Functional characterization of the proximal promoter of the murine pyruvate carboxylase gene in hepatocytes: Role of multiple GC boxes

Biochimica et Biophysica Acta 1809 (2011) 541–548

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / b b a g r m

Functional characterization of the proximal promoter of the murine pyruvate carboxylase gene in hepatocytes: Role of multiple GC boxes Pinnara Rojvirat, Tanit Chavalit, Sureeporn Muangsawat, Ansaya Thonpho, Sarawut Jitrapakdee ⁎ Molecular Metabolism Research Group, Department of Biochemistry, Faculty of Science, Mahidol University, Bangkok 10400, Thailand

a r t i c l e

i n f o

Article history: Received 4 February 2011 Received in revised form 26 May 2011 Accepted 23 June 2011 Available online 1 July 2011 Keywords: Pyruvate carboxylase Gluconeogenesis Transcription Sp1 Sp3 GC box

a b s t r a c t Pyruvate carboxylase (PC) catalyzes the first committed step in gluconeogenesis in the liver. The murine PC gene possesses two promoters, the proximal (P1) and the distal (P2) which mediate production of distinct tissue-specific mRNA isoforms. By comparing the luciferase activities of 5′-nested deletions of the P1promoter in the AML12 mouse hepatocyte cell line, the critical cis-acting elements required for maintaining basal transcription were located within the 166 nucleotides proximal to the transcription start site. Three GC boxes were identified within this region and shown by gel shift and ChIP assays to bind Sp1/Sp3. Overexpression of Sp1/Sp3 in AML12 and NIH3T3 cells increased P1-promoter activity, with Sp1 being a stronger activator than Sp3. Mutation of any one of the three GC boxes dramatically reduced basal promoter activity by 60–80% suggesting that all three boxes are equally strong regulatory elements. In AML12 cells, overexpression of Sp1/Sp3 restored the transcriptional activity of GC1 and GC2 but not GC3 mutants to levels similar to that of the WT construct, suggesting that GC3 is particularly critical for Sp1/Sp3-mediated induction. In NIH3T3 cells, however, the three boxes were equally important, indicating that the GC boxes differentially contribute to transcriptional regulation of the P1-promoter in the two cell lines. Mutants harboring two disrupted GC boxes showed a further decrease in promoter activity similar to the triple GC box mutant. Neither Sp1 nor Sp3 was able to fully restore the promoter activities of these mutants to that the WT level. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Glucose homeostasis is regulated by the balance between glucose utilization and glucose synthesis. Glycogenolysis is an important metabolic adaptation during short term starvation while gluconeogenesis plays an important role during prolonged starvation [1]. During starvation these two pathways ensure the continuous supply of glucose to the brain and red blood cells which rely on glucose as their primary source of fuel. Gluconeogenesis is the reverse of the pathway of glycolysis except for four steps which bypass the three thermodynamically irreversible reactions in the glycolytic pathway, viz pyruvate kinase (PK), phosphofructokinase (PFK) and hexokinase (HK). Thus pyruvate carboxylase (PC) and phosphoenolpyruvate carboxykinase (PEPCK) catalyze the formation of phosphoenolpyruvate from pyruvate, while fructose-1,6-bisphosphatase (FBPase) and glucose-6-phosphatase (G6Pase) circumvent the barriers represented by PFK and HK respectively [2,3]. PC catalyzes the first committed step of this pathway by converting pyruvate to oxaloacetate so that the remaining steps of gluconeogenesis can occur (for a review, see [4]).

Abbreviations: PC, pyruvate carboxylase; P1, proximal promoter, P2, distal promoter; Sp1, specific protein 1; Sp3, specific protein 3; CRE, cAMP-responsive element; PPRE, peroxisome proliferactor-activated receptor element ⁎ Corresponding author. Tel.: + 66 2 201 5458; fax: + 66 2 354 7174. E-mail address: [email protected] (S. Jitrapakdee). 1874-9399/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.bbagrm.2011.06.011

Gluconeogenesis has long been considered to be primarily regulated by PEPCK [5] and G6Pase [6], both of which are multi-hormonally regulated at the transcriptional level. A recent study, however, has indicated a limited role for PEPCK in the control of gluconeogenesis: liver-specific PEPCK knockout mice show only 40% reduction in gluconeogenic flux despite a 90% reduction in PEPCK protein [7]. This highlighted a possible role for PC in coordinating with PEPCK to control this process. In support of this idea, Louet et al. have shown that transcriptional activation of the PC gene is a key control point in gluconeogenesis during long term fasting [8]. PC also has non-gluconeogenic roles including lipogenesis in white adipose tissue; inhibition of PC activity in differentiating adipocytes impaired lipid accumulation [9]. PC has also been shown to be involved in the anaplerosis which supports glucose-induced insulin secretion in pancreatic β-cells [10]. Silencing of PC in insulinoma cells resulted in impaired anaplerosis and reduced glucose-induced insulin secretion [11]. Dysregulation of PC expression in liver, adipose tissue and pancreatic β-cells is also associated with obesity and type 2 diabetes in mouse and human (for review see [12,13]). Mammalian PC genes possess multiple promoters which regulate production of alternative transcripts with 5′-end heterogeneity [14–18]. In rat and mouse, two promoters are responsible for alternative transcription of the PC gene, resulting in the generation of PC mRNA isoforms that differ in their 5′-untranslated regions [14,15] (Fig. 1). The proximal (P1) promoter is active in gluconeogenic and adipose tissues

542

P. Rojvirat et al. / Biochimica et Biophysica Acta 1809 (2011) 541–548

2.3 kb

P2

P1

ATG

PC gene UTR exon

Expression

AUG

P1 mRNA

Liver and adipose tissue

P2 mRNA

Pancreatic β-cells

AUG

Fig. 1. Structural organization at the 5′-end of the mouse PC gene. Alternative transcription from the proximal (P1) and the distal (P2) promoter of the mouse PC gene [15,23] produce tissue specific species of PC mRNA (P1-mRNA and P2-mRNA) with distinct 5′-untranslated region (UTR) exons.

while the distal (P2) promoter is highly active in pancreatic β-cells [14,19]. A significant body of information is available regarding regulation of the P2 promoter under basal and glucose induction conditions [19–22], whereas little is known about the transcriptional regulation of the P1 promoter. We have previously shown that the P1 promoter is regulated by the peroxisome proliferator-activated receptor gamma (PPARγ) via the PPAR element (PPRE) in adipose tissue [23] while in hepatocytes, it is regulated by cAMP via the cAMP-responsive element (CRE) [24]. These two elements are distal from the transcription initiation site of the PC gene. The basal transcriptional machineries of the P1 promoter, however, have not yet been studied. Here, we report the functional importance of multiple GC boxes located in the P1 promoter of the murine PC gene. These GC boxes mediate strong transcription of the PC gene in the mouse hepatocyte cell line, AML12. 2. Materials and methods 2.1. Generation of PC proximal promoter-luciferase reporter constructs The 2.3 kb mouse PC proximal (P1) promoter [23] was truncated from its 5′-end using the following restriction enzymes: BglII, KpnI, HindIII, XhoI, MluI, DraI and PstI, and the resulting fragments were ligated to the equivalent restriction sites in the cloning site of the pGL3basic plasmid (Promega). The pGL3-P1Δ−166/−80 construct containing an internal deletion of nucleotides −166 to −80 was generated by looping out mutagenesis using Δ−166/−80-F and Δ−166/−80-R primers (see Table 1). The 166 bp fragment of the P1-promoter [pGL3-P1ΔDelA] was used as a template for generating various GC box mutants by site-directed mutagenesis. Mutagenesis was performed in a 50 μl reaction mixture containing 1× Pfu Turbo polymerase buffer, 0.2 mM dNTP, 125 ng of each primer, 250 ng of template and 2.5 units of

Pfu Turbo polymerase (Stratagene). PCR profiles consisted of an initial denaturation at 95 °C for 30 s, followed by 20 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 1 min, and extension at 68 °C for 8 min, followed by a final extension at 68 °C for 10 min. The PCR product was digested with 20 units of DpnI at 37 °C overnight, and 5 μl were transformed into E. coli DH5α (Stratagene). The mutagenic primers are shown in Table 1. The nucleotide sequences of the mutant constructs were verified by automated DNA sequencing using the BigDye Terminator Cycle (Applied Biosystems). 2.2. Cell culture, transient transfection and transactivation assay The mouse hepatoma cell line, AML12 (ATCC: CRL254) and mouse fibroblast cell line, NIH3T3 (ATCC: CRL1658) were used in this study. AML12 cells were grown in a 1:1 (v/v) of DMEM/Hams F12 medium (Gibco) supplemented with 10% (v/v) fetal bovine serum, and 100 units/ml penicillin and streptomycin (Gibco), 1% (v/v) insulintransferrin-selenium solution (Gibco) and 40 ng/ml of dexamethasone (Sigma). NIH3T3 cells were grown in DMEM supplemented with 10% (v/v) fetal bovine serum and 100 units/ml penicillin and streptomycin (Gibco). The cultures were maintained in an incubator at 37 °C with 5% CO2. For transient transfections, cells were plated at a density of 2 × 10 5 cells/well in 24-well plates and cultured in antibiotic-free medium for 1 day prior to transfection. On the day of transfection, 0.25 μg of luciferase reporter construct and 0.25 μg of pRSV-β-gal plasmid were mixed with 2 μg of LipofectAMINE™ 2000 reagent (Invitrogen) and diluted in 100 μl of Opti-MEM® I reduced serum medium (Invitrogen). The transfected cells were maintained at 37 °C with 5% CO2 for 48 h. For transactivation assays, the reporter constructs were co-transfected with the plasmids overexpressing Sp1 (0.25 μg) and/or Sp3 (0.25 μg).

Table 1 Oligonucleotides used for mutagenesis and EMSA. Construct

Oligonucleotide

Sequence (5′–3′)

ΔGC1

− 166/−142 GCbox1 MuF − 166/−142 GCbox1 MuR − 141/−117 GCbox2 MuF − 141/−117 GCbox2 MuR − 91/−79 GCbox4 MuF − 91/−79 GCbox4 MuR − 166/−142 and −91/−79 GCbox1-3 MuF − 166/−142 and −91/−79 GCbox1-3 MuR − 141/−117 and −91/−79 GCbox2-3 MuF − 141/−117 and −91/−79 GCbox2-3 MuR − 166/−117 and 91/−79 GCbox1-3 MuF − 166/−117 and 91/−79 GCbox1-3 MuR Δ−166/−80-F Δ−166/−80-R

TCGATAGGTACCTCTCTGGCCCGG CCGGGCCAGAGAGGTACCTATCGA CACTGCCCCGCCCCAGTGGCCCAT ATGGGCCACTGGGGCGGGGCAGTG GATGGCCTCAGGATTCCCCTGATTTC GAAATCAGGGGAATCCTGAGGCCATC GATGGCCTCAGGATTCCCCTGATTTC GAAATCAGGGGAATCCTGAGGCCATC GATGGCCTCAGGATTCCCCTGATTTC GAAATCAGGGGAATCCTGAGGCCATC GATGGCCTCAGGATTCCCCTGATTTC GAAATCAGGGGAATCCTGAGGCCATC CATCGGGACTACGATGCACCCAAATGGGCTTCCGAT ATCGGAAGCCCATTTGGGTGCATCGTAGTCCCGATG

ΔGC2 ΔGC3 ΔGC1ΔGC3 ΔGC2ΔGC3 ΔGC1ΔGC3 pGL3-P1Δ-166/-80

P. Rojvirat et al. / Biochimica et Biophysica Acta 1809 (2011) 541–548

nescent EMSA kit (Pierce). The 3′-end biotin-labeled oligonucleotides were synthesized by Biobasic Inc (Canada) and their sequences are shown in Table 2. The oligonucleotides were annealed to produce double stranded DNA probes. The annealing reaction was conducted in a 100 μl reaction mixture containing 6 pmol of each oligonucleotide and 1× annealing buffer (10 mM Tris pH 7.4, 1 mM EDTA, 100 mM NaCl). The DNA–protein binding reaction was performed at 4 °C for 20 min in a 20 μl reaction mixture containing 1× binding buffer (10 mM HEPES pH7.8, 50 mM NaCl, 1 mM DTT), 5% (v/v) glycerol, 2 μg of poly(dI–dC), 1% (v/v) NP-40, 4 μg of AML12 nuclear extract and 120 fmol of DNA probe. The DNA–protein complexes were separated by 5% native polyacrylamide gel in 0.5× TBE (0.45 M Tris pH8.0, 0.45 M boric acid, 10 mM EDTA) at 120 V at 4 °C for 90 min. The DNA–protein complexes on the gel were transferred to Biodyne® B nylon membrane (Pierce) for 90 min at 150 mA using Semi Phore™ Semi-Dry Transfer Units (Hoeffer). The DNA–protein complexes were detected with the nonradioactive chemiluminescence detection kit (Pierce). For the competition analyses 5×, 10× or 50× molar excesses of unlabeled double stranded oligonucleotides were included in the binding reactions. For the supershift assays, nuclear extracts were pre-incubated with 1 μg of anti-Sp1 (sc-59) and/or anti-Sp3 (sc-644) rabbit polyclonal antibodies (Santa Cruz Biotechnology, Inc.) before the binding was performed.

Table 2 Oligonucleotides used for EMSA. Oligonucleotide probe

Sequence (5′–3′)

WT GCbox3(− 96/−76) Fa WT GCbox3(− 96/−76) R WT GCbox2(− 141/−121) Fa WT GCbox2(− 141/−121) R WT GCbox1(− 156/−136) Fa WT GCbox1(− 156/−136) R EMSA unrelated probe-F EMSA unrelated probe-R Sp1consensus sequence F Sp1consensus sequence R

CAGGAGGCGGCGCCCCTTCCC GGGAAGGGGCGCCGCCTCCTG TCTCTGGCCCGGCCCTCCTAG CTAGGAGGGCCGGGCCAGAGA ACCCACTGCCCCGCCTCTCTG CAGAGAGGCGGGGCAGTGGGT GATGGCCTCAGGATTCCCCTGATTTC GAAATCAGGGGAATCCTGAGGCCATC TATTCGATCGGGGCGGGGCGAGC TGCTCGCCCCGCCCCGCACGAAT

a

543

Labeled with biotin at 3′ end.

2.3. Luciferase and β-galactosidase assay The transfected cells were scraped from dishes, transferred to 1.5 ml microtubes and centrifuged at 13,000 rpm for 5 min. The cells were washed with 1 ml of PBS pH 7.4 and resuspended in 100 μl of 1× reporter lysis buffer (Promega). The lysates were subjected to three cycles of freezing/thawing and centrifuged at 13,000 rpm for 5 min at 4 °C. Fifty micrograms of total protein were used for luciferase and βgalactosidase assays which were performed as previously described [21]. The firefly luciferase activity was normalized to β-galactosidase activity and expressed as the relative luciferase activity.

2.5. Chromatin immunoprecipitation (ChIP) assay Soluble chromatin was prepared from AML12 cells. In brief, 2 × 106 AML12 cells grown in a 10 cm culture dish overnight were cross-linked with 1% (v/v) formaldehyde at 37 °C for 10 min. The cells were sonicated for 15 × 30 s before centrifugation at 13,000 rpm for 10 min at 4 °C. The Sp1/Sp3-bound DNA complexes were precipitated with 5 μl of either anti-Sp1 or anti-Sp3 polyclonal antibody or 5 μl of anti-β-actin polyclonal

2.4. Electrophoresis mobility shift assay (EMSA) Nuclear extracts were prepared from AML12 cells as described previously [21]. EMSA was performed using the LightShift Chemilumi-

A pGL3-P1

+28

-2300

pGL3-P1ΔBg1II

+28

-1850 +28

-781

pGL3-P1ΔM1uI

+28

-603

pGL3-P1ΔDraI

+28

-335

pGL3-P1ΔKpnI

+28

-256

pGL3-P1ΔPstI

LUC LUC

LUC LUC LUC LUC

+28

LUC

-166

pGL3-P1ΔDe1A

+28

pGL3-P1ΔDe1B

*

LUC

-80 +28

pGL3-P1Δ-166/-80-2300

Δ-166/-80

pGL3-Basic

LUC LUC 2

4

6

8

10

12

14

Relative luciferase activity fold

B

+1

Fig. 2. Localization of minimal promoter elements required for basal transcription from the P1-promoter. A, The 2.3 kb 5′-flanking sequence of the P1-promoter was truncated from its 5′-end and cloned upstream of the luciferase reporter gene followed by transient transfection into AML12 cells. The luciferase activity of each construct was normalized to the β-galactosidase activity and expressed as relative luciferase activity. The values obtained from cells transfected with various reporter constructs were relative to that of cells transfected with empty vector (pGL-3 basic), which was arbitrarily set as 1. *p ≤ 0.05 B, Nucleotide sequence of the 166 bp proximal to the transcription start site of the P1-promoter. Three GC boxes are boxed; + 1 indicates the transcription start site.

544

P. Rojvirat et al. / Biochimica et Biophysica Acta 1809 (2011) 541–548

promoter. Alternative transcription from these two promoters produces different mRNA isoforms that are differentially expressed. We have previously isolated the 2.3 kb 5′-flanking sequence of the P1 region [23]. To identify the minimal promoter element of the P1promoter of the PC gene, we truncated the 2.3 kb flanking sequence from its 5′-end using restriction enzymes and PCR. The resultant reporter constructs were transiently transfected into AML12 mouse hepatocytes and their luciferase activities measured. The AML12 cell line was chosen because they represent highly differentiated hepatocytes [25]. As shown in Fig. 2A, deletions of the 5′-flanking region from nucleotide positions − 2.3 kb (pGL3-P1) to −781 (pGL3P1ΔMluI) did not affect reporter gene activity. However, further deletion to nucleotide position −603 (pGL3-P1ΔDraI) resulted in a 2fold increase in reporter activity, suggesting the presence of a negative regulatory element located between nucleotides −781 and −603. Further deletions from nucleotide positions −603 to − 166 did not significantly affect the luciferase activity. Deletion from −166 to −80 (pGL3-P1ΔDelB), however, caused a dramatic decrease (~90%) in

antibody overnight at 4 °C before addition of protein A-agarose beads (Upstate Biotech). The proteins were removed from DNA by digestion with 10 μg/ml proteinase K at 45 °C for 1 h. The DNA was purified by Nucleospin® Extract II kit (Nucleospin) and eluted in 100 μl of sterile water. Ten microliters of the eluted DNA were subjected to PCR with Sp1F 5′GTCTTGTAATTCTGTGTATTCCT-3′ and Sp1R 5′-TGGCCCCAGAATACAAAGTTCTC-3′ primers that flank the three GC boxes or negative control primers, Neg.PrimerF 5′-GGGCGGATCCTGTGAGGTGGCCAAAGAGAAT-3′ and Neg.PrimerR 5′-TCCCGGTACCTTAATGCACAGGATGTGAGT3′ which are located more than 10 kb downstream of the transcription start site and flank exon 13 [24]. 3. Results and discussion 3.1. The minimal promoter element of P1 contains three GC boxes Fig. 1 shows that in the genomic structure of the mouse PC gene, the distal (P2) promoter is located upstream of the proximal (P1)

A

core GC GC1 GC2 GC3 consensus U

B

5’ - ACCCACTGCCCCGCCTCTCTG-3’ 5’ - TCTCTGGCCCGGCCCTCCTAG-3’ 5’ - CAGGAGGCGGCGCCCCTTCCC-3’ A 5’ - TATTCGATCGGGGCGGGGCGAGC-3’ 5’ - GATGGCCTCAGGATTCCCCTGATTTC-3’

GC1 5x

10x

C

consensus

25x

5x

U

10x

GC2

25x

5x

10x

2

3

consensus

25x

U

5x

10

25x

C1

C1 C2

C2

1

2

D

3

4

5

6

GC3

7

8

1

4

10x

25x

U

5x

10x

6

7

8

E

consensus

GC1 5x

5

25x -

GC2

αSp3 αSp1 aSp3 αSp1

-

GC3

αSp3 αSp1 αSp3 αSp1

-

αSp3 αSp1αSp3 αSp1

C1 C1 C2

C2

1

2

3

4

5

6

7

8

1

2

3

4

5

6

7

8

9

10

11 12

Fig. 3. Electrophoretic mobility shift assay of GC1, GC2 and GC3 boxes of the P1-promoter. A, The sequences of GC1, GC2, GC3, consensus Sp1/Sp3 and an unrelated sequence (U) used for EMSA. B, The biotin-labeled double stranded oligonucleotides harboring the GC1 sequence were incubated with AML12 nuclear extract in the absence (lane 1) or presence of excess amounts of unlabeled double stranded oligonucleotide competitor (5×, 10× and 25×) (lanes 2–4), unlabeled double stranded oligonucleotide of an unrelated sequence (U) (lane 5), or consensus Sp1/Sp3 binding site competitor (consensus, 5×, 10× and 25×) (lanes 6–8). Similar experiments were performed using the GC2 and GC3 boxes as probes (C and D, respectively). E, EMSA of the GC box1 (lanes 1–4), GC box2 (lanes 5–8) and GC box3 probes (lanes 9–12) were performed in the presence of anti-Sp1 and/or anti-Sp3 antibodies. C1 and C2, specific complexes. Supershift bands are indicated with arrows.

P. Rojvirat et al. / Biochimica et Biophysica Acta 1809 (2011) 541–548

reporter gene activity, suggesting that the core promoter is located within the first 166 nucleotides upstream of the transcription start site. To verify the importance of the nucleotides between − 166 and −80 to promoter activity, the longest promoter construct with an internal deletion of this region was generated (pGL3-P1Δ−166/−80). As expected, deletion of this region produced a similar result to that of the pGL3-P1ΔDelB construct, indicating that the nucleotides between −166 and −80 primarily control transcription from the P1-promoter. Examination of the nucleotide sequence within this region reveals that the P1-promoter exhibits typical housekeeping features, i.e. it contains no TATA or CCAAT boxes but harbors multiple GC boxes located between nucleotides −148/−142 (GC box1), −136/−125 (GC box2) and −91/−81 (GC box3) (Fig. 2B). 3.2. Transcription factors Sp1 and Sp3 bind to three GC boxes in vitro and in vivo As the GC box is well known to bind to the Krüpple-like transcription factor family especially Sp1 and Sp3 [26,27], we performed electrophoretic mobility shift assays (EMSA) of probes harboring GC1, GC2 and GC3 sequences (Fig. 3A) using nuclear extracts prepared from AML12 cells. As shown in Fig. 3B–3D, the three probes similarly produced two strong DNA–protein complexes (C1 and C2). Incubation of the GC box1 probe with increasing concentrations of unlabeled GC box1 gradually decreased C1 and C2 formation, while incubation with an unrelated double stranded oligonucleotide (Fig. 3A) had no effect. This suggests that these DNA–protein complexes are highly specific. Incubation of the GC box1 probe with an unlabeled consensus Sp1/Sp3 binding sequence also decreased C1 and C2 formation in a concentration dependent manner. Similar results were observed with the GC box2 and GC box3 probes (Fig. 3C and 3D). To identify whether Sp1 and Sp3 are indeed associated with C1 and C2 formation, we performed supershift assays by pre-incubating nuclear extracts with antibodies to Sp1 and/or Sp3. As shown in Fig. 3E, incubation of the GC box1 probe with anti-Sp1 antibody eliminated approximately 70% of C1 formation concomitant with formation of the supershift band (arrow) while it had no effect on C2 formation. Conversely, incubation with anti-Sp3 antibody minimally affected C1 formation while completely eliminating C2 formation. Inclusion of both anti-Sp1 and anti-Sp3 antibodies in the reaction completely blocked formation of both C1 and C2 (Fig. 3E). These data

545

indicate that the C1 complex is mainly associated with Sp1 while the C2 complex is associated with Sp3. Similar results were observed with the GC box2 and GC box3 probes (Fig. 3E). Finally, we performed chromatin immunoprecipitation (ChIP) assays to confirm binding of Sp1 and Sp3 to the GC region of the P1-promoter in vivo. Because the three GC boxes are located within an 80 bp vicinity, it is technically impossible to determine in situ binding of the individual GC boxes by ChIP. Sp1-and Sp3-bound chromatin was precipitated with antibodies to Sp1 and Sp3 respectively, before PCR using primers which flank the three GC boxes (Fig. 4A). Upon immunoprecipitation, 223 bp PCR products spanning the three GC boxes that are associated with Sp1 or Sp3 (Fig. 4B) were observed, whereas the chromatin that was immunoprecipitated with an anti-actin antibody as a negative control did not produce any amplicon. When primers that flank exon 13 were used in the PCR as a negative control, only the input fraction produced an amplicon (Fig. 4C), indicating the specificity of the ChIP assay. These results indicate that Sp1 and Sp3 are associated with at least one of the three GC boxes in vivo. 3.3. Contribution of each GC box in basal and Sp1/Sp3-mediated transcription induction. To examine the relative contribution of each of the GC boxes to transcriptional regulation of the P1-promoter, each GC box was mutated (Fig. 5A) and the luciferase activities were measured. As shown in Fig. 5B, mutation of any one of the three GC boxes alone resulted in a dramatic 70–80% decrease in reporter activity under basal conditions, suggesting that these GC boxes are major determinants of transcriptional activity of the P1-promoter. Having demonstrated, by ChIP and EMSA, that Sp1 and Sp3 bind these GC boxes, transactivation assays were performed to assess the functional significance of the interactions. Plasmids overexpressing Sp1 and/or Sp3 were co-transfected into AML12 cells with the pGL3-P1ΔDelA construct harboring the 166 bp flanking sequence of the P1-luciferase reporter gene (WT). As shown in Fig. 5B, over-expression of Sp1 resulted in a 36-fold increase in reporter gene activity while overexpression of Sp3 resulted in just a 6-fold increase, suggesting that Sp1 contributes to regulation of the proximal promoter more than Sp3 in AML12 cells. The stronger transactivation activity of Sp1 over Sp3 is most likely because Sp1 can form oligomeric structures between

Sp1/Sp3 Ab

A

Neg. Primer F

Sp1F GC1 GC2 GC3

Neg. Primer R

Sp1 R

300 bp 223 bp

B

1 300 bp

2

3

4

5

6

7 223 bp

200 bp

C 300 bp

300 bp

200 bp Fig. 4. Chromatin immunoprecipitation assay of the P1-promoter GC region. Crosslinked chromatin prepared from AML12 cells was precipitated with anti-Sp1, anti-Sp3 or anti-β actin polyclonal antibodies and subjected to PCR. A, A schematic of the P1-promoter region showing the positions of PCR primers (Sp1F/Sp1R) that flank the three GC boxes and the negative control primers (Neg-primerF/R) that located between exon 13 [24]. B, PCR products generated from different fractions of the chromatin that had been precipitated with anti-Sp1 antibody or with anti-Sp3 antibodies using Sp1F/Sp1R primers or (C) Neg-primer F/R. Lane 1, DNA marker; 2, negative control PCR; 3, input fraction; 4, IP with no antibody; 5, IP with anti-Sp1 antibody; 6, IP with anti-Sp3 antibody; 7, IP with anti-β actin antibody.

546

P. Rojvirat et al. / Biochimica et Biophysica Acta 1809 (2011) 541–548

multiple adjacent GC boxes, whereas Sp3 cannot [28]. The formation of higher-order Sp1 complexes at adjacent GC boxes contributes to synergistic transactivation of transcription by this transcription factor [28,29]. Although also an activator of the P1-promoter, Sp3 reduced Sp1-mediated transactivation of the promoter when both proteins were co-expressed (Fig. 5B, Sp1 + Sp3). Under these conditions, it is most likely that Sp3 displaces or competes with Sp1 binding to adjacent GC boxes in the PC P1-promoter and thereby interferes with synergistic activation by Sp1. This phenomenon is not unprecedented because binding of Sp3 to adjacent GC boxes is more stable than binding of Sp1 [29]. Interestingly, over-expression of Sp1 or Sp3 in AML12 cells resulted in a similar level of reporter activity from the P1-promoter carrying ΔGC1 box or ΔGC2 box mutations (Fig. 5B) as from the WT promoter. This suggests that the remaining two GC boxes can functionally substitute for the disrupted GC box in the presence of excess amounts of Sp1 and Sp3. Unlike the first two mutants, however, over-expression of Sp1 or Sp3 only partially restored promoter activity of the ΔGC3 box mutant compared to that of the WT (38% for Sp1 and 18% for Sp3). This result suggests that the GC3 box is particularly critical for Sp1 and Sp3-mediated induction of P1promoter activity. A similar trend was observed when the GC box mutants were co-transfected with Sp1 and Sp3 together. However,

similar to WT, over-expression of Sp3 reduced Sp1-mediated transcriptional activation of all three mutants. To examine whether the pattern of activities of the above constructs is specific to the hepatocyte cell line, the reporter assays were also performed in the non-hepatocyte cell line, NIH3T3. Although reporter gene activity was markedly reduced for each of the three single GC box mutants in NIH3T3 cells, the degree of reduction was less than that observed in AML12 cells: mutations of GC1, GC2 or GC3 boxes reduced reporter activity by 60% (cf. 80-90% in AML12) (Fig. 5C). Overexpression of Sp1 and Sp3 increased the reporter gene activities of the WT promoter construct by approximately 14- and 5-fold respectively, although the transactivation was not as high as in AML12 cells. In sharp contrast with AML12 cells in which GC3 box is the most crucial for Sp1and Sp3-mediated transcription induction, mutations of GC1, GC2 and GC3 boxes all markedly reduced both Sp1- and Sp3-mediated transcriptional activation in NIH3T3 cells. Similar to AML12 cells, coexpression with Sp3 reduced Sp1-mediated activation of transcription from the WT and the three GC box mutants. We next examined whether disruptions of two GC boxes while maintaining one GC box intact (Fig. 6A) would make a further impact

A

-148/-142 -136/-127

WT

A

-148/-142 -136/-127

WT

-166

GC1

-91/-79

GC2

-148/-142 -136/-127

Δ GC1 -166

Δ GC1 Δ GC2

LUC

GC3 -91/-79

GC2

LUC

GC3

-166

GC1

-91/-79

GC2

-91/-79

-148/-142 -136/-127

-91/-79

-166

ΔGC 1 Δ GC3

-166

Δ GC2 -166

-91/-79

ΔGC 2 Δ GC3

-166

GC1

LUC

GC2 -91/-79

LUC

GC1

LUC

GC3

LUC

GC3

-148/-142 -136/-127

-148/-142 -136/-127

LUC

GC3

-148/-142 -136/-127

-148/-142 -136/-127

-91/-79

ΔGC 1 ΔGC 2 Δ GC3 -166 -148/-142 -136/-127

Δ GC3

GC1

-166

B

LUC

GC2

B

Basal 1.0

WT Δ GC1

*0.2

Δ GC2

* 0.2

Δ GC3

*

0.3

+Sp1

0.03

+Sp3

36.17

5.96

6.26

29.23

2.15

3.18

0.00

28.04

2.01

0.01

** 14.0

0.04

3.60

AML12 Construct

AML 12

Construct

LUC

-91/-79

WT

Sp1+Sp3

Δ GC1 Δ GC2

2.20

22.57

1.75

1.40

17.43

1.75

3.20

0.80

18.21

1.27

** 1.10

0.20

* 8.52

1.22

Δ GC1 Δ GC3 Δ GC2Δ GC3 Δ GC1Δ GC2Δ GC3

1.0

*

0.1

0.03 0.00

*

NIH 3T3

Construct

Basal

WT

1.0

Δ GC1

** 0.4

Δ GC2

** 0.4 ** 0.4

Δ GC3

0.03

13.9

0.08

** 8.5

0.16

** 6.8

0.03

**

5.4

+Sp3 5.3

0.89 0.28

** 3.3

1.12

*

0.68

**

2.9 1.5

Sp1+Sp3

0.63

7.8

0.26

** 6.9

0.33 0.15

* **

4.8 2.5

0.06 0.28 0.58 0.30

Fig. 5. Effect of single GC box mutations on P1-promoter activity. A single point mutation was introduced to an individual GC box in the pGL3-P1ΔDelA reporter construct such that the other two GC boxes remain intact (A). Each mutant on its own (basal) or together with plasmid(s) overexpressing Sp1 and/or Sp3 was transfected into AML12 (B) or NIH3T3 (C) cells. The luciferase activity of each construct was normalized to the β-galactosidase activity and expressed as relative luciferase activity. The value obtained from the WT (166 bp upstream sequence)-luciferase construct (pGL3-DelA) was arbitrarily set as 1 and those obtained from other constructs were relative to the WT construct, and shown as fold change. *p ≤ 0.01; **p ≤ 0.05.

9.7

5.9

0.4

0.01

0.2

0.03

** 8.2

0.1

* 0.1

0.00

** 1.8

0.3

0.1

*

0.3

6.26

** **

1.0 0.4

** 0.5 ** 0.1

Sp1+Sp3 1.0 0.1

22.57 1.75

**

6.13

**

0.07

2.4

0.11

0.1

**3.98

1.12

0.0

**0.92 0.20

0.1

NIH3T3 Construct

+Sp1

36.17

**

** 4.7

C

C

+Sp3

+Sp1

Basal

Basal

WT Δ GC1 Δ GC2 Δ GC1 Δ GC3 Δ GC2Δ GC3 Δ GC1Δ GC2Δ GC3

1.0

*

0.1

0.01 0.03

* *

+Sp1

+Sp3

2.7

0.28

0.2

0.04

** 3.5 0.30

0.2

0.07

** 3.1 0.23

* 0.2

0.05

5.3

13.9 0.89

**

** 1.1 0.12

**

1.0

Sp1+Sp3 7.8

0.63 0.03

**

0.06

1.9

0.16

1.6

0.25

** 1.7

0.32

**

**

0.9

0.02

** 0.9

0.02 0.03

**0.8

0.08

**0.6

Fig. 6. Effect of double and triple GC box mutations on activity of the P1-promoter. Combinations of double (ΔGC1ΔGC2, ΔGC1ΔGC3 or ΔGC2ΔGC3) or triple (ΔGC1ΔGC2ΔGC3) GC box mutations were introduced into the pGL3-P1ΔDelA construct (A). Each mutant on its own (basal), or together with plasmid(s) overexpressing Sp1 and/or Sp3, was transfected into AML12 (B) or NIH3T3 (C) cells. The luciferase activity of each construct was normalized to the β-galactosidase activity and expressed as relative luciferase activity. The value obtained from the WT (166 bp upstream sequence)-luciferase construct (pGL3-DelA) was arbitrarily set as 1 and those obtained from other constructs were relative to the WT construct, and shown as fold change. *p ≤ 0.01; **p ≤ 0.05.

P. Rojvirat et al. / Biochimica et Biophysica Acta 1809 (2011) 541–548

on basal and Sp1/Sp3-mediated transcription of the reporter gene. In general, mutations of two GC boxes further decreased the reporter gene activity close to that of the triple GC box mutant, i.e. all mutants exhibited 80-90% loss of their basal promoter activities in both AML12 and NIH3T3 cells (Fig. 6B and C). In contrast to the single GC box mutants, over-expression of neither Sp1 nor Sp3 could restore the promoter activities of ΔGC1ΔGC2, ΔGC1ΔGC3 or ΔGC2ΔGC3 mutants to the WT level (Fig. 6B and C). This indicates that a single GC box is unable to support maximal Sp1- and Sp3-mediated transcriptional activation of the P1-promoter. Again, over-expression of Sp3 decreased Sp1-mediated transcription induction of double GC box mutants in both AML12 and NIH3T3 cells. Mutation of all three GC boxes (ΔGC1ΔGC2ΔGC3) resulted in an almost complete loss of Sp1and Sp3-mediated reporter activity. Collectively the above results suggest that GC1, GC2 and GC3 boxes are important for basal transcription from the P1-promoter in both AML12 and NIH3T3 cells because disruption of any of these GC boxes markedly reduced promoter activity. As these GC boxes are located within an 80 bp vicinity, they probably interact with Sp1/Sp3 to form a complex that assists transcription from the P1-promoter. Disruption of one of these GC boxes most likely disrupts a larger transcription complex and thereby results in severely diminished reporter activity. However, these three GC boxes contribute to the transcriptional regulation of the P1-promoter in AML12 more so than in NIH3T3 cells. When Sp1/Sp3 were not limiting in AML12 cells, only loss of the GC3 box (and not GC1 or GC2) resulted in a significant decrease in the reporter activity. This indicates a critical role for GC3 in maintaining Sp1- and Sp3-mediated transcriptional induction of the P1-promoter in these cells. This may well be simply because GC3 box is located closest to the transcription start site and is likely to mediate the interaction with the initiation complex. In NIH3T3 cells, however, GC1 and GC2 were as important as GC3 for Sp1- and Sp3-mediated transcriptional activation. Sp1 and Sp3 often interact with other ubiquitous or tissue-specific transcription factors which in turn modulate transcription of their target genes [26,27]. It is highly likely that in AML12 cells, Sp1 and Sp3 interact with transcription factor(s) not present in NIH3T3 cells and this contributes to more robust transcriptional activation of the P1-promoter in hepatocytes. Sp1 is well recognized as a basal transcription factor because it is ubiquitously expressed in most tissues. However, several lines of evidence now indicate that Sp1 activity can be post-translationally modified under certain physiological conditions and plays an important role in transcriptional regulation of many genes involved in glucose and lipid metabolism (reviewed by Vaulont et al. [30]). The gene promoter of the biotin-dependent enzyme acetyl-CoA carboxylase (ACC), for instance, is highly regulated by Sp1 through multiple GC boxes proximal to the transcription start site. Glucose regulates Sp1 binding to the GC boxes through phosphorylation/dephosphorylation of Sp1 protein [31,32]. It appears that three distinct regions/domains within the P1promoter contribute to the transcriptional regulation of the PC gene in liver and adipose tissue. The first region, identified in the present study, is located within the 166 nucleotides proximal to the transcription start site and comprises three GC boxes. These Sp1/Sp3-bound GC boxes probably interact with each other to form a complex that assists transcription from the P1-promoter under basal conditions. The second important region is the PPRE, located at −386/−374, which provides a binding site for PPAR-γ1 and PPAR-γ2 [23]. The presence of the PPRE is consistent with the lipogenic role of PC during adipocyte differentiation [9,23,33,34]. Finally, the distal cAMP-responsive element (CRE), which is located at −1639/−1631, mediates the induction of PC by cAMP [24] in hepatocytes. In contrast, the P2 promoter possesses, proximal to the transcription start site, only one GC box and two inverted CCAAT boxes, providing binding sites for Sp1/Sp3 and nuclear factor Y (NF-Y) to drive transcription under basal conditions [20]. The P2 promoter also utilizes HNF3β/Foxa2, USF1/2 [21] and MAFA [35] to drive transcription in a

547

pancreatic β-cell specific manner. Pedersen et al. (2010) [22] have also recently identified the carbohydrate responsive element binding protein (ChREBP) as a transcriptional regulator of the P2 promoter during glucose induction. In summary we have shown that the P1-promoter of murine PC is regulated via three GC boxes located within the first 166 nucleotides upstream of the transcription start site. Sp1 and Sp3 bind to these GC boxes and differentially regulate P1-promoter activity in hepatocyte and non-hepatocyte cells. Acknowledgments This work was supported by the Thailand Research Fund and the Commission of Higher Education—grant RMU5080023 and Faculty of Science, Mahidol University. Pinnara Rojvirat was the recipient of a CHE-PhD-THA scholarship from the Commission of Higher Education, Thailand. The authors thank Dr. S. Ross, Department of Pathology, Harvard Medical School, Boston, Massachusetts, for the gift of mammalian expression plasmids encoding Sp1 and Sp3. The authors thank Prof. John Wallace and Clair Alvino, University of Adelaide for critical reading of the manuscript. References [1] E.P. Corssmit, J.A. Romijn, H.P. Sauerwein, Regulation of glucose production with special attention to non-classical regulatory mechanisms, Metabolism 50 (2001) 742–755. [2] F.P. Lemaigre, G.G. Rousseau, Transcriptional control of genes that regulate glycolysis and gluconeogenesis in adult liver, Biochem. J. 303 (1994) 1–14. [3] B. Desvergne, L. Michalik, W. Wahli, Transcriptional regulation of metabolism, Physiol. Rev. 86 (2006) 465–514. [4] S. Jitrapakdee, M. St Maurice, I. Rayment, W.W. Cleland, J.C. Wallace, P.V. Attwood, Structure, mechanism and regulation of pyruvate carboxylase, Biochem. J. 413 (2008) 369–387. [5] J. Yang, L. Reshef, H. Cassuto, G. Aleman, R.W. Hanson, Aspects of the control of phosphoenolpyruvate carboxykinase gene transcription, J. Biol. Chem. 284 (2009) 27031–27035. [6] J.C. Hutton, R.M. O'Brien, Glucose-6-phosphatase catalytic subunit gene family, J. Biol. Chem. 284 (2009) 29241–29245. [7] S. Burgess, T.T. He, Z. Yan, J. Lindner, A.D. Sherry, C.R. Malloy, J.D. Browning, M.A. Magnuson, Cytosolic phosphoenolpyruvate carboxykinase does not solely control the rate of hepatic gluconeogenesis in intact mouse liver, Cell Metab. 5 (2007) 313–320. [8] J.F. Louet, A.R. Chopra, J.V. Sagen, J. An, B. York, M. Tannour-Louet, P.K. Saha, R.D. Stevens, B.R. Wenner, O.R. Ilkayeva, J.R. Bain, S. Zhou, F. DeMayo, J. Xu, C.B. Newgard, B.W. O'Malley, The coactivator SRC-1 is an essential coordinator of hepatic glucose production, Cell Metab. 12 (2010) 606–618. [9] Y. Si, H. Shi, K. Lee, Impact of perturbed pyruvate metabolism on adipocyte triglyceride accumulation, Metab. Eng. 280 (2009) 27466–27476. [10] M.J. MacDonald, Feasibility of a mitochondrial pyruvate malate shuttle in pancreatic islets. Further implication of cytosolic NADPH in insulin secretion, J. Biol. Chem. 270 (1995) 20051–20058. [11] N.M. Hasan, M.J. Longacre, S.W. Stoker, T. Boonsaen, S. Jitrapakdee, M.A. Kendrick, J.C. Wallace, M.J. MacDonald, Impaired anaplerosis and insulin secretion in insulinoma cells caused by siRNA mediated suppression of pyruvate carboxylase, J. Biol. Chem. 283 (2008) 28048–28059. [12] S. Jitrapakdee, A. Vidal-Puig, J.C. Wallace, Anaplerotic role of pyruvate carboxylase in mammalian tissues, Cell. Mol. Life Sci 63 (2006) 843–854. [13] S. Jitrapakdee, A. Wutthisatapornchai, J.C. Wallace, M.J. MacDonald, Regulation of Insulin secretion: role of mitochondrial signaling, Diabetologia 53 (2010) 1019–1032. [14] S. Jitrapakdee, G.W. Booker, A.I. Cassady, J.C. Wallace, The rat pyruvate carboxylase gene structure. Alternate promoters generate multiple transcripts with the 5'-end heterogeneity, J. Biol. Chem. 272 (1997) 20520–20528. [15] S. Jitrapakdee, N. Petchamphai, P. Sunyakumthorn, J.C. Wallace, V. Boonsaeng, Structural and promoter regions of the murine pyruvate carboxylase gene, Biochem. Biophys. Res. Communs. 287 (2001) 411–417. [16] M.A. Carbone, N. MacKay, M. Ling, D.E. Cole, C. Douglas, B. Rigat, A. Feigenbaum, J.T. Clarke, J.C. Haworth, C.R. Greenberg, L. Seargeant, B.H. Robinson, Amerindian pyruvate carboxylase deficiency is associated with two distinct missense mutations, Am. J. Hum. Genet. 62 (1998) 1312–1319. [17] D. Wang, D.H. Yang, K.C. De Braganca, J. Lu, L. Yu-Shih, P. Briones, T. Lang, D.C. De Vivo, The molecular basis of pyruvate carboxylase deficiency: mosaicism correlates with prolonged survival, Mol. Genet. Metab. 95 (2008) 31–38. [18] S.R. Hazelton, D.M. Spurlock, C.A. Bidwell, S.S. Donkin, Cloning the genomic sequence and identification of promoter regions of bovine pyruvate carboxylase, J. Dairy Sci. 91 (2008) 91–99. [19] S. Jitrapakdee, Q. Gong, M.J. MacDonald, J.C. Wallace, Regulation of rat pyruvate carboxylase gene expression by alternate promoters during development, in

548

[20]

[21]

[22]

[23]

[24]

[25]

P. Rojvirat et al. / Biochimica et Biophysica Acta 1809 (2011) 541–548 genetically obese rats and in insulin-secreting cells. Multiple transcripts with 5'end heterogeneity modulate translation, J. Biol. Chem. 273 (1998) 34422–34428. P. Sunyakumthorn, T. Boonsaen, V. Boonsaeng, J.C. Wallace, S. Jitrapakdee, Involvement of specific proteins (Sp1/Sp3) and nuclear factor Y (NF-Y) in basal transcription of the distal promoter of the rat pyruvate carboxylase gene in β-cells, Biochem. Biophys. Res. Communs. 329 (2005) 188–196. T. Boonsaen, P. Rojvirat, K.H. Surinya, J.C. Wallace, S. Jitrapakdee, Transcriptional regulation of the distal promoter of the rat pyruvate carboxylase gene by hepatocyte nuclear factor 3β/Foxa2 and upstream stimulatory factors in insulinoma cells, Biochem. J. 405 (2007) 359–367. K.B. Pedersen, R.S. Buckley, R. Scioneaux, Glucose induces expression of rat pyruvate carboxylase through a carbohydrate response element in the distal gene promoter, Biochem. J. 426 (2010) 159–170. S. Jitrapakdee, M. Slawik, G. Medina-Gomez, M. Campbell, J.C. Wallace, J.K. Sethi, S. O'Rahilly, A. Vidal-Puig, Peroxisome proliferators activated receptor-γ regulates murine pyruvate carboxylase gene expression in vivo and in vitro, J. Biol. Chem. 280 (2005) 27466–27476. A. Thonpho, C. Sereeruk, P. Rojvirat, S. Jitrapakdee, Identification of the cyclic AMP responsive element (CRE) that mediates transcriptional regulation of the pyruvate carboxylase gene in HepG2 cells, Biochem. Biophys. Res. Communs. 393 (2010) 714–719. J.C. Wu, G. Merlino, N. Fausto, Establishment and characterization of differentiated, non-transformed hepatocyte cell lines derived from mice transgenic

[26] [27] [28] [29]

[30] [31] [32]

[33] [34] [35]

for transforming growth factor alpha, Proc. Natl. Acad. Sci. U.S.A. 91 (1994) 674–678. G. Suske, The Sp-family of transcription factors, Gene 238 (1999) 291–300. G. Suske, E. Bruford, S. Phillipsen, Mammalian SP/KLF transcription factors: bring in the family, Genomics 85 (2005) 551–556. E. Pascal, R. Tjian, Different activation domain of Sp1 govern formation of multimers and mediate transcriptional synergism, Genes Dev. 5 (1991) 1646–1656. B. Yu, P.K. Datta, S. Bagchi, Stability of the Sp3–DNA complex is promoter-specific: Sp3 efficiently competes with Sp1 for binding to promoters containing multiple Sp-sites, Nucleic Acids Res. 31 (2003) 5368–5376. S. Vaulon, M. Vasseur-Cognet, A. Kahn, Glucose regulation of gene transcription, J. Biol. Chem. 275 (2000) 31555–31558. S. Daniel, K.-H. Kim, Sp1 mediates glucose activation of the acetyl-CoA carboxylase promoter, J. Biol. Chem. 271 (1996) 1385–1392. S. Daniel, S. Zhang, A.A. Depaoli-Roach, K.-H. Kim, Dephosphorylation of Sp1 by protein phosphatase 1 is involved in the glucose-mediated activation of the acetyl-CoA carboxylase gene, J. Biol. Chem. 271 (1996) 14692–14697. S.O. Freytag, M.F. Utter, Regulation of the synthesis and degradation of pyruvate carboxylase in 3T3-L1 cells, J. Biol. Chem. 258 (1983) 6307–6312. J. Zhang, W.-L. Xia, F. Ahmed, Regulation of pyruvate carboxylase in 3T3-L1 cells, Biochem. J. 306 (1995) 205–210. H. Wang, T. Brun, K. Kataoka, A.J. Sharma, C.B. Wollheim, MAFA controls genes implicated in insulin biosynthesis and secretion, Diabetologia 50 (2007) 348–358.