The total lysate from each bacterial culture was analyzed by SDS–PAGE (8% acrylamide) and subsequent Western blotting, as described in the legend to Fig. 2. (A) mAb249 [12]. (B) Antibodies against phosphorylated JDP2. These antibodies recognized unphosphorylated JDP2 when an excess of JDP2 was loaded on the gel, as noted in the text.
To confirm the phosphorylation status of JDP2, we performed mass spectrometric analysis of the isolated JDP2. We extracted JDP2 from the rapidly and slowly migrating bands and digested the protein with trypsin. Examination of the MALDI time-of-flight spectra of the digested peptides from the slowly migrating band indicated that they were phosphorylated peptides derived from JDP2 (data not shown). Our results led us to conclude that the more slowly migrating band from the bacteria that had been transformed with pT7Cam_hsJNK1 plus pGSTmmJDP2_full was the phosphorylated form of mouse JDP2 and that mouse JDP2 had been phosphorylated in E. coli cells in the presence of pT7Cam_hsJNK1. We generated rabbit antiserum against an oligopeptide of mouse JDP2 with a phosphorylated Thr residue at the
We also generated the phosphorylated form of Histagged human JDP2 in E. coli. We detected phosphorylated JDP2 with the abovementioned antiserum against phosphorylated JDP2 (Fig. 4A). Phosphorylation of HisFLAG-tagged ATF-2 in a lysate of E. coli cells in which the JNK1 expression vector had been introduced simultaneously was also examined with antibodies against phosphorylated ATF-2 (Fig. 4B). The change in mobility of phosphorylated JDP2 during SDS–PAGE was not significant in the presence of the JNK1 expression vector (Fig. 4A, lanes 1 and 2). However, the phosphorylated form of ATF-2 was clearly detected with the antibodies against phosphorylated ATF-2 (Fig. 4B, lanes 5 and 6). We also examined modification of the GST-tagged tail region of histone H3 (GST-H3 tail) in E. coli using expression vectors derived from pACYC184. The GST-H3 tail was apparently phosphorylated at the Ser residue at position 10 (Ser10; Fig. 5B) when the GST-H3 tail expression vector was introduced into E. coli with the pT7Tet_ mmAurkb expression vector for mouse Aurora kinase B, a kinase that catalyzes phosphorylation at Ser10 of histone H3 [23]. In addition, the GST-H3 tail was acetylated in E. coli cells in the presence of pT7Cam_hsPCAF_HAT, which encoded the HAT domain of P/CAF (Fig. 5C), an
Fig. 4. Western blotting of JDP2 (A) and ATF-2 (B). E. coli cells harboring pET_His-hsJDP2 (lanes 1–4) and pET_HisFLAG-hsATF-2 (lanes 5–8) were cultured without (lanes 1, 3, 5, 7) or with (lanes 2, 4, 6, 8) pT7Cam_hsJNK1 in the presence of 1 mM IPTG for 4 h at 37 °C. The total lysate from each bacterial culture was analyzed by SDS–PAGE (A, 11% acrylamide; B, 8% acrylamide) and subsequent Western blotting, as described in the legend to Fig. 2. (A) lanes 1 and 2, JDP2-specific monoclonal antibody 249; lanes 3 and 4, antibodies against phosphorylated JDP2. (B) Lanes 5 and 6, antibodies against ATF-2; lanes 7 and 8, antibodies against phosphorylated-ATF-2.
120
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
Fig. 5. Western blotting for detection of the modification of the tail of histone H3. E. coli cells harboring pGEX-mmH3_tail alone (lanes 1–3), and with pT7Tet-mmAurkb (lanes 4–6) or with pEThsPCAF-HAT (lanes 7–9) were cultured in the presence of 1 mM IPTG for 4 h at 37 °C. The total lysate from each bacterial culture was analyzed by SDS–PAGE (10% acrylamide) and subsequent Western blotting, as described in the legend to Fig. 2. (A) Lanes 1–3, Antibodies against GST; (B) lanes 4–6, antibodies against phosphorylated histone H3; (C) lanes 7–9, antibodies against acetylated histone H3. The nucleotide sequence of pGEX-mmH3_tail was confirmed (data not shown) but, nonetheless, two bands of GST-H3 tail were evident on the gel. Aurkb, Aurora kinase B.
acetyltransferase specific for histone H3 [23]. Taken together, our results indicate that our expression systems are very useful not only for the phosphorylation of bZIP proteins, such as JDP2 and ATF-2, by JNK1 but also for phosphorylation and acetylation by other enzymes. Binding JDP2 and ATF-2 Our expression system in E. coli allows both ATF-2 and JDP2 to be phosphorylated. In an attempt to evaluate the properties of these newly generated proteins, we investigated the molecular interaction between them. Examining
the association of JDP2 with ATF-2, we found that HisFLAG-ATF-2 in a crude lysate of E. coli cells was efficiently immunoprecipitated by JDP2-specific monoclonal antibodies and His-JDP2 was efficiently immunoprecipitated by FLAG-specific (and, in this case ATF-2-specific) antibodies (Fig. 6, lanes 1 and 7), as reported previously [11]. To clarify the role of phosphorylation in the regulation of transcription, we compared the capacity of the phosphorylated and unphosphorylated forms of ATF-2 and JDP2 to associate with one another. When we compared the relative intensities of bands of either phosphorylated or unphosphorylated ATF-2 in immunocomplexes
Fig. 6. Coimmunoprecipitation (IP) with the monoclonal antibody against JDP2 176 (anti-JDP2; lanes 1–6) and with the monoclonal antibody against FLAG (anti-FLAG; lanes 7–12). Proteins that had bound to affinity beads were analyzed by SDS–PAGE (10% acrylamide) with subsequent Western blotting with His-specific antibody. The positions of ATF-2 and JDP2 are indicated. The band indicated by the asterisk corresponds to the antibody used for immunoprecipitation.
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
obtained with the JDP2-specific monoclonal antibody (Fig. 6, lanes 1–6), we noted that the intensity of the band of the phosphorylated form of ATF-2 was lower than that of unphosphorylated ATF-2 (compare lanes 1 and 2 and lanes 3 and 4, respectively). By contrast, when we compared levels of JDP2 in immunocomplexes formed with the FLAG-specific antibody (lanes 7–12), the intensities of the bands of both forms of JDP2 were similar (compare lanes 7 and 8 and lanes 9 and 10). These results suggest that, at least, phosphorylation of ATF-2 is critical for the interaction between ATF-2 and JDP2. Release of JDP2 from the ATF-2-JDP2 complex, possibly as a consequence of the phosphorylation of ATF-2, suggests a novel mechanism of interaction in the formation of heterodimers of ATF-2 and JDP2 in vitro. This report describes a compatible plasmid for overexpression of either a protein kinase or a histone acetyltransferase in E. coli, together with its respective substrate, for the simultaneous production of preparative amounts of modified recombinant protein. This approach for the production of modified proteins in E. coli is suitable for biochemical analyses of chemically modified proteins. Acknowledgments The authors thank Drs. K. Nakade, M. Kimura, and J. Pan for many helpful discussions and Ms. K. Takahashi and Ms. A. Kimura for diligent support. This work was supported by grants from the National BioResource Project of the Ministry of Education, Culture, Sports, Science and Technology, Japan and by the Bioresource Research Projects of RIKEN. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ab.2008. 01.038. References [1] C.L. Merrington, M.J. Bailey, R.D. Possee, Manipulation of baculovirus vectors, in: R. Rapley (Ed.), The Nucleic Acid Protocols Handbook, Humana Press Inc., Totowa, NJ, 2000, pp. 907–919. [2] N. Wu, M.M. Ataai, Production of viral vectors for gene therapy applications, Curr. Opin. Biotechnol. 11 (2000) 205–208. [3] A.H. Brivanlow, J.E. Darnell Jr., Signal transduction and the control of gene expression, Science 295 (2002) 813–818. [4] R. Davis, Signal transduction by the JNK group of MAP kinase, Cell 103 (2000) 239–252. [5] L. Chang, M. Karin, Mammalian MAP kinase signalling cascades, Nature 410 (2001) 37–40. [6] J.M. Kyriakis, J. Avruch, Mammalian mitogen-activated protein kinase signal transduction pathways activated by stress and inflammation, Physiol. Rev. 81 (2001) 807–869.
121
[7] W. Breitwieser, S. Lyons, A.M. Flenniken, G. Ashton, G. Bruder, M. Willington, G. Lacaud, V. Kouskoff, N. Jones, Feedback regulation of p38 activity via ATF2 is essential for survival of embryonic liver cells, Genes Dev. 21 (2007) 2069–2082. [8] M.C. Duyndam, H. van Dam, P.H. Smits, M. Verlaan, A.J. van der Eb, A. Zantema, The N-terminal transactivation domain of ATF2 is a target for the co-operative activation of the c-jun promoter by p300 and 12S E1A, Oncogene 18 (1999) 2311–2321. [9] B. Karanam, L. Wang, D. Wang, X. Liu, R. Marmorstein, R. Cotter, P.A. Cole, Multiple roles for acetylation in the interaction of p300 HAT with ATF-2, Biochemistry 46 (2007) 8207–8216. [10] A. Bhoumik, S. Takahashi, W. Breitweiser, Y. Shiloh, N. Jones, Z. Ronai, ATM-dependent phosphorylation of ATF2 is required for the DNA damage response, Mol. Cell 18 (2005) 577–587. [11] C. Jin, H. Ugai, J. Song, T. Murata, F. Nili, K. Sun, M. Horikoshi, K.K. Yokoyama, Identification of mouse Jun dimerization protein 2 as a novel repressor of ATF-2, FEBS Lett. 489 (2001) 34–41. [12] C. Jin, H. Li, T. Murata, K. Sun, M. Horikoshi, R. Chiu, K.K. Yokoyama, JDP2, a repressor of AP-1, recruits a histone deacetylase 3 complex to inhibit the retinoic acid-induced differentiation of F9 cells, Mol. Cell. Biol. 22 (2002) 4815–4826. [13] A. Aronheim, E. Zandi, H. Hennemann, S.J. Elledge, M. Karin, Isolation of an AP-1 repressor by a novel method for detecting protein-protein interactions, Mol. Cell. Biol. 17 (1997) 3094–3102. [14] K. Nakade, J. Pan, A. Yoshiki, H. Ugai, M. Kimura, B. Liu, H. Li, Y. Obata, M. Iwama, S. Itohara, T. Murata, K.K. Yokoyama, JDP2 suppresses adipocyte differentiation by regulating histone acetylation, Cell Death Differ. 14 (2007) 1398–1405. [15] S. Katz, R. Heinrich, A. Aronheim, The AP-1 repressor, JDP2, is a bona fide substrate for the c-Jun N-terminal kinase, FEBS Lett. 506 (2001) 196–200. [16] S. Katz, A. Aronheim, Differential targeting of the stress mitogenactivated protein kinases to the c-Jun dimerization protein 2, Biochem. J. 368 (2002) 939–945. [17] C. Jin, K. Kato, T. Chimura, T. Yamasaki, K. Nakade, T. Murata, H. Li, J. Pan, M. Zhao, K. Sun, R. Chiu, T. Ito, K. Nagata, M. Horikoshi, K.K. Yokoyama, Regulation of histone acetylation and nucleosome assembly by transcription factor JDP2, Nat. Struct. Mol. Biol. 13 (2006) 331–338. [18] A.C.Y. Chang, S.N. Cohen, Construction and characterization of amplifiable multicopy DNA cloning vehicles derived from the p15A cryptic miniplasmid, J. Bacteriol. 134 (1978) 1141–1156. [19] T.S. Tanaka, S.A. Jaradat, M.K. Lim, G.J. Kargul, X. Wang, M.J. Grahovac, S. Pantano, Y. Sano, Y. Piao, R. Nagaraja, H. Doi, W.H. Wood III, K.G. Becker, M.S. Ko, Genome-wide expression profiling of mid-gestation placenta and embryo using a 15,000 mouse developmental cDNA microarray, Proc. Natl. Acad. Sci. USA. 97 (2000) 9127–9132. [20] R.J. Duronio, E. Jackson-Machelski, R.O. Heuckeroth, P.O. Olins, C.S. Devine, W. Yonemoto, L.W. Slice, S.S. Taylor, J.I. Gordon, Protein N-myristoylation in Escherichia coli: reconstitution of a eukaryotic protein modification in bacteria, Proc. Natl. Acad. Sci. USA. 87 (1990) 1506–1510. [21] T. Yasukawa, C. Kanei-Ishii, T. Maekawa, J. Fujimoto, T. Yamamoto, S. Ishii, Increase of solubility of foreign proteins in Escherichia coli by coproduction of the bacterial thioredoxin, J. Biol. Chem. 270 (1995) 25328–25331. [22] Y. Uchimura, M. Nakamura, K. Sugasawa, M. Nakao, H. Saitoh, Overproduction of eukaryotic SUMO-1- and SUMO-2cojugated protein in Escherichia coli, Anal. Biochem. 331 (2004) 204–206. [23] T. Kouzarides, Chromatin modifications and their function, Cell 128 (2007) 693–705.
ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 376 (2008) 115–121 www.elsevier.com/locate/yabio
Phosphorylation of two eukaryotic transcription factors, Jun dimerization protein 2 and activation transcription factor 2, in Escherichia coli by Jun N-terminal kinase 1 Takehide Murata a,*, Yoriko Shinozuka a,1, Yuichi Obata b, Kazunari K. Yokoyama a b
a Gene Engineering Division, RIKEN BioResource Center, 3-1-1 Koyadai, Tsukuba, Ibaraki 305-0074, Japan Department of Biological Systems, RIKEN BioResource Center, 3-1-1 Koyadai, Tsukuba, Ibaraki 305-0074, Japan
Received 30 November 2007 Available online 6 February 2008
Abstract Recombinant eukaryotic proteins are frequently produced in Escherichia coli and such proteins are often used for biochemical studies in vitro. However, proteins produced in this way are not modified chemically, for example, by phosphorylation, acetylation, methylation, sumoylation, or ubiquitination, during their synthesis in bacterial cells. We constructed vectors for expression in E. coli of human Jun Nterminal kinase 1 (JNK1), mouse Aurora kinase B (Aurkb), and the histone acetyltransferase (HAT) domain of P/CAF. These expression vectors included the origin of replication of p15A and the origin of replication of pBR322 or ColE1. Using these expression vectors in E. coli, we were able to phosphorylate mouse and human Jun dimerization protein 2 (JDP2) and human activation transcription factor 2 (ATF-2) by the action of human JNK1 that was expressed simultaneously. Moreover, the tail region of mouse histone H3 was phosphorylated and acetylated, respectively, by Aurkb and by the HAT domain of P/CAF. We also observed that the interaction of ATF-2 with JDP2 was prevented when ATF-2 was phosphorylated. Our expression systems for production of enzyme-modified proteins in E. coli should be widely applicable and useful for biochemical studies of chemically modified eukaryotic proteins in vitro. Ó 2008 Elsevier Inc. All rights reserved. Keywords: Protein phosphorylation; Protein modification; Recombinant protein
Production of large amounts of recombinant proteins in Escherichia coli is a powerful tool for biochemical studies of the functions of such proteins and eukaryotic proteins are frequently produced in E. coli. However, the recombinant proteins produced in E. coli are not modified chemically by intracellular reactions, such as phosphorylation, acetylation, methylation, sumoylation, or ubiquitination, because of, in general, the absence of the machinery for the posttranslational modification of proteins in E. coli. Unmodified proteins allow easy and straightforward purification but experiments with unmodified proteins cannot *
Corresponding author. Fax: +81 29 836 9120. E-mail address: [email protected] (T. Murata). 1 Present address: Department of Immunology and Cell Biology, Graduate School of Medicine, Kyoto University, Yoshidakonoe-cho, Sakyo-ku, Kyoto 606-8501, Japan. 0003-2697/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2008.01.038
be expected to reveal full details of the native functions of these proteins in vivo. Production of eukaryotic proteins at large amounts has been achieved with viral vectors, such as baculovirus, retrovirus, and adenovirus [1,2]. However, production of eukaryotic proteins in E. coli is simpler in terms of both construction of the vector and purification of the protein of interest. It would, thus, be useful to develop expression systems for the large-scale production of eukaryotic proteins in E. coli in their chemically modified forms. The phosphorylation of transcription factors plays key roles in the regulation of their transcriptional activities [3]. Several groups have demonstrated the roles of the phosphorylation of transcription factors, in particular of ATF-2, in the regulation of transcriptional activity [4–6]. Phosphorylation of ATF-2 was reported after treatment of cells with stress-inducing reagents and UV irradiation.
116
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
A mutant from of ATF-2, with alanine residues instead of residues Thr69 and Thr71, is stable and is able to bind to the cyclic AMP response element (CRE).2 Moreover, appropriate homodimerization or heterodimerization with other leucine zipper proteins appears to occur normally. However, the mutant protein is no longer able to participate in the activation of CRE-dependent transcription and in the normal development of liver and heart when the homolog of ATF-2, ATF-7, is simultaneously deleted [7]. Furthermore, the interaction between ATF-2 and p300 (or related) proteins in vivo does not require phosphorylation of ATF-2 at Thr69 and Thr71 [8]. In addition to sites for phosphorylation, there are acetylation sites in ATF-2 at positions 357 and 374 that might potentially contribute to the ATF-2-mediated activation of transcription [9]. Significant enhancement of acetylation of ATF-2 does, however, not occur upon treatment of cells with so-called stress reagents, such as tumor necrosis factor-a, phorbol 12-myristate 13-acetate, or cycloheximide. By contrast, phosphorylation of ATF-2 is enhanced by similar reagents, suggesting that the phosphorylation reactions catalyzed by MAP kinases are not correlated with acetylation of lysine residues [9]. The protein kinase ATM also phosphorylates the serine residues of ATF-2 at positions 490 and 498 during the response of cells to DNA damage [10]. In a previous report, we demonstrated the interaction between ATF-2 and JDP2, both in vitro and in vivo, via their bZIP domain [11]. We also demonstrated the repression of transcription by JDP2, showing that JDP2 bound to the CRE to repress the CRE-dependent transcription that is mediated by ATF-2 in embryonal carcinoma (EC) F9 cells. ATF-2 and JDP2 bound to the differentiation response element (DRE) in the promoter of the gene for JUN in untreated EC F9 cells. After treatment of EC F9 cells with retinoic acid, JDP2 was released from the DRE and, thus, transcription of JUN was activated [12]. JDP2 has also been reported to be a repressor of the function of the JUN homodimer and the JUN/FOS heterodimer [13]. JDP2 recruits histone deacetylase 3 to the promoter of the gene for JUN and to the gene that encodes the CCAAT/enhancer-binding protein d and transcription of both gene is inhibited [12,14]. We and others have observed the immediate phosphorylation on JDP2 on Thr148 after exposure of cells to UV light (unpublished results) [15,16]. However, it is not yet clear whether JDP2 is activated by such phosphorylation. The details of the physiological roles of the phosphorylation of ATF-2 and JDP2 remain unknown and the mechanisms of transcriptional activation by phosphorylated ATF-2 and JDP2 remain unclarified. 2 Abbreviations used: CRE, cyclic AMP response element; EC, embryonal carcinoma; DRE, differentiation response element; PMSF, phenylmethylsulfonyl fluoride; GST, glutathione S-transferase; IPTG, isopropylb-D-thiogalactopyroside; PBS, phosphate-buffered saline; CIAP, calf intestinal alkaline phosphatase; MALDI, matrix-assited laser desorption ionization.
In this report, we describe an expression system for the production of phosphorylated proteins in E. coli, which should be very useful for studies of the biochemical functions of modified proteins in vitro. We also provide results of an initial study that demonstrate that binding of ATF-2 and JDP2 is affected by the phosophorylation of ATF-2. Materials and methods Plasmids Mouse JDP2-expression plasmids [12,17] derived from pGEX (Amersham Pharmacia Biotech Co., Piscataway, NJ, USA) were obtained from the RIKEN DNA Bank, RIKEN BioResource Center (Tsukuba, Ibaraki, Japan). pGST-mmJDP2_full (RDB 4757; RIKEN DNA Bank), including a gene for full-length murine JDP2, and pGSTmmJDP2_T148A (RDB 4778; RIKEN DNA Bank), with a mutation of Thr to Ala at position 148 of the encoded mouse JDP2, were prepared as described [17]. Expression vectors for His-tagged human JDP2 (pET_His-hsJDP2) and His-FLAG-tagged human ATF-2 (pET_HisFLAGhsATF-2) were also obtained from the RIKEN DNA Bank. The cDNA for human Jun N-terminal kinase 1 (JNK1; NCBI Accession No. NM_002750) and for the HAT domain of human P/CAF cDNA (NCBI Accession No. NM_003884) were amplified by reverse transcription polymerase chain reaction (RT-PCR) from first-strand cDNA prepared from HEK293 cells (RIKEN Cell Bank, Ibaraki, Japan) using JNK1-specific primers (forward 50 - AGC AGA AGC AAG CGT GAC AAC-30 ; reverse 50 - TCA CTG CTG CAC CTG TGC TAA-30 ) and P/CAF-specific primers (forward 50 -CTA GAA GAA GAA GTA TAT AGT C-30 ; reverse 50 -TCA CTT GTC AAT TAA TCC AGC-30 ). The amplified fragments of JNK1- and P/CAF-encoding DNA were inserted into the pET15b plasmid (Novagen, Merck KGaA, Darmstadt, Germany) to yield pEThsJNK1 and pEThsPCAF-HAT, respectively. T7 expression units of pEThsJNK1 and pEThsPCAF-HAT were isolated with BglII and HindIII and then inserted into the BamHI-HindIII site of pACYC184 [18] (from the Cloning Vector Collection, National Institute of Genetics, Mishima, Japan) to yield pT7Cam_hsJNK1 and pT7Cam-hsPCAF_HAT, respectively. The cDNA for mouse Aurora kinase B (Aurkb; NCBI Accession No. NM_011496) was prepared by PCR from the NIA/NIH mouse 15K cDNA clone H3053G07 [19] with the following primers: forward 50 -CTA AAG GGC AGA GGG AGG CAG A-30 ; reverse 50 -CCA TGG CTC AGA AGG AGA ACG CCT-30 . The resultant product of PCR was cloned into the SmaI site of pGEM3zf(+) (Promega, Madison, WI, USA). The NcoISalI DNA fragment [1040-basepair (bp)] that encoded the entire coding region of Aurkb was then introduced into the NcoI-XhoI site of pT7Tet1 (RDB 6211, RIKEN DNA Bank), which is a derivative of pACYC184 [18], to generate
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
pT7Tet-mmAurkb. pACYC184 contains the origin of replication of p15A [18], which allows pACYC184 to coexist in E. coli cells with plasmids in the ColE1 compatibility group (e.g., pBR322 and pUC19). Thus, we were able to coexpress proteins from two vectors in E. coli [20–22]. The cDNA for mouse histone H3.2 (Genbank Accession No. BC106177) was prepared by RT-PCR from a cDNA library of the mouse embryonic stem cell line E14TG2a, with the following primers: forward 50 -AGA ATT CGG CTC GTA CTA AGC AGA CCG CT-30 ; reverse 50 -ATT ACG CCC TCT CCC CGC GGA T-30 . The product of PCR was cloned into the SmaI site of pGEX-4T-3 (Amersham Pharmacia Biotech). The resultant plasmid was digested with SalI and then self-ligated to yield pGEXmmH3_tail, which encoded amino acids 2 through 57. All the plasmids constructed in this study will be available from the RIKEN DNA Bank. Preparation of a lysate of E. coli and purification of GSTtagged JDP2 Transformed E. coli BL21(DE3) cells (Novagen) were cultured in 40 ml of Luria-Bertani medium that contained 8 mM MgSO4 and appropriate antibiotics (50 lg/ml ampicillin, 12.5 lg/ml tetracyclin, and 12.5 lg/ml chloramphenicol) with shaking at 160 rpm until the absorbance at 600 nm (OD600) reached 1.0. Then production of recombinant proteins was induced by addition of isopropyl-b-Dthiogalactopyranoside (IPTG) to a final concentration of 1.0 mM. After incubation of cells for another 4 h, cells were pelleted by centrifugation and stored at 80 °C until use. Thawed cells were lyzed on ice by sonication in 600 ll of phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 1.5 mM KH2PO4, 8.0 mM Na2HPO4) that contained 1 mM phenylmethylsulfonyl fluoride (PMSF; Sigma Biochemical Co., St. Louis, MO, USA), 1 lg/ml pepstatin A (Sigma Biochemical), and 1 lg/ml leupeptin (Peptide Institute Inc., Minoh, Osaka, Japan). Glutathione–S-transferase-tagged (GST-tagged) proteins were isolated from the lysate by affinity chromatography on glutathione-Sepharose 4B (Amersham Pharmacia Biotech) according to the manufacturer’s instructions. Phosphatase treatment Purified GST-tagged JDP2 (0.5 lg) was incubated in 17 ll of reaction buffer (2 mM MgCl2, 1 mM dithiothreitol, 1 mM PMSF, 1 lg/ml pepstatin A, 1 lg/ml leupeptin, 20 mM Tris–Cl, pH 7.5) that contained 3.0 units of calf intestinal alkaline phosphatase (CIAP; Takara Bio Inc., Otsu, Shiga, Japan) at 37 °C for 30 min. For the analysis of modified JDP2 in vitro, purified modified GST-JDP2 proteins were separated by SDS–PAGE and eluted from the gel. Further analysis, including matrixassisted laser desorption ionization (MALDI) and mass spectrometry were performed by Shimadzu Biotech (Kyoto, Japan).
117
Antibodies Rabbit polyclonal antiserum specific for phosphorylated JDP2 was raised by Medical & Biological Laboratories Co., Ltd. (Nagoya, Aichi, Japan) against the phosphorylated oligopeptide NH2-CIVRTDSVR(pT)PESEG-COOH in which the Thr residue corresponding to residue 148 of mouse JDP2 [15] was phosphorylated. Excess amounts of unphosphorylated JDP2 are also detected by Western blotting with the phospho JDP2-specific antiserum. Monoclonal antibodies mAb 176 and 249, specific for mouse JDP2, were described previously [12]. Antibodies against GST conjugated with horseradish peroxidase (RPN-1236; Amersham Pharmacia Biotech), against His (H-15; sc803; Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), against ATF-2 (N-96; sc-6233; Santa Cruz Biotechnology), against phosphorylated ATF-2 (Thr71-R; sc7982R; Santa Cruz Biotechnology), against phosphorylated histone H3 (Ser10; 06-570; Upstate, Charlottesville, VA, USA), against acetylated histone H3 (06-599; Upstate), and against FLAG (M2; F-3165; Sigma) were purchased as indicated. Binding assays Two micrograms (as protein) of E. coli lysate (approx. 50 ll) from cells that expressed either His-JDP2 or HisFLAG-ATF-2 were mixed with 800 ll of binding buffer (PBS that contained 0.1% Nonidet-p40, 20% glycerol, and 1 mM PMSF; pH 7.4) and incubated at 4 °C for 4 h and then with 0.4 lg of antibody at 4 °C for 2 h. Immunocomplexes were recovered by the addition of protein GSepharose beads (Amersham Pharmacia Biotech) and incubation at 4 °C for 1 h. After the beads had been washed extensively with the binding buffer, bound proteins were eluted by the addition of 50 ll of 1 Laemmli sample buffer and incubation at 95 °C for 5 min. Results and discussion Phosphorylation of JDP2 It has been reported that JDP2 is phosphorylated by JNK1 and by p38 MAP kinase in vitro [15]. Therefore, we introduced pT7Cam_hsJNK1 and/or pGSTmmJDP2_full into E. coli BL21(DE3) cells. To examine the level of phosphorylation of a recombinant protein in E. coli, we subjected cell lysates, before and after induction by IPTG, to SDS–PAGE and then stained proteins with Coomassie brilliant blue to examine the total complement of soluble proteins (Fig. 1). We detected obvious bands of proteins with apparent molecular masses of 44.5 kDa (corresponding to human JNK1) and 47.0 kDa (GSTJDP2) in E. coli transformed with pT7Cam_hsJNK1 and with pGST-mmJDP2_full, respectively, after induction by IPTG (Fig. 1; compare lanes 1 and 2 and lanes 6 and 7). The lysates prepared from E. coli cells that had been trans-
118
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
Fig. 1. Production in E. coli of JNK1 and GST-JDP2 simultaneously. E. coli cells harboring pT7Cam_hsJNK1 alone (lanes 1 and 6), pGST-mmJDP2 alone (lanes 2, 4, 7, and 9), and both plasmids together (lanes 3, 5, 8, and 10) were cultured in the absence (lanes 1–5) and presence (lanes 6–10) of 1 mM IPTG for 4 h at 30 °C. The total lysate from each bacterial culture was analyzed by SDS–PAGE (8% acrylamide) and subsequent staining with Coomassie brilliant blue. Positions of GST-JDP2 and JNK1 are indicated.
formed with plasmids that encoded JNK1 and GST-JDP2, together yielded bands of proteins with apparent molecular masses of 44.5 kDa (human JNK1) and 47.0 kDa (GSTJDP2) (Fig. 1, lane 8). These results indicated that both recombinant proteins were produced simultaneously and efficiently in E. coli under our experimental conditions. To examine the extent of phosphorylation of JDP2 by JNK1 in E. coli, we performed Western blotting after treatment of affinity-purified GST-fused JDP2 with CIAP. A 47.0-kDa protein appeared on the gel when bacteria was transformed with pGST-mmJDP2 alone, and we then assumed that this protein was the unphosphorylated form of GST-JDP2 (Fig. 2, lane 1). We observed two
bands of proteins of approximately 47 kDa when bacteria had been transformed with pGST-mmJDP2_full plus pT7Cam_hsJNK1 (Fig. 2, lane 3). Treatment with CIAP reduced the intensity of the more slowly migrating band (lane 4). Furthermore, no slowly migrating form of JDP2 was detected in the lysate of bacteria that had been transformed with both pT7Cam_hsJNK1 and pGSTmmJDP2_T148A together (Fig. 3, lanes 9–12). Because the Thr residue at position 148 of JDP2 is known to be a site of phosphorylation by either JNK1 or p38 MAP kinase [15], we postulated that the more slowly migrating band corresponded to the phosphorylated form of JDP2.
Fig. 2. E. coli cells harboring pGST-mmJDP2_full alone (lanes 1, 2, 5, and 6) and with pT7Cam_hsJNK1 (lanes 3, 4, 7, and 8) were cultured in the presence of 1 mM IPTG for 4 h at 30 °C. The total lysate from each bacterial culture was incubated with glutathione-Sepharose beads. Proteins that had bound to the beads were treated with CIAP and then analyzed by SDS–PAGE (8% acrylamide) and subsequent Western blotting with monoclonal JDP2specific antibody (mAb249; see [12]) (A) and polyclonal antiserum against phosphorylated JDP2 (see text for details) (B). Positions of phosphorylated (GST-pJDP2) and unphosphorylated (GST-JDP2) JDP2 are indicated.
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
119
position that corresponded to residue 148 of JDP2 (designated here as antiserum against phosphorylated JDP2), as described under Materials and methods. With this antiserum, a strong signal was generated by the more slowly migrating band of JDP2 (Fig. 2, lane 7). Treatment of JDP2 protein with CIAP reduced the intensity of the reaction of the more slowly migrating band with the antiserum against phosphorylated JDP2 (Fig. 2, lane 8). Recombinant JDP2 accumulated in E. coli during a 4-h incubation. Similarly, phosphorylated GST-JDP2 accumulated over the same period of time (Fig. 3). Modification of various proteins Fig. 3. Accumulation in E. coli of GST-JDP2 and its phosphorylated form. E. coli cells harboring pGST-mmJDP2_full alone (lanes 1–4), pGST-mmJDP2_full plus pT7Cam_hsJNK1 (lanes 5–8), and pGSTmmJDP2_T148A plus pT7Cam_hsJNK1 (lanes 9–12) were cultured in the presence of 1 mM IPTG for the indicated times. The total lysate from each bacterial culture was analyzed by SDS–PAGE (8% acrylamide) and subsequent Western blotting, as described in the legend to Fig. 2. (A) mAb249 [12]. (B) Antibodies against phosphorylated JDP2. These antibodies recognized unphosphorylated JDP2 when an excess of JDP2 was loaded on the gel, as noted in the text.
To confirm the phosphorylation status of JDP2, we performed mass spectrometric analysis of the isolated JDP2. We extracted JDP2 from the rapidly and slowly migrating bands and digested the protein with trypsin. Examination of the MALDI time-of-flight spectra of the digested peptides from the slowly migrating band indicated that they were phosphorylated peptides derived from JDP2 (data not shown). Our results led us to conclude that the more slowly migrating band from the bacteria that had been transformed with pT7Cam_hsJNK1 plus pGSTmmJDP2_full was the phosphorylated form of mouse JDP2 and that mouse JDP2 had been phosphorylated in E. coli cells in the presence of pT7Cam_hsJNK1. We generated rabbit antiserum against an oligopeptide of mouse JDP2 with a phosphorylated Thr residue at the
We also generated the phosphorylated form of Histagged human JDP2 in E. coli. We detected phosphorylated JDP2 with the abovementioned antiserum against phosphorylated JDP2 (Fig. 4A). Phosphorylation of HisFLAG-tagged ATF-2 in a lysate of E. coli cells in which the JNK1 expression vector had been introduced simultaneously was also examined with antibodies against phosphorylated ATF-2 (Fig. 4B). The change in mobility of phosphorylated JDP2 during SDS–PAGE was not significant in the presence of the JNK1 expression vector (Fig. 4A, lanes 1 and 2). However, the phosphorylated form of ATF-2 was clearly detected with the antibodies against phosphorylated ATF-2 (Fig. 4B, lanes 5 and 6). We also examined modification of the GST-tagged tail region of histone H3 (GST-H3 tail) in E. coli using expression vectors derived from pACYC184. The GST-H3 tail was apparently phosphorylated at the Ser residue at position 10 (Ser10; Fig. 5B) when the GST-H3 tail expression vector was introduced into E. coli with the pT7Tet_ mmAurkb expression vector for mouse Aurora kinase B, a kinase that catalyzes phosphorylation at Ser10 of histone H3 [23]. In addition, the GST-H3 tail was acetylated in E. coli cells in the presence of pT7Cam_hsPCAF_HAT, which encoded the HAT domain of P/CAF (Fig. 5C), an
Fig. 4. Western blotting of JDP2 (A) and ATF-2 (B). E. coli cells harboring pET_His-hsJDP2 (lanes 1–4) and pET_HisFLAG-hsATF-2 (lanes 5–8) were cultured without (lanes 1, 3, 5, 7) or with (lanes 2, 4, 6, 8) pT7Cam_hsJNK1 in the presence of 1 mM IPTG for 4 h at 37 °C. The total lysate from each bacterial culture was analyzed by SDS–PAGE (A, 11% acrylamide; B, 8% acrylamide) and subsequent Western blotting, as described in the legend to Fig. 2. (A) lanes 1 and 2, JDP2-specific monoclonal antibody 249; lanes 3 and 4, antibodies against phosphorylated JDP2. (B) Lanes 5 and 6, antibodies against ATF-2; lanes 7 and 8, antibodies against phosphorylated-ATF-2.
120
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
Fig. 5. Western blotting for detection of the modification of the tail of histone H3. E. coli cells harboring pGEX-mmH3_tail alone (lanes 1–3), and with pT7Tet-mmAurkb (lanes 4–6) or with pEThsPCAF-HAT (lanes 7–9) were cultured in the presence of 1 mM IPTG for 4 h at 37 °C. The total lysate from each bacterial culture was analyzed by SDS–PAGE (10% acrylamide) and subsequent Western blotting, as described in the legend to Fig. 2. (A) Lanes 1–3, Antibodies against GST; (B) lanes 4–6, antibodies against phosphorylated histone H3; (C) lanes 7–9, antibodies against acetylated histone H3. The nucleotide sequence of pGEX-mmH3_tail was confirmed (data not shown) but, nonetheless, two bands of GST-H3 tail were evident on the gel. Aurkb, Aurora kinase B.
acetyltransferase specific for histone H3 [23]. Taken together, our results indicate that our expression systems are very useful not only for the phosphorylation of bZIP proteins, such as JDP2 and ATF-2, by JNK1 but also for phosphorylation and acetylation by other enzymes. Binding JDP2 and ATF-2 Our expression system in E. coli allows both ATF-2 and JDP2 to be phosphorylated. In an attempt to evaluate the properties of these newly generated proteins, we investigated the molecular interaction between them. Examining
the association of JDP2 with ATF-2, we found that HisFLAG-ATF-2 in a crude lysate of E. coli cells was efficiently immunoprecipitated by JDP2-specific monoclonal antibodies and His-JDP2 was efficiently immunoprecipitated by FLAG-specific (and, in this case ATF-2-specific) antibodies (Fig. 6, lanes 1 and 7), as reported previously [11]. To clarify the role of phosphorylation in the regulation of transcription, we compared the capacity of the phosphorylated and unphosphorylated forms of ATF-2 and JDP2 to associate with one another. When we compared the relative intensities of bands of either phosphorylated or unphosphorylated ATF-2 in immunocomplexes
Fig. 6. Coimmunoprecipitation (IP) with the monoclonal antibody against JDP2 176 (anti-JDP2; lanes 1–6) and with the monoclonal antibody against FLAG (anti-FLAG; lanes 7–12). Proteins that had bound to affinity beads were analyzed by SDS–PAGE (10% acrylamide) with subsequent Western blotting with His-specific antibody. The positions of ATF-2 and JDP2 are indicated. The band indicated by the asterisk corresponds to the antibody used for immunoprecipitation.
Phosphorylation of eukaryotic protein in E. coli / T. Murata et al. / Anal. Biochem. 376 (2008) 115–121
obtained with the JDP2-specific monoclonal antibody (Fig. 6, lanes 1–6), we noted that the intensity of the band of the phosphorylated form of ATF-2 was lower than that of unphosphorylated ATF-2 (compare lanes 1 and 2 and lanes 3 and 4, respectively). By contrast, when we compared levels of JDP2 in immunocomplexes formed with the FLAG-specific antibody (lanes 7–12), the intensities of the bands of both forms of JDP2 were similar (compare lanes 7 and 8 and lanes 9 and 10). These results suggest that, at least, phosphorylation of ATF-2 is critical for the interaction between ATF-2 and JDP2. Release of JDP2 from the ATF-2-JDP2 complex, possibly as a consequence of the phosphorylation of ATF-2, suggests a novel mechanism of interaction in the formation of heterodimers of ATF-2 and JDP2 in vitro. This report describes a compatible plasmid for overexpression of either a protein kinase or a histone acetyltransferase in E. coli, together with its respective substrate, for the simultaneous production of preparative amounts of modified recombinant protein. This approach for the production of modified proteins in E. coli is suitable for biochemical analyses of chemically modified proteins. Acknowledgments The authors thank Drs. K. Nakade, M. Kimura, and J. Pan for many helpful discussions and Ms. K. Takahashi and Ms. A. Kimura for diligent support. This work was supported by grants from the National BioResource Project of the Ministry of Education, Culture, Sports, Science and Technology, Japan and by the Bioresource Research Projects of RIKEN. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ab.2008. 01.038. References [1] C.L. Merrington, M.J. Bailey, R.D. Possee, Manipulation of baculovirus vectors, in: R. Rapley (Ed.), The Nucleic Acid Protocols Handbook, Humana Press Inc., Totowa, NJ, 2000, pp. 907–919. [2] N. Wu, M.M. Ataai, Production of viral vectors for gene therapy applications, Curr. Opin. Biotechnol. 11 (2000) 205–208. [3] A.H. Brivanlow, J.E. Darnell Jr., Signal transduction and the control of gene expression, Science 295 (2002) 813–818. [4] R. Davis, Signal transduction by the JNK group of MAP kinase, Cell 103 (2000) 239–252. [5] L. Chang, M. Karin, Mammalian MAP kinase signalling cascades, Nature 410 (2001) 37–40. [6] J.M. Kyriakis, J. Avruch, Mammalian mitogen-activated protein kinase signal transduction pathways activated by stress and inflammation, Physiol. Rev. 81 (2001) 807–869.
121
[7] W. Breitwieser, S. Lyons, A.M. Flenniken, G. Ashton, G. Bruder, M. Willington, G. Lacaud, V. Kouskoff, N. Jones, Feedback regulation of p38 activity via ATF2 is essential for survival of embryonic liver cells, Genes Dev. 21 (2007) 2069–2082. [8] M.C. Duyndam, H. van Dam, P.H. Smits, M. Verlaan, A.J. van der Eb, A. Zantema, The N-terminal transactivation domain of ATF2 is a target for the co-operative activation of the c-jun promoter by p300 and 12S E1A, Oncogene 18 (1999) 2311–2321. [9] B. Karanam, L. Wang, D. Wang, X. Liu, R. Marmorstein, R. Cotter, P.A. Cole, Multiple roles for acetylation in the interaction of p300 HAT with ATF-2, Biochemistry 46 (2007) 8207–8216. [10] A. Bhoumik, S. Takahashi, W. Breitweiser, Y. Shiloh, N. Jones, Z. Ronai, ATM-dependent phosphorylation of ATF2 is required for the DNA damage response, Mol. Cell 18 (2005) 577–587. [11] C. Jin, H. Ugai, J. Song, T. Murata, F. Nili, K. Sun, M. Horikoshi, K.K. Yokoyama, Identification of mouse Jun dimerization protein 2 as a novel repressor of ATF-2, FEBS Lett. 489 (2001) 34–41. [12] C. Jin, H. Li, T. Murata, K. Sun, M. Horikoshi, R. Chiu, K.K. Yokoyama, JDP2, a repressor of AP-1, recruits a histone deacetylase 3 complex to inhibit the retinoic acid-induced differentiation of F9 cells, Mol. Cell. Biol. 22 (2002) 4815–4826. [13] A. Aronheim, E. Zandi, H. Hennemann, S.J. Elledge, M. Karin, Isolation of an AP-1 repressor by a novel method for detecting protein-protein interactions, Mol. Cell. Biol. 17 (1997) 3094–3102. [14] K. Nakade, J. Pan, A. Yoshiki, H. Ugai, M. Kimura, B. Liu, H. Li, Y. Obata, M. Iwama, S. Itohara, T. Murata, K.K. Yokoyama, JDP2 suppresses adipocyte differentiation by regulating histone acetylation, Cell Death Differ. 14 (2007) 1398–1405. [15] S. Katz, R. Heinrich, A. Aronheim, The AP-1 repressor, JDP2, is a bona fide substrate for the c-Jun N-terminal kinase, FEBS Lett. 506 (2001) 196–200. [16] S. Katz, A. Aronheim, Differential targeting of the stress mitogenactivated protein kinases to the c-Jun dimerization protein 2, Biochem. J. 368 (2002) 939–945. [17] C. Jin, K. Kato, T. Chimura, T. Yamasaki, K. Nakade, T. Murata, H. Li, J. Pan, M. Zhao, K. Sun, R. Chiu, T. Ito, K. Nagata, M. Horikoshi, K.K. Yokoyama, Regulation of histone acetylation and nucleosome assembly by transcription factor JDP2, Nat. Struct. Mol. Biol. 13 (2006) 331–338. [18] A.C.Y. Chang, S.N. Cohen, Construction and characterization of amplifiable multicopy DNA cloning vehicles derived from the p15A cryptic miniplasmid, J. Bacteriol. 134 (1978) 1141–1156. [19] T.S. Tanaka, S.A. Jaradat, M.K. Lim, G.J. Kargul, X. Wang, M.J. Grahovac, S. Pantano, Y. Sano, Y. Piao, R. Nagaraja, H. Doi, W.H. Wood III, K.G. Becker, M.S. Ko, Genome-wide expression profiling of mid-gestation placenta and embryo using a 15,000 mouse developmental cDNA microarray, Proc. Natl. Acad. Sci. USA. 97 (2000) 9127–9132. [20] R.J. Duronio, E. Jackson-Machelski, R.O. Heuckeroth, P.O. Olins, C.S. Devine, W. Yonemoto, L.W. Slice, S.S. Taylor, J.I. Gordon, Protein N-myristoylation in Escherichia coli: reconstitution of a eukaryotic protein modification in bacteria, Proc. Natl. Acad. Sci. USA. 87 (1990) 1506–1510. [21] T. Yasukawa, C. Kanei-Ishii, T. Maekawa, J. Fujimoto, T. Yamamoto, S. Ishii, Increase of solubility of foreign proteins in Escherichia coli by coproduction of the bacterial thioredoxin, J. Biol. Chem. 270 (1995) 25328–25331. [22] Y. Uchimura, M. Nakamura, K. Sugasawa, M. Nakao, H. Saitoh, Overproduction of eukaryotic SUMO-1- and SUMO-2cojugated protein in Escherichia coli, Anal. Biochem. 331 (2004) 204–206. [23] T. Kouzarides, Chromatin modifications and their function, Cell 128 (2007) 693–705.