Changes in Cyclic Adenosine Monophosphate-Responsive Element Binding Proteins in Rat Hepatomas

Home , Deoxycytidine

(CANCER RESEARCH 51. 528-535. January 15. 199I| Changes in Cyclic Adenosine Monophosphate-responsive Element Binding Proteins in Rat Hepatomas

Joanna Kwast-Welfeld, Ian de Belle, P. Roy Walker, Richard J. Isaacs, James F. VVhitfield, and Marianna Sikorska1

Molecular Cell Biology Group, Institute for Biological Sciences, National Research Council, Ottawa, Ontario, Canada K 1.4 OR6

ABSTRACT associated with normal liver function (7). In some cases (i.e., hepatoma 5123tc), these enzymes are expressed at constitu- We applied Southwestern and Western blotting and gel retardation techniques to investigate the changes that occur in the cyclic adenosine tively high levels (8). Recent developments in our understanding monophosphate (cAMP)-responsive element (CRE) binding (CREB) pro of the mechanisms of transcriptional activation of cAMP-re- teins in rapidly growing, chemically induced 5123tc and 5123D Morris sponsive genes has permitted us to reexamine the basis for the hepatomas. Using the CRK sequences from the r-/Â«.v.E2A, and somato- alterations in these responses to cAMP in hepatoma cells. statin gene promoters, we identified in the nuclear proteins from normal In the search for the molecular mechanism by which cAMP unstimulated or proliferating rat liver cells six different protein factors controls gene expression, a as-acting DNA regulatory sequence of M, 34,000, 36,000, 40,000, 47,000, 56,000, and 72,000 capable of named the CRE has been discovered (9, 10). The CRE consen binding to the element. The M, 47,000 protein had the highest specificity for the core CRE, suggesting its importance in cAMP-mediated gene sus sequence found in the promoter region of a number of cAMP-inducible genes is an 8-base pair palindrome, 5'- expression. We could not lind the M, 47,000 CREB protein in the 5123tc TGACGTCA-3'. After that, a number of nuclear protein factors and 5123D hepatomas. Our efforts to detect this protein in the tumors by (a) using the CRE sequence from different gene promoters, (b) altering ranging in MTfrom 38,000 to 55,000 that are capable of binding the protocol for extracting nuclear proteins, or (<â€¢)attempting to restore to this sequence have been either predicted to exist or actually its DNA-binding property by phosphorylation [with endogenous protein identified (11-17). These transcription factors provided the kinase(s), a catalytic subunit of cAMP-dependent protein kinase, and missing link between the elevation of the intracellular level of protein kinase C/dephosphorylation (with alkaline phosphatase)) Â»ere unsuccessful. The loss of the M, 47,000 CREB protein from solid tumors cAMP and changes in gene expression. of the Morris hepatoma is likely to be related to the neoplastic properties Several genes whose products are essential for liver cell of the tumor cell rather than to cell growth because the level of this function [such as phosphoenolopyruvate carboxykinase (18), protein remained unchanged during a 6-day period of liver regeneration. lÃ¡clatedehydrogenase (15), glutamine synthetase (19), tyrosine The nuclear extract from the Morris hepatoma that did not have the M, aminotransferase (20), and c-fos (21, 22)] contain a functional 47,000 CRE-binding factor contained proteins immunologically related CRE sequence within their promoters. Of these genes, c-fos has to the CREB, c-Jun, and c-Fos proteins. We conclude that the M, 47,000 factor represents a distinct member of the CRE-binding protein family attracted the most attention in recent years because it belongs and that its absence from the hepatomas may lead to aberrant expression to a family of genes that respond rapidly to extracellular signals of cAMP-inducible genes. (23, 24) and because its product, the Fos protein, is itself a transcription factor involved in the transcriptional control of INTRODUCTION other genes (25). Therefore, abnormal expression of c-fos could Transient changes in the intracellular level of cAMP2 are initiate a cascade of events leading to, or contributing to, cell characteristic features of proliferating hepatocytes (1). They transformation (26). stimulate events leading to the initiation of DNA replication We showed recently that tissues in vivo and cultured cells in and mitosis, suggesting that cAMP-responsive genes are essen vitro contain multiple CREB proteins that are involved in tial for the proliferation of normal liver cells (2-4). A loss of conferring cAMP-responsiveness on genes.' We identified five these cAMP-mediated cell cycle controls could deregulate the different CREB factors with M, 34,000, 36,000,40,000,47,000, cell cycle and contribute to the generation of tumor cells that and 56,000 in nuclear extracts from rat liver using the CREs of have different growth rates and are blocked at various stages of the c-fos and somatostatin genes as probes. Among these fac differentiation (5). Moreover, alterations in the cells' ability to tors, the MT47,000 CREB protein showed the highest specific respond to changes in cAMP are likely to be involved in the ity for the CRE sequence, suggesting that this protein could development of the phenotypic diversity that is observed in the play a very important physiological role in mediating the effect Morris series of experimental hepatomas (6). For example, the of cAMP on gene expression. The M, 47,000 protein band amino acid transport systems and cAMP-inducible gluconeo- observed here was originally identified by Kwast-Welfeld et al. genic enzymes (such as tyrosine aminotransferase, phosphoen- (15) in rat ovarian tissue, rat liver, and C6 glioma cells as a olpyruvate carboxykinase, and glutamine synthetase) do not protein that specifically binds to the lactate dehydrogenase A undergo the normal periodic oscillations in activity that are subunit gene promoter, which contains a CRE octamer. To better understand the changes in cAMP-controlled gene Received 5/9/90: accepted 10/26/90. The costs of publication of this article were defrayed in part by the payment expression in tumors we analyzed the nuclear proteins from 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. rapidly growing neoplastic cells of chemically induced 5123tc 1To whom requests for reprints should be addressed, at Institute for Biological and 5123D Morris hepatomas for their ability to bind to the Sciences, National Research Council. M-54 Montreal Road. Ottawa K1A OR6. DNA fragments containing the CRE sequence using the South Canada. : The abbreviations used are: cAMP. cyclic AMP; CRE. cAMP-responsive western hybridization technique and gel retardation assays. element: CREB. CRE-binding; HPX. hepatectomized: poly(dl-dC). alternating copolymer of deoxyinosine and deoxycytidine; SDS-PAGE. sodium dodecyl 'J. Kwast-Welfeld. P. R. Walker. J. F. Whitfield, and M. Sikorska. Influence sulfate-polyacrylamide gel electrophoresis; TPA, 12-0-tetradecanoylphorbol 13- of flanking sequences on DNA-protein interactions at the cyclic AMP-responsive acctate. element. J. Biol. Chem.. submitted for publication. 528

MATERIALS AND METHODS Synthesis of Peptides, Production of Antibodies, and Immunoblotting of Nuclear Proteins. The amino acid sequence of human placentai Tissue Sources. Normal liver tissue was obtained from 200-250-g CREB (11), mouse c-Jun (34), and mouse c-Fos (35) proteins were male, specific pathogen-free Sprague-Dawley rats bred in this labora scanned for likely antigenic sites using the University of Wisconsin tory. Partial hepatectomy (70% HPX) was performed as previously Genetic Computer Group software package and the algorithm of Jame described (27). Hepatomas 5123D and 5123tc were propagated by s.c. son and Wolf (36). The peptides. representing amino acids 134-150 of inoculation into the inguinal regions of 200-300-g male Buffalo rats CREB, 301 -314 of c-Jun, and 77-90 of c-Fos proteins, were synthesized (Harlem Sprague Dawley Co., Indianapolis, IN) as described by by the simultaneous multiple peptide synthesis method (37). New MacManus (28). The original malignant (metastasized to lung) 5123 Zealand rabbits were immunized with unconjugated peptides according hepatoma was produced in a female Buffalo rat that had been fed a to the following schedule. The first i.m. injection of 0.5 mg peptide slowly acting carcinogen, 7V-2-fluorenylphthalamic acid (29). The emulsified with complete Freund's adjuvant (1:1 v/v) was followed by 5123tc hepatoma originated from 5123 tumor cells that had been two consecutive s.c. injections of peptides emulsified with incomplete passaged in tissue culture, and the 5123D hepatoma is a slower-growing adjuvant at 1-wk intervals. The rabbits were test bled at 7 and 14 days subline of the 5123 tumor after the 16th serial transplant generation. after the last vaccination, and the titers of their sera were determined Preparation of Nuclear Protein Extracts. Nuclei were isolated and by enzyme-linked immunosorbent assay. Nuclear proteins were resolved purified following the low speed centrifugation and Triton X-100 by 10% SDS-PAGE, electrotransfered onto nitrocellulose membranes washing procedure as described previously by Sikorska and Whitfield (Hybond C, Amersham), and immunoblotted with the appropriate (30). Proteins were extracted from nuclei with 25 mM 4-(2-hydroxy- antisera. The antigen-antibody complexes were visualized by autoradi ethyl)-l-piperazineethanesulfonic acid buffer, pH 7.9, containing 25% ography on Kodak SB5 film after labeling with ['25I]-protein A (DuPont v/v glycerol, 1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM phenylmethylsul- NEN). fonyl fluoride, 0.5 mM dithisthreitol. and 0.42 M NaCl for 30 min on Other Procedures. SDS-PAGE electrophoresis was carried out as ice as described in Ref. 31. described by Laemmli (38). Autoradiographs were scanned on a Bio- Preparation of Oligonucleotide Probes. Complementary strands of Rad model 620 video densitometer, and molecular weights were cal oligonucleotides containing the cyclic AMP-responsive element were culated from a polynomial standard curve drawn through the positions prepared by the phosphoroimidate method on an Applied Biosystem of l4C-labeled markers (Amersham). Protein kinase C was purified from 380 synthesizer. A 78-base pair (â€”97to â€”20)restriction fragment rat brain as described by Kitano et al. (39) and its activity was tested (BssHU/Pvull) was cut out from a pc-fos (mouse)-3 plasmid (American using two synthetic peptides one as a substrate (40) and one as a specific Type Culture Collection no. 41041). The probes were 3'-end labeled inhibitor (41). The activity of the catalytic subunit of cAMP-dependent with [Â«-"PjdATP and [a-32P]dGTP (specific activity, 800 Ci/mmol; protein kinase (Sigma) was measured using the Kemptide fragment New England Nuclear, Boston, MA) using the Klenow fragment of (42). The protected peptide resins were synthesized using a p-methyl- DNA polymerase. The labeled probes were purified through Nick benzhydrylamine resin and A'-re-tertiary butoxycarbonyl (t-boc) amino columns (Pharmacia, Uppsala, Sweden) and used in the Southwestern acids (Bachern Inc., Torrance, CA) by the method of Houghten (37). blotting and gel retardation assays. All of the oligonucleotides used in the study are listed in Table 1. RESULTS Gel Retardation Assay. The gel retardation assays were carried out as described by SivaRaman et al. (32). Briefly, 20 ng of nuclear proteins Hybridization of Nuclear Proteins with DNA Probes Contain were preincubated for 10 min with 2 ing poly(dl-dC) (Pharmacia) at ing the CRE Sequence. A modified CRE has been found in the room temperature; then 0.2 ng of labeled oligonucleotides (approxi promoter region of all c-fos genes cloned thus far, such as the mately 10,000 cpm) was added, and the incubation was continued for a further 30 min. The DNA-protein complexes were resolved on 4% human, chicken, and mouse genes. The highly conserved regu latory element contains the sequence -TGACGTA-. In an in polyacrylamide gels and were visualized by autoradiography on Kodak X-OMAT Him. vitro transcription assay the element mediates cAMP inducibil- Southwestern Blotting Procedure. The Southwestern hybridization ity through both heterologous and homologous promoters (21, technique was adopted from the original method of Silva et al. (33) as 22, 35). To identify nuclear protein factors that can interact described by Kwast-Welfeld et al. (15). Nuclear proteins (150-200 ^g) with the c-fos CRE and regulate the expression of the gene in were resolved by 8.5% w/v SDS-PAGE gels, which were then incubated vivo, we analyzed nuclear proteins from normal and neoplastic for 2.5 h at room temperature with gentle agitation in renaturing buffer liver tissues by the Southwestern hybridization technique (Fig. consisting of IO mM Tris. pH 7.4, 50 mM NaCl, 2 mM EDTA, 0.1 imi 1). A 78-base pair BssHH/Pvull restriction fragment of the dithisthreitol. and 4 M urea. After renaturation, the proteins were mouse pc-fos-3 plasmid (â€”97to â€”20)containing the CRE electrotransfered (0.1 A, 16 h) onto nitrocellulose filters (Schwarz- sequence (â€”66to â€”60)as well as CAAT (â€”49to â€”45)and Mann) and the filters were blocked for 45 min with 5% w/v nonfat TATA (â€”33to â€”29)boxes was 3'-end labeled, hybridized to milk powder in 10 mM Tris, pH 8.0, 2 mM MgCl2, 1 mM mercaptoeth- anol, and 50 mM NaCl. This was followed by an additional 45-min the nuclear proteins resolved by SDS-PAGE, and blotted onto incubation with blocking buffer containing 25 Mg/ml of poly(dl-dC). nitrocellulose (Fig. \A). The 78-base pair probe recognized After this time 0.05 ng/m\ of radioactive oligonucleotide probe (ap several protein species of M, 34,000, 36,000, 38,000, 40,000, proximately 2-5 x IO6cpm/ml) was added and incubation was contin 45,000, 47,000, and 56,000 in the nuclear extract from normal ued for a further 3.5 h. After the hybridization, the filters were washed liver at 24 h after HPX (lane I). A similar overall pattern of twice for periods of 10 min each with blocking buffer and once with 10 bands was obtained with nuclear proteins extracted from he- mM Tris. pH 8.0, containing 2 mM MgCl:. The filters were dried and patomas 5123tc (lane 2) and 5123D (lane 3), but the intensity autoradiographed on Kodak X-OMAT film. of the protein labeling was stronger than in normal liver. However, the probe did not detect in either of the two hepato- Table 1 Synthetic oligonucleotide probes with the native CRE sequence The oligonucleotides contain a 4-base pair /(Â«Â»/111linkerthat was used for 3'- mas the protein doublet of M, 45,000 and 47,000 that was end labeling of the probes. identified in the regenerating liver. The probe also labeled a few proteins of higher molecular weight, from approximately Ref. Somato- galcGCGCCTCCTTGGCTGACGTCAGAG (-60 to -37) 9 M, 90,000 to over M, 200,000, particularly in nuclear extracts statin from hepatomas (Fig. IA, lanes 2 and 3). c-fos gatcCAGTTCCGCCCAGTGACGTAGG (-79 to -58) 36 To eliminate the possible influence of the CAAT and TATA, E2A gatcCGCTGGAGATGACGTAGTTCCCG (-85 to -63) 33 boxes on the number of proteins recognized by the c-fos restric- 529

was identical to that obtained with the c-fos CRE (Fig. I, A and A B B). However, the perfect CRE octomer of the somatostatin promoter had a higher affinity than the c-fos CRE for the M, .â€¢ 47,000 protein from normal regenerating liver (Fig. 1C, lane 1). Moreover, as with the c-fos CRE, the somatostatin probe -100 did not detect the M, 47,000 protein in either the 5123tc (Fig. -97 1C, lane 2) or the 5123D (Fig. 1C, lane 3) hepatomas. 72- The adenoviral early gene promoter (E2A) contains the same -69 CRE sequence as the c-fos gene, and it is also stimulated by cAMP (32, 43). An oligonucleotide probe representing a frag -56- ment of the E2A gene promoter (â€”85to â€”63,Table 1) recog -47- 45- -46 nized and bound to the M, 47,000 CREB protein in the nuclear extract from normal liver (Fig. \D, lane /), but again this probe 38- could not detect the M, 47,000 CRE-binding protein in the hepatomas (lanes 2 and 3). -30 Effect of Flanking Sequences on Protein Binding at the CRE. Although several proteins can bind to the CRE sequence, it is likely that the affinity of the proteins for the CRE are 1 2 3 1 2 3 1 2 3 1 2 3 mw influenced by the flanking sequences.' To find out which one Fig. I. Hybridization of nuclear proteins with DNA probes containing the of the six proteins recognized by the CRE-containing DNA CRE sequence.Two hundred jig of nuclear protein extracts from normal liver at probes had the highest specificity for the regulatory element, 24 h after HPX (lane I) and from Morris hepatoma 5123tc (lane 2) and 5123D (lane .?) solid tumors were separated by 8.5% SDS-PAGE, renatured, and clec- we probed nuclear extracts with the somatostatin CRE contain trotransfered onto nitrocellulose membranes. The blots were probed with the 78- ing 5' flanking sequences of various lengths (Fig. 2). The longest base pair '"P-labeled BssHlÃ¬/PvuÃ¬lrestriction fragment of the c-fos gene (A), the probe was the 24-base pair fragment (used throughout this 22-base pair synthetic fragment of the c-fos gene promoter (A), the 24-base pair oligonucleotide of the somatostatin gene (C), and the 23-base pair promoter study), and the other probes were shortened from the 5'-end by fragment of the E2A gene (fl), all containing the CRE sequence. The labeled 4 bases to a 20-mer and by 8 bases to a 16-mer. We hybridized complexes were visualized by autoradiography, and the molecular weights of the DNA-binding proteins were estimated by densitometric scanning, as described in these probes with the protein extract from proliferating rat liver "Materials and Methods." (lane H) and with the nuclear proteins from the 5123tc hepa toma (lane tc). Very clearly, the labeling of all of the other tion fragment, we synthesized a 22-base pair oligonucleotide nuclear proteins, except the M, 47,000 protein depended on the fragment (-79 to -58) that contained only the CRE (Table 1) length of the probe, with the intensity of their labeling decreas and hybridized it to the same nuclear proteins (Fig. \B). The ing with the length of the probe. At the same time, the binding oligonucleotide recognized and bound specifically to six nuclear of the Mr 47,000 protein from the normal liver extract (lane proteins of M, 34,000, 36,000, 40,000, 47,000, 56,000, and 72,000 in the regenerating liver extract (lane 1). Their molec ular weights corresponded very well to those recognized and labeled by the native 78-base pair c-fos restriction fragment. B However, the synthetic c-fos probe additionally labeled a M, 72,000 protein, whereas it did not bind to the M, 38,000 and 45,000 proteins. All of the proteins present in normal liver, except the M, 47,000 and 72,000 CREB proteins, were also present in both hepatomas (Fig. \B, lanes 2 and 3). Of these two proteins the one of Mr 47,000 was of greatest interest to us in this study because it was detected in a variety of cells and 72- tissues (15)' and recognized only the CRE sequence and not the homologous TPA-responsive element,' whereas the M, 72,000 protein is liver specific and not always reproducibly detectable. The labeling of the M, 34,000-56,000 nuclear pro 56- teins by the synthetic oligonucleotide was more intense than that by the restriction fragment, but there was very little inter 47- action of this probe with the proteins of higher molecular 40- weights (the M, 90,000 and 200,000 proteins). 36- The rat somatostatin gene promoter contains a perfect pal- 34- indromic CRE sequence -TGACGTCA- which differs from the C-/0S-CRE by the presence of a C at position 7 of the element and which confers full cAMP inducibility on the somatostatin gene (9, 10). Perfect palindromic CREs are also found upstream of other inducible liver genes such as lÃ¡clate dehydrogenase tc H te H tc H (15). We used this fragment of the somatostatin gene promoter Fig. 2. Effect of the length of the DNA probe on protein binding at the CRE. (-60 to â€”31,Table 1) to probe nuclear proteins from regener Two hundred pg of nuclear proteins from proliferating liver at 24 h after HPX ating liver and hepatomas in search for protein(s) that may not (lane //) and from hepalomu 5123lc (lane ic) were separated by 8.5'/Â¿SDS- PAGE, renatured, and electrotransfered onto nitrocellulose filters. The filters be recognized by the imperfect CRE of c-fos (Fig. 1C). The were hybridized with the radioactive 24-base pair (A), 20-base pair (fi), and 16- pattern of protein bands recognized by the somatostatin CRE basc pair (C) oligonucleotide fragments of the somatostatin gene promoter. 530

up to 6 days after partial hepatectomy (Fig. 4). Liver nuclear proteins from laparotomized and partially hepatectomized rats were prepared by the standard extraction procedure (0.42 M NaCI, 30 min) and hybridized with the somatostatin CRE oligonucleotide on the Southwestern blots. All extracts con tained five major bands corresponding to proteins of M, 34,000, 36,000, 40,000, 47,000, and 56,000. There were no significant differences in the intensities of the labeling between laparotom 56 ized (lane L) and partially resected (lane H) livers during the first 6 days of liver regeneration, but most importantly there was no reduction in the DNA-binding property of the M, 47,000 CRE-binding protein. Effect of Phosphorylation on the DNA-binding Property of CREB Proteins from Hepatoma Tissue. It is well established that phosphorylation plays a very important role in the modi fication of the affinity of fra/is-acting nuclear protein factors for DNA fragments containing the CRE sequence (12, 44). To test whether the absence of the M, 47,000 CRE-binding protein in hepatomas could be due to an impairment of protein phos phorylation resulting in the loss of its DNA-binding property, we preincubated nuclear proteins from 5123D tumors (Fig. 5-4, 12345 lane 1) with magnesium and ATP to activate endogenous Fig. 3. Effect of the nuclear protein extraction procedure on the CRE-binding nuclear kinase(s) (lane 2), with purified protein kinase C (lane proteins from the 5123D hepatoma. Nuclei prepared from 5123D hepatoma were 3), or with the catalytic subunit of cAMP-dependent protein incubated (on ice) with the extraction buffer containing 0.42 M NaCI for 45 min (lane /); with a doubled volume of extraction buffer for 45 min (lane 2); or with kinase (lane 4). There was no difference in the pattern of the the extraction buffer for 1.5 h (lane 3): the extraction with 0.42 M NaCI was proteins that bound to the somatostatin CRE sequence under repeated two times for 45 min each and the extracts were combined (lane 4); or any of these conditions. Moreover, the binding of the somato the extraction was done with 0.6 M NaCI for 1.5 h (lane 5). One hundred fifty /jg of protein extracts were separated by 8.5% SDS-PAGE, renatured. electrotrans- statin CRE sequence to the M, 47,000 protein could not be fered. and hybridized with the 3!P-labeled 24-base pair somatostain CRE as described in "Materials and Methods." restored in hepatomas by phosphorylation. On the other hand.

H) to the somatostatin CRE was not affected by the deletion of the 5'-end flanking sequences, suggesting a high specificity of this protein for the core regulatory element. Again, none of these probes detected the M, 47,000 protein in the hepatoma. Effect of the Nuclear Protein Extraction Procedure on CRE- binding Proteins in Hepatomas. Because the morphology of the tumor cells is different from the morphology of the normal cells, there was a possibility that the procedure used for the extraction of nuclear proteins from normal tissue was not stringent enough for extracting proteins from the solid tumors. Therefore, we examined the nuclear proteins obtained from hepatomas by modified extraction protocols for the presence of the M, 47,000 CREB protein (Fig. 3). In this and in all subse quent experiments, we used the somatostatin CRE sequence because this probe had the highest affinity for the M, 47,000 protein. Nuclear proteins were prepared from the 5123D tumor by extraction with 0.42 M NaCI for 45 min (lane /), extraction with double the volume of 0.42 M NaCI for 45 min (lane 2), extraction with 0.42 M NaCI for 1.5 h (lane 3), extraction with 0.42 M NaCI repeated two times for 45 min each time (lane 4), or extraction with 0.6 M NaCI for 1.5 h (lane 5). None of these more drastic extraction procedures resulted in the appearance of the M, 47,000 CREB protein in the nuclear extracts from L HL HL HL HL H the 5123D hepatoma. The same was true for the 5123tc tumor (data not shown), demonstrating that the standard extraction 345 protocol developed for normal tissue extracted all of the CREB proteins present in tumor cells. Days after surgery CRE-binding Proteins from Normal Proliferating Rat Fig. 4. The CRE-binding proteins from normal proliferating rat liver. Male Liver. To determine whether the absence of the M, 47,000 CRE- Sprague-Dawley rats (200 g) were subjected to 70% partial hepatectomy or binding protein from tumors was related to cell transformation laparotomy, and liver samples were collected at 2 to 6 days after surgery. Southwestern hybridization with the radioactive 24-base pair somatostatin probe or related simply to the general growth properties of the cells, was performed with 200 jjg of liver nuclear proteins from either hepatectomized we probed nuclear proteins from normal regenerating livers for (lane //) or laparotomized (lane /.) rats. 531

B Phosphorytatlon Dephosphorytatton Phosphorytolton

Free

HtoDHtoDHtoD

Control PKC PKA

WITHOUT WITH IWIBfTOR INHIBITOR

P-pepÂ«* i

PKA PKC Fig. 5. Effect of phosphorylation/dephosphorylation on the DNA-binding properties of the CRE-binding proteins from Morris hepatomas. A. Nuclear proteins from 5123D tumor cells were incubated for 20 min at 30Â°Cin the extraction buffer (lane I); in the extraction buffer containing 1.5 m\i MgCI2 and 1.25 mM ATP (lane 2): in the presence of magnesium, ATP, and catalytic subunits from cAMP-dependent protein kinase (lane 3)\ or in extraction buffer containing magnesium, ATP, 1 mM CaO2, 1 JIM TPA, 43 /JM phosphatidylserine, and protein kinase C (lane 4). Nuclear proteins from 5123D tumor were also treated with 10 units of intestinal phosphatase (Sigma) for 10 min (lane 5) or 30 min (lane 6) at 3TC in the presence of 5 mM MgCl2 and 0.5 mM phenylmethylsulfonyl fluoride (Sigma). One hundred ng of phosphorylated or dephosphorylated proteins were separated by 8.5% SDS-PAGE, renatured, electrotransfered, and hybridized to the 24-base pair fragment of the somatostatin promoter. B. Twenty /ig of protein extracts from liver (lane //), 5123tc (lane tc), and 5123D (lane D) hepatomas, either untreated (Control) or phosphorylated with protein kinase C (PKC) or the catalytic subunit of cAMP-dependent protein kinase (PKA), were incubated with poly(dl-dC) and 0.2 ng (approximately 10.000 cpm) of the 24-base pair somatostatin CRE in a gel retardation assay as described in "Materials and Methods." The DN A-protein complexes were resolved on 4% polyacrylamide gels and visualized by autoradiography. C. Left panel, ten units of catalytic subunits (PKA) were incubated for 15 min at 30'C with 20 mM Tris-HCI buffer, pH 7.5. 100 MMKemptide fragment (39), 2.5 mM MgCl2, and 100 pM ATP (5-8 x 10' cpm of |Â«-"P]ATP) in a total volume 20 pi. The reaction was inhibited by the addition of protein kinase A inhibitor (Sigma) at a concentration 1 mg/ml. The reaction mixture was analyzed by chromatography on polyethyleneimine-cellulose plates (Polygram CEL 300 PEI/UV2S4; Brinkmann Instruments Inc. USA) developed with distilled water to separate phosphopeptide (P- peptide) from the components of the reaction mixture. The radioactive spots were visualzed by autoradiography on Kodak SB5 film (3). Right panel, approximately 100 ng of purified protein kinase C (PKC) (36) were incubated for 10 min at 37'C with the reaction mixture consisting of 50 mM Tris-HCI, pH 7.5, 75 /JM peptide substrate (37), 50 pM ATP (0.5 nCi/assay of [n-32P]ATP), 1 mM CaCl2, 43 pin phosphatidylserine (PS) and l JIMTPA and with or without 10 /jg/assay of peptide inhibitor (38).

treatment of the tumor nuclear extract with alkaline phospha procedure did not allow us to study the effect of the absence of tase for 10 to 30 min abolished almost completely the DNA- the M, 47,000 protein on the interactions of the other CREB binding property of the CREB proteins (lanes 5 and 6). proteins with the CRE sequence in the tumor cells. Such studies The Southwestern blotting technique allowed us to identify are better executed with gel retardation assays where the DNA DNA binding proteins that retained their binding activity after fragment containing the regulatory element is incubated with a being separated by SDS-PAGE (i.e., those proteins that bind mixture of nuclear proteins in solution. We performed this to the DNA as either monomers or possibly homodimers). This assay (Fig. 5ÃŸ)using nuclear proteins from normal HPX liver 532

(lane H) and from both the hepatomas (lanes tc and D) with a This protein was present as a single band in the nuclear extract radioactively labeled somatostatin CRE sequence. Each of these from normal proliferating liver (24 h after HPX, lane H) and nuclear protein extracts formed complexes with the probe, but as a doublet in 5123tc hepatoma (lane tc). Both tissues also the mobilities of these complexes in the gel were very different contained a protein of M, 47,000 that immunoreacted with the (Fig. SB, control). The slowest-migrating complex was formed antibody raised against amino acids 301-314 of the c-Jun by nuclear proteins from normal liver (lane H). By contrast, sequence (34) (Fig. 6B: liver, lane H, and 5123tc hepatoma, two DNA-protein complexes of different mobilities were lane tc). Similarly, the antibody raised against a fragment of formed by nuclear proteins from the 5123D hepatoma (lane the c-Fos protein sequence (amino acids 77-90; see Ref. 35) D). and predominantly one complex was formed by the nuclear recognized a protein of M, 56,000 (Fig. 6C) that was also proteins from 5123tc tumor (lane tc). Despite the fact that present in both tissues (liver, lane //, and 5123tc, lane tc). Both nuclear extracts from both of the hepatomas displayed appar the anti-CREB and the anti-ywn polyclonal antibodies recog ently the same pattern of CRE-binding proteins (Fig. 1), these nized a number of other proteins that must have some homology proteins can form different complexes in solution. Phosphor- to CREB and jun because none of these proteins immuno ylation of the nuclear proteins as described above did not affect reacted with preimmune sera as shown with the CREB preim significantly either the mobility or the pattern of shifted bands mune serum in Fig. 6D. These results demonstrate that the (Figs. 5B, PKA and PKC). The phosphorylating activities of nuclear extract from 5123tc Morris hepatoma that did not have both protein kinases toward their specific peptide substrates are the M, 47,000 CRE-binding factor contained the proteins im- shown in Fig. 5C, demonstrating that the kinases were indeed munologically related to the CREB, c-Jun, and c-Fos proteins. active. Thus, even though the hepatomas did not have the M, 47,000 protein, the other CREB proteins were still able to bind DISCUSSION to the CRE and shift the mobility of the probe. On the other hand, dephosphorylation of tumor nuclear proteins abolished All of the CRE-containing DNA probes that were used in their affinity for the DNA probe, as seen on Southwestern blot this study recognized and reproducibly bound to five nuclear (Fig. 5A, lanes 5 and 6), and dephosphorylated proteins also proteins from normal proliferating rat liver under Southwestern did not interact with the CRE sequence in solution and did not blotting conditions. Their molecular weights were estimated as shift the probe's mobility in the gel (data not shown). This M, 34,000, 36,000, 40,000, 47,000, and 56,000. We showed clearly suggested that in vivo these CRE-binding proteins ex previously that these proteins were present in other rat tissues, isted in fully phosphorylated forms. as well as in a variety of cultured cells (15).' Some of these, Comparison of Nuclear Protein Patterns from Proliferating such as the family of proteins of M, 34,000-36,000 seem to Rat Liver and 5123tc Morris Hepatoma That Were Recognized bind nonspecifically to DNA. On the other hand, the M, 40.000, by Anti-CREB, Anti-Jun, and Anti-Fos Sera. To further char 47,000, and 56,000 proteins bind only to DNA fragments acterize the M, 47,000 protein and to assess whether it belongs containing the CRE sequence/' In our hands the M, 47,000 to a family of transcription factors known to interact with the protein had the highest specificity for the CRE, and its inter CRE sequence, such as the CREB (11-13), c-Jun (34, 45), or action with the element was not affected by either the length of c-Fos proteins (24, 25), we prepared peptide-derived antibodies the flanking sequences or their composition, indicating that against these proteins and immunoblotted nuclear protein ex this protein factor plays a critical role in the regulation of tracts from liver and hepatoma (Fig. 6). Using rabbit polyclonal expression of cAMP-responsive genes. serum we were able to detect a protein of M, 45,000 (Fig. 6A) To compare the CRE-binding proteins described here to the that was recognized by the anti-CREB antibody (raised against CREB proteins identified in other laboratories (11-17) and to amino acids 134-150; see Ref. 11). Its molecular weight is other transcription factors known to interact with the CRE equivalent to the CREB protein described by others (12, 13). sequences (45), we prepared peptide-derived antibodies against CREB, c-Jun, and c-Fos proteins. We compared the pattern of B proteins labeled by the CRE-containing probes on Southwest ern blots to that of those recognized by the appropriate anti 97- 97 bodies on Western blots. 97- The Mr 56,000 CRE-binding protein, which we observed here, has a molecular weight identical to that of the protein 69- recognized by the c-Fos antibody. However, it is very unlikely that these are the same proteins because it is known that c-Fos by itself does not bind to DNA (25) and, therefore, it would not 46- 46- , -47 46- 46- be seen on the Southwestern blot. On the other hand, it is possible that this CRE-binding factor is identical with the CRE- â€” BP1 protein identified from a complementary DNA library of human fetal brain and reported to have a molecular weight of 30- 30 30- 3C M, 54,500 (14, 17, 45). H tc H tc H tc H tc The peptide-derived CREB antibody recognized two bands Fig. 6. Comparison of nuclear protein patterns from normal rat liver and of M, 45,000 and 43,000 (Fig. 6A) which correspond to the M23tc Morris hcpatoma that are recognized by anti-CREB, anti-c-Jun, and anti- CREB and Â¿-CREBproteins from PCI 2 cells (12, 46), rat brain c-Fos sera. One hundred fifty to 200 ^g of nuclear proteins from normal rat li\er (13), and human placenta (11). Thus, CREB and 5-CREB are at 24 h after HP\ (lanes //) and from 5123tc Morris hepatoma (lanes te) were separated by I0r; SDS-PAGE. electrotransfered. and immunoblotted with anti- present in hepatomas that do not contain the M, 47,000 CRE- CREB (A). anti-c-Jun (B). and anti-c-Fos (C) sera and with the CREB preimmune binding protein. The molecular weight of the M, 47,000 form serum (D) diluted to a final concentration of 2-3 ^g/ml. The antigen-antibody of the c-Jun-related protein (Fig. 6B) is the closest to our M, complexes were visuali/ed by autoradiography on Kodak SB5 film after being labeled with '"l-labeled protein A (Du Pont N'EN). 47,000 factor; however, this protein was also present in the 533

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 1991 American Association for Cancer Research. CYCLIC AMP-RESPONSIVE ELEMENT BINDING PROTEINS tumor nuclear extract and yet the same nuclear extract did not ACKNOWLEDGMENTS contain the M, 47,000 CRE-binding protein (Figs. 1 to 3 and 5). All of the above results are consistent with the identification The authors thank Sue MacLean and Jack Daoust for excellent of a novel cAMP-responsive element binding protein. technical assistance and Bob Richards for preparing illustrations. We thank Dr. Bayar Thimmappaya for the gift of the nucleotide sequence Because we were unable to detect the M, 47,000 protein in of the E24 early promoter fragment and Dr. Balu Chakravarthy for the 5123tc and 5123D solid tumors of the Morris hepatoma testing the activity of protein kinase C. This is NRCC publication No. line and because there were no changes in any other of the 31918. CRE-binding proteins, it was possible that the M, 47,000 CREB protein had lost its DNA-binding properties in the neoplastic hepatocytes. Although it is known that the DNA-binding activ REFERENCES ity of CREB proteins depends on their state of phosphorylation (12, 47), the DNA-binding activity of the M, 47.000 protein 1. MacManus, J. P., Franks. D. J.. Youdale. T.. and Braccland. B. M. Increases in rat liver cyclic AMP concentrations prior to the initiation of DNA could not be restored upon phosphorylation by either endoge synthesis following partial hepatectomy or hormonal infusion. Biochem. nous kinases or cAMP-dependent protein kinase or protein Biophys. Res. Commun.. 49: 1201-1207, 1972. 2. Boynton. A. L.. and Whitfield. J. F. The role of cyclic AMP in cell prolifer kinase C. Moreover, the loss of the M, 47,000 CREB protein ation: a critical assessment of the evidence. Adv. Cyclic Nucleotide Res., 15: from the hepatomas seems to be related to the transformation 193-294. 1983. 3. Sikorska. M.. Whitfield. J. F.. and Walker. P. R. The regulatory and catalytic properties of tumor cells rather than to normal cell growth suhunits of cAMP-depcndent protein kinases are associated with transcrit- because the level of the protein and its DNA-binding capacity pionally active chromatin during changes in gene expression. J. Biol. Chem.. remained unchanged during liver regeneration. Experiments 26J: 3005-3011. 1988. 4. Roesler, W. J.. Vandenbark. G. R., and Hanson, R. \V. Cyclic AMP and the currently under way suggested that this protein disappears induction of cukaryotic gene transcription. J. Biol. Chem.. 263: 9063-9066. during the early stages of neoplastic development in the resist 1988. ant hepatocyte model. However, the lack of the M, 47,000 5. Walker. P. R., and Potter, V. R. Isozymc studies on adult, regenerating, prccancerous and developing liver in relation to findings in hepatomas. Adv. protein is not a universal feature of tumor cells inasmuch as Enzyme Regul.. 10: 339-364, 1972. the highly malignant mouse S49T-lymphoma cells in culture 6. Potter. V. R. Phenotypic diversity in experimental hepatomas: the concept have a high content of the M, 47,000 CREB protein.'A number of partially blocked ontogeny. Br. J. Cancer, 38: 1-23, 1978. of CREB proteins are known to exist (11-17),' but their specific 7. Walker. P. R. The Molecular Biology of Enzyme Synthesis, pp. 296-323. New York: Academic Press, 1983. functions and their interactions at the regulatory element that 8. Scott, D. F.. Butcher, F. R.. Potter. V. R.. and Morris. H. P. Naturally occurring and induced levels of amino acid transport and tyrosine amino leads to a change in the expression of a gene in response to transferase in rats bearing Morris hepatomas. Cancer Res., 34: 2127-2133, cAMP are still not clear. Regardless of the exact mechanism of 1972. 9. Montminy. M. R.. Sevarino. K. A., Wagner, J. A.. Mandel. G., and Good gene activation, the lack of M, 47,000 CREB protein in the man. R. H. Identification of a cyclic-AMP-responsive element within the rat hepatomas could profoundly affect the metabolism of the tumor somatostatin gene. Proc. Nati. Acad. Sci. USA, 83: 6682-6686, 1986. cells. 10. Deutsch, P. J.. Hcoffler. J. P., Jameson. J. L.. Lin, J. C.. and Habener, J. F. Structural determinants for transcriptional activation by cAMP-responsive Earlier studies on precancerous livers and chemically induced DNA elements. J. Biol. Chem., 263: 18466-18472. 1988. hepatomas showed very significant aberrations in the endocrine 11. Hoeffler, J. P., Meyer, T. E., Yun, Y'.. Jameson, J. I... and Habener. J. F. Cyclic AMP-responsive DNA-binding protein: structure based on a cloned control of the level of cAMP (48) leading to changes in the placental cDNA. Science (Washington DC). 242: 1430-1433, 1988. isoenzymes of carbohydrate metabolism (5), in amino acid 12. Gonzalez, G. A., Yamamoto. K. K., Fischer, W. H.. Karr. D.. Menzel. P., transport (8), in the inducibility of tyrosine aminotransferase Biggs. W.. III. Vale. W. W.. and Montminy, M. R. A cluster of phosphor (8, 49), and most recently in the expression of the c-fos gene ylation sites on the cyclic AMP-regulated nuclear factor C'REB predicted by its sequence. Nature (Lond.). 337: 749-752, 1989. (50, 51). An elevated level of c-fos mRNA has been found 13. Zhu, Z., Andrisani, O. M., Pot, D. A., and Dixon, J. E. Purification and particularly in hyperplastic foci of precancerous livers in the characterization of a 43-kDa transcription factor required for rat somato statin gene expression. J. Biol. Chem.. 264: 6550-6556. 1989. resistant hepatocyte model (50) as well as in hepatomas (51). 14. Maekawa. T., Sakura. H.. Kanei-lshii. C.. Sudo. T.. Yoshimura. T.. Fujisawa, In normal cells the expression of the c-fos gene is tightly J., Yoshida, M., and Ishii. S. Leucine zipper structure of the CRE-BP1 controlled and the protein plays a very important role in the binding to the cyclic AMP response element in brain. EMBO J.. 8: 2023- 2028, 1989. transcriptional activity of the cell (25). However, it has also 15. Kwast-Wclfeld. J.. Soong, C.. Short. M. I... and Jungmann. R. A. Identifi been shown that constitutive expression of the c-fos gene in cation of rat ovarian nuclear factors that interact with the cAMP-inducible bone marrow (52) and in cultured cells (26) can lead to their lÃ¡clatedehydrogenase A subunit promoter. J. Biol. Chem.. 264: 6941-6947, 1989. transformation. Although we do not have direct evidence of the 16. Andrisani. O.. and Dixon. J. E. Identification and purification of a novel participation of the M, 47,000 liver nuclear protein in the 120-kDa protein that recognizes the c.\MP-rcsponsive element. J. Biol. Chem.. 265: 3212-3218. 1990. regulation of the c-fos gene, the absence of this protein seems 17. Kara, C. J.. Liou. H-C.. Ivashkiv. L. D.. and Glimcher. L. H. cDNA for a to be correlated with the elevated expression of the gene in human cyclic AMP response element-binding protein which is distinct from precancerous livers and liver tumors (50, 51). Similarly, the CREB and expressed preferentially in brain. Mol. Cell. Biol.. 10:1347-1357, 1990. elevated, constitutive expression of cAMP-responsive gluconeo- 18. Short. J. M., Wynshaw-Boris. A., Short, H. P.. and Hanson, R. W. Charac genic enzymes such as tyrosine aminotransferase in hepatoma terization of the phosphoenolpyruvate carboxykinase (GTP) promoter-regu latory region. J. Biol. Chem.. 261: 9721-9726. 1986. 5123C (8, 49) may be due to the loss of the M, 47,000 tran 19. Crook, R. B.. Louie. M.. Deuel, T. F.. and Tomkins, G. M. Regulation of scription factor. Therefore, we might speculate that in normal glutamine synthctase by dexamethasone in hepatoma tissue culture cells. J. cells the protein constitutively occupies the CRE sequence and Biol. Chem.. 25.ÃŽ.-6125-6131.1978. 20. Noguchi. T.. Diesterhaft. M.. and Granner. D. Dibutyryl cyclic AMP in represses the genes. Upon transcriptional activation, modifica creases the amount of functional messenger RNA coding for tyrosine ami tion of the protein (probably by phosphorylation) could lead to notransferase in rat liver. J. Biol. Chem., 253: 1332-1335. 1978. 21. Sassone-Corsi. P.. Visvader. J., Ferland. L.. Mellon. P. L., and Verma, I. M. a change in its DNA-binding properties that would allow other Induction of proto-oncogene/av transcription through the adenylate cyclase positive transcription factors, either sequence specific or gen pathway: characterization of a cAMP-responsive element. Genes & Dev., 2: eral, to operate. Without the M, 47,000 protein, the normally 1529-1538. 1988. 22. Fisch. T. M., Prywes. R., Simon. C., and Roeder. R. G. Multiple sequence tight on-and-off control of these genes may not function, lead elements in the c-fos promoter mediate induction by cAMP. Genes & Dev., ing to their aberrant expression. 3: 198-211, 1989. 534

23. Oilman. M. Z.. Wilson, R. N., and Weinberg. R. A. Multiple prolein-binding of individual amino acids. Proc. Nati. Acad. Sci. USA. 82: 5131-5135. 1985. siles in the 5'-flanking region regulate c-fos expression. Mol. Cell. Biol.. 6: 38. Laemmli, U. K. Cleavage of structural proteins during the assembly of the 4305-4316. 1986. head of bacteriophage T4. Nature (Lond.), 227.-680-685. 1970. 24. Verma, I. M.. and Sassone-Corsi. P. Proto-oncogene fas: complex but ver 39. Kitano, T.. Go, M.. Kikkawa. U.. and Nishizuka, Y. Assay and purification satile regulation. Cell. 51: 513-514. 1987. of protein kinase C. Methods Enzymol.. 124: 349-352, 1986. 25. ChiÃ¹.R.. Boyle, W. J., Meek. J.. Smeal. T. Hunter. T.. and Karin. M. The 40. Vasuda. I., Kishimoto. A., Tanaka, S., Tominaga, M., Sakurai. A., and c-fos protein interacts with c-jun/\P-\ to stimulate transcription of AP-I Nishizuka. V. A synthetic peptide substrate for selective assay of of protein responsive genes. Cell. 54: 541-552. 1988. kinase C. Biochem. Biophys. Res. Commun.. 166: 1220-1227, 1990. 26. Miller. A. D.. Curran. T.. and Verma. I. M. c-fos protein can induce cellular 41. House, C.. and Kemp. B. E. Protein kinase C contains a pseudosubstrate transformation: a novel mechanism of activation of a cellular oncogene. Cell. prototype in its regulatory domain. Science (Washington DC). 238: 1726- 36:51-60. 1984. 1728. 1987. 27. Sikorska. M.. WhitfÃ¬cld.J. F.. and Rixon. R. H. The effects of thyroparathy- 42. Kemp. B. E.. Graves, D. J., Benjamini. E.. and Krebs. E. G. Role of multiple roidectomy and 1.25-dihydroxyvitamin D3 on changes in the activities of basis residues in determining the substrate specificity of cyclic AMP-depend- some cytoplasmic and nuclear protein kinases during liver regeneration. J. ent protein kinase. J. Biol. Chem.. 252: 4888-4894. 1977. Cell. Physiol.. 115: 297-304. 1983. 43. Sassone-Corsi. P. Cyclic-AMP induction of early adenovirus promoters in 28. MacManus, J. P. Occurrence of a low-molecular-weight calcium-binding volves sequences required for EIA trans-activation. Proc. Nati. Acad. Sci. protein in neoplastic liver. Cancer Res.. 39: 3000-3005. 1979. USA. Â«5:7192-7196. 1988. 29. Morris. H. P. Studies on the development, biochemistry and biology of 44. Gonzalez. G. A., and Montminy, M. R. Cyclic AMP stimulates somatostatin experimental hepatomas. Adv. Cancer Res., 9: 227-302. 1965. gene transcription by phosphorylation of CREB at serine 133. Cell. 59: 675- 30. Sikorska. M., and Whitfield. J. F. Isolation and purification of a new 105 680. 1989. kDa protein kinase from rat liver nuclei. Biochim. Biophys. Acta. 703: 171- 45. Macgregor. P. F., Abate. C.. and Curran. T. Direct cloning of leucine zipper proteins: Jun binds cooperatively to the CRE with CRE-BP1. Oncogene, 5: 179, 1982. 451-458. 1990. 31. Dignam. J. D.. Lebowitz. R. M., and Roeder. R. G. Accurate transcription 46. Vaniamolo. K. K.. Gonzalez. G. A.. Menzel. P.. Rivier. J.. and Montminy. initiation by RNA polymerase II in a soluble extract from isolated mamma lian nuclei. Nucleic Acids Res.. //: 1475-1489, 1983. M. R. Characterization of a bipartite activator domain in transcriptional factor CREB. Cell. 60: 611-617, 1990. 32. SivaRaman. L., Subramanian. S.. and Thimmappaya. B. Identification of a 47. Merino. A., Buckbinder. L.. Mermelstein. F. H., and Reinberg. D. Phos factor in HeLa cells specific for an upstream transcriptional control sequence phorylation of cellular proteins regulates their binding to the cAMP respon of an ElA-inducible adenovirus promoter and its relative abundance in sive element. J. Biol. Chem., 264: 21266-21276. 1989. infected and uninfcctcd cells. Proc. Nati. Acad. Sci. USA. 83: 5914-5918. 48. Butcher. F. R., Scott, D. F., Potter. V. R., and Morris. H. P. Endocrine 1986. control of cyclic adenosine 3'.5'-monophosphate levels in several Morris 33. Silva, C. M.. Tully, D. B.. Petch, L. A.. Jewell. C. M., and Cidlowski, J. A. hepatomas. Cancer Res., 32: 2135-2140, 1972. Application of a protein-blotting procedure to the study of human glucocor- 49. Reynolds. R. D.. Scott. D. F.. Potter. V. R.. and Morris. H. P. Induction of licoid receptor interactions with DNA. Proc. Nati. Acad. Sci. USA. 84: tyrosine aminotransferase and amino acid transport in Morris hepatomas 1744-1748. 1987. and in adult and neonatal rat liver. Cancer Res., 31: 1580-1589, 1971. 34. Angel. P.. Allegretto. E. A.. Okino. S. T.. Hattori. K.. Boyle. W. J.. Hunter. 50. Porsch-Hallstrom. I., Blanck. A., Eriksson. L. C.. and Gustafsson. J. A. T.. and Karin. M. Oncogene jun encodes a sequence-specific fra/is-activator Expression of the c-myc. c-fos and c-ras"a protooncogenes during sex-differ similar to AP-1. Nature (Lond.). 332: 166-171. 1988. entiated rat liver carcinogenesis in the resisant hepatocyte model. Carcino- 35. Van Beveren. C.. Van Straalen. F.. Curran. T.. Muller. R., and Verma. I. M. genesis (Lond.), 10: 1793-1800. 1989. Analysis of FBJ-MuSV provirus and c-fos (mouse) gene reveals that viral 51. Corral, M., Defer. N., Paris, B.. Raymondjean. M., Coreos, D., Tichonicky. and cellular/Ã².vgene products have different carboxy termini. Cell, 32: 1241- L., KIuh, J.. Glaise, D.. Kneip, B., and Guguen-Guillouzo, C. Isolation and 1255. 1983. characterization of complementan DNA clones for genes overexpressed in 36. Jameson. B. A., and Wolf. H. The antigenic index: a novel algorithm for chemically induced rat hepatomas. Cancer Res.. 46: 5119-5124. 1986. predicting antigenic determinants. Comput. Appi. Biosci.. 4: 181-186. 1988. 52. Ruther, 0.. Garbcr, C., Komitowski, D., Muller. D., and Wagner. E. F. 37. Houghten. R. A. General method for the rapid solid-phase synthesis of large Deregulated c-fos expression interferes with normal bone development in numbers of peptides: specificity of antigen-antibody interaction at the level transgenic mice. Nature (Lond.), 325:412-416. 1987.

535

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 1991 American Association for Cancer Research. Changes in Cyclic Adenosine Monophosphate-responsive Element Binding Proteins in Rat Hepatomas

Joanna Kwast-Welfeld, Ian de Belle, P. Roy Walker, et al.

Cancer Res 1991;51:528-535.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/51/2/528

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/51/2/528. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.