A cDNA Encoding Protein Kinase C Identifies Two Species of mRNA in Brain and GH3 Cells"

Antiserum raised against purified protein kinase C (the Ca2+/phospholipid-dependent enzyme) (Ballester, and Rosen, 0. M. (1985) Biol. 260, was used to screen a rat brain cDNA library in the prokaryotic expression vector Xgtll. Three positive clones were isolated and shown to have overlapping restriction endonuclease maps. The posi- tive recombinant phage with the longest cDNA insert (1.4 kilobases (kb)) was used for production of a 8- galactosidase fusion protein. Rabbit antiserum raised against the fusion protein recognized a single rat brain polypeptide of M, 80,000 which was identified as protein kinase C by the following criteria: (i) electropho- retic eo-migration with purified protein kinase C, (ii) partial co-purification with protein kinase C, and (iii) disappearance from the cytosol of phorbol 12-myris-tate 13-acetate-treated GH, cells. The nick-translated cDNA hybridized with two mRNAs, 8 kb and 3.5 kb, whose tissue distribution was in agreement with that reported for protein kinase C activity. Hybrid selection with immobilized cDNA identified mRNA encoding a protein of M, 80,000 that could be precipitated by antibody to purified protein kinase C. Treatment of GH3

this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1) Supported by National Institutes of Health training grant 5T32
The abbreviations used are: PMA, phorbol 12-myristate 13-acetate (12-0-tetradecanoylphorbol 13-acetate); kb, kilobase; IPTG, isopropylthio-@-D-galactoside; SDS, sodium dodecyl sulfate; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesuIfonic acid; EGTA, [ethylene bis(oxyethylenenitrilo)]tetraacetic acid. release of diacylglycerol from phosphatidylinositol (6, 7) trigger association of the predominantly cytosolic kinase with the plasma membrane. The translocation, activation, and subsequent turnover of the enzyme have been assessed by measuring enzyme activity (8-lo), phorbol ester binding (see Ref. 11 for review), and, most recently, by immunoprecipitation of metabolically labeled enzyme with an antibody elicited to purified rat brain protein kinase C (12). Since all studies of the modulation of protein kinase C have of necessity analyzed only biochemical and immunological properties of the enzyme, it was important to develop probes to evaluate potential transcriptiona1 regulation. For this reason, the antiserum prepared to purified protein kinase C (121, which recognizes only protein kinase C in an immunoblot of rat brain cytosol, was used to screen a rat brain cDNA library in the prokaryotic expression vector Xgtll. We now report isolation of a cDNA clone encoding a portion of protein kinase C and the use of this probe to detect two mRNAs whose cellular distribution is consistent with that of protein kinase C activity.

EXPERIMENTAL PROCEDURES
Materials-RNasin, Xgtll bacteriophage and Escherichia coli host strains Y1090 and Y1089 were purchased from Promega Biotec. Nucleoside triphosphates and PMA were from P-L Biochemicals. DNA polymerase I and restriction enzymes were obtained from New England Biolabs. International Biotechnologies, Inc. was the source of cesium chloride and IPTG. Guanidine isothlocyanate was purchased from Fluka and oligo(dT)-cellulose was from Collaborative Research. Proteinase K was from Bethesda Research Laboratories.
Nitrocellulose sheets (BA 85,0.45 pm) were obtained from Schleicher & Schuell. All radionucleotides were from Amersham Cop. Grand Island Biological Co. was the source of RPMI 1640 media. Protein A-Sepharose CL-4B, bovine serum albumin (fraction V), Nonidet P-40, pepstatin, leupeptin, and aprotinin were all from Sigma. Soybean trypsin inhibitor was supplied by Boehringer Mannheim.
Screening of a cDNA Library-The rat brain cDNA library (a gift of Dr. David Colman of New York University) was screened with rabbit antisera directed against protein kinase C by the second antibody-biotin system (ABC) of Vector Laboratories. Positive clones were plaque-purified twice, DNA was prepared (13) and subcloned into pUC19 (14), and restriction endonuclease sites were identified.
Preparation of @-Galactosidase Fusion Proteins and Fusion Protein Antisera-E. coli strain Y1089 was infected for 20 min a t 32 "C with recombinants immunoselected with antibody to either protein kinase C or, as a control, to the human erythrocyte glucose transporter.
Infected cells were plated at low density and incubated 16 h at 32 "C. Single colonies were replica-plated, and lysogens that grew at 32 "C but not 42 "C were selected and grown to an ODsoo of 0.5. Lysogens were induced at 45 "C for 20 min, followed by incubation for 1 h at 37 "C in the absence or presence of 10 mM IPTG (15).
Rabbits were immunized perilymphnodally with SDS-polyacrylamide gel pieces containing 100 pg of fusion protein homogenized in complete Freund's adjuvant. Subsequent injections of either gelpurified fusion protein or E. coli lysate were delivered intramuscularly with incomplete adjuvant. Data shown here were obtained using antisera from one rabbit following the third boost.
For immunoblotting, proteins were subjected to SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose paper (16). The blots were incubated for 2 h with antiserum directed against either purified protein kinase C (1:lOO dilution), human glucose transporter (1:75 dilution), fusion protein (125 dilution), or preimmune serum of appropriate dilution. Following incubation with 9protein A for 30 min, the blots were dried and subjected to radioautography at -70 "C with an intensifying screen.
Cell Culture and Preparation of Cell Ertracts-GHs cells were grown in RPMI 1640 media supplemented with 10% fetal bovine serum, 2 m M glutamine, and 7.5 mM Hepes buffer, pH 7.5. Cells were main-13389 Protein Kinase C mRNA tained in monolayer culture for 5-10 days before use. Prior to PMA treatment, cells were incubated with fresh media for 18 h. Either 400 nM PMA or dimethyl sulfoxide (the vehicle for PMA) was then added directly to the culture medium. Cell extracts were prepared by scraping the monolayers into homogenization buffer (5 mM NaC1, 5 mM Tris buffer (pH 7.5), 10 mM EGTA, 5 mM MgC12,2 mM dithiothreitol) in the presence of protease inhibitors (leupeptin, aprotinin, pepstatin, and soybean trypsin inhibitor) as previously described (12). Cells were disrupted by repeated passage through a 26-gauge needle. Lysates were centrifuged at 100,000 X g for 1 h to obtain the high speed supernatant fluid. Partial purification of protein kinase C was achieved by applying this fraction to a DEAE-52 column (equilibrated with 20 mM Hepes buffer (pH 7.5), 5 mM EGTA, 2 mM EDTA, and 2 mM dithiothreitol) and collecting the 0.1 M NaCl eluate as previously described (17).
Analysis of RNA and DNA-RNA was isolated from tissues by homogenization in guanidine isothiocyanate and centrifugation through 5.7 M cesium chloride (18) or from GHa cells by Nonidet P-40 lysis in the presence of RNasin, followed by proteinase K digestion and phenol/chloroform extraction (19). Further purification was achieved by oligo(dT)-cellulose chromatography (20).
For the hybidization selection, DNA was denatured, immobilized on nitrocellulose filters, and hybridized to poly(A+) RNA at 50 "C for 3 h (22). The selected RNA was eluted and translated with a rabbit reticulocyte lysate kit from Promega Biotec. Immunoprecipitation of the translated protein was performed as previously described (12).
For the Southern hybridization, high molecular weight DNA (23) was digested with the indicated restriction endonuclease, subjected to electrophoresis in 1% agarose, transferred to nitrocellulose, and hybridized to the nick-translated probe (24). Hybridization was carried out at 42 "C, and the filter was washed at 50 "C as described above.

RESULTS AND DISCUSSION
Ten positive clones were identified by screening 1.5 x lo5 plaques with the second antibody-biotin system of Vector Laboratories. DNA was prepared from three of the bacteriophage, and the inserts were shown to have overlapping restriction endonuclease maps. To confirm the identity of the positive clones, a P-galactosidase fusion protein was prepared by infecting E. coli Y1089 with the Xgtll bacteriophage containing the largest insert (1.4 kb) that produced immunoreactive protein. The Coomassie Blue staining pattern of the fusion protein in a lysate from E. coli induced by IPTG is shown in  C (a and b ) or to the human erythrocyte glucose transport (c). Lysogens were induced in the absence (-) or presence (+) of 10 mM IPTG. a shows Coomassie Blue stain of lysogens infected with the protein kinase C-immunoselected phage containing the 1.4-kb insert. Lysates were subjected to electrophoresis in a 7.5% SDS-polyacrylamide gel. b is a radioautogram of an immunoblot of lysogens from panel a. Proteins were transferred to nitrocellulose (16), and the blot was incubated with antibody to purified protein kinase C (12) followed by ''51-protein A. c is a radioautogram of an immunoblot of a @-galactosidase fusion protein containing the C-terminal region of the rat brain glucose transporter.' The blot was incubated with an antibody to human erythrocyte glucose transporter (25). Each lane in panels a-c contains 100 pg of E. coli lysate. The arrow indicates the largest fusion protein synthesized by each culture. Immunoblot of rat brain proteins with antibody raised against the protein kinase C-&galactosidase fusion protein. a, immunoblot of a 100,000 X g supernatant fluid from rat brain homogenate (100 pg of protein per lane); b, immunoblot of partially purified rat brain protein kinase C (30 pg of protein per lane). The protein kinase C was purified by chromatography on DEAE-cellulose, gel filtration and DEAE-trisacryl ion-exchange chromatography as previously described (17). Lunes I and 3, preimmune and immune serum, respectively, from a rabbit immunized with the P-galactosidase fusion protein; lane 2, antibody raised against purified protein kinase C. purified by SDS-polyacrylamide gel electrophoresis (100-200 pg of fusion protein per injection). An immunoblot of rat brain cytosol (Fig. 2a) showed that a protein of M , 80,000 was recognized by immune serum (lune 3 ) , but not by preimmune serum from the same rabbit (lune 1 ) or antisera raised against the transporter fusion product (data not shown). The protein of M , 80,000 co-migrated with the protein recognized by the antibody prepared to purified protein kinase C (lune 2). Several minor bands (including one of M , 44,000) recognized by immune but not preimmune serum were also detected with sera from rabbits immunized against the glucose transporter fusion protein. Immune but not preimmune serum continued to recognize a protein of M , 80,000 in a 10-fold purified protein kinase C preparation (Fig. 2b, lunes 1 and 3 ) . Antisera from rabbits immunized with the glucose transporter fusion protein did not cross-react.
To confirm the identification of the immunoreactive 80kilodalton protein, GH3 cells were treated with 400 nM PMA, a condition which drastically and as far as is known, specificially reduces cellular protein kinase C (12). Cytosolic extracts were immunoblotted with antibody to either purified protein kinase C (Fig. 3u) or the 150-kilodalton fusion protein (Fig.  3b). Both antisera detected the loss of an immunoreactive species, M, 80,000, in response to PMA treatment.
The identification of the cDNA was further supported by Northern blot hybridizations using a nick-translated probe prepared from the 1.4-kb insert that had been subcloned into the EcoRI site of pUC19 (14). The probe hybridized with two RNA species of 3.5 and 8 kb in rat brain poly(A+) RNA (Fig.  4A, lune 1). Significantly, the tissue distribution of both messages is consistent with that of protein kinase C activity (26)(27)(28). Message levels were highest in brain (lune 5 ) and 5-10-fold lower in the other tissues surveyed (lung, adipose, kidney, and liver, lunes 3, 4, 6, and 7, respectively). The protein kinase C activity of GH3 cells (12) is about 3-5-fold lower than rat brain (17). This is consistent with the Northern blot of GH3 cells (lune 2 ) which reveals message levels intermediate between those of brain and the other tissues. Using RNA prepared from PMA-treated (Fig. 4B, lunes 2-4, 6 ) and untreated (Fig. 4B, lunes 1 and 5 ) GH3 cells, it was evident that treatment with 400 nM PMA did not significantly affect protein kinase C mRNA levels following exposures of up to 1 h (lune 4 ) or as long as 24 h (lunes 5 and 6). This conclusion is in agreement with the report that synthesis of protein kinase C (as judged by pulse labeling with [35S]methionine) is a.
b. unchanged in PMA-treated cells (12). The observation that there appear to be two mRNAs is of interest. A message of 3.5 kb would be sufficient to encode a protein of M , 80,000.
The two messages might be alternately spliced forms of the same primary transcript or messages from two closely homologous genes. The relative abundance of the two messages appears similar in brain and GH3 cells; it remains to be determined whether this is the case for other tissues in which the level of protein kinase C mRNA is low.
Hybrid selection was performed to show that the mRNA identified by the 1.4-kb insert in fact encodes an 80-kilodalton both mRNAs remained unaffected by phorbol ester-induced down-regulation of the enzyme in GH, cells. Res. Commun. 120,1053-1059 protein (Fig. 5 ) . pUC19 DNA into which the 1.4-kb insert had been subcloned, pUC19 DNA alone, and a control plasmid into which the full length rat brain glucose transporter cDNA had been subcloned were immobilized on nitrocellulose and hybridized with rat brain poly(A+) RNA. The RNA thus selected was eluted, translated, and immunoprecipitated with antibody raised against purified protein kinase C. As shown in Fig. 5, the plasmid containing the 1.4-kb insert selected mRNA encoding an M , 80,000 protein identified as protein kinase C by immunoprecipitation. The M , 80,000 protein was not observed when either the control plasmid or the plasmid containing the glucose transporter cDNA was used for hybrid selection. Thus one or both of the messages observed on Northern blot analysis encode protein kinase C. Southern analysis of high molecular weight DNA prepared from rat brain using the 1.4-kb cDNA indicates a simple pattern but one with multiple large hybridizing fragments (Fig. 6). This is consistent with either a single large gene or several closely related homologous sequences in the rat genome.
Although the portion of the kinase molecule encoded by the 1.4-kb insert is not known with certainty, it is likely that it includes the phospholipid-and Caz+-binding domain rather than the kinase domain, since the antibody used to identify the clone recognizes only the intact M, 80,000 species and not the M , 50,000 phospholipid-and Ca2+-independent trypsingenerated fragment of protein kinase C, designated protein kinase M (29)(30)(31).
In conclusion, we have isolated a clone for protein kinase C in a Xgtll cDNA library from rat brain. Initial studies indicate that there are two predominant messages that hybridize with this cDNA (3.5 and 8 kb) and that both are more prevalent in brain and a neural-derived cultured cell line than in other tissues examined. This distribution is consistent with that reported for protein kinase C activity. The content of