A protein kinase Cdelta-binding protein SRBC whose expression is induced by serum starvation.

West-Western screening of a cDNA expression library using 32P-labeled, autophosphorylated protein kinase Cδ (PKCδ) as a probe, led us to identify cDNA clones encoding a PKCδ-binding protein that contains a leucine zipper-like motif in its N-terminal region and two PEST sequences in its C-terminal region. This protein shows overall sequence similarity (43.3%) to the serum deprivation response (sdr) gene product, and we named it SRBC (dr-elated gene product that inds to -kinase). PKCδ binds to the C-terminal half of SRBC through the regulatory domain and phosphorylates it in vitro. In COS1 cells, the phosphorylation of over-expressed SRBC is stimulated by 12-O-tetradecanoylphorbol-13-acetate and further enhanced by the over-expression of PKCδ. The mRNA for SRBC is detected in a wide variety of cultured cell lines and tissues and is strongly induced by serum starvation. Furthermore, SRBC mRNA is induced during retinoic acid-induced differentiation of P19 cells. These results suggest that SRBC serves as a substrate and/or receptor for PKC and might be involved in the control of cell growth mediated by PKC.

West-Western screening of a cDNA expression library using 32 P-labeled, autophosphorylated protein kinase C␦ (PKC␦) as a probe, led us to identify cDNA clones encoding a PKC␦-binding protein that contains a leucine zipper-like motif in its N-terminal region and two PEST sequences in its C-terminal region. This protein shows overall sequence similarity (43.3%) to the serum deprivation response (sdr) gene product, and we named it SRBC (sdr-related gene product that binds to c-kinase). PKC␦ binds to the C-terminal half of SRBC through the regulatory domain and phosphorylates it in vitro. In COS1 cells, the phosphorylation of over-expressed SRBC is stimulated by 12-O-tetradecanoylphorbol-13-acetate and further enhanced by the over-expression of PKC␦. The mRNA for SRBC is detected in a wide variety of cultured cell lines and tissues and is strongly induced by serum starvation. Furthermore, SRBC mRNA is induced during retinoic acid-induced differentiation of P19 cells. These results suggest that SRBC serves as a substrate and/or receptor for PKC and might be involved in the control of cell growth mediated by PKC.
Protein kinase C (PKC) 1 is a serine/threonine kinase thought to act in diverse cellular processes such as the secretion of hormones and neurotransmitters and the regulation of cell proliferation and differentiation. So far, more than 10 PKC isozymes have been reported, and these can be divided into three distinct classes based on differences in their structures and biochemical properties, conventional PKC (cPKC) members (␣, ␤I, ␤II, and ␥), novel PKC (nPKC) members (␦, ⑀, , , and ), and atypical PKC (aPKC) members ( and /). All PKC members consist of an N-terminal regulatory domain and a C-terminal catalytic domain; the co-factor binding site has been identified in the regulatory domain (1)(2)(3)(4).
PKC␦ is an nPKC member that is expressed in a variety of tissues and cultured cell lines (5)(6)(7)(8). We have previously shown that a constitutively active mutant of PKC␦ acts as a potent inducer of transcription factor activator protein 1/Jun and that a kinase-deficient mutant of PKC␦ can inhibit the activity of the mutant (9). This dominant-negative effect of kinase-deficient PKC␦ can be explained by the titration of an effector, substrate, or receptor molecule(s) that binds stably to PKC␦.
The stable interaction of PKC and its substrate has been demonstrated by using MARCKS, a well characterized PKC substrate (10). Furthermore, our kinetic analysis of the phosphorylation of MARCKS by PKC isozymes revealed a very low K m value (10 -20 nM), supporting the high affinity interaction between MARCKS and PKC isozymes (␣, ␦, and ⑀) (11). Some other PKC-binding proteins have been identified by the screening of a cDNA expression library using a purified brain PKC mixture as a probe. One such protein is RACK1, which is thought to serve as a receptor for activated PKC (12). Others include F52/Mac-MARCKS and 35H/␥-adducin, which is a known substrate for PKC in vivo and in vitro (13,14). A yeast two-hybrid system has also been used to identify PKC␣-binding protein PICK1 (15). In addition to these proteins, several proteins are known to bind to a brain PKC mixture in vitro. These include annexin (I, II, and VI), vinculin, talin, and 72/53ORIG (16,17). In addition, Bruton tyrosine kinase binds to cPKC (␣, ␤I, and ␤II), nPKC (⑀), and aPKC () in vitro (18). Actin binds to PKC⑀ in vitro (19) and PKC␤II in vitro and in vivo (20). Recently, it has been shown that AKAP79, an A kinase anchoring protein, can also bind to PKC␣ and PKC␤II in vitro and in vivo (21), and human immunodeficiency virus Nef protein can bind to PKC in vitro and in vivo (22).
In the present study, we used West-Western screening to isolate cDNA clones encoding PKC␦-binding proteins. Here we describe a clone encoding as sdr-related gene product named SRBC, which serves as a substrate for PKC␦ and whose expression is induced by serum starvation of NIH3T3 cells or differentiation of mouse embryonal carcinoma P19 cells.

EXPERIMENTAL PROCEDURES
Purification and 32 P Labeling of PKC␦-PKC␦ was purified from recombinant baculovirus-infected Sf21 cells as described elsewhere (11). Briefly, 3 days after infection, Sf21 cells were lysed, and PKC␦ was purified by a series of chromatographic steps on DEAE-cellulose (Tosoh), hydroxyapatite (Koken), and MonoQ (Pharmacia Biotech Inc.) columns. No additional proteins were obvious in the final PKC␦ fraction as judged by SDS-PAGE, and the specific activity was 380 units/mg. One unit of PKC␦ activity was defined as 1 nmol of 32 P incorporated into * This work was supported in part by research grants from the Ministry of Education, Science, Sports and Culture of Japan, The Cell Science Research Foundation, and the Uehara Memorial Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The Amino acid residues included in the leucine zipper-like motif are marked by a black background; the PEST regions are underlined. The asterisks indicate potential PKC phosphorylation sites. The lower panels show schematic drawings of the protein structures of SRBC and the clone 53 product. Clone 53 encodes the C-terminal part of mouse SRBC, which was obtained from the NIH3T3-EXlox expression library using PKC␦ as a probe. B, amino acid alignment of rat SRBC, mouse Sdr, and the CHKESTFLLE protein encoded by a chicken mRNA for the expressed sequence tag. Identical residues are shaded and SRBC conserved region 1 (SCR1) and SRBC conserved region 2 (SCR2) are indicated by double-underlining. Asterisks indicate amino acid residues included in the leucine zipper-motif. NaCl, 0.1 mM dithiothreitol, and 0.5% Triton X-100 and used as a probe for screening a EXlox cDNA expression library.
Isolation of a cDNA Clone Encoding a PKC␦-binding Protein-Poly(A) ϩ RNA was isolated from NIH3T3 cells, and randomly primed cDNA was synthesized using HindIII primer adapter (Novagen). cDNA was ligated to EXlox vector arm (Novagen) and packaged in phage particles using a Phage Marker System (Novagen). This library contains approximately 0.95 ϫ 10 6 independent clones. Approximately 2 ϫ 10 5 clones were plated at 10,000 phages per plate using Escherichia coli strain BL21 (DE3) pLysE as host cells. After incubation for 6 h at 37°C, the plaques were covered with Hybond-C extra (Amersham Corp.) impregnated with 10 mM isopropyl-1-thio-␤-Dgalactopyranoside and allowed to grow for an additional 3.5 h at 37°C. The filters were then washed with TBS buffer (50 mM Tris-HCl, pH 7.5, 0.5 M NaCl), blocked with 5% skim milk at 4°C overnight, and incubated with 1 mM ATP solution (TBS buffer containing 1 mM ATP) at room temperature for 2 h to saturate any auto-phosphorylation sites or ATP-binding sites of the proteins. The filters were then incubated with 32 P-labeled PKC␦ (10 6 cpm/ml) in 50 mM Tris-HCl, pH 7.5, 0.5 M NaCl, 50 g/ml phosphatidylserine, and 1% bovine serum albumin for 5 h at room temperature. After washing in TBS buffer, the filters were exposed for autoradiography. Positive phage clones were converted to plasmids (pEXlox ϩ cDNA insert) using E. coli strain BM25.5 according to the manufacturer's instructions. A cDNA insert was used as a probe to screen a 3Y1-ZAP II cDNA library to obtain cDNA clones encoding the full SRBC coding sequence. Plaque hybridization with a ZAP II cDNA library and the isolation of positive phage clones were performed according to standard procedures.
Phosphorylation of Proteins Fixed on PVDF Membranes-PKC␦binding proteins were subjected to SDS-PAGE, blotted onto PVDF membranes, and treated with a 5% skim milk solution. The phospho-rylation reaction was carried out by incubating the PVDF membranes with phosphorylation buffer (83 ng/ml PKC␦, 20 mM Tris-HCl, pH 7.5, 5 mM MgCl 2 , 10% glycerol, 50 g/ml phosphatidylserine, 400 ng/ml TPA, and 2.4 Ci/ml [␥-32 P]ATP) at room temperature for 2 h. After washing with TBS containing 1% SDS buffer, the membranes were exposed for autoradiography.
Overlay Assay Using 32 P-Labeled PKC␦ as a Probe-Overlay assay was performed based on the procedure of Wolf and Sahyoun (23). PKC␦-binding proteins were subjected to SDS-PAGE and blotted onto a PVDF membrane. After treatment with a 5% skim milk solution, the PVDF membrane was incubated with 32 P-labeled PKC␦ (10 6 cpm/ml) diluted in 50 mM Tris-HCl, pH 7.5, 0.5 M NaCl, 50 g/ml phosphatidylserine, and 1% bovine serum albumin at room temperature for 5 h. Excess ligand was removed by washing the PVDF membrane with TBS buffer, and the membrane was exposed for autoradiography.
Phosphatidylserine Overlay Assay-Protein-blotted PVDF membrane was overlaid with 20 g/ml [ 14 C]phosphatidylserine (specific activity ϭ 1 Ci/15 g, Amersham Corp.) diluted in 50 mM Tris-HCl, pH 7.5, containing 0.5 M NaCl and 10 mg/ml bovine serum albumin at room temperature for 1 h. Membrane was washed briefly in phosphatebuffered saline and exposed for autoradiography.
Overlay Assay Using MBP Fusion Proteins as Probes-cDNAs encoding mouse PKC␦ and its deletion mutant (lacking amino acid residues 299 -654, 209 -654, and 111-654) were subcloned into E. coli expression vector pMAL-c2 (New England BioLabs) for expression of these proteins as fusion proteins with maltose-binding protein (MBP-PKC␦, MCD299, MCD209, and MCD111). The proteins were purified on amylose resin (New England BioLabs) and used as probes for the overlay assay.
The protein-transferred PVDF membrane was incubated with 0.7 M MBP-PKC␦ or 1 M MCD299, MCD209, or MCD111 in 50 mM Tris-HCl, FIG. 1-continued pH 7.5, 0.5 M NaCl, 50 g/ml phosphatidylserine, and 1% bovine serum albumin for 3 h at room temperature. After washing the membrane with TBS buffer, bound protein was fixed by incubation with 0.5% formaldehyde. Excess formaldehyde was removed by washing with 2% glycine, and MBP fusion proteins were detected using anti-MBP antibody (New England BioLabs), alkaline phosphatase-conjugated second antibody (TAGO, Inc.), and artificial substrate for alkaline phosphatase (VECTOR).
In Vitro Phosphorylation of GST-SRBC Protein by PKC␦-An E. coli expression vector for GST-SRBC was constructed using expression vector pGEX-3X (Pharmacia) and cDNA encoding SRBC, and the fusion protein was purified on glutathione-Sepharose 4B (Pharmacia). The reaction mixture contained 20 mM Tris-HCl, pH 7.5, 5 mM Mg(OAc) 2 , 10 g/ml leupeptin, 50 ng/ml TPA, 25 g/ml phosphatidylserine, 3.74 ng of PKC␦, and various concentrations of GST-SRBC proteins in a total volume of 20 l. The reaction was started by the addition of 20 M ATP and 0.5 Ci of [␥-32 P]ATP and incubated at 30°C. After 5 min of incubation, the reaction was stopped by the addition of 5 l of 5 ϫ Laemmli's SDS-sample buffer. Proteins were then separated on SDS-PAGE, and the 32 P incorporated in the GST-SRBC protein was quantified using a Bio-Image analyzer (BAS 2000 FUJI).
Phosphorylation of SRBC in COS1 Cells-An expression vector for tag-SRBC⌬15 (SRHis-SRBC) encodes a SRBC protein whose N-terminal 15 amino acids are replaced by six histidine residues and a 12amino acid sequence from the T7 gene 10 leader sequence derived from pBLUE Bac (Invitrogen). The PKC␦ expression plasmid (M241) was described previously (6). COS1 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS) and seeded 24 h before transfection at a concentration of 3 ϫ 10 5 cells/6-cm dish. Transfection by calcium phosphate co-precipitation was carried out for 6 h using the appropriate expression vectors, as described in the figure legends. The culture medium was then changed to serum-deprived DMEM to starve the cells. After 48 h, the medium was changed to phosphate-free DMEM, and the cells were further incubated for 2 h. The cells were then incubated in medium containing [ 32 P]orthophosphate (125 Ci/ml) for 4 h prior to stimulation. After treatment for 10 min with TPA (50 ng/ml) or vehicle (dimethyl sulfoxide) alone, the cells (in 6-cm dishes) were harvested and suspended in 100 l of lysis buffer containing 20 mM Hepes, pH 7.5, 150 mM NaCl, 1 mM EDTA, 50 mM NaF, 1 mM Na 3 VO 4 , 10 g/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, 1.8 g/ml aprotinin, 1% Triton X-100, 0.1% deoxycholate, and 0.1% SDS. After 30 min incubation on ice, the lysate was clarified by centrifugation at 14,000 rpm for 30 min and incubated with anti-T7-tag antibodies (Novagen) preabsorbed with Pro-tein G-Sepharose (Pharmacia), for 1 h at 4°C. The immunocomplexes on Sepharose were washed 5 times with lysis buffer and 2 times with final wash buffer containing 20 mM Hepes, pH 7.5, 1 mM EDTA, 50 mM NaF, 1 mM Na 3 VO 4 , 10 g/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, 1.8 g/ml aprotinin, and 0.1% Triton X-100. The proteins were then separated by SDS-PAGE, transferred to a PVDF membrane, and probed with anti-T7 antibody to estimate the amount of tag-SRBC precipitated. The amount of 32 P incorporated into tag-SRBC was measured by a Bio-Image analyzer (BAS 2000 FUJI).
Cell Culture Conditions for the Analysis of SRBC mRNA-NIH3T3 cells were routinely cultured in DMEM supplemented with 10% fetal calf serum (FCS). For serum starvation, the medium was changed to 0.5% FCS when the cells were still subconfluent. For serum stimulation, fresh medium containing 20% FCS was added to the starved cells. Cells were harvested at the desired times for RNA isolation. Differentiation of P19 cells was performed following a standard procedure (24,25).
Poly(A) ϩ RNA was isolated using a QuickPrep Micro mRNA Purification Kit (Pharmacia), and Northern blot analysis was performed according to the standard protocol.

RESULTS
Cloning of SRBC-A mouse NIH3T3-EXlox expression library was screened for proteins that bind to PKC␦ in the presence of phosphatidylserine. Using 32 P-labeled PKC␦ as a probe, two clones, clone 53 and 91, that appeared to originate from the same mRNA were isolated. To isolate full-length cDNAs, a rat 3Y1-ZAPII cDNA library was screened using clone 53 as a probe. The nucleotide sequence of a rat homologue of clone 53/91, and the structural features of the protein are shown in Fig. 1A. The ATG at position 1 is most likely an initiation codon because 1) the surrounding nucleotide sequence fulfills Kozak's criteria (26), and 2) there is an in-frame TAG codon in the 5Ј non-coding sequence. The open reading frame starting with this initiation codon encodes a protein with 263 amino acid residues. The calculated molecular mass of the protein, named SRBC, is 27,879. Mouse clone 53 encodes 105 amino acid residues corresponding to the C-terminal part of rat SRBC (Fig. 1A, lower panel), and the amino acid sequence identity in this region is 86%. SRBC contains a leucine zipperlike motif in the N-terminal part and two PEST regions, which are commonly found in short-lived proteins (27,28), in the C-terminal part (Fig. 1A). A comparison of the SRBC nucleotide sequence with the GenBank data base revealed high homologies to mouse sdr (29) and a chicken mRNA for the expressed sequence tag (CHKESTFLLE) (Fig. 1B).The overall amino acid sequence identities for SRBC to Sdr and CHKESTFLLE are 43.3 and 25.9%, respectively. Amino acid residues conserved among the proteins are clustered in two regions, SRBC conserved region 1 (SCR1) and SRBC conserved region 2 (SCR2). SCR1 contains the leucine zipper-like motif in its N-terminal part, and SCR2 contains one or two PKC phosphorylation sites (Fig. 1B). Amino acid identity in SCR1 is 43.8% between SRBC and Sdr and 29.0% between SRBC and CHKESTFLLE. In SCR2, it is 60.9% between SRBC and Sdr and 52.2% between SRBC and CHKESTFLLE. In addition, all of these proteins contain PEST regions even though their primary structures are not conserved. It is noteworthy that the sequence from 1 to 185 amino acids of SRBC shows 90% identity with the sequence from 1 to 185 amino acids of the human 7N-1 clone, identified as a gene fused to c-RAF-1 that results in the constitutive activation of the c-RAF-1 gene (30). In this clone, the N-terminal regulatory domain of c-RAF-1 is replaced by the human SRBC homologue.
PKC␦ Binds to SRBC-The original pEXlox clone, clone 53, encodes only a C-terminal part of SRBC (156 to 263 amino acids) as fusion proteins with the T7 gene 10 product, indicating that the C-terminal part is sufficient for the interaction with PKC␦. The overlay assay showed that PKC␦ binds to the product of the T7 gene 10 fused clone 53 product in the pres- ence of phosphatidylserine, a common activator of PKCs (Fig. 2,  lanes 3 and 5). On the other hand, no such interaction was observed with the T7 gene 10 product itself (Fig. 2, lanes 3 and  4). Furthermore, the fusion protein but not the T7 gene 10 product itself can be phosphorylated by PKC␦ on the membrane (Fig. 2, lanes 7 and 8), in agreement with the presence of putative PKC-phosphorylation sites in the clone 53 product (Fig. 1A).
To test whether full-length SRBC binds to PKC␦, SRBC was expressed as a GST-fusion protein in E. coli and tested for PKC␦ binding by overlay assay. MARCKS, a known binding protein/substrate for PKC, and GST alone were also tested for PKC␦ binding. GST-SRBC appeared as a 65.4-kDa band on SDS-PAGE, and MARCKS and GST appeared as 80-and 26-kDa bands, respectively (Fig. 3A, lanes 1-3). The positions of GST and GST-SRBC were confirmed by Western blot analysis using anti-GST antibody (Fig. 3A, lanes 4 and 5). PKC␦ bound to GST-SRBC to the same extent as to MARCKS (Fig. 3A, lanes  6 and 7) but did not bind to GST (Fig. 3A, lane 8). Again, the binding depends entirely on the presence of phosphatidylserine (Fig. 3A, lanes 9 and 10). When PKC␣ instead of PKC␦ was used as a probe, essentially the same results were obtained, whereas the signal was barely detectable when PKC was used as the probe (data not shown).

Phosphatidylserine Binds to SRBC and Clone 53 Product-
Previous studies showed that most of the PKC-binding proteins are phosphatidylserine-binding proteins (10,13,16,23,31). Thus, phosphatidylserine may form a bridge between PKC and PKC-binding proteins and stabilize the binding. Since binding of SRBC to PKC␦ also depends on phosphatidylserine, we next tested whether SRBC binds to phosphatidylserine by using [ 14 C]phosphatidylserine overlay assay. As shown in Fig. 3B, GST-SRBC binds to phosphatidylserine, and clone 53 encoding only C-terminal part of SRBC also binds to phosphatidylserine. Note that the C-terminal part of SRBC was sufficient for the binding to PKC␦. These results suggest that the binding of SRBC and PKC␦ depends on phosphatidylserine-bridging as suggested for other PKC-binding proteins.
The Regulatory Domain of PKC␦ Binds to SRBC-Since the binding of PKC␦ to SRBC depends on the presence of phosphatidylserine, the co-factor binding site of PKC␦ located in its N-terminal regulatory domain might be involved in the interaction with SRBC. It has been reported that the pseudosubstrate region, in the regulatory domain, directly mediated phosphatidylserine-dependent PKC binding to some PKC-binding proteins (16,31). To test whether the N-terminal regulatory domain is sufficient for this interaction, whole PKC␦ and its regulatory domain were expressed as MBP (maltose-binding FIG. 3. SRBC binds to PKC␦ in a phosphatidylserine-dependent manner. A, GST and GST-SRBC were produced in E. coli, and MARCKS was produced in Sf21 cells using baculovirus vector (11). These proteins were subjected to SDS-PAGE and/or transferred to a PVDF membrane for the overlay assay. The proteins were detected on the gel by silver staining (lanes 1-3). GST-SRBC and GST were also detected by Western blot analysis using anti-GST antibody (lanes 4 and 5). Overlay assay was performed in the presence (lanes 6 -8) or absence (lanes 9 -11) of phosphatidylserine using 32 P-labeled PKC␦ as a probe. Arrowheads indicate the position of each protein. B, protein-transferred membrane was overlaid with [ 14 C]phosphatidylserine. Phosphatidylserine-binding proteins were detected by autoradiography. Samples used are purified GST-SRBC (same as A) and crude extract of E. coli containing clone 53 products (same as Fig. 2). protein) fusion proteins in E. coli (Fig. 4A) and used as probes for the overlay assay. In these experiments, bound PKC␦ was detected by anti-MBP antibodies. As shown in Fig. 4B, MBP-PKC␦ bound to GST-SRBC or MARCKS similarly to intact PKC␦ (Fig. 4B, lanes 1 and 2). Furthermore, MBP fused to the regulatory domain of PKC␦ (MCD299) also bound to GST-SRBC and MARCKS to the same extent (Fig. 4B, lanes 4 and  5). The regulatory domain of PKC␦ includes a cysteine-rich domain conserved among all cPKC and nPKC members. When the cysteine-rich domain was totally deleted (MCD111), no binding to SRBC or MARCKS was observed (Fig. 4B, lanes 10  and 11), whereas the C-terminal half of the cysteine-rich domain was dispensable for binding (MCD209 in Fig. 4B, lanes 7  and 8). These results suggest that amino acid residues 111-209 on PKC␦, a region that includes a pseudosubstrate and half of the cysteine-rich domain, are essential for the binding to GST-SRBC and MARCKS. The binding of SRBC to the regulatory domain of PKC␦ raises the possibility that GST-SRBC modulates the activity of PKC␦. However, we could not detect any effect of GST-SRBC on the myelin basic protein kinase activity of PKC␦ in vitro (data not shown).
SRBC Is a Substrate for PKC␦ in Vitro-As already shown in Fig. 2, the C-terminal part of mouse SRBC, the clone 53 product, can be phosphorylated by PKC␦ in vitro. Full-length rat SRBC fused to GST is also phosphorylated by PKC␦, whereas GST alone is barely phosphorylated (Fig. 5A). We next exam-ined the kinetics for SRBC phosphorylation by PKC␦ in vitro. GST-SRBC (5-150 nM) was incubated with PKC␦ (0.7 nM), in the presence of phosphatidylserine, TPA, and [␥-32 P] ATP, and the phosphorylated proteins were analyzed by SDS-PAGE followed by quantitative autoradiography (Fig. 5B). The K m and V max values estimated from the saturation curve are 60 nM and 0.69 nmol of ATP/min/nmol, respectively. The K m value is roughly comparable with that for MARCKS (20.7 nM) obtained by similar experiments (11), whereas the V max is considerably lower than MARCKS (5.2 nmol of ATP/min/nmol). GST-SRBC is also phosphorylated by PKC␣, with K m and V max values similar to those for PKC␦, but GST-SRBC is a poor substrate for PKC with a V max value one-eighth that for PKC␦ (data not shown). The maximum incorporation of phosphate into SRBC by PKC␦ was 0.78 mol per 1 mol of GST-SRBC.
SRBC Is Phosphorylated Upon PKC Activation in Vivo-To monitor the phosphorylation of SRBC in vivo, we designed an epitope-tagged SRBC⌬15 expression vector. In this construction, the N-terminal 15 amino acids of SRBC are replaced by a T7 gene 10 epitope tag. The apparent molecular mass of this protein is 43 kDa on SDS-PAGE. The SRBC⌬15 was spontaneously phosphorylated in serum-starved COS1 cells, and TPA stimulation caused a ϳ2-fold increase in the level of phosphorylation (Fig. 6). We next examined the effect of the co-expression of PKC␦ in the presence or absence of TPA. No significant effect of PKC␦ over-expression was observed in the absence of TPA. But a more enhanced induction of phosphorylation by TPA (2.5-3-fold) was observed when PKC␦ was over-expressed. These results are consistent with the idea that SRBC is phosphorylated in vivo as a consequence of PKC activation.
The Expression of SRBC mRNA Is Induced by Cell Growth Arrest-Northern blot analysis showed that SRBC is ubiquitously expressed in almost all tissues tested except liver, and higher levels of expression are observed in uterus and ovary (Fig. 7A). In cultured cell lines, SRBC mRNA was detected in NIH3T3 cells and 3Y1 cells but not in COS1 cells or mouse embryonal carcinoma cell lines including F9 and P19 (Fig. 7B).
It has been reported that the expression of sdr mRNA is strongly induced by serum starvation in NIH3T3 cells and down-regulated within 6 h after the addition of serum to starved cells (29). The considerable structural similarity between SRBC and Sdr allowed us to test whether SRBC mRNA is also induced upon serum starvation. The level of SRBC mRNA was relatively low in growing NIH3T3 cells (Fig. 7C,  time 0), and was induced by serum deprivation reaching a maximum level within 12 h. The induced mRNA expression level was retained at least for 48 h (Fig. 7C). The amount of SRBC mRNA in starved cells (Fig. 7D, time 0) decreased rapidly after the addition of serum, reaching a minimum level after 3 h (Fig. 7D). When the amount of mRNA was normalized to that of glyceraldehyde-3-phosphate dehydrogenase mRNA, the SRBC mRNA level was ϳeight times higher in serumstarved cells compared with serum-stimulated or exponentially growing cells. The induction of SRBC mRNA was also observed upon retinoic acid-induced differentiation of mouse embryonal carcinoma P19 cells to neuron-like cells (Fig. 7B). Taken together, these results show that the level of SRBC mRNA expression correlates with cell growth suppression. DISCUSSION We cloned a cDNA encoding a PKC-binding protein by West-Western screening using 32 P-labeled, autophosphorylated PKC␦ as a probe. This protein, called SRBC, binds to and is phosphorylated by PKC␦ in vitro. In COS1 cells, the phosphorylation of over-expressed SRBC is stimulated by TPA and is further enhanced by the over-expression of PKC␦.
The K m value for the phosphorylation of SRBC by PKC␦ is quite low (60 nM), which may reflect a strong interaction between the two molecules. On the other hand, the V max value is quite low indicating that SRBC is not a good substrate for PKC␦. However, 1 mol of SRBC incorporates nearly 1 mol of phosphate after prolonged phosphorylation reaction (data not shown). Therefore, the low V max value could reflect the low exchange rate of the substrate, SRBC, as a result of stable binding to PKC␦. These features of the PKC-SRBC interaction are very similar to those for the interactions between PKC isozymes (␣, ␦, and ⑀) and MARCKS (11).
The binding of SRBC to PKC␦ depends on the presence of phosphatidylserine; therefore, the active conformation of PKC␦ might be necessary for the interaction. In addition, the fact that SRBC binds to phosphatidylserine suggests that phosphatidylserine mediates or stabilizes the interaction between SRBC and PKC␦ through its bridging. Interaction with SRBC was also observed for MCD209, a deletion mutant of PKC␦ lacking the kinase domain and the C-terminal half of the cysteine-rich domain that is responsible for the binding to phorbol ester or diacylglycerol (32,33). This indicates that these functional domains are dispensable for the interaction with SRBC. Since MCD111 cannot bind to SRBC, the N-terminal half of the cysteine-rich domain and the pseudosubstrate region may be essential for binding. Previous studies demonstrated that the  6. Phosphorylation of SRBC⌬15 in response to PKC activation in vivo. COS1 cells (3 ϫ 10 5 cells/6-cm dish) were transiently transfected with the expression vector encoding epitope-tagged SRBC⌬15 (5 g) alone or together with the PKC␦-encoded expression vector (2 g). Cells were labeled with 32 P and exposed to TPA (50 ng/ml) for 10 min before harvest. SRBC⌬15 was immunoprecipitated from cell extracts using anti-T7 tag antibody as described under "Experimental Procedures." Values represent arbitrary units normalized to the amount of immunoprecipitated SRBC⌬15 protein. DMSO, dimethyl sulfoxide.
pseudosubstrate region is one of several motifs that mediate PKC binding (16,31). However, from our results, we cannot conclude whether the pseudosubstrate region is important to the binding to PKC␦ and SRBC or not. We need further experiments to determine the SRBC binding region on the PKC␦. SRBC binds to PKC␣, a cPKC family member, to the same extent as PKC␦. And it barely binds to PKC, an aPKC family member (data not shown). Thus, the binding of PKC to SRBC does not clearly show isotype specificity. We identified several PKC␦-binding proteins in addition to SRBC by subsequent screening of the cDNA expression library using PKC␦ as a probe. These include MARCKS and several other proteins that do not include any of previously reported PKC-binding proteins. 2 Among these novel PKC-binding proteins, we could find a PKC␦-specific binding protein. 3 The primary structure of SRBC shows considerable similarity to Sdr and one of the chicken expression sequence tags (CHKESTFLLE). sdr has been identified as a gene induced in NIH3T3 cells by serum starvation but whose function remains unknown. Most of the conserved amino acids in these proteins are clustered in the N-terminal region named SCR1, which includes a "leucine zipper"-like motif. Although this region is not required for the binding of SRBC to PKC␦, it might be involved in the formation of complexes with themselves or other proteins. The rest of the conserved amino acids are clustered in a small region called SCR2. This region in SRBC includes two putative PKC phosphorylation sites, one of which is conserved among the three proteins. Since only the C-terminal half of SRBC was encoded by the cDNA clones first identified in the expression library, SCR1 must be dispensable for the binding to PKC␦. No striking sequence homology is found among known PKC-binding proteins, AKAP79 (21), MARCKS (10), and the C-terminal part of SRBC including SCR2. However, it is noteworthy that SCR2 as well as the PKC-binding region in AKAP79 and MARCKS show a high content of basic amino acids and both SRBC and MARCKS show similar binding patterns to PKC␦ (Figs. 3 and 4). Therefore, SCR2 could be responsible for the binding to PKC in the same way as the basic amino acid region of MARCKS.
The expression pattern of the srbc gene shares some common features with that of sdr; both srbc and sdr are induced by serum starvation of NIH3T3 cells and down-regulated by the addition of serum. This suggests that sdr and srbc form a family of genes that play a role in cell-growth control. In addition, other genes induced by growth arrest of cells have been identified in different systems (34,35) but show no structural homology with SRBC. Noteworthy, the N-terminal 185 amino acids of SRBC, including SCR1, have been found in the Nterminal part of the oncogenic c-RAF-1 gene, replacing the N-terminal regulatory domain of c-RAF-1 (30). The significance of the SRBC sequence fused to c-RAF-1 is unclear since a variety of genes can activate the c-RAF-1 gene by similar gene fusion (36). However, since the SRBC-RAF fusion gene seems to be under the control of the srbc promoter, the fusion gene is most likely expressed at high levels in quiescent cells, and this could explain some of the oncogenic properties of the fusion gene. The ubiquitous expression of srbc in different tissues and its induction upon retinoic acid-induced differentiation of mouse embryonal carcinoma P19 cells suggest the involvement of SRBC in a variety of the cellular events that accompany growth arrest.
A mutant PKC␦ lacking kinase activity shows a dominantnegative effect on the TPA-induced activation of the TRE-tk-CAT reporter gene in NIH3T3 cells (9). Furthermore, the same dominant negative mutant suppresses TRE-tk-CAT expression caused by the ectopic expression of a PKC␦-active mutant (9). Thus, we examined the effect of SRBC on TRE-tk-CAT expres- FIG. 7. Northern blot analysis of SRBC mRNA. A, expression of SRBC mRNA in mouse tissues. Total RNA was extracted from mouse organs and analyzed on Northern blot (10 g) using 32 P-labeled SRBC cDNA as a probe. The positions of ribosomal RNAs (18 S and 28 S) are indicated. The same blot was sequentially probed with SRBC (upper panel) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (lower panel) cDNAs. B, expression of SRBC mRNA in various cultured cell lines. Poly(A) ϩ RNA from NIH3T3 cells, 3Y1 cells, and F9 cells (0.4 g each), total RNA from COS1 cells (10 g), and poly(A) ϩ RNA (1.6 g) from undifferentiated P19 cells (P19 cells) and differentiated, retinoic acid-treated, P19 cells (P19 cells-RA) were applied. C, SRBC mRNA level after serum starvation was analyzed by Northern blot using equal amounts of poly(A) ϩ RNA (0.5 g) isolated from growing NIH3T3 cells (time 0) and cells at the indicated times after the start of starvation with 0.5% FCS/DMEM. D, analysis of SRBC mRNA level during G 0 and G 1 transition. NIH3T3 cells were starved for 48 h in 0.5% FCS and stimulated to reenter the cell cycle by the addition of 20% FCS. At the indicated times, RNA was isolated and 0.5 g of poly(A) ϩ RNA was analyzed by Northern blot with SRBC and glyceraldehyde-3-phosphate dehydrogenase probes. sion in NIH3T3 cells. However, we failed to detect any significant effect (data not shown). Another series of preliminary experiments to examine the effect of SRBC overexpression on the growth of NIH3T3 cells demonstrated the inhibition of DNA synthesis and colony formation, 4 supporting the idea that SRBC is involved in the regulation of cell growth. We cannot conclude that PKC is involved in these events; however, the further characterization of SRBC might provide a clue to understanding the signaling pathway by which PKC controls cell growth and differentiation.