INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Herpesvirus capsid assembly involves the coordinated interaction of at least five viral proteins. Three of these ultimately interact strongly to form the outer shell, and the other two are internal and interact more transiently with the outer shell and with each other to facilitate capsid formation. The internal proteins form a scaffolding array that is proteolytically freed from the outer shell and largely eliminated from the cavity of the capsid to enable DNA packaging. These internal proteins, in cytomegalovirus (CMV) called the assembly protein precursor (pAP) and proteinase precursor (pNP1), are genetically related and are distinguished from the other capsid proteins by posttranslational modifications (reviewed in references 10, 32, and 41).

CMV pAP and its homologs in other herpesviruses, most notably pVP22a in herpes simplex virus (HSV UL26.5 protein), has at least two important functions in capsid assembly. One is to escort the major capsid protein (MCP; human CMV UL86 protein) into the nucleus by serving as a nuclear localization signal (NLS)-bearing escort (29, 33, 48). Another is to act as a molecular scaffold in guiding formation of the procapsid shell within the nucleus (3, 17, 21). Once procapsid formation is complete, and possibly in conjunction with DNA packaging, pAP is freed from its interaction with MCP by proteolytic cleavage near its carboxyl end and eliminated from the capsid cavity (12, 17, 34). This cleavage is made by a genetically related proteinase that contains the entire pAP sequence as its carboxyl end (23, 31, 47). The proteinase is essential for the production of infectious virus (7, 30) and has received considerable attention as a potential antiviral target (11, 15).

Phosphorylation is a second modification that distinguishes the CMV pAP and its homologs in other herpesviruses (6, 8, 13, 16, 35). Although the significance of pAP phosphorylation and the nature of the modifying enzyme(s) is unknown, both are of interest mechanistically and as they may lead to new antiviral targets. The work described here represents an initial step in studying this modification. Our findings identify the principal phosphorylation site on the CMV pAP, show that smaller amounts of pAP isoforms are generated by phosphorylation at an additional site(s), and provide initial evidence that these modifications may be made by more than one enzyme.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cells and virus. Human foreskin fibroblasts were cultured, grown, and infected with simian cytomegalovirus (SCMV), strain Colburn, as described before (8). Human embryonic kidney (HEK) cells (line 293, ATCC CRL-1573) were grown in 35-mm-diameter wells (product no. 3001; Becton Dickinson, Lincoln Park, N.J.), each containing 3 ml of Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum.

Plasmids and cloning. Standard DNA techniques (36) were used to clone and propagate plasmids. Phosphorylation-site mutants of pAP were made from AW1, the pAP coding sequence in vector RSV.5neo (47). They were made by excising a wild-type sequence and replacing it with the appropriate mutant sequence.

Transfections and immunoprecipitations. HEK cells (10⁶ cells/35-mm-diameter well) were transfected with 6.0 µg of plasmid per well by the calcium phosphate method (2), as before (14). Twenty-four hours later the culture medium was replaced with fresh medium, where appropriate, containing ³²P (400 µCi/ml; Amersham, Arlington Heights, Ill.) or [³⁵S]methionine (50 µCi/ml; ICN, Costa Mesa, Calif.). Two days later, cells were harvested and processed for immunoprecipitation by aspirating the medium from the dish, adding 200 µl of lysis buffer (0.5 M KCl, 1% NP-40, 0.5% deoxycholate, and 1 mM phenylmethylsulfonyl fluoride in phosphate-buffered saline (pH 7.4) to each dish, and scraping the lysate to an edge for transfer to a 500-µl tube. The resulting lysate was subjected to vortex mixing (four vigorous pulses) and clarified by centrifugation (16,000 × g, 10 min, 4°C), and the supernatant fraction was transferred to a new tube and either processed immediately or stored at 80°C until needed (typically not more than 1 week). Immunoprecipitation reactions were done with the rabbit antipeptide antisera specified in Results and protein A beads, as described before (45).

AP isolation from B-capsids by HPLC. Intranuclear B-capsids were recovered from infected human foreskin fibroblasts by freezing and thawing, essentially as described before (17). The particles were concentrated by pelleting (39,000 rpm, 90 min, 4°C, SW41 rotor) and either solubilized in preparation for polyacrylamide gel electrophoresis (PAGE) or disrupted by heating at 80°C for 3 min in a solution of aqueous 6 M guanidine and 1% -mercaptoethanol in preparation for reverse-phase high-performance liquid chromatography (RP-HPLC).

SDS-PAGE and Western immunoassays. Standard procedures were used for protein separation by SDS-PAGE, staining with Coomassie brilliant blue (CBB), and autoradiography; details have been described before (8, 35, 45). In preparation for SDS-PAGE, samples were solubilized in 2× protein sample buffer (4% SDS, 2.86 M -mercaptoethanol, 10% glycerol, 100 mM Tris [pH 7.0], 0.01% bromophenol blue). Resolving and stacking gels were 10 and 4% acrylamide, respectively; the ratio of acrylamide to methylene bisacrylamide was 28:0.735; SDS was from Bio-Rad (Richmond, Calif.; catalog no. 161-0300), and intensifying screens were Kodak Biomax MS in combination with Kodak Biomax MS film.

Thin-layer separation of phosphopeptides. Two-dimensional (2-D) separations of tryptic phosphopeptides on thin-layer cellulose (TLC) plates was done essentially as described before (12). Protein bands were located by CBB staining or by autoradiography, excised from the gel, treated with trypsin (Worthington Biochemical Corp., Freehold, N.J.), and subjected to electrophoresis at pH 1.9 followed by ascending chromatography. The chromatography buffer was isobutyric acid-butanol-pyridine-acetic acid-water (65:5:3:2:25) (38). Phosphopeptides were detected by fluorography using Biomax film and intensifying screens.

Peptide analysis by mass spectrometry. ³²P-labeled AP (25 µg) was prepared by RP-HPLC as described above and treated with 15 µg of trypsin (no. 1418475; Boehringer Mannheim) per sample overnight at room temperature; the resulting peptides were subjected to RP-HPLC using a Vydak C₁₈ column eluted with a gradient of acetonitrile in aqueous 0.1% TFA, essentially as described before (49).

Alkaline phosphatase treatment of pAP. ³²P- or [³⁵S]methionine-labeled pAP was immunoprecipitated from transfected HEK cells with anti-C1 (39) as described above. The bead-bound immune complexes were washed twice in calf intestinal alkaline phosphatase (CIAP) buffer (50 mM Tris-HCl, 0.1 mM EDTA [pH 8.5]) containing 1 mM phenylmethylsulfonyl fluoride, suspended in 200 µl of CIAP buffer (per 20 mg of beads), and aliquoted equally into five tubes. After pelleting the beads again and removing the supernatant, CIAP (10 U/tube, in final volume of 100 µl; Boehringer Mannheim, Indianapolis, Ind.) was added and allowed to react for the specified times. Mock-treated control samples were incubated in the same volume of reaction buffer with no CIAP. Nontreated control samples (time zero) were prepared by adding 2× protein sample buffer to the pelleted beads, instead of CIAP reaction buffer. Reactions were shaken at 37°C for the times indicated and spun for 30 s at 16,000 × g to pellet the bead-bound immune complexes, and the supernatant was removed. The beads were then suspended in an equal volume of 2× protein sample buffer and frozen at 80°C until analyzed, or immediately subjected to SDS-PAGE followed by autoradiography and phosphorimaging.

Detection of P_i. The method used has been described before (40). Following CIAP reactions, the bead-bound immune complexes were pelleted from the preparation by centrifugation (16,000 × g, 1 min, 4°C); the supernatant was subjected to trichloroacetic acid precipitation, followed by addition of KH₂PO₄ and extraction with isobutanol-toluene and ammonium molybdate. The upper phase resulting from centrifugation (12,000 × g, 5 min, room temperature) of the mixture was removed and analyzed for Cerenkov radiation in a scintillation spectrometer.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Identification of primary phosphopeptides of AP from CMV-infected cells. ³²P-labeled AP was purified by HPLC from CMV B-capsids (Fig. 1A) and digested with trypsin, and the resulting peptides were resolved by HPLC. A major peak of ³²P radioactivity was detected (Fig. 1B), and a sample of fraction 25 was subjected to mass spectrometry. Seven of the observed peaks correlate with three predicted AP tryptic fragments (Fig. 1C). The peak with the largest mass-to-charge ratio, 2,044, corresponds to the AP tryptic peptide, Gly 124 to Arg 141 (no phosphate). The peaks at 2,153, 2,232, and 2,314 correspond to the AP tryptic peptide, Asp 154 to Lys 173 (designated casein kinase IIa [CKIIa]; Fig. 1C, inset), in its non-, mono-, and diphosphorylated states, respectively. The peaks at 2,438, 2,517, and 2,602 correspond to an incompletely cleaved form of the same peptide, Glu 152 to Lys 173 (designated CKIIb; Fig. 1C, inset), in its respective non-, mono-, and diphosphorylated states. Because CKIIa is a limit tryptic peptide containing three serines, and CKIIb is an amino-terminal extension of CKIIa, these data localize the primary sites of phosphorylation to one or more of three serines (i.e., Ser156, Ser157, and Ser164; Fig. 1C, inset) in the carboxyl half of AP. Differential ionization properties (22, 49) of the di-, mono-, and nonphosphorylated forms of the peptides limit use of these data to qualitative interpretation only.

View larger version (31K): [in this window] [in a new window]   FIG. 1.   AP is phosphorylated at its CKII-site serines in CMV-infected fibroblasts. 32P-containing AP was purified by HPLC from nuclear B-capsids (A) and digested with trypsin, and the resulting phosphopeptides were resolved by HPLC (B) and analyzed by mass spectrometry (C), all as described in Materials and Methods. Shown are the distributions of 32P radioactivity in fractions collected from the chromatographic separations (A and B) and a mass spectrum obtained in the negative ion mode from a sample of panel B HPLC fraction 25 (C). Panel C (inset) also shows two pAP tryptic peptides, CKIIa (Asp154-Lys173) and CKIIb (Glu152-Lys173), that contain a CKII consensus sequence (underlined) and three serine residues (larger letters).

AP is similarly phosphorylated in virus-infected and plasmid-transfected cells. When ³²P-labeled AP recovered from B-capsids was compared by 2-D peptide separation with ³²P-labeled AP expressed by transfection (Fig. 2A and B, respectively), the patterns were essentially the same (e.g., mixture [Fig. 2C]). Two predominant (i.e., spots 1 and 2) and several minor (e.g., spots a, b, and c) phosphopeptides were present in both. Upon redigestion with trypsin, isolated peptide 1 was unchanged but about 50% of isolated peptide 2 was converted to peptide 1 (data not shown). These results are compatible with phosphopeptide 2 being an incomplete cleavage product, corresponding to the mono- and/or diphosphorylated CKIIb species detected by mass spectrometry, and peptide 1 being the mono- and/or diphosphorylated form of the limit tryptic peptide, CKIIa (additional evidence below). Thus, AP appears to be similarly phosphorylated whether recovered from B-capsids or transfected cells, indicating that these modifications can be made by a cellular enzyme.

View larger version (71K): [in this window] [in a new window]   FIG. 2.   Phosphopeptide patterns of AP from transfected and infected cells are the same. 32P-labeled AP recovered from infected (A) or transfected (B) cells was digested with trypsin, and the resulting hydrolysates, individually or in combination (C), were subjected to 2-D separation on TLC plates. Shown is a collage of autoradiographic images prepared from the resulting plates. The position of the origin (Ori) is indicated, electrophoresis was from left to right, and chromatography was from bottom to top. Phosphopeptide designations are described in the text. Radioactivity in the region of each plate containing the cluster of labeled spots was quantified following phosphorimaging; the photo-stimulated luminescence (psl) measured in panels A + B approximated that of the mixture on plate C.

Primary sites of pAP phosphorylation are within a CKII consensus sequence. Given that AP phosphorylation appears to be the same in plasmid-transfected cells as in CMV-infected cells (e.g., Fig. 2), we used the more amenable transfection system to fine map and further study AP phosphorylation. In the first experiment, each of the three serines in the CKIIa peptide was mutated to determine which is phosphorylated. The AP precursor, pAP (see Fig. 8B), was used for these and subsequent transfection/immunoprecipitation experiments because it could be more efficiently immunoprecipitated than AP, and its 2-D phosphopeptide pattern was indistinguishable from that of AP (data not shown).

View larger version (47K): [in this window] [in a new window]   FIG. 3.   Substitution of CKII-site serines 156 and 157, or glutamic acids 159 and 160, reduced pAP phosphorylation. pAP was immunoprecipitated from cells that had been transfected with wild-type (wt) or mutant-encoding plasmids and grown in medium containing 32Pi. pAP immunoprecipitates were subjected to SDS-PAGE, the proteins were stained with CBB, and the gels were dried and autoradiographed. Shown are images of the CBB-stained proteins (top windows in panels A and B) and autoradiograms prepared from the dried gels (middle windows in panels A and B). The radioactivity/stain intensity of mutant pAP normalized to radioactivity/stain intensity of wild-type pAP is given (bottom windows) for the serine mutants in panel A and for the glutamic acid or serine/glutamic acid mutants in panel B. pAP* and pAP** are electrophoretically slower-moving isoforms of pAP, not included in specific activity calculations of pAP. Abbreviations for the mutant proteins: S156,S157 (pAP/S156A,S157A), S157 (pAP/S157A), S156 (pAP/S156A); S156,E160 (pAP/S156A,E160A); and E159,E160 (pAP/E159A,E160A).

Phosphorylation at serines 156 and 157 is influenced by key acidic residues predictive of CKII substrates. CKII phosphorylation sites characteristically have a specificity-determining acidic amino acid at the +3 position (28). Both Ser 156 and 157 have +3 position acidic residues (i.e., SSDE159E160ED), and these were mutated to test their influence on pAP phosphorylation (Fig. 3B). The mutant proteins were expressed in ³²P-labeled cells, immunoprecipitated, and analyzed by SDS-PAGE/CBB staining followed by phosphorimaging. When both Glu 159 and 160 were mutated to alanines, pAP phosphorylation was reduced by 83% relative to wild-type pAP (Fig. 3B, mutant pAP/E159A,E160A, middle and bottom windows). When Glu 160 alone (i.e., +3 to S157) was changed to Ala, but in the pAP/S156A mutant to test its specific effect on phosphorylation of S157, phosphorylation in the double Ser156/Glu160 mutant was reduced 30% more than observed in the single S156A mutant (compare Fig. 3A, S156, with Fig. 3B, S156,E160). These predicted effects of the +3 acidic residues on phosphorylation of Ser 156 and 157 are consistent with the substrate specificity of CKII or a related enzyme.

Electrophoretic mobility isoforms of pAP due to differential phosphorylation. Eliminating CKII serines 156 and 157 abolished phosphorylation of the primary pAP band but did not eliminate phosphorylation of pAP* and pAP**, indicating that the slower-migrating species are phosphorylated at other sites. To determine whether these additional phosphorylations are responsible for the slower electrophoretic mobilities of pAP* and pAP**, preparations of immunoprecipitated ³²P- or [³⁵S]methionine-labeled pAP were treated with CIAP for 1, 2, or 3 h, and the products of the reactions were separated by SDS-PAGE (Fig. 4A) and quantified by phosphorimaging, or processed to measure CIAP-released ³²P (Fig. 4B). Phosphatase treatment reduced the amount of ³²P labeling of all three bands (Fig. 4A, ³²P lanes; Fig. 4B, ³²P_b). This loss is attributed to dephosphorylation, rather than proteolytic degradation, because (i) it correlated with an increased amount of released ³²P (Fig. 4B, ³²P_r) and (ii) the combined amount of [³⁵S]methionine-labeled protein in the three bands remained nearly constant (Fig. 4A, ³⁵S lanes; Fig. 4B, ³⁵S_b).

View larger version (33K): [in this window] [in a new window]   FIG. 4.   Susceptibility of assembly protein phosphorylation to CIAP treatment. Parallel samples of [35S]methionine-labeled and 32P-labeled pAP were immunoprecipitated from transfected cells and subjected to treatment with 10 U of CIAP for 1, 2, or 3 h, and the resulting products were resolved by SDS-PAGE and analyzed by phosphorimaging. Shown is an autoradiographic image of the CBB-stained gel (A) and a graph (B) showing, for each time point, the bound 32P (32Pb) and 35S (35Sb) (both in units of photo-stimulated luminescence [psl]) in the combined three pAP isoforms, and the total 32P released (32Pr). Time-zero samples were nontreated starting material, and mock-treated samples (M) were incubated for 3 h in CIAP reaction buffer lacking the phosphatase, all as described in Materials and Methods.

View larger version (32K): [in this window] [in a new window]   FIG. 5.   pAP* and pAP** are differentially phosphorylated forms of pAP. Parallel samples of 35S-methionine- and 32P-labeled CKII-site mutant pAP/S156A,S157A were immunoprecipitated from transfected cells and subjected to treatment with 10 U of CIAP for 1, 2, or 3 h, and the resulting products were resolved by SDS-PAGE and analyzed by phosphorimaging. Shown is an autoradiographic image of the CBB-stained gel (A) and a graph (B) showing, for each time point, the bound 32P (32Pb) and 35S (35Sb); displaced downward approximately 70 psl for presentation) (both in units of photo-stimulated luminescence [psl]) in the combined three pAP isoforms, and the total 32P released (32Pr). Time-zero samples were nontreated starting material, and mock-treated samples (M) were incubated for 3 h in CIAP reaction buffer lacking the phosphatase, all as described in Materials and Methods.

Identification of pAP* phosphopeptides containing secondary phosphorylation sites. Phosphopeptide patterns of wild-type and CKII-site mutants of pAP* were compared to correlate the different phosphorylation sites with specific peptides. Respective ³²P-labeled pAP* bands were identified following SDS-PAGE of transfected-cell lysates, cleaved with trypsin, and compared by 2-D peptide analysis. Wild-type pAP* contained two phosphopeptides with low mobility in the chromatography phase (Fig. 6A, spots 1 and 2; doublets, probably due to differential oxidation of Met 163 [43]) and a group of about three phosphopeptides with high chromatographic mobility (i, ii, and iii) (Fig. 6A). Peptides 1 and 2 contain CKII-site serines 156 and 157, as deduced from the data in Fig. 1 and 2 and illustrated by their changes in the mutant proteins. Loss of either Ser 156 and 157 alone caused peptides 1 and 2 to shift toward the cathode (right, relative to spot iii), consistent with the reduced negative charge that loss of a phosphate would cause (Fig. 6B and C). Peptides 1 and 2 were not phosphorylated in the mutant lacking both Ser 156 and 157, establishing that these peptides contain the CKII site.

View larger version (73K): [in this window] [in a new window]   FIG. 6.   Verification of CKII-site, and identification of secondary-site, tryptic phosphopeptides. 32P-labeled wild-type pAP, or pAP from three CKII-site mutants (S156A, S157A, and S156A,S157A), was separated by SDS-PAGE; the pAP* band from each was digested with trypsin; and the resulting phosphopeptides were resolved by two-dimensional separation on TLC plates. Shown is a collage of autoradiographic images prepared from each plate. The position of the origin is indicated (Ori), electrophoresis was from left to right, and chromatography was from bottom to top. Phosphopeptide designations are described in the text; lines in panels A to C help demonstrate shift of peptides 1 and 2 relative to i, ii, and iii.

Secondary sites of phosphorylation are the same in transfected and infected cells. AP in B-capsids isolated from infected cells also migrates as three bands when resolved by SDS-PAGE (17, 21). Phosphopeptide comparisons of AP from B-capsids and pAP from transfected cells showed that their patterns were essentially the same, with spots 1 and 2 being predominant (Fig. 2C and data not shown).

View larger version (69K): [in this window] [in a new window]   FIG. 7.   Comparisons of secondary-site phosphopeptides from pAP and AP isoforms. 32P-labeled pAP isoforms immunoprecipitated from transfected HEK cells, and 32P-labeled AP isoforms from B-capsids, were separated by SDS-PAGE, and the two slower-moving isoforms from each were subjected to digestion with trypsin, followed by 2-D separations of the resulting peptides and fluorographic detection of the phosphopeptides. Shown are the patterns obtained for capsid AP* (A), transfection pAP* (B), and a mixture of the two (C); capsid AP** (D), transfection pAP** (E), and a mixture of the two (F); and capsid AP* (G), capsid AP** (H), and a mixture of the two (I). Some variation was observed between separations. The position of the origin (Ori), for each plate is indicated, electrophoresis was from left to right, and chromatography was from bottom to top. Phosphopeptide designations are described in the text.

Maturational proteinase is also phosphorylated at CKII-site serines. Considering that the entire pAP sequence is contained as the carboxyl half of the proteinase precursor, pNP1 (Fig. 8B), due to the 3'-coterminal, nested arrangement of their genes (46), we tested whether the corresponding CKII-site serines in pNP1 are also phosphorylated. This was done by ³²P-labeling proteins in transfected cells, immunoprecipitating them, subjecting the immunoprecipitates to SDS-PAGE, and visualizing the bands by autoradiography. As observed before (Fig. 3A, left lane) and included here for reference, pAP resolved into three bands (pAP, pAP*, and pAP** [Fig. 8A, lane 1]), only the slower-migrating two being radiolabeled in the CKII-site mutant, pAP/S156A,S157A (Fig. 8A, lane 2).

View larger version (41K): [in this window] [in a new window]   FIG. 8.   Maturational proteinase, pNP1, is phosphorylated at its CKII-site serines. (A) 32P-labeled wild-type (lanes 3 and 4) or inactive (lanes 5 and 6) precursor proteinase bearing either S156 and S157 (lanes 3 and 5) or their alanine substitutions (lanes 4 and 6) were immunoprecipitated (anti-C1 for lanes 1, 2, 5, and 6; anti-N1 for lanes 3 and 4) from transfected HEK cells and separated by SDS-PAGE. Shown here is an autoradiographic image prepared from the resulting gel. Wild-type pAP (lane 1) and CKII-site mutant pAP/S156A,S157A (lane 2) were processed in parallel with the proteinase samples as controls. The asterisk denotes a faint band of NP1 in lane 3. The presence (+) or absence () of CKII-site serines 156 and 157 is indicated beneath each lane. Bars to the left of lanes 1, 3, and 5 indicate positions of the primary and two slower-migrating isoforms of pAP, NP1c, and pNP1, respectively. (B) Schematic showing the primary translation products, pNP1 (proteinase precursor) and pAP (assembly protein precursor), and their corresponding cleavage products: pNP1NP1NP1c + NP1n (assemblin); pAPAP + Tail. M and R, maturational and release cleavage sites; ×, the inactivated serine nucleophile in mutant S118A; SS, CKII-site serines, 156 and 157; N1 (rectangle) and C1 (oval), locations of amino acid sequences used to prepare the anti-N1 and anti-C1 antipeptide antisera; filled circles, positions of NLS1 and NLS2.

Phosphorylation of the CKII serines is not required for enzymatic activity of pNP1 or for substrate suitability of pAP. The observed M- and R-site cleavage of CKII-site mutant, pNP1/CKII (e.g., presence of NP1_c* and NP1_c** [Fig. 8A, lane 4]), demonstrated that CKII-site phosphorylation is not essential for the autoproteolytic activity of the enzyme. To test the possibility that this modification may have a more subtle effect on proteolytic activity, or on pAP as a substrate, we used Western immunoassays to reveal these reactions in more detail. Wild-type proteinase or its CKII mutant, pNP1/CKII, was expressed alone or in combination with the corresponding wild-type or CKII-mutant form of pAP. Lysates of the transfected cells were subjected to SDS-PAGE followed by Western immunoassay using the anti-N1 antiserum (39), an antipeptide antiserum that recognizes all but two of the major pNP1 and pAP cleavage products (i.e., NP1_n and Tail [Fig. 8B]).

View larger version (103K): [in this window] [in a new window]   FIG. 9.   Neither proteolytic activity nor substrate suitability is noticeably altered by CKII-site phosphorylation. Wild-type (wt; lanes 3, 5, and 9) and CKII-mutant (lanes 4, 6, and 10) proteinases were tested for self-cleavage (lanes 3 and 4) and processing of wild-type (lanes 5 and 6) and CKII-mutant substrate (lanes 9 and 10) by transfection assays. Cleavages were monitored by Western immunoassays with anti-N1 (Fig. 8B) and [125I]protein A as described in Materials and Methods. Shown is an autoradiographic image of the resulting blot. Bands of interest are indicated by abbreviations at the right; proteins expressed are specified at the bottom. Presence (+) or absence () of the CKII-site serines 156 and 157 in the proteinase tested (or pAP alone in lanes 7 and 8) is indicated below lanes 3 to 10. Wild-type pAP was coexpressed with the proteinase constructs in lanes 5 and 6; pAP lacking CKII serines 156 and 157 was coexpressed with the proteinase constructs in lanes 9 and 10. Asterisks indicate positions of AP, pAP, and NP1c isoforms. Cytoplasmic (C) and nuclear (N) fractions of virus-infected (Inf.) cells (lanes 1 and 2) are shown for reference. Dots to the left of lane 1 indicate bands coincident with NP1c, pAP, AP**, AP*, and AP (top to bottom).

Nuclear localization is not required for pNP1 phosphorylation. The proximity of the CKII motif to two flanking NLSs (Fig. 8B, NLS1 and NLS2; reference 29), suggested that nuclear transport and phosphorylation of these proteins might be interrelated. To test whether nuclear translocation was required for pNP1 phosphorylation, we compared the level of phosphorylation of NLS⁺ S118A (i.e., S118A; 64 kDa), which localizes to the nucleus, with that of NLS S118A (i.e., S118A/NLS1,2), which is restricted to the cytoplasm (29). This was done by expressing and ³²P-labeling the proteins in transfected cells, lysing the cells and recovering the proteins by immunoprecipitation, and analyzing the immunoprecipitates by SDS-PAGE and autoradiography.

View larger version (59K): [in this window] [in a new window]   FIG. 10.   Evidence for cytoplasmic CKII-site phosphorylation of NLS pNP1. 32P-labeled, inactive precursor proteinase (S118A mutant) bearing either intact (NLS+; lane 1) or mutated (NLS; lanes 2 to 4) NLS (29) were immunoprecipitated with anti-C1 (Fig. 8B) from transfected cells and separated by SDS-PAGE. Shown is an autoradiographic image of the resulting gel. 32P-labeled pAP (lane 5) and mock-transfected cell (lane 6) samples were processed in parallel as controls for expression and immunoprecipitation specificity, respectively.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Phosphorylation is a distinguishing feature of the CMV capsid assembly protein (9, 16, 21, 35) and its homologs in other herpesviruses (1a, 6, 13) that may be of functional importance. As a first step in developing assays to study this modification, and to identify the protein kinase responsible for it, we have determined the primary sites of pAP phosphorylation. Two adjacent serines in a CKII consensus sequence were identified as the only residues phosphorylated in the predominant pAP isoform. Two less abundant isoforms were found to contain one or more secondary sites whose phosphorylation correlated with electrophoretic mobility shifts. Phosphorylated counterparts were identified for the proteinase precursor, pNP1, and its carboxyl cleavage product, NP1_c, both of which contain the entire AP amino acid sequence (Fig. 8B and reference 46). Mutation of the CKII-site serines had no detected effect on the proteolytic activity of pNP1 or on the substrate behavior of pAP. Initial evidence that CKII- and secondary-site phosphorylations may happen in different compartments of the cell raises the possibility that they result from different enzymes. The justification and implications of these conclusions are discussed below.

CKII-site phosphorylation. Our main objective was to determine whether phosphorylation of pAP is limited to a small number of sites and, if so, to identify them. Our primary finding is that the only residues phosphorylated in the most abundant pAP isoform are two adjacent serines (Ser 156 and 157) in a CKII consensus sequence. Although the most compelling evidence for this conclusion came from transfection experiments showing that pAP phosphorylation was eliminated in a mutant lacking these residues (i.e., S156,S157 [Fig. 3A]), it was demonstrated by mass spectrometry (e.g., Fig. 1C) and 2-D phosphopeptide comparisons (Fig. 2 and 7) that the same CKII-site serines are also the primary phosphate acceptors on AP from infected cells.

Secondary-site phosphorylation. A small amount of pAP is phosphorylated at additional sites, correlating with the formation of isoforms pAP* and pAP** (previously called 38- and 39-kDa proteins [17]), which have reduced electrophoretic mobilities. Although pAP* and pAP** are phosphorylated at their CKII serines (Fig. 6 and 7), this modification is not required either for secondary-site phosphorylation or for the mobility shift, as shown by pAP* and pAP** phosphorylation in the CKII-site mutant, pAP/S156A,S157A (Fig. 3A and 6D). Corresponding isoforms of the proteinase (i.e., pNP1* and pNP1**, and its cleavage products, NP1_c* and NP1_c**) were identified, indicating that it shares the same secondary-site modifications (Fig. 8, lanes 3 to 6).

Enzyme(s) involved. It is likely that CKII or a CKII-like activity is responsible for phosphorylating the most abundant of the three pAP and AP isoforms. Phosphorylation of this predominant species is limited to serines 156 and 157 within a CKII consensus sequence, and site-directed mutagenesis of acidic residues +3 of the target serines reduced phosphorylation at these sites, as predicted for CKII (28, 37). Also consistent with these primary phosphorylations being made by CKII, normal (if not enhanced) phosphorylation was observed for a proteinase mutant (S118A/NLS) that has no NLS and is consequently confined to the cytoplasm (Fig. 10)the generally accepted location of CKII during cell cycle interphase (50). There are other examples of herpesvirus proteins phosphorylated by CKII or CKII-like activities (4, 18, 20, 24-27), and at least some of these modifications are known to be functionally important (5, 27).

Function of AP phosphorylation? Our evidence that the CKII-site phosphorylations occur in the cytoplasm is compatible with their involvement in one or more of the earliest steps in the capsid assembly pathway, such as interaction of pAP with itself, pNP1 or MCP, or nuclear translocation of these complexes. Conversely, it could be dephosphorylation of these residues later in the assembly process that is important. We found no evidence for influences of CKII-site phosphorylation on the properties of these proteins that were tested, including proteolytic activity of pNP1 (Fig. 9), substrate effects on pNP1 or pAP (Fig. 9), interactions of pAP with itself and with MCP (1), and self-localization or MCP translocation to the nucleus (29, 48) by pNP1 (i.e., S118A/S156,S157) or pAP (immunofluorescence data not shown). Our inability to demonstrate such differences, however, does not rule out the possibility of more subtle effects on these interactions (e.g., kinetic; conformational) that may have escaped detection by the assays used but, nevertheless, have important consequences in the context of other viral proteins and processes in CMV-infected cells.