Phosphorylation of Heterochromatin Protein 1 by Casein Kinase II Is Required for Efficient Heterochromatin Binding inDrosophila *

Heterochromatin-associated protein 1 (HP1) is a nonhistone chromosomal protein with a dose-dependent effect on heterochromatin mediated position-effect silencing. It is multiply phosphorylated in vivo. Hyperphosphorylation of HP1 is correlated with heterochromatin assembly. We report here that HP1 is phosphorylated by casein kinase II in vivo at three serine residues located at the N and C termini of the protein. Alanine substitution mutations in the casein kinase II target phosphorylation sites dramatically reduce the heterochromatin binding activity of HP1, whereas glutamate substitution mutations, which mimic the charge contributions of phosphorylated serine, have apparently wild-type binding activity. We propose that phosphorylation of HP1 promotes protein-protein interaction between HP1 and target binding proteins in heterochromatin.

Heterochromatin-associated protein 1 (HP1) is a nonhistone chromosomal protein with a dose-dependent effect on heterochromatin mediated position-effect silencing. It is multiply phosphorylated in vivo. Hyperphosphorylation of HP1 is correlated with heterochromatin assembly. We report here that HP1 is phosphorylated by casein kinase II in vivo at three serine residues located at the N and C termini of the protein. Alanine substitution mutations in the casein kinase II target phosphorylation sites dramatically reduce the heterochromatin binding activity of HP1, whereas glutamate substitution mutations, which mimic the charge contributions of phosphorylated serine, have apparently wild-type binding activity. We propose that phosphorylation of HP1 promotes protein-protein interaction between HP1 and target binding proteins in heterochromatin.
In higher eukaryotes, chromosomes are composed of euchromatin and heterochromatin. Heterochromatin is distinguished from euchromatin in that it stays condensed throughout the cell cycle (1). In addition, it replicates late in S phase, is enriched in repetitive DNA, and is relatively poor in classical genes (2)(3)(4).
In Drosophila, when a euchromatic gene is placed next to or within heterochromatin by chromosomal rearrangement or transposition, the euchromatic gene usually undergoes cellspecific silencing, called "position-effect variegation" (5). This process indicates that heterochromatin interferes with euchromatic gene expression, which biochemical data suggest is caused by the acquisition of a distinct chromatin structure with reduced accessibility to DNA-binding proteins (6, 7). More than 50 different position-effect variegation regulators have been identified, including genes encoding chromatin proteins and protein modifiers (8). One such regulator encodes heterochromatin-associated protein 1 (HP1), 1 a nonhistone chromosomal protein enriched in the pericentric heterochromatin of interphase nuclei (9,10). HP1 exerts dosage-dependent effects on position-effect variegation (11,12). HP1 localization during the cell cycle is complex, perhaps reflecting the changes in chromosome structures that accompany the cell cycle (13).
Current models suggest that HP1 functions as part of a chromosomal protein complex (14). Proteins thought to bind to HP1-like proteins include transcriptional intermediary factors (15), lamin B receptors (16), and origin recognition complex proteins (17,18).
HP1 is multiply phosphorylated in vivo (19). Using twodimensional gel electrophoresis, HP1 can be resolved into as many as eight charged isoforms. HP1 phosphorylation occurs predominantly at serines and threonines, and increased phosphorylation of HP1 is correlated with heterochromatin assembly during development. Phosphorylation may be correlated with the oligomeric state of HP1 in vivo (18). Phosphorylation of chromosomal proteins is implicated in the regulation of several nuclear functions. For example, histones are multiply phosphorylated during mitotic and developmentally programmed chromatin condensation (20), specific forms of phosphorylated histone H1 are correlated with specific heterochromatic satellite DNA sequences (21), and several transcription factors are regulated by phosphorylation (22).
As a first step toward defining the role of HP1 phosphorylation in heterochromatin assembly and position-effect variegation silencing, we mapped three of the phosphorylation sites at the N-and C-terminal domains of HP1. We present biochemical evidence for casein kinase II (CKII) phosphorylation of HP1 in vitro and in vivo. Alanine substitution mutation in the Nterminal CKII target site dramatically reduces heterochromatin binding activity of HP1, whereas glutamate substitutions mutations at either or both of the N-and C-terminal sites have apparently wild-type binding activity, suggesting that CKII phosphorylation is required for heterochromatin binding. We propose that phosphorylation of HP1 promotes protein-protein interaction between HP1 and target binding proteins within heterochromatin.

EXPERIMENTAL PROCEDURES
Fly Culture and Tissue Preparation-Fly stocks were maintained at room temperature on standard cornmeal-yeast-sucrose-agar medium containing methylparaben as a mold inhibitor. Third-instar larvae were collected immediately before dissection and kept on ice until dissection.
Expression and Purification of Recombinant HP1-A XbaI-BamHI fragment containing Drosophila HP1 cDNA was cloned into expression vector pET11a, and the recombinant plasmid was transformed into Escherichia coli BL21(DE3) cells. A single colony of transformed cells was grown in LB medium until A 600 ϭ 0.6 -1.0. Expression of HP1 was induced by 1 mM isopropyl-1-thio-␤-D-galactopyranoside at 30°C for 2-3 h. Cells were spun down at 5,000 ϫ g for 5 min in 4°C, and the cell pellet was frozen and stored at Ϫ70°C. For protein purification, the pellet was thawed on ice and suspended in 40 ml of lysis buffer (100 mM Tris-HCl, pH 7.3, 4 mM EDTA, 0.4 mM EGTA, 4 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride). All subsequent steps were done on ice or at 4°C unless indicated otherwise. 24 mg of lysozyme was added to the sample and incubated on ice for 20 min. Cells were lysed either by freezing and thawing or by sonication. NaCl was added to a final concentration of 0.3 M to get maximum HP1 solubilization. After centrifugation at 18,000 rpm for 30 min in a Sorvall SS-34 rotor, the supernatant was diluted 3-fold and applied to a DEAE-Sepharose col-umn. The DEAE column was eluted with 0.4 M NaCl. Eluted fractions containing HP1 were pooled, diluted 4-fold, and applied to a 6-ml Resource Q FPLC (Amersham Pharmacia Biotech FPLC system) column. The Resource Q column was developed with a 0.1-0.4 M NaCl gradient, and HP1 was eluted in the 0.2-0.25 M fractions. These fractions were pooled and loaded on an S-200 Sephacryl FPLC gel filtration column. HP1 was collected in an elution volume corresponding to a molecular mass of ϳ40 kDa. HP1-containing fractions were pooled again and applied to a 1-ml Mono Q FPLC column. The Mono Q column was developed with a 0.1-0.4 M NaCl gradient, and HP1 was purified as a single sharp, symmetrical peak at ϳ0.25 M NaCl. At each chromatographic step, fractions containing HP1 were determined by 12% SDS-PAGE and Coomassie Brilliant Blue staining. Relative purification at each step is shown in Fig. 1.
Preparation of Drosophila Nuclear Extract-Five g of frozen 0 -6-hold embryos (gift of Dr. S. C. R. Elgin) were used to make the nuclear extract. Drosophila embryo nuclei were prepared according to Wu et al. (23). The nuclear extract was made as described (9), with modifications. Briefly, nuclei were suspended in 4 ml of extraction buffer (10 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1 mM ␤-mecaptoethanol, 0.25 M sucrose, 1 mM phenylmethylsulfonyl fluoride, 0.4% Triton X-100) and homogenized at 4°C with a glass-Teflon homogenizer. Homogenate was centrifuged in an Eppendorf 5415 C centrifuge at top speed for 10 min in 4°C. The supernatant was aliquotted and stored at Ϫ70°C as nuclear extract. Protein concentration of the extract was found to be about 1 mg/ml with a BCA protein assay (Pierce).
In Vitro Phosphorylation Assay-In a 20-l reaction mixture, purified recombinant HP1 (1 g/l) was mixed with 1 l Drosophila embryo nuclear extract and 2 Ci of [␥-32 P]ATP at 100 Ci/mol in 20 mM Tris-HCl, pH 7.4, 10 mM MgCl 2 , and 100 mM KCl or 100 mM NaCl (both salts work equally well). The reaction proceeded at room temperature for 20 min and was stopped by adding 5 l of 5ϫ SDS-PAGE sample buffer and boiling for 3 min. Products were separated by a 12% SDS-PAGE, and radiolabeled proteins were visualized by exposing the gel to x-ray film or imaging using a Molecular Dynamics PhosphorImager.
Two-dimensional Gel Analysis and Western Blotting-[ 32 P]Phosphate metabolic labeling of larval tissue and separation of in vivo labeled HP1 by two-dimensional gel electrophoresis was done as described (19). For analysis of in vitro phosphorylated HP1, protein was precipitated with 10% trichloroacetic acid, washed successively with 100% ethanol and chloroform:methanol (1:1), dried, and resuspended in 80 ml of IEF sample buffer (9.5 M urea, 2% Triton X-100, 5% ␤-mercaptoethanol, 1.6% Bio-Lyte 5-8 ampholytes (Bio-Rad), 0.4% Bio-Lyte 3-10 ampholytes (Bio-Rad)) before being applied to an isoelectric focusing gel. After two-dimensional gel electrophoresis, proteins were transferred to nitrocellulose (Millipore) membrane using a Bio-Rad mini-Trans-Blot electrophoretic transfer cell according to manufacturer's instructions. After transfer, the membrane was blocked with 1% bovine serum albumin (fraction V, Sigma) in TBST buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.05% Tween 20) at room temperature for 30 min. The anti-HP1 serum was a gift of Drs. R. F. Clark and S. C. R. Elgin (Washington University, St. Louis, MO) and was used at a 1:10,000 dilution in TBST buffer. The secondary antibody was an anti-rabbit IgG-alkaline phosphatase conjugate (Promega) and was used at a 1:7,500 dilution in TBST buffer. Detection was with 5-bromo-4-chloro-3-indoyl phosphate and nitro blue tetrazolium (Promega) as described (24). To detect radiolabeled protein, Western blots were either subjected to autoradiography using x-ray film (Kodak XAR-5) or imaged using a Molecular Dynamics PhosphorImager.
Trypsin Digestion and Radio-peptide Mapping-In situ trypsin digestion of protein immobilized on nitrocellulose membrane was done according to Fernandez et al. (25). Briefly, the radiolabeled proteins were transferred electrophoretically to nitrocellulose membrane after SDS-PAGE. Radiolabeled HP1 was visualized by Western blotting and autoradiography. The region of membrane containing immunolocalized HP1 was excised with a surgical blade and blocked in 1.0 ml of 0.5% polyvinylpyrrolidone, M r 40,000 (PVP-40, Sigma), 100 mM acetic acid (w/v) in a 1.5-ml Eppendorf tube at 37°C for 30 min. The sample was washed six times with 1 ml of Milli-Q water, cut into approximately 1 x 1-mm squares, and returned to the same tube. A solution of 0.1 M NH 4 HCO 3 and 2 M urea was added to just immerse the membrane strips. Trypsin (0.5 g/g of substrate) was added, and the digestion was incubated overnight at 37°C. Following the digestion, the sample was centrifuged at top speed in an Eppendorf centrifuge for 5 min. The supernatant was stored at Ϫ20°C before peptide mapping. Tryptic peptide mapping was done on a 40% PAGE under alkaline conditions as described (26,27). The gel was dried and subjected to autoradiography or PhosphorImager analysis to visualize the phosphopeptides.
Sequencing of Radiolabeled Tryptic Peptides and Identification of Radioactive Amino Acids-Trypic phosphopeptide bands on the dried PAGE gel were excised with a surgical blade. Peptide was eluted from the gel slices with Milli-Q water. Peptides in the extract were purified and desalted with a C-18 reverse phase HPLC column (130A Separation System, Applied Biosystems) and subjected to protein sequencing (477A Protein Sequencer, Applied Biosystems). Cleaved phenylthiohydantoin derivatives from each cycle were collected, and radioactive amino acid residues were determined by: 1) counting in a scintillation counter (1500 Tri-Carb liquid scintillation analyzer, Packard Instrument Co.) for the radioactivity and 2) concentrating to Ͻ10 l and spotting onto a 0.1-mm cellulose thin layer electrophoresis plate (EM Separations) for autoradiography.
Anti-CKII Immunodepletion of Extracts-Antibody depletion of CKII was done according to Birnbaum et al. (28) with modifications. 5 l of nuclear extract was incubated with 4 l of rabbit antiserum against native Drosophila CKII, 16 l of NET buffer (140 mM NaCl, 50 mM Tris-HCl, pH 8.0, 5 mM EDTA, and 0.05% Nonidet P-40), and 25 g of bovine serum albumin at 4°C for 1 h. 25 l of protein A-acrylic beads suspension (Sigma) was added, and the mixture was incubated at 4°C for an additional 20 min. Following a brief centrifugation, the supernatant was used in the in vitro phosphorylation assay. As a control, 5 l of the extract was treated with rabbit nonimmune serum (no preimmune serum remains), 2 and the sample was also used in the in vitro phosphorylation assay.
Site-directed Mutagenesis and Construction of the Transgene Expression Vector-Site-directed mutagenesis was performed using the Transformer site-directed mutagenesis kit (CLONTECH) according to manufacturer's instructions. Mutagenesis was done in a pBluescript SK(ϩ) plasmid (Stratagene) with Drosophila HP1 cDNA inserted between the XbaI and BamHI sites. Mutant cDNA was then excised with SacI and BglII and cloned into vector SP1 (29). A KpnI fragment from the SP1 vector was cloned into transformation vector pV␤206 in which HP1 cDNA is fused downstream of and in frame with E. coli lacZ under the Drosophila Hsp70 heat-shock promoter (29), inserted in the P-element vector, pYC1.8 (30).
Expression and Localization of ␤-Galactosidase Fusion Protein-Germline transformation was performed by injecting v 36F ;ry 506 embryos with each of the constructs together with the helper plasmid p25.7wc (31), essentially as described (32). G o survivors were mated to v 36F ;ry 506 flies. F 1 adults were screened for rescue of the vermilion eye color. Somatic transformation was performed essentially as described (33,34). Host strain in both cases was a v 36F ;ry 506 stock. Third-instar larvae were collected from each of the germline and somatic transformants carrying the pv␤206 mutant fusion constructs, heat-shocked at 37°C for 30 min, and recovered at room temperature for 1 h. Salivary glands were dissected and stained with X-gal (5-bromo-4-chloro-3-indolyl ␤-Dgalactopyranoside) as described (35). Stained tissue was placed on a microscope slide and mounted in 95% glycerol, 5% phosphate-buffered saline.

Drosophila Embryo Nuclear Extract Phosphorylates Recombinant HP1 in Vitro-HP1 is multiply phosphorylated in vivo,
and hyperphosphorylation is correlated with heterochromatin assembly (19). As a first step toward understanding the functional significance of HP1 phosphorylation, we mapped phosphorylation sites in HP1. Our strategy was to use recombinant HP1 (rHP1) as a substrate for phosphorylation by Drosophila nuclear extract in vitro and then map the sites of in vitro phosphorylation corresponding to sites used in vivo. Drosophila HP1 cDNA was expressed in E. coli and purified to Ͼ95% homogeneity (Fig. 1). To determine whether rHP1 is phosphorylated in bacteria, we compared its mobility in two-dimensional gel electrophoresis to that of HP1 from Drosophila (dHP1). Multiple dHP1 isoforms are detected in Western blots of total Drosophila protein after two-dimensional gel electrophoresis (Ref. 19; Fig. 2A, top). Similar Western blots of rHP1 revealed a single major HP1 isoform ( Fig. 2A, middle). When rHP1 is added to total Drosophila protein and co-electrophoresed, the rHP1 migrates at the extreme basic end of the dHP1 pattern ( Fig. 2A, bottom), consistent with unmodified HP1.
Hyperphosphorylated isoforms of HP1 begin to appear by 2 h of embryonic development (19). Therefore, we used a nuclear extract from 0 -6-h Drosophila embryos to phosphorylate rHP1 in vitro. When rHP1 is added to nuclear extract in the presence of [ 32 P]ATP, rHP1 is efficiently radiolabeled. Co-electrophoresis with total Drosophila protein demonstrates that radiola-beled rHP1 co-migrates with the second and third most basic isoforms of dHP1 (Fig. 2B), consistent with mono-and diphosphorylated protein. To verify that sites of rHP1 phosphorylation by embryo extract correspond to sites of dHP1 phosphorylation in vivo, the major tryptic phosphopeptides of in vitro labeled rHP1 were compared with those of dHP1 metabolically labeled with [ 32 P]orthophosphate. In vitro phosphorylated rHP1 was fractionated by SDS-PAGE and treated with trypsin. Separately, total protein from larval tissue metabolically labeled with [ 32 P]orthophosphate was fractionated by two-dimensional gel electrophoresis, and dHP1 protein was digested with trypsin. Tryptic peptides from both preparations were resolved electrophoretically on a 40% alkaline polyacrylamide gel. Three major radioactive bands were common to both preparations (Fig. 2C). Thus, these major tryptic phosphopeptides contain targets of phosphorylation used in vivo.
HP1 Is Phosphorylated by CKII-The fact that embryo extract can efficiently phosphorylate rHP1 at sites used in vivo allowed us to map the residues phosphorylated by the extract. The major tryptic peptides from in vitro phosphorylated rHP1 (peptides 1, 2, and 3 in Fig. 2C) were size-fractionated by 40% alkaline PAGE, and each phosphopeptide was separately eluted from the gel, desalted and purified by reverse phase HPLC, and subjected to amino acid sequencing. As each amino acid was determined, the eluted amino acid derivative was assayed for radioactivity. Results are summarized in Fig. 2D. Peptides 1 and 2 are fragments of the HP1 N terminus and contain the same phosphorylated serine residue, with peptide 2 having one more undigested lysine at the N terminus. Peptide 3 is a C-terminal fragment containing a major phosphoserine and a minor one three residues away on its N-terminal side. The two most terminal serines occur within CKII recognition motifs, which is S(T)XXE(D). These are the only consensus CKII site motifs in HP1. In peptide 3, the N-terminal serine resembles a CKII target only after its C-terminal partner is phosphorylated and becomes acidic.
To confirm the identity of the HP1 kinase activity in embryo extract, we tested the sensitivity of the kinase to inhibitors and stimulators of CKII. rHP1 phosphorylation by embryo extract is stimulated by spermine and inhibited by heparin (Fig. 3). Both of these effects are also seen with CKII (36,37). As further confirmation, we tested whether antibody to Drosophila CKII (a gift of Dr. C. V. C. Glover III) inhibited in vitro phosphorylation. As shown in Fig. 3, the CKII antisera almost completely eliminates HP1 phosphorylation by embryo extract. Taken together, the results strongly implicate CKII in HP1 phosphorylation in vitro and in vivo.
Mutations in CKII Sites Affect Heterochromatin Binding-To test the functional significance of HP1 phosphorylation, the serines in the CKII target sites were replaced by either alanine (to prevent phosphorylation) or glutamate (to mimic the spatial and charge contributions of phosphorylated serine; Refs. 38 and 39). Five mutations were assayed (Fig. 4): replacement of the N-terminal serine by alanine (S15A) or by glutamate (S15E), replacement of the very C-terminal serine by alanine (S202A) or by glutamate (S202E), and a combination of both Ser 3 Ala mutations (S15A,S202A). The mutant HP1 cDNA were fused downstream from, and in frame with, E. coli lacZ, and the fusion cDNA was placed under the control of the Drosophila Hsp70 heat shock promoter. P-element-mediated germline transformations were generated for the two glutamate substituted mutants. Transgenic larvae were subjected to heat shock, and larval polytene tissue was stained with X-gal to localize the fusion protein. ␤-Galactosidase alone is a cytoplasmic protein when expressed in Drosophila cells (Fig. 4a). As reported previously (29,40), full-length HP1⅐␤-galactosidase fusion protein targets ␤-galactosidase activity to the nucleus and decorates the heterochromatin with ␤-galactosidase; this results in blue-staining nuclei with X-gal, with one or two intensely staining dark blue spots corresponding to the pericentric heterochromatin (Fig. 4c). For both the S15E and S202E mutations, X-gal staining showed that these two mutant HP1 fusion proteins resulted in nuclear localization and heterochromatin binding indistinguishable from wild-type HP1 fusions (Compare Fig. 4, e and g-c; see also Refs. 29 and 40).
For the S15A, S202A, and S15A,S202A mutations, comparable efforts to obtain germline transformants resulted in no transgenic lines (occasionally, transformed phaerate adults were observed, but these failed to hatch or died shortly after hatching). To assay fusion protein targeting for these mutations, we resorted to somatic transformation. Plasmid DNA injected into early embryos stably endures in somatic tissues through the third-instar larval stage, displaying correct tissuespecific expression (33,34). We had employed somatic transformation in previous studies to assay HP1⅐␤-galactosidase fusions (29,40). Somatically transformed third-instar larvae were heat-shocked, and polytene tissues were dissected out and stained with X-gal. By this assay, the S202A mutant appeared similar to the wild type (Fig. 4f), but the S15A mutant showed decreased heterochromatin binding activity visible only in highly expressing cells (Fig. 4d). Only in overstained nuclei did we observe the darkly staining spots that represent heterochromatin binding. For the double mutation (S15A,S202A), the loss of heterochromatin targeting was even more dramatic. Although the fusion protein still appeared to concentrate in the nuclei, it no longer bound to heterochromatin (Fig. 4h). These results demonstrate a requirement for CKII phosphorylation for efficient heterochromatin binding.

Drosophila HP1 Is Phosphorylated by CKII in Vivo-
The methods we used to identify HP1 phosphorylation sites involved direct comparison of the in vivo and in vitro tryptic peptide map by high concentration PAGE, rHP1 phosphopeptide sequencing, and radioactivity detection of each amino acid derivative. This strategy has been used to identify phosphorylation sites on other proteins (26,27,41,42). For all three sites common to phosphorylated rHP1 and dHP1, the targets are good fits to CKII consensus motifs, which, together with the sensitivity of rHP1 phosphorylation to spermine, heparin, and anti-Drosophila CKII serum, strongly suggests that HP1 is a substrate for CKII. CKII is an ubiquitous cyclic nucleotideindependent protein kinase that appears not to directly mediate known signaling pathways (43). CKII activity has been found to increase in response to some mitogens, and its substrates include a number of transcription factors involved in growth control (44). Because CKII is found both in the nucleus and the cytoplasm (36), and because we found that alanine substitution had no effect on nuclear targeting, HP1 phosphorylation by CKII could occur in either compartment.
CKII consensus target sites are found at the N and/or C terminus of HP1 homologs from Drosophila virilis, Schizosaccaromyces pombe, mealybug, mouse, and human. Not all HP1 homologs have CKII targets at both ends (some have neither), but in several such cases the homologous position is occupied by glutamate. Little or nothing is known about the functional homology between Drosophila melanogaster HP1 and its structural homologs in other species, but such apparent structural conservation suggests functional conservation. Nevertheless, the data presented here showing that CKII phosphorylation is required for efficient heterochromatin targeting by the unique D. melanogaster HP1 suggest that such structural conservation is likely to be functionally significant.
The only detailed structural information for any HP1 homo-  (29) with wild-type CKII sites, showing heterochromatin binding; d, S15A⅐␤-galactosidase fusion protein, showing nuclei from highly expressing cells in which reduced heterochromatin binding can sometimes be seen; e, S15E⅐␤-galactosidase fusion protein; f, S202A⅐␤-galactosidase fusion protein; g, S202E⅐␤-galactosidase fusion protein; h, S15A,S202A⅐␤-galactosidase fusion protein. In the nuclei of highly expressing cells, a faint spot of staining is sometimes seen (arrowhead in the right nucleus).
log is a solution NMR peptide structure based on the N-terminal chromo domain of a single mouse HP1 homolog (45), and the sequences corresponding to the targets of CKII lie outside of the solved structure. So far, we have found no effect of CKII phosphorylation on HP1 multimerization in solution. 3 Although two of the CKII sites occur within a previously identified nuclear targeting domain (40), we observed no impairment in nuclear targeting for any of the CKII site mutant fusions. CKII phosphorylation could contribute to heterochromatin binding by HP1 by promoting a conformational shift that permits 1) additional kinases to phosphorylate internal targets in, for example, the HP1 linker region between the chromo domains; or 2) the exposure of sites for protein-protein interactions. Either of these results could facilitate heterochromatin assembly. As previously noted (19), this interval is serine/ threonine-rich and includes two consensus targets for protein kinase A and one for protein kinase C.
Phosphorylation on Both the N-and the C-terminal CKII Sites Is Required for Heterochromatin Binding-Although the Ser 3 Ala mutation on the C-terminal site did not discernibly alter the heterochromatin binding activity of the mutant fusion protein, the Ser 3 Ala mutation on the N-terminal site conspicuously reduces heterochromatin binding. The double Ser 3 Ala mutation (S15A,S202A) almost completely eliminated heterochromatin binding, although the protein could still get into the nucleus. The double mutant appeared to have a generally more severe effect. However, care should be taken in interpreting quantitative differences, because levels of fusion protein expression vary from cell to cell in these assays. Although the effect of the single C-terminal substitution was not detectable by the X-gal staining method, it is possible that each mutation exerts some effect on HP1 heterochromatin binding activity because the combined mutations had the most dramatic effect on heterochromatin binding. Although there are two CKII sites at the C terminus of HP1, we chose to mutate the more downstream site. The upstream site is dependent on the phosphorylation of the downstream serine, so when we mutated the first serine to alanine, we also disabled the second one as a CKII target. Thus, in our assay, any effect attributable to the downstream serine could also reflect a requirement for phosphorylation of the upstream serine. Experiments are in progress to identify additional sites of HP1 phosphorylation and to test their role in HP1 localization and silencing activity.
Although transformants were recovered with Ser 3 Glu mutations in the N or C terminus, we were unable to recover germline transformants with Ser 3 Ala mutations. The significance of this finding is unclear, but it may represent a kind of "dominant negative" phenotype (46). In the absence of heat shock, basal levels of wild-type HP1⅐␤-galactosidase fusion protein are not toxic (29,40), although such transgenic lines are not as healthy as wild-type flies. 3 Mutant HP1 fusion protein, on the other hand, may be toxic at low basal levels. A reasonable speculation is that the nonphosphorylated HP1 participates in only some HP1-dependent activities or sequesters heterochromatin factors in an inactive form.
CKII Phosphorylation and HP1 Function-The most basic HP1 isoforms in vivo are phosphorylated at CKII sites. Thus, CKII phosphorylation does not directly account for the hyperphosphorylation that accompanies the appearance of heterochromatin in the early embryonic development. Indeed, it probably accounts for the maternally loaded HP1 isoforms seen in unfertilized oocytes (19). Nevertheless, the mutational analysis shows that CKII phosphorylation is essential for heterochromatin binding.
CKII is an ubiquitous eukaryotic serine/threonine protein kinase that phosphorylates more than 100 substrates, many of which control cell division or signal transduction. These substrates include a striking number of nuclear proteins involved in DNA replication and transcription (47). CKII modifies protein-DNA binding (48 -51) and protein-protein interaction (52,53). In Drosophila, CKII is present in both the cytoplasmic and nuclear compartments. CKII phosphorylation enhances the DNA binding activity of the engrailed protein (51) and modulates Antennapedia activity (39) and dorsoventral patterning (54). Drosophila DNA topoisomerase II is stimulated by CKII phosphorylation (55). We showed previously that significant HP1 phosphorylation still occurs in vivo in tissues treated with sufficient cycloheximide to block all detectable nascent protein synthesis (19). This turnover of phosphate uncoupled from a new synthesis suggested that HP1 phosphorylation could regulate its chromatin association, an example being the dynamic dissociation and reassociation of HP1 that reportedly takes place during mitosis (13). Alternatively, phosphorylation-dephosphorylation may be regulated during decondensation of heterochromatin to permit DNA replication in late S phase.