Portal de Periódicos da CAPES

Constitutive Phosphorylation of the Acidic Tails of the High Mobility Group 1 Proteins by Casein Kinase II Alters Their Conformation, Stability, and DNA Binding Specificity

1999; Elsevier BV; Volume: 274; Issue: 29 Linguagem: Inglês

10.1016/s0021-9258(19)72624-6

1083-351X

Jacek R. Wiśniewski, Zbigniew Szewczuk, Inga Petry, Ralf Schwanbeck, Ute Renner,

Enzyme Structure and Function

The high mobility group (HMG) 1 and 2 proteins are the most abundant non-histone components of chromosomes. Here, we report that essentially the entire pool of HMG1 proteins inDrosophila embryos and Chironomus cultured cells is phosphorylated at multiple serine residues located within acidic tails of these proteins. The phosphorylation sites match the consensus phosphorylation site of casein kinase II. Electrospray ionization mass spectroscopic analyses revealed thatDrosophila HMGD and Chironomus HMG1a and HMG1b are double-phosphorylated and that Drosophila HMGZ is triple-phosphorylated. The importance of this post-translational modification was studied by comparing some properties of the native andin vitro dephosphorylated proteins. It was found that dephosphorylation affects the conformation of the proteins and decreases their conformational and metabolic stability. Moreover, it weakens binding of the proteins to four-way junction DNA by 2 orders of magnitude, whereas the strength of binding to linear DNA remains unchanged. Based on these observations, we propose that the detected phosphorylation is important for the proper function and turnover rates of these proteins. As the occurrence of acidic tails containing canonical casein kinase II phosphorylation sites is common to diverse HMG and other chromosomal proteins, our results are probably of general significance. The high mobility group (HMG) 1 and 2 proteins are the most abundant non-histone components of chromosomes. Here, we report that essentially the entire pool of HMG1 proteins inDrosophila embryos and Chironomus cultured cells is phosphorylated at multiple serine residues located within acidic tails of these proteins. The phosphorylation sites match the consensus phosphorylation site of casein kinase II. Electrospray ionization mass spectroscopic analyses revealed thatDrosophila HMGD and Chironomus HMG1a and HMG1b are double-phosphorylated and that Drosophila HMGZ is triple-phosphorylated. The importance of this post-translational modification was studied by comparing some properties of the native andin vitro dephosphorylated proteins. It was found that dephosphorylation affects the conformation of the proteins and decreases their conformational and metabolic stability. Moreover, it weakens binding of the proteins to four-way junction DNA by 2 orders of magnitude, whereas the strength of binding to linear DNA remains unchanged. Based on these observations, we propose that the detected phosphorylation is important for the proper function and turnover rates of these proteins. As the occurrence of acidic tails containing canonical casein kinase II phosphorylation sites is common to diverse HMG and other chromosomal proteins, our results are probably of general significance. High mobility group (HMG) 1The abbreviations HMGhigh mobility groupHMG1-BDHMG1 DNA-binding domaincHMG1ChironomusHMG1HMGDDrosophila HMG protein DHMGZDrosophila HMG protein ZCKIIcasein kinase IIHPLChigh pressure liquid chromatography 1The abbreviations HMGhigh mobility groupHMG1-BDHMG1 DNA-binding domaincHMG1ChironomusHMG1HMGDDrosophila HMG protein DHMGZDrosophila HMG protein ZCKIIcasein kinase IIHPLChigh pressure liquid chromatographyproteins are abundant components of chromatin. A subfamily of the HMG proteins containing the HMG1 box domain (HMG1-BD) is widely distributed in eukaryotic cells from yeast to man (for a review, see Ref. 1Bustin M. Reeves R. Prog. Nucleic Acids Res. Mol. Biol. 1996; 54: 35-100Crossref PubMed Google Scholar). The members of this group are thought to have various functions related to modulation of transcription, DNA integration, and recombination. Since these proteins have an ability to induce strong bends and unwind DNA, they are called architectural components of chromatin. The most abundant of the HMG1 box proteins are the HMG1 and HMG2 proteins. They are composed of one or two HMG1-BDs, which are primarily responsible for contacts with DNA. HMG1-BDs are amino- and/or carboxyl-terminally flanked by stretches of positively or negatively charged residues. These regions modulate the binding affinity of HMG1-BDs (2Sheflin L.G. Fucile N.W. Spaulding S.W. Biochemistry. 1993; 32: 3228-3248Crossref Scopus (115) Google Scholar, 3Wiśniewski J.R. Schulze E. J. Biol. Chem. 1994; 269: 10713-10719Abstract Full Text PDF PubMed Google Scholar, 4Štros M. Štokrová J. Thomas J.O. Nucleic Acids Res. 1994; 22: 1044-1051Crossref PubMed Scopus (144) Google Scholar, 5Payet D. Travers A. J. Mol. Biol. 1997; 266: 66-75Crossref PubMed Scopus (82) Google Scholar, 6Ritt C. Grimm R. Fernández S. Alonso J.C. Grasser K.D. Biochemistry. 1998; 37: 2673-2681Crossref PubMed Scopus (49) Google Scholar, 7Štros M. J. Biol. Chem. 1998; 273: 10355-10361Abstract Full Text Full Text PDF PubMed Google Scholar), but do not influence the extent of DNA distortion (8Heyduk E. Heyduk T. Claus P. Wiśniewski J.R. J. Biol. Chem. 1997; 272: 19763-19770Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). Deletion of these regions, in particular those of the negatively charged carboxyl-terminal tails, alters binding specificity of the HMG1 proteins (5Payet D. Travers A. J. Mol. Biol. 1997; 266: 66-75Crossref PubMed Scopus (82) Google Scholar). Moreover, the C-terminal portion of the HMG1 proteins is important for stimulation of transcription (9Aizawa S. Nishino H. Saito K. Kimura K. Shirakawa H. Yoshida M. Biochemistry. 1994; 33: 14690-14695Crossref PubMed Scopus (70) Google Scholar) and nuclear retention (10Shirakawa H. Tanigawa T. Sugiyama S. Kobayashi M. Terashima T. Yoshida K. Arai T. Yoshida M. Biochemistry. 1997; 36: 5992-5999Crossref PubMed Scopus (28) Google Scholar). high mobility group HMG1 DNA-binding domain ChironomusHMG1 Drosophila HMG protein D Drosophila HMG protein Z casein kinase II high pressure liquid chromatography high mobility group HMG1 DNA-binding domain ChironomusHMG1 Drosophila HMG protein D Drosophila HMG protein Z casein kinase II high pressure liquid chromatography Two abundantly expressed proteins of this family were found in each of the dipteran insects Chironomus (cHMG1a and cHMG1b (11Wiśniewski J.R. Schulze E. J. Biol. Chem. 1992; 267: 17170-17177Abstract Full Text PDF PubMed Google Scholar)) andDrosophila (HMGD (12Wagner C.R. Hamana K. Elgin S.C.R. Mol. Cell. Biol. 1992; 12: 1915-1930Crossref PubMed Scopus (73) Google Scholar) and HMGZ (13Ner S.S. Churchill M.E.A. Searles M.A. Travers A.A. Nucleic Acids Res. 1993; 21: 4369-4371Crossref PubMed Scopus (40) Google Scholar)). They are composed of a single HMG1-BD that is C-terminally flanked by a positively and a negatively charged region (for a review, see Ref. 14Wiśniewski J.R. Zool. Pol. 1998; 43: 5-24Google Scholar). Pulse-labeling studies on phosphorylation of the Chironomus proteins showed that they are phosphorylated within their positively charged regions by protein kinase C (15Wiśniewski J.R. Schulze E. Sapetto B. Eur. J. Biochem. 1994; 255: 687-693Crossref Scopus (44) Google Scholar). Phosphorylation of the ChironomusHMG1 proteins by protein kinase C reduces the strength of their binding to DNA and affects their nucleocytoplasmic distribution (15Wiśniewski J.R. Schulze E. Sapetto B. Eur. J. Biochem. 1994; 255: 687-693Crossref Scopus (44) Google Scholar). In mammals, mouse testis-specific HMG1 protein was found to be phosphorylated by protein kinase C (16Alami-Ouahabi N. Veilleux S. Meistrich M.L. Boissonneault G. Mol. Cell. Biol. 1996; 16: 3720-3729Crossref PubMed Scopus (35) Google Scholar). This modification appears to be required for both DNA binding and the topoisomerase I-dependent supercoiling activities of testis-specific HMG (16Alami-Ouahabi N. Veilleux S. Meistrich M.L. Boissonneault G. Mol. Cell. Biol. 1996; 16: 3720-3729Crossref PubMed Scopus (35) Google Scholar). Here, we report that in addition to this regulatory phosphorylation by protein kinase C, HMG1 proteins in insects are constitutively phosphorylated by casein kinase II within their acidic C-terminal tails. This 2- or 3-fold modification alters the conformation, stability, and DNA binding properties of these proteins and therefore appears to be essential for their function. Our results are probably of general importance because putative CKII phosphorylation sites occur in acidic stretches of HMG proteins and other chromosomal proteins The Drosophila HMGD and HMGZ proteins were isolated from 0–18-h embryos of an Oregon R strain. The embryos were collected for the desired period on apple juice-agar plates. They were washed off with water, thoroughly rinsed over a nylon filter, and stored at −80 °C. The frozen embryos were ground together with solid carbon dioxide in a laboratory grinder and thawed after addition of 5% (v/v) HClO4. The resulting suspension was centrifuged at 10,000 × g for 10 min, and the proteins were precipitated from the supernatant with 33% Cl3CCOOH at 0 °C for 1 h. The precipitate was washed with 0.2% HCl in acetone and subsequently two times with pure acetone. The pellets were dried under vacuum and stored at −20 °C. The Chironomus HMG1a and HMG1b proteins were isolated from cultured cells by extraction with 5% (v/v) HClO4 in three freezing-thawing cycles (3Wiśniewski J.R. Schulze E. J. Biol. Chem. 1994; 269: 10713-10719Abstract Full Text PDF PubMed Google Scholar). The cell supernatants were acidified with HCl to 0.35 m, precipitated with 6 volumes of acetone, and dried. Crude extracts were separated on a reverse-phase C18 Zorbax SB-300 column using an H3CCN linear gradient in 0.1% F3CCOOH/H2O as described previously (3Wiśniewski J.R. Schulze E. J. Biol. Chem. 1994; 269: 10713-10719Abstract Full Text PDF PubMed Google Scholar). The isolated proteins were rechromatographed on a second column of the same type and lyophilized. 100 μg of cHMG1a were incubated with 10 units of calf intestine alkaline phosphatase (Calbiochem) in 0.2 m Tris-HCl (pH 9.6) containing 10 mm MgCl2 and 1 mm ZnCl2at 37 °C for 2–3 h. The dephosphorylated protein was purified by HPLC as described above. The dephosphorylated protein was free from native or partially dephosphorylated forms as judged by gel electrophoresis and mass spectroscopy. The proteins were digested with proteinase Glu-C in 25 mmNH4HCO3 at 20 °C for 24 h. The mass ratio of protein to proteinase was 20:1. Tryptic digestion of the C-terminal peptides was carried out in 0.1 m Tris-HCl (pH 7.5) at 37 °C for 3 h. The peptides were resolved on a reverse-phase C18 Zorbax SB-300 column using an H3CCN linear gradient in 0.1% F3CCOOH/H2O as described previously (15Wiśniewski J.R. Schulze E. Sapetto B. Eur. J. Biochem. 1994; 255: 687-693Crossref Scopus (44) Google Scholar,17Schwanbeck R. Wiśniewski J.R. J. Biol. Chem. 1997; 272: 27476-27483Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar). Mass spectra were recorded on a Finnigan MAT TSQ 700 triple-stage quadrupole mass spectrometer equipped with an electrospray ion source. Samples were typically dissolved in methanol/water/acetic acid (47:48:5, v/v/v) solution at a concentration of 50 pmol/μl and introduced into the electrospray needle by mechanical infusion through a microsyringe at a flow rate of 1 μl/min. A potential difference of 4.5 kV was applied between the electrospray needle. Nitrogen gas was used to evaporate the solvent from the charged droplets. At least 20 scans were averaged to obtain each spectrum. Transformations of the resulted spectra were performed using the BioWorks software package (Finnigan MAT). Fluorescence measurements were carried out on a Kontron SFM-25 spectrofluorometer using 10-nm slits for excitation and emission, and the fluorescence values were registered every 2 nm. The buffer used was 50 mm NaCl and 10 mm sodium phosphate (pH 6.9), and the protein concentration was 5 μm. The temperature dependence of the fluorescence intensity was corrected using l-tryptophan as a standard. The fluorescence intensity values were transformed into fraction of unfolded protein (f u) molecules and used to calculate the free energies of unfolding using the relationship shown in Equation 1, ΔGu=−RTlnKu=−RTln[fu/1−fu]Equation 1 where K u is the unfolding constant,R is the gas constant, and T is the temperature. The melting temperature (T m) was atΔG = 0 (18Peace C.N. Shirley B.A. Thomson J.A. Creighton T.E. Protein Structure: A Practical Approach. IRL Press, Oxford1989: 311-330Google Scholar). The protein spectra were recorded and processed on a Kontron Uvikon 940 UV-visible spectrophotometer in 50 mm NaCl and 10 mmsodium phosphate (pH 6.9). The slit with was 2 nm; the scanning speed was 20 nm/min; and the absorption was registered every 0.1 nm. The measurements were done at 20 °C. The protein concentration was 27 μm. The relative tyrosine unfolding was calculated according to Equation 2, Yu=αdp/αnEquation 2 where αdp and αn are the fractional exposures of tyrosine in dephosphorylated and native proteins, respectively. The αdp and αn values were estimated from their second-derivative spectra according to Equations 3and 4, αn=(Rn−Ra)/(Ru−Ra)Equation 3 αdp=(Rdp−Ra)/(Ru−Ra)Equation 4 where R n, R dp, andR u are numerical values related to changes in the tyrosyl microenvironments (19Ragone R. Colona G. Belestrieri C. Servillio L. Irace C. Biochemistry. 1984; 23: 1871-1875Crossref PubMed Scopus (245) Google Scholar) on the native, dephosphorylated, and fully unfolded protein, respectively. R a is the value obtained for the mixture ofN-acetyl-Tyr-NH2 andN-acetyl-Trp-NH2 dissolved in ethylene glycol. The molar ratio of the amino acid derivatives was 2:3, as it has the cHMG1a protein. The R values were calculated from Equation5, R=(A″287−A″283)/(A″295−A″290.5)Equation 5 where A“ is the second-derivative absorbance at 283, 287, 290.5, and 295 nm (19Ragone R. Colona G. Belestrieri C. Servillio L. Irace C. Biochemistry. 1984; 23: 1871-1875Crossref PubMed Scopus (245) Google Scholar). A mixture of native and dephosphorylated cHMG1a proteins (1:1 molar ratio) was digested with chymotrypsin (treated withN α-tosyl-l-lysine chloromethane), thermolysin, or trypsin in 30 mm NaCl and 25 mmTris-HCl (pH 7.5) at 20 °C. The ratio of protein to enzyme was 50:1 (w/w). The reactions were terminated by mixing with an equal volume of 10 m urea solution containing 5% (v/v) acetic acid, 4% (v/v) 2-mercaptoethanol, 10 mm EDTA, 0.2 mmphenylmethylsulfonyl fluoride, and 0.2 mm N α-tosyl-l-lysine chloromethane. The reaction products were separated on urea-acetic acid-Triton X-100–15% polyacrylamide gels (20Zweidler A. Methods Cell Biol. 1978; 17: 223-233Crossref PubMed Scopus (312) Google Scholar). The 32P-labeled four-way junction DNA and AT-rich Chironomus satellite DNA were prepared as described previously (3Wiśniewski J.R. Schulze E. J. Biol. Chem. 1994; 269: 10713-10719Abstract Full Text PDF PubMed Google Scholar, 21Wiśniewski J.R. Ghidelli S. Steuernagel A. J. Biol. Chem. 1994; 269: 29261Abstract Full Text PDF PubMed Google Scholar). Briefly, the proteins were incubated together with labeled DNA in 80 mm NaCl, 1 mm MgCl2, 0.01% bovine serum albumin, 8% glycerol, and 10 mm Tris-HCl (pH 7.9) at 20 °C for 10 min. The complexes of proteins with DNA were run on 6% polyacrylamide gels containing 2.5% (v/v) glycerol, 6.75 mm Tris-HCl, 3.3 mm sodium acetate, and 1 mm EDTA (pH 7.9). The gels were dried and autoradiographed. The mass spectra of the entire HMGD, cHMG1a, and cHMG1b proteins revealed that each protein is twice phosphorylated and carries a single acetyl group because their M r values were ∼202 higher than the M r values calculated from their sequences. Relatively smaller portions of the proteins were found to be monophosphorylated, and a negligible amount possesses no phosphoryl group at all. Deacetylated species were not detected. A typical example of a transformed spectrum of the native cHMG1a protein is shown in Fig. 1 A. TheM r of 13,116 corresponds to acetylated and double-phosphorylated cHMG1a. Two other signals withM r values 13,036 and 12,956 differ by 80 and 160 units, respectively. The spectra of the other HMG proteins show a similar pattern, although their M r values differ from that of cHMG1a. To confirm that the 80-unit shift is due to phosphorylation, the native proteins were treated with alkaline phosphatase. Fig. 1 B shows that the enzyme treatment reduced the molecular weight of the phosphoprotein back to that of acetylated cHMG1a (M r 12,956). To obtain more detailed information about the location of phosphorylation and acetylation sites in the proteins, cHMG1a, cHMG1b, and the mixture of HMGD and HMGZ (Table I) were digested with proteinase Glu-C. The cleavage products were separated by HPLC (data not shown), and the obtained fractions were analyzed by electrospray ionization mass spectrometry. The molecular weights of peptides 13, 20, and 28 (Table I) indicate that the N termini of cHMG1a, cHMG1b, and HMGD are acetylated. The analogous N-terminal segment of HMGZ has not been detected. The spectra revealed that cHMG1a, cHMG1b, and HMGD were phosphorylated twice within their C-terminal peptide, whereas three phosphate groups were localized in the C-terminal peptide of HMGZ (Table I).Table IMass spectrometric identification of peptides obtained by the endoproteinase Glu-C digestion of the HMG proteinsProteinPeptide No.Elution timeaPeptides were separated by reverse-phase HPLC as described under “Experimental Procedures.”ObservedM r bThe M r values were calculated using monoisotopic masses for peptides less than 1000, whereas for larger peptides, average values were used.Sequence positionCalculatedM r bThe M r values were calculated using monoisotopic masses for peptides less than 1000, whereas for larger peptides, average values were used.,cThe values were calculated based on known protein sequence (11-13) and considering the concluded post-translational modification.Concluded post-translational modificationmincHMG1a112701.634–40701.4None214605.564–68605.3None314.52853.891–1132853.8Double phosphorylation4152005.371–902005.2None515543.3110–113544.1Single phosphorylation6162774.091–1132773.8Single phosphorylation716.51004.953–611005.2None8192281.269–902281.5None9205117.669–1135117.3Double phosphorylation10211297.053–631297.5None11271534.921–331534.7None12321564.641–521564.8None13422402.81–202402.8AcetylationcHMG1b1411701.634–40701.4dWe corrected the previously published sequence of cHMG1b (11) by replacing Val-34 with Leu.None15164056.371–1094056.2Double phosphorylation16174349.469–1094348.5Double phosphorylation17261562.921–331562.7None18291776.053–681776.0None19331546.541–521546.7None20422388.01–202388.7AcetylationMixture of HMGD and HMGZ21111074.3103–110eFragment of HMGZ.1074.8Single phosphorylation22204628.069–111fFragment of HMGD.4627.7Double phosphorylation23215984.052–103eFragment of HMGZ.5984.6Double phosphorylation24221779.253–68fFragment of HMGD.1779.0None25251470.921–33fFragment of HMGD.1470.7None26331578.541–52fFragment of HMGD.1578.8None27341560.842–53eFragment of HMGZ.1560.8None28412404.01–20fFragment of HMGD.2404.8Acetylationa Peptides were separated by reverse-phase HPLC as described under “Experimental Procedures.”b The M r values were calculated using monoisotopic masses for peptides less than 1000, whereas for larger peptides, average values were used.c The values were calculated based on known protein sequence (11Wiśniewski J.R. Schulze E. J. Biol. Chem. 1992; 267: 17170-17177Abstract Full Text PDF PubMed Google Scholar, 12Wagner C.R. Hamana K. Elgin S.C.R. Mol. Cell. Biol. 1992; 12: 1915-1930Crossref PubMed Scopus (73) Google Scholar, 13Ner S.S. Churchill M.E.A. Searles M.A. Travers A.A. Nucleic Acids Res. 1993; 21: 4369-4371Crossref PubMed Scopus (40) Google Scholar) and considering the concluded post-translational modification.d We corrected the previously published sequence of cHMG1b (11Wiśniewski J.R. Schulze E. J. Biol. Chem. 1992; 267: 17170-17177Abstract Full Text PDF PubMed Google Scholar) by replacing Val-34 with Leu.e Fragment of HMGZ.f Fragment of HMGD. Open table in a new tab The molecular weight of the C-terminal peptide of cHMG1a (peptide 3;M r 2853.8) corresponds to the molecular weight calculated from its sequence. This indicates that there are two phosphorylations in this fragment. The peptide includes three serines, making it difficult to identify the phosphorylation sites. Subdigestion of the peptide with trypsin resulted in the peptide (MH+1572) that corresponds to the double-phosphorylated fragment 102–113, containing just two serine residues (data not shown). This result clearly places the phosphorylation sites at Ser-103 and Ser-112 (TableII) and is in agreement with the postulated single phosphorylation of peptide 5 (fragment 110–113), containing one serine only.Table IISelected examples of chromosomal proteins that are substrates of casein kinase II Open table in a new tab The C-terminal peptide 21 of HMGZ (fragment 103–110) contains only one site of phosphorylation, allowing us to map the phosphorylation position to Ser-109 (Table II). All mapped phosphorylation sites match consensus substrate sites for CKII (22Glover C.V.C. Hardie G. Hanks S. The Protein Kinase Facts Book: Protein-Serine Kinases. Academic Press Ltd., London1995: 243-248Google Scholar, 23Meisner H. Czech M.P. Hardie G. Hanks S. The Protein Kinase Facts Book: Protein-Serine Kinases. Academic Press Ltd., London1995: 240-242Google Scholar). Since Ser-100 and Ser-101 in HMGZ, Ser-102 and Ser-108 in cHMG1b, and Ser-102 and Ser-110 in HMGD also match the phosphorylation sites for CKII, it is very likely that these sites are phosphorylated in these proteins (Table II). Furthermore, we found that in vitro, the cHMG1a protein is efficiently phosphorylated by human CKII (data not shown). This supports additionally the possibility that insect CKII is responsible for phosphorylation of the proteins. To obtain insight in the biological meaning of the observed constitutive phosphorylation of the HMG1 proteins, we compared the biophysical and biochemical properties of the native (phosphorylated) and alkaline phosphatase-dephosphorylated proteins. The HMG1-BDs of the insect proteins contain three tryptophanyl residues. In previous fluorescence studies using recombinant cHMG1a proteins, we showed that one of these residues, Trp-14, is exposed to solvent (3Wiśniewski J.R. Schulze E. J. Biol. Chem. 1994; 269: 10713-10719Abstract Full Text PDF PubMed Google Scholar). The maximum of the fluorescence emission of this residue is 350 nm. In contrast, the two other Trp residues are buried in the protein interior and exhibit a maximum of fluorescence at 320 nm. In addition, we found that deletion of the acidic tail of the cHMG1a protein results in an increase in fluorescence intensity, suggesting that the C-terminal part of the protein quenches or alters the environment of tryptophan residues (3Wiśniewski J.R. Schulze E. J. Biol. Chem. 1994; 269: 10713-10719Abstract Full Text PDF PubMed Google Scholar). Since double phosphorylation within the acidic tail of the protein might contribute to a specific conformation, we compared the Trp fluorescence spectra of the native and dephosphorylated cHMG1 proteins (Fig. 2 A). The emission spectra of the tryptophans of the native and dephosphorylated proteins exhibited fluorescence maxima at 329 and 337 nm, respectively. Furthermore, a measurable increase in fluorescence intensity was observed as a result of the protein dephosphorylation. The observed red shift of 8 nm suggests strong changes in protein conformation that might mainly involve the spatial arrangement of the C terminus in respect to the HMG1-BD. In thermal denaturation experiments, we observed that the native protein exhibits a higher melting temperature (T m = 46.3 °C) than the dephosphorylated protein (T m = 43.9 °C) (Fig. 2, B and C). The difference in the transition temperatures of 2.4 °C shows that the phosphates importantly contribute to the protein stability. The dephosphorylation of the protein leads to a substantial reduction of the free energy of unfolding (ΔG u). At 20 °C theΔG u values for native and dephosphorylated proteins were 13.6 and 12.8 kJ/mol, respectively (Fig. 2 C,inset). These relatively low values are in good agreement with previously reported moderate conformational stability of the cHMG1a protein (24Wiśniewski J.R. Heßler K. Claus P. Zechel K. Eur. J. Biochem. 1997; 243: 151-159Crossref PubMed Scopus (18) Google Scholar). Second-derivative near-UV absorption spectroscopy is a useful tool for examining the state of tyrosyl residues in proteins also containing tryptophan (25). We used this technique to analyze the changes upon protein dephosphorylation in the microenvironments of tyrosyl residues in cHMG1. Fig. 3 shows the second-derivative spectra of native and dephosphorylated proteins. A value of relative change upon protein dephosphorylation of the solvent exposition of tyrosyl residues was calculated. The valueY u = 2.1 suggests a 2-fold increase in tyrosyl residue exposition in the dephosphorylated protein compared with its native form. This result is in good agreement with the spectral properties of both forms observed in fluorescence emission spectra. Since Tyr-11 is located adjacent to Trp-14, it is likely that perturbations in the tyrosine component are mainly due to changes within the microenvironment of Tyr-11. Proteolytic enzymes are useful tools in the detection and characterization of changes in the tertiary structure of proteins. The mixture of native and dephosphorylated cHMG1a proteins was partially digested by chymotrypsin, thermolysin, and trypsin (Fig.4). In the presence of chymotrypsin and thermolysin, the dephosphorylated protein was digested more rapidly than the native protein (Fig. 4, B, C, andE). Chymotrypsin specifically hydrolyzes peptide bonds at hydrophobic residues, whereas thermolysin does it preferentially; therefore, these data suggest an increased exposition of apolar residues upon dephosphorylation. These results confirm our spectroscopic data showing an increased exposition of tryptophanyl and tyrosyl residues in the dephosphorylated cHMG1a protein. In contrast, trypsin, which specifically cuts peptide bonds at carboxyl termini of arginyl and lysinyl residues, digested the native protein form more rapidly (Fig. 4, D and E). Thus, it is likely that the accessibility of the basic residues to trypsin also changes upon protein phosphorylation. HMG1-BD proteins bind preferentially to the DNAs in a non-B conformation. This includes intrinsically prebent, cruciform, bulged, and cis-platinated DNAs. Previously, we have demonstrated that the acidic tail inhibits the binding affinity of the protein for linear and four-way junction DNAs (3Wiśniewski J.R. Schulze E. J. Biol. Chem. 1994; 269: 10713-10719Abstract Full Text PDF PubMed Google Scholar) and that phosphorylation at protein kinase C sites additionally weakens the interaction of cHMG1a and cHMG1b with DNA (15Wiśniewski J.R. Schulze E. Sapetto B. Eur. J. Biochem. 1994; 255: 687-693Crossref Scopus (44) Google Scholar). More recently, Payet and Travers (5Payet D. Travers A. J. Mol. Biol. 1997; 266: 66-75Crossref PubMed Scopus (82) Google Scholar) demonstrated that the presence of the acidic tail in the recombinant HMGD protein is essential for its structure-specific recognition of such DNAs. Because the phosphorylation might contribute to protein specificity, we compared the binding properties of the native and dephosphorylated cHMG1a proteins using four-way junction and linear AT-rich DNAs, which possess multiple binding sites. The native proteins produced two shifts with the four-way junction DNA. The one with the higher mobility reflects interaction with the central portion of the junction (Fig. 5 C,arrowhead), whereas the second more slowly migrating complex (arrow) corresponds to protein binding to the arm of the junction (26Hill A.D. Reeves R. Nucleic Acids Res. 1997; 25: 3523-3531Crossref PubMed Scopus (88) Google Scholar). Both protein-DNA complexes appeared for the first time at protein concentrations in the range of 30–100 nm. This suggests a similar binding affinity of the protein for both duplex arm(s) and the junction. The protein dephosphorylation essentially increased the binding affinity of the protein for the junction (Fig.5 D). At 1 nm protein, essentially the entire DNA was bound. Thus, an ∼2 orders of magnitude increase in protein affinity for the junction was found, but not for the arm. In experiments in which the binding of both proteins to linear DNA was compared, apparently no difference in the binding affinity was found (Fig. 5, A and B); however, an altered binding specificity was observed. The native protein was able to bind at different sites of the DNA, whereas the dephosphorylated protein preferentially bind to a single site on the DNA. Our results show that Drosophila andChironomus HMG1 proteins are constitutively phosphorylated within their C-terminal tails by CKII. This phosphorylation is important for their proper folding, thermodynamic and metabolic stability, and DNA binding specificity. CKII is a structurally and functionally conserved enzyme that is widely distributed among eukaryotic organisms (22Glover C.V.C. Hardie G. Hanks S. The Protein Kinase Facts Book: Protein-Serine Kinases. Academic Press Ltd., London1995: 243-248Google Scholar). However, the biological role of this kinase is only poorly understood; it appears to be involved in the modulation of properties of some transcription factors and as well as in the regulation of cell proliferation (23Meisner H. Czech M.P. Hardie G. Hanks S. The Protein Kinase Facts Book: Protein-Serine Kinases. Academic Press Ltd., London1995: 240-242Google Scholar). Furthermore, the enzyme is essential for viability ofSaccharomyces cerevisiae (27Padmanabha R. Chen-Wu J.L. Hanna D.E. Glover C.V.C. Mol. Cell. Biol. 1990; 10: 4089-4099Crossref PubMed Scopus (306) Google Scholar). This stresses the importance of this kinase in eukaryotic cells. In Drosophila and Chironomus cells, almost the entire population of HMG1 proteins was found to be double- or triple (HMGZ)-phosphorylated. Since only small amounts of partially dephosphorylated species were detected, it appears that the modification of the acidic tails of the HMG proteins is important for their proper function. The extent of phosphorylation ofDrosophila HMG1 proteins remains constant during the entire development 2J. R. Wiśniewski and U. Renner, unpublished results. ; and therefore, it is likely that the modification of these proteins by CKII is constitutive. The phosphorylation of the tails changes the DNA binding properties of these proteins with respect to their structure specificity. The phosphorylation of chromosomal proteins by CKII appears to be a common property found in evolutionarily distant organisms. In plant HMG proteins (28Grasser K.D. Maier U.-G. Feix G. Biochem. Biophys. Res. Commun. 1989; 162: 256-261Crossref Scopus (28) Google Scholar) and Drosophila protein D1 (similar to HMGI/Y) (29Glover C.V.C. Shelton E.R. Brutlag D.L. J. Biol. Chem. 1983; 258: 3258-3265Abstract Full Text PDF PubMed Google Scholar), substrates for CKII isolated from these organisms were found. Members of the HMGI/Y family were found to be phosphorylated in vivo by CKII within their acidic C-terminal tails (Table II) (30Palvimo J. Linnala-Kankkunen A. FEBS Lett. 1989; 257: 101-104Crossref PubMed Scopus (49) Google Scholar,31Ferranti P. Malorni A. Marino G. Pucci P. Goodwin G.H. Manfioletti G. Giancotti V. J. Biol. Chem. 1992; 267: 22486-22489Abstract Full Text PDF PubMed Google Scholar). Multiple phosphorylation sites were found in Drosophilaheterochomatin-associated protein (HP-1) (32Eissenberg J.C. Ge Y. Harnett T. J. Biol. Chem. 1994; 269: 21315-21321Abstract Full Text PDF PubMed Google Scholar). Inspection of the primary structure of this protein reveals at least two possible CKII phosphorylation sites (Table II). Multiple putative CKII sites are also present in the acidic tails of structurally related HP-1 proteins in mammals (protein M31). Moreover, HMG1 box-containing, structure-specific recognition proteins, upstream binding factor, and plant HMG1 proteins possess long acidic tails with canonical phosphorylation sites of CKII (Table II). What might be the functional significance of phosphorylation of HMG and other chromosomal proteins by CKII? The levels of the HMG proteins are variable between different types of cells (33Mosevitsky M.L. Novitskaya V.A. Iogannsen M.G. Zabezhinsky M.A. Eur. J. Biochem. 1989; 185: 303-310Crossref PubMed Scopus (113) Google Scholar). Usually undifferentiated and rapidly proliferating cells contain higher amounts of these proteins compared with terminally differentiated cells. However, the biological meaning of these differences as well as the mechanisms regulating the titers of the HMG proteins are not clearly understood. In the insect systems of Drosophila andChironomus, HMG proteins in vivo are metabolically relatively stable, and their turnover rates extend over many cell generations (34Ghidelli S. Claus P. Thies G. Wiśniewski J.R. Chromosoma (Berl.). 1997; 105: 369-379Crossref PubMed Google Scholar). The entire population of HMG1 proteins is CKII-phosphorylated, suggesting that intermediate or dephosphorylated forms are only short-living. The results presented show that dephosphorylation of cHMG1a causes a partial denaturation and reduction of the stability of the protein against proteinase in vitro. Taking these facts together, we suggest that CKII phosphorylation is essential for the metabolic stability of these HMG1 proteins in vivo; the dephosphorylation of these proteins might be a part of the mechanism regulating their titer, in particular, during cell differentiation. However, the phosphorylation of the acidic tails of the HMG proteins seems to be widely distributed; some groups of these proteins appear to be not modified by CKII. The HMG1 and HMG2 proteins containing two HMG1-BDs from vertebrate organisms, Drosophila DSP-1 (dorsal-switch protein1), and yeast ACP2 (acidic protein2) do not possess canonical CKII phosphorylation sites in their C-terminal acidic tails. In the HMG14/17 family, only the HMG14 protein is a substrate for CKII, whereas the HMG17 protein is not. Despite that fact that these proteins are very similar in their primary structures, they were localized to distinct regions of chromatin (35Postnikov Y.V. Herrera J.E. Hock R. Scheer U. Bustin M. J. Mol. Biol. 1997; 274: 454-465Crossref PubMed Scopus (51) Google Scholar). This selectivity might be due to phosphorylation of the acidic tail of HMG14. Recombinant technology (and in particular, the possibility of producing eukaryotic proteins in bacteria) has revolutionized biochemistry. However, many of these proteins, such as those described in this work, are post-translationally modified in eukaryotic cells. Unfortunately, in bacteria, these proteins are not phosphorylated and acetylated. Because these modifications are important, such proteins should be phosphorylated in vitro prior to biochemical analyses. This is easily to perform2 since CKII preparations are commercially available. In many proteins, the CKII sites are located several residues from C termini. This offers the possibility of end labeling such proteins. Their conformation and interaction with DNA could be analyzed by the protein footprinting method without the introduction of artificial phosphorylation sites (36Frank O. Schwanbeck R. Wiśniewski J.R. J. Biol. Chem. 1998; 273: 20015-20020Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar). The data presented in this work demonstrate the importance of the constitutive phosphorylation of a group of HMG proteins. This modification appears to be essential for the function of these proteins. Further modifications of HMG proteins, including phosphorylation by protein kinase C (15Wiśniewski J.R. Schulze E. Sapetto B. Eur. J. Biochem. 1994; 255: 687-693Crossref Scopus (44) Google Scholar, 16Alami-Ouahabi N. Veilleux S. Meistrich M.L. Boissonneault G. Mol. Cell. Biol. 1996; 16: 3720-3729Crossref PubMed Scopus (35) Google Scholar), Cdc2 kinase (17Schwanbeck R. Wiśniewski J.R. J. Biol. Chem. 1997; 272: 27476-27483Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar, 37Reeves R. Langan T.A. Nissen M.S. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 1671-1675Crossref PubMed Scopus (115) Google Scholar, 38Nissen M.S. Langan T.A. Reeves R. J. Biol. Chem. 1991; 266: 19945-19952Abstract Full Text PDF PubMed Google Scholar), and mitogen-activated protein kinase (17Schwanbeck R. Wiśniewski J.R. J. Biol. Chem. 1997; 272: 27476-27483Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar) and ADP-ribosylation (41Giancotti V. Bandiera A. Sindici C. Perissin L. Crane-Robinson C. Biochem. J. 1996; 317: 865-870Crossref PubMed Scopus (20) Google Scholar), facultatively change fractions of these proteins at particular events of cell life, such as mitosis, differentiation, and apoptosis. We thank Dr. J. Ziółkowski (University of Wrocław) and Dr. U. Grossbach (University of Göttingen) for the interest in and the support of this work. Dr. M. A. Schäfer (University of Göttingen) is gratefully acknowledged for the continuous supply of Drosophilaflies.