dinsong Lius, Edwards A. Park, Austin L. Gurney$, William J. Roeslerll, and Richard W. Hanson. From the Department of Biochemistry, Case Western Reserve ...
THE IC:)
JOURNAL OF BIOLOGICAL CHEMISTRY
Vol. 266, No. 28, Issue of October 5,pp. 19095-19102, 1991 Printed in U.S.A.
1991 by The American Society for Biochemistry and Molecular Biology, lnc
Cyclic AMP Induction of Phosphoenolpyruvate Carboxykinase (GTP) Gene TranscriptionIs Mediated by Multiple Promoter Elements* (Received for publication, April 9, 1991)
dinsong Lius, EdwardsA. Park, Austin L. Gurney$, William J. Roeslerll, and RichardW. Hanson From the Department of Biochemistry, Case Western Reserve University Schoolof Medicine, Cleveland, Ohio 44106
The cytosolic form of PEPCK’ is a key enzyme in glucoThe cis elements involved in the CAMPregulation of neogenesis, which is expressed predominantly in liver, kidney transcription of the gene for phosphoenolpyruvate carboxykinase (GTP) (EC 4.1.1.32) (PEPCK) were stud- cortex, and adipose tissue. The transcription of this gene is ied by introducing a series of block mutations (10-15 stimulated by several hormones, including cAMP (Lamerset base pairs of random sequence) into eight of the protein al., 1982), glucocorticoids(Magnuson et al., 1987), and thyroid binding domains in a region of the promoter between hormone (Loose et al., 1985), whereas insulin, phorbol esters, and vanadate inhibit its expression (Granner et al., 1983; Chu -490 and +73. Each mutant promoter was ligated to the structural gene for chloramphenicol acetyltransand Granner, 1986; Bosch et al., 1990). The PEPCK promoter expression of a chimericPEPCKferase (CAT) and transfected into HepG2 cells. Tran- (-460 to +73) can direct the scription of PEPCK-CAT was stimulated 4-fold by the bovine growth hormone gene in transgenic mice in a manner addition of 8-bromo-CAMP (8-Br-CAMP), whereas similar to the endogenous PEPCK gene (McGrane et al., 1988, overexpression of the catalytic subunit of protein ki- 1990). These observations suggest that this relatively small nase A in these cells increased transcription from the region of the promoter contains the information necessary for PEPCK promoter 30-fold. Several elements within the tissue-specific and hormonally regulatedexpression of the PEPCK promoter acted synergistically to mediate thisPEPCK gene. effect. These include CRE-1 (-92 to -82) and a comDNase I footprinting analysiswith proteins prepared from plex unit from -220 to -280 composed of multiple rat liver nuclei has defined at least eight binding domains, bindingsitestermed P3(I) (-250 to-234), P3(II) termed CRE-1, CRE-2, and P1 to P6 (Roesler et al., 1989). (-260 to -250), and P4 (-286 to -270). Mutation of Recent studies from this laboratory have identified several both CRE-1 and P3(I) resulted in the complete elimi- transcription factorswhich can bind to the PEPCK promoter. nation of transcriptionalinductionbyeither8-BrNuclear factor 1 (NFl/CTF) and the hepatic nuclear factorcAMP or the catalytic subunit of protein kinase A. To 1 (HNF-1) bind to theP1 and P2 sites,respectively (Roesler examine the proteins involved in this response, we et al., 1989). The functional significance of most of the indireplaced CRE-1, which binds both C/EBP and CAMP- vidual elements in the PEPCK promoter, which were identiresponsive element binding protein (CREB), with an fied by their ability to bind nuclear proteins, remains unclear. optimal C/EBP binding sequence which significantly T h e only previously described CAMP-responsive element in decreased the binding of CREB, but maintained the the PEPCK promoter maps between -94 and -72 (CRE-1). affinity for C/EBP. Transcription from this modified Thiselement wasidentified usingserialdeletions of the promoterwas induced by 8-Br-CAMP and the catalytic PEPCK promoter linked to aselectable marker gene, neo subunit of protein kinase A (PKA) to a similar extent (Short et al., 1986). The CRE-1 region is important for both as noted with the native PEPCK promoter. However, basal and cAMP induced transcription (Short et al., 1986; the results of experiments involving cotransfectionof Quinn et al., 1988; Bokar et al., 1988). The transcription factor PEPCK-CAT with expression vectors for PKA and C/EBP, which binds to promoters from a variety of genes of either C/EBP or CREB suggest that CREB is capable metabolic importance, hasbeen shown by DNase I footprintof mediating a greater responsiveness to PKA than ing analysis to bind to CRE-1 and to other regions of the C/EBP. Our results indicate that multiple cis elements PEPCK promoter at -250 to -234 (P3(I)) and at -286 to a r e involved in the CAMP induction of PEPCK gene -270 (P4). C/EBP also induces transcription of a chimeric transcription and that C/EBP and CREB are poten- PEPCK-CAT gene when it is introduced into hepatomacells tially involved in this response. (Park et al., 1990). The CAMP-responsive element binding protein(CREB)bindstotheCRE-1siteinthePEPCK promoter (Park et al., 1990). CREB has been implicated in the cAMP inductionof transcription of the somatostatingene * This work was supported in part by Grants DK 21859 and DK (Montminy et al., 1986; Yamamoto et al., 1988). Since both C/EBP and CREB bind to CRE-1, it is possible that either 24951 from National Institutes of Health. The costsof publication of this article were defrayed in part by the payment of page charges. or both of these proteins play a role in the cAMP induction This article must therefore be herebymarked “advertisement” in of PEPCK gene transcription. accordance with 18 U.S.C. Section 1734 solely to indicate thisfact. The intracellulareffect of CAMPis mediatedvia the CAMP$ Current address: Howard Hughes Medical Institute and Dept.of dependentkinase A (PKA)(Edelman et al., 1987; Taylor, Biological Chemistry, University of Michigan Medical Center, Ann 1989), which is composed of two regulatory and two catalytic Arbor, MI 48109-0650.
5 Trainee supported by the Metabolism Training Program Grant DK-07319 from the National Institutesof Health. ll Current address: Dept. of Biochemistry, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N OWO.
I The abbreviations used are: PEPCK, phosphoenolpyruvate carboxykinase; CAT, chloramphenicol acetyltransferase; PKA, protein kinase A; 8-Br-CAMP, 8-bromo-CAMP.
19095
19096
CAMPInduces PEPCK Transcription through
(C) subunits. Klemm et al. (1990) have shown, using a cellfree in vitro transcription system, that the C subunit of PKA caused a marked induction of transcription from the segments of the PEPCK promoter as short as -109.If CRE-1 was removed from the promoter by deletion to -68, instead of stimulating transcription, the C subunit of PKA caused a slight inhibition of transcription from the PEPCKpromoter. Cotransfection of the C subunit of PKA with chimeric genes has also been shown to induce the transcription of several CAMP-responsivegenes (Mellon et al., 1989). Here we report that cotransfection of an expression vector for the C subunit of PKA together withPEPCK-CAT vectors into HepG2 cells increases transcription from the PEPCK promoter 30-40-fold. Using this transfection system together with the PEPCK promoter containing a series of block mutations in specific protein binding domains, we determined that multiple sequence elements are involved in the cAMP regulation of gene transcription. Our resultsindicate that CRE-1, P3(I), and to a lesser extent, P3(II) and P4, are of importance in conferring cAMP responsiveness on the PEPCK promoter. EXPERIMENTALPROCEDURES
Materials DNA-modifying enzymes,poly[d(I.C) .d(I. C)], 8-bromo-CAMP (8Br-CAMP) were purchased from Boehringer Mannheim. [y””P] ATP (6000 Ci/mmol) and [1,2-’4C]chloramphenicol (50 mCi/mmol) were from Du Pont-New EnglandNuclear. Tissue culturemedia, sera, and supplies were from GIBCO. The cell line HepG2 was from American Type Culture Collection. Oligonucleotides were chemically synthesized using an Applied Biosystems 380A DNA Synthesizer. Escherichia coli strain NM522 (dut’ ung+), CJ236 (dut- ung-), and pTZ vectors (Mead et al., 1986) were from the laboratoryof Norman Pace, Indiana University, Bloomington, IN. The RSV-(%gal plasmid wasagenerous gift of Chen-ming Fan, Harvard University. The expression vector for the catalytic subunit of CAMP-dependent protein kinase A was kindly provided by Dr. Muramatsu, DNAX, Palo Alto, CA. The cDNA for CREB was a gift from Dr. M. Montminy, Salk Institute, SanDiego, CA and the cDNA for C/EBP was provided by Dr. Steven McKnight, Carnegie Institution, Baltimore, MD. All DNA-modifying enzymeswere from Boehringer Mannheim. Methods Construction of PEPCK-CAT-A 563-base pair XbaI-BglII fragment of the PEPCK promoter regulatory region, which was isolated from the plasmid BH1.2 (Wynshaw-Boris et al., 1984) was ligated into the XbaI-BglII site of the polylinker from poly-CAT. Poly-CAT is a derivative of pSV,CAT (Silver et al., 1987), exceptthat the XbaIHindIII site was replaced by a polylinker made with two complementary synthetic oligonucleotides, 5’CTAGACCCGGGATCGATAGAT CTAAGCT 3’ and 5’AGCTTAGATCTATCGATCCCGGGT3’. This polylinker contains XbaI, SmaI, ClaI, BglII, and Hind111 restriction sites. The resulting plasmid was digested with XbaI and PstI restriction enzymes, and the 2.5-kilobase pair XbaI-PstI fragment containing thewhole CAT structural gene and SV40 polyadenylation signal from the poly-CATwasligated into the XbaI-PstI sites of pTZ18R (Mead etal., 1986). Specific block mutations were introduced into the PEPCK promoter using a modification of the Kunkel method (Kunkel,1985) for site-directed mutagenesis (Liu et al., 1990). The individual oligonucleotides containing the mismatched sequencewere synthesized and used as the primersfor the synthesis of double-stranded DNA using uracil-containing single-stranded DNA as template. The DNA synthesized in vitro was transformed into wild type E. coli NM522. The plasmid DNA was isolated and digested with the appropriateenzyme, since the mismatchednucleotides in all of the primers encode a new restriction site. All of the positive clones were confirmed by dideoxy sequencing. Serial deletions from the 5’ end of the PEPCK promoter were generated withBa131 byShort etal. (1986). These promoter deletions were digested with XbaI-BglII and ligated into the XbaI and BglII sites of PEPCK-CAT vector after the wild type promoter (-490 to
Multiple Promoter Elements
+73) was removed with XbaI and BglII digestion (Park et al., 1990). Plasmids containing two mutations in CRE-1 and P3(I), CRE-1 and P3(II), and CRE-1 and P4 in the PEPCK promoter were constructed using the unique SauI restriction site a t -208. The XbaISauI fragment (-490 to -208) of thePEPCKpromoter, which contains either P3(I),P3(II), or P4block mutation, respectively, was ligated into the XbaI-Saul site of the PEPCK-CATwhich contains a block mutation in CRE-1. The pXSVlCATvector which consists of the SV40 early promoter linked to the CAT structural gene has beendescribedpreviously (Bokar et al., 1988). Oligonucleotides containing PEPCK promoter sequences -94/-77 (CRE-1) and -251/-234 (P3) were synthesized to containXbaI compatible ends, and two copies of these nucleotides were ligated into the XbaI siteof pXSVlCAT (Bokaret al., 1988). Construction of Expression Vectors for CREB and C/EBP-The mammalian expressionvector for CREB was created by removing the entire CREB cDNA (Gonzalez et al., 1989)by SmaI and BamHI digestion. This fragment was ligated in front of the RSV promoter which was derived from a RSV-CAT vector (Silver et al., 1987) that had been digested with HindIII andNcoI to remove the CAT coding sequences. Construction of themammalian expressionvectorfor C/EBP has been described (Landschultz et al., 1989). Construction of the plasmid for the expression of CREB and the production of CREB in E. coli by the two vector system of Tabor and Richardson (1987) has been described (Park et al., 1990). Recombinant C/EBP was purified after overexpression in E. coli (Landshultz et al., 1989). Cell Culture,DNATransfection, and the Determination of CAT Activity-HepG2 cells were grown in 10-cm platesin Dulbecco’s minimal essential medium, containing 5% fetal calf serum and 5% calf serum. Cells were grown to 60-70% confluence, treated with trypsin, centrifuged, and resuspended in 2 ml of complete medium. Ten pg of PEPCK-CAT, 5 pg of RSV-P-gal, and 10 mg of the expression vector for the C subunit of PKA were precipitated using calcium phosphate (Park et al., 1990). Two ml of the calcium phosphate-DNA precipitate was mixed thoroughly with the 2 ml of cells, and the mixtureof cells with DNAwas then transferred to two plates containing 10 ml of complete medium and incubatedfor 36 h. 8-BrcAMP was added for 12 h prior to harvesting the cells. No 8-BrcAMP was added to the cells cotransfected with the C subunit of PKA. The cell extracts containing CAT were prepared as described previously (Park et al., 1990). Theproteinconcentrationinthe extracts was determined by the method of Bradford (1976). Aquarter of the extract wassaved for measurement of @-galactosidase activity (Miller, 1972) to correct for transfection efficiency. The percentage of [’4C]chloramphenicol in the acetylated form was normalized for variation in transfectionefficiency by dividing bythe @-galactosidase activity measured in eachcell extract. Footprinting Analysis-The DNase I footprintinganalysis was carried out using a probe of a 513-base pair segment of the PEPCK promoter prepared from PEPCK-CAT (Fig. 1).The DNA fragment was labeled at the XbaI site a t -490 with [y-””PIATP and T4 DNA polynucleotide kinase as described previously (Roesler et al., 1989). Thesamerestrictionsite wasused for thepromotercontaining individualmutants.Nuclearproteins were isolatedfrom rat liver nuclei as described previously (Roesler et al., 1989). RESULTS
Introduction of Block Mutations into Specific Regionsof the PEPCK Gene-Site-directed mutagenesis was used to introduce mutations directly into those regions of the PEPCK promoter known to bind proteins isolated from rat liver nuclei (Fig. 1). Adetailed description of the technique used to introduce the block mutations which are examined in this report has been presented in a previous publication (Liu et al., 1990). We analyzed the ability of proteins isolated from rat liver nuclei to bind to the PEPCK promoter containing block mutations by footprint analysis. These experiments were designed to ensure that binding at the specific site was eliminated and to show that the mutationsdid not introduce an unexpected protein bindingsite in the promoter. Mutation of nucleotides at CRE-1, CRE-2, P3(I), andP4 disrupted the interaction of nuclear proteins from rat liver with each of these regions (Fig. 2). The other mutationsintroduced a t P1, P2, P3(11), P5, and P6 also disrupted binding of proteins to
CAMPInduces PEPCK Transcription through Multiple Promoter Elements GRU -
D
k z
Xbal P6
-490
P5 -400
-
c o
N
I
m
: : u
-
x
o
m
I
5
: Yu
L
P4 P3(111P3(1) P2 -300
Y
o
m
: : u CRE2
-200
l. A schematic diagram
Of
e
c
3
x
k?m
u
G
z
et:: U L U
Pl CREl -100
8g1,1
TRTA
+1
r73
PEPCK-CAT and the pro-
tein binding domains in the PEPCK promoter. Schematic dia-
gram of protein binding domains in the PEPCK promter from -490 t.0 +73 defined by the XbaI and BglII sites. Each protein binding domain from P1 to P6. CRE-1. and CRE-2 is shown at thebottom of the promoter and the corresponding transcription factors which hind t o these regions are outlined above. The region between -455 and -350, which was shown to be involved in glucocorticoid responsiveness, is indicated by GRU (glucocorticoid-responsive unit) (Imai et nl., 1990).
19097
their corresponding binding PEPCK the sites in promoter (data not shown). The Effect of the Catalytic Subunit of PKA on Transcription from the PEPCK Promoter Deletions-We first transfected thePEPCK-CATvectorintoHepG2 cells,aliver cell line derived from a humanhepatoblastoma which has been used to studyof variety a liver-specificfunctions (Aden et al., 1979; Javitt, 1990), and noted that 8-Br-cAMP induced the transcription of PEPCK-CAT gene about 4-fold but had no effect on the RSV-CAT gene (data not shown). Transcription of a Dromoterless Dlasmid (18R-CAT) was also included as a con'trol for nonsiecific CAT activity, and no CAT activity was observed. The level of induction of transcription from the PEPCK promoter noted above (about 4-fold) was not sufficient to analyze the role of each of the putative regulatory elements in the PEPCK promoter. was It possible that either the concentrationof CAMP-dependent PKA or the transcription factors mediating theeffect of PKA in HepG2 cells was limiting. We next cotransfected expression an vector encoding the C subunit of PKA, together with the PEPCK-CATgene, which increasedtranscriptionfromthePEPCKpromoter approximately 30-fold. This effect of the C subunit of PKA was specific for the PEPCK promoter, since itdid not affect transcription from the RSV promoter (data not shown). We analyzed the 8-Br-CAMP and PKA responsiveness of serial deletions in the5"flanking region of the PEPCK promoter (Fig. 3). Removal of the sequences between -355 and -210 resulted in a substantial loss of both CAMP and theC subunit of PKAinduction of PEPCKtranscription.This region contains the protein binding sites termed P3(I), P3(II), and P4 (Fig. l.4). The remaining induction of transcription was retainedby the PEPCK promoter containing -174, -134, and -109 deletions, butwas lost with the promoter containing a deletion to -68. The region of the PEPCK promoter between -109 and -68 contains the protein binding site termed CRE-1. Multiple Elements AreInvolved in the Inductionby PKA of Transcription from the PEPCK Promoter-Analysis of the serial deletions of the PEPCK promoter demonstrated that
t
c
,440
i
.355
.?i' .??c
,176
.134
.lo3
.68
FIG.3. Transcriptional induction by 8-Br-CAMP or the C subunit of PKA of the PEPCK-CAT gene containing serial deletions. HepG2 cells were transfected with 5 pg of PEPCK-CAT
and 2.5 pg of RSV-a-gal plasmids. After 36 h, 1 mM 8-Br-CAMP was added for 12 h. The cells were also transfected with 5 pg of PEPCKCAT and5 pg of SR-PKA andRSV-6-gal. No 8-Rr-CAMPwas added tothetransfected cells when the expressionvector for catalytic FIG.2. Footprint analysisof block mutations inthe PEPCK subunit of PKA was added. The expression vector for the C subunit of PKA contains the SRn promoter to direct the expression of the promoter. The XbaI-Rg/II fragment of the PEPCK promoter(-490 to +7B) was isolatedfrom PEPCK-CATandend-labeledonthe open reading frame for the C subunit (Muramatsu et al., 1989). The noncoding strand a t the XbaI site. The mutant PEPCK promoters cells were harvested, and CAT activity was measured as described were identically end-laheled. In the left lane is the DNaseI digestion under "Experimental Procedures." The results shown are corrected in the absense of proteins. The remaining lanes show the DNase I for@-galactosidaseexpressionfrom the RSV-@-gal. Shown in the digestion pattern in the presence of rat liver nuclear proteins for the figure are average values from two to five independent transfection wild type and mutant forms of the PEPCK promoter. The correspondexperiments. Open bar, no treatment, striped bar, 8-Rr-CAMP added, solid bar, cotransfected with the C subunit of PKA. ing binding sites are outlined by the boxes on the right.
CAMPInduces PEPCK Transcription through
19098
upstream elements are required for the induction of transcription by C subunit of PKA. To further define these elements, chimeric PEPCK-CAT genes containing individual mutations in the PEPCKpromoter at specific protein binding domains were transfected into HepG2 cells and tested for their transcriptional responsiveness to 8-Br-CAMP or the C subunit of PKA (Fig. 4). Mutations in regions P1, CRE-2, P2,and P4of the PEPCK promoter had no effect on the stimulation of transcription by 8-Br-CAMP. Disruptionof CRE-1 caused a greater reduction in CAMP-stimulated transcription from the PEPCK promoter than a block mutation in P3(I). The PEPCK promoter containing a mutation in CRE-1 could be induced %fold bythe C subunitof PKA, as compared with the 30-fold induction noted with the intact promoter (Fig. 4). Mutations in P3(I), P3(II), or P4 resulted in a 3- or 4-fold induction of transcription, respectively, from the PEPCK promoter by the C subunit of PKA. A mutation at P1, a binding site for NF-1/CTF, although adjacent to CRE1, did not affect the induction of transcription from the PEPCK promoter by the C subunit of PKA. Finally, a mutation in P2, a binding site for HNF-1, or in CRE-2, a weak binding site for C/EBP, had no significant effect on the induction of transcription from the PEPCK promoter by the C subunit of PKA. Since no single mutation completely abolished induction by cAMP or the C subunit of PKA, we created a series of double mutations at key protein binding sites within the PEPCK promoter (Fig. 4). The chimeric PEPCK-CAT gene containing a mutation in both CRE-1 and P3(I) had a low level of basal transcription and was completely unresponsive to stimulation by 8-Br-CAMPor the C subunit of PKA. The PEPCK promoters containing double mutations in both P3(II) and CRE-1or in P4 and CRE-1were induced 2-3-fold by 8-Br-CAMPor the catalytic subunit of PKA. These results suggest that cAMP responsiveness is mediated by both CRE1 and P3(I),with some additional involvement of P3(II) and P4. It is also possible that the effects of P4 and P3(II) are mediated through CRE-1, since no additional inhibition was observed with these double mutations. CRE-1 but Not P3 Can Confer 8-Br-CAMPResponsiveness to an SV40 Promoter in Jeg-3 Cells-To examine whether P3(I) or CRE-1 functioned as CAMP-responsive elements
35
Multiple Promoter Elements
when linked to a neutral promoter, two copies of the core CRE-1 sequence or two copies of P3 were ligated toan enhancerless SV40 promoter (Park etal., 1990).Transcription from the SV40 promoter in the presence of 8-Br-CAMPwas analyzed in bothHepG2 and Jeg-3 cells. Jeg-3 cells were used since they are very CAMP-responsive when transfected with a variety of CAMP-regulated genes (Deutsch et al., 1988). Both chimeric genes responded very poorly to 8-Br-CAMP when they were transfected into HepG2 cells (Fig. 5). However, transcription from a chimeric gene containing two copies of CRE-1 linked to the SV40 promoter and introduced into Jeg-3 cells was induced 30-fold by 8-Br-CAMP. In contrast, 8-Br-CAMPcaused only a %fold stimulation of transcription in these cells when two copies of P3(I) were included in the promoter. Thus, P3(I) must be in an appropriatecontext within thePEPCK promoter to contribute to the cAMP regulation of PEPCK transcription. An Optimal CIEBP Binding Site Can Function as a CRE when Introduced into the PEPCK Promoter-Since both C/ EBP and CREB can bind to CRE-1, it was not clear which protein bound to this element in uiuo, to mediate the cAMP induction of PEPCK gene transcription. We replaced the entire core CRE-1 (CTTACGTCAG) with an optimal C/EBP binding sequence (ATTGCGCAAT). Fig. 6 shows DNase I A. Oligomer SV1
1.
CAT
"----b
CAT
_____._I)
2. CRE-1x2
3. P3x2
SV1
~h
B.
c
CREl SV1 WT
~ ~ 5 . 1P ?
CRE2
P2
P31lI
P311lI
P1
CREl P1
CRE 1 P3(11l
CRE 1 P3lll
FIG. 4. The effect of 8-Br-CAMPor the C subunit of PKA on the induction of transcription from the PEPCK promoter containing block mutations. A, HepGP cells were transfected with PEPCK-CAT vectors containing specific mutations in the protein binding domains of the promoter. The cells were treated with 8-BrcAMP (striped bars) or cotransfected with a PKA expression vector (solidbars) as described in Fig. 3. Basal level of transcription is indicated by open bars. The site which was mutated is indicated on the horizontal axis.
P3
HepGP
Svi
CREl
P3
Jeg-3
FIG. 5 . Response of CRE-1 or P 3 elements ligated the SV40 promoter to 8-Br-CAMPin HepG2 or Jeq-3 cells. A, two copies of the CRE-1 or P3 sequence were ligated to the SV1 enhancerless promoter driving the CAT gene. B , 5 pg of each plasmid along with RSV-(3-galwas transfected into HepG2 cells. The cells were treated for 16 h with 8-Br-CAMP(solid bars) and the CAT activity analyzed. The vectors containing specific mutations in the PEPCK promoter, together with the cell lines used are indicated on the horizontal axis.
cAMP Induces PEPCK Transcription through Multiple Promoter Elements Wild Type
CRE1-bCIEBP
" "
2 1 4 1 6 1 4 1 6 2 1 4 1 6 1 4 16
u
3
p3(')
P4
" " " " " " "
"""-
~-
The inductionof transcription from this modified promoter was analyzed in thepresence of 8-Br-CAMP or after cotransfection with the expression vector containing the C subunit of PKA (Fig. 7). The degree of transcriptional induction from this modified form of the PEPCK promoter was about same as that noted with the native promoter. Surprisingly, the consensus CRE-1 is not absolutely required for the cAMP responsiveness of the PEPCK promoter. This finding suggested that C/EBP-like proteins may mediate the induction of PEPCK gene transcription causedby CAMP. Expression Vectors for C/EBP but Not CREB Stimulate Basal Transcription from the PEPCKPromoter-To further analyze the role of C/EBP and CREB in basal and cAMP induction of PEPCK gene transcription, theexpression vector for each protein was cotransfected into HepG2 cells with PEPCK-CAT. Fig. 8 shows the activity of CAT transcribed from modifed forms of the PEPCK promoter in the presence and absence of an expression vector for C/EBP and in the presence of 8-Br-CAMP. Cotransfectionof an expression vector for C/EBP with the PEPCK-CATvector resulted in a 4fold induction of transcription. The addition of 8-Br-CAMP increasedthestimulation of PEPCK-CAT only 3-4-fold above the level resulting from introducing C/EBP into the cells. The sequences at P3(I) and CRE-1 each contributed equally to the activation of transcription from the PEPCK promoter caused by C/EBP. When both 8-Br-CAMP andC/ E B P were present together, transcription from the PEPCK promoter containing the CRE-1block mutation was induced 4-fold, whereastranscriptionfromthepromoter with the P3(I) mutation was increased 6-fold. The PEPCK promoter, containing a double mutation at the CRE-1 and P3(I) hada low level of basal transcription andwas completelyunresponsive to stimulationby 8-Br-cAMP, C/EBP, ora combination of both. Similar experiments with an expression vector for CREB did not indicate effect an of CREB on basal expression or significant increase inresponsiveness to 8 Br-CAMP (data not shown). However, because our previous work suggested that protein kinase A might belimiting, we cotransfected eithertheC/EBPorCREB expressionvectorswith the expression vector for the C subunit of PKA. As can be seen
" "
35 "
~
"
~
0
"
"
19099
1
cAMP
subunit Catalytic
-
FIG. 6. Footprinting analysis of the intact PEPCK promoter and the promoter with a C/EBP site replacement at the CRE1. The XbaI-&/I1 fragment from the wild type PEPCK promoter sequence from -490 to +.in and the PEPCK promoter containing the o f CRE-1 replaced by an optimal C/EBP sequence were isolated and end-labeled at the XbaI site as described in Fig. 2. The footprinting analysis was carried out using recombinant C/EBP and CRER. The hinding sites for C/ERP and CREB areoutlined in boxes.
footprinting analysisof the native PEPCK promoter and the promoter with a C/EBP site replacing CRE-1. The four C/ E B P binding sites within the PEPCK promoter are shown by boxes a t CRE-1, CRE-2, P3(I), and P4. The major site protected by CREB was CRE-1, and there was weak protection WT CRE-1 WT CRE-1 at P3(II), which contains a near consensus AP-1 binding site. c C/EBP When CRE-1 was replaced by the optimal C/EBP binding CEBP sequence, thebindingaffinity of CREBtothe modified FIG. 7. Transcriptional induction by 8-Br-CAMP or the C PEPCK promoter was markedly decreased, and only weak subunit of PKA of the PEPCK-CAT gene following replacesite. The DNase I protection at the newly introduced C/EBP optimal ment of CRE-1 withan optimalC/EBPbinding relative level of induction of transcription from this modified prosequence was attained at the highest CREB concentration moter is outlined. HepG2 cells were transfected with PEPCK-CAT used. The binding affinityof C/EBP for the optimal C/EBP vectors in which the binding sites had beenswitched. Cells were binding site in themodified PEPCK promoter was similar to treated with 8-Br-CAMP (stripedbars) or cotransfected with the that observed with the native promoter. catalytic subunit of PKA (solid bars) as described in Fig. 3.
+
CAMPInduces PEPCK Transcription through Multiple Promoter Elements
19100
-
14 1 2
DISCUSSION C/EBP cAMP CIEBPIcAMI
1 0 C
.-
0
r ; 8 3 U
C
6
2 0
L
4 2
CC P-R4R 3E9(E1-0-1)1 I p3(1)
FIG. 8. The effect of overexpression of C/EBP on the cAMP responsiveness of the PEPCK-CAT vectors. A , wild type or mutant PEPCK-CATvectors were transfected into HepG2 cells with expression vectors for C/EBP or a mutantC/EBP (12V) in the presence or in the absence of CAMP. Each transfection contained 5 pg of PEPCK-CAT and 5 pg of C/EBP vector. 1 mM 8-Br-CAMPwas added for 16 h. The results are average of at least three independent experiments. Specific mutations in the PEPCK promoter are indicated on the horizontal axis.
8 0
-
70
-
60
-
50
-
u
-
40
-
2
30
-
20
-
1 0
-
.-
c
0 3
C
0
L
11
-
z
n
m
r
?j
m W
Lt
o
FIG. 9. Effect of cotransfection of the C subunit of PKA with expression vectors for CREB and C/EBP on the induction of PEPCK-CAT. HepG2 cells were transfected with 5 eg of PEPCKCAT, 5 pg of RSV-CREB or MSV-C/EBP, 5 pg SR-PKA, and 2.5 pg of RSV-0-gal. Cells were harvested after 40 h and CAT assays performed as described in the legend to Fig. 3.
in Fig. 9, the C subunit of PKA caused a 40-fold stimulation in the presense of C/EBP. However CREB increased transcription from the PEPCK promoter 80-fold in the presence of C subunitof PKA. Taken together, our results suggest that cAMP induction of PEPCK gene transcription could involve the interaction of both C/EBP and CREB.
The acute transcriptional responsiveness of the PEPCK gene to both positive and negative regulation by hormones is a unique aspect of the rapid control of hepatic gluconeogenesis. Considering the number of hormones and other effectors known to regulate PEPCK gene expression in the liver (Tilghman et al., 1976; Liu and Hanson, 1991), it is not surprising that the PEPCK promoter exhibits ahigh degree of complexity. These hormones interact with each other via multiple regulatory elements contained within a relatively small segment of the PEPCK promoter. For example, the region between -455 and -350, which is required for the glucocorticoid responsiveness of the PEPCK promoter, contains two glucocorticoid receptor binding sitesas well as two additional binding sites for accessory factors(Imai et al., 1990). In addition, a sequence in the PEPCK promoter between -415 and -405, which corresponds to one of the glucocorticoid binding domains mentioned above,was also shown to be necessary for part of the inhibitory effect of insulin on transcription from the PEPCK promoter (O’Brien et al., 1990). In this paper, we used a series of block mutations in specific protein binding domains in the PEPCK promoter to determine their functional significance within an intact segment of the promoter and show that multiple elements in the promoter are involved in the basal and inducible transcription of the PEPCK promoter by CAMP.Elsewhere, we have demonstrated that cAMP canact synergistically with thyroid hormone (T3) tostimulate PEPCK gene transcription (Giralt et al.). The thyroid hormone receptor binds at -322 to -308 within the PEPCK promoter and functions in a synergistic manner with P3(I), but notwith CRE-1. On the other hand, removal of the glucocorticoid-responsive domain by deletion at -355 or block mutations in the P6 and P5 region, which correspond to thebinding sites for glucocorticoidreceptor and accessory factor (AF) (Imai et al., 1990), did not affect induction of transcription by 8-Br-CAMPor the C subunitof PKA. This suggests that unlike T3, which shares a common regulatory element with CAMP, glucocorticoids and cAMP use entirely different elements in the PEPCK promoter to mediate their effects. One surprising resultfrom this study is the extent towhich the Csubunit of PKA-stimulatedtranscription from the PEPCK promoter when it was cotransfected into HepG2 cells with the PEPCK-CAT chimeric gene. It is likely that the concentration of free C subunit produced by cotransfection was significantly higher than the level of free subunit produced by cAMP treatment and that the transcription factorb) that mediates the cAMP effect was not limiting in HepG2 cells. Cotransfection of the expression vector for the C subunit of PKA has also been shown to stimulate transcription from the E3 promoter, c-fos promoter, and a-subunit promoter (Mellon et al., 1989). The large magnititude of transcriptional induction from the PEPCK promoter caused by the cotransfection of a PKA expression vector allowed us to define the role of the upstream elements, as well as CRE-1, in the cAMP responsiveness of the PEPCKpromoter. The most significant difference in the effect of 8-Br-CAMP and the expression of the C subunit of PKA was noted with the P4block mutation. It is possible that the C subunit of PKA induced the expression of an additional factor which binds to thissite to stimulate transcription from the PEPCK promoter. CRE-1 was among thefirst CAMP-responsive elements identified (Short et al., 1986; Bokar et al., 1988). However, this sequence alone was not sufficient for the full induction M. Giralt, E. A. Park, J. Liu, A. L. Gurney, and R. W. Hanson, J. Biol. Chem., in press, November issue.
CAMP Induces PEPCK Transcription through Multiple Promoter Elements
19101
by CAMP. When theregion from -109 to -68 was ligated to This suggests that C/EBP or a C/EBP-like protein from rat a neutral promoter and introduced into FTO-2B cells, the liver nuclei interacts with the PEPCK promoter. To date, the response to dibutyryl cAMP was much less robust than when only proteins which are known to bind to the P3(I) site are a larger segment from -416 t o -68 was utilized (Short et al., C/EBP or members of the C/EBP family of transcription 1986). It was apparent from these studies that other elements,factors. Since CREB did not bind to the P3(I) element,some 5‘ to CRE-1, were required for the appropriate level of tran- member of the C/EBP family is likely to be involved in the cAMP response.However,since CRE-1caninteract with scription from the PEPCK promoter. both C/EBP and CREB, the factorwhich binds to CRE-1in Many different genes require multiple elements for complete hormonalresponsiveness. Initial characterizationof sev- vivo is not clear. The results of the “binding site switch” eral CAMP-responsive genes identified the sequence experiment described in this paper suggest the involvement TGACGTCA as a CAMP-inducible enhancer sequence, since of C/EBP in the cAMP response. A modified form of the PEPCK promoter which has a very poor affinity for CREB a n oligonucleotide containing this consensus sequence conferred cAMP responsivness to a neutral promoter (Montminy but two high affinity C/EBP sites, was equally responsive to our experimentscotransfecting et al., 1986; Silver et al., 1987; Bokar et al., 1988; Sassone- 8-Br-CAMP.Incontrast, Corsi et al., 1988). Later studies demonstrated that the nucle-CREB and PKA suggested that a greater responsiveness to cAMP canbe obtained with CREB.Of course, more than one otides flanking this consensus sequence also affect the responsiveness to cAMP (Deutsch et al., 1988; Bokar et al., factor may be able to transmit the response to cAMP at the 1988). This consensussequence is not theonly element which CRE-1 siteof the PEPCK promoter. can mediate the effect of CAMP. AP-2, an enhancer binding It must be emphasized that our results are preliminary of a familyof transcripprotein which binds to the promoter of the metallothionein since C/EBP and CREB are members tionfactors which can form homo- and heterodimers and IIA gene, is responsiblefor boththe effect of cAMP and protein kinase C on transcription (Imagawaet al., 1987). The which are known to bind to similar sitesin the promoters of or C)TCA, which interacts with specific genes (Descombes et al., 1990; Poli et al., 1990; Hai et binding site for AP-1, TGA(G or liver the family of J u n proteins, has been reported to mediate both al., 1989). One of these transcription factors, C/EBPP activator protein (Decombes et al.,1990; Cao et al., 1991), also the effect of cAMP and phorbol ester tumor promoters (TPA) in Jeg-3 and HepG2 cells (Deutsch et al., 1988). The promoter binds to the same sites on the PEPCK promoter as C/EBP for the c-fos gene contains four elements that mediate cAMPand can stimulate the transcription of a PEPCK-CAT chiresponsiveness (Fischet al., 1989; Berkowitz et al., 1989), and meric gene promoter when it is introduced into hepatoma only one of these elements hasbeen shown to be a consensus cells.” A full understanding of the mechanism of action of gene transcription will requirecareful CRE sequence; others show homology with binding sequence cAMPonPEPCK analysis of both the C/EBP and CREB families of transcripAP-1 or AP-4 in the SV40 promoter (Angel et al., 1987; Lee et al., 1987; Mermod et al., 1988). The elements involved in tion factors for their relative binding affinity, bindingspecithe responsiveness to cAMP in the proenkephalingene con- ficity, and tissue distribution. The synergistic effect of cAMP on transcription from the tain overlapping bindingsitesfor a t leastfourfactors, PEPCK promoter is probably achieved through a complicated ENKTF-1, AP-1, AP-4, and AP-2, and these elements act interaction of factors at several levels. Thefirst level of synergistically to regulate CAMP-inducible transcription (Comb et al., 1988; Hyman et al., 1989). We have shown here interaction is within the enhancers themselves. CRE-1 can bind several transcription factors, which may compete for thatthePEPCKpromotercontains tworegions that are binding in response to physiologicalsignals.On the other required for induction of transcription by CAMP. Although CRE-1 and P3 are involved in the cAMP respon- hand, the upstream region of the PEPCK promoter is comsiveness of the intact promoter, these elements show differ- posed of several adjacent and overlapping sites (P3(I),PS(I1) ences in functionwhen two copiesof either element are linked and P4), and protein-protein interactions arevery likely int o a neutral promoter. Since neither one is sufficient for full volved within these elements.T h e second level of interaction the downstream CRE-1 and upstream transcriptionalinduction by cAMPinHepG2 cells, these mayoccurbetween element(s) either through cooperative binding to the DNA or elements must interact within the context of the PEPCK or simultaneous modulationof the protein-protein interaction promoter to transmit their transcriptional effect. It is not basal transcription factors at the TATAbox. The third level clear why in Jeg-3 cells two copies of CRE-1 ligated to the of interaction isbetween elements involved in cAMP responSV40 promoter are induced 30-fold by 8-Br-cAMP, and the siveness and other hormones such as T3 or factors involved same degree of stimulation is not observed with the HepG2 in the tissue-specificexpression of the PEPCK gene. The cellline. Most probably, CREB, which ispresent inhigh specificity of these interactions may require the maintenance concentrations in Jeg-3 cells, binds to the CRE-1 sequence of both the spacing and alignment of elements in the PEPCK causing a markedinduction of transcription.Theintact promoter. PEPCKpromoterresponds verypoorly to transcriptional It is clear that thecomplexity of transcriptional regulation induction by cAMP when transfected into Jeg-3 cells (data of the PEPCKgene requires a critical evaluationof promoter not shown), suggesting that additional liver-specific factors elements in the context of the physiological effectors known are required for the full induction of transcription from the to control transcriptionof the gene i n vivo. We have recently PEPCK promoterby CAMP. demonstrated that the PEPCK promoter, containing a block Is C/EBP Involved in the Induction of Transcription by mutation in P3(I), linked to the bGH structural gene and CAMP?-C/EBP activatedtranscription from the PEPCK introduced into thegerm line of transgenic mice is expressed promoter by interacting at CRE-1 and P3(I) (Trus et al., 1990; predominantly in the kidney rather than the liver and has a Park et al., 1990). The presence of CRE-1 and P3(I) in the blunted response to dietary and hormonal regulati~n.~ This PEPCK promoter was required for the full trans-activation reinforces the conclusions of the present manuscript impliof gene transcription by both C/EBP and 8-Br-CAMP. A protein purified from rat liver nuclei by CRE-1 affinity chroE. A. Park, A. L. Gurney, and R. W. Hanson, unpublished matographybehaved likepurified recombinantC/EBPin observations. titration and competition studies with the PEPCK promoter. Y. M. Patel, J. Liu, and R. W. Hanson, unpublished observations.
19102
CAMPInduces PEPCK Transcription through Multiple Promoter Elements
cating P3(I) as importantfor the regulated expression of the PEPCK gene in the liver. Further studies relating our observations on the interactions between elements in the PEPCK promoter are currently being tested in transgenic mice containing a variety of chimeric genes with specific mutations in the PEPCK promoter to determine the physiological relevance of each of these elements in controlling PEPCK gene transcription in the intactanimal. REFERENCES Aden, D. P., Fogel, A., Plotkin, S., Damjanov, I., and Knowles, B. (1979) Nature 282, 615-616 Angel, P., Imagawa, R., Chiu, R., Stein, B., Imhra, R. J., Rahmsdorf, H. J., Jonat, C., Herrlich, P., and Karin, M. (1987) Cell 729-739 Berkowitz, L.A., Riabowol, K. T., and Gilman, M. Z. (1989) Mol. Cell. Biol. 9, 4272-4281 Bokar, J. A., Roesler, W. J., Vandenbark, G.R., Kaetzel, D. M., Hanson R. W., and Nilson, J . H. (1988)J. Biol. Chem. 263,1974019747 Bosch, F., Hatzoglou, M., Park, E. A., and Hanson, R. W. (1990) J . Biol. Chem. 265,13677-13682 Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 Cao, Z., Umek, R. M., and McKnight, S. C. (1991) Genes & Deu., in press Chu, D. T. W., and Granner, D. K. (1986) J . Biol. Chem. 261,1684816853 Comb, M., Mermod, N., Hyman, S. E., Pearlberg, J., Ross, M. E., and Goodman, H. M. (1988) EMBO J. 7,3793-3805 Davis, L. G., Dibner, M. D., and Buttey, J. F. (1986) Basic Methods in Molecular Biology, Elsevier Science Publishing Co., Inc., New York Descombes, P., Chojkier, M., Lichtsteiner, S., Falvey, E., and Schibler, U. (1990) Genes & Deu. 4, 1541-1551 Deutsch, P. J., Hoeffler, J . P., Jameson, J . L., and Habener, J. F. (1988a) Proc. Natl. Acud. Sci. U. S. A. 8 5 , 7922-7926 Deutsch, P. J . Hoeffler, J. P., Jameson, J. L., Lin, J. C., and Habener, J. F. (198813) J. Biol. Chem. 263, 18466-18472 Edelman, A. M., Blumenthal, D. K., and Krebs, E. G. (1987) Annu. Reu. Biochem. 56,567-613 Fisch, T. M., Prywes, R., Simon, M. C., and Roeder, R. G. (1989) Genes & Deu. 3 , 198-211 Giralt, M., Park, E. A., Liu, J., Gurney, A.L., and Hanson, R. W. (1991) J. Biol. Chem. 2 6 6 , in press Gonzalez, G. A., Yamamoto, K. K., Fischer, W. H., Karr, D., Menzel, P., Biggs, W., Vale, W. W., and Montminy, M. R. (1989) Nature 337,749-752 Granner, D. K., Andreone, T., Sasaki, K., and Beale, E. (1983) Nature 305,549-551 Hai, T., F. Liu, Coukos, W. J., and Green, M. R. (1989) Genes & Deu. 3,2083-2090 Hyman, S. E., Comb, M., Pearlberg, J., and Goodman, H. M. (1989) Mol. Cell. Biol. 9, 321-324 Imagawa, M., Chiu, R., and Karin, M. (1987) Cell 51, 251-260 Imai, E., Stromstedt, P., Quinn, P. G., Carlstedt-Duke, J., Gunstafsson, J., and Granner, D. K. (1990) Mol. Cell. Biol. 1 0 , 4712-4719 Javitt, N. B. (1990) F A S E B J . 4 ,161-168 Klemm, D. J., Roesler, W. J., Liu, J., Park, E. A,,and Hanson, R.W. (1990) Mol. Cell. Biol. 1 0 , 390-395 Kunkel. T. A. (1985) Proc. Natl. Acad. Sci. U. S. A . 82.488-492 Lamars, W. H:, Hanson, R. W., and Meisner, H. M: (1982) Proc. ~~
~I
Natl. Acad. Sci. U. S. A. 7 9 , 5137-5141 Landschulz, W. H.,Johnson,P. F., and McKnight, S. L. (1989) Science 243, 1681-1688 Lee, W., Mitchell, P., and Tjian, R. (1987) Cell 49, 741-752 Liu, J., and Hanson, R. W. (1991) Mol. Cell. Biochem., in press Liu, J., Roesler, W. J., and Hanson, R. W. (1990) BioTechniques 9, 738-742 Loose, D. S., Cameron, D. K., Short, H.P., and Hanson, R. W. (1985) Biochemistry 24,4509-4512 Magnuson, M. A., Quinn, P. G., and Granner, D. K. (1987) J. Biol. Chem. 2 6 2 , 14917-14920 McGrane, M. M., devente, J., Yun, J.,Bloom, J., Park, E., WynshawBoris, A., Wagner, T., Rottman, F. M., and Hanson, R. W. (1988) J. Biol. Chem. 263,11443-11451 McGrane, M. M., Yun, J . S., Moorman, A. F. M., Lamers, W. H., Hendrick, G.K., Arafa, B. M., Park, E. A., Wagner, T. E., and Hanson, R. W. (1990) J. Biol. Chem. 265, 22372-22379 Mead, D. A., Szczesna-Skorupa, E., and Kemper, B. (1986) Protein Eng. 1,67-74 Mellon, P. L, Clegg, C . H., Correll, L. A., and McKnight, G. S. (1989) Proc. Natl. Acad. Sci. U. S. A. 86,4887-4891 Mermod, N., Williams, T. J., and Tjian, R. (1988) Nature 332,557561 Miller, J . (1972) in Experiments inMolecular Genetics (Miller, J., ed) pp. 352-356, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY Montminy, M. R., Sevarino, K. A., Wagner, J . A,, Mandel, G., and Goodman, R. H. (1986) Proc. Natl. Acud. Sci. U. S. A. 8 3 , 66826686 Muramatsu, M., Kaibuchi, K., and Arai, K. (1989) Mol. Cell. Biol. 9, 831-836 O’Brien, R. M., Lucas, P. C., Forest, C. D., Magnuson, M.A., and Granner, D. K. (1990) Science 249,533-537 Park, E. A., Roesler, W. J., Liu, J., Klemm, D. J., Gurney, A. L., Thatcher, J. D., Shuman, J., Friedman, A., and Hanson, R. W. (1990) Mol. Cell. Biol. 1 0 , 6264-6272 Poli, V., Mancini, F. P., and Cortese, R. (1990) Cell 6 3 , 643-653 Quinn, P. A., Wong, T. W., Magnuson, M. A., Shabb, J . B., and Granner, D. K. (1988) Mol. Cell. Biol. 8 , 3467-3475 Roesler, W. J., Vandenbark, G. R., and Hanson,R. W. (1988) J.Biol. Chem. 263,9063-9066 Roesler, W. J., Vandenbark, G. R., and Hanson,R. W. (1989) J . Biol. Chem. 264,9657-9664 Sassone-Corsi, P., Visvader, J., Ferland, L., Mellon, P. L., and Verma, I. M. (1988) Genes & Dev. 2,1529-1538 Short, J. M., Wynshaw-Boris, A., Short H. P., and Hanson, R. W. (1986) J. Biol. Chem. 261,9721-9726 Silver, B. J., Bokar, J. A., Virgin, J. B., Vallen, E. A., Milstead, A., and Nilson, J. H. (1987) Proc. Natl. Acad Sci. U. S. A . 8 4 , 21982202 Tabor, T., andRichardson, C. C. (1987) Proc. Natl. Acad. Sci.U. S. A . 82,1074-1078 Taylor, S. S. (1989) J. Biol. Chem. 264,8843-8446 Tilghman, S. M., Hanson, R. W., and Ballard, J . F. (1976) in Gluconeogenesis: Its Regulation in MammalianSpecies (Hanson, R. W., and Mehlman, M. A., eds) pp. 47-91 John Wiley & Sons, Inc., New York TNS, M., Benvenisty, N., Cohen, H., and Reshef, L. (1990) Mol. Cell. Biol. 10,2418-2422 Wynshaw-Boris, A., Short, J. M., Loose, D. S., and Hanson, R. W. (1986) J . Biol. Chem. 261,9414-9720 Yamamoto. K. K.. Gonzalez. G. A.. Brigns. -_ W. H.. 111, and Montminy, M. R. (1988) Nature 334; 494-498
