HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 2, 1092-1098, January 11, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/2/1092    most recent M109503200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Stemmer, C. Articles by Grasser, K. D. Search for Related Content PubMed PubMed Citation Articles by Stemmer, C. Articles by Grasser, K. D.

Protein Kinase CK2 Differentially Phosphorylates Maize Chromosomal High Mobility Group B (HMGB) Proteins Modulating Their Stability and DNA Interactions^*

Christian Stemmer, Andrea Schwander^§¶, Guy Bauw, Peter Fojan, and Klaus D. Grasser

From the  Department of Life Science, Aalborg University, Sohngaardsholmsvej 49, DK-9000 Aalborg, Denmark and the ^§ Institute for Biology III, Freiburg University, Schänzlestrasse 1, D-79104 Freiburg, Germany

Received for publication, October 2, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The high mobility group (HMG) proteins of the HMGB family are architectural factors in eukaryotic chromatin, which are involved in the regulation of various DNA-dependent processes. We have examined the post-translational modifications of five HMGB proteins from maize suspension cultured cells, revealing that HMGB1 and HMGB2/3, but not HMGB4 and HMGB5, are phosphorylated by protein kinase CK2. The phosphorylation sites have been mapped to the acidic C-terminal domains by analysis of tryptic peptides derived from HMGB1 and HMGB2/3 using nanospray ion trap mass spectrometry. In native HMGB1, Ser¹⁴⁹ is constitutively phosphorylated, whereas Ser¹³³ and Ser¹³⁶ are differentially phosphorylated. The functional significance of the CK2-mediated phosphorylation of HMGB proteins was analyzed by circular dichroism measurements showing that the phosphorylation increases the thermal stability of the HMGB proteins. Electrophoretic mobility shift assays demonstrate that the phosphorylation reduces the affinity of the HMGB proteins for linear DNA. The specific recognition of DNA minicircles is not affected by the phosphorylation, but a different pattern of protein-DNA complexes is formed. Collectively, these findings show that phosphorylation of residues within the acidic C-terminal domain of the HMGB proteins can modulate protein stability and the DNA binding properties of the HMGB proteins.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

High mobility group (HMG)¹ proteins represent a heterogeneous class of small and relatively abundant chromatin-associated proteins of eukaryotes (1, 2). Proteins belonging to the subgroup of the HMGB proteins² (previously termed HMG1/2 proteins (3)) have in common a distinctive DNA-binding motif, termed the HMG-box domain, in which the global fold is well conserved and consists essentially of three -helices arranged in an L-shape (1, 4). The HMG-box domain mediates both non-sequence-specific binding of these proteins to the minor groove of linear DNA and high affinity interactions with distorted DNA structures such as four-way junctions, minicircles, and cis-platinated DNA (2, 4, 5). In complexes with B DNA, a hydrophobic wedge on the concave surface of the HMG-box domain is inserted into the minor groove of the DNA, which contributes to the extent of DNA bending induced by the protein (5). The DNA interactions of the HMG-box domains, which occur in different plant, vertebrate, insect, and yeast HMGB proteins, are modulated by basic and acidic domains flanking the DNA-binding motif (4). HMGB proteins act as architectural components in chromatin facilitating the assembly of nucleoprotein complexes, which are involved, for instance, in the regulation of transcription and recombination (2, 4).

In contrast to other eukaryotes, which usually have two or three different HMGB proteins, (higher) plants contain several HMGB proteins (5 family members). The plant HMGB proteins have a single HMG-box domain, which is flanked by a basic N-terminal domain and an acidic C-terminal domain (6). Although the amino acid sequences of the HMG-box domains of the various plant HMGB proteins are relatively conserved, the basic and acidic flanking regions are variable in length and sequence (6). The plant HMGB proteins differ in their chromatin association and nucleosome binding (7), in their expression in the plant (8), and in some of their DNA interactions (9). Therefore, they may be adapted to act in different DNA-dependent processes in the nucleus.

Vertebrate HMGB proteins are subject to various post-translational modifications such as acetylation, methylation, ADP-ribosylation, and glycosylation, but relatively little is known about the functional significance of these modifications (10, 11). Insect HMGB proteins are phosphorylated by protein kinase C, inhibiting their DNA binding and nuclear translocation (12). More recently, it was demonstrated that insect HMGB proteins are constitutively phosphorylated by protein kinase CK2, altering their conformation, stability, and DNA binding specificity (13). Furthermore, plant CK2-type protein kinase activities were found to phosphorylate in vitro nuclear proteins from maize and broccoli that were characterized as HMG proteins by their size and solubility in 2% trichloroacetic acid (14, 15).

Here, we report that the maize HMGB1 and HMGB2/3 proteins, but not the HMGB4 and HMGB5 proteins, are phosphorylated by CK2 in vivo and in vitro. The phosphorylation sites have been mapped to the acidic C-terminal domains of the proteins. Phosphorylation by CK2 modulates the thermal stability of the plant HMGB proteins and alters their interactions with DNA.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Phosphorylation of the HMGB Proteins in Vivo-- Maize Black Mexican Sweet (BMS) suspension culture cells were grown as described previously (8). Mid-log phase cells were incubated with 15 µCi/ml [³²P]orthophosphate (Amersham Biosciences, Inc.) for 12 h. Cells were frozen in liquid nitrogen, and the HMG proteins were extracted using 2% trichloroacetic acid (16). The extracted proteins were separated by SDS-PAGE in 18% polyacrylamide gels and analyzed by silver staining and autoradiography.

Purification of Proteins-- Full-length and truncated recombinant maize HMGB proteins were expressed in Escherichia coli and purified by three-step column chromatography as described previously (16, 17). Native HMGB proteins were isolated from maize BMS cells by 2% trichloroacetic acid extraction and subsequently purified by Resource Q chromatography as described previously (16). The recombinant maize CK2 protein was expressed in E. coli using the pT7-7/BL21(DE3) system (kindly provided by Drs. B. Boldyreff and O.-G. Issinger) as described previously (18). The recombinant protein kinase was purified by three-step fast protein liquid column chromatography. The first chromatography was performed using heparin-agarose (Sigma) in HA buffer (0-1 M KCl in 50 mM Tris/HCl, pH 7.9, 1 mM EDTA, 1 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride, 100 µg/ml benzamidine), and was followed by Resource S and Resource Q (Amersham Biosciences, Inc.) chromatographies, which were performed as described previously (16). The native CK2 was purified from nuclei isolated from immature maize kernels by three-step fast protein liquid column chromatography. Heparin-agarose and Resource S chromatographies were performed as described for the recombinant protein kinase, whereas casein-agarose (Sigma) was used for the final purification step in buffer CA (0-1 M NaCl in 10 mM sodium phosphate, pH 7.0, 7.5 mM MgCl₂, 5 mM EDTA, 10 mM 2-mercaptoethanol, 0.5 mM phenylmethylsulfonyl fluoride).

CK2 in Vitro Phosphorylation Assays-- For analytical phosphorylation reactions, the different HMGB proteins (1 µM) were incubated in a total volume of 20 µl at 37 °C for 1 h with 40 ng of recombinant CK2 in the presence of 100 nCi of [-³²P]ATP (or [-³²P]GTP) (Amersham Biosciences, Inc.) in CK2 buffer (25 mM Tris/HCl, pH 8.5, 10 mM MgCl₂, 1 mM DTT). The phosphorylation reactions were monitored by separation of the proteins by SDS-PAGE in 18% polyacrylamide gels followed by autoradiography or scanning of the gels with a Typhoon 8600 phosphorimaging device (Amersham Biosciences, Inc.). The scanned data were used for quantification. For preparative phosphorylation, 20 µg of the HMGB proteins were reacted with 400 ng of CK2 in the presence of 300 µM ATP in a total volume of 50 µl. The phosphorylation status of the HMGB proteins was checked by acetic acid urea-PAGE in 18% polyacrylamide gels (which resolve the HMGB proteins according to the number of phosphates incorporated into the protein) and by mass spectrometry. To ensure specific phosphorylation of the proteins, the CK2 phosphorylation reactions of the proteins used for structural and functional assays were limited so that routinely double-phosphorylated HMGB proteins were used for further analyses.

Mass Analysis of Proteins and Tryptic Peptides-- All mass spectrometry analyses were performed on an ion trap LC-Q mass spectrometer (Finnigan) equipped with a nanospray source. Before measurement of the total mass of native and recombinant HMGB1 and HMGB2/3, the purified proteins were concentrated and desalted using a C18 Ziptip (Millipore). The proteins were eluted in 5 µl of 50% acetonitrile, 0.1% acetic acid, and directly applied in the nanospray needle. For HMGB protein dephosphorylation, native or recombinant proteins were desolved in 50 mM Tris/HCl, pH 8.5, and 2 units of alkaline phosphatase (Sigma) were added and incubated for 2 h at 28 °C. For tryptic digests, the HMGB proteins were dissolved in 25 mM Tris/HCl, pH 8.5, to which 0.5 µg of trypsin (Promega) was added. Trypsin digestion was performed at 28 °C for 3 h. One-half of the digest was acidified with acetic acid, and the other part was treated with 2 units of alkaline phosphatase for 2 h at 28 °C before the addition of acetic acid.

Circular Dichroism-- The measurements of HMGB1 and HMGB2 at a concentration of 10 µM in 10 mM sodium phosphate, pH 7.0, 1 mM EDTA, 1 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride were performed using a Jasco J700 instrument with a Peltier temperature controller. The temperature scans were performed in the range between 20 and 90 °C at a scan rate of 1.5°·min¹. All scans were base-line-subtracted, and the raw data were used for fitting, assuming a two-state unfolding process. Raw data were fitted to equation 1 (19) and all fittings were performed in Kaleidagraph,

(Eq. 1)

where T_m is the midpoint temperature of denaturation, H^Tm is the enthalpy of denaturation at this temperature, and C_p is the specific heat capacity for unfolding. A and C are the ellipticities of the native and denatured states at 298 K, and B and D are the linear dependencies of these values on T.

Electrophoretic Mobility Shift Assays-- For binding studies with linear DNA, a 98-bp DNA fragment was amplified by PCR from plasmid pCB8 (20) using fluorescein-labeled primers P1 (5'-CTTTGTAGAGTGCGGGTGCT) and P2 (5'-ACAGGCCAGGGCCAGCGCTT). Various concentrations of recombinant nonphosphorylated and CK2-phosphorylated HMGB proteins were incubated in a total volume of 20 µl with 2 ng of the fluorescein-labeled DNA fragment in a buffer containing 10 mM Tris/HCl, pH 7.5, 50 mM NaCl, 1 mM EDTA, and 1 mM DTT. Protein binding to the DNA was examined by electrophoresis on 5% polyacrylamide gels in 1× TBE buffer and scanning of the gels using a Typhoon 8600 phosphorimaging device (Amersham Biosciences, Inc.) at excitation 532 nm and emission 526 nm. DNA binding analyses using a mixture of ³²P-labeled linear and circularized 78-bp DNA fragments and various concentrations of HMGB proteins by electrophoretic mobility shift assays were performed as described previously (16, 17).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Phosphorylation of the Maize HMGB Proteins in Vivo-- A comparison of the measured masses of the five maize HMGB proteins purified from immature kernels with the masses calculated from the amino acid sequences (Fig. 1A) indicated that the proteins are post-translationally modified in vivo (8). To analyze whether the HMGB proteins are phosphorylated in maize BMS suspension cultured cells, in vivo ³²P labeling of the cells was performed. Taking advantage of the acid solubility of the HMG proteins (1, 16), the HMGB proteins were isolated by 2% trichloroacetic acid extraction of the labeled cells. Autoradiography of the trichloroacetic acid-extracted proteins separated by SDS-PAGE revealed that predominantly two protein bands were labeled with ³²P that co-migrate with the HMGB1 and HMGB2/3 proteins (Fig. 1B); this indicates that the maize HMGB1 and HMGB2/3 proteins are phosphoproteins in vivo.

View larger version (42K): [in this window] [in a new window]   Fig. 1.   Phosphorylation of maize HMGB proteins in vivo and by CK2 in vitro. A, alignment of the amino acid sequences of the five HMGB proteins from Zea mays. The HMG-box domain is indicated in bold, and the residues delineating the different recombinant HMGB1 proteins used in panel D are depicted above the HMGB1 sequence. The three serine residues, Ser133, Ser136, and Ser149, of HMGB1 that are phosphorylated by CK2 are indicated by , and the serine and threonine residues within the two C-terminal tryptic peptides of HMGB2 and HMGB3, which are the candidate sites for CK2 phosphorylation (see text following for details), are indicated by S or T. B, silver staining and autoradiography of proteins extracted by 2% trichloroacetic acid from 32P-labeled maize BMS cells after separation by SDS-PAGE. The migration positions of the HMGB1 and HMGB2/3 proteins are indicated. C, autoradiography of equal amounts of recombinant maize HMGB proteins phosphorylated in vitro by CK2 in the presence of [32P]ATP and separated by SDS-PAGE. D, autoradiography of equal amounts of full-length and C-terminally truncated HMGB proteins phosphorylated in vitro by CK2 in the presence of [32P]ATP and separated by SDS-PAGE.

CK2 Phosphorylates in Vitro HMGB1 and HMGB2/3 but Not HMGB4 and HMGB5-- Because the amino acid sequences of all maize HMGB proteins contain several consensus substrate sites for protein kinase CK2, and because some plant HMG-type proteins could be phosphorylated by CK2 activities (14, 15), we tested whether the purified recombinant maize HMGB proteins are substrates for purified recombinant maize CK2. Protein kinase CK2 (also known as casein kinase II) is an ubiquitous enzyme catalyzing the phosphorylation of certain Ser/Thr residues in the substrate protein (21, 22). The well characterized maize CK2 (18, 23) was expressed in E. coli and purified by three-step column chromatography. Incubation of the five maize HMGB proteins with recombinant CK2 in the presence of [³²P]ATP and analysis of the proteins by SDS-PAGE and autoradiography demonstrated that CK2 can catalyze the phosphorylation of the HMGB1 protein and the two closely related HMGB2/3 proteins (89% amino acid sequence identity), but the enzyme does not phosphorylate HMGB4 and HMGB5. HMGB2 and HMGB3 are more readily phosphorylated by CK2 in vitro than HMGB1. Depending on the extent of the CK2 phosphorylation reaction, the HMGB proteins occur in the single, double, or triple phosphorylated state.³ For further structural and functional studies, the double phosphorylated form of the three proteins was used. As most of the predicted CK2 consensus phosphorylation sites (as predicted by Phospho Base 2.0)⁴ are situated within the acidic C-terminal domain of the HMGB proteins, recombinant C-terminally truncated versions of HMGB1 (Fig. 1A) were examined in comparison with full-length HMGB1 in CK2 phosphorylation assays (Fig. 1D). Compared with full-length HMGB1, the phosphorylation of HMGB1(M1-D134), which has a truncated C-terminal domain, was strongly reduced (only a very faint band was visible) and completely abolished for HMGB1(M1-K123), which lacks the acidic C-terminal domain, demonstrating that recombinant maize CK2 phosphorylates residues within this acidic domain.

To examine whether the HMGB1 proteins are comparably phosphorylated by native CK2, the enzyme was purified by three-step column chromatography from immature maize kernels. The CK2 preparation isolated from maize kernels contained a protein that co-migrated in SDS-PAGE with recombinant CK2 (~36.5 kDa) and that reacted with an antiserum raised against recombinant CK2.³ In contrast to most other protein kinases, CK2 can utilize efficiently either ATP or GTP as phosphate donors (21). We compared the ability of recombinant and native maize CK2 to use ATP and GTP in HMGB2 phosphorylation assays, demonstrating that HMGB2 was phosphorylated similarly by both CK2 preparations in presence of various concentrations of ATP or GTP (Fig. 2, A and B). Another distinguishing feature of CK2 is its specific inhibition by low concentrations of heparin (21). Therefore, the phosphorylation of HMGB2 by the recombinant and native maize CK2 was performed in the presence of various heparin concentrations. Both protein kinase preparations were severely inhibited by 40 and 80 ng/ml heparin, and the reaction was completely abolished by 240 ng/ml heparin (Fig. 2C). Thus, the comparable utilization of ATP and GTP as phosphate donors and the specific inhibition of the phosphorylation reaction by heparin indicate that the recombinant and native maize CK2 preparations react almost indistinguishably with the HMGB2 protein.

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Phosphorylation of HMGB2 in vitro by native and recombinant CK2. Quantification of in vitro phosphorylation assays of HMGB2 using native () or recombinant () CK2 in the presence of various concentrations of radioactive GTP (A) or ATP (B and C). C, inhibition of CK2 by various concentrations of heparin.

Phosphorylations Occurring in the Native and the in Vitro CK2-phosphorylated HMGB Proteins-- The total masses of the HMGB1 and HMGB2/3 proteins purified from BMS cells were measured before and after treatment with alkaline phosphatase (which dephosphorylates phosphoproteins) to determine the number of phosphorylations in the native proteins. In the case of the native HMGB1 protein, the mass of the protein was reduced by 383 Da upon phosphatase treatment. Dependent on the number of Na⁺ ions removed during the phosphatase treatment, the number of phosphorylation sites on HMGB1 varies between three and five. Most likely, there may be four phosphorylations, as these together with three simultaneously removed Na⁺ ions account for a mass difference of 385 Da. In line with that, the removal of four phosphates was also observed by analysis of the dephosphorylation reaction using acetic acid urea-PAGE.³ Furthermore, the measured mass of the phosphatase-treated native HMGB1 matched the calculated average mass (Table I) indicating that HMGB1 most likely contains no additional post-translational modifications. Mass determination of the protein fraction containing the closely related HMGB2 and HMGB3 proteins clearly showed the presence of three species displaying different masses. Treatment with alkaline phosphatase reduced the number of observed mass peaks to two (Table I), one with the same molecular mass as HMGB2 and the other one derived from HMGB3, with a mass 162 Da larger than the calculated average mass. The nature of this mass difference is unknown. In principle, 162 Da may correspond to the mass of a hexose but could also be the result of a combination of other modifications and/or adhesion of Na⁺ ions. The most likely scenario to explain the mass differences before and after alkaline phosphatase treatment (HMGB2, 238 Da; HMGB3, 237 and 157 Da) is the following. Dephosphorylation of native HMGB2 (15,556 Da) gives rise to the species with a mass of 15,318 Da corresponding to the removal of three phosphate groups. Native HMGB3 is present in two different phosphorylation states containing in part two phosphorylations (resulting in a mass of 15,326 Da) or containing three phosphate groups (resulting in a mass of 15,406 Da). The mass values determined for the native untreated HMGB1 and HMGB2/3 proteins isolated from BMS cells (Table I) are similar to those determined previously for proteins purified from immature maize kernels (8), indicating that the proteins are similarly modified in both tissues.

To determine the number of sites phosphorylated by CK2 in vitro, the difference in mass between the phosphorylated recombinant HMGB proteins before and after treatment with alkaline phosphatase was analyzed by mass spectroscopy. The phosphatase treatment reduced the masses of the proteins by 152, 214, and 156 Da for HMGB1, HMGB2, and HMGB3, respectively. This finding corresponds well with the presence of two phosphorylations on each protein, which is in line with the number of phosphorylations determined by acetic acid urea gel electrophoresis.³ In the case of HMGB2 we assume that, together with the two phosphate groups, two Na⁺ ions were removed from the protein. The measured masses of the alkaline phosphatase-treated and the CK2-phosphorylated/alkaline phosphatase-treated recombinant HMGB1 fit well (Table I). HMGB2 and HMGB3 show mass differences of 95 and 90 Da, respectively, between the measured alkaline phosphatase-treated proteins and calculated masses. The nature of this difference is unknown but could be explained by the binding of four Na⁺ ions to the polypeptide chain (24). The HMGB proteins bind easily Na⁺ ions. We have recognized the tryptic C-terminal acidic peptides and other peptides as Na⁺ and 2 Na⁺ adducts in the mass spectra next to the common protonated forms.

CK2 Phosphorylates Residues within the Acidic C-terminal Domain of HMGB1 and HMGB2/3-- To determine the phosphorylation sites within the HMGB1 and HMGB2/3 proteins, the native proteins and the recombinant proteins phosphorylated by CK2 in vitro were digested with trypsin, and the resulting peptides were analyzed by mass spectrometry. Analysis of the tryptic peptides of the CK2-phosphorylated recombinant HMGB1 demonstrated conclusively that each of the two C-terminal tryptic peptides contains a single phosphorylation. MS/MS analysis of the ion with an m/z value of 1202.2, clearly showing that Ser¹⁴⁹ in the C-terminal peptide is phosphorylated. The same residue, Ser¹⁴⁹, was found exclusively in the phosphorylated form in native HMGB1 (Table II). The situation for the second to last C-terminal tryptic peptide was more complicated. In the recombinant protein this peptide contains a single phosphorylation site, but the MS/MS data were inconclusive as to whether the phosphate is situated on Ser¹³³ or on Ser¹³⁶. In the trypsin digestion of the native HMGB1, this peptide was found in the double phosphorylated and single phosphorylated states as well as in the nonphosphorylated state (Fig. 3). The MS/MS data of the peptide ion with a mass of 1873 Da (m/z 936.4) contained a phosphate group on both of the Ser residues (Ser¹³³ and Ser¹³⁶). The MS/MS data of the ion with mass of 1793 Da (m/z 896.5) showed that the peptide was uniquely phosphorylated on Ser¹³³. The tryptic peptide with a mass of 1713 Da (m/z 856.5), of which none of the two Ser residues were phosphorylated, was also present in the MS spectrum of the tryptic peptides from native HMGB1. Approximately 75% of the peptide (Glu¹²⁴-Lys¹³⁷) occurs in the double phosphorylated form, whereas ~12.5% is found in the nonphosphorylated and ~12.5% in the single phosphorylated form (Fig. 3). Therefore, HMGB1 is constitutively phosphorylated on Ser¹⁴⁹ but differentially phosphorylated on Ser¹³³ and Ser¹³⁶ (Table II). The in vitro phosphorylation experiment (Fig. 1D) with the C-terminally truncated HMGB1(M1-D134) protein is in agreement with this finding. HMGB1(M1-D134) lacks the region containing Ser¹³⁶ and Ser¹⁴⁹ and is phosphorylated only extremely weakly by CK2 (compared with the full-length protein), and the phosphorylation is completely abolished for HMGB1(M1-K123).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Mass spectrum revealing the different phosphorylation states of peptide Glu124-Lys137 of native HMGB1. The tryptic peptide was found in three different phosphorylation states (m/z: double phosphorylated, 936.4; single phosphorylated, 896.5; nonphosphorylated, 856.5). The phosphorylation states of the two serine residues Ser133 and Ser136 of the peptide are indicated by Ser (nonphosphorylated) and PO4-Ser (phosphoserine).

The analysis of the tryptic peptides derived from native and in vitro CK2-phosphorylated HMGB2 and HMGB3 revealed that the two C-terminal tryptic peptides were phosphorylated. We could not determine exactly the number and location of the phosphate groups in the amino acid sequences of these peptides, as we were unable to detect the phosphorylated peptides in the mass spectrum, which may be due to ion suppression (25, 26) or to the fact that these peptides have a negative charge. Alkaline phosphatase treatment of the phosphorylated tryptic peptides, however, resulted in the appearance of the two C-terminal nonphosphorylated tryptic peptides in the mass spectrum, which were absent in the original mass spectrum of the untreated sample. This fact indicates that the two C-terminal peptides of both proteins are phosphorylated. Therefore, the 2-3 phosphorylations of HMGB2/3 (see above) may occur on five candidate residues for HMGB2 (Thr¹¹⁴, Ser¹²⁰, Ser¹²², Ser¹³⁰, Ser¹³¹) and on four candidate residues for HMGB3 (Thr113, Ser119, Ser121, Ser130), as indicated in Fig. 1A.

CK2 Phosphorylation Increases the Thermal Stability of the HMGB Proteins-- To examine whether CK2 phosphorylation alters the stability of the HMGB proteins, the thermal denaturation of the proteins was followed by CD at 222 nm, because the HMG-box domain is largely -helical. The temperature scans for nonphosphorylated and in vitro CK2-phosphorylated HMGB1 and HMGB2 revealed that the phosphorylated proteins exhibited increased melting temperatures (Fig. 4). The melting profiles for the nonphosphorylated and phosphorylated proteins are shifted toward higher melting temperatures according to their increased thermostability. The observed melting temperatures were 48.3 and 50.3 °C for nonphosphorylated HMGB1 and HMGB2, respectively, and 52.4 and 51.6 °C for phosphorylated HMGB1 and HMGB2, respectively. Although there was only an insignificant rise (1.3 °C) in the thermal stability of HMGB2, a marked phosphorylation-induced increase (4.1 °C) in the melting temperature was observed in the case of HMGB1.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   CD spectrometry of the thermal denaturation of phosphorylated and nonphosphorylated HMGB1 and HMGB2. CD222 nm of the thermal denaturation of nonphosphorylated HMGB1 and HMGB2 and of HMGB1 and HMGB2 phosphorylated by CK2 in vitro (HMGB1-PCK2, HMGB2-PCK2).

Phosphorylation by CK2 Alters the Interactions with DNA-- The interaction of nonphosphorylated and in vitro CK2-phosphorylated HMGB1 and HMGB2 with linear DNA was compared to test whether the phosphorylation alters the DNA binding properties of the proteins. Increasing concentrations of the HMGB proteins were incubated with a 98-bp fragment, and the formation of complexes was monitored by electrophoretic mobility shift assays (Fig. 5A). As the plant HMGB proteins do not form specific complexes with linear DNA (16), protein binding to the DNA can be seen best as the disappearance of the DNA band corresponding to the unbound fragment. With both proteins the phosphorylation resulted in a reduced affinity for the linear DNA, because (compared with the nonphosphorylated proteins) higher concentrations of phosphorylated HMGB1 and HMGB2 are required to detect DNA binding.

View larger version (78K): [in this window] [in a new window]   Fig. 5.   Electrophoretic mobility shift assays reveal that phosphorylation of HMGB1 and HMGB2 reduces the affinity for linear DNA but not for DNA minicircles. A, increasing concentrations (0, 125 nM, 250 nM, 500 nM, 1 µM, 2 µM) of nonphosphorylated HMGB1 and HMGB2 and of HMGB1 and HMGB2 phosphorylated by CK2 in vitro (HMGB1-PCK2, HMGB2-PCK2) were incubated with a fluorescein-labeled 98-bp DNA fragment. The binding reactions were separated by native PAGE and scanned using a phosphorimaging device. The migration position of the unbound DNA fragment (lin) is indicated. B, increasing concentrations (0, 10 nM, 50 nM, 100 nM, 500 nM, 1 µM) of nonphosphorylated and phosphorylated HMGB1 and HMGB2 were incubated with a mixture of linear and circularized 32P-labeled 78-bp DNA fragment. The binding reactions were separated by native PAGE and scanned using a phosphorimaging device. The migration positions of the unbound linear (lin) and circularized (mc) fragments and of the protein-DNA complexes formed with the DNA minicircle (c1, c2, c3) are indicated.

DNA minicircles are high affinity binding sites for HMGB proteins of various sources (4-6). The minicircles are bound structure specifically, because HMGB proteins display a marked preference for the minicircle over the corresponding linear DNA. The interaction of nonphosphorylated and in vitro CK2-phosphorylated HMGB1 and HMGB2 with a ³²P-labeled 78-bp DNA minicircle was examined in the presence of the corresponding linear 78-bp fragment using electrophoretic mobility shift assays. In the tested protein concentration range, both proteins bound the minicircular but not the linear fragment (Fig. 5B). The HMGB proteins bind the linear DNA only when the preferred binding sites on the minicircles are occupied by the proteins (16). The phosphorylation of the proteins has no significant influence on the recognition of the structure of the minicircle. However, at higher concentrations (500 nM, 1 µM) the nonphosphorylated HMGB1 and HMGB2 form a third complex with the minicircle, which is not formed by the phosphorylated proteins.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Mass spectrometric analyses of HMGB proteins purified from immature maize kernels have previously indicated that the HMGB proteins are subject to post-translational modifications (8). In this report, we show by in vivo ³²P labeling that HMGB1 and HMGB2/3 are phosphorylated in BMS cells. HMGB1 and HMGB2/3 but not HMGB4 and HMGB5 are phosphorylated in vitro by recombinant (Fig. 1C) and native³ maize protein kinase CK2, although all five proteins theoretically contain CK2 phosphorylation sites. Moreover, the CK2 phosphorylation sites predicted for HMGB1 and HMGB2/3 are only to a certain extent in agreement with our experimental results obtained with native HMGB proteins. The minimal consensus consists of an acidic or phosphorylated residue at position +3 relative to the phosphorylation site, but efficient phosphorylation by CK2 requires the presence of clusters of acidic residues around the phosphorylation site (21, 22). These requirements are fulfilled for all the CK2 phosphorylation sites determined in the maize HMGB1 and HMGB2/3 proteins (Fig. 1A) except for Ser¹³³ in HMGB1. Ser¹³³ is phosphorylated although Ser¹³⁶ (the residue +3 relative to Ser¹³³) is not previously phosphorylated by CK2, because in the single phosphorylated tryptic peptide Glu¹²⁴-Lys¹³⁷ it is clearly Ser¹³³ that is phosphorylated (Fig. 3 and Table II). In the case of Ser¹³³, the acidic residue at position +1 (Asp¹³⁴) together with the acidic cluster upstream of the phosphorylation site probably can substitute for the usually critical acidic or phosphorylated residue at position +3 (22).

The HMGB1 and HMGB2/3 proteins, but not the HMGB4 and HMGB5 proteins, are phosphorylated by CK2 within their acidic C-terminal domains. In the HMGB1 protein isolated from BMS cells, amino acid residue Ser¹⁴⁹ is phosphorylated constitutively, whereas Ser¹³³ and Ser¹³⁶ are differentially phosphorylated (Table II). In the case of the insect HMGB proteins derived from Chironomus and Drosophila, the tested HMGB proteins are essentially phosphorylated constitutively at two or three residues within their acidic C-terminal domains (13). By contrast, it is unlikely that the vertebrate HMGB proteins are substrates for CK2 phosphorylation, as the acidic C-terminal domain of these proteins consists mainly of a consecutive stretch of aspartate and glutamate residues (1) lacking canonical CK2 phosphorylation sites. In the case of the maize HMGB1, it is likely that, in addition to CK2, another so far unidentified protein kinase contributes to the phosphorylation of HMGB1 in vivo because HMGB1 isolated from BMS cells contains four phosphate groups, but only three residues are phosphorylated by CK2.

The thermal denaturation experiments with the nonphosphorylated HMGB proteins demonstrated that HMGB2 has a slightly higher thermostability than HMGB1. CK2-mediated phosphorylation of the two proteins induced an increase in thermostability (4.1 and 1.3 °C for HMGB1 and HMGB2, respectively), which was more prominent with HMGB1. This marked increase in T_m of phosphorylated HMGB1 resulted in a slightly higher stability of phosphorylated HMGB1 compared with that of phosphorylated HMGB2. The relatively high thermostability of phosphorylated HMGB1 is in agreement with the finding that HMGB1 is the metabolically most stable maize HMGB protein in BMS cells (8). In line with our results, fluorescence studies of a HMGB protein from Chironomus have revealed that the CK2-phosphorylated protein exhibits a melting temperature that was 2.4 °C higher than that of the nonphosphorylated protein (13).

Depending on the DNA substrate, the acidic C-terminal domain can severely influence the DNA binding of animal and plant HMGB proteins. Although the acidic tail reduces the affinity of HMGB proteins for linear and supercoiled DNA and for four-way junction DNA, it has only relatively little effect on the affinity for DNA minicircles (17, 27-32). The reduced affinity of maize HMGB1 and HMGB2 phosphorylated by CK2 for linear DNA (Fig. 5A) is in line with the findings that DNA affinity of HMGB proteins is decreased by an increase in the length of the acidic tail, which correlates with a higher number of negative charges (17, 27, 31). Phosphorylation of the acidic domain of a HMGB protein from Chironomus by CK2 significantly reduced the affinity of the protein for four-way junction DNA, whereas it had no effect on the interaction with linear DNA (13). The fact that the affinity of plant HMGB proteins for linear DNA is reduced by phosphorylation of the acidic tail could be explained by stronger intramolecular interactions of the acidic region presumably with the basic N-terminal domain (6). This basic N-terminal domain (typical for plant and yeast HMGB proteins) enhances the binding to linear DNA (17, 33); and insect HMGB proteins lack this domain, which may explain the different effect of CK2 phosphorylation of the acidic tail on the affinity of insect and plant HMGB proteins for linear DNA. CK2-mediated phosphorylation of maize HMGB1 and HMGB2 had no marked effect on the affinity of the proteins for DNA minicircles (Fig. 5B), but the phosphorylated proteins formed only two complexes with the minicircle, whereas the nonphosphorylated proteins formed three complexes in the same range of protein concentration. Because the acidic tail is involved in the oligomerization of the maize HMGB1 protein in the presence of DNA (17), it is possible that phosphorylation of the acidic domain modulates the cooperativity of the binding to the minicircles, which may be critical for minicircle binding (34).

In contrast to other eukaryotes, which usually contain two or three HMGB proteins, five different HMGB proteins have been identified in maize and Arabidopsis. Moreover, the plant HMGB proteins are structurally more variable than mammalian, insect, and yeast HMGB proteins, both in size and in primary structure (6). They display different expression levels in the plant, are differently associated with chromatin, and exhibit differences in some of their DNA interactions (7-9). Therefore, the different plant HMGB proteins might have been adapted to act as accessory factors in a variety of specific nucleoprotein structures (6). The differential phosphorylation by CK2 further increases the number of HMGB variants occurring in plants, which display subtle differences in their DNA interactions and/or their protein/protein contacts, making these proteins even more versatile architectural chromatin-associated factors. In plants, CK2 phosphorylates a wide variety of substrate proteins and is involved in the regulation of various developmental programs such as circadian rhythm, control of light- and salicylic acid-induced gene expression, and plant growth (35-37). It will be interesting to examine whether the phosphorylation of HMGB proteins catalyzed by CK2 plays a role in these processes and how the recently identified maize CK2 subtypes (38) are involved in that.

	ACKNOWLEDGEMENTS

We thank Dr. W. Bessler for preparation of the CK2 antiserum, Drs. B. Boldyreff and O.-G. Issinger for the maize CK2 expression plasmid, and Dr. D. Otzen for comments on the manuscript.

	FOOTNOTES

^* This work was supported by grants from the German Research Society (DFG) and the Danish Research Council (SNF) (to K. D. G.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ present address: Dept. of Molecular Cell Biology, Heinrich-Pette-Institute, Martinistr. 52, D-20251 Hamburg, Germany.

To whom correspondence should be addressed. Fax: 45-9814 1808; E-mail: kdg@bio.auc.dk.

Published, JBC Papers in Press, November 1, 2001, DOI 10.1074/jbc.M109503200

² The nomenclature of the HMG proteins has been revised recently: informatics.jax.org/mgihome/nomen/genefamilies/hmgfamily.shtml.

³ C. Stemmer, A. Schwander, and K. D. Grasser, unpublished results.

⁴ Found on the Internet at cbs.dtu.dk/data bases/PhosphoBase/.

	ABBREVIATIONS

The abbreviations used are: HMG, high mobility group; BMS, Black Mexican Sweet; DTT, dithiothreitol; m/z, mass-to-charge ratio; MS, mass spectrometry; CK2, casein kinase II.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

This article has been cited by other articles:

				J. H. Youn and J.-S. Shin Nucleocytoplasmic Shuttling of HMGB1 Is Regulated by Phosphorylation That Redirects It toward Secretion J. Immunol., December 1, 2006; 177(11): 7889 - 7897. [Abstract] [Full Text] [PDF]

				D. Launholt, T. Merkle, A. Houben, A. Schulz, and K. D. Grasser Arabidopsis Chromatin-Associated HMGA and HMGB Use Different Nuclear Targeting Signals and Display Highly Dynamic Localization within the Nucleus PLANT CELL, November 1, 2006; 18(11): 2904 - 2918. [Abstract] [Full Text] [PDF]

				P. G. Ulery, G. Rudenko, and E. J. Nestler Regulation of {Delta}FosB Stability by Phosphorylation. J. Neurosci., May 10, 2006; 26(19): 5131 - 5142. [Abstract] [Full Text] [PDF]

				R. Samaniego, S. Y. Jeong, C. de la Torre, I. Meier, and S. M. Diaz de la Espina CK2 phosphorylation weakens 90 kDa MFP1 association to the nuclear matrix in Allium cepa J. Exp. Bot., January 1, 2006; 57(1): 113 - 124. [Abstract] [Full Text] [PDF]

				K. Woll, L. A. Borsuk, H. Stransky, D. Nettleton, P. S. Schnable, and F. Hochholdinger Isolation, Characterization, and Pericycle-Specific Transcriptome Analyses of the Novel Maize Lateral and Seminal Root Initiation Mutant rum1 Plant Physiology, November 1, 2005; 139(3): 1255 - 1267. [Abstract] [Full Text] [PDF]

				Y. Li, D. M. Keller, J. D. Scott, and H. Lu CK2 Phosphorylates SSRP1 and Inhibits Its DNA-binding Activity J. Biol. Chem., March 25, 2005; 280(12): 11869 - 11875. [Abstract] [Full Text] [PDF]

				T. Iype, D. G. Taylor, S. M. Ziesmann, J. C. Garmey, H. Watada, and R. G. Mirmira The Transcriptional Repressor Nkx6.1 Also Functions as a Deoxyribonucleic Acid Context-Dependent Transcriptional Activator during Pancreatic {beta}-Cell Differentiation: Evidence for Feedback Activation of the nkx6.1 Gene by Nkx6.1 Mol. Endocrinol., June 1, 2004; 18(6): 1363 - 1375. [Abstract] [Full Text] [PDF]

				J. E. Podrabsky and G. N. Somero Changes in gene expression associated with acclimation to constant temperatures and fluctuating daily temperatures in an annual killifish Austrofundulus limnaeus J. Exp. Biol., June 1, 2004; 207(13): 2237 - 2254. [Abstract] [Full Text] [PDF]

				S. Yanagisawa Dof Domain Proteins: Plant-Specific Transcription Factors Associated with Diverse Phenomena Unique to Plants Plant Cell Physiol., April 15, 2004; 45(4): 386 - 391. [Abstract] [Full Text] [PDF]

				B. H. Lower and P. J. Kennelly Open Reading Frame sso2387 from the Archaeon Sulfolobus solfataricus Encodes a Polypeptide with Protein-Serine Kinase Activity J. Bacteriol., June 1, 2003; 185(11): 3436 - 3445. [Abstract] [Full Text] [PDF]

				M. S. Wiebe, T. K. Nowling, and A. Rizzino Identification of Novel Domains within Sox-2 and Sox-11 Involved in Autoinhibition of DNA Binding and Partnership Specificity J. Biol. Chem., May 9, 2003; 278(20): 17901 - 17911. [Abstract] [Full Text] [PDF]

				N. M. Krohn, C. Stemmer, P. Fojan, R. Grimm, and K. D. Grasser Protein Kinase CK2 Phosphorylates the High Mobility Group Domain Protein SSRP1, Inducing the Recognition of UV-damaged DNA J. Biol. Chem., April 4, 2003; 278(15): 12710 - 12715. [Abstract] [Full Text] [PDF]

				F. MEGGIO and L. A. PINNA One-thousand-and-one substrates of protein kinase CK2? FASEB J, March 1, 2003; 17(3): 349 - 368. [Abstract] [Full Text] [PDF]

				I. Carlberg, M. Hansson, T. Kieselbach, W. P. Schroder, B. Andersson, and A. V. Vener A novel plant protein undergoing light-induced phosphorylation and release from the photosynthetic thylakoid membranes PNAS, January 21, 2003; 100(2): 757 - 762. [Abstract] [Full Text] [PDF]

				F. F. Miranda, K. Teigen, M. Thorolfsson, R. M. Svebak, P. M. Knappskog, T. Flatmark, and A. Martinez Phosphorylation and Mutations of Ser16 in Human Phenylalanine Hydroxylase. KINETIC AND STRUCTURAL EFFECTS J. Biol. Chem., October 18, 2002; 277(43): 40937 - 40943. [Abstract] [Full Text] [PDF]

				N. M. Krohn, S. Yanagisawa, and K. D. Grasser Specificity of the Stimulatory Interaction between Chromosomal HMGB Proteins and the Transcription Factor Dof2 and Its Negative Regulation by Protein Kinase CK2-mediated Phosphorylation J. Biol. Chem., August 30, 2002; 277(36): 32438 - 32444. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/2/1092    most recent M109503200v1 Alert me when this article is cited