Proc. Natl. Acad. Sci. USA Vol. 86, pp. 4887-4891, July 1989 Biochemistry Regulation of transcription by cyclic AMP-dependent protein kinase (phosphorylation/chorionic gonadotropin gene/cotransfection) PAMELA L. MELLON*, CHRISTOPHER H. CLEGGt, LESLAY A. CORRELLt, AND G. STANLEY MCKNIGHTt *The Regulatory Biology Laboratory, The Salk Institute, 10010 North Torrey Pines Road, La Jolla, CA 92037; and tDepartment of Pharmacology, SJ-30, University of Washington, Seattle, WA 98195 Communicated by Edwin G. Krebs, April 3, 1989 (receivedfor review March 8, 1989) ABSTRACT cAMP-dependent protein kinase (PKA; ATP: and the cDNA encoding this protein has recently been cloned protein phosphotransferase; EC 2.7.1.37) appears to be the (14, 15). Evidence that the C subunit of the kinase may major mediator of cAMP responses in mammalian cells. We directly modify this protein includes its phosphorylation in have investigated the role of PKA subunits in the regulation of vitro by purified kinase (16). Though the holoenzyme may be specific genes in response to cAMP by cotransfection of wild- cytoplasmic, treatment of bovine epithelial cells with cAMP type or mutant subunits ofPKA together with cAMP-inducible has been shown to translocate C subunit immunoreactivity to reporter genes. Overexpression of catalytic subunit induced the nucleus which would be required for direct phosphory- expression from three cAMP-regulated promoters (a-subunit, lation ofa nuclear protein (17). In addition, an inhibitor ofthe c-fos, E1A) in the absence of elevated levels of cAMP but did C subunit kinase activity has been shown to block cAMP not affect expression from two unregulated promoters (Rous induction of the enkephalin and prolactin genes (18, 19). sarcoma virus, simian virus 40). Cotransfection ofa regulatory To further investigate the role ofthe individual subunits of subunit gene containing mutations in both cAMP binding sites the kinase in gene regulation, we have used cDNA expression strongly repressed both basal and induced expression from the vectors that encode R and C subunits to perturb the balance cAMP-responsive a-subunit promoter without affecting and function of the kinase in transfected cells. We demon- expression from the Rous sarcoma virus promoter. These strate that overexpression ofC subunit can induce expression experiments indicate that cAMP induces gene expression of several known cAMP-responsive genes, completely sub- through phosphorylation by the catalytic subunit and that the stituting for elevated cAMP. In contrast, the R subunit failed ambient degree ofphosphorylation dictates the level ofbasal as to significantly induce gene expression, supporting the con- well as induced expression of the cAMP-regulated a-subunit clusion that phosphorylation plays a key role in mediating the gene. effects ofcAMP on gene expression. We have also examined the role of the basal level of kinase activity on gene expres- In mammalian cells, cAMP is generated in response to sion in uninduced cells. R subunit expression vectors with ligand-receptor interaction at the cell surface, transducing mutated cAMP binding sites (20) blocked the cAMP response that signal by binding to the regulatory subunits (R subunits) and also lowered the basal transcription level of the cAMP- of protein kinase A (PKA; ATP:protein phosphotransferase; responsive a-subunit promoter. This latter finding demon- EC 2.7.1.37) (1). The PKA holoenzyme is an inactive tetra- strates that even in the absence of elevated cAMP, the mer of two catalytic subunits (C subunits) and two R sub- ambient phosphorylation activity of the C subunit of PKA units, and cAMP binding to the R subunits causes the release plays an important role in the maintenance of transcription of active C subunits. Many of the biological effects of cAMP levels of genes regulated by cAMP. are thought to be caused by the phosphorylation of specific substrates by the C subunits (2). MATERIALS AND METHODS cAMP can also influence gene expression. In Escherichia coli, cAMP acts as a messenger in regulatory pathways by Plasmid Constructions. Reporter plasmids with the pro- binding directly to a transcription factor, catabolite gene moter of the a subunit of the glycoprotein hormones (a activator protein (CAP), altering its ability to bind promoter promoter) contained sequences from -168 to +45 of the sequences and activate gene transcription (3, 4). In Dictyo- human gene (8) linked to the coding sequences of either the stelium, cAMP regulates cellular events, in part, by binding luciferase (a-luc) (21) or the chloramphenicol acetyltrans- to a cell surface receptor (5). The role of cAMP and PKA in ferase (a-CAT) (22) genes. Expression vector plasmids con- regulating transcription in mammalian cells is not clearly tained the mouse metallothionein 1 promoter (Mt-1) (23) defined. It has been postulated that the R subunit or some cloned onto the cDNAs for the a isoform of the mouse PKA other cAMP binding protein might play a direct role in C subunit (MtC) (24), or the mouse type I PKA R subunit transcription analogous to the cAMP binding protein of [MtR(wt)] (20) with a human growth hormone segment to bacteria (6). Alternatively, the C subunit of the kinase may provide the poly(A) site (25). A control expression vector regulate transcription through direct phosphorylation oftran- contained the same Mt-1 promoter fragment cloned onto a scription factors or activation of specific transcription factor mouse-human hybrid /3-globin gene (Mtglobin) (26). Mutant kinases. The transcriptional response of several genes to cDNAs of the RIa regulatory subunit [MtR(B), MtR(AB)] cAMP has been localized to a specific DNA sequence termed were cloned as described (20). MtR(B) contains two amino a cAMP response element (CRE) (7-9). A protein that binds acid substitutions in the site B cAMP binding site at amino to this element has been identified and characterized as a 42-kDa phosphoprotein present in many cells and tissues that Abbreviations: CAT, chloramphenicol acetyltransferase; PKA, pro- are known to respond (CREB) (10). This protein may be tein kinase A (ATP:protein phosphotransferase); R subunit, regula- involved in regulating many cAMP-responsive genes (11-13), tory subunit; C subunit, catalytic subunit; CRE, cAMP response element; CREB, cAMP-responsive element binding protein; a pro- moter, promoter of the a-subunit gene of the human glycoprotein The publication costs of this article were defrayed in part by page charge hormones; Mt-1, mouse metallothionein 1 promoter; RSV, Rous payment. This article must therefore be hereby marked "advertisement" sarcoma virus; SV40, simian virus 40; CAP, catabolite gene activator in accordance with 18 U.S.C. §1734 solely to indicate this fact. protein. 4887 Downloaded by guest on October 2, 2021 A288 Biochemistry: Mellon et al. Proc. Natl. Acad. Sci. USA 86 (1989) acid positions 324 and 332, and MtR(AB) contains an addi- tional substitution at position 200 in the site A cAMP binding Control * site. E3CAT contained sequences between -487 and +65 of Zinc Cu M adenovirus type 5 E3 gene (27) and fosCAT contained 711 base pairs of the 5' flanking DNA of human c-fos (12) and 0) were kindly provided by P. Sassone-Corsi (The Salk Insti- 0 tute, La Jolla, CA). Other plasmids have been described: RSV-.Bgal (28), RSVluc and SVluc (21), RSVCAT (22). Cell Culture and Transfections. JEG-3 cells were main- Eagle's medium supplemented tained in Dulbecco's modified Mtglobin Mtglbin MtC WtC with 10%o fetal calf serum and 4.5 mg of glucose per ml. +Fsk +Fsk Transfections were performed by using calcium phosphate precipitates (29) containing 0.1-2.5 ,ug of reporter plasmid Cotransfected expression vector DNA, 7.5-10 gg of total Mt expression vector DNA (as FIG. 1. Transcription ofthe a-subunit promoter is induced by the stated for individual experiments), and 2 ,ug of the internal C subunit of PKA. One-tenth microgram of a-luc was cotransfected control plasmid, RSV-f8gal (28). Cells were incubated for 5 hr into JEG-3 cells with 10 ,.g of MtC, a Mt-i expression vector that followed by a medium change; they were harvested 36-48 hr produces the C subunit of PKA, or Mtglobin, the same expression later. For experiments using forskolin, the drug was added vector expressing a P-globin protein. Forskolin (Fsk) was added to 16-18 hr prior to harvest, to a final concentration of 10 ,uM. increase cAMP, and zinc was added to increase expression of the C For experiments using zinc, 80 /LM zinc was added directly subunit as indicated. Values shown are the average of two identical after the medium change. experiments and are expressed normalized to the control expression Luciferase, (3-Galactosidase, and CAT Enzymatic Assays. level with Mtglobin in the absence of forskolin or zinc. Protein extracts were prepared by freeze-thawing as de- -168 to +45) was used to determine the response of a scribed (ref. 22; except harvesting buffer was 1 mM cAMP-inducible gene to excess C subunit. Cotransfection dithiothreitol/100 mM potassium phosphate) and protein with a Mt-1 promoter expression vector that produces the C concentrations were determined using Bio-Rad protein assay subunit (Ca) of PKA (MtC) induced expression of the a- reagent (8). CAT assays were performed as described (8, 22). luciferase reporter gene by 6.5-fold, an effect that was Acetylated chloramphenicol was excised from the chromato- increased to 15-fold when MtC was induced by treatment of gram and the radioactivity was quantitated by scintillation transfected cells with zinc (Fig. 1). A control expression spectroscopy. Luciferase assays were performed as de- vector that expressed the neutral protein, 83-globin, was used scribed (21). Background radioactivity was determined by to equalize any effects of the Mt-1 promoter and showed no performing luciferase or CAT assays without added extract, effect on a-luc activity.