- Email: [email protected]
Molecular Cloning and Characterization of a Novel Casein Kinase II Substrate, HASPP28, from Rat Brain
ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS
Vol. 327, No. 1, March 1, pp. 131–141, 1996 Article No. 0101
Molecular Cloning and Characterization of a Novel Casein Kinase II Substrate, HASPP28, from Rat Brain1 Li Shen, Kuo-Ping Huang, Hao-Chia Chen, and Freesia L. Huang2 Endocrinology and Reproduction Research Branch, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892
Received August 2, 1995, and in revised form November 28, 1995
HASPP28 (heat- and acid-stable phosphoprotein of 28 kDa) has been purified to near homogeneity from the acid-stable protein fraction of rat brain extract. Based on the N-terminal 40 amino acid sequence, a pair of highly degenerate primers was used to generate a 107-bp probe from rat brain RNA by RT–PCR. From the rat brain lgt11 library, this probe identified two positive clones that together provided a cDNA of 837 bp with an open reading frame of 546 bp. This cDNA was extended by 3*RACE to 1.2 kb that included a polyadenylation signal and a poly(A) tail. The 180amino-acid sequence derived from the open reading frame, which did not correspond to any known protein, was predicted to have phosphorylation sites for protein kinase C, casein kinase II (CKII), and protein kinase A. Indeed, both the purified rat brain HASPP28 and the recombinant HASPP28 (rHASPP28) can be phosphorylated by these kinases. Northern blot analysis indicated that HASPP28 was present in all rat tissues tested, including those from the brain, lung, spleen, kidney, liver, heart, and muscle, in decreasing order of abundance. Phosphopeptide analysis of rHASPP28 phosphorylated in vitro by various kinases showed different tryptic peptide patterns on two-dimensional mapping and isoelectric focusing gels. From [32P]PO4-labeled N1E115 neuroblastoma cells, HASPP28 can be immunoprecipitated with a polyclonal antiserum raised against rHASPP28. The immunoprecipitated protein showed a phosphopeptide pattern similar to that of rHASPP28 phosphorylated by CK II in vitro. Furthermore, the immunoprecipitates from cells treated with phorbol 12-myristate 13-acetate or 8-bromo-cAMP did not show any increased phosphorylation over those of untreated ones, and the
phosphopeptide patterns of the immunoprecipitates again were similar to that of CK II phosphorylated protein. These results suggest that HASPP28 is a novel phosphoprotein that can be phosphorylated by several kinases in vitro. In intact cells, CK II seems to be solely responsible for the phosphorylation of HASPP28. q 1996 Academic Press, Inc.
Protein kinases and kinase-catalyzed reactions play major roles in the regulation of a great variety of cellular processes. Several well-studied protein kinases, such as protein kinase C (PKC),3 cAMP-dependent protein kinase (PKA), Ca2//CaM dependent protein kinases, casein kinases, MAP kinases, and ligand receptor-associated tyrosine kinases, mediate the responses of eukaryotic cells to external stimuli by phosphorylating their specific protein substrates (for review see Ref. 1). There is, however, only limited knowledge available in the identification of specific protein substrate(s) relevant to a particular cellular response regulated by a kinase. For example, the involvement of PKC or PKA in numerous cellular functions has been demonstrated through the enhancement of the specific responses by the treatment of phorbol 12-myristate 13-acetate (PMA), an activator of PKC, or permeable cAMP analogues, activators of PKA, respectively (2–5). Most of the time, the target substrate(s) being phosphorylated by PKC or PKA was not identified. On the other hand, for many purified proteins, whose phosphorylations by kinases were well characterized in vitro, evidence for in vivo participation of such reactions is still unavailable (6, 7). 3
1
Sequence data from this article have been deposited with the EMBL/GenBank Data Libraries under Accession No. U26541. 2 To whom correspondence should be addressed at Building 49, Room 6A36, National Institutes of Health, Bethesda, MD 208924510. Fax: (301) 480-8010.
Abbreviations used: PKC, protein kinase C; PKM, the proteasedegraded PKC; PKA, cAMP-dependent protein kinase; CK, casein kinase; PMA, phorbol 12-myristate 13-acetate; TFA, trifluoroacetic acid; AEBSF, 4-(2-aminoethyl)benzenesulfonyl fluoride; RT–PCR, reverse transcription–polymerase chain reaction; RACE, rapid amplification of cDNA end; ECL, enhanced chemiluminescence. 131
0003-9861/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
/ 6b11$$9300
01-29-96 13:40:16
arca
AP: Archives
132
SHEN ET AL.
In our studies to identify additional potential PKC substrates, we have purified several proteins from the heat- and acid-stable fraction of rat brain extract. In addition to the well-characterized proteins, including MARCKS (8), neuromodulin (9), and neurogranin (10, 11), we have purified a Mr Å 28k PKC substrate to near homogeneity. Since the sequence of its N-terminal 40 amino acids did not correspond to any known protein in the Genetics Computer Group (GCG, Madison, WE) data bank (SwissProt, release 31), we anticipated it to be a novel PKC substrate. This paper describes both the cDNA cloning of this protein and investigations of its properties of phosphorylation in vitro and in intact cells. This protein, designated as HASPP28 (for heatand acid-stable phosphoprotein of 28 kDa), or PP28 for short, is an excellent substrate of casein kinase II (CK II), capable of also being phosphorylated by PKC, PKA, and CK I in vitro. Under normal culturing conditions, PP28 is present as a phosphoprotein in intact cells, and CK II or CK II-like kinase is the kinase responsible for its phosphorylation. EXPERIMENTAL PROCEDURES Materials. [g-32P]ATP and [32P]PO4 were from DuPont NEN. [aS]ATP and ECL reagent were from Amersham. [a-32P]GTP for probe labeling was from ICN. Ampholine for isoelectric focusing gels was from LKB-Pharmacia. Protein A–agarose was from Sigma. Restriction enzymes were from New England Biolab. PKC (12), protease-degraded PKC (PKM) (13), PKA and CK I (14), and CK II (15) were prepared as described. Sources of other materials used are indicated in the following procedures at their first appearances.
35
Purification of PP28 from rat brain. Initial extraction of PP28 from frozen rat brain and acid precipitation step were similar to methods described for the purification of neurogranin and other PKC substrates (11, 16). The concentrated acid-stable protein fraction was applied to a DEAE–cellulose column, packed in a Pharmacia HR 16/ 10 column, and eluted with a 0–1 M KCl gradient in buffer A (20 mM Tris–Cl, pH 7.5, containing 1 mM DTT, 0.5 mM EDTA, and 0.5 mM EGTA), delivered by a Pharmacia FPLC unit. The effluent fractions were assayed for PKC substrate by phosphorylation with PKC, SDS–PAGE, and autoradiography. Fractions eluted between 0.1–0.2 M KCl, where PP28 was located, were concentrated by ultrafiltration. PP28 was purified to near homogeneity by subsequent chromatography on a C4 reverse-phase column (Vydac 214TP510, 10 1 250 mm), where it was eluted at 33–35% acetonitrile from a linear gradient composed of 0.1% trifluoroacetic acid (TFA)/0.1% TFA / 100% acetonitrile. The purified protein showed an apparent Mr Å 28,000 on SDS–PAGE (10–20% gradient gel). The N-terminal amino acid sequence of the purified protein was determined using a gasphase protein sequencer with an on-line PTH–amino acid analyzer (Applied Biosystems Model 470A). cDNA library screening and DNA sequence analysis. A pair of PCR primers based on the N-terminal amino acid sequence of the purified protein were synthesized. The forward primer, 5*-CGI AA(AG) GGI GGN CA(TC) AA(AG) GG-3* (where N is A/T/G/C), corresponding to the sequence of RKGGHKG (residues 5–11), and the reverse primer, 5*-TG (TC)TG (TC)TC (AG)TC (TC)TC (TC)TC (AG)TT-3*, corresponding to the sequence of NEEDEQQ (residues 34–40), were used to produce a cDNA probe of expected 107 bp in length from rat brain RNA by RT–PCR (see Fig. 2). The PCR product was directly cloned into the pCRvector (Invitrogen) and the nucleo-
/ 6b11$$9300
01-29-96 13:40:16
arca
tide sequence was determined by dideoxy-mediated chain termination method (17) using Sequenase version 2.0 (U.S. Biochemical). Primers of SP6 and T7 were used in the sequencing reaction. The lgt11 cDNA library of rat brain (Clontech) was screened as described by the manufacturer. Using [a-32P]GTP-labeled 107 bp cDNA as a probe, two strongly positive clones were obtained from about 1 1 106 recombinant phage plaques. The inserts of both clones were excised by EcoRI and each was subcloned into the plasmid pGEM-3Zf(0) (Promega). The DNA sequence was determined from both directions (see strategy in Fig. 2). One clone, l7, having a 503bp insert, contained the translational initiation codon of ATG, and the other, l8, having an 808-bp insert, contained a TAA stop codon. The pGEM–pp28 plasmid, containing the entire open reading frame, was constructed by ligation of a 217-bp EcoRI/XbaI DNA fragment (excised from the ATG containing clone) with a 652-bp XbaI/BamHI fragment (excised from the TAA containing clone) and cloned into EcoRI and BamHI sites of plasmid pGEM-3Zf(0). 3*RACE–PCR. A clone containing the 3*-noncoding region was generated by rapid amplification of cDNA end method (18) using 3*-AmpliFINDER RACE kit from Clontech. The NN01oligo(dT)CDS primer provided in the kit was used in the initial reverse transcriptional reaction. The forward primer used in the primary PCR reaction was 5*CTGTGGGAGGAGATGCC (553–569), for the secondary PCR, 5*AGCTGCTATCTTTGAGA (736–752). In both PCR reactions, the provided Anchor Primer served as the reverse primer. The secondary PCR product was directly cloned into pCRvector and sequenced. Northern blot analysis. Total RNA from different rat tissues were prepared with RNAzol (Tel-Test, Inc.) using the protocol described by the manufacturer. The uniformly labeled cDNA of 666 bp, excised from pGEM-pp28 with EcoRI/PstI, was used as a probe to hybridize the immobilized RNA (19). Expression and purification of recombinant PP28 protein. The bacterial expression plasmid, pRSETC–pp28, was constructed by PCR amplification of the insert of pGEM–pp28, using the oligonucleotide with a NdeI site fused to the 5* end of the first seven codons of the PP28 cDNA as the forward primer and the SP6 oligonucleotide as the reverse primer. The resulting PCR product was digested with NdeI/PstI and then cloned into the NdeI/PstI sites of pRSETC (Invitrogen). Escherichia coli strain BL21(DE3) (Novagen) was used to express the recombinant PP28 protein (rPP28). The transformed bacteria were grown in 2xYT medium containing 2% glucose and 100 mg/ml ampicillin. Cells from 500 ml of culture were lysed in 25 ml of 50 mM Tris–Cl, pH 7.5, containing 1 mM EDTA, 1 mM EGTA, 1 mM DTT, and 50 mM AEBSF by the addition of lysozyme to a final concentration of 200 mg/ml. The suspension was treated in boiling water for 15 min and centrifuged to remove the precipitated protein; the supernatant fluid was then lyophilized. The sample was redissolved in 0.1% TFA and first applied to a C4 reverse-phase precolumn, washed with 0.1% TFA, and then eluted with 100% acetonitrile containing 0.1% TFA and was lyophilized again. The solid sample was again dissolved in 0.1% TFA and separated on a HPLC C4 reversephase column as described for the purification of the rat brain PP28. Final purification was achieved by chromatography on a Mono Q column (FPLC system, Pharmacia-LKB) that was equilibrated with buffer A, containing 10% glycerol, and eluted by a 0–0.45 M KCl gradient in the same buffer. Recombinant PP28 was eluted at about 0.23 M KCl. The eluent was desalted by C4 reverse-phase precolumn and lyophilized. The purified rPP28 was used to immunize rabbit for raising polyclonal antibody. Protein kinase phosphorylation reactions in vitro. Phosphorylations by CK I and II, PKA, and PKC were carried out in reaction mixtures containing 30 mM Tris–Cl, pH 7.5, 6 mM Mg(OAc)2 , and 120 mM [g-32P]ATP. Reactions for PKC also contained 0.4 mM CaCl2 , 100 mg/ml phosphatidylserine, and 20 mg/ml dioleoylglycerol, while those for PKA contained 10 mM cAMP. Reactions were carried out at 307C. For autoradiographic detection of proteins, reactions were terminated by the addition of 0.25 vol of five times concentrated
AP: Archives
CLONING OF A NOVEL CASEIN KINASE II SUBSTRATE, HASPP28 SDS–PAGE sample buffer and were heated for 5 min at 1007C and analyzed by SDS–PAGE. Endogenous phosphorylation and immunoprecipitation. Mouse neuroblastoma N1E115 cells were grown in DMEM medium containing 10% fetal calf serum in 35-mm dishes. After reaching Ç70% confluence, cells were washed three times with phosphate-free DMEM medium and incubated with 400 mCi carrier-free [32P]PO4 in 0.8 ml of the same medium. After 3 h of incubation, either PMA or 8-Br-cAMP was added to give a final concentration of 1 mM or 1 mM, respectively, and incubation was continued for another 15 min. Cells were washed three times with 10 mM Tris–Cl buffered saline and lysed in 200 ml of 20 mM Tris–Cl, pH 7.5, containing 1 mM EDTA, 1 mM EGTA, 150 mM NaCl, 100 mM NaF, 0.5% NP-40, and 50 mM AEBSF. Cell lysates were centrifuged in an Eppendorf centrifuge for 5 min, and SDS was added to the supernatant (containing 250 mg protein) to give a final concentration of 0.1%. Antiserum (10 ml), raised against purified rPP28, was added and incubated at 47C overnight with gentle rocking. Protein A cross-linked agarose bead slurry (40 ml) was added to the mixture and incubation was continued at room temperature for 2 h more. The agarose beads were washed four times with the cell lysis buffer containing 0.1% SDS and then heated with 50 ml of five strength SDS–PAGE sample buffer at 1007C for 5 min. After brief centrifugation to remove the agarose beads, supernatant was subjected to SDS–PAGE and proteins were electrophoretically transferred to Immobilon-P membrane (Millipore). Labeled PP28 was located by autoradiography, excised after the membrane was blocked with polyvinylpyrrolidone-40 (20), and subjected to tryptic digestion for analysis. Phosphopeptide analysis by isoelectric focusing gel and 2-D mapping. Isoelectric focusing gel analysis was carried out with a LKB Multiphor apparatus. Polyacrylamide gels of pH range 2.5 to 6.0 was used to analyze phosphopeptides. For thin-layer 2-D mapping, the first dimension was electrophoresis in pH 6.5 buffer (acetic acid:pyridine:H2O Å 1:25:225); the second dimension was ascending chromatography in the upper phase of freshly prepared n-butanol:acetic acid:H2O (4:1:5). Determination of phosphorylation sites. Purified rPP28 (200 mg) was phosphorylated by CK II in a 200-ml reaction mixture containing 30 mM Tris–Cl, pH 7.5, 6 mM Mg(OAc)2 , and 120 mM [g-32P]ATP (3000 cpm/pmole). After the addition of 50 ml of five strength SDS– PAGE sample buffer to stop the reaction, samples were heated at 1007C for 5 min and then subjected to SDS–PAGE (10% gels). The phosphorylated protein band was located by autoradiography without staining or fixing the gel. Gel pieces were excised and homogenized using a glass-teflon homogenizer. The homogenized gel was extracted three times with acetone:water:acetic acid:triethylamine(17:1:1:1) to remove SDS (21) and then dried and reswollen in 0.5 ml of 50 mM NH4HCO3 , pH 7.8. TPCK-treated trypsin (10 mg) was added to the gel suspension and was incubated at 307C for 5 h before another 10 mg of fresh trypsin was added. The proteolysis was continued overnight. Trypsinized peptides were collected by centrifugation to remove the acrylamide gel pellet, which was washed three times with 50 mM NH4HCO3 . The combined tryptic digests were lyophilized and were separated on a Vydac C18 reverse-phase HPLC column equilibrated with 0.1% TFA. Elution was carried out with a linear gradient of 0–50% acetonitrile in 0.1% TFA in 90 min. The radioactivity was determined by Cerenkov counting and fractions of the major radioactive peak were subjected to amino acid sequence analysis and isoelectric focusing gel analysis.
RESULTS
Purification of rat brain PP28. PP28 was purified from the 2.2% perchloric-acid-soluble fraction of rat brain extract. The fraction was first chromatographed on a DEAE–cellulose column and then on a C4 reverse-
/ 6b11$$9300
01-29-96 13:40:16
arca
133
phase column. Figure 1 shows a typical fractionation profile on HPLC. Phosphorylation assay by PKC, SDS– PAGE, and autoradiography indicated that several proteins are phosphorylatable by PKC. As shown in the inset of Fig. 1, a major protein band (Mr Å 28k) appeared in fractions 81 and 82, and both were phosphorylated by PKC. Proteins in fraction 81 were separated by SDS–PAGE, transferred onto Immobilon-P, and the PP28 protein band located after staining was excised for microsequencing (sequence shown in Fig. 3). Isolation and characterization of cDNA encoding PP28. As the N-terminal 40-amino-acid sequence of the purified rat brain protein PP28 does not correspond to any known protein sequence from the protein sequence data bank (SwissProt release 31), cloning and sequencing of its cDNA were carried out to determine the complete protein sequence. In order to simplify the library screening, we first prepared a 107-bp cDNA fragment encoding the N-terminal part of the protein from rat brain RNA by RT–PCR (see Fig. 2 for cloning strategy and Fig. 3 for the corresponding sequences). In spite of the high degeneracy of the primers, the 107bp fragment turned out to be the only product of the PCR and its deduced sequence matched amino acid 5– 40 of the determined sequence. Using this cDNA fragment as a probe for library screening, two positive clones, l7 and l8, were obtained from 106 phage plaques. Clone l7 (bp 027 to 476) contained an ATG start codon, while clone l8 (bp 3 to 810) contained a TAA stop codon, collectively, they provided the complete open reading frame of PP28 cDNA (Fig. 3). The sequence, however, did not exhibit a poly(A) tail or polyadenylation signal. It should be noted that both l7 and l8 contained the GAA codon, denoting glutamic acid as their 40th codon after ATG instead of CAA, which denotes glutamine as originally determined and used in the reverse primer in the RT–PCR. During sequencing, 40 cycles of Edman degradation may have weakened the signal, thus, leading us to incorrectly assign glutamine instead of glutamic acid as the 40th residue. Incidentally, the 39th residue is also glutamine; the carryover might give rise to the ambiguity of such a sequence. In order to verify the presence of such a sequence in the tissue, RT–PCR was performed using total RNA from rat brain, kidney, and liver and mouse neuroblastoma cells N1E115. Sequences of PCR products, either RT–PCR107 or RT–PCR661 (Fig. 2), from all the RNAs tested were all identical to those originally determined from the cloned cDNA (l7 and l8). Furthermore, the RT–PCR661 products obtained from several tissue RNAs all have GAA as their 40th codon, thus, glutamic acid was assigned as the 40th residue of the protein. Since the 3* end of the clone l8 (810 bp) did not
AP: Archives
134
SHEN ET AL.
FIG. 1. Reverse-phase HPLC on C4 column for the purification of PP28 from rat brain. The perchloric-acid-soluble proteins of rat brain extract were first separated on a DEAE–cellulose column before applying to the C4 reverse-phase column as described under Experimental Procedures. Dotted line shows the profile of the gradient in percentage of acetonitrile. Effluent fractions were assayed for PKC substrate (l) as previously described (11, 16). PKC phosphorylated proteins were also analyzed by SDS–PAGE; the inset shows the protein staining and autoradiography for fractions 81 and 82, the major fractions of PP28. Subsequently, proteins in fraction 81 were separated by SDS– PAGE and transferred to an Immobilon-P membrane, and PP28 was excised for microsequencing.
contain a poly(A) tail or obvious polyadenylation signal, 3*RACE–PCR with RNAs from both brain and kidney were performed. We were able to extend 382 bp from the 3* end of clone l8 and to identify a putative polyade-
nylation signal (22) of ‘‘ATTAAA,’’ 10 bp upstream of the poly(A) tail (Fig. 3). The combined cDNA clone (Fig. 3) has approximately 1.2 kb, with a 5* noncoding region of 27 bp, a 3* noncod-
FIG. 2. Cloning and sequencing strategy. Bases are numbered from the start of translation in base pairs. l7 and l8 are independent cDNA clones obtained from rat brain lgt11 cDNA library screening. Arrows indicate the directions and lengths of individual sequencing runs. RT–PCR107 was the initial 107-bp RT–PCR product from rat brain RNA used for library screening and was later verified together with RT–PCR661 in RNAs of rat brain, kidney, liver, and neuroblastoma N1E115 cells. 3 *RACE denotes the 3*RACE–PCR product from RNAs of brain and kidney to extend the 3 * untranslated region of cDNA to the polyadenylation tail. Primers used in these PCR reactions were synthesized according to the sequences of l7 and l8 (see Experimental Procedures for sequence and location).
/ 6b11$$9300
01-29-96 13:40:16
arca
AP: Archives
CLONING OF A NOVEL CASEIN KINASE II SUBSTRATE, HASPP28
FIG. 3. Nucleotide sequence of cDNA and deduced amino acid sequence of PP28. Amino acids are shown in the single-letter code and numbered from the first codon after ATG. Underlined sequence corresponds to the 40 N-terminal amino acids of the purified rat brain PP28 as determined by automatic Edman degradation. The 40th residue (E), was originally determined as Q (see text). The sequence of 107 bp obtained from RT–PCR, which was used in the library screening, is indicated in bold; among these nucleotides, 16– 35 and 103–122 were regions used for the degenerate forward and reversed primers (see Experimental Procedures for sequences). A portion of the 3* untranslated sequence, from 811 bp to poly(A) tail, obtained from 3*RACE–PCR, is presented in lowercase. A polyadenylation signal of ‘‘attaaa’’ is indicated by a double underline.
ing region of 646 bp, and a coding region of 546 bp. Such an open reading frame encodes a polypeptide of 181 amino acids. As neither the purified PP28 from rat brain nor rPP28 started with methionine, the second residue of proline was counted as the first amino acid and the polypeptide was predicted to contain 180 amino acids. Sequence analysis. Although the molecular weight of the polypeptide was calculated to be 20,491, both the purified rat brain protein and the recombinant one migrated on SDS–PAGE as having 28 kDa. Among the 180 amino acids, there are 42 acidic and 42 basic residues, distributed as 26 glutamic acid, 16 aspartic acid, 27 lysine, and 15 arginine, and the isoelectric point is 7.77. Interestingly, there are no cysteine, phenylalanine, or tryptophane residues in the molecule. As expected from the fact that the protein can be phosphorylated by PKC, the sequence analysis (Motif in GCG) revealed four potential phosphorylation sites for PKC: Thr91, Thr92, Ser107, and Ser169 (23, 24). In addi-
/ 6b11$$9300
01-29-96 13:40:16
arca
135
tion, it also predicted five phosphorylation sites for CK II at Thr17, Ser18, Ser62, Thr96, and Ser107 (25). One of these sites, Thr96, is also a predicted site for PKA (26). Since no consensus site for N-glycosylation was found in the protein sequence, the abnormality of the retarded migration on gel electrophoresis cannot be attributed to N-glycosylation. The protein is extremely hydrophilic and the hydrophobicity profile analysis (data not shown) revealed no obvious hydrophobic region that would indicate a membrane-spanning domain. We also subjected the nucleotide sequence to a computer-assisted homology search against a DNA data bank (GCG/FASTA, GenBank release 89) without finding a significant match to any known sequence. The closet similarities as revealed by the ‘‘Bestfit’’ program were 45% to troponin T (27), 43% to a 70-kDa neurofilament (28), 48% to calpain inhibitor precursor calpastatin (29), and 49% to EMB-5 protein (30), whereas the identities were only 21.3, 17.5, 26.8, and 18.3%, respectively. It is worth mentioning that the 10 amino acids, residues 48–57 of PP28, are identical to residues 125–134 of calpastatin. It is also interesting to note that between the two asymmetric domains of troponin T, T1 and T2 (residues 1–158, 159–259, respectively), it was T1 that shared similarity with PP28 and was responsible for the interaction between troponin T and tropomyosin (31). RNA blotting analysis. From RT–PCR107, RT– PCR661, and 3*RACE–PCR reactions we have already learned that the message of the gene appeared to be present in various rat tissues besides the brain, from which the protein was first purified. The Northern blot analysis (Fig. 4) showed that there was only one Ç1.5kb species of messenger RNA present in all the tissues tested. It was most abundant in the cerebrum and cerebellum, moderate in the lung and spleen, and less in the kidney, liver, heart, and muscle (in decreasing order). These results thus confirm the findings from the RT–PCR reactions that the PP28 protein seems to be distributed widely in various tissues. Immunoblot analysis of PP28. Polyclonal antiserum raised against rPP28 and the IgG purified from this serum recognized both the purified rat brain protein and rPP28 in an immunoblot analysis (Fig. 5) while the preimmune serum failed in such an analysis (not shown). The antibodies also cross-reacted with a protein band of similar mobility as the purified PP28 in the supernatants of the heat-treated extracts of N1E115, COS, and NIH3T3 cells. It appears that PP28 is also expressed in various cultured cells of different origins. Phosphorylation of PP28 and tryptic peptide analysis. In vitro, both purified PP28 and rPP28 can be phosphorylated by CK I, CK II, PKA, and PKC; the relative amounts of phosphate incorporations are 0.36,
AP: Archives
136
SHEN ET AL.
FIG. 4. Northern blot analysis for tissue distribution. RNAs (20 mg each) isolated from various rat tissues were denatured with formaldehyde, separated on 1% agarose gel and capillarily transferred to nitrocellulose membrane, and analyzed by hybridization. Probe used in the hybridization is the 666-bp EcoRI/PstI fragment from pGEM– pp28 that covers the entire open reading frame.
1.63, 0.21, and 1, respectively, taking that by PKC is 1 (data not shown). Tryptic phosphopeptide patterns of these phosphorylations were analyzed in order to be able to determine the occurrence of any of such phosphorylation in vivo. Both two-dimensional peptide mapping and isoelectric focusing gel analysis (Fig. 6) revealed different tryptic phosphopeptide patterns when rPP28 was phosphorylated by CK I, CK II, PKA, and PKCa. The peptide map (Fig. 6A) of PKA or PKCa phosphorylated proteins contained only one major spot that migrated toward the cathode, while those of CK I or CK II phosphorylated proteins consisted of multiple spots that migrated toward the anode. The phosphopeptides resulting from CK I or CK II phosphorylation were much more acidic than those from PKA or PKC phosphorylation as demonstrated by the pI’s of these peptides obtained from isoelectric focusing gel analyses (Fig. 6B). PKM, the partially proteolyzed PKC that does not require Ca2/ or lipid for activity, gave a peptide pattern similar to that of PKCa. Phosphorylation of endogenous PP28 in intact cells. Mouse neuroblastoma N1E115 cells were metabolically labeled with [32P]PO4 , and the extracted proteins immunoprecipitated with antiserum and analyzed on SDS–PAGE. Figure 7A shows that whether unstimulated or stimulated with 8-Br-cAMP or PMA, all the cell extracts produced a prominent precipitated protein band having a similar migration as the in vitro phosphorylated rPP28. Thus, under the present culturing
/ 6b11$$9300
01-29-96 13:40:16
arca
conditions, either stimulated or unstimulated, PP28 in N1E115 cells appears to be phosphorylated. Treatment with cAMP or PMA, however, did not seem to increase the radioactivity of the immunoprecipitated protein. To determine which kinase was responsible for the phosphorylation, the labeled protein bands were excised from the membrane, trypsinized, and then analyzed by isoelectric focusing gels. Figure 7B demonstrates that immunoprecipitated PP28, either from control or stimulated cells, exhibited a phosphopeptide pattern similar to that of rPP28 phosphorylated in vitro by CK II. It appears that CK II- or a CK II-like kinase is responsible physiologically for the phosphorylation of PP28. Determination of CK II phosphorylation sites. As five sites were predicted to be phosphorylated by CK II (distributed into four tryptic peptides), and there were at most three major phosphopeptides bands detected by isoelectric focusing gel analysis, it would be interesting to determine what were the sites being phosphorylated. The labeled tryptic peptides were fractionated by reverse-phase HPLC on a C18 column, and the fractions were counted for radioactivities. As shown in Fig. 8A, there is one major peak of radioactivity including fractions 53, 54, and 55, and a minor shoulder peak at fractions 56 and 57 (collectively, these fractions comprised ú95% of the total radioactivities). These fractions were subjected to microsequencing. Surprisingly, all these fractions yielded one single amino acid sequence corresponding to Lys52 –Lys72 (Fig. 8B), ex-
FIG. 5. Immunoblot analysis of PP28. Proteins were separated on SDS–PAGE (10% gel), transferred onto nitrocellulose membrane, blotted with anti-PP28 antiserum (1:2000 dilution), and then detected with ECL (Amersham). Lane 1, 0.25 mg of purified rat brain PP28 (fraction 81 from Fig. 1); lane 2, 0.25 mg of purified recombinant PP28; lanes 3–5 are heat stable proteins (10 min, 1007C) from 500 mg extracted proteins of N1E115 (lane 3), COS (lane 4), and NIH3T3 (lane 5) cells. The buffer used for the cell extraction was 20 mM Tris– Cl, pH 7.5, containing 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 0.5% NP-40, and 50 mM AEBSF.
AP: Archives
CLONING OF A NOVEL CASEIN KINASE II SUBSTRATE, HASPP28
137
FIG. 6. Two-dimensional peptide maps and isoelectric focusing gel analysis for phosphopeptides of rPP28. Recombinant PP28 (200 mg) was phosphorylated by various protein kinases in the presence of [g-32P]ATP as described under Experimental Procedures. As phosphorylation rates of PP28 by different protein kinases are dissimilar, unitages of the various kinases used in this experiment were adjusted to yield similar phosphate incorporation except the reaction with PKA, in which the incorporation was approximately half of those of others. The phosphorylation reaction was terminated by the addition of TCA to a final concentration of 15%. The precipitated proteins were dissolved in cold 0.1 N NaOH and precipitated again to eliminate free ATP before trypsin digestion. Phosphopeptides were analyzed by two-dimensional peptide mapping (A) and isoelectric focusing gels (B). In A, ‘‘x’’ denotes the origin of sample application; in B, the pI indicates the pH value at the end of focusing, as measured by the surface electrode.
cept that fraction 56 contained Arg73 as its C-terminal. All these peptides contained a phosphorylated Ser62, while fraction 53 also had a phosphorylated Ser59 based on the high ratio of PTH–Ser* (Å PTH-dehydro-Ala/ DTT adduct) to PTH–Ser at the presumed site, as compared with other serine residues in the sequenced peptide. These phosphopeptides were also analyzed by isoelectric focusing on polyacrylamide gel as shown in Fig. 8C. The major band of fraction 53, focused toward the
/ 6b11$$9300
01-29-96 13:40:16
arca
most acidic side of the gel with a pI of 3.8, is probably the species containing two phosphorylated serines, Ser59 and Ser62. The other band of fraction 53, focused at pH 4.0 where the major band of fraction 54 is located, is probably the species containing a single phosphorylated Ser62. The major band in fraction 56, having a pI of 4.3, is most likely the species having a phosphorylated Ser62 and a C-terminal Arg73. It appears that Ser62 of PP28 is most favorable for CK II phosphorylation,
AP: Archives
138
SHEN ET AL.
peptide of pI 3.8. Also noted is that the accumulation of the peptide with pI of 4.3 depended on the amount of trypsin used in digestion. In the experiments of Figs. 6, 7, and 9, where the trypsin/protein ratio was 1/5, more extensive proteolysis occurred and resulted in less accumulation of the pI 4.3 peptide, whereas in the experiment of Fig. 8, trypsin to protein ratio was lowered to 1/20 for sequencing purposes, and the accumulation of this peptide was quite significant. DISCUSSION
FIG. 7. Immunoprecipitation of endogenous PP28 from [32P]PO4labeled cells and tryptic peptide analysis. Subconfluent N1E115 cells were metabolically labeled with [32P]PO4 for 3 h (lane 1 for untreated cells) and were treated with 8-Br-cAMP (1 mM, lane 2) or PMA (1 mM, lane 3) for 15 min. The extracted proteins were immunoprecipitated with anti-PP28 antiserum, separated by SDS–PAGE (10% gel), and transferred to an Immobilon-P membrane. (A) autoradiogram revealed the immunoprecipitated PP28. The labeled PP28 was excised from the membrane, trypsinized, and analyzed by isoelectric focusing gel (B) as described under Experimental Procedures. The lane marked ‘‘rPP28’’ represents the phosphopeptides from in vitro CK II phosphorylated rPP28 for comparison.
while Ser59, which was not a predicted site from the computer search, is the next favorable (see next paragraph). Somehow, the phosphorylation of the other four predicted sites, Thr17, Ser18, Thr96, and Ser107, is either too low for detection or not taking place at all. A closer examination of these residues, it is rather clear that Ser107 could never be phosphorylated by CK II having Arg–Arg immediately at its C-side instead of acidic residues. Sequential phosphorylation of Ser 62 and Ser 59 by CK II. According to Pinna’s convention (25), Ser62 is a good CK II consensus site due to the presence of acidic residues of N1-6 on its C-terminal side, especially the critical N / 3 residue. Although Ser59 also contains acidic residues, it lacks the critical N / 3 one and is not predicted to be a potential site for CK II. However, phosphorylated Ser62 could provide an acidic environment at N / 3 of Ser59 and set off its phosphorylation. The experiments shown in Fig. 9 supported such a notion. During the time course of phosphorylation, the accumulation of a single phosphate-containing peptide, including those having Lys72 and Arg73 at its C-terminal, focusing at pI 4.0 and 4.3, respectively, occurred significantly ahead of that of doubly phosphorylated
/ 6b11$$9300
01-29-96 13:40:16
arca
This paper demonstrates the cDNA cloning of a novel phosphoprotein HASPP28. Initially, we identified this protein from the acid- and heat-stable pool of rat brain extract as an in vitro PKC substrate; the cDNA-deduced amino acid sequence of PP28 contains consensus phosphorylation sites for PKA and CK II as well. Indeed, both the purified rat brain and recombinant PP28 can serve as substrates for these kinases in vitro, albeit with different stoichiometry of phosphorylation. Importantly, in mouse neuroblastoma N1E115 cells, metabolically [32P]PO4-labeled and immunoprecipitated PP28 revealed a tryptic peptide pattern similar to that of the in vitro CK II phosphorylated rPP28. Even when cells were treated with the tumor-promoting phorbol ester PMA, an activator of PKC, or with a cell-permeable 8Br-cAMP, an activator of PKA, and both reagents caused characteristic changes in cell morphology and kinase activities (H. Li and F. L. Huang, unpublished data), we consistently observed a CK II-reaction derived phosphopeptide pattern. It seemed that PKA and PKC, with or without stimulation, do not phosphorylate PP28l in the N1E115 cells. Although the biological function of PP28 is not presently known, phosphorylation probably plays a regulatory role, and most likely CK II or CK II-like kinase is responsible for its phosphorylation in vivo. CK II is ubiquitously present in all types of eukaryotic cells. It constitutively phosphorylates a variety of nuclear and cytosolic substrates involved in all aspects of transcription, translation, and metabolic regulation (see review of Ref. 32). It also phosphorylates many cell-cycledependent proteins such as P53 (33) and has been implicated as a critical component for regulating the cell cycle (34). This notion is further supported by the recent findings that the cell-cycle-dependent cdc2 kinase can phosphorylate and activate CK II in vitro (35). In intact cells, cdc2 kinase activity reaches a peak level during G2/M transition, and at this stage CK II becomes extensively phosphorylated by cdc2 kinase and is activated (36). Increased phosphorylation of topoisomerase by CK II during the same period of the cell cycle has been demonstrated (37). It has also been reported that stimulation of intact cells with mitogenic agents, including insulin, insulin-like growth factor, and epidermal growth factor
AP: Archives
CLONING OF A NOVEL CASEIN KINASE II SUBSTRATE, HASPP28
139
FIG. 8. Determination of casein kinase II phosphorylation sites of PP28. (A) Reverse-phase HPLC profile of tryptic phosphopeptides derived from CK II phosphorylated PP28 as separated on a C18 column. The dotted line indicates the elution gradient of acetonitrile. The continuous tracing is the absorbance at 214 nm. Radioactivities in the fractions were determined by Cerenkov counting (l). Four short bars under the radioactivity peak represent fractions 53–56, whose peptide sequences were determined as shown in B. The bold S in the sequences represents a phosphorylated serine residue based on the high ratio of PTH-dehydro-Ala/DTT adduct to PTH-Ser at the presumed cycle. (C) Autoradiogram of isoelectric focusing gel analysis of these fractions; rPP28 is the total phosphopeptides before HPLC fractionation.
(38), increases CK II activity. Correlation of PP28 phosphorylation by CK II under the influence of these hormones and at different stages of the cell cycle may help us to elucidate the function of this protein. It is interesting to note that the phosphorylation of Ser59 by CK II takes place later than that of Ser62 in
FIG. 9. Sequential phosphorylation of Ser62 and Ser59 of rPP28 by casein kinase II. At various time points during the course of phosphorylation, aliquots of sample were removed. Proteins in the samples were TCA precipitated as described in the legend of Fig. 6, trypsin digested, and analyzed by isoelectric focusing gel.
/ 6b11$$9300
01-29-96 13:40:16
arca
PP28 (Fig. 9). Presumably, the presence of a phosphate on Ser62 facilitated the phosphorylation of Ser59 by providing CK II with the critical N / 3 acidic residue, as has been described for many other CK II substrates such as topoisomerase and DARPP-32 (37, 39). This facilitatory effect, likely operating in intact cells (37, 39 and the present study), could be physiologically important, as phosphorylation greatly increases the total negative charge of the protein. Phosphorylation of Ser59 and Ser62 in PP28, for example, results in a stretch of 11 consecutive acidic residues that may generate a high Ca2/-binding site (see next paragraph about PEST). Another facilitatory effect has also been described, in that phosphorylation of a substrate by CK II often triggers subsequent phosphorylation of the substrate by a second kinase (32). For example, phosphorylation of inhibitor-2 of phosphoprotein phosphatase by CK II ensures its further phosphorylation by GSK-3, resulting in the expression of its full function (40). It is tempting to speculate that the phosphorylation of PP28 can also lead to phosphorylation by other kinases. In vitro, PP28 can be phosphory-
AP: Archives
140
SHEN ET AL.
lated by PKA and PKC without prior phosphorylation assuming D Å E, and phosphorylated S Å D or E, except by CK II; however, such reactions did not seem to take I of residue 46 of SPT6. It would certainly be worthplace in N1E115 cells under our present testing condi- while to probe the possible nuclear localization of PP28 tions. Since phosphopeptides resulting from CK II phos- as well as its cell-cycle-related activities. phorylation are the only products observed from the metabolic labeling, it is reasonable to conclude that CK REFERENCES II, or its like, is the most likely regulator of PP28. 1. Krebs, E. G. (1987) Annu. Rev. Biochem. 56, 567–613. There is evidences which points to the regulation of 2. Hu, Z.-W., Shi, X.-Y., Sakaue, M., and Hoffman, B. B. (1993) J. protein turnover by CK II phosphorylation (41). Many Biol. Chem. 268, 3610–3615. proteins with PEST sequences (41), regions rich in pro3. Ozawa, K., Szallasi, Z., Kazanietz, M. G., Blumberg, P. M., line (P), glutamic acid (E), serine (S), and threonine Mischak, H., Mushinski, J. F., and Beaven, M. A. (1993) J. Biol. (T), contain CK II phosphorylation sites and often have Chem. 268, 1749–1756. short half-lives (41). PP28 has a very high PEST score 4. Hordijk, P. L., Verlaan, I., Jalink, K., van Corven, E. J., and Moolenaar, W. H. (1994) J. Biol. Chem. 269, 3534–3538. of 15.82 (program PC/GENE) at the region of residues 5. Bubien, J. K., Jope, R. S., and Warnock, D. G. (1994) J. Biol. Ala33 –Pro50 which locates immediately N-terminal to Chem. 269, 17780–17783. its CK II phosphorylation sites of Ser59 and Ser62. Ac6. Nemenoff, R. A., Winitz, S., Qian, N.-X., Van Putten, V., Johncording to the PEST model, CK II phosphorylation genson, G. L., and Heasley, L. E. (1993) J. Biol. Chem. 268, 1960– 2/ erates a high affinity binding site for Ca , which sub1964. sequently activates the protease calpain resulting in 7. Rascon, A., Degerman, E., Taira, M., Meacci, E., Smith, C., protein degradation (34, 41). In this regard, it is worth Manganiello, V., Belfrage, P., and Tornqvist, H. (1994) J. Biol. mentioning that we found a stretch of 10 identical Chem. 269, 11962–11966. amino acids between PP28 and calpastatin, the endoge8. Blackshear, P. J. (1993) J. Biol. Chem. 268, 1501–1504. nous calpain inhibitor precursor. Although this se9. Chapman, E. R., Estep, R., and Storm, D. R. (1992) J. Biol. Chem. 267, 25233–25238. quence segment within calpastatin (Gly125 –Leu134) (29) is not located in the calpain-binding domain, it may 10. Baudier, J., Deloulme, J. C., Dorsselaer, A. V., Black, D., and Mattes, H. W. D. (1991) J. Biol. Chem. 266, 229–237. influence its interaction with the protease. This stretch 48 57 11. Huang, K.-P., Huang, F. L., and Chen, H.-C. (1993) Arch. Bioof residues in PP28 (Gly –Leu ) is located between its chem. Biophys. 305, 570–580. CK II phosphorylation sites Ser59 and Ser62 and PEST region (Ala33 –Pro50). While it is interesting to test if 12. Huang, K.-P., Nakabayashi, H., and Huang, F. L. (1986) Proc. Natl. Acad. Sci. USA 83, 8535–8539. phosphorylation of PP28 would control its turnover 13. Huang, K.-P., and Huang, F. L. (1986) Biochem. Biophys. Res. rate, it would also be important to test for a regulatory Commun. 139, 320–326. role of these identical residues. 14. Huang, K.-P., Akatsuka, A., Singh, T. J., and Blake, K. (1983) PP28 also shares some homology with neuron-speJ. Biol. Chem. 258, 7094–7101. cific filaments of 70/60 kDa (28), troponin T (27), and 15. Huang, K.-P., Itarte, E., Singh, T., and Akatsuka, A. (1982) J. Biol. Chem. 257, 3236–3242. EMB-5 (30) (SPT6 in yeast). Since the Northern blot analysis did not point to its exclusive neuronal location, 16. Huang, F. L., Huang, K.-P., Sheu, F.-S., and Osada, K.-I. (1993) in Methods in Neuroscience (Fain, J. N., Ed.), Vol. 18, Lipid PP28 probably does not assume a neuron-specific funcMetabolism in Signaling System, pp. 127–137, Academic Press, tion like neurofilament. Troponin T is considered to San Diego. 2/ be required for the Ca -dependent ATPase activity in 17. Sanger, F. S., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. muscle contraction through interaction with tropomyoAcad. Sci. USA 74, 5463–5467. sin. The molecule of troponin T can be subdivided into 18. Borson, N. D., Salo, W. L., and Drewes, L. R. (1992) PCR Methtwo domains, T1 (residues 1–158) and T2 (residues ods Appl. 2, 144–148. 159–259). It is T1 that interacts with tropomyosin (31), 19. Lehrach, H., Diamond, D., Wozney, J. M., and Boedfker, H. and it is T1 that exhibits homology with PP28. It would (1977) Biochemistry 16, 4743–4751. be interesting to test if PP28 interacts with tropomyo- 20. Kao, P. N., Chen, L., Brock, G., Ng, J., Kenny, J., Smith, A. J., and Corthesy, B. (1994) J. Biol. Chem. 269, 20691–20699. sin. EMB-5 protein of C. elegans is similar to the yeast nuclear protein SPT6, in that both affect transcription 21. Konigsberg, W. H., and Henderson, L. (1983) Methods Enzymol. 91, 254–259. in the control of the cell cycle. Importantly, both EMB22. Ishizaki, J., Hanasaki, K., Higashino, K., Kishino, J., Kikuchi, 5 and SPT6 have a stretch of acidic amino acids at the N., Ohara, O., and Arita, H. (1994) J. Biol. Chem. 269, 5897– N-terminal portion which contain CK II phosphorylata5904. ble serine residues. In fact, residues 58–68 of PP28, 23. Woodget, J. R., Gould, K. L., and Hunter, T. (1986) Eur. J. residues 44–54 of EMB-5, and residues 39–49 of SPT6 Biochem. 161, 177–184. are in good alignment as shown below: 24. Kishimoto, A., Nisshiyama, K., Nakanishi, H., Uratsuji, Y., No58 DSDESEDEDDD68 of PP28 mura, H., Takeyama, Y., and Nishizuka, Y. (1985) J. Biol. 44 Chem. 260, 12492–12499. SSDEDEDDDDD54 of EMB-5 39 SSEEDEDIDED49 of SPT6, 25. Pinna, L. A. (1990) Biochim. Biophys. Acta 1054, 267–284.
/ 6b11$$9300
01-29-96 13:40:16
arca
AP: Archives
CLONING OF A NOVEL CASEIN KINASE II SUBSTRATE, HASPP28 26. Glass, D. B., El-Maghrabi, M. R., and Pilkis, S. J. (1986) J. Biol. Chem. 261, 2987–2993. 27. Pearlstone, J. R., Carpenter, M. R., and Smillie, L. B. (1986) J. Biol. Chem. 261, 16795–16810. 28. Szaro, B. G., Pant, H. C., Way, J., and Battey, J. (1991) J. Biol. Chem. 266, 15035–15041. 29. Emori, Y., Kawasaki, H., Imajoh, S., Imahori, K., and Suzuki, K. (1987) Proc. Natl. Acad. Sci. USA 84, 3590–3594. 30. Nishiwaki, K., Sano, T., and Miwa, J. (1993) Mol. Gen. Genet. 239, 313–322. 31. Heeley, D. H., Golosinska, K., and Smillie, L. B. (1987) J. Biol. Chem. 262, 9971–9978. 32. Tuazon, P. T., and Traugh, J. A. (1991) Adv. Second Messenger Phosphoprotein Res. 23, 123–164. 33. Deppert, W., Buschhausen-Denker, G., Patschinsky, T., and Steinmeyer, K. (1990) Oncogene 5, 1701–1706.
/ 6b11$$9300
01-29-96 13:40:16
arca
141
34. Meisner, H., and Czech, M. P. (1991) Curr. Opin. Cell Biol. 3, 474–483. 35. Mulner-Lorillon, O., Cormier, P., Labbe, J., Poulhe, R., Osborne, H., and Belle, R. (1990) Eur. J. Biochem. 93, 529–534. 36. Litchfield, D. W., Luscher, B., Lozeman, F. J., Eisenman, R. N., and Krebs, E. G. (1992) J. Biol. Chem. 267, 13943–13951. 37. Cardenas, M. E., and Gasser, S. M. (1993) J. Cell Sci. 104, 219– 225. 38. Klarlund, J. K., and Czech, M. P. (1988) J. Biol. Chem. 263, 15872–15875. 39. Girault, J., Hemmings, H., Jr., Williams, K. R., Nairn, A. C., and Greengard, P. (1989) J. Biol. Chem. 264, 21748–21759. 40. DePaoli, A. A. (1984) J. Biol. Chem. 259, 12144–12152. 41. Rogers, S., Wells, R. M., and Rechsteiner, M. (1986) Science 234, 364–368.
AP: Archives
Vol. 327, No. 1, March 1, pp. 131–141, 1996 Article No. 0101
Molecular Cloning and Characterization of a Novel Casein Kinase II Substrate, HASPP28, from Rat Brain1 Li Shen, Kuo-Ping Huang, Hao-Chia Chen, and Freesia L. Huang2 Endocrinology and Reproduction Research Branch, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland 20892
Received August 2, 1995, and in revised form November 28, 1995
HASPP28 (heat- and acid-stable phosphoprotein of 28 kDa) has been purified to near homogeneity from the acid-stable protein fraction of rat brain extract. Based on the N-terminal 40 amino acid sequence, a pair of highly degenerate primers was used to generate a 107-bp probe from rat brain RNA by RT–PCR. From the rat brain lgt11 library, this probe identified two positive clones that together provided a cDNA of 837 bp with an open reading frame of 546 bp. This cDNA was extended by 3*RACE to 1.2 kb that included a polyadenylation signal and a poly(A) tail. The 180amino-acid sequence derived from the open reading frame, which did not correspond to any known protein, was predicted to have phosphorylation sites for protein kinase C, casein kinase II (CKII), and protein kinase A. Indeed, both the purified rat brain HASPP28 and the recombinant HASPP28 (rHASPP28) can be phosphorylated by these kinases. Northern blot analysis indicated that HASPP28 was present in all rat tissues tested, including those from the brain, lung, spleen, kidney, liver, heart, and muscle, in decreasing order of abundance. Phosphopeptide analysis of rHASPP28 phosphorylated in vitro by various kinases showed different tryptic peptide patterns on two-dimensional mapping and isoelectric focusing gels. From [32P]PO4-labeled N1E115 neuroblastoma cells, HASPP28 can be immunoprecipitated with a polyclonal antiserum raised against rHASPP28. The immunoprecipitated protein showed a phosphopeptide pattern similar to that of rHASPP28 phosphorylated by CK II in vitro. Furthermore, the immunoprecipitates from cells treated with phorbol 12-myristate 13-acetate or 8-bromo-cAMP did not show any increased phosphorylation over those of untreated ones, and the
phosphopeptide patterns of the immunoprecipitates again were similar to that of CK II phosphorylated protein. These results suggest that HASPP28 is a novel phosphoprotein that can be phosphorylated by several kinases in vitro. In intact cells, CK II seems to be solely responsible for the phosphorylation of HASPP28. q 1996 Academic Press, Inc.
Protein kinases and kinase-catalyzed reactions play major roles in the regulation of a great variety of cellular processes. Several well-studied protein kinases, such as protein kinase C (PKC),3 cAMP-dependent protein kinase (PKA), Ca2//CaM dependent protein kinases, casein kinases, MAP kinases, and ligand receptor-associated tyrosine kinases, mediate the responses of eukaryotic cells to external stimuli by phosphorylating their specific protein substrates (for review see Ref. 1). There is, however, only limited knowledge available in the identification of specific protein substrate(s) relevant to a particular cellular response regulated by a kinase. For example, the involvement of PKC or PKA in numerous cellular functions has been demonstrated through the enhancement of the specific responses by the treatment of phorbol 12-myristate 13-acetate (PMA), an activator of PKC, or permeable cAMP analogues, activators of PKA, respectively (2–5). Most of the time, the target substrate(s) being phosphorylated by PKC or PKA was not identified. On the other hand, for many purified proteins, whose phosphorylations by kinases were well characterized in vitro, evidence for in vivo participation of such reactions is still unavailable (6, 7). 3
1
Sequence data from this article have been deposited with the EMBL/GenBank Data Libraries under Accession No. U26541. 2 To whom correspondence should be addressed at Building 49, Room 6A36, National Institutes of Health, Bethesda, MD 208924510. Fax: (301) 480-8010.
Abbreviations used: PKC, protein kinase C; PKM, the proteasedegraded PKC; PKA, cAMP-dependent protein kinase; CK, casein kinase; PMA, phorbol 12-myristate 13-acetate; TFA, trifluoroacetic acid; AEBSF, 4-(2-aminoethyl)benzenesulfonyl fluoride; RT–PCR, reverse transcription–polymerase chain reaction; RACE, rapid amplification of cDNA end; ECL, enhanced chemiluminescence. 131
0003-9861/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
/ 6b11$$9300
01-29-96 13:40:16
arca
AP: Archives
132
SHEN ET AL.
In our studies to identify additional potential PKC substrates, we have purified several proteins from the heat- and acid-stable fraction of rat brain extract. In addition to the well-characterized proteins, including MARCKS (8), neuromodulin (9), and neurogranin (10, 11), we have purified a Mr Å 28k PKC substrate to near homogeneity. Since the sequence of its N-terminal 40 amino acids did not correspond to any known protein in the Genetics Computer Group (GCG, Madison, WE) data bank (SwissProt, release 31), we anticipated it to be a novel PKC substrate. This paper describes both the cDNA cloning of this protein and investigations of its properties of phosphorylation in vitro and in intact cells. This protein, designated as HASPP28 (for heatand acid-stable phosphoprotein of 28 kDa), or PP28 for short, is an excellent substrate of casein kinase II (CK II), capable of also being phosphorylated by PKC, PKA, and CK I in vitro. Under normal culturing conditions, PP28 is present as a phosphoprotein in intact cells, and CK II or CK II-like kinase is the kinase responsible for its phosphorylation. EXPERIMENTAL PROCEDURES Materials. [g-32P]ATP and [32P]PO4 were from DuPont NEN. [aS]ATP and ECL reagent were from Amersham. [a-32P]GTP for probe labeling was from ICN. Ampholine for isoelectric focusing gels was from LKB-Pharmacia. Protein A–agarose was from Sigma. Restriction enzymes were from New England Biolab. PKC (12), protease-degraded PKC (PKM) (13), PKA and CK I (14), and CK II (15) were prepared as described. Sources of other materials used are indicated in the following procedures at their first appearances.
35
Purification of PP28 from rat brain. Initial extraction of PP28 from frozen rat brain and acid precipitation step were similar to methods described for the purification of neurogranin and other PKC substrates (11, 16). The concentrated acid-stable protein fraction was applied to a DEAE–cellulose column, packed in a Pharmacia HR 16/ 10 column, and eluted with a 0–1 M KCl gradient in buffer A (20 mM Tris–Cl, pH 7.5, containing 1 mM DTT, 0.5 mM EDTA, and 0.5 mM EGTA), delivered by a Pharmacia FPLC unit. The effluent fractions were assayed for PKC substrate by phosphorylation with PKC, SDS–PAGE, and autoradiography. Fractions eluted between 0.1–0.2 M KCl, where PP28 was located, were concentrated by ultrafiltration. PP28 was purified to near homogeneity by subsequent chromatography on a C4 reverse-phase column (Vydac 214TP510, 10 1 250 mm), where it was eluted at 33–35% acetonitrile from a linear gradient composed of 0.1% trifluoroacetic acid (TFA)/0.1% TFA / 100% acetonitrile. The purified protein showed an apparent Mr Å 28,000 on SDS–PAGE (10–20% gradient gel). The N-terminal amino acid sequence of the purified protein was determined using a gasphase protein sequencer with an on-line PTH–amino acid analyzer (Applied Biosystems Model 470A). cDNA library screening and DNA sequence analysis. A pair of PCR primers based on the N-terminal amino acid sequence of the purified protein were synthesized. The forward primer, 5*-CGI AA(AG) GGI GGN CA(TC) AA(AG) GG-3* (where N is A/T/G/C), corresponding to the sequence of RKGGHKG (residues 5–11), and the reverse primer, 5*-TG (TC)TG (TC)TC (AG)TC (TC)TC (TC)TC (AG)TT-3*, corresponding to the sequence of NEEDEQQ (residues 34–40), were used to produce a cDNA probe of expected 107 bp in length from rat brain RNA by RT–PCR (see Fig. 2). The PCR product was directly cloned into the pCRvector (Invitrogen) and the nucleo-
/ 6b11$$9300
01-29-96 13:40:16
arca
tide sequence was determined by dideoxy-mediated chain termination method (17) using Sequenase version 2.0 (U.S. Biochemical). Primers of SP6 and T7 were used in the sequencing reaction. The lgt11 cDNA library of rat brain (Clontech) was screened as described by the manufacturer. Using [a-32P]GTP-labeled 107 bp cDNA as a probe, two strongly positive clones were obtained from about 1 1 106 recombinant phage plaques. The inserts of both clones were excised by EcoRI and each was subcloned into the plasmid pGEM-3Zf(0) (Promega). The DNA sequence was determined from both directions (see strategy in Fig. 2). One clone, l7, having a 503bp insert, contained the translational initiation codon of ATG, and the other, l8, having an 808-bp insert, contained a TAA stop codon. The pGEM–pp28 plasmid, containing the entire open reading frame, was constructed by ligation of a 217-bp EcoRI/XbaI DNA fragment (excised from the ATG containing clone) with a 652-bp XbaI/BamHI fragment (excised from the TAA containing clone) and cloned into EcoRI and BamHI sites of plasmid pGEM-3Zf(0). 3*RACE–PCR. A clone containing the 3*-noncoding region was generated by rapid amplification of cDNA end method (18) using 3*-AmpliFINDER RACE kit from Clontech. The NN01oligo(dT)CDS primer provided in the kit was used in the initial reverse transcriptional reaction. The forward primer used in the primary PCR reaction was 5*CTGTGGGAGGAGATGCC (553–569), for the secondary PCR, 5*AGCTGCTATCTTTGAGA (736–752). In both PCR reactions, the provided Anchor Primer served as the reverse primer. The secondary PCR product was directly cloned into pCRvector and sequenced. Northern blot analysis. Total RNA from different rat tissues were prepared with RNAzol (Tel-Test, Inc.) using the protocol described by the manufacturer. The uniformly labeled cDNA of 666 bp, excised from pGEM-pp28 with EcoRI/PstI, was used as a probe to hybridize the immobilized RNA (19). Expression and purification of recombinant PP28 protein. The bacterial expression plasmid, pRSETC–pp28, was constructed by PCR amplification of the insert of pGEM–pp28, using the oligonucleotide with a NdeI site fused to the 5* end of the first seven codons of the PP28 cDNA as the forward primer and the SP6 oligonucleotide as the reverse primer. The resulting PCR product was digested with NdeI/PstI and then cloned into the NdeI/PstI sites of pRSETC (Invitrogen). Escherichia coli strain BL21(DE3) (Novagen) was used to express the recombinant PP28 protein (rPP28). The transformed bacteria were grown in 2xYT medium containing 2% glucose and 100 mg/ml ampicillin. Cells from 500 ml of culture were lysed in 25 ml of 50 mM Tris–Cl, pH 7.5, containing 1 mM EDTA, 1 mM EGTA, 1 mM DTT, and 50 mM AEBSF by the addition of lysozyme to a final concentration of 200 mg/ml. The suspension was treated in boiling water for 15 min and centrifuged to remove the precipitated protein; the supernatant fluid was then lyophilized. The sample was redissolved in 0.1% TFA and first applied to a C4 reverse-phase precolumn, washed with 0.1% TFA, and then eluted with 100% acetonitrile containing 0.1% TFA and was lyophilized again. The solid sample was again dissolved in 0.1% TFA and separated on a HPLC C4 reversephase column as described for the purification of the rat brain PP28. Final purification was achieved by chromatography on a Mono Q column (FPLC system, Pharmacia-LKB) that was equilibrated with buffer A, containing 10% glycerol, and eluted by a 0–0.45 M KCl gradient in the same buffer. Recombinant PP28 was eluted at about 0.23 M KCl. The eluent was desalted by C4 reverse-phase precolumn and lyophilized. The purified rPP28 was used to immunize rabbit for raising polyclonal antibody. Protein kinase phosphorylation reactions in vitro. Phosphorylations by CK I and II, PKA, and PKC were carried out in reaction mixtures containing 30 mM Tris–Cl, pH 7.5, 6 mM Mg(OAc)2 , and 120 mM [g-32P]ATP. Reactions for PKC also contained 0.4 mM CaCl2 , 100 mg/ml phosphatidylserine, and 20 mg/ml dioleoylglycerol, while those for PKA contained 10 mM cAMP. Reactions were carried out at 307C. For autoradiographic detection of proteins, reactions were terminated by the addition of 0.25 vol of five times concentrated
AP: Archives
CLONING OF A NOVEL CASEIN KINASE II SUBSTRATE, HASPP28 SDS–PAGE sample buffer and were heated for 5 min at 1007C and analyzed by SDS–PAGE. Endogenous phosphorylation and immunoprecipitation. Mouse neuroblastoma N1E115 cells were grown in DMEM medium containing 10% fetal calf serum in 35-mm dishes. After reaching Ç70% confluence, cells were washed three times with phosphate-free DMEM medium and incubated with 400 mCi carrier-free [32P]PO4 in 0.8 ml of the same medium. After 3 h of incubation, either PMA or 8-Br-cAMP was added to give a final concentration of 1 mM or 1 mM, respectively, and incubation was continued for another 15 min. Cells were washed three times with 10 mM Tris–Cl buffered saline and lysed in 200 ml of 20 mM Tris–Cl, pH 7.5, containing 1 mM EDTA, 1 mM EGTA, 150 mM NaCl, 100 mM NaF, 0.5% NP-40, and 50 mM AEBSF. Cell lysates were centrifuged in an Eppendorf centrifuge for 5 min, and SDS was added to the supernatant (containing 250 mg protein) to give a final concentration of 0.1%. Antiserum (10 ml), raised against purified rPP28, was added and incubated at 47C overnight with gentle rocking. Protein A cross-linked agarose bead slurry (40 ml) was added to the mixture and incubation was continued at room temperature for 2 h more. The agarose beads were washed four times with the cell lysis buffer containing 0.1% SDS and then heated with 50 ml of five strength SDS–PAGE sample buffer at 1007C for 5 min. After brief centrifugation to remove the agarose beads, supernatant was subjected to SDS–PAGE and proteins were electrophoretically transferred to Immobilon-P membrane (Millipore). Labeled PP28 was located by autoradiography, excised after the membrane was blocked with polyvinylpyrrolidone-40 (20), and subjected to tryptic digestion for analysis. Phosphopeptide analysis by isoelectric focusing gel and 2-D mapping. Isoelectric focusing gel analysis was carried out with a LKB Multiphor apparatus. Polyacrylamide gels of pH range 2.5 to 6.0 was used to analyze phosphopeptides. For thin-layer 2-D mapping, the first dimension was electrophoresis in pH 6.5 buffer (acetic acid:pyridine:H2O Å 1:25:225); the second dimension was ascending chromatography in the upper phase of freshly prepared n-butanol:acetic acid:H2O (4:1:5). Determination of phosphorylation sites. Purified rPP28 (200 mg) was phosphorylated by CK II in a 200-ml reaction mixture containing 30 mM Tris–Cl, pH 7.5, 6 mM Mg(OAc)2 , and 120 mM [g-32P]ATP (3000 cpm/pmole). After the addition of 50 ml of five strength SDS– PAGE sample buffer to stop the reaction, samples were heated at 1007C for 5 min and then subjected to SDS–PAGE (10% gels). The phosphorylated protein band was located by autoradiography without staining or fixing the gel. Gel pieces were excised and homogenized using a glass-teflon homogenizer. The homogenized gel was extracted three times with acetone:water:acetic acid:triethylamine(17:1:1:1) to remove SDS (21) and then dried and reswollen in 0.5 ml of 50 mM NH4HCO3 , pH 7.8. TPCK-treated trypsin (10 mg) was added to the gel suspension and was incubated at 307C for 5 h before another 10 mg of fresh trypsin was added. The proteolysis was continued overnight. Trypsinized peptides were collected by centrifugation to remove the acrylamide gel pellet, which was washed three times with 50 mM NH4HCO3 . The combined tryptic digests were lyophilized and were separated on a Vydac C18 reverse-phase HPLC column equilibrated with 0.1% TFA. Elution was carried out with a linear gradient of 0–50% acetonitrile in 0.1% TFA in 90 min. The radioactivity was determined by Cerenkov counting and fractions of the major radioactive peak were subjected to amino acid sequence analysis and isoelectric focusing gel analysis.
RESULTS
Purification of rat brain PP28. PP28 was purified from the 2.2% perchloric-acid-soluble fraction of rat brain extract. The fraction was first chromatographed on a DEAE–cellulose column and then on a C4 reverse-
/ 6b11$$9300
01-29-96 13:40:16
arca
133
phase column. Figure 1 shows a typical fractionation profile on HPLC. Phosphorylation assay by PKC, SDS– PAGE, and autoradiography indicated that several proteins are phosphorylatable by PKC. As shown in the inset of Fig. 1, a major protein band (Mr Å 28k) appeared in fractions 81 and 82, and both were phosphorylated by PKC. Proteins in fraction 81 were separated by SDS–PAGE, transferred onto Immobilon-P, and the PP28 protein band located after staining was excised for microsequencing (sequence shown in Fig. 3). Isolation and characterization of cDNA encoding PP28. As the N-terminal 40-amino-acid sequence of the purified rat brain protein PP28 does not correspond to any known protein sequence from the protein sequence data bank (SwissProt release 31), cloning and sequencing of its cDNA were carried out to determine the complete protein sequence. In order to simplify the library screening, we first prepared a 107-bp cDNA fragment encoding the N-terminal part of the protein from rat brain RNA by RT–PCR (see Fig. 2 for cloning strategy and Fig. 3 for the corresponding sequences). In spite of the high degeneracy of the primers, the 107bp fragment turned out to be the only product of the PCR and its deduced sequence matched amino acid 5– 40 of the determined sequence. Using this cDNA fragment as a probe for library screening, two positive clones, l7 and l8, were obtained from 106 phage plaques. Clone l7 (bp 027 to 476) contained an ATG start codon, while clone l8 (bp 3 to 810) contained a TAA stop codon, collectively, they provided the complete open reading frame of PP28 cDNA (Fig. 3). The sequence, however, did not exhibit a poly(A) tail or polyadenylation signal. It should be noted that both l7 and l8 contained the GAA codon, denoting glutamic acid as their 40th codon after ATG instead of CAA, which denotes glutamine as originally determined and used in the reverse primer in the RT–PCR. During sequencing, 40 cycles of Edman degradation may have weakened the signal, thus, leading us to incorrectly assign glutamine instead of glutamic acid as the 40th residue. Incidentally, the 39th residue is also glutamine; the carryover might give rise to the ambiguity of such a sequence. In order to verify the presence of such a sequence in the tissue, RT–PCR was performed using total RNA from rat brain, kidney, and liver and mouse neuroblastoma cells N1E115. Sequences of PCR products, either RT–PCR107 or RT–PCR661 (Fig. 2), from all the RNAs tested were all identical to those originally determined from the cloned cDNA (l7 and l8). Furthermore, the RT–PCR661 products obtained from several tissue RNAs all have GAA as their 40th codon, thus, glutamic acid was assigned as the 40th residue of the protein. Since the 3* end of the clone l8 (810 bp) did not
AP: Archives
134
SHEN ET AL.
FIG. 1. Reverse-phase HPLC on C4 column for the purification of PP28 from rat brain. The perchloric-acid-soluble proteins of rat brain extract were first separated on a DEAE–cellulose column before applying to the C4 reverse-phase column as described under Experimental Procedures. Dotted line shows the profile of the gradient in percentage of acetonitrile. Effluent fractions were assayed for PKC substrate (l) as previously described (11, 16). PKC phosphorylated proteins were also analyzed by SDS–PAGE; the inset shows the protein staining and autoradiography for fractions 81 and 82, the major fractions of PP28. Subsequently, proteins in fraction 81 were separated by SDS– PAGE and transferred to an Immobilon-P membrane, and PP28 was excised for microsequencing.
contain a poly(A) tail or obvious polyadenylation signal, 3*RACE–PCR with RNAs from both brain and kidney were performed. We were able to extend 382 bp from the 3* end of clone l8 and to identify a putative polyade-
nylation signal (22) of ‘‘ATTAAA,’’ 10 bp upstream of the poly(A) tail (Fig. 3). The combined cDNA clone (Fig. 3) has approximately 1.2 kb, with a 5* noncoding region of 27 bp, a 3* noncod-
FIG. 2. Cloning and sequencing strategy. Bases are numbered from the start of translation in base pairs. l7 and l8 are independent cDNA clones obtained from rat brain lgt11 cDNA library screening. Arrows indicate the directions and lengths of individual sequencing runs. RT–PCR107 was the initial 107-bp RT–PCR product from rat brain RNA used for library screening and was later verified together with RT–PCR661 in RNAs of rat brain, kidney, liver, and neuroblastoma N1E115 cells. 3 *RACE denotes the 3*RACE–PCR product from RNAs of brain and kidney to extend the 3 * untranslated region of cDNA to the polyadenylation tail. Primers used in these PCR reactions were synthesized according to the sequences of l7 and l8 (see Experimental Procedures for sequence and location).
/ 6b11$$9300
01-29-96 13:40:16
arca
AP: Archives
CLONING OF A NOVEL CASEIN KINASE II SUBSTRATE, HASPP28
FIG. 3. Nucleotide sequence of cDNA and deduced amino acid sequence of PP28. Amino acids are shown in the single-letter code and numbered from the first codon after ATG. Underlined sequence corresponds to the 40 N-terminal amino acids of the purified rat brain PP28 as determined by automatic Edman degradation. The 40th residue (E), was originally determined as Q (see text). The sequence of 107 bp obtained from RT–PCR, which was used in the library screening, is indicated in bold; among these nucleotides, 16– 35 and 103–122 were regions used for the degenerate forward and reversed primers (see Experimental Procedures for sequences). A portion of the 3* untranslated sequence, from 811 bp to poly(A) tail, obtained from 3*RACE–PCR, is presented in lowercase. A polyadenylation signal of ‘‘attaaa’’ is indicated by a double underline.
ing region of 646 bp, and a coding region of 546 bp. Such an open reading frame encodes a polypeptide of 181 amino acids. As neither the purified PP28 from rat brain nor rPP28 started with methionine, the second residue of proline was counted as the first amino acid and the polypeptide was predicted to contain 180 amino acids. Sequence analysis. Although the molecular weight of the polypeptide was calculated to be 20,491, both the purified rat brain protein and the recombinant one migrated on SDS–PAGE as having 28 kDa. Among the 180 amino acids, there are 42 acidic and 42 basic residues, distributed as 26 glutamic acid, 16 aspartic acid, 27 lysine, and 15 arginine, and the isoelectric point is 7.77. Interestingly, there are no cysteine, phenylalanine, or tryptophane residues in the molecule. As expected from the fact that the protein can be phosphorylated by PKC, the sequence analysis (Motif in GCG) revealed four potential phosphorylation sites for PKC: Thr91, Thr92, Ser107, and Ser169 (23, 24). In addi-
/ 6b11$$9300
01-29-96 13:40:16
arca
135
tion, it also predicted five phosphorylation sites for CK II at Thr17, Ser18, Ser62, Thr96, and Ser107 (25). One of these sites, Thr96, is also a predicted site for PKA (26). Since no consensus site for N-glycosylation was found in the protein sequence, the abnormality of the retarded migration on gel electrophoresis cannot be attributed to N-glycosylation. The protein is extremely hydrophilic and the hydrophobicity profile analysis (data not shown) revealed no obvious hydrophobic region that would indicate a membrane-spanning domain. We also subjected the nucleotide sequence to a computer-assisted homology search against a DNA data bank (GCG/FASTA, GenBank release 89) without finding a significant match to any known sequence. The closet similarities as revealed by the ‘‘Bestfit’’ program were 45% to troponin T (27), 43% to a 70-kDa neurofilament (28), 48% to calpain inhibitor precursor calpastatin (29), and 49% to EMB-5 protein (30), whereas the identities were only 21.3, 17.5, 26.8, and 18.3%, respectively. It is worth mentioning that the 10 amino acids, residues 48–57 of PP28, are identical to residues 125–134 of calpastatin. It is also interesting to note that between the two asymmetric domains of troponin T, T1 and T2 (residues 1–158, 159–259, respectively), it was T1 that shared similarity with PP28 and was responsible for the interaction between troponin T and tropomyosin (31). RNA blotting analysis. From RT–PCR107, RT– PCR661, and 3*RACE–PCR reactions we have already learned that the message of the gene appeared to be present in various rat tissues besides the brain, from which the protein was first purified. The Northern blot analysis (Fig. 4) showed that there was only one Ç1.5kb species of messenger RNA present in all the tissues tested. It was most abundant in the cerebrum and cerebellum, moderate in the lung and spleen, and less in the kidney, liver, heart, and muscle (in decreasing order). These results thus confirm the findings from the RT–PCR reactions that the PP28 protein seems to be distributed widely in various tissues. Immunoblot analysis of PP28. Polyclonal antiserum raised against rPP28 and the IgG purified from this serum recognized both the purified rat brain protein and rPP28 in an immunoblot analysis (Fig. 5) while the preimmune serum failed in such an analysis (not shown). The antibodies also cross-reacted with a protein band of similar mobility as the purified PP28 in the supernatants of the heat-treated extracts of N1E115, COS, and NIH3T3 cells. It appears that PP28 is also expressed in various cultured cells of different origins. Phosphorylation of PP28 and tryptic peptide analysis. In vitro, both purified PP28 and rPP28 can be phosphorylated by CK I, CK II, PKA, and PKC; the relative amounts of phosphate incorporations are 0.36,
AP: Archives
136
SHEN ET AL.
FIG. 4. Northern blot analysis for tissue distribution. RNAs (20 mg each) isolated from various rat tissues were denatured with formaldehyde, separated on 1% agarose gel and capillarily transferred to nitrocellulose membrane, and analyzed by hybridization. Probe used in the hybridization is the 666-bp EcoRI/PstI fragment from pGEM– pp28 that covers the entire open reading frame.
1.63, 0.21, and 1, respectively, taking that by PKC is 1 (data not shown). Tryptic phosphopeptide patterns of these phosphorylations were analyzed in order to be able to determine the occurrence of any of such phosphorylation in vivo. Both two-dimensional peptide mapping and isoelectric focusing gel analysis (Fig. 6) revealed different tryptic phosphopeptide patterns when rPP28 was phosphorylated by CK I, CK II, PKA, and PKCa. The peptide map (Fig. 6A) of PKA or PKCa phosphorylated proteins contained only one major spot that migrated toward the cathode, while those of CK I or CK II phosphorylated proteins consisted of multiple spots that migrated toward the anode. The phosphopeptides resulting from CK I or CK II phosphorylation were much more acidic than those from PKA or PKC phosphorylation as demonstrated by the pI’s of these peptides obtained from isoelectric focusing gel analyses (Fig. 6B). PKM, the partially proteolyzed PKC that does not require Ca2/ or lipid for activity, gave a peptide pattern similar to that of PKCa. Phosphorylation of endogenous PP28 in intact cells. Mouse neuroblastoma N1E115 cells were metabolically labeled with [32P]PO4 , and the extracted proteins immunoprecipitated with antiserum and analyzed on SDS–PAGE. Figure 7A shows that whether unstimulated or stimulated with 8-Br-cAMP or PMA, all the cell extracts produced a prominent precipitated protein band having a similar migration as the in vitro phosphorylated rPP28. Thus, under the present culturing
/ 6b11$$9300
01-29-96 13:40:16
arca
conditions, either stimulated or unstimulated, PP28 in N1E115 cells appears to be phosphorylated. Treatment with cAMP or PMA, however, did not seem to increase the radioactivity of the immunoprecipitated protein. To determine which kinase was responsible for the phosphorylation, the labeled protein bands were excised from the membrane, trypsinized, and then analyzed by isoelectric focusing gels. Figure 7B demonstrates that immunoprecipitated PP28, either from control or stimulated cells, exhibited a phosphopeptide pattern similar to that of rPP28 phosphorylated in vitro by CK II. It appears that CK II- or a CK II-like kinase is responsible physiologically for the phosphorylation of PP28. Determination of CK II phosphorylation sites. As five sites were predicted to be phosphorylated by CK II (distributed into four tryptic peptides), and there were at most three major phosphopeptides bands detected by isoelectric focusing gel analysis, it would be interesting to determine what were the sites being phosphorylated. The labeled tryptic peptides were fractionated by reverse-phase HPLC on a C18 column, and the fractions were counted for radioactivities. As shown in Fig. 8A, there is one major peak of radioactivity including fractions 53, 54, and 55, and a minor shoulder peak at fractions 56 and 57 (collectively, these fractions comprised ú95% of the total radioactivities). These fractions were subjected to microsequencing. Surprisingly, all these fractions yielded one single amino acid sequence corresponding to Lys52 –Lys72 (Fig. 8B), ex-
FIG. 5. Immunoblot analysis of PP28. Proteins were separated on SDS–PAGE (10% gel), transferred onto nitrocellulose membrane, blotted with anti-PP28 antiserum (1:2000 dilution), and then detected with ECL (Amersham). Lane 1, 0.25 mg of purified rat brain PP28 (fraction 81 from Fig. 1); lane 2, 0.25 mg of purified recombinant PP28; lanes 3–5 are heat stable proteins (10 min, 1007C) from 500 mg extracted proteins of N1E115 (lane 3), COS (lane 4), and NIH3T3 (lane 5) cells. The buffer used for the cell extraction was 20 mM Tris– Cl, pH 7.5, containing 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 0.5% NP-40, and 50 mM AEBSF.
AP: Archives
CLONING OF A NOVEL CASEIN KINASE II SUBSTRATE, HASPP28
137
FIG. 6. Two-dimensional peptide maps and isoelectric focusing gel analysis for phosphopeptides of rPP28. Recombinant PP28 (200 mg) was phosphorylated by various protein kinases in the presence of [g-32P]ATP as described under Experimental Procedures. As phosphorylation rates of PP28 by different protein kinases are dissimilar, unitages of the various kinases used in this experiment were adjusted to yield similar phosphate incorporation except the reaction with PKA, in which the incorporation was approximately half of those of others. The phosphorylation reaction was terminated by the addition of TCA to a final concentration of 15%. The precipitated proteins were dissolved in cold 0.1 N NaOH and precipitated again to eliminate free ATP before trypsin digestion. Phosphopeptides were analyzed by two-dimensional peptide mapping (A) and isoelectric focusing gels (B). In A, ‘‘x’’ denotes the origin of sample application; in B, the pI indicates the pH value at the end of focusing, as measured by the surface electrode.
cept that fraction 56 contained Arg73 as its C-terminal. All these peptides contained a phosphorylated Ser62, while fraction 53 also had a phosphorylated Ser59 based on the high ratio of PTH–Ser* (Å PTH-dehydro-Ala/ DTT adduct) to PTH–Ser at the presumed site, as compared with other serine residues in the sequenced peptide. These phosphopeptides were also analyzed by isoelectric focusing on polyacrylamide gel as shown in Fig. 8C. The major band of fraction 53, focused toward the
/ 6b11$$9300
01-29-96 13:40:16
arca
most acidic side of the gel with a pI of 3.8, is probably the species containing two phosphorylated serines, Ser59 and Ser62. The other band of fraction 53, focused at pH 4.0 where the major band of fraction 54 is located, is probably the species containing a single phosphorylated Ser62. The major band in fraction 56, having a pI of 4.3, is most likely the species having a phosphorylated Ser62 and a C-terminal Arg73. It appears that Ser62 of PP28 is most favorable for CK II phosphorylation,
AP: Archives
138
SHEN ET AL.
peptide of pI 3.8. Also noted is that the accumulation of the peptide with pI of 4.3 depended on the amount of trypsin used in digestion. In the experiments of Figs. 6, 7, and 9, where the trypsin/protein ratio was 1/5, more extensive proteolysis occurred and resulted in less accumulation of the pI 4.3 peptide, whereas in the experiment of Fig. 8, trypsin to protein ratio was lowered to 1/20 for sequencing purposes, and the accumulation of this peptide was quite significant. DISCUSSION
FIG. 7. Immunoprecipitation of endogenous PP28 from [32P]PO4labeled cells and tryptic peptide analysis. Subconfluent N1E115 cells were metabolically labeled with [32P]PO4 for 3 h (lane 1 for untreated cells) and were treated with 8-Br-cAMP (1 mM, lane 2) or PMA (1 mM, lane 3) for 15 min. The extracted proteins were immunoprecipitated with anti-PP28 antiserum, separated by SDS–PAGE (10% gel), and transferred to an Immobilon-P membrane. (A) autoradiogram revealed the immunoprecipitated PP28. The labeled PP28 was excised from the membrane, trypsinized, and analyzed by isoelectric focusing gel (B) as described under Experimental Procedures. The lane marked ‘‘rPP28’’ represents the phosphopeptides from in vitro CK II phosphorylated rPP28 for comparison.
while Ser59, which was not a predicted site from the computer search, is the next favorable (see next paragraph). Somehow, the phosphorylation of the other four predicted sites, Thr17, Ser18, Thr96, and Ser107, is either too low for detection or not taking place at all. A closer examination of these residues, it is rather clear that Ser107 could never be phosphorylated by CK II having Arg–Arg immediately at its C-side instead of acidic residues. Sequential phosphorylation of Ser 62 and Ser 59 by CK II. According to Pinna’s convention (25), Ser62 is a good CK II consensus site due to the presence of acidic residues of N1-6 on its C-terminal side, especially the critical N / 3 residue. Although Ser59 also contains acidic residues, it lacks the critical N / 3 one and is not predicted to be a potential site for CK II. However, phosphorylated Ser62 could provide an acidic environment at N / 3 of Ser59 and set off its phosphorylation. The experiments shown in Fig. 9 supported such a notion. During the time course of phosphorylation, the accumulation of a single phosphate-containing peptide, including those having Lys72 and Arg73 at its C-terminal, focusing at pI 4.0 and 4.3, respectively, occurred significantly ahead of that of doubly phosphorylated
/ 6b11$$9300
01-29-96 13:40:16
arca
This paper demonstrates the cDNA cloning of a novel phosphoprotein HASPP28. Initially, we identified this protein from the acid- and heat-stable pool of rat brain extract as an in vitro PKC substrate; the cDNA-deduced amino acid sequence of PP28 contains consensus phosphorylation sites for PKA and CK II as well. Indeed, both the purified rat brain and recombinant PP28 can serve as substrates for these kinases in vitro, albeit with different stoichiometry of phosphorylation. Importantly, in mouse neuroblastoma N1E115 cells, metabolically [32P]PO4-labeled and immunoprecipitated PP28 revealed a tryptic peptide pattern similar to that of the in vitro CK II phosphorylated rPP28. Even when cells were treated with the tumor-promoting phorbol ester PMA, an activator of PKC, or with a cell-permeable 8Br-cAMP, an activator of PKA, and both reagents caused characteristic changes in cell morphology and kinase activities (H. Li and F. L. Huang, unpublished data), we consistently observed a CK II-reaction derived phosphopeptide pattern. It seemed that PKA and PKC, with or without stimulation, do not phosphorylate PP28l in the N1E115 cells. Although the biological function of PP28 is not presently known, phosphorylation probably plays a regulatory role, and most likely CK II or CK II-like kinase is responsible for its phosphorylation in vivo. CK II is ubiquitously present in all types of eukaryotic cells. It constitutively phosphorylates a variety of nuclear and cytosolic substrates involved in all aspects of transcription, translation, and metabolic regulation (see review of Ref. 32). It also phosphorylates many cell-cycledependent proteins such as P53 (33) and has been implicated as a critical component for regulating the cell cycle (34). This notion is further supported by the recent findings that the cell-cycle-dependent cdc2 kinase can phosphorylate and activate CK II in vitro (35). In intact cells, cdc2 kinase activity reaches a peak level during G2/M transition, and at this stage CK II becomes extensively phosphorylated by cdc2 kinase and is activated (36). Increased phosphorylation of topoisomerase by CK II during the same period of the cell cycle has been demonstrated (37). It has also been reported that stimulation of intact cells with mitogenic agents, including insulin, insulin-like growth factor, and epidermal growth factor
AP: Archives
CLONING OF A NOVEL CASEIN KINASE II SUBSTRATE, HASPP28
139
FIG. 8. Determination of casein kinase II phosphorylation sites of PP28. (A) Reverse-phase HPLC profile of tryptic phosphopeptides derived from CK II phosphorylated PP28 as separated on a C18 column. The dotted line indicates the elution gradient of acetonitrile. The continuous tracing is the absorbance at 214 nm. Radioactivities in the fractions were determined by Cerenkov counting (l). Four short bars under the radioactivity peak represent fractions 53–56, whose peptide sequences were determined as shown in B. The bold S in the sequences represents a phosphorylated serine residue based on the high ratio of PTH-dehydro-Ala/DTT adduct to PTH-Ser at the presumed cycle. (C) Autoradiogram of isoelectric focusing gel analysis of these fractions; rPP28 is the total phosphopeptides before HPLC fractionation.
(38), increases CK II activity. Correlation of PP28 phosphorylation by CK II under the influence of these hormones and at different stages of the cell cycle may help us to elucidate the function of this protein. It is interesting to note that the phosphorylation of Ser59 by CK II takes place later than that of Ser62 in
FIG. 9. Sequential phosphorylation of Ser62 and Ser59 of rPP28 by casein kinase II. At various time points during the course of phosphorylation, aliquots of sample were removed. Proteins in the samples were TCA precipitated as described in the legend of Fig. 6, trypsin digested, and analyzed by isoelectric focusing gel.
/ 6b11$$9300
01-29-96 13:40:16
arca
PP28 (Fig. 9). Presumably, the presence of a phosphate on Ser62 facilitated the phosphorylation of Ser59 by providing CK II with the critical N / 3 acidic residue, as has been described for many other CK II substrates such as topoisomerase and DARPP-32 (37, 39). This facilitatory effect, likely operating in intact cells (37, 39 and the present study), could be physiologically important, as phosphorylation greatly increases the total negative charge of the protein. Phosphorylation of Ser59 and Ser62 in PP28, for example, results in a stretch of 11 consecutive acidic residues that may generate a high Ca2/-binding site (see next paragraph about PEST). Another facilitatory effect has also been described, in that phosphorylation of a substrate by CK II often triggers subsequent phosphorylation of the substrate by a second kinase (32). For example, phosphorylation of inhibitor-2 of phosphoprotein phosphatase by CK II ensures its further phosphorylation by GSK-3, resulting in the expression of its full function (40). It is tempting to speculate that the phosphorylation of PP28 can also lead to phosphorylation by other kinases. In vitro, PP28 can be phosphory-
AP: Archives
140
SHEN ET AL.
lated by PKA and PKC without prior phosphorylation assuming D Å E, and phosphorylated S Å D or E, except by CK II; however, such reactions did not seem to take I of residue 46 of SPT6. It would certainly be worthplace in N1E115 cells under our present testing condi- while to probe the possible nuclear localization of PP28 tions. Since phosphopeptides resulting from CK II phos- as well as its cell-cycle-related activities. phorylation are the only products observed from the metabolic labeling, it is reasonable to conclude that CK REFERENCES II, or its like, is the most likely regulator of PP28. 1. Krebs, E. G. (1987) Annu. Rev. Biochem. 56, 567–613. There is evidences which points to the regulation of 2. Hu, Z.-W., Shi, X.-Y., Sakaue, M., and Hoffman, B. B. (1993) J. protein turnover by CK II phosphorylation (41). Many Biol. Chem. 268, 3610–3615. proteins with PEST sequences (41), regions rich in pro3. Ozawa, K., Szallasi, Z., Kazanietz, M. G., Blumberg, P. M., line (P), glutamic acid (E), serine (S), and threonine Mischak, H., Mushinski, J. F., and Beaven, M. A. (1993) J. Biol. (T), contain CK II phosphorylation sites and often have Chem. 268, 1749–1756. short half-lives (41). PP28 has a very high PEST score 4. Hordijk, P. L., Verlaan, I., Jalink, K., van Corven, E. J., and Moolenaar, W. H. (1994) J. Biol. Chem. 269, 3534–3538. of 15.82 (program PC/GENE) at the region of residues 5. Bubien, J. K., Jope, R. S., and Warnock, D. G. (1994) J. Biol. Ala33 –Pro50 which locates immediately N-terminal to Chem. 269, 17780–17783. its CK II phosphorylation sites of Ser59 and Ser62. Ac6. Nemenoff, R. A., Winitz, S., Qian, N.-X., Van Putten, V., Johncording to the PEST model, CK II phosphorylation genson, G. L., and Heasley, L. E. (1993) J. Biol. Chem. 268, 1960– 2/ erates a high affinity binding site for Ca , which sub1964. sequently activates the protease calpain resulting in 7. Rascon, A., Degerman, E., Taira, M., Meacci, E., Smith, C., protein degradation (34, 41). In this regard, it is worth Manganiello, V., Belfrage, P., and Tornqvist, H. (1994) J. Biol. mentioning that we found a stretch of 10 identical Chem. 269, 11962–11966. amino acids between PP28 and calpastatin, the endoge8. Blackshear, P. J. (1993) J. Biol. Chem. 268, 1501–1504. nous calpain inhibitor precursor. Although this se9. Chapman, E. R., Estep, R., and Storm, D. R. (1992) J. Biol. Chem. 267, 25233–25238. quence segment within calpastatin (Gly125 –Leu134) (29) is not located in the calpain-binding domain, it may 10. Baudier, J., Deloulme, J. C., Dorsselaer, A. V., Black, D., and Mattes, H. W. D. (1991) J. Biol. Chem. 266, 229–237. influence its interaction with the protease. This stretch 48 57 11. Huang, K.-P., Huang, F. L., and Chen, H.-C. (1993) Arch. Bioof residues in PP28 (Gly –Leu ) is located between its chem. Biophys. 305, 570–580. CK II phosphorylation sites Ser59 and Ser62 and PEST region (Ala33 –Pro50). While it is interesting to test if 12. Huang, K.-P., Nakabayashi, H., and Huang, F. L. (1986) Proc. Natl. Acad. Sci. USA 83, 8535–8539. phosphorylation of PP28 would control its turnover 13. Huang, K.-P., and Huang, F. L. (1986) Biochem. Biophys. Res. rate, it would also be important to test for a regulatory Commun. 139, 320–326. role of these identical residues. 14. Huang, K.-P., Akatsuka, A., Singh, T. J., and Blake, K. (1983) PP28 also shares some homology with neuron-speJ. Biol. Chem. 258, 7094–7101. cific filaments of 70/60 kDa (28), troponin T (27), and 15. Huang, K.-P., Itarte, E., Singh, T., and Akatsuka, A. (1982) J. Biol. Chem. 257, 3236–3242. EMB-5 (30) (SPT6 in yeast). Since the Northern blot analysis did not point to its exclusive neuronal location, 16. Huang, F. L., Huang, K.-P., Sheu, F.-S., and Osada, K.-I. (1993) in Methods in Neuroscience (Fain, J. N., Ed.), Vol. 18, Lipid PP28 probably does not assume a neuron-specific funcMetabolism in Signaling System, pp. 127–137, Academic Press, tion like neurofilament. Troponin T is considered to San Diego. 2/ be required for the Ca -dependent ATPase activity in 17. Sanger, F. S., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. muscle contraction through interaction with tropomyoAcad. Sci. USA 74, 5463–5467. sin. The molecule of troponin T can be subdivided into 18. Borson, N. D., Salo, W. L., and Drewes, L. R. (1992) PCR Methtwo domains, T1 (residues 1–158) and T2 (residues ods Appl. 2, 144–148. 159–259). It is T1 that interacts with tropomyosin (31), 19. Lehrach, H., Diamond, D., Wozney, J. M., and Boedfker, H. and it is T1 that exhibits homology with PP28. It would (1977) Biochemistry 16, 4743–4751. be interesting to test if PP28 interacts with tropomyo- 20. Kao, P. N., Chen, L., Brock, G., Ng, J., Kenny, J., Smith, A. J., and Corthesy, B. (1994) J. Biol. Chem. 269, 20691–20699. sin. EMB-5 protein of C. elegans is similar to the yeast nuclear protein SPT6, in that both affect transcription 21. Konigsberg, W. H., and Henderson, L. (1983) Methods Enzymol. 91, 254–259. in the control of the cell cycle. Importantly, both EMB22. Ishizaki, J., Hanasaki, K., Higashino, K., Kishino, J., Kikuchi, 5 and SPT6 have a stretch of acidic amino acids at the N., Ohara, O., and Arita, H. (1994) J. Biol. Chem. 269, 5897– N-terminal portion which contain CK II phosphorylata5904. ble serine residues. In fact, residues 58–68 of PP28, 23. Woodget, J. R., Gould, K. L., and Hunter, T. (1986) Eur. J. residues 44–54 of EMB-5, and residues 39–49 of SPT6 Biochem. 161, 177–184. are in good alignment as shown below: 24. Kishimoto, A., Nisshiyama, K., Nakanishi, H., Uratsuji, Y., No58 DSDESEDEDDD68 of PP28 mura, H., Takeyama, Y., and Nishizuka, Y. (1985) J. Biol. 44 Chem. 260, 12492–12499. SSDEDEDDDDD54 of EMB-5 39 SSEEDEDIDED49 of SPT6, 25. Pinna, L. A. (1990) Biochim. Biophys. Acta 1054, 267–284.
/ 6b11$$9300
01-29-96 13:40:16
arca
AP: Archives
CLONING OF A NOVEL CASEIN KINASE II SUBSTRATE, HASPP28 26. Glass, D. B., El-Maghrabi, M. R., and Pilkis, S. J. (1986) J. Biol. Chem. 261, 2987–2993. 27. Pearlstone, J. R., Carpenter, M. R., and Smillie, L. B. (1986) J. Biol. Chem. 261, 16795–16810. 28. Szaro, B. G., Pant, H. C., Way, J., and Battey, J. (1991) J. Biol. Chem. 266, 15035–15041. 29. Emori, Y., Kawasaki, H., Imajoh, S., Imahori, K., and Suzuki, K. (1987) Proc. Natl. Acad. Sci. USA 84, 3590–3594. 30. Nishiwaki, K., Sano, T., and Miwa, J. (1993) Mol. Gen. Genet. 239, 313–322. 31. Heeley, D. H., Golosinska, K., and Smillie, L. B. (1987) J. Biol. Chem. 262, 9971–9978. 32. Tuazon, P. T., and Traugh, J. A. (1991) Adv. Second Messenger Phosphoprotein Res. 23, 123–164. 33. Deppert, W., Buschhausen-Denker, G., Patschinsky, T., and Steinmeyer, K. (1990) Oncogene 5, 1701–1706.
/ 6b11$$9300
01-29-96 13:40:16
arca
141
34. Meisner, H., and Czech, M. P. (1991) Curr. Opin. Cell Biol. 3, 474–483. 35. Mulner-Lorillon, O., Cormier, P., Labbe, J., Poulhe, R., Osborne, H., and Belle, R. (1990) Eur. J. Biochem. 93, 529–534. 36. Litchfield, D. W., Luscher, B., Lozeman, F. J., Eisenman, R. N., and Krebs, E. G. (1992) J. Biol. Chem. 267, 13943–13951. 37. Cardenas, M. E., and Gasser, S. M. (1993) J. Cell Sci. 104, 219– 225. 38. Klarlund, J. K., and Czech, M. P. (1988) J. Biol. Chem. 263, 15872–15875. 39. Girault, J., Hemmings, H., Jr., Williams, K. R., Nairn, A. C., and Greengard, P. (1989) J. Biol. Chem. 264, 21748–21759. 40. DePaoli, A. A. (1984) J. Biol. Chem. 259, 12144–12152. 41. Rogers, S., Wells, R. M., and Rechsteiner, M. (1986) Science 234, 364–368.
AP: Archives